Fresenius Environmental Bulletin Founded Jointly by F. Korte and F

FEB – Fresenius Environmental Bulletin founded jointly by F. Korte and F. Coulston

Production by PSP – Parlar Scientific Publications, Angerstr. 12, 85354 Freising, Germany in cooperation with Lehrstuhl für Chemisch-Technische Analyse und Lebensmitteltechnologie, Technische Universität München, 85350 Freising - Weihenstephan, Germany Copyright © by PSP – Parlar Scientific Publications, Angerstr. 12, 85354 Freising, Germany. All rights are reserved, especially the right to translate into foreign language. No part of the journal may be reproduced in any form- through photocopying, microfilming or other processes- or converted to a machine language, especially for data processing equipment- without the written permission of the publisher. The rights of reproduction by lecture, radio and television transmission, magnetic sound recording or similar means are also reserved.

Printed in GERMANY – ISSN 1018-4619

FEB - EDITORIAL BOARD Environmental Toxicology: Prof. Dr. H. Greim Chief Editor: Senatskomm. d. DFG z. Prüfung gesundheitsschädl. Arbeitsstoffe TU München, 85350 Freising-Weihenstephan, Germany Prof. Dr. H. Parlar Institut für Lebensmitteltechnologie und Analytische Chemie Prof. Dr. A. Kettrup TU München - 85350 Freising-Weihenstephan, Germany Institut für Lebensmitteltechnologie und Analytische Chemie e-mail: [email protected] TU München - 85350 Freising-Weihenstephan, Germany

FEB - ADVISORY BOARD Co-Editors: Environmental Analytical Chemistry: Environmental Analytical Chemistry: K. Ballschmitter, D - K. Bester, D - K. Fischer, D - R. Kallenborn, N Dr. D. Kotzias D.C.G. Muir, CAN - R. Niessner, D - W. Vetter, D – R. Spaccini, I Via Germania 29 21027 Barza (Va) ITALY Environmental Proteomic and Biology:

D. Adelung, D - G.I. Kvesitadze, GEOR A. Reichlmayr-Lais, D - C. Steinberg, D Environmental Proteomic and Biology:

Environmental Chemistry: Prof. Dr. A. Görg Fachgebiet Proteomik J.P. Lay, D - J. Burhenne, D - S. Nitz, D - R. Kreuzig, D TU München - 85350 Freising-Weihenstephan, Germany D. L. Swackhammer, U.S.A. - R. Zepp, U.S.A. – T. Alpay, TR V. Librando; I

Prof. Dr. A. Piccolo Università di Napoli “Frederico II”, Environmental Management: Dipto. Di Scienze Chimico-Agrarie Via Università 100, 80055 Portici (Napoli), Italy L.O. Ruzo, U.S.A - U. Schlottmann, D

Prof. Dr. G. Schüürmann UFZ-Umweltforschungszentrum, Environmental Toxicology: Sektion Chemische Ökotoxikologie Leipzig-Halle GmbH, K.-W. Schramm, D - H. Frank, D - D. Schulz-Jander, U.S.A. - Permoserstr.15, 04318 Leipzig, Germany H.U. Wolf, D – M. McLachlan, S

Environmental Chemistry: Prof. Dr. M. Bahadir Managing Editor: Institut für Ökologische Chemie und Abfallanalytik TU Braunschweig Dr. G. Leupold Hagenring 30, 38106 Braunschweig, Germany Editorial Chief-Officer:

Prof. Dr. M. Spiteller Selma Parlar Institut für Umweltforschung Universität Dortmund PSP- Parlar Scientific Publications Otto-Hahn-Str. 6, 44221 Dortmund, Germany Angerstr.12, 85354 Freising, Germany

e-mail: [email protected] - www.psp-parlar.de Prof. Dr. Ivan Holoubek RECETOX_TOCOEN Marketing Chief Manager: Kamenice 126/3, 62500 Brno, Czech Republic Max-Josef Kirchmaier MASELL-Agency for Marketing & Communication, Public-Rela- tions Environmental Management: Angerstr.12, 85354 Freising, Germany Dr. H. Schlesing e-mail: [email protected] - www.masell.com

Secretary General, EARTO, Rue de Luxembourg,3, 1000 Brussels, Belgium

Prof. Dr. F. Vosniakos T.E.I. of Thessaloniki, Applied Physics Lab. Abstracted/ indexed in: Biology & Environmental Sciences, P.O. Box 14561, 54101 Thessaloniki, Greece BIOSIS, C.A.B. International, Cambridge Scientific Abstracts, Chemical Abstracts, Current Awareness, Current Contents/ Agricul- Dr. K.I. Nikolaou ture, CSA Civil Engineering Abstracts, CSA Mechanical & Trans- Organization of the Master Plan & portation Engineering, IBIDS database, Information Ventures, NISC, Environmental Protection of Thessaloniki (OMPEPT) Research Alert, Science Citation Index Expanded (SCI Expanded), 54636 Thessaloniki, Greece SciSearch, Selected Water Resources Abstracts

0 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

CONTENTS

ORIGINAL PAPERS

POPULATION STRUCTURE ANALYSIS OF RED BARRACUDA, 263 SPHYRAENA PINGUIS (PERCIFORMES, SPHYRAENIDAE) IN THE NORTH-EASTERN MEDITERRANEAN SEA USING MORPHOMETRIC AND MERISTIC CHARACTERS Deniz Yaglioglu, Cemal Turan and Mevlut Gurlek

ASSESSMENT OF RIPARIAN ZONE AND RIVER ISLAND CONDITIONS 269 IN A TRANS-BOUNDARY GREENBELT: THE EVROS/MERIÇ RIVER (GREECE-TURKEY) Stamatis Zogaris, Vassiliki Markogianni, Saniye Cevher Özeren and Elias Dimitriou

THE EFFECTS OF THE INSECTICIDE PYRIPROXYFEN ON GERMINATION, 278 DEVELOPMENT AND GROWTH RESPONSES OF MAIZE SEEDLINGS Yasemin Coskun, Semra Kilic and Ragbet Ezgi Duran

DUST RETENTION EFFECTS OF POPULUS ALBA VAR. 285 PYRAMIDALIS (BUNGE) IN ARID OASIS CITIES, NORTHWEST CHINA Aliya Baidourela, Ümüt Halik, Tayierjiang Aishan, Abdulla Abliz and Aygvl Elyas

ADSORPTION OF ESTRONE BY SELECTED CONSTRUCTED WETLAND SUBSTRATES 291 Xiao-Li Yang, Lu Liang, Hai-Liang Song, Yun-Chen Gao, Shi-Bei Huang and Bao-Lin Dai

REMOVAL OF CADMIUM IONS FROM AQUEOUS 302 SOLUTIONS BY MICROORGANISMS OF ACTIVATED SLUDGE Sureyya Altin, Ahmet Altin, Bekir Fatih Kahraman, Sonay Alemdar and Elif Alaydin

TOXICITY OF MOSQUITO LARVICIDES ON NON-TARGET 311 MOSQUITO PREDATOR INSECT, BACKSWIMMER (Notonecta sp.) Onder Ser and Huseyin Cetin

AIRBORNE MICROFUNGUS FLORA DETERMINED IN THE DIFFERENT UNITS 317 OF THE DEPARTMENT OF TULAY AKTAS ONCOLOGY HOSPITAL, EGE UNIVERSITY Ozlem Abaci Gunyar, Alev Haliki-Uztan, Mustafa Ates, Aysegul Yoltas and Erdem Goker

PHOTOREDUCTION OF MERCURY (II) IN AQUEOUS SUSPENSIONS OF DIFFERENT ALGAE 324 Lin Deng, Jincheng Shi, Caiqian Yang and Nansheng Deng

PHYSICAL PROPERTIES AND FIBER DIMENSION 335 IN STEM, BRANCH AND ROOT OF ALDER WOOD Majid Kiaei and Roger Moya

INDUCED MUTAGENESIS AND GENOTOXICITY BY ACCUMULATED RADIONUCLIDES IN SOME 343 EDIBLE PLANTS CULTIVATED IN BLACK SAND SOIL DETECTED BY RAPD AND SDS-PAGE Ahmad K. Hegazy, Mohammad Faisal, Abdulrahman A. Alatar, Hanan F. Kabiel and Mona H. Emam

261 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

CHANGES IN PHOTOTROPHIC COMMUNITY STRUCTURE IN THE 355 VERTICAL PROFILE DURING SUMMER STRATIFICATION IN EUTROPHIC LAKE Elżbieta Szeląg-Wasielewska*, Natalia Jakubowska and Agnieszka Kaźmierska

EFFECT OF SEASONAL CHANGES ON ANTIOXIDANT 365 VITAMINS (A, E, C) AND MDA IN LEVELS OF THE RAINBOW TROUT (ONCORHYNCHUS MYKISS WALBAUM, 1792) GROWN IN VAN CITY, CATAK DISTRICT İbrahim Hakkı Yörük, Asli Çilingir Yeltekin and Halit Demir

RESPONSES OF ENZYME ACTIVITIES AND MICROBIAL COMMUNITY STRUCTURE 371 IN AN AERATED INTEGRATED VERTICAL-FLOW CONSTRUCTED WETLAND Min Tao, Feng He, Sheng Fang Li, Zhen-Bin Wu and Qi Shuo Wang

BIOLEACHING OF ARSENIC-CONTAINING GOLD ORE INFLUENCED BY CYSTEINE 379 Hongai Zheng, Chen Zhang,Yang Wu, Penghui Shi, Zhonghong Wang, Xuejuan Wang, Bingyue Li, Jianshe Liu and Xuehui Xie

PHYSIOLOGICAL AND BIOCHEMICAL RESPONSES OF 386 CUCURBITA PEPO L. MEDIATED BY PORTULACA OLERACEA L. ALLELOPATHY Hamed M. El-Shora and Ahmed M. Abd El-Gawad

COMBINING SATELLITE AND GIS DATA TO ANALYZE CHANGES 394 IN TROPICAL FORESTS ON CENTRAL HAINAN ISLAND IN RESPONSE TO THE NATIONAL LOGGING BAN AND ECONOMIC DEVELOPMENT Shudong Wang, Yuhe Ji, Zhiyun Ouyang, Lifu Zhang and Liu Jia

NOTICE

A REVIEW OF THE PROGRESS IN CHINA’S RENEWABLE ENERGY 405 GENERATION, AND ITS LINKS WITH ENERGY CONSERVATION, ENVIRONMENTAL DEGRADATION AND ECONOMIC GROWTH Qianyu Dong and Tohru Futawatari

INDEX 412

262 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

POPULATION STRUCTURE ANALYSIS OF RED BARRACUDA, SPHYRAENA PINGUIS (PERCIFORMES, SPHYRAENIDAE) IN THE NORTH-EASTERN MEDITERRANEAN SEA USING MORPHOMETRIC AND MERISTIC CHARACTERS

Deniz Yaglioglu1,*, Cemal Turan2 and Mevlut Gurlek2

1 Department of Biology, Faculty of Arts and Science, Duzce University, 81620, Duzce, Turkey 2 Marine Science and Technology Faculty, Mustafa Kemal University, Iskenderun, Hatay, Turkey.

ABSTRACT duction of Indo-Pacific species to the Mediterranean [1, 2]. Sphyraena pinguis (Günther, 1874) was recorded first in Population structure of red barracuda, Sphyraena pin- Palestine waters as a misidentification of Belone acus [3] guis, in north-eastern Mediterranean waters was investi- in the Mediterranean Sea. S. chrysotaenia is considered a gated using morphometric and meristic characters. In dis- synonyms of S. pinguis after the studies of Doiuchi and criminant function analysis, first, second and third discri- Nakabo [4, 5]. minant functions explained 48.6%, 30.2% and 21.2% of the S. pinguis has a wide distribution and distributed from between-group variation respectively. Plotting first and the Indo-Pacific, Red Sea, Persian Gulf and East Africa second discriminant functions explained 81.1% of the be- throughout the Indian Ocean to Australia and Japan [6]. So tween-group variation, and revealed the existence of four far this species has been recorded in the Mediterranean Sea morphologically distinct populations of S. pinguis in the off the coasts of Turkey [7, 8], Lebanon [9], Lybia [10], north-eastern Mediterranean coastal waters. A correct clas- Malta [11], and in the Adriatic Sea [12] and the Gulfs of sification of individuals into their original population was Gabes and Tunisia [13]. This species now constitutes an ranged from 93.3% to 96.7%, and the overall random as- important component of the inshore fisheries in the eastern signment was high (96.7%). The proportion of correctly Mediterranean [14-17]. S. pinguis is a bentho-pelagic spe- classified individuals into their original group was the same cies, inhabiting shallow waters not exceeding 50 m depth and greatest (96.7%) for the Iskenderun Bay and Syrian [2-17]. Its eggs and larvae are planktonic, and juveniles are samples. In hierarchical cluster analysis, the Mersin Bay often observed in very shallow waters. This species is car- sample was clustered as the first clade, while the Iskende- nivorous and feeds with other pelagic species, such as clu- run Bay and Syrian samples were in the neighbouring clades, peids and anchovy. and the Antalya Bay sample was clustered as the most diver- Effective management of exploited fish resources re- gent. Examination of the contribution of each morphomet- quires information on population structure of a species. So ric and meristic character to the first and second principal far no information is available for the population structure components revealed that the observed differences were of the lessepsian S. pinguis. Such information facilitates mainly from the characters associated on head length and our understanding of management strategies for conserv- eye diameter, indicating that these characters are important ing biodiversity. Ineffective understanding of the fish and in the detected differentiation of the populations. fishery management can lead to dramatic changes in the biological attributes and productivity of a species [18-21]. KEYWORDS: Red barracuda, population structure, North-eastern Morphological variation between fish populations is influ- Mediterranean, morphometric, meristic enced by a mixture of environmental and genetic factors [21-26]. When morphological variation is found between stocks it is not certain that there are genetic bases in the variation. However that variation is useful for initial under- 1. INTRODUCTION standing of status of stocks due to its simplicity and low- cost. Studies of morphologic variation among populations The opening of the Suez Canal in 1869 connected the continue to have an important role to play in stock identi- Red Sea to the Mediterranean Sea and allowed the intro- fication [27]. Morphometric and meristic characters have been most frequently used to delineate stocks of a variety * Corresponding author of exploited fish species [28-31]. Morphometric characters

263 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

are continuous characters describing aspects of body ble 1, Fig. 1). Taxonomic identification of the samples was shape. Meristic characters are the number of discrete, seri- carried out according to Turan (2007) [32]. Abbreviations ally repeated, countable structures that are fixed in em- of the sampling areas, sampling size and relevant infor- bryos or larvae [31]. mation for the collected samples are given in Table 1. Fol- The aim here is to investigate the population structure lowing capture, samples were placed individually into of the Indo-Pacific immigrant S. pinguis in the eastern plastic bags and were kept on ice until transportation to the Mediterranean Sea based on morphometric characters and laboratory. meristic characters. The truss network system described for fish body morphometrics [33] was used to construct a network of 14 landmarks determining 28 distances produced and measured as 2. MATERIALS AND METHODS illustrated in Fig 2. After collection of samples from different areas digital images from thawed samples were taken. Sphyraena pinguis were collected by commercial fish- MorFISH [34], the image processing tool specially devel- ing vessels from four fishing areas, comprising Iskenderun oped for morphometric measurements and analysis of fish Bay, Mersin Bay, Antalya Bay (Turkey) and Lattakia Port populations, was used for morphometric analysis. The (Syria Coast) in the north-eastern Mediterranean Seas (Ta- MorFISH automates the measurements of distances and

FIGURE 1 - The map of the sampling of S. pinguis. ● indicates sampling location.

TABLE 1 - Sampling details of S. pinguis used in this study. MSL, mean standard length (cm). Standard deviations of MSL are given in brackets.

Sampling area Abbreviation Sample size Sampling Area Coordinates MSL Collection date

360 36 29 N Iskenderun Bay NMS2 30 21.67 (1.75) 20.10.2011 350 59 18 E 360 30 20 N Mersin Bay NMS3 30 21.94 (1.57) 25.10.2011 340 32 47 E 360 14 39 N Antalya Bay NMS4 30 20.95 (1.23) 19.01.2011 310 10 25 E Syrian Coast 350 27 10 N NMS1 30 22.33 (1.27) 01.10.2011 (Lattakia Port) 350 38 35 E

264 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

accumulates them in a project file for each population and films of fish. Forward stepwise discriminant function anal- is capable of carrying out the statistical analysis for allome- ysis (DFA) based on the generalized Mahalanobis distance tric eliminations [35]. The standardized truss measure- was used to determine the similarity between populations ments showed no significant correlation with standard [36]. Population centroids with 95% confidence ellipses length. Therefore, the size effect has been successfully re- derived from the DFA were used to visualize relationships moved with the allometric transformation. Additional data, between the groups. The resultant discriminant functions such as head length (HL), eye diameter (ED), lengths of were used to assign individuals into the samples. The clas- anal fin (AL), first dorsal fin (DL1), second dorsal fin sification success rate was evaluated based on the percent- (DL2), pectoral fin (PL) and ventral fin (VL) were also rec- age of individuals correctly assigned into the original sam- orded. Only head width (HW) was manually measured. ple. The Mahalanobis distance for the morphometrics and Only undamaged fish were included in the analyses. meristics was compared to geographic distance using the Mantel (1967) [37] test to determine whether there was any correlation between morphologic and geographic distances. All statistical analyses were performed using SPSS v12 and SYSTAT v10.

3. RESULTS

Univariate statistics (ANOVA) showed that all of the FIGURE 2 - Locations of the 14 landmarks defining the truss network 34 morphometric measurements, excluding 7-10, head length on S. pinguis. and eye diameter, and 12 meristic characters were signifi- Meristic characters were examined under a binocular cantly different between the samples (P < 0.001). ANOVA microscope using the number of first dorsal soft and hard also revealed no statistical differences between males and fe- fin rays (DFSR1 and DFHR1), second dorsal soft and hard males for morphometric and meristic variables (P > 0.05). fin rays (DFSR2 and DFHR2), ventral fin soft and hard In discriminant function analysis, three discriminant rays (VFSR and VFHR), anal fin soft and hard rays (AFSR functions (DFs) were produced: the first (DF1) accounted and AFHR), pectoral fin rays (PFR), gill rakers (GR), num- for 48.6%, the second (DF2) accounted for 30.2% and the ber of intestinal villi, (IV) and scales in the lateral line (LS). third (DF3) accounted for 21.2% of the between group var- Vertebrae numbers (VN) were counted after taking X-ray iability among populations.

FIGURE 3 - Contribution of morphometric and meristic variables to the discriminant functions.

265 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

TABLE 2 - Correct classification results of populations. 94% of cross-validated grouped cases correctly classified. The sample abbreviations are given in Table 1.

Populations NMS1 NMS2 NMS3 NMS4 Total

NMS1 96.7 0 3.3 0 100

NMS2 3.3 93.3 3.3 0 100 Percent % NMS3 3.3 3.3 93.3 0 100

NMS4 3.3 3.3 0 93.3 100

NMS1 29 0 1 0 30

Individual NMS2 1 28 1 0 30 number NMS3 1 1 28 0 30

NMS4 1 1 0 28 30

The contribution of variables to the first discriminant showed a clear between-sample differentiation (Fig. 4). In function was mostly from morphometric measurements (5- the discriminant space, all the samples were clearly distinct 11, 5-10, 3,11, 6,11, 11,12) . Besides examination of the from each other. Remarkably, the Iskenderun Bay (NMS2) contribution of each variable to the second discriminant and Syria (NMS1) samples were close to each other, and function showed high contributions from measurements the Antalya Bay and Mersin Bay samples were mostly iso- DFHR2, 2-3, 12-13, 8-9 and 6-7 (Fig. 3). lated from each other and from all other samples.

The morphologic relationship between samples based A correct classification of individuals into their origi- on Mahalanobis distance is summarized in the form of a nal population varied between 93.3% and 96.7% by classi- neighbor joining tree (Fig. 5). In the first group, Mersin fication results (Table 2). The proportion of correctly clas- Bay population (NMS3) was most isolated and the Isken- sified Syria (NMS1) samples into their original group was derun Bay (NMS2) and Syrian samples (NMS1) clustered greatest (96.7%) among the samples. in the neighbouring clades, while the Antalya Bay popula-

tion (NMS4) clustered in another group as the most diver- gent (Fig. 5). High bootstrapping values were detected for each node on the neighbour joining tree. The correlation between geographic distance and Euclidean distance of meristic and morphometric characters of S. pinguis was also significant.

FIGURE 5 - Neighbour joining tree phenogram of morphologic relationships among populations of S. pinguis. The sample abbreviations are given in Table 1.

4. DISCUSSION

FIGURE 4 - 95% confidence ellipses of DFA scores for morphologic Morphometric and meristic data showed that there are analysis. discrete populations of S. pinguis in eastern Mediterranean waters. Morphometric and meristic data revealed four pop- The discriminant function 1 (DF1) and discriminant ulations of S. pinguis along the Turkish and Syrian coasts. function 2 (DF2) were plotted to allow visual examination The present morphometric study revealed evidence of of the distribution of each sample along the DF axis that highly significant morphometric heterogeneity among S.

266 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

pinguis populations. Only the neighbouring populations S. pinguis in the north-eastern Mediterranean coastal wa- NMS2 and NMS1 showed a morphologically closer rela- ters. The detected significant morphological differences do tionship than that of the others (Fig. 4, Fig. 5). This dis- not necessarily prove restricted gene flow between popula- tance indicates that geographic proximity is playing a role tions, but do suggest that fish in each group do not mix ex- between Mersin and Antalya populations. The isolation by tensively. Therefore, the detected high morphometric and distance model was supported for population differentia- meristic differentiation may indicate reproductive isolation tion of S. pinguis since the correlation between geographic between the populations of S. pinguis in the northeastern distance and Euclidean distance of the meristic and mor- Mediterranean. Genetic basis of observed morphologic dif- phometric was significant. ferentiation between the samples should be investigated with molecular genetic markers such as mtDNA and/or The close morphological relationship of the NMS2 (Is- nDNA markers. kenderun Bay population) and NMS1 (Syrian population) samples in DFA may suggest a limited degree of immigra- The authors have declared no conflict of interest. tion between these populations. Also, S pinguis may have immigrated to the Iskenderun bay from the Syrian coast and there has been not enough time to generate high phenotypic differentiation among populations since these REFERENCES coasts have similar environmental conditions. Interestingly there were clear differences between the Iskenderun, Mer- [1] Por, F.D. (1978) Lessepsian migration – the influx of Red Sea biota into the Mediterranean by way of the Suez Canal. Eco- sin and Antalya populations. These morphometric differ- logical Studies, Springer Verlag Berlin, pp. 228 ences may be attributed to the fact that the S. pinguis population along the Antalya shore does not only come [2] Golani, D. and Ben-Tuvia, A. (1995) Lessepsian migration and the Mediterranean fisheries of Israel. In, Armantrout NB through Mersin Bay. Since the lessepsian migration from (Ed): Conditions of the world’s aquatic habits. Proceedings of the Red Sea to the Mediterranean generally takes place on the World Fishery Congress, Theme 1, 279–289, New Delhi: the Syria-Iskenderun Bay-Mersin Bay route [8-38-39]. Oxford & IBH Publishing Moreover the Antalya Bay population is self-recruiting and [3] Spicer, I.J. (1931) Fisheries. In: Report of the Department of independent from the other populations. On the other hand Agriculture and Forests for the years 1927–30. Printing Office, the detected difference may be due to only environmental Russian Building, Jerusalem, 159–160 differences, since the Antalya, Mersin and Iskenderun [4] Doiuchi, R. and Nakabo, T. (2005) The Sphyraena obtusata Bays have different topographic and benthic structures. group (Perciformes: Sphyraenidae) with a description of a new species from southern Japan. Ichthyol Res, 52 (2), 132-151 The detected body shape differences between the pop- [5] Doiuchi, R. and Nakabo, T. (2007) Molecular evidence for the ulations may be related to different habitat characteristics, taxonomic status of three species of the Sphyraena obtusata such as temperature, turbidity, food availability, water group (Perciformes: Sphyraenidae) from East Asia Ichthyol depth and flow. For example, differences in the landmarks Res, 54 (3), 313-316 of 5-11, 2-3, 12-13 and second dorsal fin hard rays [6] Ben-Tuvia, A. (1986) Sphyraenidae. Fishes of the north-east- (DFHR2) may be attributed to water flow and food availa- ern Atlantic and the Mediterranean. In, (Eds): Whitehead PJP, bility. Antalya bay has different environmental conditions Bauchot ML, Hureau JC, Nielsen J, Tortonese E, UNESCO, Paris. Vol. 3, 1194-1196 compared to Iskenderun Bay, and such environmental differences may cause morphological differences among S. [7] Akyüz, E. (1957) Observations on the Iskenderun red mullet pinguis populations. The surface water salinity in the east (Mullus barbatus) and its environments. General Fisheries Council for the Mediterranean. Proceedings and Technical Pa- Mediterranean is 38,5‰ and the temperature varies be- pers FAO, Rome, 4, 305–326. tween 16- 27 °C [38]. These values are greater in İskende- [8] Ben-Tuvia, A. (1966) Red Sea fishes recently found in the run Bay, resulting in extreme salinity and temperature val- Mediterranean. Copeia, 2, 254-275 ues of, respectively 39.5‰ and 29.3°C [40]. In addition, the topographic structure and food availability in the Mer- [9] George, C.J., Athanassiou, V.A., Boulos, I. (1964) The fishes of the coastal waters of Lebanon. Misc Pap Nat Sci Am Univ sin and Iskenderun Bays are different and have an unusual Beirut, 4, 1–27, character of shoaling and expansion tendencies simultane- [10] Strin, J. (1970) Some notes on western trends of lessepsian mi- ously [41]. grations. Journees ichthyol Rome, CIESM, 187–190

[11] Lanfranco GG: (1993) The Fish around Malta (Central Medi- terranean). Progress Press Co., Malta, pp 132 5. CONCLUSION [12] Pallaoro, A. and Dulcic, C. (2001) First record of the Sphy- raena chrysotaenia (Klunzinger, 1884) (Pisces, Sphyraenidae) In conclusion, Population structure of red barracuda, from the Adriatic Sea. J Fish Biol 59, 179–182 Sphyraena pinguis, in north-eastern Mediterranean waters, including the Antalya, Mersin, Iskenderun Bays and Syrian [13] Bradai, M.N., Saidi, B., Ghorbel, M., Jarboui, O., Bouain, A., El Abed, A., Hamza, A. (2002) Statut des espèces exotiques waters, was investigated using morphometric and meristic signalées dans les eaux tunisiennes. Bull. Inst. Nat. Sci. Tech. characters. The discriminant function analysis revealed the de la Mer. Actes des 5émes Journees Tunisiennes des Sci- existence of four morphologically distinct populations of ences de la Mer, 7, 68–69

267 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

[14] Golani, D., Orsi-Relini, L., Massut´ı, E., Quignard, J.P. [32] Turan, C. (2007) Atlas and systematic of marine bony fishes CIESM Atlas of exotic species in the Mediterranean: Fishes. of Turkey. pp. 549, Nobel Publishing House, Adana pp. 254, CIESM: Monaco [33] Strauss, R.E. and Bookstein, F.L. (1982) The truss: body form [15] Wadie, W.F. and Riskallah, S.I. (2001) Fisheries for the genus reconstruction in morphometrics. Syst Zool, 31, 113–135 Sphyraena (Perciformes, sphyraenidae) in the southeastern part of the Mediterranean Sea. Pak J Mar Sci, 10: 21–34 [34] Turan, C., Oral, M. (2005) A computer package program for morphometric identifications of fish populations:MorFISH. [16] Zouari-Ktari, R., Ghorbel, M., Bradai, M.N., Ouannes-Ghor- In: ITAFE’05- International Congress on Information Tech- bel, A., Jarboui, O., Neifar, A. (2003) A. Analyse de deux cam- nologies in Agriculture, Food and Environment, Cukurova pagnes de chalutage benthiques effectuées dans la region du University, October, 12–14, Adana, Turkey, 143–147 golfe de Gabes. Bull. Inst. Nat. Sci. Techn. Mer. Actes des [35] Elliott, N.G., Haskard, K., Koslow, J.A. (1995) Morphometric 6èmes Journées de l’ATS Mer. Tunis (Tunisie) analysis of orange roughy (Hoplostethus atlanticus) off the [17] Zouari-Ktari, R., Ennejjar, S., Bradai, M.N., Ghorbel, M., continental slope of southern Australia. J Fish Biol, 46, 202– Bouain, A. (2007) Age and Growth of the Lessepsian Migrant 220 “Sphyraena chrysotaenia” ( Klunzinger, 1884) from the Gulf [36] Hair, J.F., Anderson, R., Tahtam, R., Black, W. (1996) Multi- of Gabes (Eastern Mediterranean), Rev Fish Sci, 15 (3), 169- variate Data Analysis with Readings. Prentice Hall Inc., pp. 181 449, New Jersey [18] Altukhov, Y.P. (1981) The stock concept from the viewpoint [37] Mantel, N. (1967) The detection of disease clustering and a of population genetics. Can. J. Fish. Aquat. Sci., 38, 1523– generalized regression approach. Cancer Res, 27, 209–220 1538 [38] Avşar, D. (1999) Physico-chemical characteristics of the East- [19] Ricker, W.E. (1981) Changes in the average size and age of ern Mediterranean in relation to distribution of the new Scy- pacific salmon. Can J Fish Aquat Sci, 1115, 1–117 phomedusae (Rhopilema nomadica). Turk J Zool, 23 (2), 605- 616 [20] Smith, P.J., Francis, R., McVeagh, M. (1991) Loss of genetic diversity due to fishing pressure. Fish Res, 10, 309–316 [39] Yaglioglu, D. and Turan, C. (2012) Colonization and genetic changes of Indo-Pacific immigrant Saurida undosquamis [21] Turan, C., Oral, M., Ozturk, B., Duzgunes, E. (2006) Morpho- (Richardson, 1848) (Lizardfish) in the Mediterranean Sea. J metric and meristic variation between stocks of Bluefish Black Sea/Medit Environ, 18 (3), 329-340 (Pomatomus saltatrix) in the Black, Marmara, Aegean and northeastern Mediterranean Seas. Fish Res, 79, 139–147 [40] Iyiduvar, O. (1986) Hyprograptic Characteristics of Iskende- run Bay. MSc. Thesis, Instute of Marine Sciences, Middle [22] Hubbs, C.L. (1926) The structural consequences of modifica- Eeas Technical University, Tr 33731, Erdemli- Mersin/ Tü- tions of the developmental rate in fishes, considered in refer- rkiye, pp. 157 ence to certain problems of evolution. Am Nat, 60, 57–81 [41] Serpin, D. (2007) Differences in the metric, merıstic characters [23] Vladykov, V.D. (1934) Environmental and taxonomic charac- and otolith element composition of the Brushtooth Lizardfish, ters of fishes. Trans R Can Inst, 20, 99–140 Saurida undosquamis (Richardson, 1848), from the different locations. Master Thesis, Mustafa Kemal University, Antakya, [24] Smith, G.R. (1966) Distribution and evolution of the North Hatay, Turkey, pp. 36 American catostomid fishes of the subgenus Pantosteus, genus Castostomus. pp.129, Miscellaneous Publications, Museum of Zoology, University of Michigan

[25] Lindsey, C.C. (1988) Factors controlling meristic variation. In, Hoar WS, Randall DJ (Eds): Fish Physiology, vol. 11-B. 197– 274, Academic Press, San Diego, CA [26] Turan, C. (2000) Otolith shape and meristic analysis of Her- ring (Clupea harengus) in the northeast Atlantic. Arch Fish Received: March 25, 2014 Mar Res, 48 (3), 283–295 Revised: June 02, 2014 Accepted: June 11, 2014 [27] Swain, D.P. and Foote, C.J. (1999) Stocks and chameleons: the use of phenotypic variation in stock identification. Fish Res, 43, 1123–1128 CORRESPONDING AUTHOR [28] Murta, A.G. (2000) Morphological variation of horse mackerel (Trachurus trachurus) in the Iberian and North Africa At- lantic: implications for stock identification. ICES J Mar Sci, Deniz Yaglioglu 57, 1240–1248 Department of Biology Faculty of Arts and Science [29] Silva, A. (2003) Morphometric variation among sardine (Sar- dina pilchardus) populations from the northeastern Atlantic Duzce University and the western Mediterranean. ICES J Mar Sci, 60, 1352– 81620 Duzce 1360 TURKEY [30] O’Reilly, K.M., Horn, M.H. (2004) Phenotypic variation among populations of Atherinops affinis (Atherinopsidae) Phone: +90 380 542 11 33 with insights from a geometric morphometric analysis. J Fish Fax: +90 380 541 24 03 Biol, 64, 1117–1135 E-mail: [email protected] [31] Turan, C. (2004) Stock identification of mediterranean horse [email protected] mackerel (Trachurus mediterraneus) using morphometric and meristic characters. ICES J Mar Sci, 61, 774–781 FEB/ Vol 24/ No 1b/ 2015 – pages 263 – 268

268 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

ASSESSMENT OF RIPARIAN ZONE AND RIVER ISLAND CONDITIONS IN A TRANS-BOUNDARY GREENBELT: THE EVROS/MERIÇ RIVER (GREECE-TURKEY)

Stamatis Zogaris1, Vassiliki Markogianni1,*, Saniye Cevher Özeren2 and Elias Dimitriou1

1 Hellenic Centre for Marine Research, Institute of Marine Biological Resources and Inland Waters, Anavissos Attiki Greece. 2Ankara University, Faculty of Sciences, Department of Biology, Hydrobiology Division Tandoğan, Ankara, Turkey.

ABSTRACT tween Greece and Turkey. This 218 km frontier is one of Satellite imagery is vital for assessing river riparian the most heavily militarized river borders in Europe and a conditions in areas that are inaccessible or rarely visited by notorious hotspot of irregular immigrant movement [1]. scientists. The lower Evros/Meriç River, shared by Greece Much of the riparian area is off limits to scientists and con- and Turkey, is one of the most heavily militarized river sequently baseline knowledge of the river and riparian area borders in Europe and access to the river corridor is strictly is incomplete. Serious gaps in aquatic biodiversity research restricted. This 218 km proposed green-belt conservation are shown by several new faunal discoveries in recent years area has never been completely surveyed for its riparian bi- [2,3]. However, the entire river border area was proposed odiversity values. Here we provide an assessment and evi- as a green-belt conservation area in IUCN publications dence for green-belt conservation justification for the en- since the late 1980s [4] and there is an ongoing reference tire portion of the Greek-Turkish river corridor for the first to its trans-frontier biodiversity conservation importance time. We employ satellite imagery to assess riparian corri- [5]. The wider Evros/Meriç valley’s international value for dor conditions and the distribution of natural habitat refuge wildlife, especially for birds, is remarkable and rather well environments focusing on riparian woodlands and river is- documented [6,7]. Two national parks in Greece and one lands. The analysis found that nearly half of the area of the in Turkey on the river valley as well as a Special Protected riparian green-belt is under cultivation; approximately Area along the riparian area in the Greek northern part of 22% of the riparian zone is wooded, while some of these the valley are all designated due to their outstanding orni- woodlands are taken over by artificial poplar plantations. thological and biodiversity values. The main reasons for The natural woodland areas are severely fragmented and the Evro’s/Meriç’s rich biodiversity is the position of this usually small in extent and width being affected by fre- river valley on the Thracian land-bridge; a biological cross- quent logging and clearing. The most extensive remaining roads between two continents and, the extensive aquatic woodlands are mapped; most of the 12 largest contiguous and riparian habitats associated with this large Balkan river natural woodland are in the northern half of the river val- system and its delta [7,8]. The management of trans-bound- ley. Important natural habitat refuges are the river’s is- ary zones raises many challenges due to different ap- lands; 219 such island features are mapped from recent sat- proaches in strategic planning and decision making. Espe- ellite images, and most are heavily wooded, again most are cially at the river basin area level, where the countries in- found in the northern half of the river valley. Based on best- volved are not all member states of the European Union, available information we interpret these patterns and en- the challenges are even more complex. Few remote sensing courage an international green-belt conservation initiative. applications have been conducted at the wider region of the Evros river, with most of them concerning flood monitoring [9,10], groundwater and surface water quality [11, 12] KEYWORDS: remote sensing, green-belt conservation, trans- boundary river, Greece, Turkey and land uses [13]. In this study, we apply the first rapid assessment of the condition of its riparian and river island features primarily through a remote sensing application. 1. INTRODUCTION

The river Evros (Maritsa in Bulgarian and Meriç in 2. MATERIALS AND METHODS Turkish) is the second largest river basin in the Balkans with a total main-stem river length of about 528 km. The Despite initial difficulty of access to the riparian zone, lower part of the Evros/Meriç river delimits the border be- many on-site field observations by the authors of this paper have recently been accomplished in related projects for wa- * Corresponding author ter monitoring and impact assessment [14, 15]. On these

269 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

surveys, important aquatic and riparian habitat areas (i.e. 2.2 Riparian land-use classification large woodland plots) have been identified. With the back- K-MEANS Classification algorithm was used to detect ground knowledge of the topography and local conditions, the basic land cover types [19-21] and the results were specific satellite imagery was acquired to analyze the con- compared with recent land use maps (Corine 2000 data- ditions of the general riparian zone. base) to assess their accuracy. This process was repeated by altering the number of classes until the output to be ac- 2.1 Satellite imagery manipulation curately classified according to the aforementioned supple- Riparian land cover classification of the Evros/Meriç mentary data (Corine 2000 Land Cover). K-MEANS clas- was conducted through satellite images from Ikonos and sification algorithm has been selected after several trials Landsat 5 remote sensors. Seven Ikonos imageries of with the appropriate imagery and comparison with the re- 06/09/2007 and two Landsat of 17/07/2009 with cell size sults of Iterative Self-Organizing Data Analysis (ISO- (x,y) 4 m and 30 m respectively, were used from the Euro- DATA). The classification result was a land cover map pean Satellite Agency (ESA) and the United States Geo- (LULC), in a raster format which was converted to a shape logical Survey (USGS). The imagery used in this study has file. Then, the extent of the land uses were quantified in been selected so as the acquisition date was in the summer- hectares (ha) and percentages of the total riparian zone area autumn season, when the vegetation cover in the study area under study. Land cover map categories comprised of the is dense (growing season). Unfortunately, recent Ikonos following four major generic features: water body, agricul- imageries cover only a specific part of Evros/Meriç river; tural land, woodland vegetation, and low scrub vegetation. Landsat imageries were used to map riparian land uses in the northernmost and southernmost parts of the river corri- 2.3 River islands detection and classification dor. These poorer resolution images concern the northern After the delineation of riparian land cover units in the part of the study area where the tri-point of Bulgaria, Greece riparian corridor, the next step was to detect the major sta- and Turkey, and the Evros/Meriç Delta in the extreme south. ble islands which appear within the river. The classification The data manipulation and analysis was conducted in of the riparian land uses resulted in a shape file with four ESRI’s ArcGIS 10.1 software while for the analysis of the attribute-classes, of which one was water body (i.e. the Ev- satellite imagery, ENVI 5.0 software was used. After select- ros/Meriç River). Through processing in ArcGis 10.1, the ing the study area scenes and the appropriate dates, the digi- river water body was extracted as a separate shapefile from tal data were submitted to the following procedures: the previous complete file of landuses. Further corrections a) Georeferencing of the imageries and geographical were conducted in order that riparian attributes, which were conversion from WGS’84 to EGSA ’87 coordinate not islands but represented some individual pixels of other system (National Datum) were performed using Beam classes, were extracted. Ikonos images though supplied no 4.7 software. information about the islands that are located near the tri- point frontier of Greece, Bulgaria and Turkey and at Evros/ b) Radiometric correction for the conversion of actual ra- Meriç Delta. These islands located at the northern and south- diance values. ern parts of Evros/Meriç were digitized by using Google The formula used in this process is as follows: Earth images and were transferred into ArcGis 10.1. Lλ = {(L - L ) / (Q - Q )}* MAXλ MINλ CALMAX CALMIN (Q – Q ) + L [16] Where L is the cell CAL CALMIN MINλ λ value as radiance 3. RESULTS QCAL is the digital number L is spectral radiance scales to Q MINλ CALMIN 3.1 Riparian land uses classification LMAXλ is spectral radiance scales to QCALMAX Land use classification was executed in the immediate QCALMIN is the minimum quantized calibrated pixel value (typically = 1) riparian zone along the main stem of the river with high and medium analyses (spatial resolution 4 m and 30 m). QCALMAX is the maximum quantized calibrated pixel value (typically = 255) The primary aim here focused on the detection of all riparian woodland vegetation cover, accurately delineated and c) Atmospheric correction through the darkest-pixel sub- quantified in the river’s immediate riparian zone. The ri- traction technique [17, 18] via the relevant ENVI 5.0 parian band width studied in detail was arbitrarily placed software tool. This method is one of the oldest and at approximately 10 km width along the river; the area this widely used procedures for adjusting digital remote covers is 112,448.21 ha in total. This riparian zone breadth sensing data for effects of atmospheric scattering. The is at least 10 times the river channel’s wetted width and is method's limited capabilities, relative to more sophisti- deemed adequate for an initial screening of riparian condi- cated methods, are at least partially offset by its wide tions in an agriculturally developed and embanked river applicability, due its requirement for little information corridor [22]. The derived land use map shows that there beyond the image itself. were relatively few errors after it was compared carefully d) Satellite bands of each imagery with the same spatial with the background of the satellite image. In a small per- resolution were joined in a single layer (layer stack- centage (less than 1%) confusion was observed between ag- ing) and saved in image format (Tiff, Geotiff). riculture and natural vegetation, especially in cases where

270 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

crops were at the peak of their development and demonstrated similar optical properties to the neighbouring riparian woodland, wetland or scrub vegetation according to the satellite images (09 / 2007). Examples of these delineations are presented here in three areas on the Evros/Meriç (Fig- ure 1; see three sites a,b,c); the exemplar land-use maps are depicted in Figure 2. Table 1 provides the quantitative areal cover results from the land use delineations. The most important woodland sites are also shown on the orientation map (Figure 1) and tabulated in terms of their location and areal extent in Table 2. As aforementioned, agricultural land is the dominant land use in the riparian zone of Evros/Meriç river while woodland vegetation covers approximately 22% of riparian corridor under study (Table 1). Concerning woodland vegetation class in general, natural forests dominate but areas of natural high scrub-woodland pastures, including regenerating degraded woodlands and high tamarisk and artificial poplar plantations are also included in this initial screening level category. The distribution of these woodlands, natural, semi-natural and artificial plantations, is extremely fragmented. What remains today as riparian woodland vegetation covers a small section of both sides of the river with a width which typically ranges from 0 – 250 m.

TABLE 1 - Evros/Meriç riparian land use areal coverage and areal percentages derived from the Ikonos imagery combination of 06/09/2007 and 17/07/2009.

Land use category Area (ha) % cover FIGURE 1 - The lower Evros/Meriç river study area with site classi- 52587.6 44.7 Agricultural land fication areal exemplars in red letters (see Fig 2) and the 12 most Woodland vegetation 25790.2 21.9 important surviving woodland sites (green patches with green nu- merals). Existing protected area locations are shown by the ‘tree Low scrub vegetation 33041.2 28.1 symbol’ and named in dark green. Water bodies 6167.8 5.3

TABLE 2 - Twelve significant woodland patches as delineated using recent 2013 Google Earth images (numeric codes are designated on the Map in Figure 1).

Code Lat Lon Nearest habitation Area (ha) River (km) 1 40°49'27.76" 26°12'8.90" Monastiraki-Yenikarpuzlu 94.4 17.7 2 41° 1'24.60" 26°21'40.76" Tichero-Balabancik 85.8 43 3 41°10'5.93" 26°18'42.27" Kornofolea-Kadidondurma 43.8 54.8 4 41°11'32.23" 26°19'27.69" Soufli-Umurca 34.2 57.5 5 41°14'57.67" 26°21'49.78" Lavara-Alibey 36.9 65.5 6 41°20'11.64" 26°31'56.13" Didimoticho-Gemici 95.9 80.3 7 41°27'58.55" 26°36'23.98" N. Chimonio-Elcili 204.3 97.8 8 41°29'58.42" 26°36'8.29" Orestiada-Doyran 39.7 99 9 41°36'28.01" 26°35'31.73" N. Vissa-Tayakadin 171 106.7 10 41°39'46.52" 26°32'26.29" Andrianoupoli-Karaagac 38.6 113.4 11 41°39'31.80" 26°29'38.13" Kastanies-Rizia 354.6 108.4 12 41°42'32.40" 26°21'39.16" Dilofos-Kemal 109.7 111.6

271 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

FIGURE 2 - Evros/Meriç riparian land uses at selected areas (as shown in Figure 1 at respective locations: a, b and c)

272 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

3.2 Important woodland sites and least fragmented forests strips are in the northern half Despite the fragmented nature of remaining woodland of the valley. formations, there are several notable exceptions where the riparian forest seems almost untouched in recent decades 3.3 River island detection and land-use classification by human activities. These patches were easily apparent in The largest islands and most islands are present mainly the initial satellite screening and they were carefully delin- at the northern part of Evros/Meriç main stem. All islands eated on the most recent Google Earth images (2013 im- are mapped and presented according to their location and ages). In these delineations the large river tributary conflu- area (Figure 3). Our inventory based on the available sum- ence area of the Ardas-Evros is also included. The resulting mer-period images defines 219 islands with a total area of delineations identify the largest surviving forest patches; 277.5 ha and total perimeter 90902.4 m (Table 3). 33.8% these are extremely important because they maintain true of the islands are very small, with an area between 47.3 and ‘inner forest’ attributes and higher longitudinal continuity 1000 m2; 31.1% islands are between 1000 and 5000 m2; along the river. The width of these forested corridor 11.4% between 5000 and 10000 m2; 14.1% between 10000 patches usually ranges from 50 to 350 m (from bank to in- and 30000 m2; and 9.6% are much larger than 30000 m2. land). The most extensive such forest site is on the Ardas tributary near its confluence with the Evros (in Greek ter- TABLE 3 - Classification of Evros/Meriç’s islands (N: 219) according ritory), covering an area of 800 hectares. A tally of the al- to areal size categories. located forests shows approximately 3,300 hectares of re- Area (m2) Number of islands maining riparian forest of significant width within Greek 47.3-1000 74 territory. On the side of Turkey there are slightly smaller 1000-5000 68 areas of riparian forest of significant width (approximately 5000-10000 25 2,500 hectares). Apart from these larger patches, it is re- 10000-30000 31 markable that along most of the river the riparian forest >30000 21 strip is usually very narrow. 90% of the satellite calculated TOTAL 219 vegetation in the riparian zone covers narrow aspects of TOTAL PERIMETER (m) 90902.4 vegetation (from the river bank up to 50 m of both sides of TOTAL AREA (ha) 277.5 the river). On both the Greek and Turkish sides the largest

FIGURE 3 - The Evros/Meriç river valley with all island locations (left) and their areal class size classification (right).

273 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

An attempt was made to visually assess, via recent 4. DISCUSSION Google Earth (2013) imagery, the percentage of woodland that covers each island. It was decided that only the islands 4.1 The screening process evaluation with area greater than 1000 m2 would be reliably assessed Riparian zones host complex vegetation patterns and because it is quite difficult to visually define substantial often change rapidly under human land-use pressures cre- “woodland” conditions on the really small islets. The is- ating a state of “perpetual succession” [23]. The natural lands to be assessed were 145. Three levels of woodland complexity and varied anthropogenic disturbance on vege- cover classes are visually estimated: from 0% to 30%; 30% tation patterns along the corridor led us to define and map to 80 %; and more than 80 % of woodland cover, respec- only four generic land use categories, and these subsume tively (see Table 4). Figure 4 shows that the distribution of various riparian habitat formations. The woodland cate- the most wooded islands is scattered throughout the north gory necessarily includes mature hybrid poplar plantations and mid-section of the river’s main stem. Interesting very which in many cases are planted among natural native ri- few sand bars or beaches exist with a main exception being parian forest. Agricultural land includes the land uses of immediately after the confluence of the Tundza and Ardas, permanently irrigated and non-irrigated arable lands. Since where we observe a concentration of islands with low per- extensive rice paddies are present in the Turkish portion centages of woodland cover. and large floods frequently take place beyond the older embankments in both countries, determination of natural wet- TABLE 4 - Island woodland cover classification pertaining to islands with areal extent greater than 1000 m2 (N: 145). Three classification lands is especially challenging and should be done in a sep- categories are defined based on percentage of woodland cover. arate application. The water body category includes only the Evros/Meriç River channel and major deep-water ca- Woodland Percentages (%) Number of islands classification nals. The low scrub vegetation category includes a variety 1 0-30 32 of habitat formations many of which are in transitional veg- 2 30-80 46 etation succession stages: low degraded scattered riparian 3 >80 67 shrubland, previously cultivated patches, grazed scrub, rock and gravel deposits, sand-mud flats, bars and beaches, and sparse artificial constructions. Despite the limited number of categories used and the necessary breadth of the definitions, the results of this screening process are considered as highly accurate delineations.

In its natural state the Evros/Meriç corridor would have been covered with extensive riparian forests, floodplain wetlands and river islands as documented in reference sites on the nearby Nestos river [24] and other Balkan rivers [7, 25]. However, the Evros/Meriç has changed considerably since the beginning of the 20th century when it was navigable up to the town of Soufli [6]. The river flow re- gime has been degraded by extensive water over-exploitation and upstream dam development, one of the prominent aspects of hydrological change [14]. The expansion of crops and the simultaneous deforestation of the riparian zone intensified in the 1960s and continued with declining rate after 1986 [26]. In localized situations a vigorous regeneration of riparian woodland patches near the river and in drainage ditch banks was observed (e.g. near Orestiada, in the northern part of the corridor). However Google Earth historic satellite photos (2002-2010) also show that extensive and frequent woodland clear-cutting still occurs near the river and this may be for crop expansion, local logging, and military reasons. Today deforested conditions on and embanked river may also affect hydromorphological qualities. Woodland vegetation plays an active role in developing heterogeneous and natural channel forms through both biotic processes (vegetative regeneration) and abiotic effects such as increasing flow resistance affecting sedimentation, protection from bank erosion and sediment transport FIGURE 4 - River islands categorized in three woodland cover clas- dynamics [23, 27, 28]. ses.

274 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

4.2 Interpreting riparian forest and river island condition of the river. Limited access to scientists and conservationists Identifying remaining woodland cover is important be- mainly due to the military zone restrictions has impeded cause this is the natural vegetation along the riparian corri- aquatic and riparian biodiversity surveys, and this is also a dor in the northern Mediterranean [25] and in most south- barrier to wider conservation proposals. eastern European perennial rivers [23]. The regenerative An outstanding feature of the Evros/Meriç is that this power of this type of forest cover is affected by anthropo- lower segment of a large lowland river maintains an ab- genic disturbance, river processes and local watershed con- sence of dams or any artificial barriers in its longitudinal ditions. Limiting factors that may effect forest cover in- connectivity. This condition is very rare in lowland Euro- clude salinity. In the delta of Evros/Meriç extensive tama- pean rivers and it is important for giving the river outstand- risk (Tamarix spp.) shrublands dominate because they tol- ing ichthyological value [14]. Compared to other rivers in erate saline soil conditions which have increased and ex- Greece, the Evros/Meriç holds a remarkable area of high panded due to ground water degradation by anthropogenic quality and extensive riparian woodland [7, 24, 33]. The aquifer over-abstraction [29]. In the upper delta though, wildlife values of the entire river are also apparent in studies there are in some places riverside tree-stands with high wil- concerned with the isolated existing protected areas [6, 8, lows (Salix alba) but tree species diversity is lower than 32]. Because of its position on an international border the further upstream. Apart from river flooding disturbances corridor is well suited as an area where semi-natural river and water-logged wetland conditions only direct anthropo- and riparian conditions can be preserved for their biodiver- genic disturbance limits forest cover upstream of the delta. sity at relatively low cost (i.e. no navigation issues, no dams Islands are characteristic biogeomorphic features of etc). But could this river corridor become important for con- large rivers and are often created by woody debris which servation as a riparian green-belt, similar to other borders in has its source in riparian forests [30]; they are important for Eastern Europe since the end of the Cold War? lowland river functioning and have been effectively used Mention of the opportunity for a trans-frontier pro- to monitor hydromorpholgical status [31]. River reaches tected area between Greece and Turkey was promoted at with vegetated islands have high in-stream habitat diversity an IUCN conference in 1988 [4] and the international in- [27]. Most islands, islets and bars, including the largest is- terest in a green-belt protected area is well documented, the lands, are located in the northern half of the Evros/Meriç Evros/Meriç has often been repeatedly cited as an im- river corridor which coincides with the area where the larg- portant green-belt opportunity [5, 33, 34]. Other successful est and most extensive riparian woodlands are located. Im- approaches to international green-belt cooperation and portant areas with large floodplain and island formations planning in the immediate vicinity have shown effective are near the river confluences of the Tundza, Ardas and conservation cooperation between Greece and neighbour- Erythropotamos. It is remarkable that the southern half of ing Bulgaria [35,36]. However, despite these preliminary the Evros/Meriç has very few islands, even though there is proposals, no lasting cooperative initiatives and any pol- a major river confluence (Ergene Tributary) and the large icy-based biodiversity conservation on this border exist be- river meanders in former floodplains. This nearly ‘island- tween Greece and Turkey [1, 36, 37]. less’ condition may be due to the more intensive river en- For Turkey and Greece, the Evros/Meriç river is an gineering efforts to drain floodplains, straighten and em- area of important flood protection issues [37], while quality bank the river near this sensitive part of Greek-Turkish bor- and quantity of waters are also deemed of outstanding im- der-crossing. Significant changes due to water abstraction, portance [36, 38]. Unfortunately, biodiversity values have embankment building and severe hydro-peaking floods not been considered in this area’s trans-boundary water (from upstream dams) may have degraded island condi- management issues. This border zone is an area of interna- tions in the river as well [27]. The ecological history of the tional border friction, so sensitive multidisciplinary ap- river’s changes has been poorly studied [6] and more re- proaches are needed if conservation initiates are to be de- search is needed to better interpret the pattern of floodplain veloped in serious long-term investments by both countries formations and river island conditions. [37, 38]. Ecosystem and biodiversity conservation along a proposed river green-belt is an initiative that emphasizes a 4.3 Conservation implications unique identity for the Evros/Meriç corridor and this may Biodiversity conservation has focused for many years provide the enabling environment for dialogue on ecosys- on protection of specific show-case parks and not on the tem-based management, as has been done in many other wider environment or on large-scale restoration initiatives. Eastern European borders [39]. The example of Dadia Na- In conservation terms, the Evros/Meriç valley lies in the tional Park, as a focus for eco-development and ecotourism ‘shadow’ of three high-level protected areas, Dadia National in the Greek part of the Evros valley which has supported Park, Evros Delta National Park and Gala Lake National ecosystem-based conservation, is of global importance [8, Park. The wider riparian zone of this large lowland river has 40, 41] and is immediately adjacent to the Evros/Meriç pro- not been given due conservation attention. With the excep- posed green-belt. A coordination structure including the tion of a recently enacted Natura 2000 site in the northern creation of a panel of international experts of the Evros’ part of the riparian corridor in Greece [32], very little is ever three riparian countries may be considered as an initial step reported on aquatic and riparian values along the main-stem for green-belt conservation planning.

275 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

An important issue in the Evros/Meriç is an oppor- following ad hoc structural solutions (e.g. anthropocentric tunity for ecological restoration that may benefit flood pro- management approaches). Special attention to riparian tection, water body ecological quality, biodiversity conser- zone conditions and promoting forest and wetland conservation and international cooperation. The EU Water vation in floodplains can mean a more integrative conser- Framework Directive also demands these international co- vation approach that would assist in applying EU policy- operation efforts. If a strategic plan is promoted that will relevant conservation and water management frameworks. include biodiversity conservation, a golden opportunity for restoring sensitive habitats such as woodlands and islands The authors have declared no conflict of interest. can be initiated beyond routine river engineering and flood protection which threatens to further degrade the natural integrity of this lowland river [32]. REFERENCES

[1] Demetriou, O. (2012) Capricious Borders: Minority, Population, and 5. CONCLUSIONS Counter-Conduct between Greece and Turkey, Berghan Journals. [2] Koutrakis, E., Sapounidis, A., Favre-Krey, L., Krey, G. and Econ- This initial survey is the first inventory of the Ev- omidism, P.S. (2011) Incidental catches of Acipenseridae in the estuary of the River Evros, Greece. Journal of Applied Ichthyol- ros/Meriç river’s riparian woodlands and islands, many of ogy, 27: 366-368. which are part of the ‘shared’ frontier river-border. Access [3] Zogaris, S. and Apostolou, A. (2012) First record of Pontian Mon- to much of the riparian zone and the river is restricted so key Goby, Neogobius fluviatilis (Pallas, 1814) in the Evros River remote sensing approaches are instructive for a first assess- (Greece); Is it an alien species? Mediterranean Marine Science, ment. This survey shows that the riparian zone of a large 12(2), 454-461. portion of the Evros/Meriç river is degraded by human [4] Thorsell, J. (1990) Parks on the Borderline: Experience in Transfron- tier Conservation. Cambridge, IUCN, Publication Services, UK. pressures acting in both countries. Intensive agriculture, [5] Biró, E., I. Bouwma, I. and V. Grobelnik, V. (Eds.) (2006) Indicative forestry and associated wetland drainage, river engineering map of the Pan-European Ecological Network in South-Eastern Eu- works (especially embankments) and wood-cutting have rope. Technical background document. European Centre for Nature effectively cleared and degraded natural vegetation. How- Conservation, ECNC technical report series, Tilburg. ever, extensive woodland refuges do exist. Most of the larger [6] Handrinos, G., Athanasiadis, A. and Rigas, Y. (2005) Evros Delta – 219 islands identified in this assessment are also densely princess of Thrace’s nature, ANEE A.E., Oikos LTD, KOAN Press, Athens. wooded. So these are important natural biodiversity refuges [7] Skoulikidis, N., Economou, A., Gritzalis, K. and Zogaris S. (2009) on the Evros/Meriç river enhancing the corridor’s ecological Rivers of the Balkans. In: Tockner, K., Uehlinger, U., Robinson, integrity and rich biodiversity. Dead woody debris from ri- C.T. (Eds.), Academic Press, first ed., Rivers of Europe, pp. 421- parian forests is critical in creating river islands and bars. 466, UK. Since flooding is frequent on the wider floodplain, wetlands [8] Catsadorakis, G. and Kallander, H (Eds.) (2009) The Dadia- are also important constituents although of limited size with Lefkimi-Soufli Forest National Park, Greece: Biodiversity, Man- agement, and Conservation. WWF-Greece, Athens. the exception of the Delta area. Detailed satellite-based hab- [9] Avci, D. U., Göral, B., Akkartal, A. and Sunar, F. (2006) Flood itat surveys must include wetlands and all natural riparian monitoring using multi-temporal radarsat-1 images. In Proceed- habitat types in future work. ings of the Remote Sensing and Photogrammetry Society (RSPSoc) Annual Conference, 2006, Newcastle, UK. Proceedings, The Evros/Meriç provides a unique trans-boundary op- pp. 12-14. portunity for conservation and restoration work which [10] Mentzafou, A., Markogianni, V. and Dimitriou, E. (2014) Flood should initiate dialogue along a border of political friction. risk mapping and assessment in a transboundary river by using sat- A renewed initiative for a trans-boundary green-belt pro- ellite data and GIS techniques, In Proceedings of the 22nd Carto- graphic School: Geoinformatics and Atmospheric Science, tected area should be promoted between Greece and Tur- Wroclaw-Ksiaz, Poland. Proceedings p. 21. key since functioning protected areas already exist on ei- [11] Dimitriou, E., Mentzafou, A., Markogianni, V., Tzortziou, M. and ther side of the border. A wider trans-boundary green-belt Zeri C. (2012). Geospatial investigation into groundwater pollu- along the entire stretch of this river corridor should be de- tion and water quality supported by satellite data: A case study lineated and managed as a protected area. Trans-boundary from the Evros River (Eastern Mediterranean). Pure and Applied Geophysics, DOI 10.1007/s00024-012-0621-2. conservation areas are increasingly important in protecting [12] Markogianni, V., Dimitriou, E. and Tzortziou M. (2013) Monitor- and maintaining complex ecosystems on a larger scale than ing of chlorophyll-a and turbidity in Evros River (Greece) using can be accomplished in a single protected area within one Landsat imagery, /Proc. SPIE/ 8795, First International Confer- country. Trans-boundary conservation initiatives and asso- ence on Remote Sensing and Geoinformation of the Environment ciated linked efforts can bring many benefits that will af- (RSCy2013), doi:10.1117/12.2027047. fect research opportunities, shared management and moni- [13] Silleos, N., Perakis K. and Petsanis G. (2002) Assessment of crop damage using space remote sensing and GIS. International Journal toring, international cooperation and international peace. of Remote Sensing, 23(3): 417-427. We urge more effort be applied at the international level of [14] Dimitriou, E., Mentzafou, A., Zogaris, S., Tzortziou, M., Gritzalis, integrated river management, not just on simplistic techno- K., Karaouzas, I. and Nikolaidis, Ch. (2012) Assessing the environ- cratic resource management of the Evros/Meriç. This river mental status and identifying the dominant pressures of a trans- boundary river catchment, to facilitate efficient management and corridor has important flooding, water management and mitigation practices. Environmental Earth Sciences, 66(7): 1839- conservation problems that cannot be adequately managed 1852.

276 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

[15] Özsoy, E.D., Erkmen, B., Özeren, S. C. and Kolankaya, D. [34] Schneider-Jacoby, M. (2004) Conservation without Frontiers - To- (2012) Detection of aquatic pollution in Meriç River by a measure wards a new Image for the Balkans. A Strategic Plan for the IUCN of developmental instability, fluctuating asymmetry in the South-Eastern European Programme. Compiled by EU- fish. Journal of Animal and Veterinary Science, 11 (8): 1213-1216. RONATUR for the IUCN Regional Office for Europe (ROFE) in cooperation with IUCN/WCPA. [16] YCEO - Yale Center for Earth Observation, Yale Institute of Bio- spheric Studies (2010), Index of CEO Documentation [35] Popov, D. (2011) Cross-sectoral Cooperation for Bird Conserva- (http://www.yale.edu/ceo/Documentation/). tion - Green Balkans NGO and Bulgarian Chief Directorate Border Police. In: Vasilijević, M., Pezold, T. (eds.), Crossing Borders for [17] Keiner, L.E. and Yan, X. (1998) A neural network model for esti- Nature: European examples of transboundary conservation. Gland, mating sea surface chlorophyll and sediments form Thematic Map- Switzerland and Belgrade, Serbia: IUCN Programme Office for per imagery. Remote Sensing of Environment, 66: 153–165. South-Eastern Europe, pp. 50-54. [18] Lathrop, R.G., Lillesand, T.M. and Yandell, B.S. (1991) Testing [36] Schindler, S. Curado, N., Nikolov, S.C., Kret, E., Carcamo, B, the utility of simple multi-date thematic mapper calibration algo- Catsadorakis, G., Poirazidis, K., Wrbka, T. and Kati, V. (2011) rithms for monitoring turbid inland waters. International Journal of From research to implementation: Nature conservation in the East- Remote Sensing, 10: 2045–2063. ern Rhodopes mountains (Greece and Bulgaria), European Green [19] Han, K.S., Champeaux, J.L., and Roujean, J.L. (2004) A land Belt. Journal for Nature Conservation, 19: 193– 201. cover classification product over France at 1 km resolution using [37] Sakcali, M. S., Yilmaz, R., Gücel, S., Yarci, C. and Ozturk, M. SPOT4/VEGETATION data. Remote Sensing of Environment, (2009) Water pollution studies in the rivers of the Edirne Region- 92: 52–66. Turkey. Aquatic Ecosystem Health & Management, 12(3), 313 – [20] Zalik, K.R. (2008) "An Efficient K-Means Clustering Algorithm", 319. Pattern Recognition Letters, 29: 1385–1391. [38] Kramer, A. and Schellig, A. (2011) Meriç River Basin: Trans- [21] Triloki, P., Singh, D. and Srivastava, T. (2010) Advanced fractal boundary Water Cooperation at the Border between the EU and approach for unsupervised classification of SAR images. Ad- Turkey. In Kibaroglu, A., Kramer, A., Scheumann, W., Turkey's vances in Space Research, 45: 1338–1349. Water Policy: National Frameworks and International Coopera- tion. Heidelberg, Springer, pp 229-249. [22] Verry, E.S., Dolloff, C.A. and Manning, M.E. (2004) Riparian eco- tone: a functional definition and delineation for resource assess- [39] Vasilijević, M., and Pezold, T. (eds.) (2011) Crossing Borders for ment. Water, Air, and Soil Pollution 4: 67-94. Nature. European examples of transboundary conservation. Gland, Switzerland and Belgrade, Serbia: IUCN Programme Of- [23] Naiman, R.J., Decamps, H. and McClain, M.E. (2005) Riparia - fice for South-Eastern Europe. ecology, conservation and management of streamside communities. Elsevier Academic Press Publications, London UK. [40] Zogaris, S., Chatzinikolaou, Y. and Dimopoulos P. (2009) As- sessing environmental degradation of montane riparian zones in [24] Vasilopoulos, G., Tsiripidis, I., and Karagiannakidou, V. (2007) Do Greece. Journal of Environmental Biology, 30(5): 719-726 abandoned tree plantations resemble natural riparian forests? A case study from northeast Greece. Botanica Helvetica, 117: 125-142. [41] Skias, S. and Kallioras, A. (2007) Cross border co-operation and the problem of flooding in the Evros Delta. In Verwijmeren, J. and [25] Dimopoulos, P. and Zogaris, S. (2008) Vegetation and flora of ri- Wiering, M. (Eds), Many rivers to cross: cross border co-operation parian zones. In “Sustainable Riparian Zones: A Management in river management. Eburon Academic Publishers, Delft, The Guide (eds: D.Arizpe, A.Mendes, & J.Rabaca), pp. 66-82, Gener- Netherlands, pp. 119-145. alitat Valenciana, Valencia, Spain. (English Edition ISBN:978-84- 482-4967-0). [26] Dimitriou, E., Moussoulis, E., Mentzafou, A., Gritzalis, K., Zog- garis, S., Karaouzas, I., Tzortziou, M., Zeri, Ch., Colombari, E., Markogianni, V. and Fotopoulou A. (2010) Environmental status assessment for Evros river basin‘, Final Technical Report, IIW-

HCMR. Available at: http://imbriw.hcmr.gr/en/environmental-status-of-evros-river-catchment-pressures/ [27] Ligon, F.K., Dietrich, W.E. and Trush, W.J. (1995) Downstream ecological effects of dams: a geomorphic perspective. BioScience, 45:183–192.

[28] Gurnell, A.M. and Petts, G.E. (2002) Island-dominated landscapes Received: April 02, 2014 of large floodplain rivers, a European perspective. Freshwater Bi- Revised: June 06, 2014 ology, 47(4): 581–600. Accepted: June 25, 2014 [29] Diamantis, J.B. and Pealas, C.P. (1989) Seawater intrusion into coastal aquifers of Thrace and its impact on the environment. Tox- CORRESPONDING AUTHOR icological and Environmental Chemistry, 20-21(1): 291-305. [30] Corenblit, D., Steiger, J., Gurnell, A.M. and Naiman, R.J. (2009) Plants intertwine fluvial landform dynamics with ecological suc- Vassiliki Markogianni cession and natural selection: a niche construction perspective for Hellenic Centre for Marine Research (HCMR), Insti- riparian systems. Global Ecology and Biogeography, 18: 507–520. tute of Marine Biological Resources and Inland Waters [31] Ling, F. Cai, X., Li, W., Xiao, F., Li, X., and Du, Y. (2012) Mon- 46.7 km Athens-Sounio Ave. itoring river discharge with remotely sensed imagery using river 19013 Anavissos, Attica island area as an indicator. Journal of Applied Remote Sensing, 6(1) doi:10.1117/1.JRS.6.063564 GREECE

[32] Ormerod, S.J. (2004) A golden age of river restoration science? Aquatic Conservation: Marine and Freshwater Ecosystems, 14: Phone: +30-2291076349 543–549. Fax:+30-2291076419 [33] Kortelainen, J. (2010) The European Greenbelt: Generating envi- E-mail: [email protected] ronmental governance reshaping border areas. Quaestiones Geo- graphicae, 29(4): 27-40. FEB/ Vol 24/ No 1b/ 2015 – pages 269 – 277

277 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

THE EFFECTS OF THE INSECTICIDE PYRIPROXYFEN ON GERMINATION, DEVELOPMENT AND GROWTH RESPONSES OF MAIZE SEEDLINGS

Yasemin Coskun*, Semra Kilic and Ragbet Ezgi Duran

Suleyman Demirel University, Faculty of Arts and Sciences, Department of Biology, Isparta, Turkey

ABSTRACT 1. INTRODUCTION

Phytotoxicity is defined as the detrimental effects of Pesticides are chemical substances that are used to elim- chemical products used for pest eliminating or growth reg- inate or remove pests which damage agricultural products ulating purposes on some morphological, anatomical and during production, consumption and storage of nutrients, physiological processes of plants. We examined the effects causing product loss [1-3]. Chemical control of harmful or- of pyriproxyfen, which is used to protect crops against ganisms is being strictly managed. Pesticides are applied to whitefly, to determine whether it adversely affects maize crops throughout the world, but extensive application in ag- plants. For this purpose, maize (Zea mays L. saccharata ricultural areas, horticulture and home gardens leads to en- Sturt.) seeds were treated with several pyriproxyfen con- vironmental pollution, especially in soil, water and air. Pes- centrations (0.0, 0.1, 0.2, 0.4, and 0.6 ppm) for 72 h under ticide residue, left after treatment, may penetrate into plant controlled conditions. The results showed that, while low tissues, and can cause various deformations in their struc- concentrations of the insecticide had little effect on the tures [4-7]. More importantly, the involvement of pesticide growth of maize seedlings, increasing the concentration into the food chain may have adverse effects on human and gradually had harmful effects on the growth. The highest other species [8], and especially the accumulation of pesticide insecticide concentration led to a remarkable decrease in residues is known to have potential cancerogenic effects [9]. all the growth parameters of maize seedlings. It was found Insecticides have become more widespread as a new that maize growth was significantly inhibited by increasing way to harvest more crops. As a result, living things have the concentration of pyriproxyfen, which adversely af- been exposed to a wide variety of insecticides lately. fected seed germination and seedling growth. Stomata in- Pyriproxyfen [4-phenoxyphenyl (RS)-2-(2-pyridyloxy) dex decreased in both adaxial and abaxial surfaces of propyl ether] is an influential insecticide used to protect leaves treated with increasing concentrations compared crops against whitefly [10]. It is a pyridine-based pesticide with the control group. The adaxial surfaces of maize which easily dissolves in water, and harms both animals leaves always had fewer stomata than the abaxial surfaces. and plants in aquatic systems [11]. Thus, it is readily trans- As insecticide concentration increased, photosynthetic pig- located in plant tissue. Pyriproxyfen prevents larvae from ment contents decreased except for anthocyanin, but the developing into maturity (inhibits insect maturation pro- proline accumulation increased as an adaptation to toxic cess from cell division to elongation) and thus their repli- treatment. This work suggests that excessive and uncon- cation is inhibited [12, 13], but its activity against non-tar- trolled usage of pesticide pyriproxyfen results in phyto- get organisms such as plants has not been tested. However, toxic effects by inducing some morphological, anatomical, it is known that the presence of certain functional groups (- physiological and metabolic processes. In addition, as COOR, -CO.NH2, -NH2, -NHR, -NR3, -0H,) in the chemi- chemical compounds accumulate, pyriproxyfen is thought cal structure of the pesticides further facilitate adsorption to adversely affect all living beings feeding on plants that of the soil humus [14]. Pesticide residues dissolved in wa- are exposed to chemicals, and also negatively affects the ter reduces nutrient uptake from soil [15, 16] and depolar- environmental factors such as, soil, air and water. izes the plasma membrane of the root cells [17]. This can cause abnormalities in different growth parameters [18- 21]. Therefore, it is necessary to consider the possible haz- KEYWORDS: Insecticide, photosynthetic pigment, proline, pyriproxyfen, maize, stomatal index ardous effects of these insecticides used extensively in agriculture on ecological balance.

In the present work, the adverse effects of different

concentrations of pyriproxyfen have been evaluated on

growth parameters (morphological, anatomical and physi- * Corresponding author ological) of maize seedlings. In this way, ecotoxicological

278 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

profiles of pyriproxyfen can be determined in relation to samples of 100 mg FW with 2 ml of 40% methanol. 1 ml the structural modification of pyriproxyfen. extract was mixed with 1 ml of a mixture of glacial acetic acid and orthophosphoric acid (6 M) (3: 2, v/v) and 25 mg ninhydrin. After 1 h incubation at 100 °C, the tubes were 2. MATERIALS AND METHODS cooled and 5 ml toluene was added. The absorbance of the upper phase was spectrophotometrically determined at 528 2.1 Plant material and chemicals nm and total proline amount was calculated with the help -1 Uniform-sized seeds (n=25) of a commercial variety of of standard curve and expressed as µg proline g FW. maize (Zea mays L. saccharata Sturt.) were used as the test For Chlorophyll a, chlorophyll b, total chlorophyll, ca- plant. The Pyriproxyfen was obtained from Bayer. Other rotenoids and anthocyanin measurement, the tissue sam- chemicals were purchased from Sigma (USA) and Fluka ples (1.0 cm2 in most cases) were ground in 2 ml cold ace- AG (Buchs, Switzerland). tone/Tris buffer solution (80:20 vol:vol, pH = 7.8), centrifuged to remove particulates, and the supernatant was di- 2.2 Germination test luted to a final volume of 6 ml with additional acetone/Tris Seeds of maize were surface sterilized with 0.5% so- buffer. The absorbance of the extract solutions was meas- dium hypochlorite for 10 min, followed by extensive wash- ured with the spectrophotometer (Jenway 6300) using an ing in sterile distilled water. Treatment concentrations external cuvette holder [25]. were prepared using control (distilled water) and 0.1, 0.2, 0.4, and 0.6 ppm of original insecticide solution. The ap- 2.5 Statistical analysis plication doses were prepared as recommended by the The statistical analysis of variance (ANOVA) was per- manufacturing company on the label. Uniform-sized seeds formed on all of the experimental data reported in the pre- were placed into a clean 50 ml glass beaker and were present paper and statistical significance (P < 0.05) of means treated with the pyriproxyfen concentrations for 72 hours. of at least two independent assays with three replicates was At the end of 72 h, seeds of maize were sown in 12 cm petri judged by the Duncan’s New Multiple Range Test using dishes lined with two layers of filter paper (Whatman 1) SPSS software 13.0. All data were expressed as means ± moistened with 10 ml of the distilled water, and maintained S.D. in an incubator at 25°C for 7 days. The experiment was replicated three times at each concentration of Pyriproxyfen. Seeds were considered to be germinated with the emer- 3. RESULTS gence of the radicle. At the end of the 7th day, the radicle and coleoptile lengths of seedlings were measured with a 3.1 Seed germination and seedling growth responses millimetric ruler and germination percentages were calcu- There was a clear propensity that the growth of plant lated according to Khan and Ungar [22]. decreased with increasing dosage level of pyriproxyfen (Table 1). Statistical analysis showed that the germination 2.3 Anatomical parameters and seedling growth of the seeds treated with pyriproxyfen The seedlings of each application were placed in 4-L was rather different from the control group (P<0.05), espe- pots with perlite containing Hoagland’s nutrient solution. cially at the highest concentration of pyriproxyfen. For ex- Plants were cultivated in a climate chamber with controlled ample, the germination percentage of maize seeds treated conditions (photoperiod 12-h, temperature 25±2°C, rela- by 0.1, 0.2, 0.4 and 0.6 ppm of pyriproxyfen were de- tive humidity 60±5%, light intensity 160 μmol m-2 s-1) for creased by 2%, 8%, 16% and 40% respectively, and were 45 days. Seedling growth was provided with the nutrient dose-dependent. Plants were highly damaged at the con- solution added on a regular basis. Epidermal tissue was centration of 0.6 ppm. The elongation of radicle depended stripped from the adaxial and abaxial surfaces of leaf lam- on the pyriproxyfen concentration involved. The highest ina pieces to determine leaf stomatal density, expressed as level of its dose (0.6 ppm) almost totally inhibited the rad- the number of stomata per unit leaf area, mounted on a icle length. Similarly, the highest concentration of glass slide, immediately covered with a cover slip, and then pyriproxyfen on coleoptile length and radicle number were lightly pressured with fine-point tweezers. For each inde- observed, and at day 7 they declined by 36% and 21%, re- pendent measurement, numbers of stomata (s) and epider- spectively, compared with the control. mal cells (e) for each film strip were counted from both the adaxial and abaxial surfaces. The leaf stomatal index was 3.2 Anatomical responses calculated using the formula [s / (e + s)]×100, as defined Maize plants treated with pyriproxyfen on leaf epider- by Meidner and Mansfield [23]. Stomatal size was defined mis had lower stomata cell counts than untreated ones (Ta- as width and length in micrometres. ble 2). This pattern was observed on both adaxial and abaxial surfaces of maize leaves treated with an increasing 2.4 Physiological parameters tested dose. Stomata index for adaxial and abaxial surfaces of leaf The proline content was determined using to the were found to be lower than those of the control group in method of Bates et al. [24]. Proline was extracted from leaf treated plants. Maximum decrease in stomata index was

279 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

found on adaxial surface (68%) of leaf of maize plant treated control group (Figure 1). The proline content was increased with 0.6 ppm of pyriproxyfen. It suggests that pyriproxyfen significantly with increasing pesticide concentrations decreased the number of stomata and epidermis of maize (P<0.05). For example, proline content was raised by 28%, seedlings mainly by reducing stomata index. Besides, sto- 40%, 54% and 65% when exposed to 0.1, 0.2, 0.4 and mata length and width for adaxial and abaxial surfaces of 0.6 ppm of pyriproxyfen respectively, compared to the con- leaf treated with an increasing dose showed significant dif- trol. ferences from the control group (P<0.05). This difference was due to a decrease in stomata size. For example, stomata length on upper surface of maize leaves treated with 0.1, 0.2, 0.4 and 0.6 ppm of pyriproxyfen declined by 14%, 26%, 48% and 73%, respectively, compared to the control. Similarly, observed effects of pyriproxyfen on stomata length on the lower surfaces were dependent on concentration of treatment. Moreover, stomata width on both upper and lower surfaces decreased with increasing dosage level of pyriproxyfen.

3.3 Physiological responses FIGURE 1 - Variation in proline content of maize leaves treated with Proline content showed a significant increase with in- different concentrations of pyriproxyfen. Different letters indicate creasing concentration of pyriproxyfen as compared to the statistically significant difference (one-way ANOVA, P < 0.05).

TABLE 1 - Effect of different concentration of pyriproxyfen on germination and seedling growth

Pyriproxyfen Germination Radicle length Coleoptile length Radicle number (ppm) Percentage (%) (mm) (mm)

0 100±0.0a 9.2±4.8a 11.4±1.4a 4.4±0.6a

0.1 98±0.0a 8.9±3.8a 10.0±2.0a 4.2±1.1a

0.2 92±1.8b 8.3±4.6a 9.7±2.1a 3.8±0.7a

0.4 84±2.0c 7.1±3.5ab 8.5±1.8a 3.6±0.8a

0.6 60±2.1d 5.5±2.0b 7.3±2.1b 3.5±0.7a *The values that are followed by the same letter do not differ statistically at a significance level at 0.05

TABLE 2 - The structure of adaxial and abaxial epidermis in leaves of maize plants treated with different concentrations of pyriproxyfen

Pyriproxyfen Stomata Epidermal cells (num- Stomata Stomata (ppm) (number mm2) ber mm2) width (µm) length (µm) Stomatal index

0.0 *10.5±0.4a 66.4±0.7a 11.6±0.5a 29.4±0.8a 13.72±0.5a

0.1 9.2±1.1b 63.8±0.8ab 10.4±0.6ab 25.5±0.4b 12.82±0.7ab Adaxial 0.2 8.2±0.5bc 61.3±0.2bc 8.6±0.4b 21.8±0.6bc 11.95±0.4bc

0.4 4.8±0.8c 45.2±0.6c 5.6±0.5c 15.3±0.5c 9.63±0.4c

0.6 1.3±0.8d 27.4±0.2d 2.2±0.7d 8.2±0.9d 4.52±0.1d

0.0 16.5±0.2d 68.5±0.6d 13.5±0.9c 32.8±0.3d 19.41±0.4a

0.1 15.4±0.8d 67.2±0.8cd 12.8±0.4c 26.6±0.5c 18.64±0.7ab Abaxial 0.2 13.1±0.8c 65.7±0.3ab 9.2±0.5b 23.8±1.2b 16.62±0.5abc

0.4 9.3±0.3b 58.8±0.9ab 5.8±0.8b 17.6±0.8a 13.65±0.4c

0.6 4.9±1.2a 51.5±0.4a 2.5±0.5a 12.2±0.8a 8.53±0.8d

*The values that are followed by the same letter do not differ statistically at a significance level at 0.05

280 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

TABLE 3 - Effects of different concentration of pyriproxyfen on leaf pigments contents

Pyriproxyfen Total Chlorophyll content Chlorophyll a Chlorophyll b Carotenoids Anthocyanin (ppm) (µmol m-1 fw) (µmol m-1 fw) (µmol m-1 fw) (µmol m-1 fw) (µmol m-1 fw) 0.0 *0.0078±0.0001a 0.0047±0.0009a 0.0031±0.0008a 0.0019±0.0001a 0.0003±0.0001d 0.1 0.0067±0.0002ab 0.0042±0.0007ab 0.0025±0.0004ab 0.0017±0.0002ab 0.0007±0.0007c 0.2 0.0052±0.0002b 0.0032±0.0002bc 0.0020±0.0001bc 0.0013±0.0001bc 0.0011±0.0001bc 0.4 0.0032±0.0004c 0.0021±0.0008c 0.0011±0.0006c 0.0008±0.0001c 0.0015±0.0005b 0.6 0.0017±0.0004d 0.0015±0.0002c 0.0002±0.0002d 0.0003±0.0006d 0.0023±0.0003a *The values that are followed by the same letter do not differ statistically at a significance level at 0.05

The effects of pyriproxyfen on photosynthetic pigment methrin, chlorpyriphos and pirimicarb) on tomato seed- contents in leaves of maize seedlings demonstrated signif- lings [31]. The obtained results indicate that plant growth icant differences with increasing concentration, compared is a sensitive variable for determining the toxic effects of with the control (Table 3). While chlorophyll a, chlorophyll pyriproxyfen in seed germination. The toxic effects of b, total chlorophyll, carotenoids contents declined with in- pyriproxyfen caused an inhibition in root growth of maize. creasing dosages, anthocyanin content increased all the The plants investigated were substantially damaged during treatment with respect to the control. The rise in anthocya- germination, they were sensitive to environmental condi- nin content was 58%, 73%, 80%, and 87% with 0.1, 0.2, tions most and it was significantly delayed after pyri- 0.4, and 0.6 ppm of pyriproxyfen, respectively, compared proxyfen application. The effects of pesticide on seed ger- with control. However, the lowest dose of pyriproxyfen in- mination inhibit the growth process of plant, depending on creased the contents of chlorophyll a, chlorophyll b and ca- the concentration of the treatment solution [32]. However, rotenoids in leaves of maize seedlings. Moreover, under the reduction in the growth and the seed germination rate pyriproxyfen stress, the amount of chlorophyll b content of may inhibit cell division, as reported in maize plants ex- leaves was more affected than chlorophyll a and carote- posed to pesticide stress [33]. On the other hand, radicle noids content, compared to the control. On the other hand, and coleoptile lengths also decreased significantly by the pyriproxyfen applied at the highest concentration signifi- addition of pyriproxyfen as stated by the reduction of radi- cantly decreased their content. The chlorophyll a, chloro- cle number. The reduction of the radicle length has already phyll b and carotenoid contents of maize seedlings de- been noted for the roots of soybean with different concen- creased approximately by 69%, 94% and 85%, respectively trations of some pesticides which caused serious damage at the highest pyriproxyfen dosage. The negative relation- to the growth of root [34]. ship between the pigments content and the pesticide concentration was statistically supported for chlorophyll a, b Stomata densities may vary in response to several en- and carotenoids contents (P<0.05). vironment stresses like drought, light intensity and atmospheric levels of carbon dioxide [35]. Therefore changes in the density and the morphology of the stomata on a leaf 4. DISCUSSION surface may be considered as a bioindicator of different environmental factors. Moreover, the interaction between There are only a few reports about the toxic effects of plant and pesticide concentrations occurs mainly via sto- pyriproxyfen, especially on rats and dogs [26, 27]. It causes mata changes. After the treatment with different concentra- toxic effects through stimulation of the free radicals in sen- tions of pyriproxyfen, we observed that stomata move- sitive plants, and also causing chromosomal abnormalities ments decreased in parallel with the increase in concentra- [28] and oxidative stress [29] in living cells. To our tions (Table 2). It was found that pyriproxyfen reduced sto- knowledge, there are no data about their phytotoxicologi- matal index of especially the upper and lower surfaces of cal effects on target plants, non-target plants and human leaves with increasing concentrations with respect to the health. The present study determined for the first time the control. The adaxial surfaces of maize leaves always had morphological and physiological impacts of pyriproxyfen fewer stomata than the abaxial surfaces. These results sup- on living organisms. In our study, pyriproxyfen was found port previous findings that plants express negative correla- to adversely affect seed germination and seedling growth tion between stomata index and pesticide concentration of maize. It is clear from our results that even low concen- [36]. The reduction in stomata density can be explained by trations of pyriproxyfen decreases seed germination and the cell division which decreased with increasing concen- seedling growth (Table 1). However, maize growth was trations of sulcotrione [37] and diclofop-methyl and lin- significantly inhibited by the increasing concentration of dane [38]. On the other hand, stomata size has shrunk with pyriproxyfen, which is in agreement with previous experi- increasing doses, compared to control. Stomata play vital ments with thiamethoxam and thiophanate-methyl on roles in CO2 diffusion pathway in leaves. The reduction in wheat seedlings [30] and three insecticides (alpha-cyper- the number and the size of stomata, which provide the dif-

281 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

fusion of atmospheric CO2 [35], will adversely affect pho- throughout the life cycle. Although the highest concentra- tosynthesis and respiration, determining plant growth and tions of pyriproxyfen are not generally found in the envi- development rate [39, 40]. This incident can be evaluated ronment, uncontrolled or accidental application of it may as an adaptation to stress conditions. suppress the growth and development of non-target beings. Therefore, it is essential to test the toxic effects of insecti- Proline levels can increase substantially in plant tis- cides on plants and other organisms before using them for sues under stress conditions such as drought, low tempera- agricultural purposes. In addition, insecticides may have tures and high salinity [41, 42]. In normal conditions, the toxic effects on farmers and humans who consume the content of proline in plants is low. The effects of pesticide plants exposed to insecticides. This information also en- stress on the proline content in maize leaves was investi- courages further studies to determine the impact of heavy gated and the results showed that the proline content in- pyriproxyfen contamination on soil and water. creased substantially in relation to the concentration of insecticide stress (Figure 1), and were significantly higher in The authors have declared no conflict of interest. the pyriproxyfen treatments than in the control. Proline is known to act as an osmolyte for osmotic adjustment, and affects the solubility of various proteins [43]. However, proline can protect plant cells from stress conditions by REFERENCES maintaining the integrity of cell membranes [44]. The present observations show a positive relationship between in- [1] Cabras, P. and Angioni, A. (2000) Pesticide residues in grapes, wine, and their processing products. Journal of Agricultural creasing insecticide dose and proline accumulation. This and Food Chemistry 48(4), 967-973. accumulation could be due to a protective role of blocking [2] Nwadinigwe, A.O. (2010) Effects of the insecticide, lambda- protein denaturation under pyriproxyfen stress. cyhalothrin on the growth, productivity and foliage anatomical characteristics of Vigna unguiculata (L) Walp. Bio-Research The contents of the photosynthetic pigments in the 8(1), 583-587. plants subjected to pesticide stress are known to decline [3, [3] Horbowicz, M., Sempruch, C., Kosson, R., Koczkodaj, D. and 45]. After the treatment with different concentrations of Walas, D. (2013) Effect of fluazifop-p-butyl treatment on pig- pyriproxyfen, we observed that photosynthetic pigment con- ments and polyamines level within tissues of non-target maize tents decreased in parallel with increasing insecticide con- plants. Pesticide Biochemistry and Physiology 107(1), 78-85. centrations (Table 3). This may be due to the destruction of [4] Aktar, Md.W., Sengupta, D. and Chowdhury, A. (2009) Im- the chloroplast structure of the maize leaves which causes pact of pesticides use in agriculture: their benefits and haz- reduction in the lamellar content of the light harvesting pig- ards. Interdisciplinary Toxicology 2(1), 1-12. ments. Photosynthetic pigments harvest the light energy and [5] Miliou, A., Petrus-Vancea, A., Timar, A.V., Borza, I.M. and convert them to chemical products, and they are of great im- Brata, A. (2009) Morpho-anatomical changes of plant vegeta- portance as indicators for determining photosynthesis rate. tive organs of olive (Olea europaea) tree culture treated with pesticides. Analele Universitatii din Oradea, Fascicula Biolo- On the other hand, the decrease in the pigment content led to gie 2, 85-90. adverse effects on PS II and photosynthetic electron [6] Mishra, V., Mishra, P., Srivastava, G. and Prasad, S.M. (2011) transport activities [46]. Maize seedlings treated with the Effects of dimethoate and UV-B irradiation on the response of pyriproxyfen underwent drastic changes in the photosyn- antioxidant defense systems in cowpea (Vigna unguiculata L.) thetic pigment content, thereby raising the probability of seedlings. Pesticide Biochemistry and Physiology 100(2), greater damage and ultimately resulting in the reduction of 118-123. growth. However, the increasing concentrations of pyri- [7] Wang, S., Li, H. and Lin, C. (2013) Physiological, biochemical proxyfen also increased accumulation of anthocyanin in and growth responses of Italian ryegrass to butachlor exposure. Pesticide Biochemistry and Physiology 106(1-2), 21-27. maize leaves. These substances accumulate under the influence of various environmental stresses [47]. The major func- [8] Swarnam, T.P. and Velmurugan, A. (2013). Pesticide residues in vegetable samples from the Andaman Islands, India. Envi- tion of anthocyanin seems to be its photoprotective role [48, ronmental Monitoring Assessment 185(7), 6119-6127. 49]. This function might explain the increasing anthocyanin [9] Sutherland, T.D., Home, I., Weir, K.M., Russell, R.J. and level under high doses of insecticides. Oakeshott, J.G. (2004). Toxicity and residues of endosulfan isomers. Reviews Environmental Contamination and Toxicol- ogy 183, 99-113. 5. CONCLUSION [10] Ishaaya, I. and Horowitz, A.R. (1995) Pyriproxyfen, a novel insect growth regulator for controlling whiteflies: mechanisms The stress-induced changes lead to a decline in all and resistance management. Pesticide Science 43(3), 227-232. growth parameters [50]. Based on the results presented, it [11] US-EPA. (1999) U.S. Environmental Protection Agency, Pol- is clearly indicated that the effects of pyriproxyfen on all icy on a common mechanism of action: the organophosphate growth parameters (germination, plant growth and devel- pesticides. Fed. Register, 64, pp. 5795–5799. opment) of maize plant depend on its dose, and they can be [12] Invest, J.F. and Lucas, J.R. (2008) Pyriproxyfen as a mosquito considered as a sensitive indicator for pyriproxyfen tox- larvicide. (Paper presented at the Sixth International Confer- ence on Urban Pests) (In: W.H. Robinson, D. Bajomi (Eds.), icity. Reduction in growth caused by accumulative effects Printed by OOK-Press Kft., H-8200 Veszprém, Pápai út 37/a, of increasing pyriproxyfen concentrations is manifested Hungary).

282 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

[13] Ohashi, K., Nakada, K., Ishiwatari, T., Miyaguchi, J., Shono, [30] Nicuta, D., Badaluta, N., Lazar, G. and Lazar, I.M. (2013) Y., Lucas, J.R. and Mito, N. (2012) Efficacy of pyriproxyfen- Changes in growth of wheat plantlets induced by the action of treated nets in sterilizing and shortening the longevity of Thiamethoxam and Thiophanate-Methyl on seeds. Environmen- Anopheles gambiae (Diptera: Culicidae). Journal of Medical tal Engineering and Management Journal 12(1), 85-96. Entomology 49(5), 1052-1058. [31] Chahid, K., Laglaoui, A., Zantar, S. and Ennabili, A. (2013) Ef- [14] Misra, S.G. and Mani, D. (1994) Adverse effects of Pesticides: fect of three insecticides on tomato (Solanum lycopersicum) Agricultural pollution II. (In: S.G. Misra and D. Mani (Eds.), seedling germination and early plants growth. Biological Diver- Ashish Publisher, New Delhi. sity and Conservation 6(1), 57-61. [15] Wilkinson, R.E. and Ohki, K. (1991) Effect of PCMBS and [32] Pereira, C.R., Monterroso, C. and Macías, F. (2010) Phytotoxi- EPTC on mineral nutrient uptake and translocation in city of hexachlorocyclohexane: Effect on germination and early wheat. Journal of Plant Nutrition 14(1), 93-108. growth of different plant species. Chemosphere 79(3), 326-333. [16] Amtmann, A., Troufflard, S. and Armengaud, P. (2008) The [33] Aksakal, O., Erturk, F.A., Sunar, S., Bozari, S. and Agar, G. effect of potassium nutrition on pest and disease resistance in (2013) Assessment of genotoxic effects of 2,4- dichlorophenox- plants. Physiologia Plantarum 133(4), 682-691. yacetic acid on maize by using RAPD analysis. Industrial Crops and Products 42, 552-557. [17] Wright, J.P. (1994) Use of membrane potential measurements to study mode of action of diclofop-methyl. Weed Science [34] Aksoy, O. and Deveci, A. (2012) The investigation of the cyto- 42(2), 285-292. toxic effects of some pesticides on soybean (Glycine max L.). Cytologia 77(4), 475-483. [18] Du, Q., Jia, X. and Yuan, B. (2006) Toxic effects of 1,2,4-tri- chlorobenzene on rice seed germination and seedling growth. [35] Driscoll, S.P., Prins, A., Olmos, E., Kunert, K.J. and Foyer, C.H. Chinese Journal of Applied Ecology 17(11), 2185-2188. (2006) Specification of adaxial and abaxial stomata, epidermal structure and photosynthesis to CO2 enrichment in maize leaves. [19] Mishra, V., Srivastava, G. and Prasad, S.M. (2009) Antioxi- Journal of Experimental Botany 57(2), 381-390. dant response of bitter gourd (Momordica charantia L.) seedlings to interactive effect of dimethoate and UV-B irradiation. [36] Tort, N., Ozturk, I. and Tosun, N. (2004) The effects of fun- Scientia Horticulturae 120(3), 373-378. gucides applications on anatomical structures and physiology of tomato (Lycopersicon esculentum Mill.). The Journal of Ege [20] Zhang, G.L., Chen, W.J., Qiu, L.M., Sun, G.R., Dai, Q.G. and University Faculty of Agriculture 41(2), 111-122. Zhang, H.C. (2009) Physiological response to 1,2,4-trichloro- benzene stress of different rice genotypes in seedlings. Acta [37] Sta, C., Ledoigt, G., Ferjani, E. and Goupil, P. (2012) Exposure Agronomica Sinica 35(4), 733-740. of Vicia faba to sulcotrione pesticide induced genotoxicity. Pes- ticide Biochemistry and Physiology 103(1), 9-14. [21] Perveen, K., Haseeb, A. and Shukla, P.K. (2010) Effect of Sclerotinia sclerotiorum on the disease development, growth, [38] Mesi (Dizdari), A. and Kopliku, D. (2013) Cytotoxic and geno- oil yield and biochemical changes in plants of Mentha arven- toxic potency screening of two pesticides on Allium cepa L. Pro- sis. Saudi Journal of Biological Sciences 17(4), 291-294. cedia Technology 8, 19-26. [22] Khan, M.A. and Ungar, I.A. (1984) The effect of salinity and [39] Sances, F.V., Ting, I.P., Hogenson, R.O. and McDonald, J.E. temperature on the germination of polymorphic seeds and (1982) Impact of certain pesticides on avocado. California Avo- growth of Atriplex triangularis Willd. American Journal of cado Society Yearbook 66, 145-154. Botany 71(4), 481-489. [40] Yakir, E., Hilman, D., Harir, Y. and Green, R.M. (2007) Regu- [23] Meidner, H. and Mansfield, T.A. (1968) Physiology of sto- lation of output from the plant circadian clock. FEBS Journal mata. New York: McGraw Hill, 178 p. 274(2), 335–345. [24] Bates, L.S., Waldren, R.P. and Teare, I.D. (1973) Rapid deter- [41] Kavi Kishor, P.B., Sangam, S., Amrutha, R.N., Sri Laxmi, P., mination of free proline for water-stress studies. Plant and Soil Naidu, K.R., Rao, K.R.S.S., Sreenath, Rao, Reddy, K.J., The- 39(1), 205-207. riappan, P. and Sreenivasulu, N. (2005). Regulation of proline biosynthesis, degradation, uptake and transport in higher plants: [25] Sims, D.A. and Gamon, J.A. (2002) Relationships between Its implications in plant growth and abiotic stress tolerance. Cur- leaf pigment content and spectral reflectance across a wide rent Science 88(3), 424-438. range of species, leaf structures and developmental stages. Re- mote Sensing of Environment 81(2-3), 337-354. [42] Ma T T, Christie P, Luo Y M, Teng Y. 2014. Physiological and Antioxidant Responses of Germinating Mung Bean Seedlings to [26] Choi, S.M. and Lee, B.M. (2004) An alternative mode of action Phthalate Esters in Soil. Pedosphere, 24(1), 107–115. of endocrine disrupting chemicals and chemoprevention. Jour- nal of Toxicology and Environmental Health Part B 7(6), 451- [43] Delauney, A.J. and Verma, D.P.S. (1993) Proline biosynthesis 463. and osmoregulation in plants. The Plant Journal 4(2), 215-223. [27] Franc, M., Genchi, C., Bouhsira, E., Warin, S., Kaltsatos, V., [44] Zhang, Y.X. and Li, X.L. (2003) Toxicity of three organophos- Baduel, L. and Genchi, M. (2012) Efficacy of dinotefuran, per- phorus pesticides on Hordeum vulgare seedling. Journal of mithin and pyriproxyfen combination spot-on against Aedes Agro-Environment Science 22, 754–757. aegypti mosquitoes on dogs. Veterinary Parasitology 189(2- [45] Chris, A., Luxmisha, G., Masih, J. and Abraham, G. (2011) 4), 333-337. Growth, photosynthetic pigments and antioxidant responses of [28] Sial, A.A. and Brunner, J.F. (2010) Lethal and sublethal ef- Azolla filiculoides to monocrotophos toxicity. Journal of Chem- fects of an insect growth regulator, pyriproxyfen, on oblique- ical and Pharmaceutical Research 3(3), 381-388. banded leafroller (Lepidopetra: Tortricidae). Journal of Eco- [46] Doganlar, Z.B. (2012) Physiological and genetic responses to nomic Entomology 103(2), 340-347. pesticide mixture treatment of Veronica beccabunga. Water, Air and Soil Pollution 223(9), 6201-6212. [29] Shimabukuro, R.H., Davis, D.G. and Hoffer, B.L. (2001) The effect of diclofop-methyl and its antagonist, vitamin E, on [47] Gould, K.S., Markham, K.R., Smith, R.H. and Goris, J.J. membrane lipids in oat (Avena sativa L.) and leafy spurge (Eu- (2000) Functional role of anthocyanins in the leaves of Quin- phorbia esula L.). Pesticide Biochemistry and Physiology tinia serrata A. Cunn. Journal of Experimental Botany 69(1), 13-26. 51(347), 1107-1115.

283 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

[48] Chalker-Scott, L. (1999) Environmental significance of anthocyanins in plant stress responses. Photochemistry and Photo- biology 70(1), 1-9.

[49] Steyn, W.J., Wand, S.J.E., Holcroft, D.M. and Jacobs, G. (2002) Anthocyanins in vegetative tissues: a proposed unified function in photoprotection. New Phytologist 155(3), 349– 361.

[50] Bashir, F., Mahmooduzzafar, Siddiqi, T.O. and Iqbal, M. (2007) The antioxidative response system in Glycine max (L.) Merr. exposed to Deltamethrin, a synthetic pyrethroid insecticide. Environmental Pollution 147(1), 94-100.

Received: April 03, 2014 Revised: May 26, 2014 Accepted: June 11, 2014

CORRESPONDING AUTHOR

Dr. Yasemin Coskun Suleyman Demirel University Faculty of Arts and Sciences Department of Biology 32260 Isparta TURKEY

Phone: +90-532-4061438 Fax: +90-246-2371106 E-mail:[email protected]

FEB/ Vol 24/ No 1b/ 2015 – pages 278 - 284

284 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

DUST RETENTION EFFECTS OF POPULUS ALBA VAR. PYRAMIDALIS (BUNGE) IN ARID OASIS CITIES, NORTHWEST CHINA

Aliya Baidourela1, Ümüt Halik1,2,*, Tayierjiang Aishan1,2, Abdulla Abliz1 and Aygvl Elyas1

1College of Resources & Environmental Science, Xinjiang University/ Key Laboratory of Oasis Ecology Xinjiang Uyghur Autonomous Region, Urumqi 830046, China; 2Faculty of Geography and Mathematics, Catholic University of Eichstaett-Ingolstadt, Eichstaett 85071 Germany

ABSTRACT PM10 and calculates Air Pollution Index (API) in accordance with the national standards [7]. It was shown that the The present study examined dust deposition on single mean daily concentrations of particulate matter (PM) with leaves of Populus alba var. pyramidalis (Bunge) at various the size ranging from 84 μgm−3 to 600 μgm−3 seem to be heights (1m, 2m, and 4m), analyzed dust particle size dis- closely associated with human health [8]. Dust events were tribution and its contents in arid oasis city Aksu, Northwest found to be a high risk factor for respiratory and cardiovas- China. The results showed that the amount of dust retained cular diseases in central China [9]. It was also shown that per leaf area of P. alba var. pyramidalis (Bunge) differed reduced visibility caused by increased Saharan dust input significantly between sites, as follows: CTA> IA> P> RA> was closely associated with an increased daily hospital ad- SFA; Leaves at a height of 1 meter retain the most dust mission rate in the Caribbean Island of Trinidad [10]. A compared to other two heights, as follows: 1m > 2m > 4m; more recent study showed that pediatric hospital admis- Dust particles larger than PM10 account for more than 90% sions remained at high rate for up to 7 days following a of all dust particles in all sites and the dust accumulated on peak atmospheric dust concentration [11]. tree leaves is mainly of local urban origin; The variation of To reduce the health risks for urban population, we heavy metal concentrations in different sites is significant. should control pollutant sources and reduce their effects. Vegetation, as the lung of the city, can absorb particulate

matter (PM) through adhering PM to leaf surfaces, thereby KEYWORDS: Oasis Cities, Dust retention effect, Heavy metal con- reducing the adverse effects of air pollution on people [12- centrations, Populus alba var. pyramidalis (Bunge) 14]. Trees are particularly effective in absorbing PM because of their larger leaf areas and the air turbulence cre-

ated by their vertical structure [12]. Particularly, the hairy 1. INTRODUCTION and sticky leaf surfaces of trees can filter dust pollutants by interception and absorption [7, 15-17]. However, the dust Air pollution poses a long-term threat to the health and retention of tree leaves is usually affected by rainfall, quality of life for urban residents. In particular, dust pollu- winds, and dust storms. Therefore, dust retention of plant tion in oasis cities located in arid areas is a contributing leaves is cyclical, being dependent on weather cycles (e.g., factor of reduced life expectancy. The adverse effects of rainfall, winds, and dust storms) [18, 19]. anthropogenic particulate matter (PM) on human health This study examined dust deposition on single leaves have been widely documented in the recent literature [1-4], of P. alba var. pyramidalis (Bunge) at various heights (1m, but so far very little attention has been given to the possible 2m, and 4m) to estimate the quantity of dust deposited on adverse effects on human health of naturally occurring par- an entire tree. The characteristics of the removed dust particulates (such as PM emanating from dust storms) that are ticles were analyzed to evaluate the effectiveness of P. alba the major atmospheric pollutants in many cities around the var. pyramidalis (Bunge) in removing atmospheric dust in world [5-6]. As one of the naturally occurring forms of at- urban areas. The study addressed the following questions: mospheric pollution, dust mainly exists in three forms: dust- How much dust can one unit leaf area of P.alba var. py- fall (diameter ≥ 10 μm), total suspended particles or TSP (di- ramidalis (Bunge) remove under natural conditions? How ameter ≤ 100μm) and repairable particles or PM10 (diame- does this vary with site and leaf height? Where does the ter ≤ 10 μm). China's Daily Air Quality Report monitors dust come from? We hope that the results can be used as references in future ecological design and optimized con-

figuration of urban tree species in arid oasis cities, North- * Corresponding author west China.

285 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

2. MATERIALS AND METHODS (i.e., No. 8 Xi Da Jie Road), as a representing Commercial Transportation Area (CTA); 2) Aksu TV station courtyard 2.1. Field conditions and plant material (No. 5 Jiao Yu Road), as a representing Residential Area Aksu (41°10′N, 80°15′E, 1115 m above sea level), one (RA); 3) Cement factory (No. 9 Nan Jiao Road), as a rep- of the typical oasis city in arid northwest China, is situated resenting Industrial Area (IA); 4) Dolan Park (Xi Da Jie in the northern edge of the Taklimakan Desert (Fig. 1). The Road), as a representing Park Area (PA); and 5) Kökyar mean annual temperature is 9.9–11.5°С and the mean an- Forest management station, as a representing Suburban nual precipitation is 43.9–65.3mm with a potential annual Forest Area (SFA). Sampling began in 04.05.2013 and fin- evaporation of over 2000 mm. Scarce precipitation and ished in 03.06.2013. At each site, 2 to 3 P. alba var. py- strong wind are the perfect mechanisms to incur dust ramidalis trees were selected and leaf samples were col- storms in this oasis city. We selected P. alba var. pyrami- lected from each tree at heights of 1m, 2m, and 4m. The dalis (Bunge), one of the most popular tree species planted fieldwork was conducted on sunny mornings with little in the city, as the object of this study. P. alba var. pyrami- wind, and metrological parameters were recorded by a dalis (Bunge) is a variant of P. alba and a native in arid portable meteor graph (M307592/NK4000). areas due to its drought tolerance. The front side of leave We collected healthy and lookalike leaf samples from is smooth and back side is hairy with shape being oval or trees and dust samples from non-vegetated areas near the elliptical [20]. selected tree with iron containers. The samples were stored in plastic bags (for leaf) and iron containers (for dust) at 2.2. Field sampling and laboratory analyses room temperature in a laboratory. Dust was washed from Five sites were selected in Aksu urban area and the sur- the leaf surfaces with distilled water, according to the re- rounding Kökyar shelterbelt forest area (Fig. 1) and each ported method [21]. Filters were weighed before and after site was plotted in relation to the various functional areas filtration by electronic scales with an accuracy of 1μg (industrial area, commercial transportation area, residential (PTX-FA-210, Shanghai). The dust-collecting potential of areas, parks, suburban shelter, and forest area) and also to each individual leaf was measured according to surface greening types. The five sites are: 1) Urban basins center area, using a laser leaf area meter (CI-203, USA). The ave-

FIGURE 1 - Sketch map of field measurement sites in Aksu City and the relative location of Aksu within the context of China (upper right inset) and the Xinjiang autonomy region (lower right inset)

286 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

rage weight of dust retained per unit leaf area was calcu- 3.2. Height differences lated as the ratio of dust removed from leaves over the total Table 1 and Figure 3 show that the dust retention is de- leaf area. The sediment washed from the leaves and the pendent on the sampling heights. In any case (RA, CTA, IA, dust samples collected from the non-vegetated sites were SFA, PA) the leaves at 1m height retain the most dust and analyzed by a laser particle size analyzer (Microtrac the leaves at 4m height retained the least dust with the 2m S3500, USA). height being the intermediate. For example, at the CTA site, the weight of dust per unit leaf area was 12.46 g/m2 at 1m height, 10.37 g/m2 at 2mheight and 8.6 g/m2 at 4m height. 3. RESULTS 3.3. Dust particle distribution 3.1. Site differences The particulate matter (i.e., dust) washed from the leaf The average dust concentration on the sampled leaves samples was classified into three categories based on their was heavily dependent on the sampling sites. As shown in sizes: 1) smaller than PM10 (<10 μm); 2) PM10 (=10 μm); Figure 2, the trees located in the commercial-transportation and 3) larger than PM10 (>10 μm). Unexpectedly, the per- area (CTA) retained the highest amount of dust per unit leaf centage of >10 μm was much higher than other two cate- area and the second highest was the industrial area (IA). gories (Fig. 4), reaching over 90% at all sites including the The park area (PA) and the residential area (RA) are more suburban forest area (SFA). But, the detailed curve of the or less similar in retaining the dust and the suburban forest particle size distribution demonstrates a distinguishable area had the lowest retention of dust. It is apparent that the feature among different sites. Specifically, suburban forest commercial-transportation area (CTA) is the most severely area (SFA) has the finest size with a peak spanning from dust-polluted area and the suburban forest area is the least 13 to 74 μm and the industrial area (IA) has coarsest size dust-polluted area. with a peak spanning from 31 to >1000 μm (Fig. 4). The

8.0 40 ） 2 7.0 35 g/m ）（

6.0 30 2

5.0 25 g/m （

4.0 20

3.0 15

2.0 10 Dust Loading

1.0 5 Foliar dust -retention amount -retention dust Foliar 0.0 0 4th Day 8th Day 12th Day 16th Day 20th Day

DL IA CTA RA SFA PA

FIGURE 2 - Average dust weight retained per unit leaf area at different sites. DL = Dust Loading; IA = industrial area; CTA = commercial- transportation area; RA = residential area; SFA = Suburban forest area; PA = park area.

TABLE 1 - Average weight of dust retained per unit leaf area at different heights and sites

1m 2m 4m Residential areas (RA) 8.14±0.25a 5.27±0.22b 4.14±0.14c Commercial transportation area (CTA) 12.46±0.66a 10.37±0.57b 8.6±0.60c Industrial area (IA) 13.48±0.31a 9.74±0.49b 1.46±0.29c Suburban forest area (SFA) 4.99±0.16a 3.21±0.18b 0.78±0.08c Park area (PA) 5.47±0.24a 4.35±0.19b 1.27±0.45c Letters a–c indicate significant differences between three different heights, P<0.05 Differences are statistically significant between different heights.

287 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

SFA Smaller than PM10 IA PM10 CTA Larger than PM10

0246810(g)

FIGURE 3 - Dust particle size distribution at different sites

12.0

10.0 RA 8.0 PA

6.0 SFA

4.0 IA Volume fraction (%)

2.0 CTA

0.0 0.972 1.635 2.750 4.62 7.78 13.08 22.00 37.00 62.22 104.6 176.0 296.0 497.8 837.1 1408

Particle size (µm) FIGURE 4 - Frequency and cumulative probability curve of dust particle size distribution at different sites

TABLE 2 - Comparison of heavy metal contents of foliar dust in different size (mg /kg)

Sites / Heavy metals Cu Pb Zn Cd No. Samples 20 19 20 18 Industrial area (IA) 9 ± 1.3 43.2 ± 5.2 161.2 ± 12.3 1.941 ± 0.2 Commercial transport area (CTA) 11.5 ± 1.8 48.4 ± 4.6 236.0 ± 22.1 1.920 ± 0.3 Residential area (RA) 6.5 ± 0.9 43.2 ± 3.9 111.5 ± 11.3 1.670 ± 0.1 Shelterbelt forest area (SFA) 4.5 ± 0.6 35.9 ± 4.5 23.6 ± 2.4 1.655 ± 0.1 Park area (PA) 5.6 ± 0.8 45.8 ± 5.3 11.2 ± 1.6 1.780 ± 0.2 Soil Background Value* 23.6 25.4 102.6 0.097 * China Environmental Monitoring Station, 1990 [28].

rest three (i.e., RA, PA. and CTA) have similar peaks span- of all dust measured, respectively. The distribution curves ning from 13 to 42 μm. for the CTA, RA, and PA sites are bimodal, with the first mode of 74~150μm, accounting for 4~7%; and the second At the CTA, IA, RA, PA, and SFA sites, the curve rep- mode of 209~352μm accounting for 6~10% of the content. resenting the dust particle size distribution peaked at The SFA site has only one mode, in the range 22~37μm, 248.9μm, 296.0μm, 296.2μm, 296.0μm, and 31.11μm, ac- representing 9~11% of the total dust volume. The two counting for 10.61%, 6.73%, 7.66%, 8.57%, and 11.05% modes of the distribution curve for the IA site are greater

288 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

than the corresponding values for CTA, RA, and PA: the the framework of the Sino-German joint research project- first at 104~296μm, 3%~6%; the second at 591~995μm, SuMaRiO (Grant-No.: 01LL0918C). 6~7%. The authors have declared no conflict of interest. 3.4. Element distribution We have also analyzed the heavy metals (including Cu, Pb, Zn, Cd) and the results are listed in Table 2. It can be REFERENCES seen that Pb and Cd concentrations at all functional areas (i.e., sampling sites) are larger than soil background values [1] Bousquet, J., Ndiaye, M., Aït-Khaled, N., Annesi-M. I. and Vi- gnola, A.M.(2003) Management of chronic respiratory and al- and Cu concentration is smaller than soil background value. lergic diseases in developing countries. Focus on sub-Saharan It is also notable that the concentrations of Cu, Zn, Pb are the Africa, Allergy 58, 265–283. highest in the commercial-transportation area (CTA) and the [2] Urch,B., Silverman, F., Corey, P., Brook, J.R., Lukic, K.Z., Ra- lowest in the suburban forest area (SFA). jagopalan, S. and Brook, R.D. (2005) Acute blood pressure responses in healthy adults during controlled air pollution expo- sures. Environ. Health Prospect 113 (8), 1052-1055. 4. DISCUSSION AND CONCLUSIONS [3] Romieu, I., Samet, J.M., Smith, K.R. and Bruce, N. (2002) Outdoor air pollution and acute respiratory infections among It is apparent that the commercial-transportation area children in developing countries. Journal of Environ Med 44, (CTA) is most severely dust-polluted area and the subur- 640-649. ban forest area is the least dust-polluted area. The greatest [4] Annette, C., Rohr, R.E. and Wyzga. (2012) Attributing health amount of dust in CTA means that a greater amount of at- effects to individual particulate matter constituents. Atmos- mospheric pollutants was emitted in the commercial-trans- pheric Environment 62, 130-152. portation area where more in-situ dust was locally gener- [5] Langner, M. (2006) Exponierter innerstaedtischer Spitzahorn ated. The smallest amount of dust in SFA means that the (Acer platanoides) - eine effiziente Senke fuer PM10? In: Karls- dust generated from the urban settings did not significantly ruher Schriften zur Geographie und Geooekologie, Band 21. affect the suburban area and also lends a support that the [6] Langner, M., Kull M. and Endlicher, W.R. (2011) Determina- dust was not from large-scale atmospheric processes. Our tion of PM10 deposition based on antimony flux to selected results are in agreement with reports showing a positive urban surfaces. Environmental Pollution 159, 2028-2034. correlation between traffic flow and dust load [23, 24]. Our [7] Kang, X.F., Zhang, J.H. and Liu, H.H. (2002) Influences of results also show that in any case (RA, CTA, IA, SFA, PA) Sandy Weather on City`s Air Quality in China: Scope And the leaves at 1m height retain the most dust and the leaves Scale. Resources Science 24(4), 1-4. at 4m height retained the least dust and also lent a support [8] Hwang, S., Jang, Y. and Kim, H. (2008) The Asian Dust Events to the report by Liu [19] who found a negative correlation and Emergency Department Visits with Atopic Asthma in between dust load and tree height. Seoul, Korea. Epidemiology 19, 360-360. [9] Cheng, M.F., Hu, S.C., Chiu, H.F., Wu, H.F., Chen, P.S. and Being similar with the dust load differences among the Yang, C.Y. (2008). Consequences of exposure to Asian dust different functional areas (i.e., sampling sites), the concen- storm events on daily pneumonia hospital admissions in Tai- trations of Cu, Zn, Pb are the highest in the commercial- pei, Taiwan. J Toxicol Environ Health A 71, 1295-1299. transportation area (CTA) and the lowest in the suburban [10] Gyan, K., Henry, W., Lacaille, S., Laloo, A., Lamsee, E.C., forest area (SFA). McKay, S., Antoine, R.M. and Monteil, M.A. (2005) African dust clouds are associated with increased pediatric asthma ac- Being different from the dust load differences among cident and emergency admissions on the Caribbean island of the different functional areas (i.e., sampling sites), a de- Trinidad. Int J Biometeorol 49, 371-376. tailed curve of the particle size distribution demonstrates [11] Florence, L., Pierre, O., Seydou, D. and Sabine, H. (2013) De- that suburban forest area (SFA) has the finest size and the sert dust impacts on human health: an alarming worldwide re- industrial area (IA) has coarsest size. It implies that the in- ality and a need for studies in West Africa. International J Bio- meteorology 57, 1–19. dustrial activities in Aksu City locally generate more coarse dust than other activities including the commercial [12] Beckett, K. P., Freer, S., P. H. and Taylor, G. (2000) Effective and transportation. tree species for local air- quality management. J. Arboricult 26(1), 12-19.

[13] Ottelé, M., Bohemen, H.D. and Fraij, A.L.A. (2010) Quantify- ing the deposition of particulate matter on climber vegetation on living walls. Ecol Eng 36(2), 154-62. ACKNOWLEDGMENTS [14] Prajapati, N. and Tripathi, B. (2008) Seasonal Variation of Leaf Dust Accumulation and Pigment Content in Plant Species We thank Kahar Zayim, Akram Ubul for useful dis- Exposed to Urban Particulates Pollution. Journal of Environ- cussions and technical supports on the manuscript. This mental Quality 37, 865-870. study was funded by National Natural Science Foundation [15] Shi, G.Y. and Zhao, S.W. (2003) Several Scientific Issues of of China (Grant-No.: 31270742), and by the German Fed- Studies on the Dust Storms. Chinese Journal of Atmospheric eral Ministry of Education and Research (BMBF) within Sciences 27(4), 591-602(In Chinese).

289 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

[16] Halik, U. (2003) Urban greening in arid regions: the example of the oasis in Southern Xinjiang/China, with special consideration of ecological, socio-economical, and cultural aspects, Diss., TU Berlin.

[17] Thorsteinsson, T., Gísladóttir, G., Bullard, J. and Mctainsh, G. (2011) Dust storm contributions to airborne particulate matter in Reykjavík. Iceland Atmospheric Environment 45, 5924- 5933. [18] Ghafghazia, S., Sowlatia, T. and Sokhansanj S. (2011) Partic- ulate matter emissions from combustion of wood in district heating applications. Renewable and Sustainable Energy Re- views 15, 3019- 3028. [19] Liu, C.H. and Cen, K. (2007) Particle size characteristics and possible sources of street dust in Beijing. Acta Scientiae Cir- cumstantiate 27 (6), 1006 -1012. [20] Han, R.S.H. (2006) The research on the photosynthesis and water eco-physiological characteristics of platycladus oriental- ist, Juniperus rigida and Populus alba var. Forest College, In- ner Mongolia Agricultural University, Huhhot.

[21] Chai, Y.X., Zhu, N. and Han, H.X. (2002) Dust removal effect of urban tree species in Harbin. Chinese Journal of applied ecology 13(9), 1121- 1126(in Chinese).

[22] Prusty, B.A.K., Mishra, P.C. and Azeez, P.A.(2005) Dust accumulation and leaf pigment content in vegetation near the national highway at Sambalpur, Orissa, India．Ecotoxicology and Environmental Safety 60(2), 228-235.

[23] Khaiwal, R., Ranjeet, S. and Rene, V. G. (2008) Atmospheric polycyclic aromatic Hydrocarbons: Source attribution，emission factors and regulation. Atmospheric Environment 42(13), 2895-2921. [24] Zhu, L. X. (2003) Significance and idea for urban dust geo- chemical survey. Acta Geologica Sichuan 23 (3), 174-175( in Chinese).

Received: April 11, 2014 Revised: July 17, 2014 Accepted: July 23, 2014

CORRESPONDING AUTHOR

Ümüt Halik

Xinjiang University

Key Laboratory of Oasis Ecology

14 Shengli Road

Urumqi 830046

P.R. CHINA

Phone: +86-991-8586533

Fax: +86-991-8582318

E-mail: [email protected]

FEB/ Vol 24/ No 1b/ 2015 – pages 285 - 290

290 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

ADSORPTION OF ESTRONE BY SELECTED CONSTRUCTED WETLAND SUBSTRATES

Xiao-Li Yang1, Lu Liang2, Hai-Liang Song2,*, Yun-Chen Gao2, Shi-Bei Huang2 and Bao-Lin Dai2

1School of Civil Engineering, Southeast University, Nanjing 210096, China 2School of Energy and Environment, Southeast University, Nanjing 210096, China

ABSTRACT while others have reported that 17β-estradiol (E2) and estrone (E1) caused roaches to be intersexual and promoted The adverse health effects on the endocrine function of vitellogenin induction in rainbow trout [2], and PCBs endocrine-disrupting chemicals (EDCs) have attracted caused harpacticoid copepods near sewage discharge to be worldwide attention. The present study reports the adsorp- in abnormal sex ratios in the Firth of Forth [3]. tion of EDCs from the perspective of choosing constructed Most EDCs are man-made chemicals, including indus- wetland substrates to help achieve a high removal effi- trial chemicals, crop protection products, antifouling com- ciency of EDCs. Batch studies were performed for the con- pounds, personal care products and pharmaceuticals [4]. In tact time, initial concentration, temperature and particle addition, EDCs can be naturally produced in the environ- size of the substrates. The results demonstrated that the ad- ment. For example, E1, a ubiquitous estrogen in the aquatic sorption capacity (qe) increased with an increase in the tem- environment, is not only a natural female hormone but is perature from 278 K to 308 K and a decrease in the sub- also excreted by women. E1 is one of the most potent ster- strate diameter from 75 to 590 μm. Along with the increase oid estrogens that can cause adverse effects on fish near the of initial concentration of estrone (E1) (from 100 μg/L to sewage effluent. 2000 μg/L), qe increased, while the removal efficiency decreased. The equilibrium adsorption data were analyzed The removal of EDCs, such as estrone, has been widely using two isotherm models (Langmuir and Freundlich) and studied. Many processes have been studied to remove these three kinetic models (pseudo-first-order, pseudo-second- estrogenic compounds at trace concentrations in water, order and intra-particle diffusion models). The Freundlich such as advanced oxidation [5-7], biological process [8, 9], model provided a better fit to the data, and the pseudo-sec- membrane retention [10], and adsorption [11-14]. Adsorp- ond-order kinetic model may be the best description of the tion is considered to be one of the preferred choices for the adsorption process. The FTIR spectra of three substrates removal of trace concentrations of contaminants in water before and after adsorption of estrone revealed that the ad- streams at high efficiency generally. Constructed wetland sorption was physical. Based on these results, it can be con- (CW) is widely used for the removal of endocrine disrup- cluded that ceramsite is an effective constructed wetland tors [15-17] as an inexpensive environmentally friendly substrate for use in the removal of estrone. water treatment process. The mechanisms of CW include the accumulation of plants and microorganisms [18], the degradation of the roots of plants and microorganisms and also their synergistic effects [19], some light degradation KEYWORDS: [20], in addition to the adsorption of the substrates. The Estrone; Adsorption; Constructed wetland substrates. function of the substrates in phosphorus removal is quite widespread [21-23]; nevertheless, research studies on estrogens sorbed by CW substrates have been very limited. 1. INTRODUCTION The present study, to the best of our knowledge, is the first of its kind to focus on the adsorption of EDCs from There is an increasing concern over the endocrine- the point of choosing the CW substrates to help achieve a modulating characteristics of some chemicals (EDCs). The high removal efficiency of EDCs. The commonly used CW adverse health effects of EDCs on endocrine receptors, or substrates, including ceramsite, zeolite and slag, were ap- its progeny, especially on endocrine function, have at- plied to the removal of EDCs (setting E1 as an example) tracted worldwide attention. Colborn observed that various from aqueous solution as a function of contact time, initial disturbing consequences were related to the alteration and estrone concentration, temperature and particle size. E1 disturbance of endocrine functions in humans and wildlife [1], was selected as the target compound, due to its widespread presence in wastewater effluents as a result of its relatively * Corresponding author low removal rate in sewage treatment plants and because it

291 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

is a major and relatively persistent biotransformation prod- 2.3. Batch study uct of E2. The equilibrium of this study was modeled using The batch adsorption experiments were conducted in Langmuir and Freundlich isotherm models, with the ki- a water-bathing orbital shaker, experimental variables, netic studies matched using pseudo-first-order, pseudo- such as contact time (0.25, 0.5, 1, 6, 12, 24, 36, and 48 h), second-order and intra-particle diffusion kinetic models. initial estrone concentration (100, 200, 400, 1000, 1500, and 2000 μg/L), temperature (278, 288, 298 and 308K), and particle size (#30, #60, #100) were used for the adsorp- 2. MATERIALS AND METHODS tion of E1 onto constructed wetland substrates. The adsorption experiments were performed by mixing 50 mL of so- 2.1. Preparation and characterization of the adsorbents lution with 2 g of substrate powder at natural pH in flasks The constructed wetland substrates, ceramsite, zeolite and shaken at 150 rpm in the shaker for a specified period and steel slag were chosen in present study for absorbing of time interval (equilibrium contact time). As a control, the estrone. The wetland padding was obtained from the estrone solution was shaken during the experimental period Yusong water supply and the drainage equipment factory in the absence of the constructed wetland substrates. After in Gongyi, Henan province, People’s Republic of China, agitation, the substrate-E1 mixture solution was filtered and then washed using tap-water and deionized water sev- through 0.45-μm Whatman glass fiber papers. The filtrates eral times to remove all of the loose materials from the wet- were analyzed for the final estrone concentration by using land packing surface. The substrates were then air-dried at an HPLC with a UV detector fitted with an Agilent Extend- a temperature of approximately 298 K for 2-3 days. After C18 reverse-phase column (4.6 μm × 250 μm × 5 μm), and drying, the wetland padding was pulverized. Uneven diam- 20 μL of the water was analyzed by HPLC at a wavelength eters of the wetland substrates were sieved by different di- of 200 nm. The mobile phase contained deionized water mensions of meshes to obtain particles of the diameter and acetonitrile with the 55/45 (v/v) ratio at the first 5 min, from 75 to 590 μm and the powder was later used in the and it was adjusted to 50/50 (v/v) from 5 min to 10 min, estrone removal experiment. In the present study, #30 rep- and at this time the flow rate was 0.8 mL min-1 and the col- resents the substrate powder with particle diameters be- umn temperature was 30. The method has a limit of tween 250 μm and 590 μm, #60 represents particle diame- quantification of 10 μg/L under these conditions. All of the ters between 149 μm and 250μm, and #100 represents par- tests were repeatedly performed three times in this study. ticle diameters between 75 μm and 149μm, which are all The amount of E1 adsorbed per unit substrate (q, according to the ASTM standard [24]. μg/kg) and the removal efficiency (R,%) of E1 were calcu- The physical properties of three substrates (#30) were lated from Eqs. (1) and (2), respectively: measured by using automatic specific surface area and po- rosity analyzer and the results were shown in Table 1. C0 Ct q  V (1) TABLE 1 - The physical properties of the substrates. m

BET Pore Volume Pore Radius Substrates (#30) C0  Ct （m2/g）（cm3/g）（nm） R  100% (2) Ceramsite 2.1 9.2×10-4 11.8 C0 Zeolite 0.7 3.9×10-4 8.7 where C0 is the E1 initial concentration (μg/L), Ct is the Slag 0.1 1.6×10-4 2.3 concentration at time t (μg/L), V is the volume of the E1 solution of 50 mL, and m is the weight of adsorbent of 2 g. 2.2. Stock solution 2.4. Kinetic studies Estrone was obtained from Sigma-Aldrich located in In the kinetic study, 50 mL of 1000 μg/L E1 solution St. Louis, MO, USA. The main physical-chemical proper- was mixed with 2 g of the #30 wetland substrate with rotary ties were listed in Table 2 below. The stock solution of agitation at 298 K natural pH surroundings. The aqueous E1(100mg/L) was prepared in an HPLC grade acetonitrile samples were withdrawn at a certain time from 15 min to because it has limited solubility in water. When it was nec- 48 h and the E1 concentration was identically measured as essary, the stock solution could be diluted to different con- previously described. centrations with distilled water. All solutions contained -1 0.01 M CaCl2 background electrolyte and 100 mg L NaN3 2.5. FTIR spectroscopy analysis acting as a biocide to prevent microbial degradation. Three substrates before and after adsorption of estrone TABLE 2 - The main physical-chemical properties of estrone were characterized using FTIR spectroscopy to determine the adsorption mechanism. The FTIR spectrophotometer Molecular Solubility Octanol-Water Properties was made by Thermo Nicolet Company and the model was Formula in water Partition Coefficient IR200. The dried substrates (before and after adsorption) E1 C18H22O2 0.03g/L 3.43 prepared as KBr discs were analyzed in the wavenumber

292 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

range from 4000 to 400 cm-1 using a Fourier transform in- curred after the external surface adsorption occurred [26]. frared (FTIR) spectrophotometer. The equilibrium was achieved (approximately 24 h), and the adsorption system remained nearly constant until no further considerable adsorption was noted. The maximum 3. RESULTS AND DISCUSSION adsorption amount was 21494μg/kg, 971μg/kg and 527μg/kg for ceramsite, zeolite and slag, respectively. Four 3.1. The effect of contact time steps that occurred in the adsorption process could be seen The effect of contact time on the adsorption of estrone in Fig. 1. An initial instantaneous adsorption step, which (E1, 1000 μg/L) on the three different constructed wetland lasted for 15 min, followed by the desorption step, which substrates considered in this study are presented in Fig. 1. lasted for approximately 45 min, and then a rapid adsorp- It can be observed from Fig. 1 that ceramsite exhibited the tion step, which lasted for approximately 6 h, followed by maximum adsorption capacity, followed by zeolite and the slower fourth phase, which continued until the equilib- slag in turn. At the beginning of the adsorption, the system rium was reached within 24 h. A further increase in the showed an instantaneous adsorption, it was due to the high contact time did not exhibit any increase in adsorption. The concentration disparity between the solution and the sur- adsorption rate decreased in the third stage might be due to face of the substrates and the high mass transfer power. In the intra-particle diffusion dominating over adsorption the first 15 min, the adsorption amount of three substrates [27], while the surface pores were occupied in the previous were relatively high (approximately 60%-80% of the max- stages. The rate of percent estrone removal was higher in imum adsorption amount), while between 15 and 60 min, the beginning, due to a larger surface area of the adsorbents the adsorption amount were decreased, and the adsorbed being available for the adsorption of E1. E1 desorbed back into the aqueous solution. The adsorption amount decreased to 17372μg/kg for ceramsite, 3.2. The effect of the initial estrone concentration 359μg/kg for zeolite and 312μg/kg for slag. The possible The effect of the initial E1 concentration on the adsorp- reason for the decreasing might be that at the beginning of tion amount and the removal efficiency is presented in Fig. 2. adsorption, the adsorbing force between the adsorbed E1 The adsorption amount of E1 on the substrate increased by and the substrate was weak, but the rotation speed was rel- increasing the initial estrone concentration in the range of atively high, which resulted in the bound E1 returning back 100 to 2000 μg/L, but the removal efficiency was de- into the aqueous solution. After 60 min, the amount of each creased. The adsorption capacity increased from 2391 to substrate adsorption increased rapidly, and then the adsorp- 38178 μg/kg, while the removal efficiency decreased from tion rate became slower, to the point where intra-particle 95.6% to 76.4% for ceramsite adsorbing E1 by increasing diffusion was the rate limiting step [25]. This stage oc- the E1 concentration from 100 μg/L to 2000 μg/L. With

25000 g/kg) μ 20000

ceramsite zeolite 15000 slag 1500

1000

500

Adsorbed amount of E1 on adsorbent ( adsorbent on E1 of amount Adsorbed 0 0 5 10 15 20 25 30 35 40 45 50 55 Time（） h

FIGURE 1 - The relationship between the E1 adsorption amount and the time. 2 g of #30 substrates, 1000 μg/L E1 and the temperature of 298K were used in this experiment. All of the tests were repeated three times. Experimental errors < 5%.

293 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

100 40000

90 g/kg) μ 30000

80 ceramsite zeolite 70 slag 20000 10

10000 Removal efficiency (%) Removal efficiency Removal efficiency (%) Removal efficiency 5

0 0 ( on adsorbent E1 of amount Adsorbed 0 250 500 750 1000 1250 1500 1750 2000 2250 Initical Conc （）μg/kg

FIGURE 2 - Effect of initial concentration on E1 adsorption. The dashed line stands for removal efficiency of E1, and the real line for the amount of E1 adsorbed per unit of substrate.2 g of #30 substrates and the temperature of 298K were used in this experiment. All of the tests were repeated three times. Experimental errors < 5%.

100

ceramsite zeolite 60 slag

4 Removal efficiency (%) 2

0 275 280 285 290 295 300 305 310 Temperature (K)

FIGURE 3 - Effect of temperature on E1 adsorption. 2 g of #30 substrates and the concentration of 1000 μg/L were used in this experiment. All of the tests were repeated three times. Experimental errors < 5%.

294 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

regard to the other two substrates, zeolite and slag, a similar to 308K, the E1 removal efficiency for ceramsite also in- trend was observed, while the adsorption capacities and the creased from 64.0% to 89.7%, and the removal efficiency removal efficiencies were much lower than that of cerams- for zeolite and slag increased by 1.7% and 1.0%. It can be ite. This result was due to the higher initial concentration concluded that the temperature can affect the E1 adsorption of E1, the larger number of moles available to the surface on ceramsite significantly. As the temperature goes up, the area, i.e., the adsorption characteristics indicated that the energy of the system increased, and the ratio of the active surface saturation was dependent on initial concentration site on the surface of adsorbent became higher, which led [28]. The initial concentration provided the driving force to to an increased percentage of effective collision between overcome the resistance of the aqueous and solid phases. E1 and active site, so the removal efficiency of E1 in- At lower E1 concentration, the ratio of the surface active creased. When the temperature grows above 298K, the sites of substrates might be high and the E1 molecules were growth increased not remarkable and the growth rate easily restrained by the adsorbent and removed from the slowed down. It was probably because all of the active sites solution. had sufficient energy to satisfy the activation energy for the adsorption and the increasing temperature could only in- The ceramsite exhibited the highest efficiency for the crease the adsorption rate. Due to the above observations, removal of E1, and we supposed that it was concerned with it was believed that E1 adsorption onto the substrates was the physical properties. From Table 1, the physical proper- an endothermic process. What’s more, further thermody- ties, including the BET, which is an important property for namic analysis or molecular level experiments are required adsorption, the pore volume and the pore radius, indicated to reveal that the adsorption process is either physical or that ceramsite had the maximum BET, which was 3 times chemical adsorption. the value of zeolite BET, and 16 times the value of slag BET. And the pore volume and the pore radius followed 3.4. The effect of particle size the order of ceramsite > zeolite > slag. A positive correla- The effect of particle size upon estrone adsorption on tion was found between adsorption capacity and the phys- the constructed wetland substrates was investigated in the ical properties of the substrates. range from 75 to 590 μm. The research results of ceramsite, zeolite and slag are presented in Fig. 4(a, b and c), which 3.3. The effect of temperature indicated that the E1 adsorption amount of three substrates The effect of temperature upon E1 adsorption of the were all increased as the particle size decreased, this in- three constructed wetland substrates was investigated in crease might be due to the increase of the surface area when this study. As noted above, four temperatures were used. the particle size decreased, and the amount of active sites Fig. 3 showed that as the temperature increased from 278K on the surface of substrates increased. Within the range of

a 105 50000 g/kg) 100 μ 40000

（） 30000 90 #30 #60 20000 85 #100

80 Removal efficiency % 10000

75 Adsorbed amount of E1 on adsorbent ( adsorbent on E1 of amount Adsorbed 0 0 250 500 750 1000 1250 1500 1750 2000 2250 Initical Conc（）μg/L

295 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

b 40 10000

35 #30 g/kg)

#60 8000 μ 30 #100

25 （） 6000

20 4000 15

Removal efficiencyRemoval % 2000 5

0 0 ( adsorbent on E1 of amount Adsorbed 0 250 500 750 1000 1250 1500 1750 2000 2250 Initical Conc（）μg/L

c 18 8000

16 7000 g/kg) μ 14 6000

12 #30 5000 （） #60 10 #100 4000

8 3000

6 2000 Removal efficiency % 4 1000 2 0 Adsorbed amount of E1 onAdsorbed amount E1of adsorbent ( 0 500 1000 1500 2000 Initical Conc（）μg/L

FIGURE 4 - The effect of particle size, ceramsite (a), zeolite (b) and slag (c) in different particle sizes (75~590 μm) were studied in the adsorption tests, 2 g substrates and the concentration of 100-2000 μg/L were used in 298 K. All of the tests were repeated three times. Experimental errors < 5%.

the experiment, the E1 adsorption onto ceramsite was amount of E1 in the smaller particle size. The possible rea- found to be the highest among the three substrates, which son might be that the amount of adsorption active site of indicated more affinity of ceramsite for estrone compared larger particles was smaller, and the adsorption of the sub- to zeolite and slag (Fig. 4a-4c). Fig. 4 also reveals that the strates was saturated at a relatively high concentration and concentration had a noticeable effect on the adsorption could not adsorb additional estrone.

296 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

3.5. Adsorption isotherms the Langmuir model can not fit the adsorption at all. Hater Adsorption isotherms are an important method to un- [33] pointed out that the linear form of Langmuir model derstand the mechanism of an adsorption system, and they could lead a big error, and this can also be found in this provide valuable information for the selection of adsor- study. bents. Experimental data of different concentrations of estrone adsorbed onto the substrates were analyzed by regres- 3.6. Kinetic study sion analysis to fit two commonly used isotherms: Lang- The rate of the adsorption process is probably the most muir [29] and Freundlich isotherms [30]. The adsorption important parameter in an adsorption system, and the reac- isotherms are characterized by definite parameters, whose tion kinetics controls the reaction dimension and the resi- values represent the surface properties and affinity of the dence time [34]. Consequently, it is necessary to study the adsorbents for E1. The linear form of the Langmuir equa- reaction kinetics in the designed system. The experimental tion and the log form of the Freundlich isotherm are pre- results presented in Fig. 1 were fitted to pseudo-first-order sented in Eqs. (3) and (4), respectively. [Eq. (5)] [35, 36], pseudo-second-order [Eq. (6)] [37] and C 1 C intra-particle diffusion models [Eq. (7)] [38, 39] by apply- e   e (3) ing a linear regression analysis using the Origin 8.5 pro- qe qmaxb qmax gram for Windows. The kinetic models can be expressed as follows: logCe logq   logK (4) ln(q  q )  lnq  k t (5) e n e t e 1 where q (mg/kg) denotes the theoretical monolayer t 1 t max   (6) saturation capacity, which indicates the maximum adsorp- q k q2 q tion amount, b (L/mg) denotes a constant related to the en- t 2 e e ergy of adsorption, qe (mg/kg) denotes the equilibrium up- 1/ 2 qt  k3t  C (7) take of E1, Ce denotes the estrone concentration remaining in solution (mg/L), and K (mg/kg) and n are constants re- where qe and qt are the amounts of estrone adsorbed lated to the adsorption capacity and the adsorption inten- (μg/kg) at equilibrium and at time t (h), respectively, k1 is sity, respectively [31]. the adsorption rate constant (h-1) of the pseudo-first-order -1 -1 The Freundlich and Langmuir isotherm parameters kinetic model, k2 is the rate constant (kg μg h ) of the for the adsorption system are presented in Table 3. The pseudo-second-order model, and k3 is the rate constant (μg higher regression coefficients suggested that the Freun- kg-1 h-1/2) of the intra-particle diffusion model. The plot of dlich model fits the data better than the Langmuir model. log(qe-qt) versus t [Eq. (5)] should exhibit a linear relation- Similar behavior of EE2 adsorption on different size of soil ship, whose slope and intercept of the plot determine k1 and was previously reported [32]. K was related to the bound- qe, respectively. The plot of (t/qt) and t [Eq. (6)] should ex- ing strength between E1 and substrates and the adsorption hibit a linear relationship, from which qe and k2 can be de- capacity. From Table 3, it was obvious that the turn of K termined by the slope and intercept of the plot, respec- 1/2 value was ceramsite>zeolite>slag, which reflected that the tively. Similarly, qt vs. the square root of time (t ) should bounding strength with E1 of substrates also followed this exhibit a linear relationship, and k3 can be obtained from turn. The Freundlich constant “n” values were greater than the slope of such a plot. 1, indicating that the adsorbent surface was heterogeneous The adsorption kinetic parameters of the three models and that the adsorption was favorable. and experimental qe values are given in Table 4. The ki- In Table 3, it was easily found that the coefficient of netic data suggested that the pseudo-second-order kinetic Langmuir model was all greater than 0.94, which probably model with a higher value of regression coefficient (R2) fit- can tell that the Langmuir model can also fit the isotherm ted best to the adsorption data, which demonstrated that the adsorption. But there is a big gap(>50%) between the max- adsorption might be the rate limiting step, and the rate-lim- imum adsorption amount obtained by model fitted and the iting step might be adsorption involving valence forces adsorption experiment. This showed that the fitting degree through the sharing of estrone [40]. According to the could not been measured only by the coefficient. Actually, pseudo-second-order model, the adsorption capacity of

TABLE 3 - Isotherm parameters of estrone adsorption onto constructed wetland substrates.

Langmuir isotherm Freundlich isotherm Substrates 2 2 b(L/mg) qmax（mg/kg） R K n R Ceramsite 0.1171 42.37 0.9699 1.5174 1.8278 0.9716 Zeolite 0.0168 56.73 0.9442 0.7836 1.0625 0.9943 Slag 0.1113 35.01 0.9580 0.2106 1.1463 0.9768 (Experimental conditions: 2 g substrates and the concentration of 1000 μg/L were used in 298 K.)

297 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

TABLE 4 - Parameters for kinetic models of pseudo-first-order, pseudo-second-order, and intra-particle diffusion model.

Intra-particle diffusion Pseudo first-order model Pseudo second-order model model Substrates 2 2 2 qecal k1 R qe k2 R k3 R (mg/kg) (1/h) (mg/kg) (kg/mg/h) (h1/2) Ceramsite 5.25 0.21 0.9444 28.18 3.0103×10-4 0.9999 1575.84 0.2556 Zeolite 0.53 0.13 0.8636 1.08 0.0010 0.9990 82.61 0.7218 Slag 0.17 0.06 0.7486 0.53 0.0084 0.9991 31.04 0.7138 qecal means the value of qe calculated from pseudo first-order model.

-1 -1 ceramsite, zeolite and slag are 21.18mg/kg, 1.08mg/kg and 532 cm and 463 cm represent Fe2O3 in the three sub- 0.53mg/kg, the relative error with the true value was within strates. The transmittance of each substrate between before 10%, so it is possible to estimate the adsorption process and after adsorption was different at 3444 cm-1 for different with pseudo-second-order kinetic model. The intra-particle moisture contents. The intensity of the characteristic peaks diffusion model exhibited poor fitting to the experimental of every substrate remained the same after adsorption, ex- data (Table 4), which implied that the intra-particle diffu- cept for the peak intensity. The transmittance exhibited an sion was not the only rate-limiting step. Other kinetic overall decrease after the adsorption of E1 for ceramsite mechanisms might also play a role in the E1 adsorption and zeolite in the range of 2350 cm-1~400 cm-1, and an in- process onto the substrates. crease for slag in the range of 1600 cm-1~400 cm-1. The difference in the absolute intensity of the IR peaks might 3.7. FTIR spectroscopy analysis be caused by the thickness of the KBr discs. This experi- The Fourier transform infrared spectra (FTIR) of the ment did not use quantitative methods in the tablet pressing three substrates before and after adsorption of estrone are process. The different dilution ratio of the substrates by shown in Fig. 5a-c to provide information regarding the KBr would cause different background values of the IR functional groups. The three figures all exhibited a broad spectra and lead to changes of the absolute intensity of the strong band at 3444 cm-1 (Fig. 5a-c), which indicated the characteristic peaks, which eventually results in different OH stretching vibration of H2O in the substrates. The IR peak transmittance values of infrared spectrum. In addition, E1 at 1620 cm-1 in the three figures might be due to the bending is a complex form of organic matter with various vibrations -1 -1 vibration of H2O. The bands at 1080 cm , 1012 cm , in the molecule, and the chemisorption process would re- -1 779 cm represent the SiO2 vibration, and the peaks at sult in a change in the peaks, similar to the results of a study

a 100

40 ceramsite E1 adsorbed ceramsite Transmittance (%) 20

0 4000 3600 3200 2800 2400 2000 1600 1200 800 400 Wavenumber (cm-1)

298 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

b 100

40 Transmittance (%)

zeolite 20 E1 adsorbed zeolite

0 4000 3600 3200 2800 2400 2000 1600 1200 800 400 Wavenumber (cm-1)

c 100

60 slag E1 adsorbed slag

Transmittance (%) Transmittance 20

0 4000 3600 3200 2800 2400 2000 1600 1200 800 400 Wavenumber (cm-1)

FIGURE 5 - Infrared spectrometry for ceramsite (a), zeolite (b), slag (c) before and after adsorption of E1.

from Parab and Sudersannan [41], in which a sharp and that the adsorption of estrone by the constructed wetland strong peak at 2110 cm-1 was found after adsorption of ce- substrates studied in this paper, ceramsite, zeolite, and slag, sium by CP-NIHCF. Consequently, it could be concluded was physisorption.

299 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

4. CONCLUSION REFERENCES

The present study investigated the E1 adsorption char- [1] Colborn, T., Dumanoski, D. and Myers, J. P. (1996) Our stolen acteristics on three constructed wetland substrates: cerams- future. Dutton, New York. ite, zeolite and slag. The adsorption capacities were studied [2] Rodgers-Gray, T.P., Jobling, S., Kelly, C., Morris, S., Brighty, in detail and found to follow the order of ceramsite > zeo- G., Waldock, M.J., Sumpter, J.P. and Tyler, C.R. (2001) Ex- lite > slag, suggesting the potential of ceramsite to act as a posure of juvenile roach (Rutilusrutilus) to treated sewage effluent induces does-dependent and persistent disruption in constructed wetland substrate for the removal of E1. Based gonadal duct development. Environ. Sci. Technol. 35, 462- on batch experiments, it was found that the estrone adsorp- 470. tion capacity of the constructed wetland substrates in- [3] Matthiessen, P. and Gibbs, P.E. (1998) Critical appraisal of the creased with an increase of the temperature and/or with a evidence for tributyltin-mediated endocrine disruption in mol- decrease of the particle size of the substrates. The experi- lusks. Environ. Toxicol. Chem. 17, 37-43. mental data were analyzed by using the Langmuir and [4] Zhang, Y.P. and Zhou, J.L. (2005) Removal of estrone and 17 Freundlich models. According to the Langmuir isotherm beta-estradiol from water by adsorption. Water Res. 39, 3991- analysis, the monolayer adsorption capacities were deter- 4003. mined to be 42.37, 56.73 and 35.01 mg/kg for ceramsite, [5] Coleman, H.M., Routledge, E.J. and Sumpter, J.P. (2004) zeolite and slag, respectively. The reason for the differ- Rapid loss of estrogenicity of steroid estrogens by UVA pho- ences in the capacities might be the differences of the phys- tolysis and photocatalysis over an immobilised titanium diox- ical characteristics of the substrate surface because the ide catalyst. Water Res. 38, 3233-3240. physical parameters (BET, pore volume and pore radius) [6] Tanizaki, T., Kadokami, K. and Shinohara, R. (2002) Catalytic also followed the order of ceramsite > zeolite > slag. By photodegradation of endocrine disrupting chemicals using ti- comparing the regression coefficients of the isotherm anal- tanium dioxide photosemiconductor thin films. Bull. Environ. Contam. Toxicol. 68, 732-739. ysis, the Freundlich isotherm matched better for the adsorption data of the three substrates. In addition, the E1 up- [7] Ohko, Y., Iuchi, K.I., Niwa, C., Tatsuma, T., Nakashima, T., take studies indicated that the adsorption process consisted Iguchi, T., Kubota, Y. and Fujishima, A. (2002) 17 beta-estro- dial degradation by TiO2 photocatalysis as means of reducing of four stages: the instantaneous adsorption stage, the de- estrogenic activity. Environ. Sci. Technol. 36, 4175-4181. sorption stage, the rapid adsorption stage, and the equilib- [8] Joss, A., Andersen, H. and Ternes, T. (2004) Removal of es- rium stage. The adsorption kinetics study indicated that the trogens in municipal wastewater treatment under aerobic and pseudo-second-order kinetic model fitted best to the exper- anaerobic conditions: Consequences for plant optimization. imental data. The functional groups of three substrates be- Environ. Sci. Technol. 38, 3047-3055. fore and after adsorption of E1 were identified via FTIR [9] Servos, M.R., Bennie, D.T., Burnison, B.K., Jurkovic, A., spectroscopy. There were few differences after the adsorp- Mclnnis, R., Neheli, T., Schnell, A., Seto, P., Smyth, S.A. and tion, such as for some of the peak intensities, which were Ternes, T.A. (2005) Distribution of estrogens, 17 beta-estra- caused by the difference of the KBr ratio in the tablet press- diol and estrone, in Canadian municipal wastewater treatment plants. Sci. Total Environ. 336, 150-155. ing process. The results indicated that the adsorption process of estrone for the studied substrates was physisorption. [10] Nghiem, L.D. and Schafer, A.I. (2006) Critical risk points of nanofiltration and reverse osmosis processes in water recy- According to the results presented in this study, cling applications. Desalination. 187, 303-312. ceramsite can be a promising constructed wetland substrate [11] Fukuhara, T., Iwasaki, S., Kawashima, M., Shinohara, O. and for the removal of E1. The results of this study will provide Abe, I. (2006) Adsorbability of estrone and 17β-estradiol in a basis for choosing a constructed wetland substrate for the water onto activated carbon. Water Res. 40, 241-248. removal of E1. [12] Hildebrand, C., Londry, K.L. and Farenhorst, A. (2006) Sorp- tion and desorption of three endocrine disrupters in soils. J. Environ. Sci. Heal. B. 41, 907-921.

[13] Zhang, Z.B. and Hu, J.Y. (2010) Effect of environmental fac- ACKNOWLEDGMENTS tors on estrogenic compounds adsorption by MIP. Water Air Soil Pollut. 210, 255-264. The authors acknowledge the financial support pro- [14] Han, J., Qiu, W., Cao, Z., Hu, J.Y. and Gao, W. (2013) Ad- vided by the Natural Science Foundation of China sorption of ethinylestradiol (EE2) on polyamide 612: Molecu- (51109038), the National Key Technology R&D Program lar modeling and effects of water chemistry. Water Res. 47, (2013BAJ10B12-02), and the Major Science and Technol- 2273-2284. ogy Program for Water Pollution Control and Treatment [15] Song, H.L., Yang, X.L., Nakano, K., Nomura, M., Nishimura, (2012ZX07101-005). O. and Li, X.N. (2011) Elimination of estrogens and estrogenic activity from sewage treatment works effluents in subsurface and surface flow constructed wetlands. Int. J. Environ. Anal. The authors have declared no conflict of interest. Chem. 91, 600-614.

[16] Reyes-Contreras, C., Matamoros, V., Ruiz, I., Soto, M. and Bayona, J.M. (2011) Evaluation of PPCPs removal in a combined anaerobic digester-constructed wetland pilot plant treating urban wastewater. Chemosphere. 84, 1200-1207.

300 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

[17] Matamoros, V. and Bayona, J.M. (2006) Elimination of phar- [35] Mane, V.S., Mall, I.D. and Srivastava, V.C. (2007) Use of ba- maceuticals and personal care products in subsurface flow gasse fly ash as an adsorbent for the removal of brilliant green constructed wetlands. Environ. Sci. Technol. 40, 5811-5816. dye from aqueous solution. Dyes. Pigm. 73, 269-278. [18] Imai, S., Shiraishi, A., Gamo, K., Watanabe, I., Okuhata, H., [36] Nandi, B.K., Goswami, A. and Purkait, M.K. (2009) Adsorp- Miyasaka, H., Ikeda, K., Bamba, T. and Hirata, K. (2007) Re- tion characteristics of brilliant gree dye on kaolin. J. Hazard. moval of phenolic endocrine disruptors by Portulacaoleracea. Mater. 161, 387-395. J. Biosci. and Bioeng. 103, 420-426. [37] Ho, Y.S. and McKay, G. (1999) Pseudo-second order model [19] Toyama, T., Yu, N., Kumada, H., Sei, K., Ike, M. and Fujita, for sorption processes. Process Biochem. 34, 451-465. M. (2006) Accelerated aromatic compounds degradation in aquatic environment by use of interaction between Spi- [38] Pokhrel, D. and Viraraghavan, T. (2008) Arsenic removal rodelapolyrrhiza and bacteria in its rhizosphere. J. Biosci. and from aqueous solution by modified A. nigerbiomass: batch ki- Bioeng. 101, 346-353. netic and isotherm studies. J. Hazard. Mater. 150, 818-825. [20] He, Y.Q., Xu, G., Tang, L., Qian, D.Y., Ren, L.T., Hu, G.Y., [39] Gerente, C., Lee, V.K.C., Le Cloirec, P. and McKay, G. (2007) Lei, J.Q. and Wu, M.H. (2012) Research on photolysis of ster- Application of chitosan for the removal of metals from oid estrogens in aquatic system. Adv. Mater. Res. 343-344, wastewaters by adsorption-mechanisms and models review. 241-245. Crit. Rev. Environ. Sci. Technol. 37, 41-127. [21] Zhang, B.H., Wu, D.Y., Wang, C., He, S.B., Zhang, Z.J. and [40] Feng, N.C., Guo, X.Y., Liang, S., Zhu, Y.S. and Liu, J.P. Kong, H.N. (2007) Simultaneous removal of ammonium and (2011) Biosorption of heavy metals from aqueous solutions by phosphate by zeolite synthesized from coal fly ash as influ- chemically modified orange peel. J. Hazard. Mater. 185, 49- enced by acid treatment. J. Environ. Sci. 19, 540-545. 54. [22] Chen, J.G., Kong, H.N., Wu, D.Y., Chen, X.C., Zhang, D.L. [41] Parab, H. and Sudersanan, M. (2010) Engineering a lignocel- and Sun, Z.H. (2007) Phosphate immobilization from aqueous lulosic biosorbent-Coir pith for removal of cesium from aque- solution by fly ashes in relation to their composition. J. Haz- ous solutions: Equilibrium and kinetic studies. Water Res. 44, ard. Mater. 139, 293-300. 854-860. [23] Zeng, L., Li, X.M. and Liu, J.D. (2004) Adsorptive removal of phosphate from aqueous solutions using iron oxide tailings. Water Res. 38, 1318-1326. [24] ASTM Standard D422-63. (2007) Standard Test Methods for Particle-Size Analysis of Soils. ASTM International. West Conshohocken, PA. doi: 10.1520/D0422-63R07. [25] Hameed, B.H., Krishni, R.R. and Sata, S.A. (2009) A novel agricultural waste adsorbent for the removal of cationic dye from aqueous solutions. J. Hazard. Mater. 162, 305-311. [26] Ramesh, S.T., Gandhimathi, R., Joesun, J.H. and Nidheesh, P.V. (2013) Novel agricultural waste adsorbent, Cyperusro- tundus, for removal of heavy metal mixtures from aqueous solutions. Environ. Eng. Sci. 30, 74-81. [27] Zubair, A., Bhatti, H.N., Hanif, M.A. and Shafqat, F. (2008) Kinetic and equilibrium modeling for Cr(Ⅲ) and Cr(Ⅵ) removal from aqueous solutions by Citrus reticulate waste biomass. Water Air Soil Pollut.191, 305-318. [28] Manzoor, Q., Nadeem, R., Iqbal, M., Saeed, R. and Ansari, T.M. (2013) Organic acids pretreatment effect on Rosa bour- boniaphyto-biomass for removal of Pb(Ⅱ) and Cu(Ⅱ) from aqueous media. Bioresour. Technol. 132, 446-452. Received: April 11, 2014 [29] Langmuir, I. (1918) The adsorption of gases on plane surfaces Revised: May 15, 2014 of glass, mica and platinum. J. Am. Chem. Soc. 40, 1361-403. Accepted: June 11, 2014 [30] Freundlich, H. (1906) Over the adsorption in solution. Z Phys. Chem. 57, 385-471. CORRESPONDING AUTHOR [31] Tang, Y.K., Chen, L., Wei, X.R., Yao, Q.Y. and Li, T. (2013) Removal of lead ions from aqueous solution by the dried aquatic plant, Lemnaperpusilla Torr. J. Hazard. Mater. 244, Hai-Liang Song 603-612. School of Energy and Environment [32] Sun, K., Jin, J., Gao, B. and Zhang, Z. (2012) Sorption of 17α- Southeast University ethinyl estradiol, bisphenol A and phenanthrene to different Nanjing 210096 size fractions of soil and sediment. Chemosphere. 88, 577-583. P.R. CHINA [33] Harter, R.D. and Baker. D.E. (1997) Applications and misap- plications of the Langmuir equation to soil adsorption phenomena. Soil Sci. Soc. Am. J. 41, 1077-1080. Phone: 86-25-83794171 Fax: 86-25-83795618 [34] Mohan, S.V., Ramanaiah, S.V., Rajkumar, B. and Sarma, P.N. (2007) Removal of fluoride from aqueous phase by biosorp- E-mail: [email protected] tion onto algal biosorbent Spirogyra sp.-IO2. Sorption mechanism elucidation. J. Hazard. Mater. 141, 465-474. FEB/ Vol 24/ No 1b/ 2015 – pages 291 – 301

301 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

REMOVAL OF CADMIUM IONS FROM AQUEOUS SOLUTIONS BY MICROORGANISMS OF ACTIVATED SLUDGE

Sureyya Altin1,*, Ahmet Altin1, Bekir Fatih Kahraman1, Sonay Alemdar2 and Elif Alaydin1

1Bulent Ecevit University, Department of Environmental Engineering, Zonguldak, 67100, Turkey 2Gunluler Waste Collection and Recycling Company, Zonguldak, 67100, Turkey

ABSTRACT cation processes limit its large scale usage [3]. Many researchers have focused on cheap and easily applicable al- In this study, the utilization of dried activated sludge ternative materials instead of activated carbon [3]. Biomass for removal of Cd2+ from aqueous solution in a batch sys- is one of the promising alternative sorption materials due tem was investigated. Initial pH, biosorbent dosage, con- to several advantages including reduced cost, environmen- tact time, and initial metal concentration parameters were tal acceptability, regenerability and simple application [5]. selected to determine optimal process conditions. The bi- For example, some biological materials (biosorbents) such osorption mechanism was examined by SEM, FTIR and as agricultural wastes, sugar beet pulp, fungi, algae, bacte- EDX results. The optimum conditions for Cd2+ biosorption ria and yeasts can be used for heavy metal removal from were found to be 6.0, 120 min., 1.2 g.L-1 and 80 mg.L-1, industrial wastewater, effectively [5-9]. respectively for initial pH, contact time, biosorbent dosage 2+ Biosorbents can be used either in alive or dead form. and initial Cd concentration. Langmuir and Freundlich However, living cells suffer from heavy metals toxicity isotherms were used to model the biosorption equilibrium which causes cell death. Besides, COD and BOD concen- data, and it was determined that the system followed the trations raise in the wastewater due to the nutrient addition Langmuir isotherm, and the sorption capacity of the bio- -1 in order to supply the necessary nutrient requirements of sorbent was found to be 15.43 mg.g . Biosorption fol- living cells [10-12]. However, dead biomass is not affected lowed a pseudo-second-order rate model. Two main mech- 2+ by heavy metals toxicity; it is cheap and easily controlled. anisms of Cd biosorption onto the dried activated sludge In addition, non-living biosorbents can be regenerated and were adsorption to the C-H bonds and ion exchange with reused [10,13]. Therefore, dead biomass may be consid- Na+, K+ and Ca2+ ions. ered more appropriate for heavy metal removal. There is a growing interest in use of activated sludge (which is mix-

KEYWORDS: ture of bacteria, fungi, yeast, algae and protozoa) as a bio- Cadmium removal, Biosorption, Adsorption Kinetics sorbent [3, 7, 10, 14-16].

Biosorption processes include extracellular mass transfer of metal onto the binding regions of biomass and 1. INTRODUCTION diffusion into the particulates. These processes are generally fast and reversible [17]. Extracellular polymeric sub- Wastewater discharges into the aquatic environment stances protect living cells from metal toxicity via some have been on the increase as a result of urbanization. Most mechanisms such as ion exchange, complexation with neg- of these discharges contain several toxic substances, espe- atively charged groups, adsorption and precipitation [15, cially heavy metals. The presence of heavy metals in the 17-19]. Interaction between metals and biomass is related environment causes major concern because of their toxicity with biopolymers of biomass or functional groups located and bio-accumulation tendency [1]. in the cell wall. These functional groups are; carboxylic, There are several techniques developed for the re- amino, phosphate and sulphate groups [7]. Physicochemi- moval of heavy metals from wastewater such as chemical cal properties of these organic compounds can be modified precipitation, solvent extraction, membrane separation, re- by changing solution characteristics. Therefore, some pa- verse osmosis, evaporation, electrolysis and ion exchange. rameters such as pH, ionic strength, redox potential, metal These techniques are not cost effective and metal recovery concentration, complexing properties of anions, are known capacities are insufficient [2-4]. Adsorption onto activated as important factors in the biosorption process [20]. carbon is effectively used for the removal of heavy metals. Amount of adsorbate on adsorbent, at fixed tempera- However, high regeneration costs and losses in the appli- ture, is a function of concentration and this function is called the adsorption isotherm. Adsorption isotherms are * Corresponding author also representative for biosorption processes [15, 21, 22].

302 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

It is not likely to infer mechanisms of biosorption from iso- cadmium (II) chloride salt (Merck, Germany) in deion- therms [17]. In order to understand the mechanisms of bi- ized water (18.2 µS.cm-1). A total of 0.05 g biosorbent in osorption, the physical and chemical changes on the sur- 50 mL volume of the stock solution (initial concentrations face of the biosorbent also have to be understood. ranged from 25 to 175 mg.L-1) was adjusted to desired pH -1 Activated sludge is an abundant, freely available mat- value with 1.0 mol.L HCl and NaOH solutions, and was ter because it is a by-product of wastewater treatment pro- agitated until equilibrium state was reached using an incu- cesses. Utilization of activated sludge can significantly re- bator shaker (Gerhardt, Germany) at 125 rpm and room duce the cost of heavy metal removal from wastewater. The temperature (25°C). Sample solution, which was in equi- objective of the present work is to investigate the biosorp- librium state, was centrifuged at 5000 rpm for 20 minutes 2+ tion potential of the dried activated sludge for the removal and biomass was separated from the liquid. Cd amount in of cadmium (Cd2+) from the aqueous solution. Optimum the supernatant was analyzed by atomic adsorption spec- biosorption conditions were determined as the function of trophotometer (AAS) (1100B, Perkin Elmer, USA). Ad- contact time, initial pH, biosorbent dosage and the initial sorbed metal quantity was calculated by the following metal concentrations. Physical and chemical properties of equation: biosorbent, surface structure before and after biosorption, V (C  C ) changes in functional groups were also analyzed. In addi- qe  0 e (1) tion, Langmuir and Freundlich models were used to de- M -1 scribe equilibrium isotherms. In order to define kinetic con- where qe (mg.g ) is the equilibrium adsorption capac- stants of the biosorption, Lagergren and Ho kinetic models ity, C0 and Ce are the initial and equilibrium concentrations were applied to the experimental data. (mg.L-1) of Cd2+ in solution, V (L) is the volume and M (g) is the weight of dry biosorbent. Removal efficiencies in the experiments were determined by the following expression: 2. MATERIALS AND METHODS C C Efficiency (%) 0 e (2) 2.1 Preparation of biosorbent C0 A reactor was set up for activated sludge production in the laboratory. The reactor was filled with 5 L of domestic 2.4 Biosorption isotherms wastewater and continuously aerated by air pumps. Microor- ganism production was enhanced by the addition of a nutrient Biosorption isotherm is an important component for solution into the reactor at a loading rate of 5 mL.L-1.day-1. the clarifying of adsorption mechanism. In order to determine the equilibrium state of the biosorption, experiment Microorganism growth was observed for two weeks. At results were applied to the well known Langmuir (Eq.3) the end of two weeks, two third of the activated sludge flocs and Freundlich (Eq.4) adsorption isotherms. was removed from the reactor and centrifuged at 3000 rpm for 5 min. More wastewater was added to remaining activated C 1 1 e  C  (3) sludge in the reactor to maintain the microorganism q q e q b growth. For the preparation of biosorbent, the centrifuged e max max material was dried overnight in an oven at 60°C. It was then 1 grounded into powder and sieved to size ≤ 150 μm. The log qe  log K f  log Ce (4) prepared biosorbent was stored in a desiccator for all ex- n -1 2+ periments. where (mg.g ) is the amount of Cd adsorbed -1 -1 per unit mass of the biosorbent, , (L.g ) and (L.mg ) 2.2 Characterization of the biosorbent values are constants. Elemental composition of the biosorbent was determined using Leco TruSpec (CHNS) elemental analyzer 2.5 Biosorption kinetics while the specific functional groups in the biosorbent were Biosorption kinetics is a useful tool to explain the fac- identified using PerkinElmer Frontier FTIR (Fourier Trans- tors affecting sorption mechanism and efficiency. Sorption form Infra-Red) spectroscopy system. The surface morphol- process is related with the nature of the adsorbent, physical ogy of the biosorbent was examined using FEI Quanta and chemical properties of the adsorbate and operating con- FEG450 SEM/EDX (scanning electron microscopy/ en- ditions of the process. It is important to know the kinetics ergy dispersive x-ray spectroscopy) system. All these anal- data in order to determine the effective step of the adsorp- yses were conducted on the biomass samples before and tion [23]. after the adsorption process. In order to investigate the mechanisms of adsorption, there are numerous kinetic models such as first-order and 2.3 Biosorption experiments second-order reversible ones, and first-order and second- Biosorption tests in this study were performed at order irreversible ones. On the other hand, reaction orders room temperature. Stock Cd2+ solution used in the tests based on the capacity of the adsorbent have also been pre- was prepared by dissolving analytical-grade anhydrous sented, such as Lagergren’s pseudo-first-order equation

303 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

(Eq. 5) and Ho’s pseudo-second-order expression (Eq. 6) 3. RESULTS AND DISCUSSION [24, 25]. 3.1 Effect of the contact time k1  t logqe  qt  log qe  (5) The contact time between adsorbent and adsorbate has 2.303 great importance in adsorption process. In general, the bi- t 1 t osorption capacity and the removal efficiency of metal ions   (6) become higher when the contact time is prolonged. In this 2 2+ qt k2  qe qe study, to determine the effect of the contact time, Cd concentrations were analysed after biosorption experiments -1 and (mg.g ) are the sorption capacity at equilib- for different contact times (15, 30, 45, 60, 75, 90, 105, 120, -1 rium and at time , respectively and (L.min ) and 140 minutes). (g.mg-1.min-1) are reaction rate constants of the models, re- The results in Fig.1 (A) showed that equilibrium was spectively. established within a contact time of 120 min. and sorption First one of these models is based on the assumption did not change with further increase of contact time up to that the rate of occupation of adsorption sites is propor- 20 min. The probability of confrontation between Cd2+ ions tional to the number of unoccupied sites. [5] Second model and active sites on the dry biomass increased with increas- assumes that biosorption capacity is proportional to the ing contact time. Thus, concentration of Cd2+ ions de- number of occupied active sites on the biosorbent. Here, creased in solution at equilibrium. However, metal adsorp- rate limiting step is chemisorption. Several chemical inter- tion decreased over time due to the occupation of all active actions contribute to chemisorption such as electron shar- points or reduction in the number of unoccupied active ing or the exchange between biosorbent and adsorbate, sites. When whole capacity of the adsorbent was used, ad- complexation, coordination and chelation. [26]. sorption reached equilibrium.

1,0 100

0,8 80 ) 0

0,6 60 removal, (%) removal,

removal, (C/C removal, 0,4 40 2+ 2+ Cd Cd

0,2 20

0,0 0 0 20 40 60 80 100 120 140 23456789 Time, (min.) pH value (A) (B)

100 100

80 80

60 60 removal, (%) removal, 40 removal,(%) 40 2+ 2+ Cd Cd

20 20

0 0 0501001502000,0 0,5 1,0 1,5 2,0 2,5 2+ -1 Initial Cd concentration, (mg.L ) Biosorbent concentration, (g.L-1) (C) (D)

FIGURE 1 - Effect of contact time (Cd2+ = 100 mg.L-1, biosorbent = 1.0 g.L-1, pH = original solution pH) (A), Effect of pH (Cd2+ = 100 mg.L-1, biosorbent = 1.0 g.L-1, contact time = 120 min.) (B), Effect of initial Cd2+ concentration (pH = 8, biosorbent = 1.0 g.L-1, contact time = 120 min.) (C) Effect of biosorbent concentration on the Cd2+ removal efficiency (Cd2+ = 80 mg.L-1, initial pH = 8, contact time = 120 min.) (D).

304 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

3.2 Effect of pH ome less in comparison to the moles of metal ion present The interaction between the metal ions and the func- in solution. [27] tional groups present on the biomass depends on the nature of the biosorbent as well as the aqueous chemistry of the 3.4 Effect of biosorbent concentration metal ions. It is well known that pH is one of the major pa- The biosorbent concentration experiments were per- rameters controlling the metal sorption of the biosorbents. formed at various biomass doses from 0.3 to 2.0 g.L-1. As 2+ In this study, the effect of solution pH on the biosorp- it is seen from Fig.1 (D), the Cd removal efficiency rapidly increases from 44% to 80% by increasing used bio- tion was investigated in pH range of 3.0-8.0. In the experi- -1 ments, pH 8.0 was taken as the upper limit to avoid possi- sorbent dose from 0.3 to 1.2 g.L , respectively. Several re- ble hydroxide precipitation. As seen in Fig.1 (B), biosorp- searchers also reported that the increase in the biosorbent dosage results in more adsorption sites which in turn leads tion rate of Cd2+ in the solution increased by increasing the initial pH of the solution, and decreased sharply under the to higher removal efficiencies [2]. In other words, the pres- initial pH value 4. According to these results, it can be con- ence of high amounts of biosorbent in solution causes a rapid surface biosorption. Increase in the sorption with the cluded that optimum initial pH range for biosorption of Cd2+ is between 6.0 and 8.0. The low removal efficiencies increase in the biosorbent dose can be attributed to in- seen under initial pH of 6 can be attiributed to competition creased sorbent surface area of mesopores and the availability of more sorption sites. Thus, active sites needed for of a large number of protons with the metal cations for ac- 2+ tive exchange sites on the biomass. Cd uptake can be rapidly filled. The decreasing metal removal at high biosorbent dosage could be explained by considering a partial cell aggregation. This results in the 3.3 Effect of initial metal concentration three dimensional structure of the cell wall and the internal 2+ - + The influence of the initial Cd concentrations on ad- linkages between the reactive groups (COO and NH3 ), sorption was investigated by changing the concentrations thus reducing Cd2+ diffusion through the structure and the (175, 150,120,100, 80, 60, 40, 25 mg.L-1) at the initial pH 8 accessibility of the active sites for the sorption [28]. and the contact time 120 min. In this work, Langmuir and Freundlich isotherms were As it is seen from Fig.1 (C), when the initial metal con- used to optimize biosorption process parameters. Experi- centrations increased up to 20 mg.L-1, removal efficiency ments were carried out at initial pH 8.0, 1.2 g.L-1 bio- increased aggressively up to 90%, and then the efficiency sorbent dose and 120 min contact time. Initial Cd2+ con- decreased slightly. These results can be explained on the centrations varied from 25 to 175 mg.L1. The changes in basis that at the lower initial concentrations, metal ions eas- equilibrium concentrations for various initial metal concen- ily reach and bind onto the active sites but at higher con- trations were given in Fig.2 and the obtained isotherm pa- centrations, the available active sites for the sorption be- rameters and correlation coefficients were listed in Table 1.

0,4 2,4

2,2

0,3 2,0

e 1,8 e

/ q 0,2 e Log q Log C 1,6

1,4 0,1

1,2

0,0 1,0 0 1020304050 0,2 0,4 0,6 0,8 1,0 1,2 1,4 1,6 1,8 C Log C e e (a) (b)

FIGURE 2 - (a) Langmuir and (b) Freundlich isotherms for the Cd2+ biosorption.

TABLE 1 - Langmuir and Freundlich isotherm parameters for biosorption of Cd2+ ions.

Langmuir Isotherm Freundlich Isotherm -1 2 -1 2 qmax (mg/g) b (L.mg ) R Kf (L.g ) 1/n R 15.43 0.068 0.9917 1.37 0.614 0.9640

305 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

The fact that metal biosorption can be modeled by 3.5 Biosorption kinetics Langmuir and Freundlich model, have been identified by In order to investigate the adsorption kinetics of Cd2+ many researchers [3, 29-30]. Similarly, in our results seen on the biosorbent, pseudo-first-order and pseudo-second- in Fig.2, the correlation coefficients obtained from both order models were taken into consideration. Langmuir and Freundlich isotherms are higher than 95%. However, Langmuir model results are relatively better than It can be seen from the results presented in Fig.3 and the Freundlich model judging by higher R2 values. This Table 2 that, correlation coefficients of both kinetic models are higher than 94%. However, the use of pseudo-second- provides an initial indication that the governing mechanism 2+ is chemisorption with high possibility of monolayer ad- order kinetics for the Cd sorption by dried activated sorption of cadmium ions occurring on the surface of the sludge may be more appropriate due to having slightly high biosorbent. correlation coefficient.

It can be concluded from the results that the governing 3.6 FTIR analysis mechanisms of Cd2+ biosorption by using dry activated Fig.4 shows the FTIR spectra of biosorbent before and sludge are both chemical and physical sorptions. Phy- after adsorption of cadmium ions. The FTIR spectra of raw sisorption is a result of Van der Waals forces and adsorbed and experimented biomass were taken in the range of 450– molecules are not fixed on a specific site. In contrast, 4000 cm−1 and functional groups responsible for the bio- chemisorption is a result of the chemical interaction with sorption were investigated. the adsorbent, and the movement of the adsorbed molecules is not possible [7,15]. Adsorbate that is bonded with The cell walls of microorganisms contain large mole- chemisorption usually does not accumulate more than a cules (peptidoglycan) combined with teichoic acid and pol- molecular layer. Chemisorption changes only according to ysaccharides. These molecules and intracellular substances the specific sites and functional groups. Thus, chemisorption can adsorb heavy metals. Especially, functional groups degree is different for each biomass. Due to the broad diver- such as carboxylate (COO-), hydroxyl (-OH), and others sity of microorganisms, there is a heterogeneous structure in (-NH), (-C-N), (-C-O), (-C-H), (-C=O) are responsible for the activated sludge. Cadmium ions may be adsorbed by the sorption. Peak values on FTIR spectrum higher than chemical or physical binding onto the heterogeneous regions 1500 define the organic bond structures that can chemically of biosorbent surface as a monolayer. This explains both the adsorb heavy metals. surface heterogeneity and the monolayer uptake.

2,5 1,5 a) b)

2,0

1,0 ) t 1,5 t -q e t / q

1,0 Log ( q Log 0,5

0,5

0,0 0,0 0 20 40 60 80 100 120 0 20406080100120 Time, (min.) Time, (min.) FIGURE 3 - (a) Pseudo-first-order, and (b) Pseudo-second-order models

TABLE 2 - Pseudo-first-order and pseudo-second-order kinetics parameters for Cd2+ biosorption

Pseudo-first-order Pseudo-second-order -1 -1 qe (mg.g ) 10.36 qe (mg.g ) 9.157 -1 -1 k1 (min. ) 0.38 k2 (g.(mg. min) 0.0038 R2 0.9482 R2 0.9892

306 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

(A)

(B)

FIGURE 4 - The FTIR spectras of biosorbent before (A) and after (B) the biosorption.

TABLE 3 - The changes in biosorbent structure before and after the biosorption study.

Analysis Before biosorption After biosorption Multipoint BET, m2.g-1 (R2=0,99) 8,786 13,432 Langmuir surface area, m2.g-1 (R2=0,99) 13,739 19,831 Pore volume , cc.g-1 2.900x10-3 4.868x10-3 (for pore diameter < 1.94 nm) Average pore diameter, nm 1.320 1.450 Density, g.cc-1 2.0936 1.3236 Cd2+, mg.kg-1 22.560 24 060 Elemental analysis (C-H-N-S), (%) 7.39-3.72-1.75-3.16 6.47-3.39- 1.62-3.83 Volatile solids, (%) 41 34

3.7 Surface analysis results these results, it can be inferred that some portion of the Cd+2 In order to find out the changes in biosorbent structure ions undergo physical adsorption onto the biomass surface. before and after the biosorption process, some surface and Carbon, hydrogen, nitrogen and volatile solids values given elemental analyses were conducted, and results are pre- in Table 3 also decrease after the biosorption experiments. sented in Table 3. As it is seen, from the FTIR results (Fig 4.), this may be due to the binding of Cd+2 chemically to the adsorbate. Both BET analysis and Langmuir analysis results in Table 3. show that the surface area, the pore volume and Scanning electron microscopy, SEM, micrographs of the pore diameter of the biosorbent increase after the sorp- the biosorbent before and after the sorption are shown in tion experiments. Increase in the surface area supports the Fig. 5. As it is seen from Fig.5, before the sorption, the bi- idea that adsorption occurs onto the biosorbent surface. osorbent surface appears to be rough and sponge-like Furthermore, increases in the pore volume and pore diam- which indicates a typical biomass texture. After the adsorp- eter also indicate the sorption of Cd+2 ions. According to tion experiment, its surface experiences a slight change in

307 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

(A) (B)

FIGURE 5 - The SEM micrographs of biosorbent before (A) and after (B) the biosorption.

(a)

(b) FIGURE 6 – The EDX spectra of biosorbent before (A) and after (B) the biosorption.

morphology. Energy dispersive X-ray (EDX) analysis was (K), calcium (Ca), magnesium (Mg), phosphorus (P), chlo- also conducted to evaluate the adsorption of Cd2+ on the ride (Cl), carbon (C), and oxygen (O), but does not show biosorbent (Fig.6). the characteristic signal of Cd2+ ions on the surface of the biosorbent. After the biosorption experiment, a signiﬁcant The EDX spectrum for the raw biosorbent seen in amount of K+, Na+ and Ca2+ releases from the biosorbent. Fig.6(a) indicates the presence of sodium (Na), potassium As it is expected, Cd2+ ions detected after the biosorption

308 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

process suggest a possible ion exchange occurrence be- [8] Kuyucak, N., Volesky, B. (1988) Biosorbents for recovery of tween these ions and Cd2+ ions. According to these obser- metals from industrial solutions. Biotechnology letters, 10(2), 137-142 vations, it can be concluded that the ion exchange mechanism plays an important role in cadmium uptake. [9] Mohan, D., Singh, K.P. (2002) Single- and multi-component adsorption of cadmium and zinc using activated carbon derived from bagasse—an agricultural waste. Water Research, 36(9), 2304-2318 4. CONCLUSION [10] Al-Qodah, Z. (2006) Biosorption of heavy metal ions from aqueous solutions by activated sludge. Desalination, 196(1–3), Based on all results, it can be concluded that dried ac- 164-176. tivated sludge can be evaluated as an alternative biosorbent [11] Dilek, F.B., Gokcay, C.F., Yetis, U. (1998) Combined effects for treatment of Cd+2 containing wastewater, since it is of Ni(II) and Cr(VI) on activated sludge. Water Research, low-cost biomass and has a considerably high biosorption 32(2), 303-312. capacity. Initial pH, biosorbent concentration and removal [12] Yan, G., Viraraghavan, T. (2000) Effect of pretreatment on the time can be taken into account as important process param- bioadsorption of heavy metals on Mucor rouxii. Water sa-pre- eters. The physical and chemical sorption forces are effec- toria-, 26(1), 119-124 tive at the excessive Cd+2 accumulations on the biosorbent. [13] Moffat, A.S. (1995) Plants Proving Their Worth in Toxic The Langmuir isotherm fits better than the Freundlich iso- Metal Cleanup. Science, 269 (5222), 302-303 therm indicating the applicability of monolayer coverage [14] Aksu, Z., Akpınar, D. (2000) Modelling of simultaneous bio- of Cd+2 on biosorbent surface. The pseudo-second-order sorption of phenol and nickel(II) onto dried aerobic activated sludge. Separation and Purification Technology, 21(1–2), 87- kinetic model describes the adsorption kinetics more accu- 99 rately. [15] Arican, B., Gokcay, C.F., Yetis, U. (2002) Mechanistics of nickel sorption by activated sludge. Process Biochemistry, 37(11), 1307-1315 ACKNOWLEDGEMENT [16] Sağ, Y., Tatar, B., Kutsal, T. (2003) Biosorption of Pb(II) and Cu(II) by activated sludge in batch and continuous-flow stirred This research was financially supported by the Re- reactors. Bioresource Technology, 87(1), 27-33 search Fund of Bulent Ecevit University (Project code: [17] Tsezos, M., Volesky, B. (1982) The mechanism of uranium 2010-45-10-01). biosorption by Rhizopus arrhizus. Biotechnology and Bioen- gineering, 24(2), 385-401. The authors have declared no conflict of interest. [18] Brown, M.J., Lester, J.N. (1982) Role of bacterial extracellular polymers in metal uptake in pure bacterial culture and activated sludge—II Effects of mean cell retention time. Water Research, 16(11), 1549-1560. REFERENCES [19] Rudd, T., Sterritt, R., Lester, J. (1984) Formation and condi- tional stability constants of complexes formed between heavy [1] Igwe, J., Abia, A. (2006) A bioseparation process for remov- metals and bacterial extracellular polymers. Water Research, ing heavy metals from waste water using biosorbents. Afri- 18(3), 379-384 can Journal of Biotechnology, 5(11). [20] Ledin, M., Pedersen, K., Allard, B. (1997) Effects of pH and [2] Baysal, Z., Çinar, E., Bulut, Y., Alkan, H., Dogru, M. (2009) ionic strength on the adsorption of Cs, Sr, Eu, Zn, Cd and Hg Equilibrium and thermodynamic studies on biosorption of byPseudomonas putida. Water, Air, & Soil Pollution, 93(1), Pb(II) onto Candida albicans biomass. Journal of Hazardous 367-381. Materials, 161(1), 62-67. [21] Chang, D., Fukushi, K., Ghosh, S. (1995) Stimulation of acti- [3] Wang, X., Chen, L., Xia, S., Zhao, J., Chovelon, J.-M., Re- vated sludge cultures for enhanced heavy metal removal. Wa- nault, N.J. (2006) Biosorption of Cu(II) and Pb(II) from aque- ter environment research, 67(5), 822-827. ous solutions by dried activated sludge. Minerals Engineering, [22] Lawson, P.S., Sterritt, R.M., Lester, J.N. (1984) Adsorption 19(9), 968-971. and complexation mechanisms of heavy metal uptake in acti- [4] Xu, H., Liu, Y. (2008) Mechanisms of Cd2+, Cu2+ and Ni2+ bi- vated sludge. Journal of Chemical Technology and Biotech- osorption by aerobic granules. Separation and Purification nology. Biotechnology, 34(4), 253-262. Technology, 58(3), 400-411. [23] Ho, Y.S., McKay, G. (1999) Pseudo-second order model for [5] Li, X., Xu, Q., Han, G., Zhu, W., Chen, Z., He, X., Tian, X. sorption processes. Process Biochemistry, 34(5), 451-465. (2009) Equilibrium and kinetic studies of copper(II) removal [24] Ho, Y., McKay, G. (1998) A comparison of chemisorption ki- by three species of dead fungal biomasses. Journal of Hazard- netic models applied to pollutant removal on various sorbents. ous Materials, 165(1–3), 469-474. Process safety and environmental protection: transactions of the Institution of Chemical Engineers, part B, 76(4), 332-340. [6] Özer A., Tümen F. (2003) Cd(II) Adsorption from aqueous solution by actived carbon from sugar beet pulp imprenated with [25] Fu, Y., Viraraghavan, T. (2000) Removal of a dye from an phosphoric acid, Fresenius Environmental Bulletin, 12(9), aqueous solution by the fungus Aspergillus niger. Water Qual- 1050-1058. ity Research Journal of Canada, 35(1), 95-111 [7] Hammaini, A., González, F., Ballester, A., Blázquez, M.L., [26] Lesmana, S.O., Febriana, N., Soetaredjo, F.E., Sunarso, J., Muñoz, J.A. (2007) Biosorption of heavy metals by activated Ismadji, S. (2009) Studies on potential applications of biomass sludge and their desorption characteristics. Journal of Environ- for the separation of heavy metals from water and wastewater. mental Management, 84(4), 419-426. Biochemical Engineering Journal, 44(1), 19-41.

309 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

[27] Srivastava, P., Hasan, S.H. (2011) Biomass of Mucor heimalis for the biosorption of cadmium from aqueous solutions: equilibrium and kinetic studies. BioResources, 6(4), 3656-3675.

[28] El-Sayed, M.T. (2012) The use of Saccharomyces cerevisiae for removing cadmium (II) from aqueous waste solutions. Af- rican Journal of Microbiology Research, 6(41), 6900-6910.

[29] Aksu, Z. (2001) Biosorption of reactive dyes by dried activated sludge: equilibrium and kinetic modelling. Biochemical Engineering Journal, 7(1), 79-84.

[30] Özer, A., Özer, D., Dursun, G., Bulak, S. (1999) Cadmium(II) adsorption on Cladophora crispata in batch stirred reactors in series. Waste Management, 19(3), 233-240.

Received: April 03, 2014 Accepted: July 30, 2014

CORRESPONDING AUTHOR

Sureyya Altin Bülent Ecevit University Department of Environmental Engineering Zonguldak, 67100 TURKEY

Phone: +90-372-257-4010/1565 E-mail: [email protected]

FEB/ Vol 24/ No 1b/ 2015 – pages 302 - 310

310 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

TOXICITY OF MOSQUITO LARVICIDES ON NON-TARGET MOSQUITO PREDATOR INSECT, BACKSWIMMER (Notonecta sp.)

Onder Ser1,* and Huseyin Cetin2

1Malaria Control Unit of the Infectious Disease Division of Antalya Public Health Directory, Antalya, Turkey 2Akdeniz University, Faculty of Science, Department of Biology, Antalya, Turkey

ABSTRACT Integrated Pest Management Programs, that use biological, physical (mechanical) and cultural methods in The toxicity of some mosquito larvicides (Bacillus combination with chemicals, are conducted to control mos- thuringiensis ssp. israelensis, spinosad, diflubenzuron and quito populations. The most effective control methods are pyriproxyfen) on mosquito predator insect, backswimmer those targeted against the larval stage of the life cycle. Ef- (Notonecta sp.) was investigated by exposing the predator fective predator organisms feed on mosquito larvae and pu- to doses recommended by World Health Organization in pae live in breeding areas of mosquitoes. Some fish species laboratory conditions. At least ten nymphs (3rd and 4th in- (Gambusia affinis S. F. Baird & Girard) and insects (im- stars) of backswimmers were released in polyethylene mature dragonflies and backswimmers) can help to reduce drums containing tap water treated with test doses by using the number of larvae in breeding sites if initial larval num- a fish net. For testing each level of dose, four replicates bers are not high [4, 5]. were used and one control group was set up as negative Products containing synthetic pyrethroids (permethrin, (only tap water). The number of dead nymphs was counted deltamethrin), microbial compounds (Bacillus thurin- in one day intervals for the 7 days exposure period, during giensis ssp. israelensis (Bti), Bacillus sphaericus, spi- which the nymphs were fed with non-larvicide-exposed nosad) and insect growth regulators (diflubenzuron, mosquito larvae and pupae. According to the results, the pyriproxyfen) are currently available in many countries four larvicides showed no or low toxic effect to Notonecta [6]. The application of these chemical and biological insec- sp. nymphs in 7-day exposure time except for the highest ticides to adult or larval habitats can prove effective in con- doses of spinosad (500 g ai/ha, 68.75% mortality) and trolling the mosquito populations [7-9]. However, most of diflubenzuron (100 g ai/ha, 44.32% mortality). these insecticides are also likely to be toxic to non-target organisms [10-12]. If the predators are unintentionally destroyed as a result of these applications, the number of KEYWORDS: mosquitoes could increase, proving counterproductive to Larvicide, Mosquito, Notonecta, Toxicity mosquito control.

Backswimmers have been recognized as the most promising predators of mosquito larvae. Notonecta is a ge- 1. INTRODUCTION nus of these insects in the family Notonectidae [13]. When backswimmer resting at the surface of water, its body is There are over 3500 species of mosquitoes (Diptera: typically tilted with the head downward. They are highly Culicidae) in the world, and at least 50 of them have been effective predators that because their nymphs feed on small documented in Turkey [1, 2]. These insects are distributed invertebrates, including mosquitoes. In addition, density of very widely in all zoogeographical regions and aquatic Notonecta or concentration of its kairomone affects ovipo- habitats. They are vectors of many important infectious sition behavior of many mosquito species [14-16]. For in- diseases, such as malaria, West Nile Virus infections, yel- stance, Aldemir and Bosgelmez [17] reported that No- low fever and dengue fever [2]. Over one million people tonecta viridis Delcourt can consume more than twenty worldwide die from mosquito-borne diseases every year. mosquito larvae per day in semi-field conditions. Blaustein Therefore, mosquito control is a vital public-health meas- et al. [18] showed that Notonecta maculata Fabricius had a ure throughout the world [3]. very large negative impact on populations of Culiseta longiareolata Macquart (Diptera: Culicidae) in natural lo- * Corresponding author cal pools in the Negev Desert, Israel. The efficacy of Bti

311 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

and the backswimmer Notonecta irrorata Uhler were eval- WHO Pesticide Evaluation Scheme (WHOPES) were ap- uated both individually and in combination to control mos- plied with hand sprayer to the water surface [6]. The ap- quito larvae in plastic containers in Monterrey, Mexico by plied doses ranged between the WHO recommended mini- Neri-Barbosa et al. [19]. These authors concluded that the mum and maximum rates per hectare. For the test, four rep- combined strategy proved to be the most effective one. licates were used and one control group (four replicates) was set up as negative (tap water). The numbers of dead Since little is known about the toxicity of some mos- nymphs was counted after every day for 10 days, during quito larvicides to backswimmers, the present investigation which three non-larvicide-exposed 3rd and 4th instars Culex was carried out with four mosquito larvicides (Bti, spi- sp. mosquito larvae or pupae/per nymph were given to the nosad, diflubenzuron and pyriproxyfen) the non-target or- nymphs. Nymphs were considered dead when they failed ganism; Notonecta sp. in laboratory conditions. to move after probing with a needle to the posterior region of abdomen. Moribund nymphs were those incapable of rising to the surface or of showing the characteristic diving 2. MATERIALS AND METHODS reaction of live nymphs when the water was disturbed.

2.1 Insects 2.3 Statistical analysis Third and fourth instar nymphs of Notonecta sp. were Corrected mortality was calculated using Abbott's for- used in the toxicity tests originated from Cakirlar, Kon- mula when mortality rate in the control was between 5- yaalti, Antalya. These were collected several days prior to 20% [20]. Differences between controls and treatments at the experiment from greenhouse irrigation pools between given exposure times and application doses were compared April and June 2013. At their arrival in the laboratory, the by using one-way ANOVA and Duncan’s multiple range insects were transferred to an aquarium and acclimated to test [21]. test water for at least 24 h before the start of the tests. The nymphs were reared at 12 h Light: 12 h Dark photoperiod, 60 ± 10% Relative Humidity, and 26 ± 2 °C in an insectary 3. RESULTS AND DISCUSSION in the Department of Biology, Faculty of Science, Akdeniz University. In this research, we compared the toxicity of four mosquito larvicides (Bti, spinosad, diflubenzuron and pyri- 2.2 Toxicity assays proxyfen) to non-target organism Notonecta sp. in labora- Water-dispersible granule formulation of Bti (Bti sero- tory conditions. Percentage mortality of the nymphs exposed type H14, 3000 ITU/mg), pyriproxyfen (Emulsion concen- to larvicides in the laboratory conditions are shown in Tables trate, 200 g ai/L), diflubenzuron (Suspension concentrate, 1-4. Results of this research indicated that generally all in- 150 g ai/L) and spinosad (Suspension concentrate, 120 g secticides had no significant or low toxic effects to No- ai/L) were used in toxicity assays. tonecta sp. at the lower doses recommended by the WHO.

At least ten nymphs (3rd and 4th instars) were released Seven days after treatment of insect growth regulators; in a 15 L polyethylene drum containing 12 L tap water (wa- the highest doses of pyriproxyfen (50 g ai/ha) and difluben- ter surface area; 660 cm2, water deep 25 cm) with the help zuron (100 g ai/ha) resulted in 31.26% and 44.32% mortal- of a fish net. The larvicide doses recommended by the ity of backswimmers, respectively.

TABLE 1 - Toxicity of Diflubenzuron on Notonecta sp. nymphs in laboratory conditions (Corrected mean % mortality ± Standard Error)

Doses (g active ingredient/ha) Days Control 25 50 75 100 1 0.0 ± 0.00 0.00 ± 0.00 2.50 ± 2.50 2.50 ± 2.50 2.27 ± 2.27 axAy aA aA aA aA 2 4.16 ± 2.49 7.08 ± 4.73 5.00 ± 2.88 6.92 ± 2.37 7.27 ± 2.43 aA aA aA aA aA 3 4.16 ± 2.49 7.08 ± 4.73 14.54 ± 2.65 6.92 ± 2.37 12.04 ± 4.56 aA aA bcA aA abA 4 13.33 ± 4.71 2.88 ± 1.84 5.25 ± 1.81 4.44 ± 1.94 18.36 ± 13.35 bA aA aA aA abcA 5 18.33 ± 1.67 2.54 ± 1.92 4.26 ± 2.22 2.98 ± 0.94 27.64 ± 10.92 bAB aA aA aA abcB 6 18.33 ± 1.67 11.73 ± 6.41 10.11 ± 4.92 8.39 ± 3.67 39.61 ± 9.44 bA aA abA aA bcB 7 19.99 ± 0.00 13.54 ± 5.20 20.74 ± 2.94 18.27 ± 3.35 44.32 ± 11.37 bA aA cA bA cB x Followed by the same lower case letters within a column are not significantly different (DMRT, P < 0.05) y Followed by the same capital letters at the same dose (row) are not significantly different (DMRT, P < 0.05)

312 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

TABLE 2 - Toxicity of Pyriproxyfen on Notonecta sp. nymphs in laboratory conditions (Corrected mean % mortality ± Standard Error)

Doses (g active ingredient/ha) Days Control 10 20 35 50 1 0.00 ± 0.00 5.00 ± 2.88 2.50 ± 2.50 2.50 ± 2.50 5.00 ± 2.88 axAy aA aA aA aA 2 4.16 ± 2.49 7.50 ± 4.78 5.00 ± 2.88 10.00 ± 4.08 7.50 ± 4.78 aA aA aA abcA aA 3 4.16 ± 2.49 7.50 ± 4.78 20.00 ± 4.08 10.00 ± 4.08 12.50 ± 7.50 aA aA abA abcA aA 4 13.33 ± 4.71 15.38 ± 10.05 16.34 ± 2.88 5.77 ± 1.92 14.42 ± 4.80 bA aA abA abA aA 5 18.33 ± 1.67 13.77 ± 8.91 17.35 ± 7.70 19.90 ± 8.30 11.22 ± 3.06 bA aA abA abcA aA 6 18.33 ± 1.67 16.83 ± 11.82 29.59 ± 7.70 26.53 ± 8.65 14.29 ± 0.00 bA aA bA bcA aA 7 19.99 ± 0.00 21.88 ± 10.67 28.13 ± 7.86 31.26 ± 10.82 31.26 ± 3.60 bA aA bA cA bA x Followed by the same lower case letters within a column are not significantly different (DMRT, P < 0.05) y Followed by the same capital letters at the same dose (row) are not significantly different (DMRT, P < 0.05)

TABLE 3 - Toxicity of Bacillus thuringiensis ssp. israelensis on Notonecta sp. nymphs in laboratory conditions (Corrected mean % mortality ± Standard Error)

Doses (g formulation/ha) Days Control 125 200 300 400 750 1 0.00 ± 0.00 0.00 ± 0.00 2.50 ± 2.50 4.58 ± 2.66 2.50 ± 2.50 2.50 ± 2.50 axAy aA aA aA abA abA 2 4.16 ± 2.49 2.50 ± 2.50 6.66 ± 4.08 11.24 ± 5.15 5.00 ± 5.00 7.50 ± 2.50 aA aA aA aA abcA abA 3 4.16 ± 2.49 5.00 ± 5.00 9.16 ± 3.43 13.33 ± 7.06 7.27 ± 4.75 10.00 ± 0.00 aA aA aA aA abcA bcA 4 13.33 ± 4.71 4.80 ± 4.80 3.36 ± 3.36 10.57 ± 6.15 1.92 ± 1.92 3.85 ± 2.22 bA aA aA aA aA abA 5 18.33 ± 1.67 6.63 ± 6.63 2.54 ± 1.92 9.18 ± 4.20 6.82 ± 3.44 1.53 ± 0.51 bB aAB aA aAB abcAB aA 6 18.33 ± 1.67 7.65 ± 6.31 2.54 ± 1.92 9.18 ± 4.20 18.50 ± 7.30 8.16 ± 3.53 bB aAB aA aAB bcB abAB 7 19.99 ± 0.00 6.25 ± 6.25 7.81 ± 5.91 13.54 ± 5.20 20.46 ± 7.00 15.63 ± 3.12 bA aA aA aA cA cA x Followed by the same lower case letters within a column are not significantly different (DMRT, P < 0.05) y Followed by the same capital letters at the same dose (row) are not significantly different (DMRT, P < 0.05)

TABLE 4 - Toxicity of Spinosad on Notonecta sp. nymphs in laboratory conditions (Corrected mean % mortality ± Standard Error)

Doses (g active ingredient/ha) Days Control 20 50 100 200 250 500 1 0.00 ± 0.00 0.00 ± 0.00 0.00 ± 0.00 5.00 ± 5.00 2.50 ± 2.50 2.50 ± 2.50 2.50 ± 2.50 axAy aA aA aA aA aA aA 2 4.16 ± 2.49 7.50 ± 4.78 5.00 ± 2.88 7.50 ± 7.50 5.00 ± 2.88 5.00 ± 2.88 27.50 ± 13.76 aA aAB abA aAB aA aA abB 3 4.16 ± 2.49 7.50 ± 4.78 12.50 ± 2.50 10.00 ± 7.07 12.50 ± 4.78 7.50 ± 4.78 47.50 ± 14.93 aA aA bA aA aA aA abB 4 13.33 ± 4.71 3.85 ± 2.22 3.85 ± 2.22 4.80 ± 4.80 3.85 ± 2.22 10.58 ± 2.88 59.61 ± 19.13 bA aA abA aA aA aA bB 5 18.33 ± 1.67 4.08 ± 3.43 4.59 ± 3.26 13.77 ± 5.42 1.53 ± 0.51 14.28 ± 12.24 60.20 ± 17.58 bA aA abA aA aA aA bB 6 18.33 ± 1.67 7.14 ± 6.48 4.59 ± 3.26 13.77 ± 5.42 7.65 ± 3.85 23.47 ± 17.58 63,26 ± 19.36 bA aA abA aA aA aA bB 7 19.99 ± 0.00 6.25 ± 6.25 9.38 ± 5.98 18.75 ± 6.25 12.51 ± 7.21 31.25 ± 18.74 68.75 ± 14.87 bA aA abA aA aA aA bB x Followed by the same lower case letters within a column are not significantly different (DMRT, P < 0.05) y Followed by the same capital letters at the same dose (row) are not significantly different (DMRT, P < 0.05)

313 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

The corrected percentage mortality of the nymphs at the and size structure of the D. pulex population was showed lowest dose of pyri-proxyfen (10 g ai/ha) was 21.88% for by Duchet et al [28]. They reported that both survival and the seven days exposure period, while the corrected per- size structure were affected. However, at the lowest con- centage mortality of backswimmers at the lowest dose of centration (8 ppm), population recovered after the first diflubenzuron (25 g ai/ha) was 13.54% for the same exposure week. In microcosms treated with Bti, the abundance of D. period in laboratory conditions (Tables 1 and 2). pulex was not affected but the size structure of the population changed after 21 days. When Notonecta sp. was exposed to all doses (125-750 g formulation/ha) of Bti there were no significant mortalities The toxicity of formulations of eleven synthetic insec- compared with the control group after 7 days exposure. Bti ticides used in desert locust control was monitored on the at the dose of 400 and 750 g formulation/ha, caused 20.46% backswimmer, Anisops sardeus Herrich-Schaeffer adults. and 15.63% corrected mortality of test insects, respectively Low toxic effects of the insect growth regulators; difluben- (Table 3). Spinosad was found statistically non-toxic to zuron, teflubenzuron, and triflumuron were reported [29]. Notonecta sp. at the doses of 20, 50, 100, 200, 250 g ai/ha. The use of lambda-cyhalothrin and fipronil in pest manage- However, the highest dose of spinosad (500 g ai/ha) was ment system against the rice water weevil, Lissorhoptrus found toxic to Notonecta sp. and mean mortality was rec- oryzophilus Kuschel in Arkansas rice have adverse effects orded as 68.75% after 7 days exposure (Table 4). on some nontarget insects, Tropisternus lateralis Say and N. indica. Fipronil was found less harmful to both preda- Mortality in the controls was lower than 5% within 72 h. tors than lambda-cyhalothrin [30]. The toxicity of neem oil Average mortality in the controls after three days of expo- to Notonecta sp. was reported by Anjaneyulu et al. [31]. sure was 4.16%. Between the 4th and 7th day control mortalities were 13.33 to 19.99% (Table 1-4). After 8 days, Toxicity to backswimmers of commonly used organo- more than 20% mortality was recorded in the control group phosphorus (OP) insecticides; DDVP (0.05-0.3 ppm), mal- and the test was stopped. athion (0.1-0.4 ppm), fenitrothion (0.01-0.05 ppm) and tri- chlorfon (0.8-1 ppm) was reported by Perschbacher and Bti products are toxic to mosquitoes, black flies and Sarkar [32]. Results of field test in this research showed certain midges [9, 22]. Many studies have examined the ef- that the four tested insecticides had toxic effects on nymphs fect of Bti toxins on non-target organisms in the laboratory, of backswimmers. One hundred percent mortality was de- under semi-field and field conditions. Notonecta indica L. tected only at the highest concentration of tested insecti- may have died from eating larvae that had ingested Bti tox- cides assayed on the nymphs of Notonecta sp. OP insecti- ins. However, others have reported that such exposure does cides are active against a wide range of public health pests not affect backswimmers [23]. In our experiments, the Bti but the use of them against public health pests is prohibited formulation as bioinsecticide showed no significant tox- by the Ministry of Health of Turkey after 2009 and many icity on Notonecta sp. (Table 1). The toxicity of Teknar countries in Europe. For this reason, OP insecticides have HP-D, an improved biolarvicidal formulation of Bti, to G. not been used in this research. affinis (larvivorous fish), Notonecta sp. and Diplonychus indicus Venkatesan & Rao (water bugs) was evaluated by In conclusion, many researchers have examined the ef- Gunasekaran et al. [24] in laboratory conditions. This for- fects of mosquito control agents (adulticides and larvicides) mulation was found non-toxic to G. affinis and also both on non-target organisms in the laboratory and the field con- water bugs that fed on the surviving mosquito larvae ex- ditions. Some insecticide sprays applied to control larval posed to the sub-lethal doses of the same formulation were mosquito populations may have negative effects on non- safe without having any mortality. Lagadic et al. [25] re- target insects and related organisms. However, according ported that long-term use of Bti formulation (VectoBac to the literature survey, up to date, effects of spinosad and WG) in French Atlantic coastal wetlands had no influence pyriproxyfen on backswimmers have not been studied. on the temporal evolution of the taxonomic structure and Backswimmers showed differences in sensitivity to the lar- taxa abundance of non-target aquatic invertebrate commu- vicides tested. We found that, lowest or median recommen- nities. dation doses of WHO recommended mosquito larvicides that used in this research were non-toxic to backswimmers. Toxic effects of two mosquito larvicides (Bti and spi- However, additional field and laboratory studies should be nosad) on two non target species, Daphnia pulex and D. made to determine lethal and sub-lethal toxic effects of manga, were evaluated by Duchet et al. [26] under labora- mosquito larvicides on backswimmers. tory and field microcosm exposure conditions. They reported that D. magna was more sensitive than D. pulex to spinosad. The results confirmed the significant negative effect of spinosad on survival, mean time at death, and fe- ACKNOWLEDGMENTS cundity as compared to controls and Bti-treated groups. In other study, spinosad caused a sharp decrease of D. magna The authors are grateful to the Scientiﬁc Projects Ad- abundance within the first two days following treatment ministration Unit of Akdeniz University (Antalya, Turkey) whereas Bti had no effect [27]. The impact of spinosad (8, for financial support. 17 and 33 ppm) and Bti (0.16 and 0.50 ppm) on abundance

314 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

The authors have declared no conflict of interest. [17] Aldemir, A., Boşgelmez, A. (2004). Predation by Alburnus orontis (Cypriniformes: Cyprinidae) and Notonecta viridis (Hemiptera: Notonectidae) on Mosquito Larvae. Turkish J. Aqua. Life 2, 1-10 (in Turkish).

REFERENCES [18] Blaustein, L., Kotler, B.P., and Ward, D. (1995). Direct and indirect effects of a predatory backswimmer (Notonecta maculata) on community structure of desert temporary pools. Ecol. [1] Ramsdale, C.D., Alten, B., Caglar, S.S., and Ozer, N. (2001). Entomol. 20, 311-318. A revised, annotated checklist of the mosquitoes (Diptera, Cu- licidae) of Turkey. European Mosq. Bull. 9, 18–28. [19] Neri-Barbosa, J.F., Quiroz-Martinez, H., Rodriguez-Tovar, M.L., Tejada, L.O., and Badii, M.H. (1997). Use of Bactimos [2] Becker, N., Petric, D., Zgomba, M., Boaseâ, C., Dahl, C., Ma- (R) briquets (Bti formulation) combined with the backswim- donâ, M., and Kaiser, A. (2010). Mosquitoes and Their Con- mer Notonecta irrorata (Hemiptera: Notonectidae) for control trol. Second Edition, Springer, Heidelberg Dordrecht London of mosquito larvae. J. Am. Mosq. Control Assoc. 13, 87-89. New-York, 577 p. [20] Abbott, W.S. (1925). A method for computing the effective- [3] W.H.O. (2013). World Malaria Report 2013, 284 p. ness of an insecticide. J. Econ. Entomol. 18, 265-267. [4] Rodriguez-Castro, V.A., Quiroz-Martinez, H., Solis-Roja, C., and Tejada L.O. (2006). Mass rearing and egg release of [21] S.P.S.S. (1999). SPSS for Windows, version 15.0. SPSS, Chi- Buenoa scimitra bare as biocontrol of larval Culex quinque- cago, IL. fasciatus. J. Am. Mosq. Control Assoc. 22, 123-125. [22] Molloy, D.P. (1992). Impact of the black fly (Diptera: Simuli- [5] Chandra, G., Bhattacharjee, I., Chatterjee, S.N., and Ghosh, A. idae) control agent Bacillus thuringienses var. israelensis on (2008). Mosquito control by larvivorous fish. Indian J. Med. chironomids (Diptera: Chironomidae) and other nontarget in- Res. 127, 13-27. sects: results of ten field trials. J. Am. Mosq. Control Assoc. 8, 24-31. [6] W.H.O. (2006). Pesticides and their application for the control of vectors and pests of public health importance. Sixth Edition. [23] Poulin, B. (2012). Indirect effects of bioinsecticides on the Department of Control of Neglected Tropical Diseases, WHO nontarget fauna: The Camargue experiment calls for future re- Pesticide Evaluation Scheme (WHOPES), 125 p. search. Acta Oecol. 44, 28-32.

[7] Cetin, H., Yanikoglu, A., and Cilek, J.E. (2005). Evaluation of [24] Gunasekaran, K., Doss, P.S.B., and Vaidyanathan, K. (2004). the naturally-derived insecticide spinosad against Culex Laboratory and field evaluation of Teknar HP-D, a biolarvi- pipiens L. (Diptera: Culicidae) larvae in septic tank water in cidal formulation of Bacillus thuringiensis ssp israelensis, Antalya, Turkey. J. Vector Ecol. 30, 151–154. against mosquito vectors. Acta Trop. 92, 109-118.

[8] Cetin, H., Yanikoglu, A., and Cilek, J.E. (2006). Efficacy of [25] Lagadic, L., Roucaute, M., and Caquet, Th. (2014). Bti sprays diflubenzuron, a chitin synthesis inhibitor, against Culex do not adversely affect non-target aquatic invertebrates in pipiens larvae in septic tank water. J. Am. Mosq. Control As- French Atlantic coastal wetlands. J. Appl. Ecol. 51, 102-113. soc. 22, 343–345. [26] Duchet, C., Coutellec, M.A., Franquet, E., Lagneau, Ch., and [9] Cetin, H., DeChant, P., and Yanikoglu, A. (2007). Field trials Lagadic, L. (2010). Population-level effects of spinosad and with tank mixtures of Bacillus thuringiensis subsp. israelensis Bacillus thuringiensis israelensis in Daphnia pulex and Daph- and Bacillus sphaericus formulations against Culex pipiens L. nia magna: comparison of laboratory and field microcosm ex- (Diptera: Culicidae) larvae in septic tank water in Antalya, posure conditions. Ecotoxicol. 19, 1224-1237. Turkey. J. Am. Mosq. Control Assoc. 23, 161-165. [27] Duchet, C., Caquet, Th., Franquet, E., Lagneau, Ch., and [10] Federle, P.F., Collins, W.J. (1976). Insecticide toxicity to three Lagadic, L. (2010). Influence of environmental factors on the insects from Ohio ponds. Ohio J. Sci. 76, 19-24. response of a natural population of Daphnia magna (Crusta- cea: Cladocera) to spinosad and Bacillus thuringiensis is- [11] Karasu Benli, C., Selvi, M., Sepici Dinçel, A., Sarıkaya, R., raelensis in Mediterranean coastal wetlands. Environ. Poll. Yıldırım M.Z., Özkul, A., and Erkoc, F. (2009). Acute toxicity 158, 1825-1833. of beta-cypermethrin on Nile tilapia, Oreochromis niloticus L., fingerlings. J. Environ. Prot. Ecol. 10, 104-109. [28] Duchet, C., Franquet, E., Caquet, Th., Larroque, M., Lagneau, Ch., and Lagadic, L. (2008). Effects of spinosad and Bacillus [12] Stevens, M.M., Burdett, A.S., Mudford, E.M., Helliwell, S., thuringiensis israelensis on a natural population of Daphnia and Doran, G. (2011). The acute toxicity of fipronil to two pulex (Crustacea: Cladocera) in field microcosms. Chemo- non-target invertebrates associated with mosquito breeding sphere, 74, 70-77. sites in Australia. Acta Trop. 117, 125-130. [29] Lahr, J., Badji A., Marquenie, S., Schuiling, E., Ndour, K.B., [13] Hungerford, H.B. (1933). The Genus Notonecta of the World Diallo, A.O., and Everts, J.W. (2001). Acute toxicity of locust (Notonectidae-Hemiptera). Univ. Kans. Sci. Bull. 21, 5-195. insecticides to two indigenous invertebrates from sahelian [14] Eitam, A., Blaustein, L. (2004). Oviposition habitat selection temporary ponds. Ecotoxicol. Environ. Saf. 48, 66-75. by mosquitoes in response to predator (Notonecta maculata) density. Physiol. Entomol. 29, 188-191. [30] Dennett, J.A., Bernhardt, J.L., and Meisch, M.V. (2003). Ef- fects of fipronil and lambda-cyhalothrin against larval Anoph- [15] Silberbush, A., Markman, S., Lewinsohn, E., Bar, E.E., Co- eles quadrimaculatus and nontarget aquatic mosquito preda- hen, J.E., and Blaustein, L. (2010). Predator-released hydro- tors in Arkansas small rice plots. J. Am. Mosq. Control As- carbons repel oviposition by a mosquito. Ecol. Lett. 13, 1129- soc. 19, 172-174. 1138. [31] Anjaneyulu, G.V.S.R., Nayak, V., Rao, U.D.V.P.P., Sateesh, [16] Silberbush, A., Blaustein, L. (2011). Mosquito females quan- T.V.R., and Mishra, K.D. (1999). Acute toxicity of neem oil tify risk of predation to their progeny when selecting an ovi- to aquatic hemipteran predatory insect Notonecta sp. Environ. position site. Funct. Ecol. 25, 1091-1095. Ecol. 17, 57-61.

315 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

[32] Perschbacher, P.W., Sarkar, J. (1989). Toxicity of selected or- ganophosphorus insecticides to the backswimmer, Notonecta sp. Asian Fish. Sci. 2, 265-268.

Received: April 14, 2014 Revised: June 19, 2014 Accepted: June 25, 2014

CORRESPONDING AUTHOR

Onder Ser Malaria Control Unit Infectious Disease Division Antalya Public Health Directory Antalya TURKEY

Phone: +90-505-3745649 Fax: +90-242 -2278911 E-mail: [email protected]

FEB/ Vol 24/ No 1b/ 2015 – pages 311 - 316

316 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

AIRBORNE MICROFUNGUS FLORA DETERMINED IN THE DIFFERENT UNITS OF THE DEPARTMENT OF TULAY AKTAS ONCOLOGY HOSPITAL, EGE UNIVERSITY

Ozlem Abaci Gunyar1,*, Alev Haliki-Uztan1, Mustafa Ates1, Aysegul Yoltas1 and Erdem Goker2

1Ege University, Faculty of Science, Department of Biology, Basic & Industrial Microbiology Section, 35100 Bornova-İzmir, Turkey 2Ege University, Medical School, Department of Chest Disease, 35100 Bornova-İzmir, Turkey

ABSTRACT 1. INTRODUCTION

Recently, there has been a dramatic increase in the Immune suppressed patients with impaired body re- prevalence of nosocomial fungal infections. Especially sistance comprise the biggest risk group for nosocomial in- with the increase in the number of immunosuppressed pa- fections. Especially in patients receiving chemotherapy due tients, rare fungal pathogens which were difficult to detect to their hematologic malignancies or having undergone in the past are now frequently isolated. In this study, the transplantation, major opportunistic fungal infection is inva- level of endogenous airborne fungal concentrations and the sive Aspergillus infections which usually develop as a result types of fungi were determined at Ege University, Faculty of spore inhalation. High mortality rates in immune sup- of Medicine, Tulay Aktas Oncology Hospital. For this pur- pressed individuals with hematological disorders (95-100%) pose, air samples were collected at 34 different locations and having undergone bone marrow transplant (84%), and monthly during six months with the Merck MAS 100 air sam- high costs of their treatments have drawn attention to studies pler. As a result, a total of 5590 fungal colonies in 204 petri conducted on the prevention of nosocomial infections dishes were counted. Thirty-two fungal species belonging to caused by Aspergillus [1-4]. In recent years, in addition to 10 genera were identified. The most prevalent genera were Candida and Aspergillus species, fungi such as Mucor, Rhi- the common moulds; Aspergillus (3.91±1.01), Penicillium zopus, Fusarium, Alternaria, Curvularia, and Trichosporon (16.08±3.10), Cladosporium (40.82±4.15) and Alternaria species which were formerly considered as contaminators or (7.51±1.82). The results revealed that the most common harmless colonies in hospitals have been found to cause in- genus was Cladosporium. Fungal concentration levels in vasive infections in immune suppressed patients [5]. different parts of the stem cell transplantation department ranged from 0 to 340 cfu/m3. In other units of the hospital, Monitoring the concentration of airborne fungal spores fungal load was found to be between 0 and 45600 cfu/m3. and being aware of the incidence of microbiota in hospitals When the fungal density of the stem cell transplantation play an important role in recognizing the types of infection department and that of the other units were compared, fun- likely to develop and in understanding the problems they gal density of the stem cell transplantation department was might cause. As a result, it would be possible to take ap- found to be significantly lower (p = 0.000 <0.05). The propriate protective measures and thus to decrease the in- highest values in our study were determined during the 5th cidence of such mycoses. These measures include regular month of this study, because natural gas pipelines were in- monitorization of fungal spore load in the environment stalled around the hospital in this month. In this study, we around the patient during his / her stay in the hospital and aimed to emphasize the importance of monitoring the pres- the control of exogenous sources and potential exogenous ence of airborne fungal flora, particularly Aspergillus spp, sources [6]. Considering all these reasons, in our study, we in the departments of a hospital where patients of high-risk aimed to determine the level of endogenous air-borne fungi groups are hospitalized. at Ege University, Faculty of Medicine, Tulay Aktas On- cology Hospital.

KEYWORDS: 2. MATERIALS AND METHODS fungi, fungus, microfungus, Izmir, hospital, biomass This study was carried out at the Ege University Hos- pital, Faculty of Medicine, Department of Internal Dis-

eases, Tülay Aktaş Oncology Hospital. The Ege University * Corresponding author Hospital is located in the center of Izmir.

317 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

Tulay Aktas Oncology Hospital has three floors. Sam- (SPSS Inc., Chicago, IL, USA). Statistical significance was pling was conducted at 34 different points: the ground floor defined as p < 0.05. (entrance, corridor, lounge, adult outpatient clinics, pediatric outpatient clinics, one-patient chemotherapy room, two-patient chemotherapy room, chemotherapy room 7, 3. RESULTS drug preparation part, tea room), 1st Floor – Adult Inpatient Services (corridor, interns’ room, physicians’ room, room In our study, the sampling procedure was carried out 102, room 105, Room 109, room 112, room 118, room each month between February and July for 6 months. Via- 121), 2nd Floor- pediatric inpatient service (Corridor, assis- ble fungi present in the air in different areas ofTülay Aktaş tant physicians’ room, room 205, room 209, room 216, Oncology hospital were studied. The number of fungal room 223, nursery class), stem cell transplantation depart- spores recovered from the air varied from month to month ment (stem cell transplantation unit 1, stem cell transplan- (Table 1). tation unit 3, the area between the units), drug preparation laboratories (pediatric and adult), 3rd Floor (classrooms, A total of 5590 fungal colonies in 204 petri dishes were corridor). counted. All colonies were identified to genus level and different colonies were isolated from each petri dish and 2.1 Air sampling then identified to species level. Thirty-two fungal species belonging to 10 genera were identified. Only one isolate Air sampling was carried out to determine microscopic was identified in genus level which belongs to the genus fungi in the air. During each monitoring period 100 lt. of Phoma. The most prevalent genera were the common air were taken from 34 sites. A Merck Mas 100 air sampler moulds; Aspergillus (3.91±1.01), Penicillium (16.08±3.10), was used for air sample collection. Samples were taken be- Cladosporium (40.82±4.15) and Alternaria (7.51±1.82) tween 11:00 a.m and 13:00 p.m. Air sampler was generally (Table 2). located in the center of the room away from open windows or doors at 1.40 m above floor level. When the units and fungal density were randomly compared, the difference was found to be significant (p = 2.2 Isolation methods 0.000 <0.05). The points with the lowest fungal concentra- Rose-Bengal chloramphenicol medium was adopted tion were the stem cell transplantation department (81.25 ± for the collection of fungi. The number of cfus was counted 15.96) and the pediatric inpatient service (130.24 ± 28.04). after 5 days of incubation at 25 C and cfus per cubic meter When the stem cell transplantation department was were calculated. compared with the other units in terms of fungal concen-

2.3 Identification tration, statistically significant differences were found between the stem cell transplantation department and the The spores of each colony were picked and placed un- ground floor (p = 0.003 <0.05), and between the stem cell der the high-resolution microscope for the identification of transplantation department and the adult inpatient unit (p = genus according to references available [7]. 0.000 <0.05). The comparison between the stem cell trans- Cultures were identified to species level using the fol- plantation department and the pediatric inpatient revealed lowing literature: Penicillium link species were identified no significant difference (p=0.422>0.05). The difference using colony diameters, macro- and micromorphology ac- between the stem cell transplantation department and drug cording to the standardized conditions in the monograph by preparation rooms was not significant either (p=0.062> Pitt. [8]. So, Czapek Yeast Extract Agar (CYA), Malt Ex- 0.05). tract Agar (MEA), % 25 Glycerol Nitrate(G25N) and Crea- When genus concentrations determined at regular in- tine Sucrose agar (CSN) were used for the cultivation of tervals were compared, the highest counts were as follows: Penicillium species. CYA culture was inoculated in triplicate genus Cladosporium at the 3rd month (p = 0.026 <0.05), and incubated at three different temperatures (5 C, 25 C, genus Penicillium at the 4th month (p = 0.006 <0.05), genus 37 C) for a period of 7 days in the dark during which the Cladosporium at the 5th month (p = 0.000 <0.05), genus colony diameters of all the Penicillium species were meas- Cladosporium at the 6th month (p = 0.004 <0.05) and genus ured. Cladosporium months at the 7th month (p = 0.002 <0.05). The identification of Aspergillus species was based on The results revealed that the most common genus was Klich [9]. The Cladosporium Link and Alternaria Nees co- Cladosporium. nidia were present and the other structures on the microscope slides were identified as described by Hasenekoğlu [10]. 4. DISCUSSION 2.4 Statistical analysis Time differences were tested for statistical signifi- Although many of the air-borne fungi are not patho- cance by Friedman test. Mann-Whitney U test was used to genic, they can cause sensitive or immune suppressed indi- compare units (intensive care and other units). All statisti- viduals to develop an illness. Due to the changes in host cal analyses were performed using SPSS 13.0 for Windows factors, saprophytic fungus species are encountered more

318 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

and more often as an opportunistic infection factor. Kal- plant pathogen was considered important since it caused kancı et al. isolated Valsa sordida from a patient with im- infections in humans for the first time and it was presented munodeficiency due to acute myeloid leukemia and they as a case report [11]. confirmed the diagnosis with molecular methods. This

TABLE 1 - Monthly counts of airborne fungi determined in Tülay Aktaş oncology hospital cfu/m3.

February March April May June July 1) The ground floor Entrance 70 200 200 1060 840 540 Corridor 60 80 370 1100 220 560 Lounge 50 80 200 1520 240 180 Adult outpatient clinics 60 30 130 1400 160 300 Pediatric outpatient clinics 0 570 250 1220 0 340 One-patient chemotherapy room 30 100 150 460 0 160 Two-patient chemotherapy room 70 20 20 760 440 140 Chemotherapy room 7 20 110 220 320 60 100 Drug preparation part 40 30 20 80 0 80 Tea room 80 30 400 2560 400 300 2) 1st Floor – Adult Inpatient Services Corridor 140 60 1340 1340 320 240 Interns’ room 40 50 520 580 660 80 Physicians’ room 120 190 560 100 380 200 Room 102 110 50 370 140 100 280 Room 105 110 50 380 320 220 120 February March April May June July Room 109 20 40 850 100 480 200 Room 112 40 80 540 80 180 160 Room 118 70 30 500 580 120 120 Room 121 60 150 360 60 20 60 3) 2nd Floor- pediatric inpatient service Corridor 30 100 240 1080 240 180 Assistant physicians’ room 10 50 210 180 100 20 Room 205 0 20 120 340 400 60 Room 209 0 30 70 160 200 340 Room 216 10 50 70 60 240 120 Room 223 0 20 130 60 180 100 Nursery class 0 0 30 140 60 20 4) Stem cell transplantation department The area between the units 20 20 180 340 120 40 Stem cell transplantation unit 1 20 20 30 140 80 140 Stem cell transplantation unit 3 20 30 60 0 60 40 Stem cell transplantation unit 5 60 140 30 160 160 40 5) Drug preparation laboratories Pediatric drug prepation laborotary 50 50 260 2120 160 40 Adult drug prepation laborotary 70 70 420 4560 100 20 6) 3rd Floor Classrooms 10 20 50 160 260 120 Corridor 0 20 0 120 100 100

TABLE 2 - Distribution of the genera

Genus Minimum Maximum Mean Std. Error Alternaria 0.00 50.00 7.51 1.82 Aspergillus 0.00 24.39 3.91 1.01 Cladosporium 0.00 100.00 42.82 4.15 Penicillium 0.00 89.48 16.08 3.10 Other Fungi 0.00 80.00 15.13 2.86 Yeast 0.00 100.00 10.05 2.55

319 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

TABLE 3 - List of fungal species isolated from Ege University, Faculty of the Medicine, Tülay Aktaş Oncology Hospital, sampling time and sampling points.

Species Time (station) Aspergillus alliaceus Thom & Church, Manual of the Aspergilli: 244 (1945) 6(III), 12(IV), 15(III, V) A. flavus Link, Magazin der Gesellschaft Naturforschenden Freunde Berlin 3: 16 4(V,VI), 7(II,), 8(I,VI), 9(V), 14(IV), 22(II, IV), 29(VI) (1809) A. fumigatus Fresen., Beiträge zur Mykologie 3: 81 (1863) 17(V), 24(III), 26(I) A. niger Tiegh., Annales des Sciences Naturelles Botanique 8: 240 (1867) 16(I,II,VI), 17(I), 26(II,III, V, VI), 29(IV) A. parasiticus Speare, Bull. Div. Pathol. Physiol., Hawaiian Sugar Planters' Assoc. 16(I) Exp. Sta.: 38 (1912) A. restrictus G. Sm., J. Textile Res. Inst.: 115 (1931) 13(II), 19(II, VI), 20(IV,V), 24(III,VI), 25(V, VI) A. terreus Thom, American Journal of Botany 5 (2): 85 (1918) 34(V) A.versicolor (Vuill.) Tirab., Annali Bot.: 9 (1908) 13(V), Penicillium brevicompactum Dierckx, Annales de la Société Scientifique de Brux- 6(I), 8(IV), 29(II), 30(II), 31(II) elles 25 (1): 88 (1901) P. camemberti Thom, U.S.D.A. Bureau of Animal Industry Bulletin 82: 33 (1906) 5(III), 13(I,II), 31(I), P. citreonigrum Dierckx, Annales de la Société Scientifique de Bruxelles 25 (1): 13(VI), 14(VI) 86 (1901) P. chrysogenum Thom, U.S.D.A. Bureau of Animal Industry Bulletin 118: 58 1(I),1(III), 4(I), 4(III), 14(III, V), 16(I, II, III), 17(V), 18(II, (1910) III), 19(I, II, III), 20(I,III), 22(III, V), 23(III), 24(IV,V), 25(II, III, V, VI), 26(V), 27(II, III, V), 28(III), 29(III), 30(III), 31(III), 32(III), 33(V), 34(III) P. commune Thom, U.S.D.A. Bureau of Animal Industry Bulletin 118: 56 (1910) 31(V), 32(V), P. digitatum (Pers.) Sacc., Fung. Ital.: tab. 894 (1881) 6(1, ), 9(I,V) P. expansum Link, Magazin der Gesellschaft Naturforschenden Freunde Berlin 3: 3(III), 3(V), 4(III), 4(V), 5(II), 5(III), 5(V), 6(II, IV), 7(II, 17 (1809) III), 8(I, II), 9(IV), 10(I, II, III, IV), 11(II, IV), 12(III), 13(IV), 14(I, II, III), 15(II, III), 17(I, II, III), 21(III), 22(II), 23(II) P. glabrum (Wehmer) Westling, Arkiv før Botanik 11 (1): 131 (1911) 3(II), 4(VI), 7(III), 14(III), 16(VI), 19(V), P. hirsutum Dierckx, Ann. Soc. Sci. Bruxelles 25: 89 (1901) 13(III), 17(V) P. paxilli Bainier, Bulletin de la Société Mycologique de France 23: 94 (1907) 12(I), 21(I) P. solitum Westling, Arkiv før Botanik 11 (1): 65 (1911) 20(II) P. spinulosum Thom, U.S.D.A. Bureau of Animal Industry Bulletin 118: 76 9(I), 23 (IV) (1910) P. viridicatum Westling, Arkiv før Botanik 11 (1): 88 (1911) 24(III), 26(III) Alternaria alternata (Fr.) Keissl., Beihefte zum Botanischen Zentralblatt 29: 433 6(IV,VI), 10(III, IV), 19(III, IV) (1912) A. tenuissima (Nees) Wiltshire, Transactions of the British Mycological Society 18 1(I),1(IV),1(V),1(VI), 5(1, V,), 9(IV,V), 12(IV), 13(IV), (2): 157 (1933) 14(V), 16(IV), 17(IV), 18(V), 31(V). A.citri Ellis & N. Pierce, Botanical Gazette Crawfordsville 33 (3): 234 (1902) 3(I), 3(II), 3(IV), 3(V), 3(VI), 4(III), 4(IV), 4(VI), 7(IV, VI), 8(IV), 10(VI), 12(III), 13(V), 15(IV,V), 20(IV), 22(II, IV, V), 25(IV), 26(IV.V), 27(V, VI), 33(IV,VI), 34(II, IV,VI). Cladosporium herbarum (Pers.) Link, Magazin der Gesellschaft Naturforschenden 18(V) Freunde Berlin 8: 37 (1816) C. oxysporum Berk. & M.A. Curtis Botanical Journal of the Linnean Society 10: 2(I), 2 (II),2(III),2(IV),2(V),2(VI), 3(I), 3(II), 3(III), 3(IV), 362 (1869) 3(V), 3(VI), 4(1, II, III, IV, V, VI), 5(I, II, III, IV, V, VI), 6(1, II, III, IV, V, VI), 7(II, III, IV, VI), 8(I, III, IV, VI), 9(I, II, IV, V, VI), 10(II, III, IV, V, VI), 11(I, II, IV, VI), 12(I, II, III, IV), 13(I, II, III, IV, V, VI), 14(I, II, III, IV, V, VI),, 15(I, III, IV, V, VI), 16(I, II, III, IV, V, VI), 17(I, II, III, IV, V, VI), 18(II, III, IV, V, VI), 19(II, III, IV, V, VI), 20(III, IV, V, VI), 21(I, IV, V, VI), 22(II, III, IV, V, VI), 23(II, III, IV, V, VI), 24(II, III, IV, V, VI), 25(II, III, IV, V, VI), 26(II, III, IV, V, VI), 27(II, III, IV, V, VI), 28(III,IV,V,VI), 29(I, II, III, IV, V, VI), 30(I, II, III, V, VI), 31(III), 32(IV, V, VI), 33(II, IV, V, VI) Helminthosporium acaciae M.B. Ellis, Mycological Papers 82: 9 (1961) 3(VI), 5(VI), 6(VI), 11(III), 13(I,VI), 16(III, VI), 18(VI), 20(VI), 25(VI), 27(VI), 28(VI), 34(VI) Drechslera hawaiiensis (Bugnic.) Subram. & B.L. Jain, Current Science 35 (14): 6(III,VI), 7(VI), 10(VI), 13(V), 29(VI) 354 (1966) Sclerotinia sclerotiorum (Lib.) de Bary, Vergleichende Morphologie und Biologie 10(II), 11(III), 29(I), der Pilze Mycetozoen und Bacterien: 56 (1884) Trichoderma hamatum (Bonord.) Bainier, Bulletin de la Société Mycologique de 26(II), 33 (II), 34(II) France 22: 131 (1906) Rhizopus oryzae Went & Prins. Geerl., Verhandelingen Koninklijke Nederlandse 9(III), 18(V), 21(II), 22(V), 28(III, IV), 32(IV) Akademie van Wetenschappen Afdeling Natuurkunde 4: 16 (1895) Cunninghamella elegans Lendn., Bulletin de l´Herbier Boissier 7: 250 (1907) 9(IV) Phoma sp. 13(III)

320 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

Mycelia sterilia 1(I),1(III), 1(IV), 2(I), 2(III), 2(IV), 4(I,II, III, V, VI), 5(I, V), 6(1, II, III, IV, VI), 7(II), 8(I, III), 9(I,V), 10(I, II, III, VI), 11(I, VI), 12(I,II), 13(I, II, IV), 14(I, II, III, V, VI), 15(I, III, IV, V), 16(I, II, V), 17(I, III, IV, V), 19(I, II, IV, V), 20(I, IV, VI), 22(I, II, III), 23(I, II, IV, V), 24(III, IV, V), 25(II, V), 26(V), 27(V), 28(II, III), 29(I, III), 30(III), 31(IV), 32(III)34(V) Yeast 1(3)1(v), 2(II), 2(III), 2(IV), 2(V), 3(II, III, V, VI), 6(III, IV, V, VI), 7(II, IV), 8(II, III), 9(IV,VI), 11(I, IV), 13(I, II, IV, V, VI), 14(III, V), 15(I, V, VI), 16(V,VI), 17(I, II, VI), 18(I,IV, V, VI), 19(II, III,IV, VI), 20(IV, VI), 21(IV, VI), 22(III, V, VI), 24(IV, V), 25(V, VI), 26(V, VI), 28(I, III), 29(IV, VI), 30(VI), 31(I, V, VI), 34(I, III, IV, V, VI) Drug preparation laboratories: 1: Adult drug preparation laboratory, 2: Pediatric drug preparation laboratory, The ground floor: 3: Entrance, 4: Corridor, 5: Lounge, 6: Adult outpatient clinics, 7: Pediatric outpatient clinics, 8: One-patient chemotherapy room, 9: Two-patient chemotherapy room, 10: Chemotherapy room 7, 11: Drug preparation part, 12: Tea room, 1st Floor: 13: Adult Inpatient Services, 14: Interns’ room, 15: Physicians’ room, 16: Room 102, 17: Room 105, 18: Room 109, 19: Room 112, 20: Room 118, 21: Room 121, Pediatric inpatient service 22: Corridor, 2nd Floor: 23: 2nd Floor Assistant physicians’ room, 24: 2nd Floor Room 205, 25: 2nd Floor Room 209, 26: 2nd Floor Room 216, 27: 2nd Floor Room 223, Stem cell transplantation department: 28: The area between the units, 29: Stem cell transplantation unit 1, 30: Stem cell transplantation unit 3, 31: Stem cell transplantation unit 5, 32: Nursery class, 3rd Floor 33: Classrooms, 34: Corridor.

It is well known that the main cause of fungal infec- In the light of the results, it is clear that they were the best- tions in hospital environments is the inhalation of fungal protected areas of the hospitals. When the fungal density of particles. Alberti et al. followed their patients staying in the the stem cell transplantation department and that of the other hematology unit for 4 years, and determined a significant units were compared, fungal density of the stem cell trans- correlation between the surface fungal contamination level plantation department was found to be significantly lower. and the invasive nosocomial aspergillosis incidence in the Besides this relatively high number of contaminants in conventionally ventilated rooms [12]. the hospital air, the species spectrum is also important. As- Therefore, it is important to be aware of the microbiota pergillus species are very common in the environment. of the hospital indoor air because this enables us to identify They colonize on soil, leaves, live plants and grains. They the types of fungal infection and allergies caused by them. are also present in the unfiltered air of hospital settings, The study aimed to determine the concentration and diver- ventilation systems, contaminated dust spreading from the sity of different genera of airborne fungi in Ege University, construction areas of hospitals, floors and food [14]. Tülay Aktaş Oncology Hospital. Flowerpots in patient rooms can also host Aspergillus In recent years, the most widely used method to measure species. Aspergillus spores can easily spread to the envi- the spore concentrations is collecting samples with auto- ronment and therefore can lead to infection. The mortality matic air samplers. With these devices, a standard volume of rate due to systemic aspergillosis is higher compared with air samples can be collected and the number of spores can the mortality rate due to other systemic mycoses especially be measured quantitatively. Because of these advantages, in in the intensive care units [15]. Among the human patho- our study Merck Mas 100 air sampler was used. genic species of Aspergillus, A. fumigatus is the primary Of the places where the study was carried out, the stem causative agent of human infections, followed by A. flavus, cell transplantation department is of particular importance. A. terreus, A. niger, A. nidulans [16]. In our study, In the Fungal concentration levels in different parts of the stem cell stem cell transplantation department, genus A. niger and transplantation department ranged from 0 to 340 cfu/m3 (Ta- genus A. flavus were isolated once during the 4th and 6th ble 1). samplings respectively. In the literature, it is emphasized that the concentration Other potential pathogenic species of Aspergillus were of fungus spores in special care units should be less than not isolated in the stem cell transplantation department. Of that in ordinary patient care rooms (<100 cfu/m3) (1). In the Aspergillus species, Alliceus A., A. flavus, A. fumigatus, another study in the literature, it was suggested that sapro- A. niger, A. parasiticus, A. versicolor, A. reistrich were iso- phytes counts should be less than 15 cfu/m3 and concentra- lated from the other departments of the hospital. Of these tions of opportunistic fungal pathogens should be kept at species, while A. flavus and A. niger were isolated from the 0.1 cfu/m3 in the operating rooms and patient isolation other departments of the hospital more frequently, A. fu- rooms, which are supposed to be the cleanest sections of a migatus was isolated in each of the three patient rooms dur- hospital [13]. ing only one sampling period. The presence of A. fumigatus When the units and fungal density were randomly in patient rooms is most likely explained by its transmission compared, the difference was found to be significant (p = through visitors’ clothes carriage on the cloths of visitors. 0.000 <0.05). The points with the lowest fungal concentra- The existence of Aspergillus species in the environ- tion were the stem cell transplantation department (81.25 ± ment is an important risk factor for the development of nos- 15.96) and the pediatric inpatient service (130.24 ± 28.04). ocomial aspergillosis. Inhalation or the direct inoculation

321 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

of the tissues with spores is the most common way for the In many hospitals, large-scale construction works such infection to be transmitted. Lutz et al. conducted a study as renovation, expansion, demolition and excavation are in a tertiary care hospital to investigate whether the fungal carried out. infection agents developing in the wounds of patients hav- It should not be overlooked that opportunistic fungi-re- ing undergone surgery were related to fungal isolates isolated infections can be isolated especially from operating lated from the operating room settings, and they found that theaters , intensive care units and neonatal units where high- clinical and environmental isolates of A. fumigatus are ge- risk patients stay during these construction works [21 - 24] . netically identical strains [17]. During the 5th month of this study, natural gas pipe- In our study, in addition to Aspergillus species, Peni- lines were installed around the hospital. cillium species, known to cause respiratory diseases due to Therefore, the highest values in our study were deter- their allergenic characteristics and toxin-producing capa- mined in this month (Table 1). Compared with the data in bilities, were frequently isolated (16.08±3.11). When the the literature, high counts we obtained were consistent with relationship between invasive infections and Penicillium the factors which contributed to them. genus is considered, Talaromyces marneffei is mostly isolated as an invasive infectious agent; however, Ly- The number of the infections caused by genus Rhizo- ratzopoulos et al. reported that they also isolated invasive pus is higher particularly among individuals hospitalized in infections due to P. chrysogenum and P. decumbens [18]. intensive care units. Penicillium species identified in our study were P. brevi- This organism can be transmitted to a patient not only compactum, P. camemberti, P. citreonigrum, P. chryso- through patient-care equipment but also through the in- genum, P. commune, P. expansum, P. digitatum, P. gla- haled air [25]. brum, P. hirsitum, P. paxilli, P. solitum, P. spinulosum and In our study, genus Rhizopus oryzae was isolated in the P. viridicatum. During the 2nd sampling, P. brevicompac- stem cell transportation department only during the 2nd and tum was isolated from 3 points in the stem cell transplanta- 4th sampling periods. Similarly, it was isolated in other tion department. 5 sampling points only once (Table 3). The fact that the Other Penicillium species isolated from the stem cell species was rarely isolated from the air oncology hospital transplantation department were P. chrysogenum and P. suggests that there was no colonization in the unit, but camemberti. P. chrysogenum was isolated at four different spread into the air in the main entrance and exit, or through points only during the 3rd sampling. P. camemberti was iso- the ventilation system, the house flora carried by the staff, lated only at one point during the 3rd sampling. or the equipment used in the treatment. Nosocomial infections have increased in recent years and have been the In the other parts of the hospital, P. chrysogenum and P. cause of morbidity and mortality to a considerable extent. expansum were frequently isolated. Like Aspergillus spe- Immunocompromised patients or patients with immunode- cies, Penicilium species are generally isolated from indoor ficiency for any reason (children and elderly) are at high air and spread into the air due to factors present in the envi- risk for nosocomial infections. It is well known that another ronment [19]. These conditions observed might be related to cause of nosocomial infections is microorganisms present variations in operating personnel and activities [20]. in the air we breathe. When genus concentrations determined at regular in- Therefore, hospitals, especially departments where pa- tervals were compared, the highest counts were as follows: tients with suppressed immune system stay, should be rd genus Cladosporium at the 3 month (p = 0.026 <0.05), monitored in terms of indoor air quality. Air sampling th genus Penicillium at the 4 month (p = 0.006 <0.05), genus works conducted in hospitals at regular intervals will make th Cladosporium at the 5 month (p = 0.000 <0.05), genus it possible to identify contamination sources and potential th Cladosporium at the 6 month (p = 0.004 <0.05) and genus disease agents likely to be present in the air [15, 26 - 28]. th Cladosporium months at the 7 month (p = 0.002 <0.05). The results revealed that the most common genus was Cladosporium. ACKNOWLEDGEMENT Cladosporium oxysporium species was isolated during all the samplings conducted in the stem cell transplantation The authors would like to extend their thanks to Ege department. Isolation of such fungi species as Cladosporium University, Science Faculty. This work was supported by and Alternaria specific to outdoor air indicate that the venti- BAP Project (03-FEN-009), Ege University. lation system was not working effectively and filters were clogged, which suggests that there was an effective barrier The authors have declared no conflict of interest. between the filters and outdoor environment [4]. This is thought to result from that fact that required precautions may not have taken when people entered or left REFERENCES the intensive care unit. A. citri and C. oxysporim were the [1] Rainer, J., Peintner, U. and Pöder, R. (2000) Biodiversity and most frequently isolated species during all the samplings in Concentration of Airborne Fungi in a Hospital Environment. every part of the hospital. Mycopathologia, 149 (2), 87–97.

322 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

[2] Perdelli, F., Cristina, M.L., Sartini, M., Spagnolo, A. M., [20] Araujo, R., Cabral, J.P., Rodrigues, A.G. and Portugal, P. Dallera, M., Ottria, G., Lombardi, R., Grimaldi, M. and Or- (2008) Air Filtration Systems and Restrictive Access Condi- lando, P. (2006) Fungal Contamination in Hospital Environ- tions Improve Indoor Air Quality in Clinical Units: Penicil- ments. Infection Control and Hospital Epidemiology 27(1), lium as a General Indicator of Hospital Indoor Fungal Levels. 44–47. American Journal of Infection Control, 36(2), 129–134. [3] Falvey, D.G. and Streifel, A.J. (2007) Ten-Year Air Sample [21] Clare, A. and Dykewicz. (2001) Summary of the Guidelines Analysis of Aspergillus Prevalence in a University Hospital. for Preventing Opportunistic Infections among Hematopoietic Journal of Hospital Infection 67 (1), 35–41. Stem Cell Transplant Recipients, Clinical Infectious Diseases, 33 (2), 139–44. [4] Sautour, M., Sixt, N., Dalle, F., L'Ollivier, C., Fourquenet, V., Calinon, C., Paul, K., Valvin, S., Maurel, A., Aho, S., Couillault, [22] Srinivasan, A., Beck, C., Buckley, T., Geyh, A., Bova, G., G., Cachia, C., Vagner ,O., Cuisenier, B., Caillot, D. and Bonnin, Merz, W. and Perl, T.M. (2002) The Ability of Hospital ven- A. (2009) Profiles and Seasonal Distribution of Airborne Fungi tilation Systems to Filter Aspergillus and Other Fungi in Indoor and Outdoor Environments at a French Hospital. Sci- Follwing a Building Implosion. Infection Control and Hospital ence of the Total Environment 407 (12), 3766–3771. Epidemiology, 23(9), 520–524. [5] Uzun, O. (2003) Yoğun Bakım Ünitesinde Fungal İnfeksiyon- [23] Singh, N. and Paterson, D.L. (2005) Aspergillus Infections in lara Yaklaşım. Yoğun Bakım Dergisi, 3(2):1, 35-144. Transplant Recipients, Clinical Microbiology Reviews, 18 (1), 44–69. [6] Vackova, M., Buchta, V., Prymula, R., Cerman, J., Kubatova, A., Hamal, P., Raclavsky, V. and Chlibeka, R. (2006) The Oc- [24] Vonberg, R.P. and Gastmeier, P. (2006) Nosocomial Aspergil- curence of Microscopic Fungi in Air Samples from a Transplant losis in Outbreak Settings. Journal of Hospital Infection, 63 Intensive Care Unit. Indoor Built Environ, 15 (1), 115–118. (3), 246–254. [7] Barnett H.L and Hunter B.B. (1999) Illustrated Genera of Im- [25] Gniadek, A. and Macura, A.B. (2007) Intensive care unit en- perfect Fungi. St. Paul: APS. 4th Ed., p. 218. vironment contamination with fungi. Adv Med Sci, 52, 283– 287. [8] Pitt, J.I. (2000) A Laboratory Guide to Common Penicillium species. Food Science Australia, 197p. [26] Lutz. B.D., Jin, J., Rinaldi, M.G., Wickes, B.L. and Huycke, M.M. (2003) Outbreak of Invasive Aspergillus Infection in [9] Klich, M.A. (2002) Identification of Common Aspergillus spe- Surgical Patients, Associated with a Contaminated Air-Han- cies. Centraalbureau voor Schimmelcultures, Utrecht, The dling System. Clinical Infectious Diseases, 37(6), 786–93. Netherlands, 116p. [27] Martins-Diniz, J.N., da Silva, R.A., Miranda, E.T. and [10] Hasenekoğlu, I. (1991) Soil fungi, Publication of Atatürk Uni- Mendes-Giannini, M.J. (2005) Monitoring of Airborne Fungus versity, Volume 7 (689). Kazim Karabekir Faculty of Educa- and Yeast species in a Hospital Unit. Revistade-Saúde-Pú- tion No: 11, Erzurum (in Turkish) blica, 39(3), 398–405. [11] Kalkanci, A., Kustimur, S., Sucak, G.T., Senol, E., Sugita, T., [28] Garnacho-Montereo, J. and Amaya-Villar, R.A. (2006) Vali- Adams, G., Verkley, G. and Summerbell, R. (2006) Fulminat- dated Clinical Approach for the Management of Aspergillosis ing Fungal Sinusitis Caused by Valsa sordida, a Plant Patho- in Critically Ill Patients: Ready, Steady, Go! Crit Care. 10, gen, in a Patient Immunocompromised by Acute Myeloid Leu- 132. kemia. Med Mycol. 44(6), 531–539. [12] Alberti, C., Bouakline, A., Ribaud, P., Lacroix, C., Rousselot, P., Leblanc, T. and Deroun, F. (2001) Relationship Between Environmental Fungal Contamination and the Incidence of In- vasive Aspergillosis in Haematology Patients. J Hosp Inf, 48 (3), 198–206.

[13] Cox, C.S. and Wathes, C.M., (1995) Bioaerosols Handbook, Macher, J.M., Streifel, A.J. and Vesley, D. Problem Building, Laboratories and Hospitals, p.505-509. [14] Frikin, S.K. and Jarvis, W.R. (1996) Epidemiology of Noso- comial Fungal Infections. Clinical Microbiology Reviews, Received: April 16, 2014 9(4), 499–511. Revised: May 22, 2014 Accepted: June 26, 2014 [15] Ortiz, G., Yague, G., Segovia, M. and Catalan, V. (2009). A Study of Air Microbe Levels in Different Areas of a Hospital. Curr Microbiol, 59(1), 53–58. CORRESPONDING AUTHOR [16] Dagenais, T.R.T. and Keller, N.P. (2009) Pathogenesis of As- pergillus fumigatus in Invasive Aspergillosis. Clin Microbiol Rev., 22(3), 447–465. Ozlem Abaci Gunyar Ege University [17] Li, C. and Hou, P. (2003) Bioaerosol Characteristics in Hospi- tal Clean Rooms. The Science of the Total Environment, 305 Faculty of Science (1–3), 169–176. Department of Biology [18] Lyratzopoulos, G., M. Ellis, M., Nerringer, R. and Denning, Basic & Industrial Microbiology Section D. W. (2002) Invasive Infection Due to Penicillium species 35100 Bornova-İzmir Other Than P. marneffei. J Infection, 45(3), 184-207. TURKEY [19] Sarica, S., Asan, A., Oktun, M.T. and Ture, M. (2002). Moni- toring Indoor Airborne Fungi and Bacteria in the Different Ar- E-mail: [email protected] eas of Trakya University Hospital, Edirne, Turkey. Indoor Built Environ, 11(5), 285–292. FEB/ Vol 24/ No 1b/ 2015 – pages 317 – 323

323 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

PHOTOREDUCTION OF MERCURY (II) IN AQUEOUS SUSPENSIONS OF DIFFERENT ALGAE

Lin Deng1,*, Jincheng Shi1, Caiqian Yang1,* and Nansheng Deng2

1School of Civil Engineering，Southeast University, Nanjing 210096, PR China 2School of Resources and Environmental Science, Wuhan University, Wuhan 430072, PR China

ABSTRACT Previous studies based on Hg(II) elimination onto natural materials had demonstrated that mercury uptake was a The effect of algae on the photoreduction of mercury (II) complex process which involved adsorption-reduction pro- was investigated under the irradiation of metal halide lamps cess [3–5]. Their studied materials implied different func- (ë ≥ 365 nm, 250W). Three algal species were used: Chlo- tional groups in their structure such as carboxyl and sulfate rella vulgarism, Anabaena cylindrical and Nitzschia groups, while these groups were located in cell wall poly- hantzschiana. The enhancement effect was influenced saccharides [6, 7], as well as other organic functional by Fe(III) ions, humic substances, algal concentration, in- groups which could also play an important role in metal itial mercury (II) concentration and pH. The photoreduc- reduction process [8]. Among different kinds of materials tion rate of mercury (II) was found to increase with increas- tested in pollutant elimination studies, high efficiency for ing the concentration of algae, pH, Fe (III) ions, and humic metal removal using algae as cost effective sorbent has substances and decrease with increasing the concentration been specially proved [9–12]. Sargassum muticum algal of initial mercury (II) concentration. The effects of differ- was selected as a potentially good material to provide high ent algae species on the photoreduction of Hg(II) followed mercury eliminations through a combined adsorption–re- the trend of Anabaena cylindrical > Nitzschia hantzschi- duction process, as it had been already proved [4]. Mercury ana > Chlorella vulgarism under the same condition. (II) and chromium (VI) were photo-reduced in the presence of algae, Chlorella vulgarism [13, 14]. The diel changes of

KEYWORDS: DGM (Dissolved gaseous mercury) upon the uptake of so- Photoreduction; Mercury (II); suspensions; algae lar radiation have been observed in natural freshwaters [15- 17]. These findings implicate an intrinsic role of sunlight

in controlling aquatic DGM dynamics. Recent field sun- 1. INTRODUCTION light incubations of northern lake waters in Teflon bottles further suggested that sunlight-induced photochemical re- Mercury, together with lead and cadmium, are consid- duction of Hg (II) in freshwater can induce the natural pro- ered as the most toxic heavy metals with the greatest po- duction of DGM [18,19]. Because the competition between tential hazard to humans and the environment. Mercury the sunlight-driven pathway for Hg(II) substrate and the poisoning effects are not immediately obvious, while the methylation pathway would reduce Hg toxic hazards in lo- two major responses to its toxicity involve neurological cal aquatic ecosystems through removal of Hg(II) as a re- and renal disturbances [1]. Mercury and its compounds are sult of its reduction to Hg and its subsequent evasion [20], used in dental preparations, thermometers, fluorescent and it is of significance to study on the aquatic Hg biogeochem- UV lamps, pharmaceuticals, fungicides in paints and in- ical cycling and its environmental impacts. Although the dustrial process waters [2]. All these human activities ren- last decades witnessed the increasing efforts on the obser- der many sources and discharges of mercury in the envi- vations of sunlight-induced aquatic production of DGM in ronment; Mercury (Hg)-containing compounds are envi- both northern and southern aquatic systems [21], the un- ronmentally extreme malign but the related toxicological derstanding of such phenomena in fundamental still re- procedure is yet still elusive. Furthermore, Hg-containing mains interesting but challenging. compounds are very difficult to be naturally degraded to yield the products that are environmentally benign, alt- The reduction rate of Hg (II) is controlled by Hg which hough the evolution of such compounds can occur based is relative to the biotic and abiotic processes of Hg-based on various biological and/or chemical reactions, forming evolutions. The abiotic reduction of Hg (II) can be initiated different chemical structures. Therefore, studying on the by humic substances in water [22]. It is assumed that inter- pollutions containing mercury is of significant importance mediates such as semiquinones presented in humic subin environmental protection. stances, are involved in the reduction of Hg (II). In addition, it is also indicated that the rate of such abiotic reduc- * Corresponding author tion can be photochemically enhanced under laboratory

324 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

conditions [23]. Among the abiotic photochemical path- and N. hantzschiana (Nitzschia hantzschiana) which ob- ways involved, Schroeder et al have suggested that hydro- tained from the Wuhan Hydrobiology Institute of Chinese gen peroxide (H2O2), mainly originated from UV-induced Academy of Sciences (Wuhan, P. R.China). The algae were transformations of dissolved organic matter in surface wa- grown in culture medium at 25℃ using 24-hour light cycle ters [24]. Some biotic processes leading to the reduction of in a culturing room equipped with constant temperature air- Hg(II) had also been identified in previous literature [13- conditioner (the room light intensity was 2000 Lux). 17]. The primary mercury- detoxifying mechanism in the For C. vulgaris, the culture medium consisted of case of mercury-resistant bacteria was explained as the re- 200 mg/L (NH4)2SO4, 30 mg/L [Ca(H2PO4) 2·2H2O + duction of mercuric ion to volatile Hg [25]. Photosynthesis CaSO ·H O], 80 mg/L MgSO ·7H O, 100 mg/L NaHCO , may also produce Hg(II) reductants, and the rates of pro- 4 2 4 2 3 25 mg/L KCl, 0.150 L/L FeCl3 (1%), 2.86 mg/L H3BO3, ducing metallic Hg have been observed during the course 1.81 mg/L MnCl ·4H O, 0.222 mg/L ZnSO ·7H O, 0.0177 of algal blooms [15-17]. 2 2 4 2 mg/LMoO3 (85%), 0.079 mg/L CuSO4·5H2O, 0.5 mL/L Although photochemical and/or photobiological pro- soil extract; the medium was adjusted to pH 7.0-7.2 by us- cesses are considered as one of the important factors to af- ing 0.1M Na2CO3. fect the reduction rate of Hg (II) in lake water [26], few For A. cylindrical, the culture medium consisted of experimental evidences, have been reported on the reduc- 200 mg/L (NH4)2SO4, 30 mg/L [Ca(H2PO4) 2·2H2O + tion of mercury (II) in the system containing algae com- CaSO4·H2O], 80 mg/L MgSO4·7H2O, 100 mg/L NaHCO3, plexes, especially those performed the effects of different 25 mg/L KCl, 0.150 L/L FeCl (1%), 2.86 mg/L H BO , algae species on the photoreduction of Hg (II) in the de- 3 3 3 1.81 mg/L MnCl2·4H2O, 0.222 mg/L ZnSO4·7H2O, 0.0177 tectable literature up to date. In this context, the present mg/LMoO3 (85%), 0.079 mg/L CuSO4·5H2O, 0.5 mL/L work will focus on the effects of algae species on the pho- soil extract; the medium was adjusted to pH 7.0-7.2 by us- toreduction of Hg (II) in natural water. The main objective ing 0.1 M Na CO . of the present work is to carry out chemical modifications 2 3 to provide information for the best mercury elimination ca- For N. hantzschiana, the culture medium consisted pacities using C. vulgarism, A. cylindrical and N. Hantzschi- of 120mg/L NaNO3, 70mg/LMgSO4·7H2O, 40mg/ ana under the irradiation of metal halide lamps. These ex- LK2HPO4·3H2O, 80mg/ LKH2PO4, 20mg/LCaCl2·2H2O, periments provide additional information about the com- 10 mg/L NaCl, 100 mg/L Na2SiO3·9H2O, 2 mg/L bined mechanism implied in mercury elimination. Moreo- MnSO4·4H2O, 2.86 mg/L H3BO3, 1.81 mg/L MnCl2·4H2O, ver, different algae treatments and their efficiency for mer- 0.22 mg/L ZnSO4·7H2O, 0.079 mg/L CuSO4·5H2O, cury removal are reported in this work. In addition, a com- 0.039 mg/L Na2MoO4·2H2O, 5 mg/L ferric citrate, 0.5 mL/L plete elimination study has been carried out to compare the soil extract. The medium was adjusted to pH 7.0-7.2 by us- capacity of different algae for mercury removal. The influ- ing 0.1M Na2CO3. ence factors of mercury elimination have been analyzed. The algae were cultured in axenic medium. After stay- This research will provide information for predicting the ing for 12-14 days, the algae were washed prior to each photochemical transformation of mercury (II) in the real measurement. To remove the colloidal ferric hydroxide environment and different algae treatments and their effi- particles that might have adsorbed on the algae cells, a ciency for mercury removal. modified version of the procedure [27] was used in experiments. This procedure involves washing the cells by gen- tle agitation for 30 min with 0.01M aqueous ascorbic acid 2 MATERIALS AND METHODS adjusted to pH 3.0. Then the algae were washed with double-distilled water three times. The resulting algae suspen- 2.1 Chemicals and reagents sion was obtained in this way. The cell counting was car- HNO3,KBr, KBrO4,NaCl, HCl, NH2OH·HCl, ried out under inverted microscope at 400× and the density -1 (NH4)2SO4, SnCl2, Ca(H2PO4)2·2H2O, CaSO4·H2O, of algae (cells L ) was calculated. As such, the algae were MgSO4·7H2O, FeCl3·6H2O, CaCl2·2H2O, CuSO4·5H2O, well-prepared for subsequent use. Different concentrations H3BO3, KCl, K2HPO4, KH2PO4, MgSO4·7H2O, K2Cr2O7, of algae were gained through diluting the washed algae MnCl2·4H2O, MoO3, NaOH, NaCl, Na2CO3, NaHCO3, with double-distilled water. These experiments were car- Na2MoO4·2H2O, NaNO3, Na2SiO3·9H2O, ascorbic acid, and ried out at a room temperature of 26 ± 2. ferric citrate were analytical grade and obtained from Beijing chemical reagents company. Humic substances were pur- 2.3 Irradiation procedure chased from Aldrich Chem. Co. (Milwaukee, WI, USA). Un- The irradiation experiments were carried out in a cy- less stated otherwise, Double-distilled water and N2 (98.5%) lindrical reactor (20.0 cm length, 10.0 cm diameter, 1.0 cm were used in all experiments. HgCl2 was analytical grade wall thickness) (see Fig. 1), with a 250 W metal halide and denoted as Hg (II) in this context. lamp (ë ≥ 365nm, Changzhou Shangzi Lamp Co. Ltd, China) placed in cooling trap for maintaining a constant 2.2 Preparation algae temperature (25±2°C) by water circulation. The cylindrical The algae used in the experiments were C. vulgaris reactor was unsealed in the uppermost part. The cylindrical (Chlorella vulgarism), A.cylindrical (Anabaena cylindrical) reactor was placed in a box. HCl and/or NaOH were used

325 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

to adjust the pH values of solutions. The cylindrical reactor of C.vulgarism, A.cylindrical and N.hantzschiana, respec- containing solutions were kept in a dark condition before tively. For comparison, the corresponding control experi- and after irradiation. Different concentrations of mercury (II) ments were performed under a dark condition. In the pres- and humic substances, Fe (III) or algae were mixed thor- ence of C.vulgarism, A.cylindrical and N.hantzschiana, re- oughly and transferred into the cylindrical reactor. The spectively, 6.7%, 11.1% and 9.9% of Hg (II) were removed light intensity at the position of the cylindrical reactor was from the suspension after 7-hour, which is due to the slow 159,000 Lux, which was detected using Digit Lux meter absorption and/or reduction by the biological and ther- (TES 1332, Taiwan, China). At different time intervals mal chemical processes [28]. Under the metal halide lamp during the irradiation, samples with mercury (II) and Fe irradiation, 13.0% of Hg (II) disappeared after 7-hour irra- (III), humic acid or algae were taken from the cylindrical diation without algae, suggesting the slow thermal reduc- reactor. For the solution with algae, samples were digested tion, absorption and the direct photoreduction of Hg (II). with HCl-HNO3 (1+1). Then the samples and a Teflon- The direct photoreduction of Hg (II) could be responsible coated stir bar were placed in a caped and secured glass for the increased disappearance of Hg (II). flask and heated to 90°C in a water bath for 1 h. or more Hg (OH) 2 + light →Hg (0) (aq) + other products (1) upon cooling, the samples were analyzed by a cold vapor atomic absorption spectrometer (AAS). The algae-free sam- In the presence of A.cylindrical, C. vulgarism and ples were analyzed directly. N.hantzschiana, respectively, 43.0%，27.1% and 33.3% of Hg (II) disappeared from the suspension after 7-hour irradiation under the metal halide lamps, indicating that the photolysis of algae could promote the reduction of Hg (II). The algae might undergo photolysis to generate smaller organic molecules and free electrons that is capable to reduce Hg (II). The enzymatic reaction at the algae cell surface could be another possible mechanism involved in the reduction of Hg (II) [29]. As shown in Figure 2, the results indicate that algae species on the photoreduction of Hg (II) followed the trend of A.cylindrical > N.hantzschiana > C.vulgarism under the irradiation of metal halide lamps. The photoreduction rate of Hg (II) was quite different in different algal suspensions. The results suggest that environmental oxidative stress effects are different among algal species: the species have different cell wall structure, which can be related to their resistance to irradiation. The FIGURE 1 - Cylindrical reactors A. cylindrica is a prokaryote; its cell walls are made of murein and peptidoglycan [30] and are thin and easily sol-

2.4 Chemical analyses uble. So A.cylindrica is easily destroyed. But C. vulgaris is eucaryotes; the cell walls consist of fibrin and chitin and The mercury (II) concentrations were determined by a are stronger than that of A. cylindrica, and so they are dif- cold vapor AAS. The calibration equation for mercury (II) ficult to destroy. The cell walls of N. hantzschiana are was A peak area = 96.24 C Hg (II) − 7.24 (r = 0.996), where C made of pectic substances (inside layer) and chitin (outside Hg (II) was the concentration of mercury (II) in the range of layer), and contain inorganic substances such as SiO2 and 0.0–200.0 µg/L. All vitreous apparatus were dipped in CaCO3 [31]. This enhances the rigidity of the cell wall and HNO3/water (the volume ratio was 1:1) overnight to re- makes it hard to destroy. As a result, the amounts of pho- move impurities. 3.0 ml reaction solution was transferred tosensitising substances (such as pigments, carboxylic ac- into a 20.0 ml reduction bottle. Then 2.0 ml HNO3 (5%) ids) released into the reaction solution by the algal cells and 2.0 ml SnCl2 (10%) were added into the reduction bot- when they were irradiated were quite different. tle. Finally, Hg (II) was determined by a cold vapor AAS. All the experiments were triplicate. The results pre- 3.2 Effect of pH on the photoreduction of mercury (II) in aqueous solution with or without algae sented were the mean values with a less than 5% total analytical error. The photoreduction experiment was performed in Hg (II) aqueous solutions at several pH values ranging from 5 to 9. 3. RESULTS AND DISCUSSION As shown in Fig.3 (a), it is clear that the photoreduction efficiency of Hg (II) increased with increasing pH. At 3.1 Photoreduction of Hg (II) both in the absence and presence pH = 5, up to 9.4% of Hg (II) was removed from the sus- of algae pension after 7-hour light irradiation. At pH = 9, up to Figure 2 shows the typical time series for the photo- 15.8% of Hg (II) was removed from the suspension after 7- chemical reduction of Hg (II) in the absence and presence hour light irradiation. The observed pH effect demon-

326 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

1.0

0.9

0.8 0 0.7 C/C UV+Hg(II) 0.6 A.cylindrica+Hg(II) A.cylindrica+Hg(II)+UV C.vulgaris+Hg(II) C.vulgaris+Hg(II)+UV 0.5 N.hantzschiana+Hg(II) N.hantzschiana+Hg(II)+UV 01234567 Exposure time(h)

FIGURE 2 - Mercury (II) concentration changes in solutions that contained 100ìg L-1 Hg (II) and/or C. vulgaris: 7.0×109 cells L-1, A.cylindrical: 1.7×109 cells L-1, N. hantzschiana: 7.0×109 cells L-1 with/without metal halide lamps (250 W) irradiation. The solution was at pH 7.0.

1.0 1.0 (a) (c)

0.9 0.9 0 0 C/C 0.8 0.8

pH=5 C/C pH=7 pH=5 0.7 pH=9 0.7 pH=7 pH=9

0.6 0.6 01234567 01234567 Exposure time(h) Exposure time(h)

1.0 1.0 (b) (d) 0.9

0.9 0.8

0.7 0

0 0.8 0.6 C/C C/C 0.5 pH=5 pH=5 pH=7 0.7 0.4 pH=7 pH=9 pH=9 0.3 01234567 0.6 Exposure time(h) 01234567 Exposure time(h)

FIGURE 3 - Effect of pH on the photoreduction of mercury (II) in solutions that contained (a) 100ìg L-1 Hg (II) (b) 100ìg L-1 Hg (II) and 7.2×109 cells L-1C. vulgarism (c) 100ìg L-1 Hg (II) and 4.5×109 cells L-1 N. hantzschiana (d) 100ìg L-1 Hg (II) and 1.7×109 cells L-1A. cylindrical under metal halide lamps (250 W) irradiation.

327 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

strated that Hg (OH) 2 could be the reactive species of Hg (II) were shown in Fig.4(c), After 7-hour irradiation, 35.0% of involved in the direct photochemical reduction of Hg (II) in Hg (II) was removed from the suspension under the initial aqueous solution, which was in consistence with those re- algal concentration 1.6×109cells L-1; 51.9% of Hg (II) was ported in previous studies [32, 33]. disappeared at initial algal concentration of 3.5×109cells L-1; Experiments were also carried out to study the pH ef- 55.7% removal of Hg (II) at the algal concentration of 9 -1 fect on the photoreduction of mercury (II) in aqueous sus- 4.6×10 cells L . A higher algal concentration resulted in a pensions of C.vulgarism, N. hantzschiana and A.cylindri- faster photoproduction of free electrons and smaller dis- cal under metal halide lamps. The results showed that the solved organic matter. The dissolved organic matter, such photoreduction of Hg (II) increased with increasing pH as humic and fulvic acid, could serve as photosensitizers value in the range of 5.0–9.0. As shown in Fig.3 (b), at pH and thus accelerate the photoreduction of mercury (II) [34]. =5, 21.6% of Hg (II) was disappeared from the C.vulgar- From Fig 4, the results indicate that algae species on the ism suspension after 7-hour irradiation; at pH = 9, 33.4% photoreduction of Hg (II) still followed the trend of A.cy- of Hg (II) was disappeared from the C.vulgarism suspen- lindrical > N.hantzschiana > C.vulgarism under the irradi- sion after 7-hour irradiation. As shown in Fig.3 (c), at pH ation of metal halide lamps. =5, 23.1% of Hg (II) was disappeared from the N. hantzschiana suspension after 7-hour irradiation; at pH = 9, 33.9% of Hg (II) was disappeared from the N. hantzschi- 1.00 ana suspension after 7-hour irradiation. As shown in Fig.3 0.95 (a) (d), at pH =5, 26.8% of Hg (II) was disappeared from the 0.90 A.cylindrical suspension after 7-hour irradiation; at pH = 0.85

9, 63.7% of Hg (II) was disappeared from the A.cylindrical 0 0.80 suspension after 7-hour irradiation. In the aqueous suspen- C/C 0.75 sion of algae, the photochemical and biological processes 1.5×1010cells/L 0.70 7.0×109cells/L 9 are more complex. Besides that the higher concentration of 0.65 5.7×10 cells/L 9 Hg(OH)2 at higher pH value leads to a higher photoreduc- 4.8×10 cells/L 0.60 tion of mercury (II), the enhanced production of dissolved 01234567 organic matter and free electrons also accelerated the con- Exposure time(h) version of Hg (II) to Hg (0). From Fig 3, the results indicate that algae species on the photoreduction of Hg (II) still fol- 1.00 lowed the trend of A.cylindrical > N.hantzschiana > C.vul- (b) garism under the irradiation of metal halide lamps. 0.95 0.90 3.3 Effects of algae concentration on the photoreduction of 0.85 Hg (II) 0 0.80

To test the effects of algae on the photoreduction of C/C 0.75 7.0×109cells/L mercury (II), suspensions of Chlorella vulgaris at four ini- 4.6×109cells/L 0.70 tial algal concentrations were 4.8×109, 5.7×109, 7.0×109, 3.1×109cells/L 9 1.5×1010 cells L-1, at pH 7.0. As shown in Fig.4 (a), the 0.65 2.6×10 cells/L faster photoreduction of mercury (II) occurred with a 0.60 higher algal concentration. After 7-hour irradiation, 30.9% 01234567 Exposure time(h) of Hg (II) was removed from the suspension under the initial algal concentration 1.5×1010cells L-1; 27.1% of Hg (II) 9 1.00 was disappeared at initial algal concentration of 7.0×10 0.95 (c) -1 cells L ; 22.0% removal of Hg (II) at the algal concentra- 0.90 9 -1 tion of 5.7×10 cells L ; 18.3% removal of Hg (II) at the 0.85 1.6×109cells/L algal concentration of 4.8×109cells L-1. When suspensions 0.80 3.5×109cells/L 9 of N. hantzschiana at four initial algal concentrations were 0.75 4.6×10 cells/L 0 0.70 2.6×109, 3.1×109, 4.6×109, 7.0×1010 cells L-1, at pH 7.0, the C/C 0.65 photoreduction of mercury (II) were shown in Fig.4(b), Af- 0.60 ter 7-hour irradiation, 28.9% of Hg (II) was removed from 0.55 the suspension under the initial algal concentration 0.50 0.45 9 -1 2.6×10 cells L ; 30.0% of Hg (II) was disappeared at ini- 0.40 tial algal concentration of 3.1×109cells L-1; 32.2% removal 01234567 of Hg (II) at the algal concentration of 4.6×109cells L-1; Exposure time(h) 33.3% removal of Hg (II) at the algal concentration of 9 -1 FIGURE 4 - Effect of algae concentrations on the photoreduction of 7.0×10 cells L . When suspensions of A. cylindrical at mercury (II) in solutions that contained 100ìg L-1 Hg (II) and (a) C. four initial algal concentrations were 1.6×109, 3.5×109, vulgarism (b) N. hantzschiana (c) A. cylindrical under metal halide 4.6×109 cellsL-1, at pH 7.0, the photoreduction of mercury (II) lamps (250 W) irradiation. The solution was at pH 7.0.

328 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

3.4 Effects of initial concentrations of mercury (II) on the pho- As shown in Fig.5 (a), the metal halide lamps could toreduction of mercury (II) in algal suspensions evidently induce photoreduction of Hg(II) in aqueous solu- The influence of initial Hg (II) concentration on the tion with C. vulgaris. After 7 hours irradiation, 34.3% of photoreduction of mercury (II) with the algae was exam- Hg(II) was removed from the suspension under the initial ined by irradiating the suspensions of C. vulgaris at initial Hg(II) concentration 50ìg/L; 31.2% of Hg(II) disappeared concentration was 4.7×109 cells L-1, at pH 7.0, with initial at initial Hg(II) concentration of 100ìg/L; 27.1% of Hg(II) concentrations (C0) of mercury (II) at 50, 100, 150 and disappeared at initial Hg(II) concentration of 150ìg/L; 200µg/L, respectively. 25.9% of Hg(II) disappeared at the Hg(II) concentration of 200ìg/L. When N.hantzschiana at initial concentration was 9 -1 7.0×10 cells L , at pH 7.0, with initial concentrations (C0) 1.0 of mercury (II) at 50, 100, 150 and 200µg/L, respectively, (a) the results were shown in Fig.5 (b). After 7 hours irradia- 0.9 tion, 36.2% of Hg(II) was removed from the suspension under the initial Hg(II) concentration 50ìg/L; 32.2% of 0.8 Hg(II) disappeared at initial Hg(II) concentration of

0 100ìg/L; 29.1% of Hg(II) disappeared at initial Hg(II) con- C =50 µgL-1 C/C 0.7 0 centration of 150ìg/L; 28.0% of Hg(II) disappeared at the C =100 µgL-1 Hg(II) concentration of 200ìg/L. When A. cylindrical at in- 0 C =150 µgL-1 itial concentration was 3.5×109 cells L-1, at pH 7.0, with in- 0.6 0 C =200 µgL-1 itial concentrations (C0) of mercury (II) at 50, 100, 150 and 0 200µg/L, respectively, the results were shown in Fig. 5 (c). 0.5 After 7 hours irradiation, 74.1% of Hg(II) was removed 01234567 from the suspension under the initial Hg(II) concentration Exposure time(h) 50ìg/L; 51.9% of Hg(II) disappeared at initial Hg(II) concentration of 100ìg/L; 45.4% of Hg(II) disappeared at initial Hg(II) concentration of 150ìg/L; 39.4% of Hg(II) dis- 1.0 appeared at the Hg(II) concentration of 200ìg/L. The pho- (b) toreduction rate of low concentration Hg(II) decreased 0.9 faster than that of high concentration of Hg(II) in aqueous solution with algae. Under these conditions, with initial 0.8 0 concentration (C0) of Hg(II) increasing, the reduction rate

C/C decreased. The disappearance of Hg(II) in terms of per- 0.7 C =50µgL-1 0 centage of the initial Hg(II) concentration was higher at C =100µgL-1 0 lower initial Hg(II) concentration. C =150µgL-1 0.6 0 C =200µgL-1 0 3.5 Effects of humic substances on the photoreduction of mer- 0.5 cury (II) in algal suspensions 01234567 Exposure time(h) Humic substances are a ubiquitous group of natural organic compounds because of their important role in the fate and transport of many pollutants. 1.0 In our case, humic substances could also affect the C =200μg/L 0 0.9 C =150μg/L photoreduction of Hg (II) as shown in Fig.6. When suspen- 0 sions of C. vulgaris at initial algal concentration was C0=100μg/L 0.8 9 -1 C0=50μg/L 7.2×10 cells L , at pH 7.0, the photoreduction of mercury

0 0.7 (II) were shown in Fig.6(a). In the presence of 8.0 mg/L

C/C humic substances, 9.7% Hg (II) was removed from a sus- 0.6 pension of C. vulgaris stood in dark for 7 hours. After 7-hour 0.5 irradiation, 37.2% Hg (II) disappeared in the presence of (c) 8.0 mg/L humic substances, while only 29.0% Hg (II) was 0.4 removed in the absence of humic substances. When sus- 0.3 pensions of N.hantzschiana at initial algal concentration 01234567 was 4.6×109cells L-1, at pH 7.0, the photoreduction of mer- Exposure time(h) cury (II) were shown in Fig.6(b), In the presence of 8.0 mg/L humic substances, 10.4% Hg (II) was removed from a sus- FIGURE 5 - Effect of initial Hg (II) concentrations on the photoreduc- pension of N.hantzschiana stood in dark for 7 hours. After tion of Hg (II) in solutions that contained (a) 4.7×109 cells L-1 Chlorella vulgarism, (b) 7.0×109 cells L-1 N. hantzschiana, (c) 3.5×109 cells L-1 An- 7-hour irradiation, 40.2% Hg (II) disappeared in the pres- abaena cylindrical under metal halide lamps (250 W) irradiation. The ence of 8.0 mg/L humic substances, while only 34.2% Hg (II) solution was at pH 7.0. was removed in the absence of humic substances. When

329 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

7-hour irradiation, 52.8% Hg (II) disappeared in the pres- 1.0 (a) ence of 8.0 mg/L humic substances, while only 44.2% Hg (II) was removed in the absence of humic substances. 0.9 These results indicated that humic substances could promote the reduction of mercury (II) in algal suspension. 0.8

0 Several research groups have reported that humic substances could promote the reduction of mercury (II) in C/C 0.7 aqueous solution [23, 33]. From Fig 6, the results indicate 0.6 that algae species on the photoreduction of Hg (II) still fol- C.vulgaris+UV+Humid acid(8mg/L) lowed the trend of A.cylindrical > N.hantzschiana > C.vul- 0.5 C.vulgaris+UV garism under the irradiation of metal halide lamps C.vulgaris+Humid acid(8mg/L) 0.4 3.6 Effects of Fe (III) on photoreduction of mercury (II) in algal 01234567 suspensions Exposure time(h) Fe (III) may also influence the photoreduction of mercury (II) as shown in Fig.7. When suspensions of C. vulgaris at initial algal concentration is 4.8×109 cells L-1, at pH 1.0 (b) 7.0, the photoreduction of mercury (II) are shown in 0.9 Fig.7(a). In the dark, 7.5% Hg (II) disappeared in the suspension of C. vulgaris containing 0.02mmol/L Fe (III). After 0.8 7-hour light irradiation, 18.3% of Hg (II) was removed from

0 suspension of C. vulgaris without Fe (III), while 26.8 % of

C/C 0.7 Hg(II) disappeared in the presence of 0.02 mmol/L Fe(III). When suspensions of N.hantzschiana at initial algal con- 0.6 centration was 4.6×109 cells L-1, at pH 7.0, the photoreduc- N.hantzschiana+UV+Humid acid(8mg/L) tion of mercury (II) were shown in Fig.7(b). In the dark, 0.5 N.hantzschiana+UV 9.4% Hg (II) disappeared in the suspension of N.hantzschi- N.hantzschiana+Humid acid(8mg/L) ana containing 0.02mmol/L Fe (III) after 7h. After 7-hour 0.4 01234567 light irradiation, 32.2% of Hg (II) was removed from sus- Exposure time(h) pension of N.hantzschiana without Fe (III), while 37.8 % of Hg(II) disappeared in the presence of 0.02mmol/L Fe(III). When suspensions of A.cylindrical at initial algal concentration was 4.6×109 cells L-1, at pH 7.0, the photore- 1.0 (c) duction of mercury (II) were shown in Fig.7(c). In the dark, 0.9 10.9% Hg (II) disappeared in the suspension of A.cylindrical containing 0.02mmol/L Fe (III). After 7-hour light ir- 0.8 radiation, 50.3% of Hg (II) was removed from suspension 0.7 of A.cylindrical without Fe (III), while 68.9% of Hg(II) dis-

0 0.6 appeared in the presence of 0.02mmol/L Fe(III). From Fig 7, the results indicate that Fe (III) could clearly accelerate C/C 0.5 the photoreduction of mercury (II) in algal suspensions un- 0.4 A.cylindrica+UV+Humid acid(8mg/L) der metal halide lamps irradiation. The involved abiotic A.cylindrica+UV mechanisms for the effects of Fe(III) on the photoreduction 0.3 A.cylindrica+Humid acid(8mg/L) of Hg(II) include light induced photochemical production 0.2 of highly reducing organic free radicals through photolysis 01234567 of Fe (III)-organo coordination compounds [Fe(III)-Org] Exposure time(h) and subsequent reaction of Hg (II) with the organic and inorganic free radicals formed [35, 36]: FIGURE 6 - Effect of humid acid on the photoreduction of mercury (II) in solutions that contained 100ìg L-1 Hg (II), 8mg L-1 humid acid Fe(III) + natural organic acids → Fe(III)-Org (2) and/or (a) 7.20×109 cells L-1 C. vulgarism (b) 4.6×109 cells L-1 N. hantzschiana (c) 1.6×109 cells L-1 A. cylindrical with/without metal hal- Fe(III)-Org + hv → Fe(II) + organic free radicals (3) ide lamps (250 W) irradiation. The solution was at pH 7.0. organic free radicals + Hg(II) → Hg(0) + products + CO2 (4) • - suspensions of A. cylindrical at initial algal concentration organic free radicals + O2 → O2 + products + CO2 (5) 9 -1 was 1.6×10 cells L , at pH 7.0, the photoreduction of mer- DOC + h v → DOC• (6) cury (II) were shown in Fig.6(c), In the presence of 8.0 DOC• + O → •DOC+ + •O - (7) mg/L humic substances, 7.8% Hg (II) was removed from a 2 2 • - • - + suspension of A. cylindrical stood in dark for 7 hours. After O2 + O2 + H → H2O2 + O2 (8)

330 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

• DOC + H2O + hv → OH + products (12) 1.0 (a) Fe (III) is highly photochemically reactive in natural waters and its role is well established in mediating photo- 0.9 chemical redox cycling of heavy metals in natural aqueous systems [38-40]. Photochemical reduction of heavy metal 0.8 ions [e.g., Cr (VI)] by highly reducing organic free radicals, 0 produced through photolysis of Fe (III)-organo coordination C/C 0.7 compounds has been implicated to be one of the most im- 3+ portant mechanisms for the redox cycling of these metals C.vulgaris+Fe (1.0 mg/L) [41,42]. The same mechanisms may also act in the observed 0.6 C.vulgaris+UV C.vulgaris+UV+Fe3+(1.0 mg/L) light-induced photochemical production of DGM in algal suspensions and natural water because both algae and Fe (III) 0.5 01234567 are commonly present in nature aquatic environment.

Exposure time(h) 3.7 Effects of Fe (III) and humic substances on photoreduction of mercury (II) in algal suspensions Fe (III) and humic substances could accelerate the pho- (b) 1.0 toreduction of mercury (II), but the synergy effect between 0.9 Fe (III) and humic substances need further experiments to investigate. When suspensions of C. vulgaris at initial algal 0.8 concentration was 4.8×109 cells L-1, at pH 7.0, the photore-

0 duction of mercury (II) were shown in Fig.8(a). In the dark, 0.7

C/C 7.8% Hg (II) disappeared in the suspension of C.vulgaris containing 1.0 mg/L Fe (III) and 8.0 mg/L humic sub- 0.6 N.hantzschiana+Fe3+(1.0 mg/L) stances in 7 hours. After 7-hour light irradiation, 18.3% of 0.5 N.hantzschiana+UV Hg (II) was removed from suspension of C.vulgaris with- 3+ N.hantzschiana+UV+Fe (1.0 mg/L) out 1.0 mg/L Fe (III), while 26.8 % of Hg (II) disappeared 0.4 in the presence of 1.0 mg/L Fe (III). 30.2 % Hg (II) disap- 01234567 peared from suspension of C.vulgaris in the presence of 8.0 Exposure time(h) mg/L humic substances. 45.2% Hg (II) disappeared from suspension of C.vulgaris in the presence of 1.0 mg/L Fe 1.0 (c) (III) and 8.0 mg/L humic substances. When suspensions of N.hantzschiana at initial algal concentration was 4.2×109 0.9 cells L-1, at pH 7.0, the photoreduction of mercury (II) were 0.8 shown in Fig.8(b). In the dark, 8.3% Hg (II) disappeared in 0.7 the suspension of N.hantzschiana containing 1.0 mg/L Fe 0.6 (III) and 8.0 mg/L humic substances in 7 hours. After 7- 0 0.5 hour light irradiation, 22.2% of Hg (II) was removed from C/C suspension of N.hantzschiana without 1.0 mg/L Fe (III), 0.4 while 28.8 % of Hg (II) disappeared in the presence of 1.0 0.3 3+ A.cylindrica+UV+ Fe (1.0 mg/L) mg/L Fe (III). 33.7 % Hg (II) disappeared from suspension 0.2 A.cylindrica+UV of N.hantzschiana in the presence of 8.0 mg/L humic sub- A.cylindrica+Fe3+(1.0 mg/L) 0.1 stances. 51.2% Hg (II) disappeared from suspension of 01234567 N.hantzschiana in the presence of 1.0 mg/L Fe (III) and 8.0 Exposure time(h) mg/L humic substances. When suspensions of A.cylindri- FIGURE 7 - Effect of Fe3+ on the photoreduction of mercury (II) in cal at initial algal concentration was 1.4×109 cells L-1, at solutions that contained 100ìg L-1 Hg (II), 1.0mg L-1 Fe3+ and/or (a) pH 7.0, the photoreduction of mercury (II) were shown in 9 -1 9 -1 4.8×10 cells L C. vulgarism, (b) 4.6×10 cells L N. hantzschiana, (c) Fig.8(c). In the dark, 9.1% Hg (II) disappeared in the sus- 3.5×109 cells L-1 A. cylindrical with/without metal halide lamps (250 W) irradiation. The solution was at pH 7.0. pension of A.cylindrical containing 1.0 mg/L Fe (III) and 8.0 mg/L humic substances in 7 hours. After 7-hour light H O + Fe (II) → •OH + OH- + Fe(III) (9) irradiation, 26.2% of Hg (II) was removed from suspension 2 2 of A.cylindrical without 1.0 mg/L Fe (III), while 41.2 % of •OH + Hg(0) → Hg(II) + OH- (10) Hg (II) disappeared in the presence of 1.0 mg/L Fe (III). • - 52.8% Hg (II) disappeared from suspension of A.cylindri- O2 + Hg(II) → Hg(0) + O2 (11) cal in the presence of 8.0 mg/L humic substances. 69.3% • - Where O2 is a superoxide free radical. In addition to Hg (II) disappeared from suspension of A.cylindrical in the reaction 9, •OH may also be produced in natural water by presence of 1.0 mg/L Fe (III) and 8.0 mg/L humic sub- direct photolysis of DOC [37]: stances. From Fig 8, the results indicate that Fe (III) and

331 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

effects of Fe (III) and humic substances on the photoreduc- 1.0 (a) tion of Hg (II) were similar to those proposed in previous studies [13-14]. From Fig 8, the results indicate that 0.9 algae species on the photoreduction of Hg (II) still fol- 0.8 lowed the trend of A.cylindrical > N.hantzschiana > C.vulgarism under the irradiation of metal halide lamps. 0.7

0 0.6

C/C 3+ 4. CONCLUSIONS 0.5 C.vulgaris+Humid acid+Fe C.vulgaris+UV 0.4 3+ C.vulgaris+UV+Fe It was demonstrated that algae, Chlorella vulgaris, 0.3 C.vulgaris+UV+Humid acid could significantly accelerate the photochemical reduction 3+ C.vulgaris+UV+Humid acid+Fe of mercury (II) under the irradiation of metal halide lamps. 0.2 01234567 The photoreduction of mercury (II) increased with increas-

Exposure time(h) ing concentration of algae, Fe (III) and humic substance. Increasing pH value also led to a higher photoreduction rate of mercury (II) in algae suspensions. The photoreduction rate of low concentration Hg(II) decreased faster than 1.0 (b) that of high concentration of Hg(II) in aqueous solution 0.9 with algae. Effects of different algae species on the photoreduction of Hg (II) followed the trend of Anabaena cylin- 0.8 drical > Nitzschia hantzschiana > Chlorella vulgarism un- 0.7

0 der the irradiation of metal halide lamps.

C/C 0.6

0.5 N. hantzschiana+Humid acid+Fe3+ ACKNOWLEDGEMENTS 0.4 N. hantzschiana+UV N. hantzschiana+UV+Fe3+ Financial support from the National Key Technology 0.3 N. hantzschiana+UV+Humid acid N. hantzschiana+UV+Humid acid+Fe3+ Support Program (No.2013BAC05B02) and the Priority Ac- 0.2 ademic Program Development of Jiangsu Higher Education 01234567 Institutions (No.1105007001) is gratefully acknowledged. Exposure time(h) The authors have declared no conflict of interest.

A.cylindrica+UV+Humid acid+ Fe3+ A.cylindrica+UV+Humid acid A.cylindrica+UV+Fe3+ A.cylindrica+UV REFERENCES 3+ 1.0 A.cylindrica+Humid acid+Fe [1] Wang L.K., Chen J.P., Hung Y., Shammas N.K. (2009) Heavy 0.9 Metals in the Environment, CRS Press Taylor & Francis, Boca 0.8 Raton, pp.25-29.

0 0.7 [2] Carro L., Barriada J.L., Herrero R., Sastre de Vicente M.E. C/C 0.6 (2013) Surface modifications of Sargassum muticum algal bi- 0.5 omass for mercury removal: A physicochemical study in batch and continuous flow conditions, Chemical Engineering Jour- 0.4 (c) nal. 229, 378-387. 0.3 [3] Carro L., Anagnostopoulos V., Lodeiro P., Barriada J.L., Her- 0.2 01234567 rero R., Sastre de Vicente M.E. (2010) A dynamic proof of Exposure time(h) mercury elimination from solution through a combined sorption–reduction process, Bioresour. Technol. 101, 8969–8974. [4] Carro L., Barriada J.L., Herrero R., Sastre de Vicente M.E. FIGURE 8 - Effect of humid acid and Fe3+ on photoreduction of mer- (2011) Adsorptive behaviour of mercury on algal biomass: cury (II) in solutions that contained 100ìg L-1 Hg (II) and/or 8mg L-1 competition with divalent cations and organic compounds, J. humid acid, 1.0mg L-1 Fe3+ and (a) 4.8×109 cells L-1 C. vulgarism, (b) Hazard. Mater. 192, 284–291. 4.2×109 cells L-1 N. hantzschiana, (c) 1.4×109 cells L-1 A. cylindrical [5] Cox M., El-Shafey E.I., Pichugin A.A., Appleton Q. (2000) with/without metal halide lamps (250 W) irradiation. The solution Removal of mercury(II) from aqueous solution on a carbona- was at pH 7.0. ceous sorbent prepared from flax shive, J. Chem. Technol. Bi- otechnol. 75, 427–435. humic substances had a synergy effect on the photoreduc- [6] Crist R.H., Oberholser K., Schwartz D., Marzoff J., Ryder D., tion of mercury (II) in algal suspensions under metal halide Crist D.R. (1988) Interactions of metals and protons with al- lamps irradiation. The abiotic mechanisms involved for the gae, Environ. Sci. Technol. 22, 755–760.

332 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

[7] Davis T.A., Volesky B., Mucci A. (2003) A review of the bi- [24] Schroeder W.H., Yarwood G., Niki H. (1991) Transformation ochemistry of heavy metal biosorption by brown algae, Water processes involving mercury species in the atmosphere—re- Res. 37, 4311–4330. sults from a literature survey, Water Air Soil Pollut. 56, 653- 666. [8] Deng B., Stone A.T. (1996) Surface-catalyzed Chromium (VI) reduction: reactivity comparisons of different organic reduct- [25] Steffan R.J., Korthals E.L., Winfrey M.R. (1988) Effects of ants and different oxide surfaces, Environ. Sci. Technol. 30, acidification on mercury methylation, demethylation, and vo- 2484–2494. latilization in sediments from an acid-susceptible lake, AppL. Environ. Microbiol. 54, 2003-2009. [9] Romera E., González F., Ballester A., Blázquez M.L., Munoz J.A. (2007) Comparative study of biosorption of heavy metals [26] Monperrus M., Tessier E., Amouroux D., Leynaert A., Hu- using different types of algae, Bioresour. Technol. 98, 3344– onnic P., Donard O.F.X. (2007) Mercury methylation, demeth- 3353. ylation and reduction rates in coastal and marine surface waters of the Mediterranean Sea, Marine Chemistry. 107, 49–63. [10] Romera E., González F., Ballester A., Blázquez M.L., Munoz [27] Zepp R.G., Schlotzhauer P.F. (1983) Influence of algae on J.A. (2008) Biosorption of heavy metals by fucus spiralis, Bi- photolysis rates of chemicals in water, Environ. Sci. Technol. oresour. Technol. 99, 4684–4693. 17, 462-468. [11] Ozudogru, Y. (2013) Biosorption of Cu (II) and Pb (II) ions [28] Lesa B., Aneggi E., Rossi G., Comuzzi C., Goi D. (2009) using marine brown algae padina pavonica. Fresenius Environ. Bench-scale tests on ultrasound-assisted acid washing and Bull. 22(12A), 3725-3729. thermal desorption of mercury from dredging sludge and other [12] Sahmurova, A. Turkmenler, H. Gulnaz, O. (2013) Heavy solid matrices, Journal of Hazardous Materials. 171, 647-653. metal removal by using dried enteromorpha compressa [29] Èernocká H., Ostatná V., Paleèek E. (2013) Enzymatic activity macroalgae from aqueous solution and industrial wastewater and catalytic hydrogen evolution in reduced and oxidized ure- application. Fresenius Environ. Bull. 22(6):1659-1665. ase at mercury surfaces, Analytica Chimica Acta. 789, 41-46. [13] Deng L., Fu D.F., Deng N.S. (2009) Photoinduced transfor- [30] Pei H.Y., Hu W.R., Guo G.J., Liu P.Q. (2004) The characters mations of mercury (II) species in the presence of algae, Chlo- of algae killing with chlorine dioxide, J. Shandong Univ. Eng. rella vulgarism, Journal of Hazardous Materials. 164, 798– Sci. 34, 105–108. 805. [31] Pei H.Y., Hu W.R. (2003) Study on the effectiveness and mu- [14] Deng L., Wang H.L, Deng N.S. (2006) Photoreduction of tagenicity of algae killing with ozone, Ind. Water Treat. 23, Chromium(VI) in the presence of algae, Chlorella vulgarism, 55–57. Journal of Hazardous Materials, 138, 288-292. [32] Wang X.H., Sun R.Z., Wang C.Y. (2014) PH dependence and [15] Zhang H., Lindberg S.E. (2000) Air/water exchanges of mer- thermodynamics of Hg(II) adsorption onto chitosan-poly(vi- cury in the Everglades I: The behavior of dissolved gaseous nyl alcohol) hydrogel adsorbent, Colloids and Surfaces A: mercury (DGM), Sci. Total Environ. 259, 123-133. Physicochemical and Engineering Aspects. 441, 51-58. [33] Matthiessen A. (1998) Reduction of divalent mercury by hu- [16] Lindberg S.E., Vette A.F., Miles C., Schaedlich F. (2000) Ap- mic substances-kinetic and quantitative aspects, The Science plication of an automated mercury analyzer to field speciation of the Total Environment. 213, 177- 183 measurements: Results for dissolved gaseous mercury in natural waters, Biogeochemistry. 48, 237-259. [34] Berkovic A.M., Bertolotti S.G., Villata L.S., Gonzalez M.C., Diez R.P., Mártire D.O. (2012) Photoinduced reduction of di- [17] Zhang H. and Lindberg S.E. (2002) Dissolved gaseous mer- valent mercury by quinones in the presence of formic acid un- cury in Whitefish bay and the Taquemenon River watershed in der anaerobic conditions, Chemosphere. 89, 1189-1194. the Michigan Upper Peninsula: Distribution and dynamics, Water, Air, Soil, Pollut. 133, 379-389. [35] Zhang H., Lindberg S. (2001) Sunlight and Iron(III)-Induced Photochemical Production of Dissolved Gaseous Mercury in [18] Amyot M., Mierle G., Lean D., McQueen D.J. (1997) Effect of Freshwater Environ, Environ. Sci. Technol. 35, 928-935 solar radiation on the formation of dissolved gaseous mercury in temperate lakes, Geochim. Cosmochim. Acta. 61, 975-987. [36] Ababneh F.A., Scott S.L., Al-Reasi H.A., Lean D.R.S. (2006) Photochemical reduction and reoxidation of aqueous mercuric [19] Brazeau M.L., Poulain A.J., Paterson A.M., Keller W., Sanei chloride in the presence of ferrioxalate and air, Science of The H., Blais J.M. (2013) Recent changes in mercury deposition Total Environment. 367, 831–839. and primary productivity inferred from sediments of lakes from the Hudson Bay Lowlands, Ontario, Canada, Environ- [37] Vaughan P.P. and Blough N.V. (1998) Photochemical For- mental Pollution. 173, 52-60. mation of Hydroxyl Radical by Constituents of Natural Wa- ters, Environ. Sci. Technol. 32, 2947-2953. [20] Vandal G.M., Mason R.P., Fitzgerald W.F. (1991) Cycling of [38] Faust B.C., (1994) In Aquatic and surface photochemistry, volatile mercury in temperate lakes, Water Air Soil Pollut. 56, Helz G.R., Zepp R.G., Crosby D.G., Eds. Lewis Publishers: 791-803. Boca Raton, FL, pp.3-37. [21] Lindberg S.E., Zhang H., Meyers T.P. (1999) Final report: Ev- [39] Stumm W. and Morgan J.J. (1996) Aquatic chemistry: Chem- erglades Mercury Air/ Surface Exchange Study (E-MASE), ical equilibria and rates in natural waters, John Wiley & Sons: South Florida Water Management District: West Palm Beach, New York, pp.726-759. FL., pp.41-48. [40] Wang Y., Zhang P.Y, Pan G., Chen H. (2008) Ferric ion me- [22] Jerzykiewicz M. (2013) The effect of Hg(II) ions on the free diated photochemical decomposition of perfluorooctanoic acid radicals of humic substances and their model compounds, (PFOA) by 254 nm UV light, Journal of Hazardous Materi- Chemosphere. 92, 445–450. als.160, 181–186. [23] Kritee K., Blum J.D., Reinfelder J.R., Barkay T. (2013) Mi- [41] Hug S.J., Laubscher H.U., James B.R. (1997) Iron (III) Cata- crobial stable isotope fractionation of mercury: A synthesis of lyzed Photochemical Reduction of Chromium (VI) by Oxalate present understanding and future directions, Chemical Geol- and Citrate in Aqueous Solutions, Environ. Sci. Technol. 31, ogy. 336, 13–25. 160-170.

333 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

[42] Zuo Y.G and Hoigne J. (1992) Formation of hydrogen peroxide and depletion of oxalic acid in atmospheric water by photolysis of iron(III)-oxalato complexes, Environ. Sci. Technol. 26, 1014-1022.

Received: April 23, 2014 Revised: June 04, 2014 Accepted: July 03, 2014

CORRESPONDING AUTHOR

Lin Deng and Caiqian Yang School of Civil Engineering Southeast University Nanjing 210096 P.R. CHINA

Phone: + 86-25-8768-6550 Fax: +86-25-87686550. E-mail: [email protected] or [email protected]

FEB/ Vol 24/ No 1b/ 2015 – pages 324 - 334

334 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

PHYSICAL PROPERTIES AND FIBER DIMENSION IN STEM, BRANCH AND ROOT OF ALDER WOOD

Majid Kiaei1,* and Roger Moya2

1Young Researcher and Elite Club, Chaloos Branch, Islamic Azad University, Chaloos, Iran 2 Instituto Tecnológico de Costa Rica; Escuela de Ingeniería Forestal;P.O. Box: 159-7050, Cartago-Costa Rica;

ABSTRACT searchers [4, 5]. There are positive relationships between- fiber length and paper burst strength [6-9], tensile strength The aim of this study was to determine physical prop- [6, 7], tear strength [6, 10] and folding endurance [9, 11]. erties and fiber dimensions in stem, branch and root wood Anatomical properties of root wood are different than stem for alder (Alnus glutinosa L) species. For this purpose, wood and branch wood. Root wood has a high vessel num- three normal alder trees were selected from Khanican for- ber, wider diameter of vessel, and low fiber volumes [12]. est in north of Iran. Disks were taken from three parts such Chemical composition analysis showed that the cellulose as stem, branch and root of trees. Testing samples were ran- content in root wood is higher than stem wood. Average of domly taken at disk surfaces to examine the physical prop- lignin and extractive component in root wood is very high erties (according to the ISO standard for oven-dry density [13]. and volumetric shrinkage) and fiber dimension (fiber length, fiber diameter and lumen diameter according to the Frank- Different results have emerged from studies carried lin method). Analysis of variance (ANOVA) indicated that out to assess the properties of root, stem and branch wood. the wood samples had significant difference on the wood Wood density in root wood is lower than other wood sam- density, volumetric shrinkage, fiber length, fiber diameter ples (stem and branch wood) in many of hardwoods species and lumen diameter. The highest wood density, volumetric [14]. Whereas, wood density and mechanical strength shrinkage and fiber length was found in stem wood, whereas properties in root wood is higher compared to branch wood root wood had a lighter density, larger fiber diameter and and stem wood in African hardwood species [15]. Okai et larger fiber lumen diameter. The branch wood had the lowest al. (2004) [16] on the physical and mechanical properties fiber dimensions (fiber length, fiber diameter and lumen of stem and branch wood of tropical trees in Ghana re- diameter) compared to other wood sample parts. The rela- ported that the branches of the two species emire (Termi- tionship between wood density and volumetric shrinkage nalia ivorensis) and asanfina (Anigeria robusta) had higher were determined by regression model. These relationships densities and modulus of rupture (MOR) than their corre- in root wood were very weaker compared to stem wood sponding stems. Amoah et al (2012) reported that the root and branch wood. wood of iroko and emire exhibited the highest basic density compared to branch and stem wood [15]. Research on the tracheid dimension in root wood of southern pine reported

KEYWORDS: Alnus glutinosa, fiber length, fiber diameter, fiber lu- that the rootwood tracheids were one-third longer, averag- men diameter, stem, branch, root. ing 3.99 mm as compared to 2.97 mm for stumpwood tracheids. In transverse dimensions root tracheids were larger in diameter and had thinner walls and larger lumens than

1. INTRODUCTION stem tracheids from the same stump. Cell diameter averaged one-third greater (52.48 μm) in root wood than in stump The cell size and relative cell dimensions have a major wood (38.76 μm). Average thickness of cell walls was influence on the quality of pulp and paper products and on 5.40 μm in roots and 6.55 μm in stumps. Lumen diameters solid wood products [1-2]. The fiber length and width, wall averaged 41.69 μm in roots and 25.66 μm in stumps [17]. thickness, and lumen size have an effect on the bulk, burst, There are no reports on wood different properties tear, fold, and tensile strengths of paper [3]. However, there among root, branch and stemwood for Iranian hardwood are definite correlations between tracheid length and cer- species especially for alder wood. Therefore, the aim of the tain lumber mechanical properties [4]. A positive and present study was: a) to determine the physical properties strong relationship between MOR and the tracheid length and fiber dimension properties in the three wood samples of hardwood and soft species was found by several re- parts such as stem, branch and root wood for alder species (Alnus glutinosa) and to examine the relationship between * Corresponding author wood different properties in Mazandran region of Iran.

335 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

2. MATERIALS AND METHODS dry weights of the specimens were measured. The wood physical properties were calculated using the following 2.1 Wood samples equations: Three alder trees (Alnus glutinosa) with breast height diameters 20, 21.5, and 22 cm, with 37, 39 and 39 growth rings at breast height, respectively were harvested in the western part of Mazandran region, north of Iran. Discs were taken from three part of each tree include stem (at breast height), branch and root parts. After removing pith part for every disc, sample testing was randomly prepared where , , , and are the oven-dry at disc surface (after 20 years – old for stem wood). Thirty density, weight and volume of the specimen (in dry state), samples for each part were used to determine the physical volumetric shrinkage and volume of the saturated speci- properties and fiber dimensions (3 part × 30 samples = mens, respectively. 90 according to ISO standard. Examination of fiber dimension was carried out after the physical properties had been 2.4 Statistical analysis determined. One-way analysis of variance (ANOVA) was carried 2.2 Fiber dimensions out to establish the effect types of wood sample parts on The material was macerated using the Franklin method physical properties and fiber dimension and the differences (1946) [18] to examine fiber dimensions. The pieces of fiber were were quantified through the Duncan test (p ≤ 0.05). In order cut with a scalpel and placed in test tubes in a solution of 1:1 acetic acid and to examine relationship between various properties were oxygenated water of 30 volumes. used from regression models. In order to examine relation- The samples were dried in a stove at 60 ºC for approximately 1 week. The ship between various properties were used from regression disaggregated particles were washed in water, stained with aqueous safranin at 1% for 3 min, dehydrated with alcohol at 96% and xylol. Subsequently, the models. fibers were dried, placed on slides and fixed with Canada balsam. From a samples, 30 fibers (3 part ×30 samples × 30 fiber =2700 fiber) were selected to determine fiber dimension (fiber 3. RESULTS AND DISCUSSION length, fiber width, and fiber lumen diameter), which used from Leica Image Analyzer System (Figure 1). 3.1 Fiber length The average fiber length in root wood, branch wood 2.3 Physical properties and stem wood are shown in Figure 2. The fiber length av- The specimens were soaked in distilled water for 72 h eraged 1.31± 0.188 mm in stem, 0.940 ± 0.128 mm in to ensure that their moisture content was above the fiber branch and 1.25 ± 0.173 mm in root wood. Maximum and saturation point, and then their dimensions were measured minimum of fiber length in stem, branch and root alder in all three principal directions, with a digital caliper to the wood were 1.02-1.64, 0.639-1.12 and 0.890-1.44 mm, re- nearest 0.001 mm. The specimens were weighed to the spectively. The analysis of variance (ANOVA) indicated nearest 0.001 g for saturated weight and the saturated vol- that the effect of wood sample parts on the fiber length was ume was calculated based on these dimension measurements. significant (F=22.346, Sig=0.002). The highest and lowest Finally, the samples were oven-dried at 103 ± 2 °C to 0% of fiber length were found in stem and branch, respectively. moisture content. After cooling in desiccators, the oven- No statistically significant difference was detected between

FIGURE 1 - Fibers of stem wood in Alnus glutinosa

336 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

FIGURE 2 - The average of fiber length in stem, branch and root wood

FIGURE 3 - The average of fiber diameter in stem, branch and root wood

stem and root wood samples in fiber length. These results 3.2 Fiber diameter were previously reported by several researchers. For exam- The average fiber diameter in root wood, branch wood ple, Manwiller (1972) [17] and Dunham et al. (2007) [19] and stem wood are shown in Figure 3. The fiber diameter reported that the branch wood have shorter tracheid/fiber averaged 32.87 ± 6.41 µm in stem, 26.92 ± 3.16 µm in than the stem wood. Generally, there are different results branch and 42.52 ± 9.15 µm in root wood. Maximum and about fiber length in stem, branch and root wood. Fegel minimum of fiber diameter in stem, branch and root alder (1941) [20] and Dunham et al. (2007) [19] reported that wood were 26.08-49.05, 20.34-31.83 and 24.14-59.72 µm, root wood may have shorter or longer tracheids than the respectively. The analysis of variance (ANOVA) indicated bole, or Psaras and Sofroniou (2004) [21] reported that the that the effect of wood sample parts on the fiber diameter fiber length in root wood (620 µm) is higher than stem was significant (F=20.873, Sig=0.004). The highest and wood (590 µm) for Phlomis fruticosa species. lowest of fiber diameter were found in root and branch wood, respectively.

337 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

3.3 Lumen diameter in root and branch wood, respectively. Results of lumen di- The average of lumen diameter in root wood, branch ameter and fiber diameter are in agreement with a study wood and stem wood are shown in Figure 4. The fiber lu- carried out by Manwiller (1972) [17]. According to the re- men diameter averaged 19.73 ± 5.92 µm in stem, 14.40 ± sults of lumen diameter and fiber diameter it can be said that 3.28 µm in branch and 30.10 ± 7.8 µm in root wood. Max- the fiber cell wall thickness in stem wood was higher com- imum and minimum of fiber lumen diameter in stem, pared to the branch wood and root wood in alder species. branch and root alder wood were 12.96-31.83, 9.68-19.52 and 12.11-49.87 µm, respectively. The analysis of variance 3.4 Wood density (ANOVA) indicated that wood sample parts had signifi- The average oven-dry density in root wood, branch cant effect on the lumen diameter (F=20.205, Sig=0.005). wood and stem wood are shown in Figure 5. Among the The highest and lowest of fiber lumen diameter were found wood types of alder, the root wood recorded the lowest av-

FIGURE 4 - The average of lumen diameter in stem, branch and root wood

FIGURE 5 - The average of oven-dry density in stem, branch and root wood

338 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

FIGURE 6 - The average of volumetric shrinkage in stem, branch and root wood

erage oven-dry density (0.333 ± 0.021 g/cm3), ranging be- 3.5 Volumetric shrinkage tween 0.294 and 0.375 g/cm3. Stem wood recorded the high- The average volumetric shrinkage in root wood, branch est (0.496 ± 0.038 g/cm3), varying between 0.444 and 0.595 wood and stem wood are shown in Figure 6. The volumetric g/cm3. The average wood density in branch wood was 0.492 shrinkage averaged 13.44 ± 1.50 % in stem, 13.23 ± 1.58 % ± 0.024 g/cm3, varied ranging between 0.458 and 0.559 g/cm3. in branch and 11.05 ± 1.71 % in root wood. Maximum and The analysis of variance (ANOVA) indicated that wood sam- minimum of volumetric shrinkage in stem, branch and root ple parts had significant effect on the oven-dry density wood were 11.15-17.41, 11.44-16.45 and 7.74-13.77 %, re- (F=202.223, Sig=0.001). There was no significant difference spectively. The analysis of variance (ANOVA) indicated that between branch wood and stem wood in oven-dry density. wood sample parts had significant effect on the volumetric Several scientific studies have reported that the shrinkage (F=17.408, Sig=0.002). The highest and lowest vessel diameter and vessel length of root wood is higher of volumetric shrinkage were found in stem and root wood, than that of branch wood and stem wood [12, 21]. There respectively. There was no significant difference between also are inverse relationship between wood density and branch wood and stem wood in volumetric shrinkage. volumetric shrinkage in alder wood [14, 22, 23]. Therefore, The relationship between wood density and volumetric decreasing the wood oven-dry density in root wood can be shrinkage in three parts of trees (stem, branch and root) are related to its higher vessel diameter. shown in Figure 7-9. Results indicated that shrinkage is

FIGURE 7 - The relationship between density and shrinkage in stem wood

339 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

FIGURE 8 - The relationship between oven-dried density and shrinkage in branch wood

FIGURE 9 - The relationship between oven-dried density and shrinkage in root wood

TABLE 1 - Pearson correlation between fiber dimension and oven dried density in alder wood

Wood samples Stem Branch Root Total Fiber length 0.114 -0.152 -0.441 -0.325* (FL) Fiber diameter (FD) -0.076 0.252 0.305 -0.615**

Lumen Diameter (LD) -0.245 -0.032 0.189 -0.647** * and ** were significant at 0.01 and 0.05%

positively and significantly related to wood density in stem and shrinkage for root wood for alder wood. The volumetric and branch wood, whereas this relationship found in root shrinkage and swelling properties are affected by several wood is positive but not significant. The correlation coeffi- wood factors, such as the heartwood to sapwood ratio or the cient between wood density and shrinkage in branch wood fibrillar angle on the S2 layer [24]. However, the most im- (R2=0.449) is lower compared to the stem wood (R2=0.586). portant parameter affecting wood shrinkage is the wood den- Greater variation was found in relationship between density sity [25]. This present research indicated that the shrinkage

340 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

in stem wood is higher than that of branch wood and root [2] Yilmaz1 M, Serdar B, Altun L and Usta A. 2008. Relationship wood. High wood density have proportionately more cell between environmental variables and wood anatomy of Quer- cus pontica C. KOCH (Fagceae). Fresenius Environmental and less lumen volume and they shrink and swell more due Bulletin 17 (7b): 902-910 to the unique nature of the microstructures. [3] Clark J (1962) Effects of fiber coarseness and length, I. Bulk, burst, tears, fold and tensile tests. Tappi J 45: 628-634. The effect of fiber dimension on the wood density of three parts of alder wood was determined by Pearson cor- [4] Zobel BJ, van Buijtenen JP (1989) Wood Variation: Its Causes relation (Table 1). Significant differences were no found and Control. Springer-Verlag, Berlin, Heidelberg, New York. between FL, FD and LD with wood density for each of [5] Omidvar AA (1987) The impact of annual ring width on tech- wood samples, while in total of wood samples, there are nological quality of pine wood (Pinus taeda). M.A. thesis. Noor University of Modares-Training, pp. 200. significantly negative relationship between fiber length (r =- 0.325), fiber diameter (r = -0.615) and lumen diameter [6] Casey JP (1952) Properties of paper and converting. Pp. 835– (r = -0.647) with wood oven-dried density. Two most ef- 837 in Pulp and Paper Chemistry and Chemical Technology. Volume 2. Interscience Publisher Inc., New York. fective variables on the oven-dry density in alder wood are fiber diameter (FD) and lumen diameter (LD). The highest [7] Miyake M (1968) Wood characteristics and kraft pulp properties of hardwood grown in Hokkaido. Japan Tappi 22: 600– of correlation coefficients between FL-density and FD – 610. density was found in root wood, while the relationship between LD-density in the stem wood is higher compared to [8] El-Hosseiny F, Anderson D (1999) Effect of fibre length and coarseness on the burst strength of paper. Tappi J 82: 202–203. branch wood and stem wood. [9] Ona T, Sonoda T, Ito K, Shibata M, Tamai Y, Kojima Y, Ohshima J, Yokota S, Yoshizawa N (2001) Investigation of relationship between cell and pulp properties in Eucalyptus by 4. CONCLUSION examination of within-tree property variations. Wood Sci Tech 35: 363–375. This research examines the effect of wood sample [10] Haygreen JG, Bowyer JL (1996) Forest Products and Wood parts on the physical properties and fiber dimensions of al- Science: An Introduction. Third edition. Iowa University der wood in north of Iran. The analysis of variance Press, Ames. (ANOVA) indicated that the wood samples had significant [11] Dinwoodie JM (1965) The relationship between fiber mor- difference on the fiber length, fiber diameter, and fiber lu- phology and paper properties: a review of literature. Tappi J men diameter. Root wood had the highest fiber diameter 48: 440–447. and fiber lumen diameter values. The lowest fiber dimen- [12] Bhat K. M (1982) A note on cellular properties and basic density sions were found in the branch wood. The highest and low- of lateral roots in birch. IAWA Journal 3 (2):89-94. est wood density and volumetric shrinkage values were [13] Howard ET (1973) physical and chemical properties of slash found in stem wood and root wood, respectively. The wood pine tree parts. Wood Science 5 (4): 312-317. density and shrinkage in branch wood is close to the stem [14] Machado SR, Angyalossy-Alfonso VA, De Morretes BL wood. There are positive relationship between wood den- (1997) Comparative wood anatomy of root and stem in Styrax sity and shrinkage, whereas these relationships in root camporum (Styracaeae). IAWA Journal 18 (1):13-25. wood were weaker than stem wood and branch wood. [15] Amoah M, Appiah-Yeboahand, Okai R. (2012). Characteriza- These results indicated that Iranian root alder wood fibers tion of physical and mechanical properties of branch, stem, are promising fibrous raw material for the fiberboard and and rootwood of Iroko and Emire tropical tress. Research Jour- paper production due to long fiber length, while had a low nal of Applied Science, Engineering and Technology 4 (12): density. There is a lack of wood resources in Iran. Root 1755-1761. wood fibers can solve these problems. [16] Okai, R., K. Frimpong-Mensah and D. Yeboah, 2004. Charac- terization of strength properties of branchwood and stemwood of some tropical hardwood species. Wood Sci. Technol., 38: 163-171. ACKNOWLEDGEMENTS [17] Manwiller F (1972) Tracheid dimensions in rootwood of southern pine. Wood Science 5(2): 122-124. This article is derived from a M. Sc thesis of Fariborz Alinejad and Hossein Mahmoodkhani. The authors appre- [18] Franklin F.L. (1946). A rapid method for softening wood for microtome ciate the support receive from the Chaloos Branch, Is- sectioning. Trop. Woods Yale Univ. Sch. For. 88, 35-36. lamic Azad University, Chaloos, Iran. [19] Dunham S. M., Lachenbruch B, and Ganio LM. 2007. Bayes- ian analysis of Douglas-fir hydraulic architecture at multiple The authors have declared no conflict of interest. scales. Trees 21:65–78. [20] Fegel A. C. 1941. Comparative anatomy and varying physical properties of trunk, branch, and root wood in certain northeastern trees. Bulletin of the N.Y. State College of Forestry at Sy- REFERENCES racuse University. Tech. Pub., 55 (14):5–20. [21] Psaras GK, Sofroniou I (2004) Stem and rootwood anatomy of [1] Bisset IJ, Dadswell HE,Wardrop B (1951) Factors influencing the shrub Phlomis fruticosa (labiatae). IAWA Journal 25 (1): tracheid length in conifer stems. Aust For 15: 17-30. 71-77.

341 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

[22] Searson JM, Thomas DS, Montagu KD and Conroy JP. 2004. Wood density and anatomy of water-limited eucalypts, Tree Physiology 24, 1295–1302.

[23] Panshin AJ and de Zeeuw C (1980) Text book of wood tech- th nology, 4 edition, Mc Graw Hill,NewYork. [24] Bektaş İ, Güler C (2001) The determination of some physical properties of beech wood (Fagus orientalis Lipsky) in the Andırın region. Turk Agric For J 25: 209-215. [25] Guler C, Copur Y, Akgul M, Buyuksari B (2007) Some chemical,physical andmechanical properties of juvenile wood from- black pine (Pinus nigra Arnold) plantations. J Appl Sci 7: 755- 758.

Received: April 28, 2014 Revised: July 07, 2014 Accepted: July 21, 2014

CORRESPONDING AUTHOR

Majid Kiaei

Young Researcher and Elite Club,

Chaloos Branch

Islamic Azad University

Chaloos, Iran

IRAN

E-mail: [email protected]

FEB/ Vol 24/ No 1b/ 2015 – pages 335 - 342

342 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

INDUCED MUTAGENESIS AND GENOTOXICITY BY ACCUMULATED RADIONUCLIDES IN SOME EDIBLE PLANTS CULTIVATED IN BLACK SAND SOIL DETECTED BY RAPD AND SDS-PAGE

Ahmad K. Hegazy1,2, Mohammad Faisal1,*, Abdulrahman A. Alatar1, Hanan F. Kabiel2 and Mona H. Emam2

1Department of Botany and Microbiology, College of Science, King Saud University, Riyadh, Saudi Arabia 2Botany Department, Faculty of Science, Cairo University, Giza, Egypt;

ABSTRACT nomically important because of high concentrations of uranium and thorium in their structures [1, 2]. Abnormal oc- Assessment of mutagenic effects and DNA damage in- currences of uranium and its decay products in rocks and duced by radioactive elements at molecular level is im- soils, and thorium in monazite sands are the main sources portant in ecotoxicology. In the present study, the radionu- of high natural background areas that have been identified clides content was determined in four crop plant species, in several areas of the world [3]. The study of U and Th namely Eruca sativa M., Solanum lycopersicum L. Pasid- transfer from soil to edible plants through root uptake is ium guajava L. and Mangifera indica L. cultivated in the very important, especially considering accumulation of black sand soil. The results revealed the ability of the plants these radionuclides in the food chains [4]. Radioactive ele- to accumulate a considerable amount of uranium and tho- ments and heavy metals are chemical elements. On the ba- rium in the edible portions. The absorbed radionuclide in- sis of their chemical toxicity, generating reactive oxygen duced damages in DNA as indicated by variations in band species, thus leading to DNA damage and lipid peroxidation intensities, loss of normal band and appearance of new of membranes [5]. Radionuclides can induce deleterious ef- bands in RAPD profiles compared to plants with low radi- fects on biota via both chemical and radiological pathways. onuclides content. The high level of uranium and thorium Because of its dual toxicity, U represents a relatively unique in plant tissues caused some changes in SDS-PAGE of pro- challenge to (eco) toxicologists. However, its toxicity has teins reflected by alterations in the band relative mobility been extensively studied for non-human biota [6]. and intensities, expression of new proteins and suppression of some proteins. It is concluded that high uranium and tho- Most studies on the plant genotoxicity related to that rium content induces mutagenesis and genotoxicity in the induced by heavy metals such as Cd, Hg, Cr and Zn but the four crop plant species. The DNA polymorphism detected studies on the genotoxicity of radionuclides is so rare and by RAPD analysis and SDS-PAGE of proteins could be used mostly carried out on mammals. Genotoxic effects of Cd as investigation tool for environmental toxicology and as a on barley (Hordeum vulgare L) seedling were studied by useful biomarker for the detection of genotoxicity caused by Liu et al. [7] and the results revealed that genomic template the radionuclides pollution in plants. stability was significantly affected at high concentration of Cd. Also DNA damage in the roots and leaves of bean (Phaseolus vulgaris L) seedling was detected by Cenkci et

KEYWORDS: Ecotoxicology; proteomics; thorium; uranium; DNA al. [8] after uptake of Hg, B, Cr and Zn. Study of genotox- fingerprinting; protein profile icity and mutagenic effects of uranium and thorium was carried out by Soliman et al. [9] on some wild plants natu-

rally grow in black sand soil and the study showed high 1. INTRODUCTION cytotoxicity and genotoxicity in the studied plant populations indicated by high percentage of chromosomal aberra- In Egypt, the occurrences of black sand deposits are tions at different mitotic stages in addition. Moreover ab- known at the Nile mouth near Damietta and Rosetta and sorbed radionuclides induced changes in SDS-PAGE of have been worked for their heavy mineral content such as- protein indicated by expression of new proteins and sup- magnetite, ilmenite, zircon and monazite. They are eco- pression of some proteins in the studied plants.

The chemical toxicity of radioactive elements may be * Corresponding author similar to that of heavy metals; however, an ionizing activ-

343 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

ity associated with radioactive decay can result in addi- (Mangifera indica L.) in the black sand of Baltim region tional toxic effects on the plant [10]. The detrimental effect and to study the mutagenic responses that may be induced on life caused by trace metals can be exerted through their by the accumulated radionuclides in the edible portions of capability of combining with biomolecules, especially pro- these plant species. teins and may so act as potent enzyme inhibitors hampering biochemical processes and compromising DNA and cell membrane integrity [11]. The relationship between chemi- 2. MATERIALS AND METHODS cal pollution and alterations in the biochemical processes of exposed organisms is of high importance in environmen- 2.1. Study sites tal research. Molecular data can provide information for The present study was conducted in the black sand de- understanding the environmental impact of chemical posits in Baltim, Mediterranean coast of Egypt. Baltim site is stressors and help to find suitable biomarkers for exposure located on one of the high hills around Lake El- Burullus be- [12]. Because even essential metals can be toxic to living tween longitudes 31° 5′ E and latitudes 31° 35′ N (Figure 1). organisms, numerous genes tightly regulate their intracel- Black sand deposits are rich in heavy minerals, e.g., zircon lular concentrations. Since metals are not metabolized, dif- (ZrSiO4), rutile (TiO2) and ilmenite (FeTiO3). They are ferent strategies of detoxification have been developed dur- economically important and are characterized by high con- ing the evolution to cope with their toxicity. In the cytosol, centrations of thorium (232Th) and uranium (238U) in their they can be conjugated to glutathione, amino acids, metal- crystalline structure [16, 17]. Physical and chemical prop- lothioneines or phytochelatins to be in the last step either erties of the soil were previously determined [18]. The sequestrated in a less sensitive compartment such as the black sand soil texture is composed of coarse sand vacuole or excreted out of the cell [5]. (78.12%) with low fine sand (2.38%), silt (0.21%) and clay Toxic chemicals induce several cellular stress re- (1.30%) content. As for control soil, the coarse amounted sponses and damage to different cellular components such 54.57%, fine sand 11.80%, silt 6.46% and clay 27.18%. as membranes, proteins and DNA [13]. Several studies The soil reaction was slightly alkaline in both black sand have used the chromosome aberration, micronucleus, or and control soils, with pH values 7.27 and 7.41, respec- comet assays to measure the effects of genotoxins on plants tively. The electric conductivity was higher in control soil -1 -1 [14]. Advantages of measuring effects of genotoxic chem- (0.43 dSm ) than in black sand soil (0.36 dSm ). Simi- icals directly on DNA are mainly related to the sensitivity larly, the organic matter content reached 3.66% in control and short response time. Recently, advances in molecular soil and 0.24% in the black sand. biology have led to the development of a number of selec- 2.2. Plant material and determination of radionuclide content tive and sensitive assays for DNA analysis in eco-genotoxicology. Random amplified polymorphic DNA (RAPD) The soil and the edible portions of the studied crop can be used to detect genotoxicity and differences in RAPD plants and fruit trees were collected from the study site of profiles can clearly be shown when comparing DNA fin- Baltim and also from a control soil in which these plant gerprints from untreated and treated individuals to geno- species cultivated. The collected edible samples were thor- toxic agents [8]. oughly washed and rinsed with distilled water. The samples were then cut into small slices and dried in oven at Proteomics, or the systematic analysis of the proteins 60°C. The dried samples were then ground into fine pow- expressed by a genome, is a powerful tool for not only de- der and stored in fresh plastic polythene bags ready for fur- scribing complete proteomes at the organelle, cell, organ or ther use. For estimation of radionuclides content in the soil, tissue levels, but also comparing proteomes as affected by the concentration of uranium and thorium was chemically different physiological conditions, e.g., those resulting determined using spectrophotometric technique as de- from exposure to heavy metals or several other stressful scribed by Marczenko [19] at Uranium & Thorium Labor- environmental factors. Most of the proteomic analyses that atory, Central Labs, Anshas, Nuclear Materials Authority. have been performed to study the effect of heavy metals in The spectrophotometric determination was based on meas- different organisms have shown that proteins related to an- uring the absorption of the complex compounds with Arse- tioxidative defense mechanisms are differentially regulated nazo III or Thorin in the visible part of the spectrum from in response to metal toxicity [15]. 400 to 800 nm. Measurements were performed according We set out the hypothesis that crop plant species or to blind samples. The samples were prepared in flasks of fruit trees raised in radionuclide-contaminated soil may 10 mL by adding of the specific sample volume and 0.8 mL take up and accumulate radionuclides in the edible organs. of 0.05% solution of Arsenazo III or 1.2 mL of 0.1% solu- These absorbed radioactive elements may cause mutagenic tion of thorin. The pH values were adjusted by adding conc. −1 effects and genotoxicity on the plants. To test the formu- 10% NaOH or 5 mol L HNO3 or HCl, and the flasks were lated hypothesis, the present study was conducted to assess refilled with water, methanol, or ethanol. Uranium and tho- the radionuclides uptake by the two commonly cultivated rium content in the edible portions of the studied plant spe- crop plants, viz. Rocket (Eruca sativa Mill.) and Tomato cies were determined by high resolution inductively cou- (Solanum lycopersicum L.), and the two commonly culti- pled plasma mass spectrometer (ICP-MS technique) model vated fruit trees Guava (Psidium guajava L.) and Mango Jeol-JMS-PLASMAX2, which installed at the Central La-

344 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

FIGURE 1 - Map of northern part of Egypt showing the black sand deposits (dotted coastal strip) and the study site of Baltim.

boratory for Elemental and Isotopic Analysis, Nuclear Re- amplification with denaturation at 95 ºC for 1 min, anneal- search Center, Atomic Energy Authority in Egypt. Before ing at 36 ºC for 1 min and extension at 72 ºC for 2 min. The analysis, the ICP-MS measuring conditions were optimised final extension was carried out at 72 ºC for 5 min. The am- using a solution containing uranium and Thorium at a con- plified DNAs were separated using electrophoresis unit centration of 1 μg L−1. Calibration standards for ICP-MS (WIDE mini-sub-cell GT Bio-RAD, USA) on 1% agarose determination were prepared by dilution of 10 mg L−1 ura- containing ethidium bromide (0.5 µg/ml) at constant volt nium and thorium. The measured uranium and thorium and determined with UV transilluminator. concentrations in soil and plant samples were obtained in ppm and converted to unite of BqKg-1 dry weight by mul- TABLE 1 - Sequence of random primers used in RAPD-PCR. tiplying the value of uranium by 12.34 and multiplying the Primer code Sequence (5ʹ-3ʹ) value of thorium by 4.0404. Primer 1 GGTGCGGGAA Primer 2 GTTTCGCTCC 2.3. Genomic DNA extraction Primer 3 GTAGACCCGT DNA was isolated from 50 mg of leaf tissues using Qi- Primer 4 AAGAGCCCGT agen Kit for DNA extraction by a modified CTAB method Primer 5 AACGCGCAAC [20]. The extracted DNA was dissolved in 100 µl of elution Primer 6 CCCGTCAGCA Primer 7 CAGTAGCGGA buffer. The concentration and purity of the obtained DNA Primer 8 GGTGGTGATG was determined by using “Gen qunta” system-pharmacia Primer 9 GACTAAGCCC Biotech. The purity of the DNA for all samples was be- Primer 10 TTCAGGGCAC tween 90-97%. Concentration was adjusted at 6 ng/µl for Primer 11 GGTCGGAGAA all samples using TE buffer pH 8. Primer 12 TGCTCTGCCC Primer 13 AGCGTGTCTG Primer 14 TCCGCTCTGG 2.4. RAPD analysis Primer 15 TCCCACGCAA A total of 18, 10 mer oligonucleotide random primers were Primer 16 GGACCCTTAC used for RAPD analysis (Table 1). DNA amplification re- Primer 17 CATCCCCCTG actions were performed in 25 µl reaction mixture consist- Primer 18 TTCCCCCGCT ing of 1 unit of Taq DNA polymerase, 0.2 mM dNTP, 1x PCR buffer, 3 mM MgCl2, 10 Pmol of each primer and ap- 2.5. Electrophoretic detection of protein by SDS-PAGE proximately 50 ng of the extracted genomic DNA. Ampli- Cytoplasmic proteins were extracted and purified from fication reactions were carried out using PCR unit II Bi- leaf tissues of the test species for SDS-PAGE analysis as ometra with the following thermal profile: 1 cycle of 95 ºC described by Nelson et al. [21]. One gm freeze dry plant for 5 min (initial denaturation), followed by 45 cycles of material was ground with liquid nitrogen with a pre-cold

345 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

mortar and pestle until the sample completely homoge- thorium content showed the values of 20.067 and 24.242 Bq nized. The grounded samples were transferred to 1.0 ml ep- Kg-1 dry weights respectively but these values were signifi- pendorf tube containing 200 µl extraction buffer (50m M cantly decreased to 7.946 and 1.912 Bq Kg-1 dry weight in the tis-HCl buffer, pH 6.8, glycerol 10% w/v, ascorbic acid control soil in which the two species were cultivated. Con- 0.1% and cysteine hydrochloride 0.1 w/v). Centrifugation centrations of uranium and thorium determined in the edi- at 18000 rpm for about 30 min was carried out at 4 0C to ble portions of the study species are shown in Figure 2. The remove debris. The protein content in the supernatant was uranium and thorium content in the plants cultivated in the estimated according to the method of Bradford [22] by us- black sand soil generally attained values higher than that in the ing bovine serum albumin as standard protein. Protein con- plants cultivated in the control soil. In the leaves of E. sativa tent was adjusted to 2 mg/ml per sample. The protein sam- uranium and thorium contents were 4.196 and 0.525 Bq Kg-1 ple was separated by SDS-PAGE on 11% polyacrylamide dry weights respectively in the plants cultivated in the control gel and electrophoresed at 50 volt for 3 hours according to the soil but these contents significantly increased to 12.957 and method of Laemmili [23]. Electrophoresis was performed in a 1.657 Bq Kg-1 dry weight respectively in the plants culti- vertical slab mold (Hoefer Scientific Instruments, San Fran- vated in the black sand. With respect to S. lycopersicum, cisco, CA, USA, model LKB 2001, measuring (16 x (18 x uranium and thorium contents in the fruits attained the val- 0.15 cm). The silver staining method as described by Sam- ues of 2.221 and 0.485 Bq Kg-1 dry weight respectively in mons et al. [24] was used for protein detection. the control soil, while in the plants of the black sand soil these values significantly increased to 16.659 and 1.535 Bq Kg-1 dry weight respectively. Uranium concentration in the 3. RESULTS fruits of P. guajava and M. indica plants cultivated in the control soil showed the values of 11.476 and 9.378 Bq Kg-1 3.1. Radionuclides content dry weight respectively in the but these contents signifi- Uranium and thorium contents in the study soil are cantly increased in the plants grown in the black sand of -1 summarized in Table 2. The concentration of uranium and Baltim (16.165 and 14.314 Bq Kg dry weight respec- thorium was significantly higher in the black sand than in tively). Fruits of these two species acquired lower levels of -1 the clay soil. With respect to uranium concentrations in the thorium (0.566 and 0.485 Bq Kg dry weight respectively) black sand of Baltim in which test crop species (Eruca sativa when cultivated in the control soil than the plants cultivated and Solanum lycopersicum) were cultivated were 76.919 in the black sand of Baltim as thorium content attained val- -1 and 95.429 BqKg-1 dry weight respectively which was sig- ues of 1.455 and 0.97 Bq Kg dry weight respectively. nificantly higher than that in the control soil in which the both species were cultivated; 45.247 and 46.892 BqKg-1 3.2. RAPD profiles of DNA dry weight respectively. Also uranium concentrations in For verifying the genetic effect of the absorbed radio- the black sand in which the test fruit tree species, P. nuclides, RAPD analysis was performed on DNA extracted guajava and M. indica, were cultivated attained the values from the tissues of the test plant species. The resulted of 86.791 and 81.855 Bq Kg-1 dry weight respectively. RAPD-DNA profiles of the four plant species are summa- While the control soil in which the two species were cultivated rized in Tables 3, 4, 5 and 6. With respect to E. sativa, of uranium content was 60.877 and 9.749 Bq Kg-1 dry weight. 18 decamer oligonucleotide primers tested, only 13 gave The same trend of results was also observed with the concen- specific and stable results (Figures 3; Table 3). The total tration of thorium, which was 32.458 and 31.246 BqKg-1 in number of bands was 94 ranging from 100 to 3000 bp in the black sand of Baltim in which E. sativa and S. lycoper- molecular size. Different polymorphic bands were detected sicum were grown respectively. While in the control soil, for different primers (P, P2, P3, P4, P5, P6, P7 and P8) thorium content was 16.027 and 11.448 BqKg-1 dry weights in the plants cultivated in black sand soil from which the respectively. In the black sand of P. guajava and M. indica plants accumulated 12.957 and 1.657 Bq Kg-1 dry weight

TABLE 2 - Uranium and thorium content (BqKg-1 dry weight) in the black sand of Baltim and the control soil in which the study plant species are grown. Data are expressed as mean ± S.D. Different superscript stars for the same species indicate a significant difference.

Study species Study Soil Uranium Thorium (Bq Kg-1) (Bq Kg-1) Eruca sativa Black sand 76.91 ± 3.10** 32.4 ± 0.61** Control 45.24 ± 1.88 16.0 ± 0.61 Solanum lycopersicum Black sand 95.42 ± 3.10** 31.24 ± 1.01** Control 46.89 ± 2.46 11.44 ± 0.61 Psidium guajava Black sand 86.79 ± 1.88** 20.06 ± 0.61** Control 60.87 ± 1.42 7.94 ± 0.61 Mangifera indica Black sand 81.85 ± 1.88** 24.24 ± 0.80** Control 9.74 ± 0.44 1.91 ± 0.10

** Significant at P ≤ 0.001.

346 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

20 Black Sand 20 *** ) 18 Control Soil 18 * -1 16 16 ** ** 14 14 12 12 10 10 8 8 6 6 4 4

Uranium Content (BqKg 2 2 0 0 Es Le Pg Mi

2,0 *** 1,6 ) 1,8 ** *** -1 1,4 1,6 1,2 1,4 *** 1,2 1,0 1,0 0,8 0,8 0,6 0,6 0,4 0,4 Thorium Content (Bq Kg 0,2 0,2 0,0 0,0 Es Le Pg Mi

FIGURE 2 - Uranium and thorium content (BqKg-1 dry weight) in the edible portions of the study crop plants (Es= Eruca sativa and Le= Solanum lycopersicum) and study fruit trees (Pg= Psidium guajava and Mi= Mangifera indica) cultivated in the black sand of Baltim and in a control soil. Vertical bar around the mean is the standard deviation. And different superscript stars for the same species indicate a significant difference, *** significant at P ≤ 0.001; ** significant at P ≤ 0.01 and * significant at P ≤ 0.05.

TABLE 3 - RAPD profiles of Eruca sativa cultivated in black sand of Baltim and in a control soil in terms of the number of amplified products; polymorphic bands; monomorphic bands and polymorphism percentage with different RAPD primers.

Primer Total amplified Number of Black Sand products bands in con- Number Polymorphic bands Monomorphic bands Polymorphism (%) trol soil of bands (+) (-) T I D Primer 1 5 4 1 1 4 - - - 100 Primer 2 8 4 5 4 3 1 - 1 87.5 Primer 3 11 5 7 6 4 1 - 1 90.9 Primer 4 3 2 3 1 - 2 1 - 33.3 Primer 5 8 5 4 3 4 1 - 1 87.5 Primer 6 7 3 4 4 3 - - - 100 Primer 7 3 1 3 2 - 1 1 - 66.6 Primer 8 7 6 5 1 2 4 - 1 42.9 Primer 9 11 11 11 - - 11 1 - 0 Primer 10 9 9 9 - - 9 2 1 0 Primer 11 7 7 7 - - 7 1 - 0 Primer 12 10 10 10 - - 10 - - 0 Primer 13 5 5 5 - - 5 - 2 0 Total 94 72 74 22 20 52 6 7 44.7

347 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

FIGURE 3 - RAPD profiles of genomic DNA from the study plant species (Eruca sativa, Solanum lycopersicum and Mangifera indica) cultivated in black sand of Baltim and in a control soil generated using different RAPD primers as indicated at the top of each gel. M: 100 bp DNA ladder (100-1000 bp), P: primer, C: control soil, B: black sand, : appearance of new band and : varied band intensities.

of uranium and thorium in the edible portions. Polymor- 55.3% of fragments were monomorphic. Out of 52 mono- phisms were due to the loss and/or gain of amplified band morphic bands intensities of 6 bands increased and 7 bands in the plants of black sand soil in comparison to plants from decreased with the increase of radionuclides content in the the control soil. Values of polymorphisms (%) were ranged plants cultivated in the black sand soil. With primer 2 the between 33.3% with primer 4 to 100% with primer 1. intensity of band with approximate molecular size 300 bp Among the 94 amplified products, 42 were polymorphic and decreased and bands of with approximate molecular size thus 44.7% of the total fragments were polymorphic while 400 bp increased when the plants cultivated in the black

348 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

sand soil. Also with primer 6 and primer 7 intensity of band DNA profile of P. guajava among the 18 RAPD primers with approximate molecular size 200 bp increased with in- used, 8 primers only have amplified and 10 primers have creased uranium and thorium contents in leaves of plants not amplified (Table 5). The eight primers generated a total from Baltim. of 40 bands of which 38 band were polymorphic bands. Eighteen 10-mer oligonucleotide primers were utilized Primer 4 and primer 8 are the only primers that showed the for screening the genome of S. lycopersicum (Figures 3; monomorphic bands and 80% polymorphism while with Table 4), whilst only 12 primers generated specific and sta- remaining 6 primers 100% polymorphism was detected. ble results. A total of 71 fragments were amplified by the Thus 95% of the total amplified products were polymor- 12 primers with molecular sizes ranged from 100 to 2500 bp. phic and only 5% of bands were monomorphic. Polymor- Out of these amplified products only 43 bands were mon- phisms observed in the amplified products were meanly omorphic but the remaining bands (28 bands) were poly- due to the appearance of 36 newly amplified fragments morphic due to appearance of 10 new bands and disappear- when the plants cultivated in the black sand of Baltim from -1 ance of 18 bands with uptake of more uranium and thorium which the fruits accumulated 16.165 and 1.455 Bq Kg dry in the fruits of the plants cultivated in black sand soil. Thus weight of uranium and thorium in comparison to accu- 39.4% of the total amplified products were polymorphic. It mulated radionuclides from the control soil; 11.476 and -1 was observed that the polymorphism was 100% with pri- 0.566 Bq Kg dry weigh respectively. mer 1, primer 2, primer 3, primer 4, primer 5 and primer 6 As indicated from resulted bands, the 18 10 - mer prim- while primer 9, primer 10, primer 11, primer 12 and primer ing oligonuleotides were used to analyze the RAPD - PCR 13 showed 100 monomorphic bands. Among the primers, products of M. indica but only 13 primers gave specific and which showed monomorphisms intensities of bands with stable results. With exception of primer 14 and primer 18, molecular sizes 843 and 771 bp produced by primer 10 in- RAPD patterns generated by DNA from the plants col- creased while with primer 12 intensities of bands having lected from black sand of Baltim were clearly different molecular sizes of 674 and 412 bp decreased (Figure 3) from those obtained from the plants cultivated in the con- when the plants cultivated in the black sand soil. In RAPD trol soil. A total of 105 bands were scored ranging from

TABLE 4 - RAPD profiles of Solanum lycopersicum cultivated in black sand of Baltim and in a control soil in terms of the number of amplified products; polymorphic bands; monomorphic bands and polymorphism percentage with different RAPD primers.

Primer Total ampli- Number of Black Sand fied products bands in con- Number of Polymorphic bands Monomorphic bands Polymorphism (%) trol soil bands (+) (-) T I D Primer 1 4 4 - - 4 - - - 100 Primer 2 9 8 1 1 8 - - - 100 Primer 3 1 - 1 1 - - - - 100 Primer 4 1 1 - - 1 - - - 100 Primer 5 6 2 4 4 2 - - - 100 Primer 6 3 1 2 2 1 - - - 100 Primer 7 ------Nil Primer 8 6 4 4 2 2 2 1 - 66.7 Primer 9 9 9 9 - - 9 - - 0 Primer 10 10 10 10 - - 10 2 1 0 Primer 11 8 8 8 - - 8 - - 0 Primer 12 9 9 9 - - 9 - 2 0 Primer 13 5 5 5 - - 5 - 1 0 Total 71 61 53 10 18 43 3 4 39.4

TABLE 5 - RAPD profiles of Psidium guajava cultivated in black sand of Baltim and in a control soil in terms of the number of amplified products; polymorphic bands; monomorphic bands and polymorphism percentage with different RAPD primers.

Primer Total ampli- Number of Black Sand fied products bands in con- Number of Polymorphic bands Monomorphic bands Polymorphism (%) trol soil bands (+) (-) T I D Primer 1 4 - 4 4 - - - - 100 Primer 2 2 - 2 2 - - - - 100 Primer 3 8 1 7 7 1 - - - 100 Primer 4 5 1 5 4 - 1 1 - 80 Primer 5 7 - 7 7 - - - - 100 Primer 6 6 1 5 5 1 - - - 100 Primer 7 3 - 3 3 - - - - 100 Primer 8 5 1 5 4 - 1 - - 80 Total 40 4 38 36 2 2 1 - 95

349 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

FIGURE 4 - SDS-PAGE of proteins from the study plant species cultivated in the black sand of Baltim and in a control soil. Lane M: molecular weight standards (marker); Lane C: control soil; Lane B: black sand; Ec: Eruca sativa, Le: Solanum lycopersicum, Pg: Psidium guajava, Mi: Mangifera indica. The arrows refer to the appearance of new bands and varied band intensities in plants cultivated in black sand.

TABLE 6 - RAPD profiles of Mangifera indica cultivated in black sand of Baltim and in a control soil in terms of the number of amplified products; polymorphic bands; monomorphic bands and polymorphism percentage with different RAPD primers.

Primer Total ampli- Number of Black Sand fied products bands in con- Number of Polymorphic bands Monomorphic bands Polymorphism (%) trol soil bands (+) (-) T I D Primer 1 12 9 4 3 8 1 - - 91.7 Primer 2 6 3 3 3 3 - - - 100 Primer 3 8 7 2 1 6 1 - 1 87.5 Primer 4 10 5 5 5 5 - - - 100 Primer 5 10 7 4 3 6 1 - - 90 Primer 6 10 5 6 5 4 1 - - 90 Primer 7 6 5 3 1 3 2 - 2 66.7 Primer 8 9 6 6 3 3 3 - - 66.7 Primer 14 8 8 8 - - 8 - - 0 Primer 15 5 5 4 - 1 4 - 4 20 Primer 16 9 6 3 3 6 - - - 100 Primer 17 6 4 3 2 3 1 - 1 83.3 Primer 18 6 6 6 - - 6 - - 0 Total 105 76 57 29 48 28 - 8 73.3

350 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

200 to 3000 bp in molecular size. The principal events ob- teins expressed in plants of the control soil were disap- served following absorption of high level of uranium and peared when the plants cultivated in the black sand soil thorium from the black sand soil were variations in inten- with expression of 10 new proteins not detected in the sities of some monomorphic bands and loss of normal plants from the control soil. Protein of molecular weight bands and appearance of new bands causing polymor- 110 K.Da its intensity decreased in the plants of black sand phisms compared with plants cultivated in the control soil. soil. Electrophoretic banding pattern of Proteins extracted Percentage of polymorphisms was ranged between 0% from fruits of M. indica showed disappearance of most pro- with primer 14 and primer 18 to 91.7% with primer 1 teins expressed in the plants cultivated in the control soil reaching to 100% with primer 2, primer 4 and primer 16. and appearance of 21 new proteins in the plants from black In general, out of 105 amplified products only 28 were sand of Baltim where the plants absorbed high level of ura- monomorphic bands (26.7%) and the remaining 77 bands nium and thorium. were polymorphic (73.3%) due to disappearance of 29 normal bands and appearance of 48 new bands. As indicated from profiles of primer 3, primer 7, primer 15 and primer 4. DISCUSSION 17 the variation in the monomorphic bands was meanly due to the decrease of the bands intensities. For instance, mono- In the present research determination of radionuclides morphic bands of molecular sizes 1314, 720, 491 and 417 bp content in the black sand of Baltim and in a control soil obtained by primer 15 and band of molecular size 803 bp revealed that uranium content is generally higher than tho- produced by primer 17 decreased in intensities with the in- rium content, and both uranium and thorium concentra- crease of uranium and thorium contents. tions in the black sand soil were significantly higher than that in the control soil. The amount of uranium and thorium 3.3. Electrophoretic banding pattern of protein accumulated in the edible portions of the plant species from The mutagenic effect of uranium and thorium taken up the black sand soil and from the control soil where the the study plant species in the edible portions on the banding study plant species were cultivated have been measured pattern of protein are shown in Figure 4. With regard to E. and the results showed that uranium and thorium were col- sativa, the expression of 11 proteins was indicated by SDS- lected with significant concentrations in the edible portions PAGE of the plants cultivated in the black sand of Baltim of the plants of the black sand soil in comparison to plants and in the control soil. But 6 proteins expressed in the from the control soil. Similarly, Hegazy and Emam [17] plants of control soil were not detected in the plants from detected higher level of uranium and thorium in the tissues the black sand soil while the expression of 5 new proteins of the plants grown in the black sand soil than those grown was detected in these plants where the levels of uranium in the control soil. The mutagenic effects of the absorbed and thorium increased from 4.196 and 0.525 Bq Kg-1 dry Uranium and Thorium by the plant species were studied weight respectively in the leaves of the plants from the con- through the study of RAPD profiles of DNA and electropho- trol soil to 12.957 and 1.657 Bq Kg-1 dry weight respec- retic banding pattern of proteins and the results revealed that tively in the plants cultivated in the black sand soil. More- high content of uranium and thorium of the plants induced over the intensities of proteins of molecular weight 88, 77 DNA damages which was reflected by changes in RAPD and 40 K.Da increased when the plants cultivated in the profiles. Changes in SDS-PAGE of proteins were alos ob- black sand with the increase of uranium and thorium con- served with the increased concentration of radioactive ele- tent in the plants leaves. ments in the edible portions of these plants indicated by the alterations in the bands intensities, appearance of new pro- The protein-banding pattern of S. lycopersicum tein band and disappearance of some bands. showed the decrease of number of the expressed proteins from 15 proteins in plants cultivated in the control soil to Contamination of ecosystems and exposure to toxic 11 proteins in the plants from the black sand soil with the metals is a major concern all over the world [16]. One of the increase of uranium and thorium content in the fruits of the important aspects of ecological and agricultural recovery plants. Only three of the expressed proteins were detected from radiation pollution is the understanding of the effect of in both plants of control soil and black sand of Baltim but radiation on organisms and their populations [25]. Under- the remaining expressed proteins were greatly different as standing the mechanisms by which higher plants perceive 12 proteins expressed in plants of control soil were not de- environmental stimuli is of vital importance to modern mo- tected in plants cultivated in black sand soil but 8 new pro- lecular biology. In practice, the key point to agriculture is teins were expressed in these plants. how to regulate the harmonious relationship between soil- In SDS-PAGE of P. guajava 22 bands were observed environment and crops and make the best of physiological in the plants grown in the control soil where uranium and potential of crops [26]. To some extent, plants could over- thorium contents were 11.476 and 0.566 Bq Kg-1 dry come environmental stress by developing efficient and spe- weight respectively. These bands decreased to 17 bands cific physio-biochemical mechanisms. On the other hand, an when the plants cultivated in the black sand soil and accu- excess of toxic heavy metal ions induces several cellular mulated more uranium and thorium in the plant tissues stress responses and damage to different cellular compo- (16.165 and 1.455 Bq Kg-1 dry weight respectively). 15 pro- nents such as membranes, proteins and DNA [7].

351 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

In “genetic-ecotoxicology” or ‘‘eco-genotoxicology”, main characteristic changes included alterations in the the effective evaluation and proper environmental monitor- band relative mobility, the band intensities, and disappear- ing of potentially genotoxic pollutants will be improved ance of some bands and appearance of new bands indicat- with development of sensitive and selective methods to de- ing that protein expression is highly modified in the four tect toxicant-induced alterations in the genomes of a wide study species. Similarly, the study of Soliman et al. [9] on range of biota [27]. As reported by some researchers [7, 10, the mutagenic effect of radionuclides absorbed by some 28, 29], the alteration in DNA fingerprinting is a useful bi- natural grown plant species in black and soil showed the omarker in eco-genotoxicology. There are limited genotox- same trend of result. High concentrations of radionuclides icity studies on the mutagenic effects of radioactive ele- in the plant tissues caused expression of new proteins, ments in the plants. while suppression of some proteins occurred at law con- In the present investigation the polymorphism in ge- centration of radionuclides with general decrease of the to- nomic DNA was detected by RAPD profiles through the tal protein content. Radionuclides absorbed by the study randomly primed PCR reactions. In this sense, the obvious species emits an internal radiation as reported by Nayar et disappearance of normal bands and appearance of new al. [34] who pointed out that the radiations from the ab- bands generated from the plants absorbed high concentra- sorbed radionuclides are much more important than the ex- tion of uranium and thorium from black sand soil of Baltim ternal radiations in the production of the biological in comparison to the plants from the control soil. The same changes. These internal radiations may be the reasons of trend of results was also detected by Dhakshanamoorthy et appearance and disappearance of protein bands of the study al. [30] on the genetic alterations in Jatrapha curcas pro- species with the increased radionuclide content of plants as duced by ionizing radiation and the study of genotoxicity earlier reported by Zhenxing et al. [35]. induced from toxic chemicals in Phaseolus vulgaris by In the present investigation, the observed changes in Cenkci et al. [8]. The disappearance of normal bands (band the protein banding pattern may be due to two mutational loss) may be related to the events such as DNA damage types i.e. gene mutation, induced by the absorbed radionu- (e.g. single- and double-strand breaks, modified bases, clides and detected by RAPD profile of DNA studied in the abasic sites, oxidized bases, bulky adducts, DNA–protein present research and have been mentioned above and cyto- cross links), point mutations and/or complex chromosomal logical aberrations, that were caused by the gamma radia- rearrangements induced by genotoxins [28, 31]. Appear- tion emitted by the soil radionuclides as reported by Soli- ance of new PCR products (extra bands) was also detected man et al. [9]. A protein-banding pattern of an organism in RAPD profiles. New PCR amplification products may represents a biochemical genetic fingerprint to that organ- reveal a change in some oligonucleotide priming sites due ism [34]. Each protein band represents the final products of to mutations [new annealing event(s)], large deletions a transcriptional and translation events occurring due to ac- (bringing to pre-existing annealing site closer), and/or ho- tive structural genes. Amino acid changes within a protein, mologous recombinations. The new bands could be at- due to mutational events, can result in altered protein mi- tributed to mutations while the disappeared bands could be gration rates when the proteins are compared on poly- attributed to DNA damage [32]. acrylamide gels [34]. Thus the change in the protein pattern Domingo et al. [33] reported that the quantity of DNA is highly correlated to the recorded cytological anomalies strand breakage increased with increasing doses of en- caused by soil radionuclides and the polymorphisms in riched uranium. Uranium is indeed chemically able to gen- DNA induced by the accumulated radionuclides in plants erate reactive oxygen species (ROS) via the redox chemis- as reflected by RAPD profiles. This correlation may lead try of transition ions. Similarly, Uranium can radiologi- to constitutive protein production or to attenuation or com- cally induce free radicals production via the ionization phe- plete suppression of the concerned genes and well-pro- nomenon induced by alpha particle emissions [6]. DNA le- duced intensive bands, faint bands or complete band disap- sions induced by ROS and free radicals can indeed result pearance [35]. in DNA strand breaks, DNA-protein cross-links and oxida- The increased expression of several proteins associ- tive base modifications. In addition, metals can cause direct ated with energy production and metabolism suggests that DNA strand breaks and inhibit DNA repair. Uranium is higher energy is required to activate the metabolic pro- hence expected to be both chemically and radiologically cesses, mainly focused to detoxification, in organisms ex- able to induce genotoxicity [6]. Most adducts in DNA in- posed to heavy metals. Organisms chronically exposed to duced by reactive oxygen species (ROS) can be removed a moderate level of a toxicant can develop acclimatization by DNA repair mechanisms and hence don’t cause muta- that confers increased resistance to the toxicant, involving tions [13]. The gain/loss or intensity differences of RAPD important metabolic reshuffling [16]. bands may be related to DNA damage, mutations or structural rearrangements induced by genotoxic agents, affecting the primer sites and/or interpriming distances [28]. 5. CONCLUSIONS The study of electrophoretic banding pattern of proteins indicated the change of SDS-PAGE of proteins with The study of radionuclides content in the selected four the increase of radionuclides content of the plants. The plants species cultivated in the black sand revealed the abil-

352 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

ity of these plants to accumulate uranium and thorium in [8] Cenkci, S., Yıldız, M., Ciğercim, I.H., Konuk, M. and Bozdag, the edible portions of the plants with significantly high A. (2009). Toxic chemicals-induced genotoxicity detected by random amplified polymorphic DNA (RAPD) in bean concentration in comparison to plants from the control soil. (Phaseolus vulgaris L.) seedlings. Chemosphere 76: 900–906. Mutagenic effects and genotoxicity induced by the absorbed radionuclides were detected by the analysis of [9] Soliman, M.S.A., Hegazy, A.K., Goda, S.K., Emam, M.H. and Alatar, A.A. (2011) Cytotoxicity and mutagenic effects of soil RAPD-DNA profiles and electrophoratic banding pattern radionuclides on some black sand plant species. J. Med. Ecol. of protein. The results showed a correlation between the 11: 5-20. radionuclides content and the level of radionuclides in- [10] Shtangeeva, I., Ayrault, S. and Jain, J. (2005). Thorium uptake duced DNA damage in the tissues of the study species. by wheat at different stages of plant growth. J. Environ. Radi- High radionuclide content in the edible portions of the oact. 81: 283-293. plants species caused changes in SDS-PAGE of proteins [11] Poli, A., Salerno, A., Laezza, G., di Donato, P., Dumontet, S. indicated by alterations in the bands relative mobility and and Nicolaus, B. (2009) Heavy metal resistance of some ther- intensities, expression of new proteins and suppression of mophiles: potential use of a-amylase from Anoxybacillus am- some normal proteins. These results underline the deep ylolyticus as a microbial enzymatic bioassay. Res. Microbiol. need for further investigations regarding radionuclides 160: 99-106. ecotoxicity. The DNA polymorphism detected by RAPD [12] Vannini, C., Domingo, G., Marsoni, M., Fumagalli, A., Ter- profiles and SDS-PAGE could be used as an investigation zaghi, R., Labra, M., De Mattiac, F., Onelli, E. and Bracale, M. (2011). Physiological and molecular effects associated tool for environmental toxicology and as a useful bi- with palladium treatment in Pseudokirchneriella subcapitata. omarker assay. Aquat. Toxicol. 102: 104–113.

[13] Jimi, S., Uchiyama, M., Takaki, A., Suzumiya, J. and Hara, S. (2004). Mechanisms of cell death induced by cadmium and ar- ACKNOWLEDGMENT senic. Annals N.Y. Acad. Sci. 1011: 325-331. [14] Yıldız, M., Ciğerci, İ., Konuk, M., Fidan, A.F. and Terzi, H. The authors extend their appreciation to the Deanship (2009). Determination of genotoxic effects of copper sulphate of Scientific Research at King Saud University for funding and cobalt chloride in Allium cepa root cells by chromosome the work through the research group project number RGP- aberration and comet assays. Chemosphere 75: 934-938. VPP-175. [15] Luque-Garcia, J.L., Cabezas-Sanchez, P. and Camara, C. (2011). Proteomics as a tool for examining the toxicity of The authors have declared no conflict of interest. heavy metals. Trend. Anal. Chem. 30: 703-716. [16] Hassan, A.A.H. (1993) Ground magnetic and spectrometric exploration of the black sand deposits in Abu Khashaba district, Rosetta area, Egypt. M.Sc. thesis Cairo Univ. REFERENCES [17] Hegazy, A.K. and Emam, M.H. (2011). Accumulation and soil-to-plant transfer of radionuclides in the Nile delta coastal [1] Tykva, R., Salahel, D.K, Pavel, C.C., Cecal, A. and Popa, K. black sand habitats. Int. J. Phytoremediat. 13: 140–155. (2009). Contribution to the external surface of a titanium-rich sand (Abou-Khashaba, Egypt) in the uranium uptake pro- [18] Hegazy, A.K., Afifi, S.Y., Alatar, A.A., Alwathnani, H.A. and cesses. J. Radioanal. Nucl. Chem. 279: 811-816. Emam, M.H. (2013). Soil characteristics influence the radionuclide uptake of different plant species. Chem. Ecol. 29: 255- [2] El-Gamal, A., Rashad, M. and Ghatass, Z. (2010). Radiologi- 269. cal responses of different types of Egyptian Mediterranean coastal sediments. Radiat. Phys. Chem. 79: 831–838. [19] Marczenko, Z. (1986). Separation and spectrophotometric determination of elements. Harwood, New York, pp. 644-650. [3] Bolcaa, M., Saçb, M.M., Çokuysala, B., Karalıb, T. and Ekdalb, E. (2007) Radioactivity in soils and various food stuffs [20] Doyle, J.J. and Doyle, J.L. (1990) Isolation of plant DNA from from the Gediz River Basin of Turkey. Radiat. Meas. 42: 263- fresh plant tissue. Focus 12: 12–15 270. [21] Nelson, T., Harps, M.H. and Mayfield, S.P. (1984). Taylor, [4] Shtangeeva, I. (2010). Uptake of uranium and thorium by na- W.C. Light regulated gene-expression during maize leaf de- tive and cultivated plants. J. Environ. Radioact. 101: 458–463. velopment. J. Cell. Biol. 98: 558-564. [5] Prévérala, S., Ansoborlob, E., Maric, S., Vavasseura, A. and [22] Bradford, M.M. (1976). Rapid and sensitive method for the Forestier, C. (2006) Metal(loid)s and radionuclides cytotoxi- quantitation of microgram quantities of protein utilizing the city in Saccharomyces cerevisiae. Role of YCF1, glutathione principle of protein-dye binding. Ann. Biochem. 72: 248-254. and effect of buthionine sulfoximine. Biochimie 88: 1651- 1663. [23] Laemmili, U.K. (1970). Cleavage of structural proteins during the assembly of the head of Bactriophage T4. Nature 227: 680- [6] Barillet, S., Adam-Guillermin, C., Palluelm, O., Porcher, J.M. 685. and Devaux, A. (2011). Uranium bioaccumulation and biological disorders induced in zebrafish (Danio rerio) after a de- [24] Sammons, D.W., Adams, L.D. and Nishizawa, E.E. (1981) Ul- pleted uranium waterborne exposure. Environ. Pollut. 159: tra-sensitive silver based color staining of polypeptides in pol- 495-502. yacrylamide gels. Electrophoresis 2: 135. [7] Liu, W., Li, PJ., Qi, X.M., Zhou, Q.X., Zheng, L., Sun, T.H. [25] Turuspekov, Y., Adams, R.P. and Kearney, C.M. (2002). Ge- and Yang, Y.S. (2005). DNA changes in barley (Hordeum vul- netic diversity in three perennial grasses from the Semipala- gare) seedlings induced by cadmium pollution using RAPD tinsk nuclear testing region of Kazakhstan after long-term ra- analysis. Chemosphere 61: 158–167. diation exposure. Biochem. Sys. Ecol. 30: 809–817.

353 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

[26] Gang, W., Zhen-Kuan, W., Yong-Xiang, W., Li-Ye, C. and Hong-BoS. (2007). The mutual responses of higher plants to environment: Physiological and microbiological aspects. Col- loid Surface B: Biointerfaces 59: 113-119.

[27] Liu, W., Yang, Y.S., Zhou, Q., Xie, L., Li, P. and Sun, T. (2007). Impact assessment of cadmium contamination on rice (Oryza sativa L.) seedlings at molecular and population levels using multiple biomarkers. Chemosphere 67: 1155–1163. [28] Wolf, D.H., Blust, R. and Backeljau, T. (2004). The use of RAPD in ecotoxicology. Mutation Research 566: 249–262.

[29] Atienzar, F.A. and Jha, A.N. (2006) The random amplified polymorphic DNA (RAPD). assay and related techniques applied to genotoxicity and carcinogenesis studies: A critical review. Mutation Research 613: 76–102. [30] Dhakshanamoorthy, D., Selvaraj, R. and Chidambaram, A.L.A. (2011). Induced mutagenesis in Jatropha curcas L. using gamma rays and detection of DNA polymorphism through RAPD marker. C. R. Biol. 334: 24-30. [31] Atienzar, F.A., Venier, P., Jha, A.N. and Depledge, M.H. (2002). Evaluation of the random amplified polymorphic DNA (RAPD) assay for the detection of DNA damage and mutations. Mutation Research 521: 151–163.

[32] Atienzar, F.A., Conradi, M., Evenden, A.J., Jha, A.N. and De- pledge, M.H. (1999). Qualitative assessment of genotoxicity using random amplified polymorphic DNA: comparison of genomic template stability with key fitness parameters in Daph- nia magna exposed to benzo[a]pyrene. Environ. Toxicol. Chem. 18: 2275–2282. [33] Domingo, J.L. (2001). Reproductive and developmental toxicity of natural and depleted uranium: Annual Review Reprod. Toxicol. 15: 603-609.

[34] Nayar, G.G., George, .KP. and Gopal-Ayengar, A.R. (1970) On the biological effects of high background radioactivity: studies on Tradescantia grown in radioactive monazite sand. Radiation Botany 10: 287-292.

[35] Zhenxing, L., Hong, L., Limin, C. and Jamil, K. (2007) The influence of gamma radiation on the allergenicity of shrimp (Penaeus vannamei). Journal of Food Engineering 79: 945- 949. [36] Abdelsalam, A.Z.E., Hassan, H.Z. and Soliman, K.H.A. (1996). The Mutagenicity of two aromatic systemic pesticides using three biological systems. Egypt. J. Genet. Cytol. 23: 185- 195. [37] Hassan, H.Z. (1995). Evaluation of mutagenic effects of two insecticides basudin and decis on Vicia faba plants. Egypt. J. Genet. Cytol. 25: 27-38. Received: May 06, 2014 Revised: June 24, 2014 Accepted: July 03, 2014

CORRESPONDING AUTHOR

Mohammad FAISAL Department of Botany and Microbiology College of Science King Saud University Riyadh SAUDI ARABIA

E-mail: [email protected]

FEB/ Vol 24/ No 1b/ 2015 – pages 343 – 354

354 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

CHANGES IN PHOTOTROPHIC COMMUNITY STRUCTURE IN THE VERTICAL PROFILE DURING SUMMER STRATIFICATION IN EUTROPHIC LAKE

Elżbieta Szeląg-Wasielewska*, Natalia Jakubowska and Agnieszka Kaźmierska

Department of Water Protection, Faculty of Biology, Adam Mickiewicz University in Poznań, Umultowska 89, 61-614 Poznań, Poland

ABSTRACT taxonomic composition and biomass of cell size fractions, especially in a strongly stratified lake. Phytoplankton often The community structure and biomass of pelagic pho- has non-random, specific distribution, but the question is: totrophs, with particular attention to the picoplanktonic how are the species compositions assembled from a com- size fraction, were studied in a eutrophic Lake Strzeszyńskie mon species pool? Phototrophs are a major component of in the Wielkopolska Lakeland (western Poland). Photo- the food web, so investigations of their species diversity is trophic community included prokaryotic cells, such as oxy- often a starting point for research on trophic links. Many photobacteria (cyanobacteria), anoxyphotobacteria, eukary- studies have demonstrated vertical variation of the phyto- otic cells representing cryptophytes, dinophytes, diatoms, plankton community within lakes. However, investigation chrysophytes, xanthophytes, haptophytes, and green algae. of all groups of phototrophs and their size fractions requires 112 taxa of the phototrophs were recorded. Most of them the use of various approaches to sample collection and vari- were eukaryotic cells (83 species), including chlorophytes ous microscopic methods, which is rarely done simultane- (48) and diatoms (15). The range of the photortophs bio- ously. As a rule, in upper water layers single-celled cyano- mass was between 0.1 and about 3 mg l-1, and its maximum bacteria are neglected, while in the deeper layers, the same was at 8 m depth, mostly due to eukaryotic cells. In the applies to phototrophic bacteria. Thus, only few studies so shallower waters layers the contributions of the cyanobac- far have described the abundance, distribution, and compo- teria and the eukaryotes to the total biomass were very sim- sition of the entire phototrophic community, i.e. consider- ilar. Starting from 10 m anoxyphotobacteria increased in ing all taxonomic groups and all size fractions in the whole importance and near the bottom their contribution to pho- water column [1-5]. Size structure is a particularly important totrophic biomass exceed 50%. The smallest phototrophs parameter used for characterization of phototrophic commu- were dominated by the picocyanobacteria. The joint con- nities, but the importance of the smallest cells, i.e. pico- tribution of pico- and nanoplankton in the water column planktonic phototrophs, as significant contributors to photo- ranged significantly. In the epilimnion it was about 30%, trophic biomass, has been increasingly recognized [6-8]. in the upper part of metalimnion it reached 50%, and in the lower part of hypolimnion they accounted for 70% of total The aim of this study was to characterize the photo- phototrophic biomass. In the microplankton, the highest bi- trophic community of the pelagic zone of a lake and to as- omass was produced by the filamentous cyanobacteria and sess changes in its structure in the vertical profile. Special phytoflagellates. These examinations are pointing out to attention was paid to picoplanktonic phototrophs and pig- the need for a holistic approach to the photosynthetic organ- mented bacterioplankton. For the comparison of photo- isms and show that at increasing depths different groups of trophs, water samples were collected during thermal strat- phototrophs play the role as major producers. ification from Lake Strzeszyńskie, which is quite deep and eutrophic. In this lake this is the first study of all phototrophs under increasing eutrophication, after a cyanobacte- KEYWORDS: phytoplankton, vertical distribution, picocyanobacterial bloom. ria, picoeukaryotes, photosynthetic bacteria

2. MATERIALS AND METHODS 1. INTRODUCTION This study was carried out in Lake Strzeszyńskie In the pelagic zone of lakes, phototrophic communities (52°27′36″N, 16°49′19″E), located in the Wielkopolska vary in the vertical profile, and the variation concerns both Lakeland (mid-western Poland). The lake area is 35 ha, max. depth 17.8 m, and a volume 2.8 million m3 [9]. Cur- * Corresponding author rently it has one significant affluent, from a plateau, and

355 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

only 2 effluents. It is a hard-water lake with summer strat- and rapidly decreased as the depth increased, to about 15% ification and anoxic hypolimnion but sometimes in late in the middle of metalimnion. In the deeper part of the wa- summer also the lower part of metalimnion has an oxygen ter column, from 8 m, oxygen was absent (Fig. 1). Hydro- deficit, as reported earlier [10]. Lake Strzeszyńskie is sub- gen sulphide was detected starting from the upper part of ject to progressing eutrophication and is strongly affected hypolimnion. Secchi depth amounted to 2.15 m. by humans. Increasing water pollution results from mostly o bathing people, anglers, and hydrological and sewerage C, mgO2/l changes in the catchment area. 0 5 10 15 20 25 Water samples for the analysis of plankton were taken in the deepest part of the pelagic zone on August 22, 2011. Depth (m) The samples were collected using an electric pump and sil- icon tube at 2-metre intervals from the surface to the bot- 2 tom. Water samples for picophytoplankton and bacterioplankton studies were preserved with buffered formalde- 4 hyde, while the remaining size fractions of phytoplankton with Lugol’s solution. Water temperature and dissolved 6 oxygen concentration were measured by a multiprobe (YSI meter) in the whole vertical column at 1-metre in- 8 tervals, whereas water transparency was measured with a Secchi disc. 10 Planktonic organisms were identified to the species level or  if this was impossible  they were only assigned 12 to a genus. Pico- (0.22.0 µm), nano- (2.020 μm) and microplankton (20200 µm) size fractions were taken into ac- 14 count. Phototrophic and heterotrophic picoplanktonic cells were distinguished on the basis of autofluorescence of photosynthetic pigments [11] and second fluorescence [12] un- 16 der an epifluorescence microscope (Olympus BX-60). For the analysis of picoplanktonic cells water samples of 1- FIGURE 1 - Vertical variation in the temperature (open circles) and 3 ml in volume were collected by filtration on black poly- the concentration of dissolved oxygen (black diamonds ) in Lake carbonate filter (pore size 0.2 m) under a low vacuum Strzeszyńskie in August 2011. pressure. A mercury lamp HBO (100 W) and standard filter sets were applied to generate green, blue, blue-violet and 3.1 Species richness of phototrophs ultra-violet excitation light. The specimens were examined In the pelagic zone of Lake Strzeszyńskie, 112 taxa of at a magnification of 1500. Nanoplankton, microplankton phototrophs were recorded in total. Most of them were eu- and larger phytoplanktonic organisms (>200 µm) were an- karyotes algae (83 species), including chlorophytes (Chlo- alysed under an inverted microscope using phase–contrast rophyta) (48 species), diatoms (Bacillariophyceae) (15), illumination after sedimentation in settling chambers of cryptophytes (Cryptophyceae) (7), dinophytes (Dinophyceae) 14 ml in volume, at a magnification of 40, 150 and 600. (6), chrysophytes (Chrysophyceae) (5), haptophytes (Hapto- As the abundance and biomass of organisms larger than phyceae) (1), and xanthophytes (Xanthophyceae) (1). Species 200 m were negligible, they were analysed jointly with richness was the highest in the epilimnion, at the depth of microplankton. Abundance was expressed as cell numbers 4 m. At the deeper layer it gradually decreased, and only per 1 ml of water. The biovolume of each species was cal- 12 taxa of eukaryotic cells were observed near the bottom. culated on the basis of cell shape, size, and number, while Cyanobacteria were also present in the whole water column their biomass was expressed assuming that the volume of and were represented by 27 taxa, whereas anoxyphotobac- 106 µm3 is equivalent to 1 µg. The biomass value are given teria, by 2 taxa. The latter group appeared at the depth of as wet weight (mg l-1). 8 m, but occurred also deeper down, to the lake bottom (Table 1).

3. RESULTS The number of phototrophic taxa in samples from various depths ranged from 24 to 60 (53-60 in the epilimnion, Lake water temperature reached 22.2ºC at the surface 35-51 in the metalimnion, and 24-34 in the hypolimnion). but declined as the depth increased, to 5.0ºC near the bot- Starting from the metalimnion, it gradually declined at in- tom. On the basis of the temperature curve, the vertical pro- creasing depths. Most of the taxa are common in inland file was divided into the epi-, meta-, and hypolimnion, waters. However, some of them, e.g. Dinobryon crenula- which were 4 m, 6 m, and 6 m thick, respectively. The sat- tum W. West et G.S. West and Tetraedriella jovetii Bour- uration of epilimnetic waters with oxygen exceeded 100% relly, are rarely reported from Polish lakes.

356 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

TABLE 1 - Floristic spectrum of pelagic plankton in Lake Strzeszyńskie in the vertical profile.

Number of taxa Total Group of phototrophs 0.2m 2m 4m 6m 8m 10m 12m 14m 16m Anoxyphotobacteria - - - - 1 2 2 2 2 2 Oxyphotobacteria 21 18 17 15 17 12 13 13 10 27 Eukaryotic cells: 1 Cryptophyceae 5 5 4 6 7 3 3 0 0 7 2 Dinophyceae 3 3 5 4 2 0 1 0 0 6 3 Chrysophyceae 4 2 2 2 0 1 0 1 1 5 4 Haptophyceae 1 1 1 ------1 5 Bacillariophyceae 4 6 6 5 7 5 3 4 2 15 6 Xanthophyceae 1 1 1 ------1 7 Chlorophyceae 21 17 23 19 16 12 12 13 9 48 Total 1-7 39 35 42 36 32 21 19 18 12 83 Total 60 53 59 51 50 35 34 33 24 112

3.2 Abundance and biomass of phototrophs Phototrophic biomass was not high, ranging from The total phototrophic abundance, considering all the 0.147 to 2.956 mg l-1. The mean value was 1.21 mg l-1 for analysed depths, ranged from 3.09105 to 1.18106 cells the whole water column, while 1.70, 1.66, and 0.28 mg l-1 ml-1 (mean 6.17105cells ml-1). It increased gradually, for the epi-, meta-, and hypolimnion, respectively. The bi- reaching a maximum at the depth of 4 m, i.e. at the base of omass of phototrophs was weakly positively correlated epilimnion, and later decreased to a minimum in the lower with their abundance. Their biomass peaked at the depth of part of metalimnion (10 m), where oxyphotobacteria were 8 m, mostly thanks to eukaryotic cells (87%). In shallower 3-fold less numerous but anoxyphotobacteria were already water layers, contributions of cyanobacteria and eukaryotic quite numerous. Deeper down, in the hypolimnion, at suc- cells to total biomass were very similar, although the latter cessive depths, mostly because of anoxyphotobacteria, to- always prevailed. Deeper down, starting from 10 m, where tal phototrophic abundance was still quite high, ranging total phototrophic biomass did not exceed 0.6 mg l-1, anox- from 3.32105 to 3.82105 cells ml-1 (Fig. 2). yphotobacteria increased in importance, and near the bot-

cells ml-1

0 200000 400000 600000 800000 1000000 1200000

0.2

Depth (m) 10

Anoxyphotobacteria Oxyphotobacteria Eukaryotic algae

FIGURE 2 - Vertical variation in the abundance of phototrophs in Lake Strzeszyńskie in August 2011.

357 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

mg l-1

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5

0.2

Depth (m) 10

Anoxyphotobacteria Oxyphotobacteria Eukaryotic algae

FIGURE 3 - Vertical variation in the biomass of phototrophs in Lake Strzeszyńskie in August 2011.

mg l-1 0.0 0.5 1.0 1.5 2.0 2.5 3.0

0.2

Depth (m) 10

Cryptophyceae Dinophyceae Chrysophyceae Haptophyceae Bacillariophyceae Xanthophyceae Chlorophyta

FIGURE 4 - Vertical variation in the biomass of taxonomic groups of eukaryotic cells in Lake Strzeszyńskie in August 2011.

358 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

tom their contribution to phototrophic biomass exceeded The proportion of eukaryotic cells (i.e. the smallest chloro- 50% (Fig. 3). In the hypolimnion, as much as 45% of the phytes), was low, as on average they accounted for 0.2% biomass was contributed by photosynthetic bacteria: pur- of picophototrophic abundance and 1.2% of the biomass ple and green sulphur bacteria (Chlorobium sp.) and their (Table 3). Eukaryotic PPP occurred in the lower part of the symbiotic consortia (Pelochromatium spp.). water column, and its abundance peaked in the metalim- In the epilimnion, among eukaryotes, the main contrib- nion, at the depth of 10 m (3.1103 cells ml-1). Eukaryotic utors to phototrophic biomass were dinophytes, crypto- PPP included chlorophytes of the genera Choricystis, phytes and green algae, but in the upper part of metalim- Chlorella and Pseudodictyosphaerium. nion, the importance of diatoms increased. Diatoms with In the hypolimnion, as much as 56-84% of photo- cryptophytes peaked in biomass at the depth of 8 m, in the trophic biomass was contributed by photosynthetic bacte- very middle of the water column. However, only 2 m ria – mostly symbiotic consortia of green sulphur bacteria deeper down, the biomass of eukaryotic cells was 18-fold (Pelochromatium spp.). The same phototrophic bacteria lower. The same eukaryotic phototrophs were found also were found also at the lower part of metalimnion, but their in hypolimnion, but their abundance and biomass were contribution to phototrophic biomass did not exceed there very small. Chlorophytes and diatoms in that layer were 26%, because the major contributors at that depth were cy- major contributors to eukaryotic phototrophic biomass anobacteria, both single-celled (29%) and colonial forms (Fig. 4). (39%).

3.3 Size structure of phototrophs 3.3.2 Nanoplankton 3.3.1 Phototrophic picoplankton (PPP) The most abundant nanoplanktonic species were: the The abundance of the smallest size fraction of photo- diatom Cyclotella ocellata, haptophyte Chrysochromulina trophs varied widely, from 2.7×105 to 1.1×106 cells ml-1, parva, cryptophytes Cryptomonas marssonii, and while its contribution to total abundance, from 87 to 97% Rhodomonas lacustris, chlorophytes of the genera (Table 2). The highest density of PPP was observed in the Tetrastrum and Scenedesmus, and phototrophic bacteria epilimnion. Its biomass was the highest there (0.314 mg l-1), Pelodictyon sp. Cyclotella ocellata was most abundant at and decreased with increasing depth to 0.06 mg l-1 at 10 m the depth of 6 m, while cryptophytes peaked slightly and then slightly increased again near the lake bottom. The deeper down, at 8 m. In contrast, Ch. parva was detected mean contribution of PPP to the total biomass of photo- only in the epilimnion. Purple and green sulphur bacteria trophs was close to 20%, and was the highest in the hypo- appeared in the lower part of metalimnion but were most limnion, where it reached 46%. numerous in hypolimnion. At all the studied depths, picocyanobacteria occurred The mean contribution of nanoplankton to photo- but were most numerous in epi- and metalimnion. They trophic abundance in the water column was only 3.2%. were mostly dominated by colonies of picoplanktonic cya- However, this did not result from a low density of the cell nobacteria (45-83.2%), primarily Aphanocapsa and Aphan- size fraction, because at 3 out of the 9 studied depths it ex- othece spp. Single-celled PPP of the genus Synechocystis ceeded 20 000 cells ml-1, but it was due to an exceptionally were more abundant in the lower part of the water column. high abundance of picoplanktonic cells. In contrast, nano-

TABLE 2 - Abundance and biomass of size fractions and their contribution to total phototrophic abundance and biomass.

Abundance Biomass Size fraction 103 cells ml-1 Contribution (%) mg l-1 Contribution (%) Range Mean Range Mean Range Mean Range Mean Picoplankton 268.7-1110 576.5 86.9-97.3 92. 9 0.060-0.314 0.153 6.3-45.8 19.6 Nanoplankton 1.7-33.2 12.0 0.2-8.2 3.2 0.045-0.687 0.251 12.4-36.3 22.1 Microplankton 2.4-67.0 28.1 0.6-7.0 3.9 0.036-2.083. 0.808 23.1-74.9 58.3

TABLE 3 - Abundance and biomass of phototrophic picoplanktonic cells and contributions of single-celled picocyanobacteria (S-Pcy), colonial picocyanobacteria (C-Pcy), eukaryotic picoplankton (E-PPP) and other photosynthetic picobacteria (PhB).

Abundance Biomass (mg l-1) 103 cells ml-1 Contribution (%) mg l-1 Contribution (%) Range Mean Range Mean Range Mean Range Mean S-Pcy 21.2–240 111.2 6.1-28.7 17.6 0.005-0.090 0.034 8.0–35.0 19.7 C-Pcy 17.5–866.5 369.5 5.0-83 53.1 0.005-0.233 0.101 7.9–82.1 54.2 E-PP nd.–3.1 0.7 nd.-1.1 0.2 nd.- 0.004 0.012 nd.–6.5 1.2 PhB nd.– 10.3 95.3 nd.-88.9 29.1 0.015-0.053 0.016 nd.–83.6 24.9 nd – not detected

359 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

0% 20% 40% 60% 80% 100%

0.2

Depth (m) 10

Picoplankton Nanoplankton Microplankton

FIGURE 5 - Contribution of size fraction to phototrophic biomass in Lake Strzeszyńskie in August 2011

plankton biomass accounted for a high proportion of photo- mass was produced by the cyanobacteria Pseudanabaena, trophic biomass (on average 22%, with a wide variation) (Ta- Gomphosphaeria and Microcystis, flagellates Ceratium hi- ble 3). In the water column, there are 2 characteristic peaks rundinella, and members of the genera Peridinium, Cryp- of the contribution of nanoplankton to total phototrophic bi- tomonas, Dinobryon. Important microplanktonic species omass (Fig. 5). A maximum of 36% was detected at the depth were also green algae of the genus Carteria and Closterium of 6 m, partly because of a high abundance of the diatom C. acutum var. variabile. ocellata. At the depth of 14 and 16 m, it reached about 31%, and purple sulphur bacteria were abundant. 4. DISCUSSION It is noteworthy that the joint contribution of pico- and nanoplankton in the water column ranged from 25 to 77%. In studies of phototrophic communities in the epilim- In the epilimnion it was about 30% (28-31%), in the upper nion, aimed to assess species richness, water quality or part of metalimnion (at the depth of 6 m) it approached trophic state, picoplanktonic cells are usually neglected, 50%, and exceeded this value only in the lower part of hy- although many reports indicate their high diversity, abun- polimnion, where they accounted for 70% (73-77%) of to- dance, and sometimes also biomass [6-7], while in the hy- tal phototrophic biomass. polimnion, phototrophic bacteria are often neglected [13]. In Lake Strzeszyńskie, they were taken into account as 3.3.3 Microplankton early as in the 1990s [10] but the former only recently, i.e. In the phototrophic biomass (but not abundance, which in the study presented here. Phototrophic communities are was similar to nanoplankton contribution to total photo- dynamic systems of interacting populations and, at a given trophic abundance), microplankton dominated in epi- and depth, a distinct assemblage of species occurs [14]. Con- metalimnion. In the epilimnion their contribution to photo- sideration of these organisms from various points of view trophic biomass was not lower than 69%, whereas in the and in various respects (e.g. in respect of species composi- metalimnion, not lower than 52%. Deeper down, in the up- tion and size structure) is fascinating, especially with the per part of hypolimnion (12 m), the contribution of micro- use of microscopic methods, and may contribute to better plankton increased suddenly to 75%, mostly due to understanding of their importance in the food web. A par- Pseudanabaena occurrence, and in the lower hypolimnion ticularly favourable period for investigations is the time of suddenly declined to 23-27% in favour of nano- and pico- lake stratification, as it gives a possibility to detect varia- plankton (Fig. 5). Within microplankton, the highest bio- tion within the water column. Interesting results achieved

360 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

in this area by examining two lakes in Siberia using micro- of metalimnion, phototrophic biomass culminated, mostly scopic techniques, analysis of photosynthetic pigments, and thanks to eukaryotic cells. Above this layer, i.e. in the epi- molecular analyzes [15-17]. limnion, the contribution of eukaryotes to phototrophic biomass was lower, as they codominated with cyanobacteria, The total number of phototrophic taxa in the water col- both small-sized and relatively large. In contrast, slightly umn (112) of Lake Strzeszyńskie during thermal stratifica- below the metalimnion, i.e. in the upper hypolimnion, pho- tion in summer did not differ remarkably, as it was only totrophic biomass declined dramatically, while contribu- slightly lower than in the epilimnion (117), where samples tions of oxyphotobacteria and eukaryotic algae to the total were collected 12 times from May to November [18]. The biomass were nearly equal. Anoxyphototrophs found fa- vertical variation, mostly of water temperature and oxy- vourable conditions in the hypolimnion. This sheds new genation, results in a high species richness of phototrophs. light on the magnitude of phototrophic biomass in the hy- In the study lake, species composition varied widely, both polimnion of this lake, indicating that it is on average 45% with depth and seasonally. Various algal groups dominated higher. The range of changes in contribution of photo- in the epilimnion in successive months, but during thermal trophic bacteria to total phototrophic biomass in hypolim- stratification in August they coexisted and their species nion was wide (11-52%), but similar to values were earlier richness was similar. Moreover, the total phototrophic di- reported from some stratified Lobelia lakes with anoxic hy- versity may be even higher in the water column because of polimnions, located in the Pomeranian Lakeland (northern the presence of phototrophic bacteria in the anoxic hypo- Poland): Czarnówek, Iłowatka, and Piekiełko. In the hypolimnion. In less fertile lakes, species richness is lower. For limnion of Lake Czarnówek, as much as 45% of biomass example, in the oligo-mesotrophic, meromictic Lake Czarne, in the first summer and about 5% in the second summer only 60 taxa were recorded in total in July at 8 depths in the was contributed by the photosynthetic green sulphur bac- water column [19]. teria and their symbiotic consortia. In Lake Piekiełko in the Pelagic phototrophs in Lake Strzeszyńskie occurred in first summer, their contribution to phototrophic biomass the whole water column but proportions between their reached 25%, but in the second summer, the contribution groups and the dominant group varied. In the upper part of of phototrophic bacteria was doubled (48%). Within the the column, oxyphotobacteria dominated in terms of abun- hypolimnion of Lake Iłowatka, studied only in 2003, pho- dance, while major contributors to biomass were eukary- totrophic biomass was very low (0.2 mg l-1) and the struc- otic cells, which precisely in the middle of the water col- ture of phototrophs was different, as photosynthetic bacte- umn formed a subsurface biomass peak. In summer, a sub- ria were the dominant group. Their contribution to total surface peak of biomass or chlorophyll concentration in phototrophic biomass reached nearly 70% [5]. However, deeper water layers was recorded in oligo- and mesotrophic the density of these microorganisms in the studied lake is lakes, e.g. in the lower metalimnion or upper hypolimnion much lower than in the highly stratified meromictic lakes, of Lakes Huron and Michigan [20] or in the metalimnion which were examined in detail various aspects of the eco- of Lakes Konstanz, Maggiore and Stechlin [21-23]. They physiology of purple sulfur bacteria [30]. are formed by one or several planktonic species, mostly diatoms, cyanobacteria or flagellates of various taxonomic The size structure of phototrophs is an important pa- groups [24-25]. In Lake Strzeszyńskie the subsurface dom- rameter characterizing food resources in the pelagic zone. inants were cryptophytes and diatoms, at a ratio of about Many consumers feed on the smallest organisms, as they 13:1. Such a maximum, but closer to the lake surface alt- are abundant in open water and their sedimentation is hough still in the metalimnion, was observed in another eu- slower than of larger microorganisms. Thus it is important trophic lake in the Wielkopolska region, but was formed by to distinguish between the larger phototrophs and their microplanktonic filamentous cyanobacteria [26]. smallest fraction, i.e. picophototrophs. Moreover, as noted by Pinel-Alloul et al. [31], size structure is needed to un- Species composition of eukaryotic phototrophs varied derstand clearly the mechanisms by which grazers and nu- significantly in the water column of Lake Strzeszyńskie, trients produce variable chlorophyll-a yields per unit of TP with green algae, dinophytes, diatoms or cryptophytes pre- in the lake. Thanks to inclusion of the smallest and usually dominating at successive depths. Cryptophytes were major also most abundant size fraction of phototrophs, i.e. pico- contributors to biomass, dominant at the depth of 8 m, in plankton, total phototrophic plankton abundance in Lake the area with cooler water and less light but higher nutrient Strzeszyńskie was much higher than in earlier reports [10]. concentrations than in the epilimnion. It is known that In the water column during lake stratification, picophoto- cryptophytes may increase rapidly at optimal conditions trophic abundance was high and varied widely (2 orders of [27-28], and are often the only abundant phytoplanktonic magnitude). Despite their small size, picoplankton contrib- cells in transitional conditions. The additional pigment uted substantially to phototrophic biomass, on average phycoerythrin enables their survival under poor light con- about 20%. On the whole, the contribution of picoplankton ditions, while their capacity for phagotrophy allows them to total phototrophic biomass was not very high in compar- to compensate for the shortage of photosynthates [29]. ison with Lobelia lakes in the Bytów Lakeland in northern Metalimnion, particularly its lower part, is a special Poland [5] but were within the range found in lakes with a level in the water column of the study lake. In the middle low trophic level [7].

361 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

Picophototrophs included both cyanobacteria and eu- both single-celled and colonial. The picochlorophytes were karyotic cells (chlorophytes) as well as phototrophic bac- a lot less frequent, they formed dense populations only in teria. Cyanobacteria were much more numerous than chlo- the upper part of the hypolimnion. The joint contribution rophytes, which confirms an earlier opinion about the prev- of pico- and nanoplankton in the water column ranged from alence of cyanobacteria in respect of abundance in fresh- 25 to 77%. In the epilimnion it was about 30%, in the upper water habitats in summer [7]. As shown by earlier research part of metalimnion it reached 50%, and exceeded this in Polish lakes, they occurred as single cells and colonies value only in the lower part of hypolimnion, where they of diverse morphology. In the upper part of the water col- accounted for 70% of total phototrophic biomass. In the umn, more picocyanobacteria were colonial than single- microplankton, the highest biomass was produced by the celled. Similarly, in summer in the water column of Lake filamentous cyanobacteria and flagellates (members of the Czarne, colonial picocyanobacteria were more numerous genera Ceratium, Peridinium, Cryptomonas, Dinobryon). than single cells in epilimnion [19]. The causes of appearance of picocyanobacterial colonies are still unclear but usually they are associated with nutrient limitation in sum- ACKNOWLEDGEMENTS mer in the epilimnion and/or consumer pressure [7]. Colo- nial forms or other cell aggregations may be treated as a This research was financed by the Polish Ministry of survival strategy in less favourable nutrient conditions [31- Education Science, grant No. NN 304 385538. Microscopy 34]. To explain the origin of multicellularity, the transition analyses were possible thanks to the sponsorship of the between colonies and unicellular forms, differential gene Foundation for Polish Science, which bought the Olympus expression analysis may supply new information, as in the BX-60 microscope (programme SUBIN). The authors case of gene identification exercise in Desmodesmus, thank Dr Tomasz Joniak for his support during sampling at which is a potential model for elucidating the origin of mul- Lake Strzeszyńskie. The manuscript was partly translated ticellularity [35]. and edited by Sylwia Ufnalska, MSc, MA.

Eukaryotic PPP was much less numerous than picocy- The authors have declared no conflict of interest. anobacteria, and its occurrence below 8 m, where water temperature was nearly 2-fold lower than at the lake surface, is consistent with reports that eukaryotic PPP prefers REFERENCES colder waters, but richer in nutrients [36-37, 18]. Very much like in the mesotrophic Lake Czarne [19], their pres- [1] Happey-Wood, C.M. (1991) Temporal and spatial patterns in ence was detected only in the bottom layer of water (at 8 m the distribution and abundance of pico, nano and microphyto- or deeper) but its abundance was lower, although Lake plankton in an upland lake. Freshwater Biology, 26, 453-480. Strzeszyńskie, located within an urban agglomeration, is [2] Beaty, M.H. and Parker, B.C. (1996) Relative importance of regarded as more fertile than Lake Czarne in the Drawa pico-, nano-, and microplankton to the productivity of moun- National Park. In contrast, in a very fertile dam reservoir, tain lake, Virginia. Hydrobiologia, 331, 121-129. eukaryotic PPP was exceptionally abundant, reaching sim- [3] Szeląg-Wasielewska, E. (1997) Picoplankton and other size ilar densities as picocyanobacteria in other eutrophic water groups of phytoplankton in various shallow lakes. Hydrobio- bodies [18]. Moreover, eukaryotic PPP cells were slightly logia, 342/343, 79-85. larger than picocyanobacteria, which is commonly ob- [4] Szeląg-Wasielewska, E. (2003) Phytoplankton community served and was already reported earlier from this lake and structure in non-stratified lakes of Pomerania (NW Poland). other Polish lakes, as well as in other countries [36, 39]. Hydrobiologia, 506-509, 229-236. The picoplanktonic chlorophytes recorded in the study lake [5] Szeląg-Wasielewska, E. (2010) Vertical distribution of photo- are taxa found in various aquatic habitats. Choricystis mi- trophs in the pelagic zone of three small Lobelia lakes. Ocean- nor, which was most abundant, is common e.g. in ponds, ological and Hydrobiological Studies, 39(2), 121-133. lakes, and rivers. It is noteworthy, however, that in respect [6] Weisse, T. (1993) Dynamics of autotrophic picoplankton in of species composition, freshwater eukaryotic PPP is less marine and freshwater ecosystems. In Advances in Microbial studied than marine eukaryotic PPP, and the list of its taxa Ecology. Edited by J.G. Jones. Plenum Press, New York, 13, 327-370. is continuously supplemented [23, 40-41]. [7] Stockner, J.G., Callieri, C. and Cronberg, G. (2000) Picoplank- In conclusion, the results of this study was aimed to ton and other non-bloom-forming cyanobacteria in lakes. In The indicate the need for a holistic approach to the photosyn- ecology of Cyanobacteria. Edited by B.A. Whittion and M. Potts. thetic organisms and to show that at increasing depths dif- Kluwer Academic Publishers, Netherlands, 195-231. frent groups of phototrophs play the role as major produc- [8] Polat, S. (2006) Size-fractionated distribution of the phytoplank- ers. In the shallower waters layers the contributions of the ton biomass in the Iskenderum Bay north-eastern Mediterranean cyanobacteria and the eukaryotes algae to the total biomass Sea. Fresenius Environmental Bulletin, 15(5), 417-423. were very similar. Starting from 10 m anoxyphotobacteria [9] Jańczak, J. and Sziwa, R. (1995) A morphometric and hydro- increased in importance and near the bottom their contri- logical characteristic of lakes and reservoirs of Poznań (in polish). In Surface waters of Poznań. Water problems of urban bution to phototrophic biomass exceed 50%. The smallest areas. Edited by A. Kanieckiand J. Rotnicka, Sorus, Poznań, phototrophs were dominated by the picocyanobacteria, 345-355.

362 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

[10] Szeląg-Wasielewska, E. (1999) Seasonal fluctuations of phy- [25] Gervais, F. (1997) Diel vertical migration of Cryptomonas and toplankton as an expression of changes in the pelagic zone of Chromatium in the deep chlorophyll maximum of a eutrophic lakes (in polish). In Natural and anthropogenic transfor- lake. Journal of Plankton Research, 19(5), 533-550. mations of lakes. Edited by A. Choiński and J. Jańczak. IM- iGW, Warszawa, 245-254. [26] Szeląg-Wasielewska, E. and Fyda, J. (2006) Pelagic biocoeno- sis in a stratified eutrophic lake: vertical variation in the mi- [11] MacIsaac, E.A. and Stockner, J.G. (1993) Enumeration of crobial loop and phytoplankton. Limnological Review, 6, 269- phototrophic picoplankton by autofluorescence microscopy. 276. In The handbook of methods in aquatic microbial ecology. Ed- ited by P.F. Kemp, B.F., Sherr, E.B. Sherr E.B. and J.J. Cole [27] Dokulil, M. (1988) Seasonal and spatial distribution of crypto- J.J. CRC Press, Boca Raton, FL, 187-97. phycean species in the deep, stratifying lake Mondsee and their role in the food web. Hydrobiologia, 161, 185-201. [12] Porter, K.G. and Feig, Y.S. (1980) The use of DAPI for identifying and counting aquatic microflora. Limnology and [28] Gervais, F. (1998) Ecology of cryptophytes coexisting near a Oceanography, 25, 943-948. freshwater chemocline. Freshwater Biology, 39, 61-78. [13] Eloranta, P. (1985) Hypolimnetic chlorophyll maximum by [29] Klaveness, D. (1988) Ecology of the Cryptomonadida: a first algae and sulphur bacteria in one eutrophic pond. Archiv für review. In Growth and reproductive strategies of freshwater Hydrobiologie - Supplement, 71(3), 459-469. phytoplankton. Edited by C.D. Sandgren. Cambridge Univer- sity Press, Cambridge, 105-133. [14] Reynolds, C.S. (1984) The ecology of freshwater phytoplankton. Cambridge University Press, Cambridge, pp. 384. [30] Rogozin, D.Yu., Zykov, V.V. and Degermendzhi, A.G. (2012) Ecology of purple sulfur bacteria in the highly stratified [15] Lunina, O.N., Bryantseva, I.A., Akimov, V.N., Rusanov, I.I., meromictic Lake Shunet (Siberia, Khakassia) in 2002-2009. Rogozin, D.Yu., Barinova, E.S., Lysenko, A.M. and Pimenov, Mikrobiology, 81(6), 727-735. N.V. (2007) Seasonal changes in the structure of the anoxy- genic photosynthetic bacterial community in Lake Shunet, [31] Pinel-Alloul, B., Bourbonnais, N. and Dutilleul, P. (1996) Khakassia. Microbiology, 76(3), 368-379. Among-lake and within-lake variation of autotrophic pico- and nanoplankton biomass in six Quebec lakes. Canadian Journal [16] Rogozin, D.Yu., Zykov, V.V, Chernetsky, M.Y., Degermen- of Fisheries and Aquatic Sciences, 53(11), 2433-2445. dzhy, A.G. and Gulati, R.D. (2009) Effect of winter conditions on distributions of anoxic phototrophic bacteria in two [32] Passoni, S. and Callieri, C. (2000) Picocyanobacteria single meromictic lakes in Siberia, Russia. Aquatic Ecology, 43, forms, aggregates and microcolonies: survival strategy or spe- 661–672. cies succession? Verhandlungen des Internationalen Verein Limnologie, 27, 1879-1883. [17] Rogozin, D.Yu., Trusova, M.Yu., Khromechek, E.B. and De- germendzhi, A.G. (2010) Microbial community of the chemo- [33] Schallenberg, M. and Burns, C.W. (2001) Tests of autotrophic cline of meromictic Lake Shunet (Khakassia, Russia) during picoplankton as early indicators of nutrient enrichment in an summer stratification. Microbiology, 79(2), 253-261. ultraoligotrophic lake. Freshwater Biology, 46, 27-37.

[18] Szeląg-Wasielewska, E. (2006) Trophic status of lake water [34] Komárková, J. (2002) Do cyanobacterial pikoplankton exist in evaluated using phytoplankton community structure – change eutrophic reservoirs? Verhandlungen des Internationalen Ver- after two decades. Polish Journal of Environmental Studies, ein Limnologie, 28, 497-500. 15(1), 139-144. [35] Lawton, S., Hayes, P. and Shubert, E. (2012) Desmodesmus is [19] Szeląg-Wasielewska, E. (2005) Autotrophic picoplankton dur- a potential model for elucidating the origin of multicellularity ing summer stratification: vertical distribution and contribu- and an ethical model for investigating cancer. In Algae in hu- tion to total phytoplankton. In Lakes and artificial water reser- man’s environment. Taxonomy, ecology and role in ecological voir – functioning, revitalization and protection. Edited by status assessment. The 31st International Conference of the A.T. Jankowski and M. Rzętała. University of Silesia – Fac- Polish Phycological Society, Olsztyn, May 17-20 2012, 30. ulty of Earth Sciences, Polish Limnological Society, Polish Geographical Society – Branch, Katowice-Sosnowiec, 231- [36] Pick, F.R. and Agbeti, M. (1991) The seasonal dynamics and 240. composition of photosynthetic picoplankton communities in temperate lakes in Ontario, Canada. Internationale Revue der [20] Fahnenstiel, G.L. and Carrick, H.J. (1992) Phototrophic picoplankton in Lakes Huron and Michigan: abundance, distribu- gesamten Hydrobiologie und Hydrographie, 76, 565-580. tion, composition, and contribution to biomass and production. [37] Sondergaard, M. (1991) Phototrophic picoplankton in temper- Canadian Journal of Fisheries and Aquatic Sciences, 49(2), ate lakes: seasonal abundance and importance along a trophic 379-388. gradient. Internationale Revue der gesamten Hydrobiologie [21] Weisse, T. and Schweizer, A. (1991) Seasonal and interannual und Hydrographie, 76, 505-522. variation of autotrophic picoplankton in a large prealpine lake [38] Stockner, J.G. and Shortreed, K.S. (1991) Autotrophic pico- (Lake Constance). Verhandlungen des Internationalen Verein plankton: Community composition, abundance and distribu- Limnologie, 24, 821-825. tion across a gradient of oligotrophic British Columbia and [22] Callieri, C. and Pinolini, M.L. (1995) Picoplankton in Lake Yukon Territory lakes. Internationale Revue der gesamten Hy- Maggiore, Italy. Internationale Revue der gesamten Hydrobi- drobiologie und Hydrographie, 76, 581-601. ologie und Hydrographie, 80, 491-501. [39] Szeląg-Wasielewska, E. (2004) Dynamics of autotrophic pico- [23] Padisák, J., Krienitz, L., Koschel, R. and Nedoma, J. (1997) plankton communities in the epilimnion of a eutrophic lake Deep-layer autotrophic picoplankton maximum in the oligo- (Strzeszyńskie Lake, Poland). Annales de Limnologie - Inter- trophic Lake Stechlin, Germany: origin, activity, development national Journal of Limnology, 40(2), 113-120. and erosion. European Journal of Phycology, 32, 403-416. [40] Hepperle, D. and Krienitz, L. (2001) Systematics and ecology [24] Moll, R.A. and Stoermer, E. (1984) A hypothesis relating of chlorophyte picoplankton in German inland waters along a tropic status and subsurface chlorophyll maxima of lakes. Ar- nutrient gradient. Internationale Revue der gesamten Hydrobi- chiv fur Hydrobiologie, 94, 425-440. ologie und Hydrographie, 86(3), 269-284.

363 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

[41] Fawley, K.P. and Fawley, M.W. (2003) Diversity of the pico- plankter Choricystis (Trebouxiophyceae, Chlorophyta) from Minnesota and north Dakota lakes. Journal of Phycology, 39(Suppl. 1), 16.

Received: May 09, 2014 Revised: July 01, 2014 Accepted: July 03, 2014

CORRESPONDING AUTHOR

Elżbieta Szeląg-Wasielewska Department of Water Protection Faculty of Biology Adam Mickiewicz University in Poznań Umultowska 89 61-614 Poznań POLAND

E-mail: [email protected]

FEB/ Vol 24/ No 1b/ 2015 – pages 355 - 364

364 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

EFFECT OF SEASONAL CHANGES ON ANTIOXIDANT VITAMINS (A, E, C) AND MDA IN LEVELS OF THE RAINBOW TROUT (ONCORHYNCHUS MYKISS WALBAUM, 1792) GROWN IN VAN CITY, CATAK DISTRICT

İbrahim Hakkı Yörük, Asli Çilingir Yeltekin and Halit Demir*

Division of Biochemistry, Department of Chemistry , Faculty of Science, University of Yuzuncu Yil, 65080, Van, Turkey

ABSTRACT organisms, fish also developed many defense mechanisms in their body to prevent oxidative stress and its damage The present study was aimed to determine the seasonal caused by these. These are known as antioxidant defense changes of malondialdehyde (MDA), vitamin retinol (vita- systems and consist of materials such as super oxide dis- min A), α-tocopherol (vitamin E) and ascorbic acid (vita- mutase, catalase, glutathione peroxidase with enzymatic min C) levels. For this purpose, fishes were collected in the qualities and non-enzymatic glutathione, A, E, C vitamins, second week of the second month in each four seasons. selenium and melatonin [5, 6]. Antioxidants function as Fishes were fed with the same food for one year. Fishes structures which prevent free radical formation or break were selected from the same pool. Statistically significant chains. The research showed that antioxidants neutralize differences were found between the levels of vitamin A, free radicals and prevent cell damage [7]. vitamin E, vitamin C and MDA (p<0.0001) levels. Conse- It was determined that β-carotene which is the precur- quently, it was determined that the levels of vitamin (A, E, sor of Vitamin A suppressed singlet oxygen, cleaned su- C) showed significant differences between seasons and peroxide radical and function as an antioxidant by interact- there was a marked increase in the vitamin levels especially ing with peroxide radicals [8]. Due to its fat soluble quality, in the winter season. The present study is the first study in it is effective as lipid flow antioxidant in the membranes of this region. More studies are required to clarify the results. cellular and subcellular structures. It reduces singlet oxy-

gen to inactive oxygen [9].

KEYWORDS: Vitamin E (-tocopherol) is a very strong antioxidant, Rainbow trout, vitamin A, vitamin E, vitamin C, MDA and it forms the first defense line which protects polyunsaturated fat acids existing in cell membrane phospholipids

from the effect of free radicals. Vitamin E reduces super- 1. INTRODUCTION oxide and hydroxyl radicals, singlet oxygen, lipid peroxide radicals and other radicals. Vitamin E is known as chain Aquaculture has been developing rapidly in recent breaking antioxidant. Lipid peroxidation chain reaction can years around the world. The rainbow trout (Oncorhynchus be ended by vitamin E [10]. mykiss) has an important place in freshwater aquaculture Ascorbic acid is also a strong antioxidant due to its [1]. In this respect, the rainbow trout became an important strong reducing activity. It goes into reaction with super - point of focus both economically and also food wise [2]. oxide radical (O2 ) and hydroxyl radical (OH) and cleans In biological systems, free radical reactions have an them from the environment [11]. important place. These compounds can both be formed dur- The most affected components from free radicals are ing the functioning of normal metabolism in the organism lipids. Lipid peroxidation, starts by removing an unsatu- and also with the effects of external factors [3]. The most rated hydrogen atom from fatty acid chain existing in the important free radicals in biological systems are the ones membrane structure as the result of the free radical effect consisting of oxygen. It was observed in recent studies re- formed in organism. With the change in the positions of active types of oxygen (ROS) maybe somewhat the cause double bond in the molecule, later lipid peroxyl radical is of damages and diseases in molecules by impairing the nor- formed as the result of other conjugants, with lipid radicals mal functions by means of oxidation [4]. Like all aerobic interacting with molecular oxygen. The most important peroxidation product is malondialdehyde. While lipid per- * Corresponding author oxyl radicals affect other unsaturated fat acids in membrane

365 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

structure and cause the formation of new lipid radicals, 200 µl plasma was taken for this purpose, 200 µl etha- they receive the hydrogen atoms which are liberated turn nol (Merck, Germany) was added and was subject to vortex into lipid hydroperoxides and the event continues by cata- for 1 minute. Later, 600 µl hexane was added (Merck Ger- lyzing itself [8, 12-15]. many) and the vortexing process was repeated for 1 min. Malondialdehyde shows its toxic effect by binding (Mixer VM20). It was centrifuged in homogenate 2000 RPM with protein amino groups, phospholipids and nucleic ac- for 15 minutes (SED 6); the lipophilic flow formed was ids. It is easily diffused from membranes, goes into reac- taken. Hexane application process was repeated twice and tion with nitrogen base existing in the structure of DNA the flow collected was dried under nitrogen gas flow. Res- and shows genotoxic effects. Since the measurement of li- idue was dissolved in 200 µl methanol (Merck, Germany) pid peroxidation is accepted as the good indicator of dam- and was injected into HPLC column. age in the tissues; the measurement of conjugated dienes To analyze muscle samples, 1 g fish muscle was taken formed during peroxidation, is an important method which into plastic tubes, 2 ml ethanol was added in order to settle will reflect the in vivo lipid peroxide level [3]. proteins on, and was vortexed for 1 minute. Then it was The present study was aimed to determine the seasonal homogenized by adding 3 ml hexane (Wiggenhauser D- changes of malondialdehyde (MDA), vitamin retinol (vita- 130) and was vortexed again for 1 minute. It was centri- min A), α-tocopherol (vitamin E) and ascorbic acid (vita- fuged in homogenated 2000 RPM for 15 minutes and the min C) levels. lipophilic flow formed was obtained. Hexane application system was repeated twice and flows collected were dried under nitrogen gas flow. Residue was dissolved in 200 µl 2. MATERIALS AND METHODS methanol and amounts of vitamin were determined with HPLC like the process applied in plasma samples. 2.1 Sampling In the separation of vitamin A and E levels with HPLC In this study, a total of 40 rainbow trout (250-300 g) (Thermo Sentific Pinligan Surveyor, USA) C 18 column, were used. Fish were collected from the trout hatchery of (25 cm x 4.6 mm Supelco, USA) methanol-water (98:2) the company “EL-FA Su ve Su Ürünleri Ith. İhr. Ltd. Şti” was used with mobile flow in 1.5 ml/min. flow speed. in Van-Çatak, Elmacı village during 2010 at the end of sec- Measurements were done with diodearray detector in 325 ond week of each season, in January, April, July and Octo- and 290 nm wavelengths [17]. ber. Existing conditions of the fish pool in which the 2 year old fish were kept are given in the following Table 1. 2.2.2 Determination of vitamin C and MDA Vitamin C and MDA levels in fish plasma and muscles TABLE 1 - Environmental conditions of the fish pool were determined with HPLC-UV method [18]. Temperature Dissolved oxygen Months pH For plasma, 250 µl 0.1 M HClO4 (Merck, Germany) (˚C) (mg/l) solution was added over 200 µl plasma and was vortexed January 8.6 10.04 7.27 for a couple of seconds and 550 µl distilled water was April 11.0 9.45 7.35 added, after repeating vortex process, it was centrifuged in July 15.8 8.98 6.01 4500 RPM for 10 minutes. After these processes flow re- October 13.2 9.12 6.91 maining on top is removed carefully and is placed in a vial, vitamin C and MDA levels were determined with HPLC.

Fish were taken out from the pool one by one without For tissue analysis, 1 g tissue was weighed, 1250 ml changing their environment at the facility and then placed 0.1M HClO4 was added on top and homogenized. Later, in a bucket, containing anesthetic MS-222 (3-aminoben- vortexing is repeated by adding 2750 ml distilled water and zoic acid ethyl Esther methane sulphonate salt) (Merck, centrifuged at 4500 rpm, for 10 minutes. After these pro- Germany) (200 mg/L). Later, bloods from the tail veins of cesses, flow remaining on top was removed and place in a the fish under anesthesia were taken into heparin tubes. vial, and level of HPLC, vitamin C and MDA were deter- Muscle tissue and blood were kept in (+4) and brought into mined. the research laboratory of the Yüzüncü Yıl University, Fac- Inertial ODS-3 column (25 cm x 4.6 mm GL Sciences ulty of Science Research Laboratory. Muscle tissue was Inc, Japan), 30 µM KH2PO4 (Merck, Germany) - methanol frozen at -55 °C. Blood samples were centrifuged and the (82.5:17.5) was used in HPLC equipment with mobile flow plasma were separated and frozen at -55 °C. The process for 1.2 ml/min flow speed. Reading was taken with UV was repeated every season. detector at 250 nm wavelength.

2.2 Biochemical analysis 2.3 Statistical analysis 2.2.1 Vitamin A and E measurement After levels of vitamin A, E, C and MDA analyzes Fish plasma and vitamin A and E (retinol and α-to- were made on samples obtained from the Rainbow trout for copherol) levels in their blood, were determined by the ex- four seasons, standard deviations in the results obtained traction method indicated by Miller & Yang (1985) [16]. were determined and confidence intervals were created.

366 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

One-way Variant analysis (One-way ANOVA) analysis mer and autumn can be seen in Table 3. The differences in and Duncan multiple comparison test were applied on the the vitamin A, vitamin E, vitamin C and MDA levels in data to determine whether there were seasonal differences the muscle tissue of rainbow trout in winter, spring, sum- between data or not [19]. mer and autumn statistically around p<0.0001 level were determined to be critical. While plasma vitamin A levels were the same in spring, summer and autumn, it reached a 3. RESULTS higher level in winter (p<0.0001). While plasma vitamin E levels were approximately at the same levels in winter, Studies during one year were conducted on fish grown spring and summer, it showed a distinct decline in autumn in same pool for one year. Feeding was done with the same (Table 3). While plasma vitamin C levels were also at the food in the farm where study was conducted during the highest level in winter, a decline was observed in spring, year. Ambient conditions of the pool are given in Table 1. summer and autumn respectively (p<0.0001). While Rainbow trout muscle tissue vitamin A, vitamin E, vit- plasma MDA levels were in their highest levels in spring amin C and MDA levels for winter, spring, summer and and summer it showed a decline in winter and autumn re- spring seasons are given in Table 2. The differences in the spectively (p<0.0001) (Table 3). vitamin A, vitamin E, vitamin C and MDA levels in the muscle tissue of rainbow trout in winter, spring, summer and autumn statistically around p<0.0001 level were deter- 4. DISCUSSION mined to be critical. Vitamin A which was at the highest level during winter season, was determined to be reduced Antioxidants are potential cleaners of free radicals and towards spring, summer and autumn (p<0.0001). While serve as inhibitors of neoplastic processes. It is argued that vitamin E level in the muscle tissue was also high in winter many synthetic and natural antioxidants have a beneficial season, it showed a reduction in summer and autumn. Vit- effect on human health and prevent diseases [20]. In the amin C level muscle tissue which was high in winter also deficiency of vitamin A which has antioxidant quality, was reduced in spring. It was reduced approximately at the some inadequacies in growth, development, sight and dif- same levels during summer and autumn, reached the lowest ferentiation in epithelia cells may occur. Additionally vita- level (p<0.0001). While the MDA levels in muscle tissue min A deficiency may also increase cancer risk [21]. in fish samples was in the lowest level during winter, they A study was conducted on the effects of vitamin A and showed an increase towards spring and reached the highest E on the growth, food assessment and in some tissues of level (p<0.0001) (Table 2). the fry under oxygen stress. In this study the A and E vita- The levels of plasma vitamin A, vitamin E, vitamin C mins which are antioxidants were added to the feed of and MDA levels in rainbow trout by winter, spring, sum- Rainbow trout and were fed for 12 weeks. Fish which were

TABLE 2 - Variation of vitamin A, vitamin E, vitamin C and MDA levels in muscle tissue of the rainbow trout by seasons (n=10)

Vitamin A Vitamin E Vitamin C MDA Season (µg/g) (µg/g) (nmol/g) (nmol/g) Winter 1.170.06a 9.220.27a 205.551.53a 3.300.42c Spring 0.860.05b 7.910.20b 194.421.88b 3.891.10a,b Summer 0.650.06c 6.090.25c 118.803.80c 3.810.95b Autumn 0.480.03d 5.380.15d 113.502.38c 4.150.98a Mean±standard deviation. The difference between the averages with different letters in the same column is statistically important (p<0.0001).

TABLE 3 - Variation of plasma vitamin A, vitamin E, vitamin C and MDA levels in rainbow trout by seasons (n=10)

Season Vitamin A Vitamin E Vitamin C MDA (µg/ml) (µg/ml) (nmol/ml) (nmol/ml) Winter 0.240.01a 2.550.31a 80.922.38a 1.380.04b Spring 0.19  0.01b 2.170.29a 58.191.67b 1.560.03a Summer 0.170.01b 1.950.20a 39.611.15c 1.500.02a Autumn 0.17 0.01b 0.110.01b 21.920.38d 0.770.02c Mean±standard deviation. The difference between the averages with different letters in the same column is statistically important (p<0.0001).

367 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

fed with these rations, in 2 different dissolved oxygen lev- Jensen et al. (1998) [28] created 17 diet groups in rain- els (4.4 mg/l O2 concentration and 7.6 mg/l O2 concentra- bow trout and implemented a diet containing certain amounts tion), the effects of vitamins A and E in some tissues (mus- of -tocopherol, astaxantin and fat. After implementing this cle, liver and kidneys) and blood serum were determined diet, they stored the fish in these groups for 12 to 18 weeks [22]. In this study, the results determined are close to the by fumigation or freezing methods. At the end of the study vitamin A and E levels in the literature [22]. they determined that in muscle tissue of the control groups A study has been conducted for 12 weeks in trout by -tocopherol level was 12.1 mg/kg and MDA level applying different proportions of propolis and vitamin E 3.1mol/kg. In the study, this information obtained sup- diets, in different flow speed (0.9 and 2.1Lmin-1). In the ports the information of literature. study vitamin A, Vitamin E, vitamin C and MDA values Since vitamin E dissolved in fat, vitamin deficiency were searched in muscle, kidney, liver and serum [23, 24]. may be observed in lipid absorption disorders. In the event In the study presented, vitamin A levels found in muscle of deficiency, sensitivity of the erythrocytes to peroxida- tissue and plasma and the serum and muscle tissue values tion increases, and therefore abnormal cell membrane is were seen to be close to each other. While vitamin E levels formed [29]. One of the most important antioxidants pre- were different in muscle tissue, in serum the values are venting lipid peroxidation is vitamin E. The study con- -1 closer in 0.9 Lmin flow rate. ducted showed parallel findings, and it was observed that In the study presented, vitamin A and E levels in mus- in seasons when α–tocopherol level increased, the decline cle tissue showed a meaningful increase in autumn, sum- in MDA level, and in seasons when it was reduced, an in- mer, spring and winder respectively. In plasma again a crease was observed. meaningful increase was seen in winter season (p<0.0001) Huang &Huang (2004) [30] researched the effect of (Table 2). These increases may make us consider the in- vitamin E diet on the development of tissue lipid peroxida- crease in fatty tissues of fish towards winter arising from tion and liver glutation level of Oreochromis niloticus x O. their being vitamins soluble in oil, and therefore the in- Aureus, one of the hybridized Tilapia fish fed by oxidized crease observed in retinol and -tocopherol levels. fat. In the study hybrid Tilapia fish was implemented a li- A study was conducted on the fat soluble vitamins, fat pid diet containing vitamin E for 14 weeks. At the end of acids and cholesterol of the farm grown trout. It was ob- the study they found out that lipid peroxidation with ascor- served that the levels of vitamin A and -tocopherol bat effect in the liver and muscles of the fish fed with a diet (12.4µg/100g, 714mg/100g) were parallel to the vitamin A containing 0 and 40 IU Vit. E/kg vitamin E (MDA control reported; however it was not compatible to the -tocoph- muscle tissue 2.16 nmol/g, liver tissue 3.92 nmol/g) was erol [25]. quite higher compared to those fed with a diet containing higher level of vitamin E. They determined that the level Köprücü & Özdemir (2002) [26] in their study they of liver glutation increased with the vitamin E content in conducted to determine the amounts of some vitamins (A , 2 the diet. MDA results found are compatible with the levels C, E) in the muscle tissue of rainbow trout, they found out of plasma and MDA levels in the tissue. that fish muscle tissue contained 12,26 μg/g vitamin A2, 17.54 μg/g vitamin C, 8.73 μg/g vitamin E on the average They analyzed the changes in antioxidant enzyme ac- and raw fat with a percentage of 4.23%. They stated that tivities in erythrocytes of rainbow trout by implementing the differences in A2, C, E vitamins and raw fat levels in melatonin in Rainbow trout (Oncorhynchus mykiss). In the the muscle tissue samples of rainbow trout were insignifi- study they researched the effects of malondialdehyde cant, (p>0.05) when the muscle tissue of rainbow trout (MDA) (plasma control 0.8-1.2 nmol/ml) which is an indi- used were compared to many other commercial aquacul- cator of lipid peroxidation. In the end they determined that ture, they contained higher percentages of A2, C and E vit- MDA level, as the result of melatonin implementation was amins. Differences are observed between the levels deter- reduced to a significant degree (p<0.01) [31]. MDA results mined in this study and the levels determined in the study. are close to the plasma MDA levels. It is believed that the differences obtained could be caused Yonar & Yonar (2010) [32] conducted a study by sub- by the conditions of the environment fish live, type of feed jecting the rainbow trout to malahite and where they re- and method of feeding, age of the fish and the conditions searched the changes in antioxidant defense systems and in of work in the laboratory. their immunity systems. In the study they subjected the fish Wahli et al. (1998) [27] formed groups of rainbow to malahite in two different percentages (1/15 66.67 mg/l, trout and infected them with Yersinia ruckeri and Ichthy- 1/150 6.67 mg/l). In the end of the implementation, they ophthirius multifiliis (VHS). By applying vitamin C and researched the changes in the lipid peroxide and catalase vitamin E in certain percentages to this fish, they conducted levels. As the result, while an increase twice as much was a study to analyze the immunity system of the fish and the seen in MD levels (control 0.65 nmol/ml) in blood tissue resistance shown by the fish to disease with these deter- compared to control group, in both groups, a higher in- mined amounts of vitamins. No diet was implemented in crease was observed in 6.67 mg/l malahite application. the control group during the study. Control group’s -to- MDA levels in this study overlap with the MDA levels in copherol level is higher compared to the - tocopherol level. the literature [32] (Table 2 and 3).

368 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

A study has been conducted by Tuna Keleştemur [33] [3] Gümüştaş MK, Atukeren P. (2008). Oxidative and nitrositive in muscle, liver and kidney tissues of the trout fry with dif- stress relationship to psychiatric disorders. IU Cerrahpaşa Faculty of Medicine continuous medical training events, Fre- ferent oxygen levels (7mg/l, 4.5mg/l, 3.5mg/l). In this quent psychiatric disorders in Turkey, symposium series, 62: study, antioxidant vitamins (A,C,E) and MDA level of 329-340. trout were examined. While the MDA levels indicated in [4] Somogyi A, Rosta K, Pusztai P, Tulassay Z. Nagy G. (2007). the study show closeness with the 4.5mg/l oxygen in MDA Antioxidant measurements. Physiol Meas, 28 (4), 41-55. levels in the work presented, it was observed that the value [5] Dautremepuits C, Betoulle S, Vernet G. (2003). Stimulation of was lower than the 3.5mg/l oxygen level, and above 7mg/l antioxidant enzymes levels in carp (Cyprinus carpio L.) in- oxygen. Vitamin C levels determined in each three oxygen fected by Ptychobothrium sp. (Cestoda). Fish and Shellfish levels were observed to be lower than the values found in Immunology. 15, 467-471. the present study. It was determined that the vitamin A lev- [6] Trenzado C, Carmen HM, Gallego MG, Morales AE, Furne els found in the study by Tuna Keleştemur [33], overlapped M, Domezain A, Domezain J, Sanz A. (2006). Antioxidant en- the values found in the present study. On the other hand, zymes and lipid peroxidation in sturgeon Acipencer naccarii vitamin E levels were found to be a little higher in the pre- and trout Oncorhynchus mykiss. A comparative study. Aqua- sent study (Tables 2 and 3). culture. 254, 758-767. The result of this research clearly shows that seasonal [7] Gökpinar Ş, Koray T, Akçiçek E, Göksan T, Durmaz Y. (2006). Algale antioxidants. E.Ü. Aqua Culture Magazine, 23: changes have an effect on the antioxidant vitamin (A, E, C) 85-89. levels and MDA level of Rainbow trout. It is believed that these differences observed may be caused by the pH of the [8] Akkuş İ. (1995). Free Radical and their physiophatological effects, 1. Mimoza Publishing, Konya, 395. water in the season, water temperature and the amount of dissolved oxygen associated with this (Table 1). The changes [9] Kayaalp O. (2000). Medical Pharmacology from the point of rational treatment . Volume 1. 9th edition. Ankara: Hacettepe observed especially in autumn, is believed to have arisen TAŞ.s.718. from its corresponding to the reproduction period of rainbow trout. It can also be considered that partial closeness [10] Dogru Pekiner B, Daş Evcimen N, Ulusu NN, Bali M, Karasu C. (2003). Effect of vitamin E on microsomal Ca(2+) ATPase of the spring values to winter season values arises from the activity and calcium levels in strepto zotocin-induced diabetic snow in the mountains in the region melting in spring and rat kidney. Cell Biochem Funct. 2:177-82. mixing to spring water and this changing the environmen- [11] Harris R. (1996). Editor. Haris R,. Ascorbic acid.: Subcellular tal conditions less. Biochemistry, 25:Springer. Today, when taking dietary antioxidant is very im- [12] Slater TF. (1984). Overwiew of methods for detecting lipid pe- portant, the opinion which can be reached with respect to roxidation. Methods in Enzymology, 105: 283–305. the study conducted is, the increase of rainbow trout con- [13] Kavas GÖ. (1989). Free radicals and their effects on organism, sumption as antioxidant vitamin source, especially its con- Turkish Clinics Magazine, 9 (1): 1-8. sumption is more beneficial in winter months. [14] Moslen MT. (1994). Reactive oxygen species in normal physiology, cell injury and phagocytosis, free radicals in diagnostic medicine, Ed D Armstrong Plenum Press, NewYork, Pp, 1-15.

ACKNOWLEDGEMENTS [15] Kaçmaz M, Atmaca M, Arslan A, Demir H, Ozbay MF. (2014). Oxidative stress in patients with thyroidectomy and thyroparathyroidectomy under replacement therapy. Endo- We thank Directorate of Scientific Research Projects crine. 2014 Apr 24. [Epub ahead of print] of Yüzüncü Yıl University for their financial support. [16] Miller KW, Yang CS. (1985). An Isocratic High-performance We thank Van-Çatak Elmacı village EL-FA Water and liquid chromatography method for the simultaneous analysis Aquaculture Import Export Ltd. Co. hatchery for supplying of plasma retinol, α-tocopherol and various carotenoids. trout and Directorate of Scientific Research Projects of Analitical Biochemistry, 145: 21-26. Yüzüncü Yıl University for their financial support. [17] Zaspel BJ, Csallary AS. (1983). Detemination of alpha-tocopherol in tissues and plasma by high-performance liquid The authors have declared no conflict of interest. chromatography. Analitical Biochemistry. 130(1): 146–150. [18] Karatepe M. (2004). Simultaneous determination of ascorbic acid and free malondialdehyde in human serum by HPLC-UV. LCGC North America, 22: 362-365.

REFERENCES [19] Özdamar K. (2004). Statistical data analysis with package programs 1.5th Edition. Kaan Publishing, Eskişehir. 352. [1] Sağlam YS, Işık N, Arslan A, Erer H. (2006). Erzurum bölge- [20] Bagchi D, Bagchi M, Stohs SJ, Das DK, Ray SD, Kuszynski sindeki gökkuşağı alabalıklarında (Oncorhynchus mykiss w CA, Joshi SS, Pruess HG. (2000). Free radicals and grape seed .1792) Aeromonas Hyrdophila ve Yersinia ruckeri İzolasyonu proanthocyanidin extract: importance in human health and dis- ve patolojik incelemeler, Atatürk Üniversitesi Veteriner Bilim- ease prevention. Toxicology, 148 (2) 3, 187-197. leri Dergisi, 1(2), 6-10. [21] Kalaycıoğlu L, Serpek B, Nizamlıoğlu M, Başpınar N, Tiftik [2] Turchini GM, Torstensen BE, Ng WK. (2009). Fish oil re- AM. (2006). Biochemistry, 3rd Edition, Nobel Publishing, Dis- placement in finfish nutrition. Rev. Aquacult. 1, 10–57. tribution Ltd. Şti., Ankara.

369 © by PSP Volume 24 – No 1b. 2015 Fresenius Environmental Bulletin

[22] Keleştemur GT. (2009). Rainbow trout under oxygen stress (Oncorhynchus Mykiss W. 1792) effect of “A” and “E” vitamins on the fry growth, feed assessment and on some tissues (Master’s Thesis). Fırat University, Science Institute, Elazığ.

[23] Tuna Keleştemur G, Seven İ. ( 2012). Effects of dietary propolis and vitamin E on growth performance and antioxidant status in juvenile rainbow trout (Oncorhynchus mykiss) under different flow rate. Aquaculture Research, 1–12. [24] Tuna Keleştemur G, Tatli Seven P, Yılmaz S. (2012). Effects of dietary propolis and vitamin E on growth performance and antioxidant status in blood of juvenile Rainbow trout, On- corhynchus mykiss (Teleostei: Salmoniformes) under different flow rates. Zoologıa 29 (2): 99–108. [25] Harlioğlu AG. (2012). Fatty Acid Composition, Fat Soluble Vitamins and Cholesterol Content of Farmed Rainbow Trout (Oncorhynchus mykiss). Pakistan J. Zool. vol. 44 (4), pp. 1013-1019. [26] Köprücü K, Özdemir Y, (2002). Amounts of A , C and E vit- 2 amins in Rainbow Trout flesh (Oncorhynchus mykiss Wal- baum, 1792). F.U. Science and Engineering Sciences Maga- zine, 14(2): 227-232. [27] Wahli T, Verlhac V, Gabaudan J, Schüep W, Meier W. (1998). Influence of combined vitamins C and E on non-specific immunity and disease resistance of rainbow trout, (Oncorhyn- chus mykiss). Journal of Fish Diseases, 21:127-137. [28] Jensen C, Birk E, Jokumsen A, Skibsted HL, Bertelsen G. (1998). Effect of dietary levels of fat, a-tocopherol and astaxanthin on colour and lipid oxidation during of frozen rainbow trout (Oncorhynchus mykiss) and during chill storage of smoked trout. Z Lebensm Unters Forsch A 207: 189–196

[29] Mert N, Bildik A, Ertekin A, Dede S. (1999). Biochemistry, Y.Y.U. Veterinary Faculty Publishing, Van.

[30] Huang CS, Huang SL. (2004). Effect of dietary vitamin E on growth, tissue lipid peroxidation, and liver glutathione level of juvenile hybrid tilapia, Oreochromis niloticus x O. aureus, fed oxidized oil. Aquaculture. 237, 381–389.

[31] Gülçin İ, Beydemir Ş, Hisar O, Köksal E, Reıter RJ. (2009). Melatonin administration increases antioxidant enzymes activities and reduces lipid peroxidation in the rainbow trout (On- corhynchus mykiss, Walbaum) erythrocytes. Turk. J. Vet. Anim. Sci. 33(3): 241-245. [32] Yonar ME, Yonar SM. (2010). Changes in selected immuno- logical parameters and antioxidant status of rainbow trout exposed to malachite green (Oncorhynchus mykiss, Walbaum, Received: May 12, 2014 1792). Pesticide Biochemistry and Physiology, 97: 19–23. Accepted: October 14, 2014

[33] Tuna Keleştemur G. (2012). The Antioxidant Vitamin (A, C, E) and the Lipid Peroxidation Levels in Some Tissues of Ju- venile Rainbow trout (Oncorhynchus mykiss, W. 1792) at Dif- CORRESPONDING AUTHOR ferent Oxygen Levels. Iranian Journal of Fisheries Sciences. 11(2) 315-324. Halit Demir Division of Biochemistry Department of Chemistry Faculty of Science University of Yuzuncu Yil 65080 Van TURKEY

Phone: +90 4322251706 Fax: +90 4322251706 E-mail: [email protected]

FEB/ Vol 24/ No 1b/ 2015 – pages 365 - 370

RESPONSES OF ENZYME ACTIVITIES AND MICROBIAL COMMUNITY STRUCTURE IN AN AERATED INTEGRATED VERTICAL-FLOW CONSTRUCTED WETLAND

Min Tao1,2, Feng He1, Sheng Fang Li2, Zhen-Bin Wu1 and Qi Shuo Wang1,3,*

1State Key Laboratory of Freshwater Ecology and Biotechnology, Institute of Hydrobiology, Chinese Academy of Sciences, Wuhan 430072, China; 2Hubei Key Laboratory of Mine Environmental Pollution Control & Remediation, School of Environmental Science and Engineering, Hubei Polytechnic University, Huangshi 435003, China; 3Huai’an Research Center, Institute of Hydrobiology, Chinese Academy of Sciences, Huai’an 223006, China.

ABSTRACT Oxygen limitation led to the incomplete nitrification and lower aerobic decomposition of organic matter in con- An aerated integrated vertical-flow constructed wet- structed wetlands [4]. land (IVCW) was used to treat domestic sewage, and In recent years, many studies have focused on improv- changes in enzyme activities and microbial community ing oxygen levels within constructed wetlands through some structure in the aerated IVCW were investigated. The re- design and manipulations such as the vertical-flow pre-treat- sults showed that aeration increased Eh of all sites, and en- ment filter [5], frequent water level fluctuation [6], passive larged the depth of aerobic zone from 33cm to 69cm in the air pumps [7, 8] and artificial aeration in the gravel bed summer and from 17cm to 54cm in the winter. Artificial [9,10], but it remained unclear how artificial aeration could aeration significantly enhanced dehydrogenase and urease improve pollutants removal and what changes were hap- activities, while phosphatase activity was not improved. pened within constructed wetlands, such as redox potential, Four kinds of saturated phospholipid fatty acid (PLFAs) enzyme activities and microbial community structure. were not determined in aerated IVCW, while artificial aer- Many researchers have pointed out that microorganism ation obviously increased the relative contents of monoun- decomposition was the main mechanism of wastewater pu- saturated PLFAs, suggesting that the proportion of aerobic rification in wetlands, and enzymes also played an im- microorganisms significantly increased. More importantly, portant effect on nutrient cycling and decomposition [11]. intermittent aeration had little effect on the relative contents of cyclopropane PLFAs at the bottom layer, indicat- Phospholipid fatty acids (PLFAs) are important coming that anaerobic denitrification and the proportion of an- ponents of a microorganism cell membrane, and can help aerobic microorganisms were not affected. These results us to characterize microorganism community structure could demonstrate why the aerated IVCW had a better [12]. PLFAs have been successfully utilized to measure the treatment performance. viable microbial biomass and community structure in soil and sediment [13, 14].

In this study, changes in redox potential, enzyme ac- KEYWORDS: integrated vertical-flow constructed wetland (IVCW); tivities and microbial community structure in aerated intermittent aeration; enzyme activity; microbial community struc- IVCW and control were studied. The objectives are: (1) to ture; responses evaluate the effect of artificial aeration on substrate redox potential and redox zones and (2) to disclose the response of enzyme activities and microbial community structure to 1. INTRODUCTION artificial aeration, as well as the relationship between different microorganism groups and redox potential. Owing to the advantages of low construction and maintenance expenses, high efficiency and better ecological services, constructed wetlands (CWs) have been widely 2. MATERIAL AND METHODS used to treat different types of wastewater [1]. Temperature and DO are two key factors determining removal efficien- 2.1. Experimental systems cies in constructed wetland systems. Pollutants removal ef- Two small-scale plots were set up in Wuhan, China. ficiencies dramatically decrease in the cold seasons, prob- Each plot (2 m2) was divided equally into two chambers: a ably due to low temperatures and poor oxygen levels [2, 3]. down-flow chamber and an up-flow chamber, which was

connected at the bottom. Each chamber was composed of * Corresponding author two different layers of substrate with different sizes: the 15 cm

Down-flow chamber Up-flow chamber Canna generalis Acorus tartarinowii O2 input

Influent Effluent S1 S6 S2 S5 S3 Water flow Sand S4 direction Gravel

FIGURE 1 - Section view of aerated integrated vertical-flow constructed wetland

layer of gravel (5-8 mm in diameter) in the bottom and the samples were freeze dried and sieved (2.1 mm) to calculate 55cm (in the down-flow chamber) or 45cm (in the up-flow the dry weight. chamber) upper layer of sand (<4 mm in diameter). The down-flow and up-flow chambers were planted with Canna 2.3. Enzyme activity assay generalis and Acorus tartarinowii, respectively (Fig.1). Dehydrogenase activity was determined by reduction One plot was equipped with an artificial aeration sys- of triphenyltetrazolium chloride (TTC) to triphenylforma- tem, which consisted of an air compressor, a PVC air dis- zone (TF). Briefly, 5 g soil was incubated in 5 ml 5 gl-1 tribution header (3 cm in diameter), and two parallel perfo- TTC and 2 ml 0.1 M glucose at 37 for 12 h. The reactions rated tubes (1 m in length) with pores (0.5 cm in diameter). were terminated by 0.25 ml 98% H2SO4 and the products The tubing was placed at the bottom of down-flow chamber were extracted 30 min with 5 ml toluene on a shaker. After (about 60 cm in depth) with an average airflow of 0.25 m3/h. centrifugation, the TF dissolved in toluene was assayed at Air was delivered to the wetland beds for about 8 hours per 492 nm. The activity was measured following a modified day (from 9 a.m. to 5 p.m.) to enhance oxygen-dependent method [16]. microbial processes and also allow anaerobic processes to Urease activity was measured using urea, firstly, 1g occur by turning off aeration during night-time. Another soil was put into a 50mL triangular flask, then 4 mL acetic plot without aeration was used as the control. The influent acid buffer solution (pH5.6), 0.3 mL methylbenzene and was domestic sewage collected from a small neighborhood 1 mL p-nitrobenzol sodium phosphate (0.115M) were and hydraulic loading was 200 L/d. The main water quality added. After shaking to uniform, the sample was cultured of the influent were (in mg/L) 105±47 TSS, 208±50 CODCr, at 37 for 1h. When culture was completed, 1 mL calcium + - 47±22 BOD5, 39±14 TN, 9.7±2.1 NH4 -N, 0.8±0.3 NO3 -N chloride (0.5M) and 4 mL sodium hydroxide (0.5M) were and 1.9±0.4 TP. added into sample liquid. Finally, the ammonium released from soil was assayed colorimetically at 460 nm [17]. 2.2. Sampling Phosphatase activity was measured following a modi- Water samples were collected from the influents and fied method [18]. For the phosphatase assay, 1 g soil (fresh the effluents every week. The total suspended solid (TSS), weight) was mixed with 4 ml acetate buffer (pH 5.8), 0.25 ml chemical oxygen demand (CODCr), five day biochemical toluene and 0.25 ml 0.115 M p-nitropheol phosphate (sub- oxygen demand (BOD5), total nitrogen (TN), ammonium + - strate). After the pretreatment, the soil from assay was gen- (NH4 -N), nitrate (NO3 -N), and total phosphorus (TP) tly vortexed, and placed in an incubator at 37 for 1 h. To were measured according to standard methods [15]. terminate the phosphatase reaction, 1 ml 0.5 M CaCl2 and The redox potential (Eh) of wetland substrate was 4 ml 0.5 M NaOH were added, and then the reactants were measured in situ using the HS-29 portable acidity meter filtrated. The released p-nitropheol in the sample was with the platinum electrode and potassium chloride (KCl) measured at 400 nm using UNICOE UV-2000 spectropho- reference system twice per hour. The electrode was pre-lay tometer. at different depths (10, 25 and 45 cm) in both downflow and up-flow chambers, labeled as S1, S2, S3, S4, S5, and 2.4. PLFAs analysis S6 along the direction of water flow (Fig. 1). Microbial community structure was characterized us- Furthermore, substrate samples were collected at dif- ing PLFA analysis of samples according to the procedure ferent depths (10, 25 and 45 cm) both at down-flow and up- [19]. Phospholipids were extracted from each subsample flow chamber, which labeled with S1, S2, S3, S4, S5 and and analyzed using GC/MS. In brief, 5 g of fresh substrate S6 along the water flow direction (Fig.1). The substrate was extracted with a chloroform: methanol: pH 7.4 phos-

phate buffer (1:2:0.8, V/V/V), and then the total lipids ex- 3. RESULTS tracted were further fractionated into neutral lipids, glycol- lipids and phosphor-lipids on prepacked silicic acid col- 3.1. Redox potential umns (SPE-Si, 500 mg/6 mL, Agilent). The polar lipids In the control, redox potential (Eh) in the surface layers were transesterified with mild alkali to form fatty acid me- of two chambers (represented by S1 and S6) remained oxic thyl esters (FAMEs). The FAMEs were separated, identi- (Eh>+300mV) around the year, while the bottom layers fied and quantified by a Hewlett-Packard 6890N GC/5873I (S3 and S4) and middle layer of up-flow chamber (S5) MS with a nonpolar capillary column (HP-5 MS, 30m× were anoxic (from -100mV to +300 mV). Meanwhile, arti- 0.25mm× 0.25μm). The MIDI peak identification software ficial aeration increased Eh measurements of all sites, es- (MIDI Inc., Newark, DE, USA) was used to identify indi- pecially the down-flow chambers (S1, S2 and S3) main- vidual fatty acids. tained oxic in the aerated IVCW during four seasons.

2.5. Data analysis In order to reveal the effect of artificial aeration on the PLFA analysis was performed in triplicates and the area of redox zones, the fitting equations between redox data were also analyzed by analysis of variance (ANOVA) potential (Eh) and depth (h) in different seasons were esti- and principal component analysis (PCA). All analyses mated, and the depth of redox zones were also calculated were considered significant at the 0.05 level. (Table 1). Artificial aeration enlarged the depth of aerobic

700 700 Summer Autumn 600 600 500 500 Non-aeration 400 400 Non-aeration Aeration 300 300 Aeration

200 200 Eh (mV) Eh (mV) 100 100 0 0 -100 -100 -200 -200 S1 S2 S3 S4 S5 S6 S1 S2 S3 S4 S5 S6

700 700 Spring 600 Winter 600 500 500 400 Non-aeration 400 Non-aeration 300 300 Aeration Aeration 200 200 Eh (mV) Eh (mV) 100 100 0 0 -100 -100 -200 -200 S1 S2 S3 S4 S5 S6 S1 S2 S3 S4 S5 S6 FIGURE 2 - Redox potential at the sampling sites in the aerated and non-aerated IVCWs

TABLE 1 - The depth of redox zones in aerated and non-aerated IVCWs

Depth (cm) 2 Fitting equation R 1st oxic zone Anoxic zone 2nd oxic zone (>300mV) (-100~300mV) (>300mV) Aerated IVCW Summer Eh=0.002h3-0.265h2+3.487h+561.08 0.92 53 51 16 Autumn Eh=0.002h3-0.283h2+3.4h+549.77 0.99 49 54 17 Winter Eh=0.0023h3-0.319h2+5.549h+478.67 0.94 44 66 10 Spring Eh=0.002h3-0.32h2+5.196h+495.27 0.95 45 60 15 Non-aerated IVCW Summer Eh=0.187h2-24.557h+747.73 0.97 22 87 11 Autumn Eh=0.151h2-20.439h+696.53 0.92 24 88 8 Winter Eh=0.203h2-26.113h+676.97 0.99 17 103 0 Spring Eh=0.152h2-20.096h+625.12 0.94 19 95 6

zone (1st and 2nd) from 33cm to 69cm in the summer and ated IVCW, and decreased with increasing depth of sub- from 17cm to 54cm in the winter, which could provide suit- strate. Dehydrogenase activity of surface layer in the able habitat for aerobic microorganisms. Moreover, the downflow chamber was the highest, which was approxi- 2nd oxic zone was not present in the winter unless artificial mately 8 times as of that at the bottom layer. The spatial aeration was added to the aerated IVCW. This aerobic zone distribution feature of dehydrogenase activity in aerated may facilitate the further decomposition of pollutants and IVCW agreed with that in non-aerated system. However, increase the growth of aerobic microorganisms. at the same substrate layer, dehydrogenase activity in aerated IVCW was higher than that in non-aerated system 3.2. Enzyme activity (p<0.05). This may be because that aeration provides suf- Substrate enzyme activities reflect the dynamic ficient oxygen source for dehydrogenase enzyme to pro- strength of various biochemical processes in constructed mote oxidation process of organic mass. wetland, which is very meaningful for carbon and nitrogen Urease activity can be used as an important index to transformation. Dehydrogenase, urease and phosphatase express removal rate of total nitrogen in constructed wet- activities were investigated to evaluate the change of mi- lands [18]. Fig. 3 showed that spatial variation of urease croorganism activity in IVCW under artificial aeration, as activity was similar to that of dehydrogenase activity. Ure- shown in Fig. 3. ase activity of surface layer in the downflow chamber was Dehydrogenase activity is responsible for conversion also highest, which was approximately 16 times as that at of soil organics, which can promote dehydrogenation reac- the bottom layer. Artificial aeration enhanced nitrification tions of carbohydrate and organic acids, etc. and play a role process and promoted substrate urease activity in the of intermediation transferring [11]. Fig. 3 showed that de- downflow chamber (p<0.05); while in the upflow chamber, hydrogenase activity in the downflow chamber was signif- nitrification process was weakened due to oxygen source icantly higher than that in the upflow chamber in non-aer- decrease.

)

-1 30 h -1 27 a Non-aeration Aeration

µgg 24 ( 21 18 15 b a 12 a 9 a a b 6 b b a 3 b b 0

Dehydrogenase activity Dehydrogenase S1 S2 S3 S4 S5 S6

350

) a Non-aeration -1

h 300 Aeration -1 250

µgg b ( 200 150 a a 100 a a 50 b Urease activity b a aaa 0 S1 S2 S3 S4 S5 S6 )

-1 100 h

-1 a a Non-aeration Aeration

µgg 80 (

60 a a 40 a a a a a a a a 20

Phosphatase activity activity Phosphatase 0 S1 S2 S3 S4 S5 S6 Sampling sites FIGURE 3 - Enzyme activities of different sampling sites in non-aerated and aerated IVCWs

TABLE 2 - Relative content (%) of PLFAs in non-aerated and aerated IVCWs

Non-aerated IVCW Aerated IVCW PLFA group S1 S2 S3 S4 S5 S6 S1 S2 S3 S4 S5 S6 Kinds of PLFAs 20 17 17 17 17 18 16 15 15 15 15 16 57.48 56.03 68.58 75.6 69.07 63.47 43.52 47.3 64.06 69.43 61.5 54.98 SATFA (13.86) (10.57) (6.98) (5.73) (7.22) (9.95) (14.61) (15.13) (12.58) (8.39) (8.74) (10.52) MUFA 23.32 24.34 14.80 9.83 15.73 19.85 29.47 28.83 17.99 14.07 14.45 26.82 (Gram-) (6.35) (8.07) (2.96) (1.88) (3.32) (3.76) (7.12) (8.46) (5.77) (4.29) (2.86) (3.90) 2.97 1.91 1.42 1.39 1.45 2.88 8.05 2.47 1.67 1.50 2.26 6.03 PUFA (0.84) (0.47) (0.21) (0.25) (0.19) (0.67) (1.79) (0.83) (0.41) (0.30) (0.36) (0.96) BRAFA 12.60 13.33 10.88 10.39 9.92 10.22 11.81 13.38 11.92 12.55 13.44 8.09 (Gram+) (2.88) (3.04) (1.79) (1.48) (0.85) (1.69) (3.45) (3.17) (3.86) (2.18) (2.64) (1.71) 3.63 4.39 4.31 2.78 3.83 3.57 7.15 8.02 4.35 2.44 8.35 4.08 CYCFA (Anaerobic) (0.78) (0.62) (0.60) (0.43) (0.81) (0.87) (1.56) (1.89) (1.07) (0.85) (2.11) (1.02) (i15:0+a15:0)/16:0 0.23 0.24 0.19 0.18 0.16 0.21 0.27 0.30 0.22 0.20 0.22 0.12 18:1ω9c/18:1ω9t 1.18 0.82 0.65 0.70 0.80 0.67 0.84 0.72 0.63 0.76 0.50 1.11 MUFA/BRAFA 1.85 1.83 1.36 0.95 1.58 1.94 2.50 2.15 1.51 1.12 1.07 3.32 Note: SATFA, MUFA, PUFA, BRAFA and CYCFA represent saturated fatty acid, monounsaturated fatty acid, polyunsaturated fatty acid, branched fatty acid and cyclopropane fatty acid.

Fig. 3 showed that phosphatase activity tended to de- higher than that in the upflow chamber, indicating that the crease with depth, but the amplitude of change was signif- abundance ratio of bacteria in the downflow chamber is icantly smaller than that of dehydrogenase and urease ac- richer than that in the upflow chamber. Furthermore, artifi- tivity. Artificial aeration had no significant influence on the cial aeration increased the ratios of three sampling sites in phosphatase activity of all sampling sites in wetland sys- the downflow chamber, suggesting that aeration increased tem (p>0.05). the abundance and living space of bacteria. It is reported that (18:1ω9c)/(18:1ω9t) can indicate the 3.3. Composition and relative contents of PLFAs degree of nutrient deficiency or environmental stress for The kinds and relative contents of PLFAs of different microorganisms in the system [22]. Compared with non- depth samples in non-aerated and aerated IVCW were aerated IVCW, this ratio is significantly lower in the down- shown in Table 2. Twenty kinds of PLFAs between C11 and flow chamber in the aerated system, probably indicating C20 were found in non-aerated IVCW, among them, satu- that the microorganisms were not under huge anoxic stress. rated fatty acid 16:0 and 18:0, branched fatty acid i15:0, However, the ratio at surface layer in the upflow chamber monounsaturated fatty acid 18:19c, 18:19t and 16:19 was suddenly increased, which probably meant that the were dominant. However, four kinds of PLFAs, saturated stress of nutrient deficiency for the microorganisms be- fatty acid 11:0, 12:0, 13:0 and 19:0, were not determined came the main limiting factor. in aerated IVCW. Although they were obviously decreased, The ratio of monounsaturated fatty acid/branched fatty the relative contents of saturated fatty acid (43.52%- acid reflects the relative dominance of aerobic bacteria/ an- 69.43%) were still present in the greatest amounts of aer- aerobic bacteria [20]. As shown in Table 2, the ratio was ated IVCW. Additionally, the relative contents of monoun- larger and significantly increased in aerated IVCW, indi- saturated fatty acid in aerated IVCW were higher than cating that aerobic bacteria were dominant and proportion those of the control, suggesting that aerobic microorgan- of aerobic bacteria obviously increased after artificial aer- isms significantly increased. In general, the composition ation. and relative contents of PLFAs between non-aerated and aerated IVCW showed variation, indicating the differences 3.5. PCA analysis in the lipid contributing communities. As shown in Fig. 4, the contribution factor of first prin- 3.4. Ratio of biomarker PLFAs cipal composite PC1 was 56.4%, which had a comparatively big positive related coefficient with monounsatu- Characteristic PLFAs are biomarkers for different mi- rated fat acid 16:1ω9, 18:1ω9c, 18:1ω9t and saturated fat croorganism populations, and their relative contents can in- acid 13:0, 17:0, 19:0, 20:0 (all >0.7), with the biggest co- dicate their relative dominance [20]. efficient of 18:1ω9t (being 0.92). Moreover, PC1 had a It has been reported that (i15:0+a15:0)/16:0 may re- comparatively big negative coefficient with saturated fat flect the ratio of bacteria to microorganisms [21]. As seen acid 11:0, 16:0, 14:0 (all >0.8). The second principal com- from Table 2, whether in non-aerated or aerated IVCW, the position PC2 had a contribution rate of 15.2%, which had ratio in middle layer of substrate in the downflow chamber comparatively positive relationship with branched fat acid were the biggest. This may be because the highly devel- i15:0, a15:0 and cyclopropyl fat acid cy17:0. oped root system of Canna generalis provided suitable In addition, the sampling sites in the downflow and up- habitat. In addition, the ratio in the downflow chamber was flow chambers in non-aerated IVCW were mainly sepa-

rated by PC1 shaft. The sampling sites at the surface layer as DO concentration, Eh, etc [12]. This may be the reason and PC1 shaft showed stronger positive relation, while that artificial aeration can enhance dehydrogenase activity sampling sites at the bottom appeared to have stronger neg- in constructed wetlands. ative relation, indicating that aerobic microorganism was Urease and phosphatase enzyme belong to hydrolyzed dominant at the surface layer while anaerobic microorgan- enzyme, which is involved the cycling process of nutrients ism was dominant at the bottom layer. Sampling sites in such as nitrogen and phosphorous in wetlands [18]. Com- non-aerated and aerated IVCW were separated in the PCA pared to non-aerated IVCW in our experiment, urease ac- diagram, indicating that microorganism structure changed tivity was obviously promoted in the aerated IVCW, but with aeration. Sampling sites in the downflow chamber sig- phosphatase activity was not improved. Moreover, com- nificantly increased along PC1 shaft positive direction pared to the results of other researchers, Wu et al. [24] compared with that in non-aerated system, and abundance found that urease and phosphatase activities in non-aerated of related monounsaturated fat acid 16:1ω9, 18:1ω9c, IVCW on January were 50 μg·g-1·h-1 and 80 μg·g-1·h-1, re- 18:1ω9t increased as well, suggesting that aeration pro- spectively, but our measurement of urease and phosphatase moted growth of aerobic microorganism. Furthermore, S4 activities at the surface layer of aerated IVCW during win- sampling site at the bottom of aerated and non-aerated sys- ter were 73.1-287.5μg·g-1·h-1 and 30.2-85.8μg·g-1·h-1, re- tem have little difference from one another along the neg- spectively. Therefore, it could be concluded that aeration ative direction of PC1 shaft, which indicated that intermit- enhanced urease activity while little effect on phosphatase tent aeration had little effect on the number of anaerobic activity. microorganism at the bottom of system. Moreover, as shown in Fig.3, spatial variation of the three substrate enzyme showed that enzyme activities in S5 cy17:0 the downflow chamber (represented by S1, S2 and S3) were significantly higher than that in the upflow chamber a15:0 (S4, S5 and S6), and enzyme activities at the surface layer i15:0 S2 S2 (S1 and S6) were obviously higher than that at the bottom S3 layer (S3 and S4), which were consistent with the detected 15:0 S3 i16:0 14:0 cy19:0 16:1ω9 Eh measurements. This observation is consistent with that 12:0 i17:0 previously reported [25]. 16:0 11:0 S1 13:0 18:1ω9t S5 19:0 S4 17:0 18:1ω9c 4.2. Relationship between microbial community structure and artificial aeration PC2(15.2%) 18:0 20:0 S1 In constructed wetlands, degradation of pollutants 18:2ω9 S6 mainly rely on substrate microorganism, therefore, microbial community structure (e.g. quantity, activity and dominant groups, etc.) directly influences treatment perfor- S6 S4 mance of wetlands. In non-aerated IVCW system, an aero-

-1.0 1.0 bic environment was formed by air diffusion and plant-me- -1.0PC1(56.4%) 1.0 diated aeration at the surface layer, while a large anaerobic zone with a depth of 87-103 cm appeared after depletion of FIGURE 4 - PCA plot of PLFAs among different sampling sites dissolved oxygen (Table 1). The Eh measurements demon- Note: Box-shaped represents sampling sites, solid (■) and hollow (□) strated that artificial aeration could enlarge the area of aer- mark represent sampling sites in non-aerated and aerated IVCW, respectively. obic zones which was twice bigger than that of the control. PLFAs analysis showed that absolute abundance of PLFAs was significantly enhanced in aerated IVCW, suggesting 4. DISCUSSION that microbial biomass increased. Total unsaturation degree of PLFAs was also obviously improved, indicating 4.1. Relationship between enzyme activities and artificial aer- that proportion of aerobic microorganism increased. ation In summary, the intermittent artificial aeration may Enzyme is an important index for evaluating activity have several advantages: (1) to directly strengthen pollu- and function of microorganism community in soil [11]. De- tants removal efficiency by suppling oxygen for aerobic hydrogenase can catalyze oxidation-reduction reaction that decomposition of organic matter and nitrification; (2) to is needed for organic degradation and is usually used as an enlarge aerobic zone to provide more suitable environment index for evaluating the level of microorganism activity in for aerobic microorganism to develop; (3) to increase quan- wetlands [12]. It was verified that the quantity of bacteria tity of microorganism, especially the proportion of aerobic agreed with dehydrogenase activity with seasonal dy- microorganism, which could utilize the additional oxygen namic, indicating that dehydrogenase activity represent to- to degrade more organic matter and oxidize more ammonia tal biomass of microorganism [23]. In addition, dehydro- to nitrate (nitrification), particularly in the winter; (4) to in- genase activity could be influenced by many factors such crease the activities of enzyme and microorganism which

could enhance pollutants removal efficiency and process; [3] Nivala, J., Wallace, S., Headley, T., Kassa, K., Brix, H., Van- (5) to create the alternate aerobic/anoxic redox conditions, Afferden, M. and Müller, R. (2013). Oxygen transfer and consumption in subsurface flow treatment wetlands. Ecological allowing the occurrence of the coupled nitrification/deni- Engineering, 61, 544-554. trification process. As a result, the addition of an intermit- [4] Van-Oostrom, A.J. and Russell, J.M. (1994). Denitrification in tent artificial aeration could improve the substrate redox constructed wastewater wetlands receiving high concentration environment, enzyme activities and microbial community of nitrate. Water Science and Technology 29, 7-14. structure, which in turn could promote the rapid aerobic [5] Pandey, M.K., Jenssen, P.D., Krogstad, T. and Jonasson, Sven. decomposition of organic matter and the coupled nitrifica- (2013). Comparison of vertical and horizontal flow planted tion/denitrification process in the aerated IVCW. and unplanted subsurface flow wetlands treating municipal wastewater. Water Science and Technology 1, 117-123. [6] Du, X.L., Xu, Z.X., Li, J.Q. (2014). Enhanced nitrogen re- 5. CONCLUSIONS moval in a non-planted modified extended tidal flow constructed wetland. Fresenius Environmental Bulletin 1, 285- Our yearly monitoring have demonstrated that artifi- 289. cial aeration significantly increased Eh of all sites, and the [7] Admon, S., Tarre, S., Sabbah, I., Lahav, O., Beliavski, M. and depth of aerobic zone was enlarged twice in the summer Green, M. (2005). Treatment of presettled municipal wastewater using a passively aerated vertical bed. Environ- and three times in the winter. Especially, the 2nd oxic zone mental Engineering Science 6, 707-715. was not present in the winter unless artificial aeration was added to the aerated IVCW. Artificial aeration signifi- [8] Du, X.L., Xu, Z.X., Li, J.Q. and Fu, K.M. (2013). Using passive atmospheric oxygenation to increase nitrification poten- cantly enhanced dehydrogenase and urease activities, tial in a non-planted vertical flow constructed bed system. while phosphatase activity was not improved. PLFAs anal- Polish Journal of Environmental Studies 6, 1685-1690. ysis revealed that the relative contents of monounsaturated [9] Tao, M., He, F., Xu, D., Li, M. and Wu Z.B. (2010). How arti- PLFAs were obviously increased by artificial aeration, sug- ficial aeration improved the sewage treatment of an integrated gesting that the proportion of aerobic microorganisms sig- vertical-flow constructed wetland. Polish Journal of Environ- nificantly increased. More importantly, the artificial aera- mental Studies 1, 181-189. tion of 8 hours per day had little effect on the relative con- [10] Butterworth, E., Dotro, G., Jones, M., Richards, A., Onunkwo, tents of cyclopropane PLFAs at the bottom layer, indicat- P., Narroway, Y. and Jefferson, B. (2013). Effect of artificial ing that anaerobic denitrification and the proportion of an- aeration on tertiary nitrification in a full-scale subsurface horizontal flow constructed wetland. Ecological Engineering 54, aerobic microorganisms were not affected. These results 236-244. demonstrate that artificial aeration improved oxygen sup- [11] Kadlec, R.H. and Wallace, S.D. (2008). Treatment wetlands ply and then enhanced enzyme activities, so pollutants re- (2nd ed.). CRC Press, New York, 267-280. moval efficiency were increased. [12] Zelles L. (1999). Fatty acid patterns of phospholipids and lipo- polysaccharides in the characterisation of microbial communities in soil: a review. Biology and Fertility of Soils 2, 111-129.

ACKNOWLEDGEMENTS [13] Shi, Y. C., Lalande, R., Hamel, C., Ziadi, N., Gagnon, B. and Hu, Z. Y. (2013). Seasonal variation of microbial biomass, activity, and community structure in soil under different tillage The research is supported by grants from National Natu- and phosphorus management practices. Biology and Fertility ral Science Foundation (51178452, 41303090), National Sci- of Soils 7, 803-818. ence and Technology Support Program (2012BAJ21B03-04), [14] Sushchik, N.N., Gladyshev, M.I., Kalachova, G.S., Kravchuk, Project of Science and Technology Department of Jiangsu E.S., Dubovskaya, O.P. and Ivanova, E.A. (2003). Particulate Province (BY2011193), Open Project Program of Hubei fatty acids in two small siberian reservoirs dominated by different groups of phytoplankton. Freshwater Biology 3, 394-403. Key Laboratory of Mine Environmental Pollution Control & Remediation (2012108), Natural Science Foundation of [15] China State Environmental Protection Administration. (2002). Hubei Province (2011CDB405) and Research Project of Monitoring and determination methods for water and wastewater (4th ed.). China Environmental Press, Beijing. (in Institute of Hydrobiology (Y25E021701). Chinese)

[16] Institute of Soil Science, Chinese Academy of Sciences. The authors have declared no conflict of interest. (1985). Methods on soil microorganisms study. Science Press, Beijing. (in Chinese)

[17] Kandeler, E. and Gerber, H. (1998). Short-term assay of soil REFERENCES urease activity using colorimetric determination of ammonium. Biology and Fertility of Soils 6, 68-72. [1] Zhang, T., Xu, D., He, F., Zhang Y.Y. and Wu Z.B. (2012). Ap- plication of constructed wetland for water pollution control in [18] Xu, G.H. and Zheng H.Y. Handbook of analysis of soil micro- China during 1990-2010. Ecological Engineering 47, 189-197. organism. (1986). Agriculture Press, Beijing. (in Chinese) [2] Akratos, C.S. and Tsihrintzis, V.A. (2007). Effect of tempera- [19] Bossio, D.A. and Scow, K.M. (1998). Impacts of carbon and ture, HRT, vegetation and porous media on removal efficiency flooding on soil microbial communities: phospholipid fatty of pilot-scale horizontal subsurface flow constructed wetlands. acid profiles and substrate utilization patterns. Microbial Ecol- Ecological Engineering 2, 173-191. ogy 3, 265-278.

[20] Rajendran, N., Matsuda, O., Rajendran, R. and Urushigawa, Y. (1997) Comparative description of microbial community structure in surface sediments of eutrophic bays. Marine Pol- lutution Bulletin 1, 26-33.

[21] Rajendran, N., Matsuda, O., Imamura, N. and Urushigawa, Y. (1992). Variation in microbial biomass and community structure in sediments of eutrophic bays as determined by phospholipid ester-linked fatty acids. Applied and Environmental Mi- crobiology 2, 562-571. [22] Vestal, J. R. and White, D. C. (1989). Lipid analysis in microbial ecology. Bioscience 8, 535-541. [23] Liang, W., Wu, Z.B., Cheng, S.P., Zhou, Q.H. and Hu, H.Y. (2003). Roles of substrate microorganisms and urease activities in wastewater purification in a constructed wetland system. Ecological Engineering 21, 191-195. [24] Wu, S.Q., Chang, J.J., Dai, Y.R., Wu, Z.B. and Liang, W. (2013). Treatment performance and microorganism community structure of integrated vertical-flow constructed wetland plots for domestic wastewater. Environmental Science and Pollution Research 6, 3789-3798.

[25] Huang, J., Yan, L., Cai, W.S., Zhong, Q.S., Wang, S.H. (2013). Distribution of DO, substrate enzyme activities and operating performance of constructed wetlands for wastewater treatment. Fresenius Environmental Bulletin 10, 2816-2822.

Received: May 23, 2014 Revised: June 18, 2014 Accepted: July 04, 2014

CORRESPONDING AUTHOR

Qi Shuo Wang

Huai’an Research Center

Institute of Hydrobiology

Chinese Academy of Sciences

223006 Huai’an

P.R. CHINA

E-mail: [email protected]

FEB/ Vol 24/ No 1b/ 2015 – pages 371 - 378

BIOLEACHING OF ARSENIC-CONTAINING GOLD ORE INFLUENCED BY CYSTEINE

Hongai Zheng1,2, Chen Zhang2,Yang Wu2, Penghui Shi2, Zhonghong Wang2, Xuejuan Wang2, Bingyue Li2, Jianshe Liu1,∗ and Xuehui Xie1,∗

1College of Environmental Science and Engineering, Donghua University, Shanghai 201620, China 2College of Environmental and Chemical Engineering, Shanghai University of Electric Power, Shanghai 200090, China

ABSTRACT creased [8, 9]. However, bacterial oxidation rate is slow, and enhancement of bacteria activity is still a major problem to Adding a small amount of cysteine to bioleaching solu- be solved in bioleaching [10]. Therefore, strategies to in- tion in contact with gold ores, largely enhanced the leaching crease bio-oxidation rate at gold ores surface have been rate of Fe, while cysteine inhibited the extraction of As. X- widely investigated. Some studies have shown that the re- ray diffraction (XRD), Scanning electron microscope and covery of gold from sulphide ores by bacterial leaching tech- Energy-dispersive X-ray spectroscopy (SEM/EDS) obser- niques is economically improved when carbohydrates, pro- vations of residues without cysteine showed bio-leaching teins and other substances of biological origin are added to products were mainly jarosite and sulfur. Yet, the peak den- the tank reactors [11]. sities of sulfur disappeared in the XRD patterns of residues Recently, cysteine (Cys), an important sulfureous amino with cysteine added. Furthermore, Fourier transform infra- acid, aroused great attention because of its capacity to ac- red spectroscopy (FTIR) led to the conclusion that more bac- celerate bioleaching. Studies have shown that cysteine can teria had adsorbed on the bioleached residues surface in me- accelerate pyrite [11] and Ni-Cu sulfide [12] bioleaching. dia with cysteine than without cysteine. Combining the re- Although cysteine has been proved highly effective to ac- sults of the bioleaching experiment and the analysis of celerate pyrite leaching, the mechanism involved was not SEM/EDS, FTIR and XRD leads us to conclude that cyste- well understood yet. Rojas-Chapana and Tributsch [11] ine addition just facilitates the electron transfer but does not speculated that the active group sulfhydryl (-SH) of cyste- change the gold ore oxidation mechanism. ine participated in a binding process with pyrite. Whereas,

Cysteine contains–HS group, and –HS group is a substrate KEYWORDS: Bioleaching; Gold ores; Cysteine; X-ray diffrac- or a reduced sulfur source for the organisms. Then, cyste- tion( XRD ); Scanning electron microscope and Energy-dispersive ine provides an additional reduced sulfur source for the X-ray spectroscopy(SEM/EDS); Fourier transform infrared spectroscopy(FTIR) bacteria which competes with the Fe2S in mineral. In this case, Rojas-Chapana and Tributsch [11] have shown that

the thiol species interacting with the sulfide play a domi- 1. INTRODUCTION nant role in the disruption of the interface and the transport of the sulfur species to the bacterial. However, evidence Global gold ore deposits are fast being depleted due to available is still little to support these hypotheses. Further- rapid industrialization and demand [1]. Low and lean grade more the reaction mechanism and intermediate have not gold ores become the main gold resource in the future. Thus, been conformed when cysteine interacts to minerals. it is increasingly necessary to exploit ores of low grade to Therefore, it is necessary and significant to investigate meet the demand, especially arsenic-rich gold ores. The re- how cysteine interacts with mineral. In this paper, Fourier covery of metals from these gold ores using conventional transform infrared spectroscopy (FTIR), X-ray diffraction pyrometallurgical technologies is not only the diseconomy (XRD), scanning electron microscope and Energy-disper- due to capital investments, but also the environmental cost sive X-ray spectroscopy (SEM-EDS) were employed to pro- due to pollution [2-4]. Bioleaching operations with environ- vide detailed insight into the interaction mechanism. mental benefits, low operating and capital costs [5] have been successfully applied in the recovery of precious metals from sulfide ores in many countries [6, 7]. When Fe, S and As of 2. MATERIALS AND METHODS the gold ores are effectively oxidized by bio-oxi-dation in pretreatment, gold will be naked and its recovery will be in- 2.1 Materials

Bacteria used in this experiment consisted of Leptospi- * Corresponding author rillum ferriphilum and Sulfobacillus thermotolerans. In the

TABLE 1 - Elemental composition of gold ore samples

Element Fe As S Al Pb Sb Cu C Au Content (%) 20.2 3.97 24.42 0.465 0.054 0.147 0.046 1.15 76.22g/t

lab, the bacteria was cultured on a medium containing constants of samples blended with a few high-purity Si pow- (NH4)2SO4 3.0 g/L, MgSO4·7H2O 0.5 g/L, K2HPO4 0.05 g/L, der, which was applied to assist calibration of measurement KCl 0.1 g/L, Ca(NO3)2 0.01 g/L, FeSO4.7H2O 43.2g/L. The error in 2θdiffraction peak, were accurately determined. culture was incubated in 500mL Erlenmeyer flask containing 200mL medium and 10% (v/v) inoculum, on a rotary shaker at 150 rpm at a constant temperature of 40°C. The 3. RESULTS AND DISCUSSION initial pH was adjusted to 2.0 using 0.5M H2SO4. The culture had been sub-cultured through several transfers in me- 3.1 Leaching study dium using gold ore instead of ferrous sulfate to adapt the Bioleaching of gold ores with moderate thermophiles bacteria to experiment condition. The gold ore sample was in the presence of different amount of cysteine at 40 °C was sampled from Axi gold ore by Changchun Gold Research carried out for 24 days. The effect of cysteine on the bio- Institute. Its chemical composition analyzed by X-ray dif- leaching of arsenic and iron, presented in Fig. 1 and Fig. 2, fraction (XRD) and ICP is presented in Fig.7 (a) and Table 1. shows the leaching of arsenic was totally different from XRD analysis showed that gold ore sample was mainly constituted of pyrite (FeS2), arsenopyrite (FeAsS) and 100 quartz (SiO2). 0M 10-4M Cysteine is a biochemical reagent. Sodium hydroxide -3 is of laboratory reagent grade. Double-distilled water was 80 10 M 5×10-3M used throughout the experiments. -2 10 M 60 2.2 Experimental Methods A 9K medium with the following composition was 40 used: (NH4)2SO4 3.0 g/L, MgSO4·7H2O 0.5 g/L, K2HPO4

0.05 g/L, KCl 0.1 g/L, Ca(NO3)2 0.01 g/L. All these rea- Fe Recovery(%) gents were analytic grade. Gold ore particle diameter was 20 less than 74 µm and the pulp density of gold ore was 8 g/ 100 mL. The moderate thermophiles consisting of Lepto- 0 spirillum ferriphilum and Sulfobacillus thermotolerans were 0 5 10 15 20 25 enriched at 40 °C with an initial pH of 2.0 in 250 mL flasks, Time(days) shaken at speed of 150 rpm. During leaching, the total ar- FIGURE 1 - The effect of cysteine concentration on the leaching of senic and iron concentration were measured by ultraviolet iron (Incubation concentration:107cells/mL; Temperature:40 °C; spectrophotometry (Hitachi U-2900) and atomic absorption Amount of gold ore:8 g/mL) spectrophotometry (Z-2000) every 4 days, respectively.

2.2.1 FTIR measurement 100

After bioleaching, a series of procedures like agitation, 0M 10-4M blending and rinsing with diluted acid were performed. 80 -3 10 M Then the mixture was filtered and natural aired for 48 h. 5×10-3M FTIR spectra of the samples in the presence and absence of 10-2M cysteine were obtained on a NEXUS670 FTIR spectrome- 60 ter in the range of 500-4000 cm-1.

40 2.2.2 SEM-EDS measurement As Recovery(%) The raw and treated samples in the presence and ab- 20 sence of cysteine were sent to the Research Center for Analysis and Measurement of Donghua University for SEM (QUANTA) and EDS (JSM-5600LV) analysis. 0 0 5 10 15 20 25 2.2.3 XRD measurement Time(days) FIGURE 2 - The effect of cysteine concentration on the leaching of The raw and treated samples were measured by Rigaku Arsenic(Incubation concentration:107cells/mL; Temperature:40 °C; 。 diffractometer in the 2θrange of 10 to 90。. Then the lattice Amount of gold ore:8g/mL)

2.4 that of iron. Certain amount of cysteine can promote iron to dissolve out; the greatest iron recovery efficiency oc- 2.2 curred when 10-3M of cysteine was added. With the in- 2.0 0M crease of cysteine added, the extraction of iron decreased. 10-4M When 5×10-3M cysteine was added, the bioleaching of Fe 1.8 10-3M was slightly inhibited. However, as shown in Fig. 2, more -3 1.6 5×10 M cysteine containing resulted in a less arsenic extraction, in- 10-2M 1.4 dicating that cysteine could inhibit arsenic leaching.

pH Value 1.2 A notable conclusion from Fig.3 was that different concentration of cysteine affected the pH value of leaching 1.0 solution. When 10-3M cysteine was added, the pH value 0.8 reached the lowest point on the twentieth day. An initial pH

0.6 increase was observed in the leaching systems, resulting 0 5 10 15 20 25 from ferrous oxidation and the acid dissolution (Eq.5). Af- Time(days) ter 4-24 days, the pH started to decrease relating to the for-

FIGURE 3 - Change of pH during gold ore bioleaching at cysteine mation of jarosite (Eq.9) and bacteria oxidation of sul- concentration (Incubation concentration:107cells/mL; Tempera- fur(Eq.6) [13]. It was suggested that the following reac- ture:40℃;Amount of gold ore:8g/100mL) tions took place in bio leaching system.

4FeAsS+13O2+6H2O 4H3AsO4+4FeSO4 (1)

2FeS2+7O2+2H2O 2FeSO4+ 2H2SO4 (2)

2FeAsS+Fe2(SO4)3+6O2+4H2O 2H3AsO4+4FeSO4+H2SO4 (3)

FeS2+7Fe2(SO4)3+8H2O 15FeSO4+ 8H2SO4 (4)

4FeSO4+ 2H2SO4+O2 2Fe2(SO4)3+2H2O (5)

2S+3O2+2H2O 2H2SO4 (6) 4FeAsS+11O2+6H2O 4H3AsO3+4FeSO4 (7) H3AsO3+Fe2(SO4)3+H2O H3AsO4+2FeSO4+ H2SO4 (8) 3+ 2- + + 3Fe +2HSO4 +K +6H2O KFe3 (SO4)2(OH)6 +8H (9)

By adding cysteine, organisms would oxidize the reduced sulfur in cysteine to produce H2SO4, reducing the pH. The acidic pH value of the leaching solution (below 2.0) favored the binding of reduced cysteine to newly exposed pyritic sulfhydryl group which in turn maintained the ongoing dissolution of gold ore.

3.2 Results of SEM and EDS To scrutinize the difference between the residues in presence and in absence of cystein, the morphology and mineralogy of bioleaching residues were examined using SEM photographs (Fig.4) and EDS microanalyses (Fig.5). SEM analysis of gold ore residues as shown in Fig.4(c) revealed that the surface of the mineral particles appeared porous structure and etched pits was also evident. Compar- atively, without cysteine (Fig.4 (b)), deep and oriented cor- rosion pores were not so evident on the surface of mineral particles.

By SEM/EDS analyses (Fig.5), a significant amount of jarosite and sulfur was found on the bioleaching residue. It (a) is likely that both jarosite and polysulfides passivate jointly

(b) (c)

FIGURE 4 - SEM images of gold ore particles.(a) Untreated gold ore, (b) After interaction with mixed moderate thermophiles, (c) After interaction with moderate thermophiles and cysteine

FIGURE 5 - EDS analysis of the solid residue (a) After interaction with mixed moderate thermophiles, (b) After interaction with moderate thermophiles and cysteine

the mineral surface [14], and some trace elements, such us N, could not be detected.

3.3 Results of FTIR Fig. 6 presents the results obtained from FTIR. They show that the predominant mineral phase is jarosite, demonstrated by the presence of the bands (anti-symmetric stretching triply degenerate vibration 1163cm-1, 1086 cm-1 and 1038 cm-1[15,16]. In addition, it could be observed typical bands of quartz (798 cm-1,778 cm-1and 694 cm-1).

100

60 ● (1)Bacteria+Cysteine (2)Bacteria 40 (b) ●Quartz ★Jarosite Reflectance(%) ★ ◆Sulfur 5000 20

◆ 0 ★ 4000 3000 2000 1000 0 Intensity(CPS) ●◆ Wavenumber(cm-1) ★● ● ★ ● FIGURE 6 - FTIR spectra of solid residue in presence and in absence ◆ ◆ ● ● ★ ◆★ ★★ ★● ● of cysteine ★ ★ ● ★ ● ● ● ★● ● ● ● ● 0 As the FTIR spectrum shows, there is an obvious 10 20 30 40 50 60 70 80 90 change within the range of 1650–1000 cm-1in the spectrum Two-Theta(deg) of gold ore. The intensity of some absorbance peaks is greater with cysteine than it is without cysteine, and other (c) ● new peaks appear after adding cysteine. Firstly, the three 12000 new bands around 1485 cm-1, 1622cm-1 and 1297cm-1 indi- - 10000 cate the strong COO and C=O, stretching group [17-19]. In addition, two new bands appeared near 847cm-1, 874 cm-1. ●Quartz ★Jarosite In addition, the sharp peak around 1650cm-1 is due to an 8000 amide group, and band near 1542 cm−1 has been interpreted 6000 as N-H bending of the amide group –CONH [20, 21]. Most of these groups are present in protein molecule in the bacte- 4000 Intensity(CPS) ria, then the FTIR spectra of treated mineral show the ab- ★★ ● ★ sorbance bands characteristic of the cells’ surface chemical 2000 composition. Hence, it can be inferred that more bacteria has ●● ★ ★ ● ★ ● ★ ●● ● ★★★● ●●★ ★ ★★ ●●●●● been adsorbed on the bioleached residues surface in media 0 ● with cysteine than without cysteine. 10 20 30 40 50 60 70 80 90 Two-Theta(deg) A previous study [22] revealed that bridge complexes formed by adsorbing –CN–, –O– or –CO–groups to inter- FIGURE 7 - XRD analysis of the gold ore samples, (a) the original gold ore sample, (b) after interaction with mixed moderate thermo- facial iron atoms in mineral, could greatly accelerate the philes, (c) after interaction with moderate thermophiles and cysteine transfer of photo generated charge carriers between mineral and soluble iron. XRD patterns of bioleached gold ores indicate that the chemical compositions of original gold ores were pyrite 3.4 Results of XRD (FeS2) and arsenopyrite (FeAsS) (Fig.7(a)), and bioleach- In order to acquire further information on the influence ing products were mainly jarosite and sulfur (Fig.7(b)). The of cysteine during the bio-oxidation process, the original ore peak densities of sulfur disappeared as shown in samples and bioleaching residues were analyzed by XRD (Fig. 7). with cysteine added (Fig.7(c)), which was due to the ability

2- of bacteria to oxidize elemental sulfur to SO4 [23, 24] and The detailed mechanism of how cysteine interacts with unevenly distribution of the sulfur in the samples. When different sulphide ore needs further investigation. In this comparing to residues untreated by cysteine, the spectrum paper, we have found some moderate thermophiles showed intensity of diffraction peaks changed scarcely, indicating synergistic response to cysteine in Fe extraction. Further that addition of cysteine just facilitated the electron transfer research can be carried out in heavy metal [28, 29], such as but did not change the gold ore bio-oxidation mechanism. Cu leaching system added thio subspecies. The general thio The results obtained by XRD, are in accordance with pre- subspecies like cysteine as promoting agents in sulfide vious studies [10, 25, 26], which shows that cysteine may leaching means an extensive application in bio-hydromet- serve as conductor or a bridge to facilitate electronic charge allurgical operation. transfer from gold ore to end-products.

The assumed action of cysteine in bioleaching is that weakening and breaking up of chemical bonds lead to com- ACKNOWLEDGMENT plex formation of cysteine with both iron and sulphur surfaces species. These complexations are FeS-R and FeS-S- The authors gratefully acknowledge the financial sup- R bridges, which are also the putative chemical energy car- port from Shanghai Engineering Research Center of En- rier for organisms with a cyclic turnover of cysteine and ergy- Saving in Heat Exchange Systems and the National metal ion, and electron transfers across these S- or S-S Natural Science Foundation of China (41073060). bridges. In addition, Rojas-Chapana and Tributsch [27] believed that the sulfhydryl group of cysteine participated in The authors have declared no conflict of interest. a binding process with FeS2. This reaction meant that cysteine was consumed by FeS2. surface with the subsequent release of iron-sulfur species. Former studies [12] demonstrated that low concentration of cysteine heavily improved REFERENCES the growth of cells, and cysteine could increase the zeta potential of the mineral surfaces and hence enhanced the [1] Rawlings, D.E. (2004) Microbially assisted dissolution of absorption of negatively charged bacteria to the mineral minerals and its use in the mining industry. Pure Appl. Chem. surfaces. Wang et al. [10] also speculated that the adsorp- 76 (4), 847–859. tion might be accomplished by the subunits of aporustucya- [2] Mishra, D., Kim, D.J., Ahn, J.G. and Lee, J.C. (2004). Bacte- nim and pyrite, iron or sulfur atom for pyrite crystal, amino rial leaching of metals from sulphide minerals and industrial acid residuals for apoprotein. Considering the cysteine's wastes. KIGAM Bull. 1 (9), 48–57. good affinity to pyrite, they predicted that cysteine located [3] Pradhan, N., Das, B., Gahan, C.S., Kar, R.N. and Sukla, L.B. in the framework of aporusticyanin could act as a mineral- (2006). Beneficiation of iron ore slime using Aspergillus niger and Bacillus circulans. Bioresour. Technol. 97, 1876–1879. specific receptor. It is still a process that deserves further investigation. [4] Fozia, A., Muhammad, S., and Ata, A. (2012). Biohydromet- allurgy techniques of low grade ores: A review on black shale. Based on the discussion above, it is of necessity to Hydrometallurgy. (117-118), 1–12. reexamine the issue of the inhibition of cysteine for As ex- [5] Zeng, W.M., Qiu, G.Z. and Zhou, H.B. (2011). Electrochemi- traction. Cysteine traps ferric iron, which is the main oxi- cal behaviour of massive chalcopyrite electrodes bioleached dizing reagent of As (Eq.5) with ferric iron being regener- by moderately thermophilic microorganisms at 48 °C. Hydro- ated by the bacteria. metallurgy. 105 (3-4), 259–263.

[6] Ehrlich, H.L. (2001). Past, present and future of biohydromet- allurgy. Hydrometallurgy. 59 (2-3), 127–134. 4. CONCLUSIONS [7] Brierley, J.A. (2008). A perspective on developments in bio- hydrometallurgy. Hydrometallurgy (1-4), 94, 2–7. This study employed bioleaching experiments together with SEM/EDS, FTIR and XRD to investigate details of [8] Wang, Y.M., Ji, S.J. and Chen, K.R. (1994). New discussion of submicroscopic micrographic gold in pyrite and arsenopy- gold ore bioleaching process in a mixed microorganism rite. Mineral Journal. 14(1), 83-87. system added cysteine at 40 °C. The dissolution of gold ore [9] Jiang ,T., Li, Q..and Yang, Y.B. (2008). Bio-oxidation of ar- involves many intermediate reactions and thus was a com- senopyrite, Nonferrous Met. Soc. China. 18 (6) , 1433-1438. plex process. Cysteine could greatly increase the bioleaching speed of Fe. However, it was inhibited for the bioleach- [10] Wang, Z.H., Xie, X.H. and Xiao, S.M. (2010). Comparative study of interaction between pyrite and cysteine by thermo- ing rate of As. These solid intermediate species could form gravimetric and electrochemical techniques. Hydrometal- on the mineral surface. Combining the results of the bi- lurgy. 101, 88–92. oleaching experiment and the analysis of SEM/EDS, FTIR [11] Rojas-Chapana, J.A. and Tributsch, H. (2000). Bio-leaching of and XRD leaded us to conclude that bioleaching products pyrite accelerated by cysteine. Process Biochemistry. 35, 815– were mainly jarosite and sulfur, cysteine addition just fa- 824 cilitated the electron transfer but did not change the gold [12] He, Z.G., Gao, F.L. and Zhong, H.(2009). Effects of L-cyste- ore oxidation mechanism, and the low As extraction rate in ine on Ni–Cu sulfide and marmatite bioleaching by Acidithio- bioleaching was due to iron trapping by cysteine. bacillus caldus. Bioresource Technology 100, 1383-1387.

[13] Qiu, G.Z., Fu ,B. and Zhou, H.B. (2007). Isolation of a strain of Acidithiobacillus caldus and its role in bio leaching of chalcopyrite. Word J Microbiol Biotechnol. 23 (6), 1217–1225.

[14] Ali, A., Mahin, S. and Jochen, P. (2011). Conventional and electrochemical bioleaching of chalcopyrite concentrates by moderately thermophilic bacteria at high pulp density. Hydro- metallurgy. 106 (1-2), 84–92.

[15] Marco, M., José, G. and Bessler, K.E. (2006). Process mineralogy of bacterial oxidized gold ore in São Bento Mine (Bra- sil). Hydrometallurgy. 83 (1-4), 114–123

[16] Gunneriusson, L., Sandström, A., Holmgren, A. and Kuz- mann, E. (2009). Jarosite inclusion of fluoride and its potential significance to bioleaching of sulphide minerals. Hydrometal- lurgy. 96 (1–2), 108–116. [17] Liu, J.S., Wang, Z.H. and Li B.M. (2006). Interaction between pyrite and cysteine. Trans. Nonferrous Met. Soc. China. 16, 943-946 [18] Liu, J.P., Li, J. and Ge, X. (2004). Synthesis and free radical inhibition rate of copper (II), iron (II), manganese (II) complexes with cysteine. Chemical World (5), 235-238. (in Chi- nese)

[19] Yue, S. (2000). Preparations and structure characterisation of the crystals of ferrous cysteine. Chemical Research and Appli- cation. 12(4), 387-390. (in Chinese)

[20] Liu, H., Gu, G.H. and Xu, Y.B. (2011). Surface properties of pyrite in the course of bioleaching by pure culture of Acidithi- obacillus ferrooxidans and a mixed culture of Acidithiobacil- lus thiooxidans and Acidithiobacillus ferrooxidans, Hydro- metallurgy. 108 (1-2), 143–148 [21] Jia, C.Y., Wei, D.Z. and Liu, W.G. (2008). Selective adsorption of bacteria on sulfide minerals surface. Transactions of Nonferrous Metals Society of China. 18 (5), 1247–1252. [22] Schubert, B., and Tributsch, H. (1990). Photoinduced electron transfer by coordination chemical pathways across pyrite/electrolyte interfaces. Inorg. Chem. 29, 5041–5046 [23] Suzuki, I. (2001). Microbial leaching of metals from sulphide minerals. Biotechnology Advances .19 (1), 119–132. [24] Fu, J.H, Qiu, G.Z. and Liu J.S. (2003). Study on bacterial leaching of oxide-sulphide copper ores. Mining and metallur- gical Engineering .23, 30–34. [25] Blake II, R.C., Sasaki, K. and Ohmura, N. (2001). Does aporusticyanin mediate the adhesion of Thiobacillus ferrooxidans to pyrite? Hydrometallurgy 59, 357–372. [26] Liu, Z., Yang, W.B., Bai, G. and Tian, W. (2003). Microbial enzyme conversion of L-cysteine and L-cystine. Microbiology 30 (6), 16–21. Received: March 10, 2014 [27] Rojas-Chapana, J.A. and Tributsch, H. (2000).Bio-leaching of Revised: June 05, 2014; July 14, 2014 pyrite accelerated by cysteine. Hydrometallurgy. 59, 291–300 Accepted: July 23, 2014

[28] Kolencik, M., Urik, M., Cernansky, S., Molnarova, M. and Matus, P. (2013).Leaching of zinc, cadmium, lead and copper from electronic scrap using organic acids and the Aspergillus CORRESPONDING AUTHOR niger strain. Fresenius Environ. Bull. 22, 3673-3679

[29] Ruan, F., Ye, H., Li, D and Du, D. (2013). Leaching charac- Jianshe Liu and Xuehui Xie teristics of arsenic-containing solid waste. Fresenius Environ. College of Environmental Science and Engineering Bull. 22, 207-212 Donghua University Shanghai 201620 P.R. CHINA

E-mail: [email protected]

FEB/ Vol 24/ No 1b/ 2015 – pages 379 – 385

PHYSIOLOGICAL AND BIOCHEMICAL RESPONSES OF CUCURBITA PEPO L. MEDIATED BY PORTULACA OLERACEA L. ALLELOPATHY

Hamed M. El-Shora and Ahmed M. Abd El-Gawad*

Botany Department, Faculty of Science, Mansoura University, Mansoura, Egypt

ABSTRACT into the soil rhizosphere by a variety of mechanisms, including decomposition of residue and root exudation [2]. These Weeds exert many biological stresses on crops, alle- allelochemicals become stressful only when they are toxic lopathy considered as one of these stresses. The present in- or when they affect the growth and development of sur- vestigation advances the possibility of using allelo-chemi- rounding plants [3]. The physiological effects on receptor cals as growth regulators and natural herbicides to promote plants or other organisms are useful in determining the role sustainable agriculture. Seed germination, activities of an- of the allelochemicals in the system. Many researchers in- tioxidant enzymes including peroxidase (POD, EC: vestigated the effect of plant extracts on the physiological 1.11.1.7), superoxide dismutase (SOD, EC: 1.15.1.1), cat- and growth aspects of other plants [2]. alase (CAT, EC: 1.11.1.6) and ascorbate peroxidase (APX, EC: 1.11.1.11), lipid peroxidation, proline, protein, Chl a, Portulaca oleracea L. (purslane) is a C3 plant and a Chl b, total chlorophyll and total carotenoids were meas- common troublesome weed worldwide [4]. This plant has ured in Cucurbita pepo leaves under treatment with ex- a wide biological activities (hypoglycemic and hypoli- tracts of roots and leaves of Portulaca oleracea L. (purs- pidemic activities, antioxidant, antibacterial and antitumor lane). Seed germination was reduced upon treatment with activities) [5, 6]. Portulaca oleracea has been character- P. oleracea extracts. The activities of the four antioxidant ized with many bioactive compounds such as hesperidin, enzymes as well as the antioxidant proline were increased caffeic acid [7], ferulic acid and p-coumaric acid [8, 9]. under the treatment. Protein, Chl a, Chl b, total chlorophyll Portulaca oleracea has been reported to be rich in α-lino- and total carotenoids were decreased depending on the ex- lenic acid, β-carotene [10], flavonoids, coumarins [11], al- tract concentration. Leaf extract of P. oleracea showed kaloids [12] and monoterpene glycoside [13]. Some of more allelopathic potential than root extract which may re- these bioactive compounds have been reported to be alle- veal the high and diverse bioactive compounds in the lochemicals [14]. leaves of this weed. Many studies have reported the allelopathic activity of Portulaca species. Silva et al. [15] and many authors [16, 17] reported various allelopathic potential against P. KEYWORDS: Allelopathy, enzyme activities, Portulaca oleracea, lipid peroxidation, Cucurbita pepo. oleracea. However, the physiological and biochemical response to P. oleracea allelopathy is still poor established.

The Cucurbita genus is one of the most common genera of Cucurbitaceae family, the second most important INTRODUCTION horticultural family in terms of economic importance after Solanaceae [18]. Cucurbita pepo L., is the most economi- Some plants have allelopathic potential by releasing cally important crop of this genus commonly known as allelochemicals to their surroundings that have either summer squashes [19]. Its fruits consumed as vegetables deleterious or beneficial effects on other plants in the rich sources of fat and vitamins in developing countries vicinity [1]. Allelopathy can be defined as a mechanism of like Egypt [20]. interference in plant growth and development mediated by the addition of plant-produced secondary products (allelo- The present study was taken to explore the physiolog- chemicals) to the soil rhizosphere. Allelochemicals are pre- ical and biochemical responses of C. pepo (tested plant) to sent in all types of plants and tissues and are released allelopathic effect of root and leave extracts of P. oleracea. This information will contribute to the understanding of al- * Corresponding author lelopathic mechanisms of P. oleracea allelochemicals.

MATERIALS AND METHODS ter blank was used in the assay. The enzyme activity was expressed in units per mg of protein per min. Preparation of P. oleracea extracts Superoxide dismutase (SOD) activity was determined Fresh P. oleracea plants were collected from the field by the nitroblue tetrazolium (NBT) method [24] by meas- at Mansoura, Egypt and separated into roots and leaves. uring the photoreduction of NBT at 650 nm. The reaction The roots and leaves were chopped into pieces. The com- mixture (3 ml) contained 50 mM sodium phosphate buffer ponents were then oven dried at 50 ○C for 5 days. Eighty (pH 7.8), 13 mM methionine, 75 μM NBT, 10 μM EDTA, grams of roots and leaves were respectively extracted by 2 mM riboflavin and enzyme extract (100 μl). The reaction soaking in 1L deionized water at 25 ○C for 24 h in a shaker. was started by placing the tubes below two 15-W fluores- The extracts were respectively filtered through four layers of cent lamps for 10 minutes and then stopped by switching cheesecloth to remove the fiber debris and centrifuged at off the light. The absorbance was measured at 650 nm. One 3000 rpm for 4 h [21]. The supernatant was filtered again unit of SOD was defined as the quantity of enzyme that using a 0.2-mm filter unit. Fresh stock extracts were kept produced 5 % inhibition of NBT reaction under the exper- in a refrigerator at 2 ○C until used. Various concentrations imental conditions. (2, 4, 6, 8 and 10% w/v) of roots and leaves were prepared. Catalase (CAT) activity was determined according to Aebi [25]. The assay mixture (3 ml) consisted of 100 μl Germination bioassays enzyme extract, 100 μl H2O2 (300 mM) and 2.8 ml 50 mM Seed of C. pepo were germinated according to El- phosphate buffer with 2 mM EDTA (pH 7.0). Catalase ac- Shora and Abo-Kassem [22]. Seeds were surface sterilized tivity was assayed by monitoring the decrease in the ab- in 10% sodium hypochlorite for 10 min and then soaked in sorbance at 240 nm as a consequence of H2O2 disappear- running tap water for 24 h. The seeds were then germinated ance (E = 39.4 mM-1 cm-1). between paper towels, moistened with distilled water (con- Ascorbate peroxidase (APX) activity was assayed ac- trol) or different extract concentrations in sterilized plastic cording to the method of Nakano and Asada [26]. The re- trays and were covered and incubated in dark at 25ºC. The action mixture in a total volume of 1 ml contained 100 μl germinated seeds were counted daily and the percentage of enzyme extract, 100 μl ascorbate (7.5 mM), 100 μl H2O2 germination was calculated. (300 mM) and 2.7 ml (25 mM) potassium phosphate buffer with 2 mM EDTA (pH7.0). The oxidation of ascorbate was Treatment experiment determined by the change in absorbance at 290 nm (E = The germinated seeds of C. pepo with well-grown root 2.8 mM-1 cm-1). were then supported on plastic bowls containing 0.2 mM CaCl2 solution and different concentrations of P. oleracea Lipid peroxidation analysis leaf and root extracts and vigorously aerated for 7 day ac- Fresh plant leaves (0.2 g) were homogenized and ex- cording to El-Shora and Abo-Kassem [22]. The experi- tracted in 10 mL of 0.5% (w/v) thiobarbituric acid (TBA) mental design was carried out with three replicates. made in 5% (w/v) trichloroacetic acid (TCA). The extract was heated at 95 ºC for 15 min and then quickly cooled on Enzymes assay ice [27]. After centrifuging at 5000 g for 10 min, the ab- All samples were prepared for enzyme analyses by ho- sorbance of the supernatant was measured at 532 nm. mogenization of the fresh leaves with a mortar and pestle Correction of nonspecific turbidity was made by subtract- and a small amount of sand in a solution (5 g) fresh weight) ing the absorbance value taken at 600 nm. The malondial- containing 50 mM potassium phosphate buffer (pH 7.0), 10 dehyde (MDA) was calculated using an extinction coeffi- g-1 polyvinylpyrrolidone (PVP), 0.2 mM EDTA and 10 ml 1- cient of 155 mM cm-1. 1 Triton X-100. After the homogenate was centrifuged at 12000 g for 20 min at 4°C, the supernatant was used for Determination of protein immediate determination of enzyme activities. Protein content of C. pepo leaves was determined according to the method of Bradford [28]. About 30 μL of C. Determination of enzyme activities pepo leaves extract was added to a tube and the volume was Assay of peroxidase (POD) activity is carried out accord- made up to 100 μL with 0.15 M NaCl. One mL Bradford’s ing to the procedure of Chance and Maehly [23]. 3.5 ml of reagent was added and mixed well. The absorbance was phosphate buffer (pH 6.5) was taken in a clean dry cuvette, measured at 595 nm. The concentration of the protein in the 0.2 ml seed extract and 0.1 ml of freshly prepared o-dianis- samples was determined from the calibration curve of bo- idine solution was added. The temperature of assay mixture vine albumin (0-100 μg/mL) as standard. was brought to 28-30°C and then place the cuvette in the Determination of proline spectrophotometer at 430 nm. Then, 0.2 ml of 0.2 M H2O2 was added and mixed. The initial absorbance was read at Free proline in C. pepo leaves was extracted according every 30 sec. intervals up to 3 min. A graph was plotted to the method of Bates et al. [29] by drying leaf samples with the increase in absorbance against time. From the lin- and extraction in 3% (w/v) aqueous sulfosalicylic acid in ear phase, the change in absorbance per min. was read. Wa- boiling water for 10 min. The obtained extract was filtered

and the filtrate was mixed with equal volumes of glacial activities of POD, SOD, CAT and APX by about 93.8%, acetic acid and ninhydrin reagent (1.25 g ninhydrin, 30 ml 82.0%, 700% and 130%, respectively. Some protective en- glacial acetic acid, 20 ml of 6 M H3PO4) and incubated for zymes are activated in plants due to production of oxygen 40 min in boiling water. The reaction was terminated by free radicals which stimulated under various stresses in- placing the test tubes in cold water. The samples were vig- cluding allelopathy. Increased SOD activity may be con- orously mixed with 3 ml toluene. The absorbance of the sidered as circumtantial evidence for enhanced production toluene phase was estimated at 520 nm. The proline con- of superoxide radical. This seems likely to be due to de centration was determined using a standard curve. novo synthesis of protein [34], which is attributed to tran- scription of SOD genes by a superoxide mediated transduc- Determination of photosynthetic pigments tion signals [35]. The enhanced SOD activity observed in Pigments content of Cicer leaves was determined in this study might support the hypothesis that the H2O2 re- 80% acetone extract. After centrifugation (20 000 g, 20 min) sulted from oxygen free radicals including superoxide rad- .- the absorbance was read spectrophotometrically at 663 and ical (O2 ). 645 nm. Total chlorophyll as well as chlorophyll a and b concentrations were calculated according to Arnon [30], 120 LSD [root] = 7.37 Root while the estimation of carotenoids was performed accord- a 0.05 LSD 0.05 [leaf] = 6.82 Leaf ing to Myers and Kratz [31]. 100 b c a Statistical analysis 80 b d All values are the mean of three replicates ± standard b 60 error. Data were subjected to ANOVA and the mean values e were separated based on Ducan test at 0.05 probability 40 c f level using COSTAT 6.3 program. (%) Germination 20 d e RESULTS AND DISCUSSION 0 0246810 Allelopathic effect of P. oleracea extracts on germination of C. Concentration (% w/v) pepo FIGURE 1 - Effect of Portulaca oleracea root and leaf extracts on ger- The allelopathic activity of P. oleracea root and leaf mination of Cucurbita pepo. aqueous extracts on germination of C. pepo is shown in Fig. 1. All concentrations significantly inhibited the germi- The increase in the activity of CAT could be attributed nation of C. pepo (P≤0.05) throughout experimental pe- to substrate induction in order to maintain low level of riod. The inhibition was concentration-dependent. It was H2O2 as an adaptive mechanism. The increased activity of observed that the higher concentration (10% w/v) of root APX may efficiently scavenge H2O2 to protect against ox- and leaf extracts inhibited the germination by 67.6% and idative damage by allelochemicals present in extracts. It 78.6%., respectively. These results coped with other stud- has been proposed that allelochemicals can cause oxidative ies reported for P. oleracea leaves and roots which inhib- stress in target plants and therefore activate the antioxidant ited the germination and growth of others crops such as Al- mechanism [2, 36]. The susceptibility to oxidative stress is lium cepa, Brassica oleracea, Raphanus oleracea and Lyco- a function of the overall balance between the factors that persicon esculentum [15]. It was observed that leaves of P. increase oxidant generation and those substances that ex- oleracea exhibited more allelopathic effect than roots, this hibit antioxidant capability [37, 38]. may be contributed to the high bioactive constituents in the leaves [32] especially phenolic acids such as chlorogenic, POD is considered as stress marker enzyme having a caffeic, p-coumaric, ferulic and rosmarinic acids [33]. broad specificity of phenolic substrates and a higher affinity for H2O2 than CAT. POD is located in cytosol, cell wall, Allelopathic effect of P. oleracea extracts on the enzymes ac- vacuole and extracellular spaces. It consumes H2O2 to gen- tivity of C. pepo leaves erate phenoxy compounds that are polymerized to produce The activities of SOD, POD, CAT and APX were in- cell wall compounds such as lignans [39]. The increase in vestigated in C. pepo leaves to determine whether P. POD is correlated with treatment by allelochemicals sug- oleracea extracts influenced these antioxidant enzymes or gesting that it is an intrinsic defense tool [40]. The in- not. All enzymes activities, estimated on a fresh weight ba- creased POD activity in the present results might be due to sis, were substantially increased under the influence of P. the release of peroxidase localized in the cell wall. oleracea root and leaf extracts (Fig. 2). P. oleracea root extract significantly (P≤0.05) enhanced the activities of Allelopathic effect of P. oleracea extracts on lipid peroxidation POD, SOD, CAT and APX by about 164.5%, 226.1%, of C. pepo leaves 638.9% and 428.6% at the highest concentration (10% MDA as marker for membrane peroxidation was en- w/v). On the other hand, P. oleracea leaf extract enhanced hanced by P. oleracea leaf and root extracts (Fig. 3). The

60 70 a LSD [root] = 3.24 LSD 0.05 [root] = 3.04 0.05 a LSD [leaf] = 4.23 (SOD) b LSD [leaf] = 4.48 (POD) 0.05 60 0.05 50 c b d a 50 c 40 e d fresh weight) f a fresh weight) 40 e b -1 30 b -1 e c c d d 30 e 20 e f Root 20 Root f Leaf 10 Leaf 10 Activity (µmole Activityg (µmole Activity (µmole Activityg (µmole 0 0 0246810 0246810 Concentration (% w/v) Concentration (% w/v)

16 16 LSD 0.05 [root] = 1.11 (CAT) d a LSD 0.05 [root] = 0.82 (APX) a 14 LSD 0.05 [leaf] = 0.97 LSD [leaf] = 0.32 bc b 14 0.05 c bc 12 12 b a b b 10 b fresh weight) 10 Root fresh weight) c -1

-1 Leaf 8 8 6 a 6 Root b d c 4 Leaf 4 d d d c 2 2 e

Activity (µmole Activityg (µmole e Activity (µmole Activityg (µmole c 0 0 0246810 0246810 Concentration (% w/v) Concentration (% w/v)

FIGURE 2 - Effect of Portulaca oleracea root and leaf extracts on the activity of antioxidant enzymes in Cucurbita pepo leaves.

70 Root High MDA production in the present study indicates the ex- LSD 0.05 [root] = 3.24 a tent of cell damage which might be due to allelochemicals in LSD [leaf] = 4.48 Leaf 60 0.05 P. oleracea extracts. Destruction of lipid components of b a c membrane by lipid peroxidation causes membrane impair- 50 ment and leakage [41, 42]. Lipid peroxidation can be stimu- d b lated by iron through Fenton reaction and also accelerates pe-

fresh weight) 40 e c roxidation by decomposing lipid hydroperoxides into peroxyl

-1 e d and alkoxyl radicals that can themselves abstract hydrogen 30 MDA content MDA e and perpetuate the chain reaction of lipid peroxidation [43]. f Qian et al. [44] reported that allelochemicals stress

(µmole g (µmole 20 cause oxidative damage and trigger the synthesis of reac- 10 tive oxygen species (ROS) to disrupt the subcellular struc- 0246810 ture. The increase of MDA formation is a direct conse- Concentration (% w/v) quence of increase ROS formation and thus unsaturated FIGURE 3 - Effect of Portulaca oleracea root and leaf extracts on fatty acid peroxidation MDA content leaves of Cucurbita pepo. Allelopathic effect of P. oleracea extracts on protein content content of MDA in C. pepo was significantly (P≤0.05) in- of C. pepo leaves creased by about 164.5% and 93.8% under the highest con- The allelopathic effect of P. oleracea leaf and root ex- centration (10% w/v) of root and leaf extracts, respectively. tracts on the protein content of C. pepo leaves are presented

in Fig. 4. Protein content at the highest concentration scavenging [52] and the increase of proline concentration (10% w/v) was significantly (P≤0.05) decreased by about can be considered as a defense reaction against allelochem- 73.5% and 63.4% under the effect of P. oleracea root and ical stress. leaf extracts, respectively. These results suggest that allelochemicals of P. oleracea may inhibit protein synthesis Allelopathic effect of P. oleracea extracts on pigments of C. [45] or stimulate the degradation of the existing protein pepo leaves in C. pepo plant [46]. The same effect of allelochemicals The phytotoxic effect of P. oleracea root extract on was reported in other plants such as Brassica napus, Trit- chlorophyll and carotenoid contents of C. pepo leaves were icum aestivum [47], Vicia faba, Zea mays [48], T. durum illustrated in Fig. (6a). Chl b and carotenoid were de- [49] and Achillea biebesteinii [50]. creased gradually depending on the concentration. It decreased by about 81.4% and 77.8%, respectively at 10% 70 (w/v). On the other hand, Chl a and total Chl were signifi- Root e LSD 0.05 [root] = 2.47 antly enhanced (P≤0.05) by about 7- and 2-fold, respec- LSD [leaf] = 4.06 Leaf 60 0.05 tively at the lowest concentration (2% w/v) of P. oleracea e root extract, while they were decreased again by about 50 37.5% and 65.6%, respectively at 10% w/v.

fresh weight) d

-1 f 40 Chl a 3,0 LSD [Chl a] = 0.08 e c Chl b 0.05 Carotenoid LSD 0.05 [Chl b] = 0.06 30 d b Total LSD 0.05 [Carotenoid] = 0.05 a LSD [Total] = 0.19 c 2,5 a a 0.05 20 b a b 2,0 b 10 fresh weight) -1 Protein content (µg g content Protein c (a) 0 1,5 0246810 a a b Concentration (% w/v) 1,0 a d a b b c FIGURE 4 - Effect of Portulaca oleracea root and leaf extracts on b b c protein content in leaves of Cucurbita pepo. d d 0,5 e 120 g (mg content Pigment c e Root d LSD 0.05 [root] = 2.47 a e Leaf 0,0 100 LSD 0.05 [leaf] = 3.54 b c 80 3,0 Chl a LSD [Chl a] = 0.07 Chl b 0.05 60 d a LSD [Chl b] = 0.05 e Carotenoid 0.05

fresh weight) b c LSD 0.05 [Carotenoid] = 0.03

-1 Total 2,5 LSD [Total] = 0.16 40 0.05 Proline content d a f e a (µg g 20 2,0

e fresh weight) b 0 -1 (b) 0246810 1,5 b Concentration (% w/v) a a FIGURE 5- Effect of Portulaca oleracea root and leaf extracts on pro- c line content in leaves of Cucurbita pepo. 1,0 b a b b c Allelopathic effect of P. oleracea extracts on proline content a c d of C. pepo leaves 0,5 b b d e d The allelopathic effect of P. oleracea leaf and root ex- g (mg content Pigment c e tracts on the proline content of C. pepo leaves were illus- 0,0 d e trated in Fig. (5). Proline content significantly increased 0246810 (P≤0.05) by about 397% and 180.6% for root and leaf ex- Concentration (% w/v) tracts at the highest concentration (10% w/v), respectively. Proline belongs to non-specific defense system against tox- FIGURE 6- Effect of Portulaca oleracea leaf (a) and root (b) extracts icity, as an inhibitor of lipid peroxidation [51] after radical on pigments in leaves of Cucurbita pepo.

The allelopathic effect P. oleracea leaf extract on [2] Weir, T., Park, S. and Vivanco, J. (2004) Biochemical and chlorophyll and carotenoid contents of C. pepo leaves were physiological mechanisms mediated by allelochemicals. Cur- rent Opinion in Plant Biology. 7, 472-479. presented in Fig. (6b). Chl b and carotenoid were decreased gradually with increasing the concentration. They de- [3] Tripathi, V.D., Venkatesh, A., Prasad, R. and Dhyani, S.K. (2013) Phyto-toxicity of Eucalyptus tereticornis clones on creased by about 72.1% and 67.2%, respectively at 10% Leucaena leucocephala L. International Journal of Avanced w/v. Chl a and total Chl were significantly enhanced Research. 1, 82-87. (P≤0.05) at the lowest concentration of P. oleracea leaf ex- [4] Miyanishi, K. and Cavers, P. (1980) The biology of Canadian tract by about 4- and 2-fold, respectively, while they were weeds: 40. Portulaca oleracea L. Canadian Journal of Plant increased again by about 37% and 53.9%, respectively at Science. 60, 953-963. 10% w/v. Generally Chl b and carotenoid contents were [5] Lim, Y., Lim, T. and Tee, J. (2007) Antioxidant properties of decreased gradually by concentration of P. oleracea root several tropical fruits: A comparative study. Food chemistry. extract. These results are in agreement with other reports 103, 1003-1008. [49, 53]. This reduction could be attributed to the reduction [6] Zhao, R., Gao, X., Cai, Y., Shao, X., Jia, G., Huang, Y., Qin, in chlorophyll biosynthesis and/or degradation of the exist- X., Wang, J. and Zheng, X. (2013) Antitumor activity of Por- ing chlorophylls [54]. Allelochemicals interfere with portulaca oleracea L. polysaccharides against cervical carcinoma phyrin biosynthesis [55] or inhibit the activity of protopor- in vitro and in vivo. Carbohydrate Polymers. 96, 376-83. phyrinogen, the enzyme involved in the chlorophyll bio- [7] Yang, Z., Zheng, Y. and Xiang, L. (2007) Study on chemical synthesis pathway [56]. Another possibility for the reduc- constituents of Portulaca oleracea. Journal of Chinese Medic- tion of chlorophyll content is the lipid peroxidation medi- inal Materials. 30, 1248-1250. ated cell damage in tissues. [8] Cheng, L., Cheng, Z., liu, H., Zhang, H., Zhang, W., Du, Y., Wang, Y., Li, H., Ying, X. and Kang, T. (2011) Liquid chro- On the other hand, Chl a drastically increased over the matographic (LC) determination of four bioactive compounds in the Portulaca oleracea L. Journal of Medicinal Plants Re- control and then gradually decreased by the concentration search. 5, 6876-6880. but still over the control. These results supported by the finding of El-Rokiek and Eid [57]. Generally, pigment con- [9] Xiang, L., Xing, D., Wang, W., Wang, R., Ding, Y. and Du, L. (2005) Alkaloids from Portulaca oleracea L. Phytochemis- tents and photosynthetic efficiency in leaves were fre- try. 66, 2595-2601. quently altered in the presence of allelochemicals [58]. Leaf extract of P. oleracea showed more allelopathic po- [10] Barbosa-Filho, J.M., Alencar, A.A., Nunes, X.P., Tomaz, A.C., Sena-Filho, J.G., Athayde-Filho, P.F., Silva, M.S., Souza, M.F. tential than root extract. Similar finding was reported by and Da-Cunha, E.V. (2008) Sources of alpha-, beta-, gamma-, recent work [59]. delta-and epsilon-carotenes: A twentieth century review. Revista Brasileira de Farmacognosia. 18, 135-154. [11] Awad, N. (1994) Lipid content and antimicrobial activity of CONCLUSION phenolic constituents of cultivated Portulaca oleracea L. Bul- letin of Faculty of Pharmacy, Cairo University. 32, 137-142. Based on the results in present work, the treatment of [12] Yang, Z., Liu, C., Xiang, L. and Zheng, Y. (2009) Phenolic C. pepo seeds with P. oleracea root and leaf extracts asso- alkaloids as a new class of antioxidants in Portulaca oleracea. ciated with significant inhibition of germination. C. pepo Phytotherapy Research. 23, 1032-1035. seedlings expressed significant enhancement of many [13] Sakai, N., Inada, K., Okamoto, M., Shizuri, Y. and Fukuyama, physiological parameters including the activity of antioxi- Y. (1996) Portuloside A, a monoterpene glucoside, from Por- dant enzymes POD, SOD, CAT, APX, lipid peroxidation tulaca oleracea. Phytochemistry. 42, 1625-1628. and proline contents. In addition, protein and pigment con- [14] Cheema, Z.A., Iqbal, M. and Ahmad, R. (2002) Response of tents were decreased under the phytotoxic effect of P. wheat varieties and some rabi weeds to allelopathic effects of sorghum water extract. International Journal of Agriculture oleracea root and leaf extracts. P. oleracea caused oxida- and Biology. 4, 52-55. tive stress, as evidenced by the increased lipid peroxidation in C. pepo leaves. Leaf extract of P. oleracea showed more [15] Silva, M., Magrico, S., Dias, A.S. and Dias, L.S. (2007) Alle- lopathic plants. 20. Portulaca oleracea L. Allelopathy Journal. allelopathic potential than root extract which may reveal 10, 275-286. the high and diverse bioactive compounds in the leaves of this weed. Further studies needed for separation and char- [16] Dadkhah, A. (2013) Allelopathic effect of sugar beet (Beta vulgaris) and eucalyptus (Eucalyptus camaldulensis) on seed acterization of the allelochemical(s) of P. oleracea. germination and growth of Portulaca oleracea. Russian Agri- cultural Sciences. 39, 117-123. The authors have declared no conflict of interest. [17] El-Rokiek, K.G., El-Nagdi, W.M. and El-Masry, R. (2012) Controlling of Portulaca oleracea and Meloidogyne incognita infecting sunflower using leaf extracts of Psidium guava. Ar- chives of Phytopathology and Plant Protection. 45, 2369-2385. REFERENCES [18] Esteras, C., Gómez, P., Monforte, A.J., Blanca, J., Vicente- Dólera, N., Roig, C., Nuez, F.a and Picó, B. (2012) High- [1] Cheema, Z.A., Farooq, M. and Wahid, A. (2013) Allelopathy: throughput SNP genotyping in Cucurbita pepo for map con- Current trends and future applications. Springer, Heidelberg, struction and quantitative trait loci mapping. BMC genomics. New York, Verlag. 13, 80.

[19] Paris, H.S. (2008) Summer squash. In: Prohens, J., Nuez, F. [38] Abu-Romman, S. (2012) Molecular cloning and expression of (ed.), Handbook of Plant Breeding Vegetables I. Vol 1, 12-oxophytodienoic acid reductase gene from barley. Austral- Springer, Heidelberg, 351-379. ian Journal of Crop Science. 6, 649-655.

[20] Ferriol, M. and Picó, B. (2008) Pumpkin and winter squash. [39] Reddy, V., Urooj, A. and Kumar, A. (2005) Evaluation of an- In: Prohens, J., Nuez, F. (ed.), Handbook of Plant Breeding tioxidant activity of some plant extracts and their application Vegetables I. Vol 4, Springer, Heidelberg, 317-349. in biscuits. Food chemistry. 90, 317-321.

[21] Chon, S., Choi, S., Jung, S., Jang, H., Pyo, B. and Kim, S. [40] Ceker, S., Agar, G., Alpsoy, L., Nardemir, G. and Kizil, H.E. (2002) Effects of alfalfa leaf extracts and phenolic allelochem- (2013) Protective role of essential oils of Calamintha nepeta icals on early seedling growth and root morphology of alfalfa L. on oxidative and genotoxic damage caused by aflatoxin b1 and barnyard grass. Crop protection. 21, 1077-1082. in vitro. Fresenius Environmental Bulletin. 22, 3258-3263.

[22] El-Shora, H. and Abo-Kassem, E. (2001) Kinetic characteri- [41] Halliwell, B. (1994) Free radicals and antioxidants: a personal zation of glutamate dehydrogenase of marrow cotyledons. view. Nutrition reviews. 52, 253-265. Plant Science. 161, 1047-1053. [42] Halliwell, B. (1997) Antioxidants and human disease: a gen- [23] Chance, B. and Maehly, A. (1955) Assay of catalases and pe- eral introduction. Nutrition reviews. 55, S44-S49. roxidases. Methods in enzymology. 2, 764-775. [43] Halliwell, B. (1991) Reactive oxygen species in living sys- [24] Becana, M., Aparicio-Tejo, P., Irigoyen, J.J. and Sanchez-Diaz, tems: source, biochemistry, and role in human disease. The M. (1986) Some enzymes of hydrogen peroxide metabolism in American journal of medicine. 91, S14-S22. leaves and root nodules of Medicago sativa. Plant physiology. 82, 1169-1171. [44] Qian, H., Xu, X., Chen, W., Jiang, H., Jin, Y., Liu, W. and Fu, [25] Aebi, H. (1984) Catalase in vitro. Methods in Enzymology. Z. (2009) Allelochemical stress causes oxidative damage and 105, 121-126. inhibition of photosynthesis in Chlorella vulgaris. Chemo- sphere. 75, 368-375. [26] Nakano, Y. and Asada, K. (1981) Hydrogen peroxide is scav- enged by ascorbate-specific peroxidase in spinach chloro- [45] Einhellig, F.A. (1996) Interactions involving allelopathy in plasts. Plant and Cell Physiology. 22, 867-880. cropping systems. Agronomy Journal. 88, 886-893. [27] Zhang, W., Zhang, F., Raziuddin, R., Gong, H., Yang, Z., Lu, [46] Mersie, W. and Singh, M. (1993) Phenolic acids affect photo- L., Ye, Q. and Zhou, W. (2008) Effects of 5-aminolevulinic synthesis and protein synthesis by isolated leaf cells of velvet- acid on oilseed rape seedling growth under herbicide toxicity leaf. Journal of chemical ecology. 19, 1293-1301. stress. Journal of Plant Growth Regulation. 27, 159-169. [47] Ullah, N., Haq, I.U., Safdar, N. and Mirza, B. (2013) Physio- [28] Bradford, M.M. (1976) A rapid and sensitive method for the logical and biochemical mechanisms of allelopathy mediated quantitation of microgram quantities of protein utilizing the by the allelochemical extracts of Phytolacca latbenia (Moq.) principle of protein-dye binding. Analytical Biochemistry. 72, H. Walter. Toxicology and Industrial Health. 25, 1-7. 248-254. [48] Saleh, A. and Madany, M. (2013) Investigation of the allelopa- [29] Bates, L., Waldren, R. and Teare, I. (1973) Rapid determina- thic potential of Alhagi graecorum Boiss. Asian Journal of Ag- tion of free proline for water-stress studies. Plant and Soil. 39, ricultural Research. 7, 1-9. 205-207. [49] Abu-Romman, S., Shatnawi, M. and Shibli, R. (2010) Alle- [30] Arnon, D. (1949) Copper enzymes in isolated chloroplasts. lopathic effects of spurge (Euphorbia hierosolymitana) on Polyphenoloxidase in Beta vulgaris. Plant Physiology. 24, 1- wheat (Triticum durum). American-Eurasian Journal of Agri- 15. cultural and Environmental Science. 7, 298-302.

[31] Myers, J. and Kratz, W. (1955) Relations between pigment [50] Abu-Romman, S. (2011) Allelopathic potential of Achillea content and photosynthetic characteristics in a blue-green alga. biebersteinii Afan.(Asteraceae). World Applied Sciences The Journal of General Physiology. 39, 11-22. Journal. 15, 947-952. [32] Yan, J., Sun, L., Zhou, Z., Chen, Y., Zhang, W., Dai, H. and [51] Mehta, S. and Gaur, J. (1999) Heavy‐metal‐induced proline Tan, J. (2012) Homoisoflavonoids from the medicinal plant accumulation and its role in ameliorating metal toxicity in Portulaca oleracea. Phytochemistry. 80, 37-41. Chlorella vulgaris. New Phytologist. 143, 253-259. [33] Erkan, N. (2012) Antioxidant activity and phenolic compounds [52] Alia, Mohanty, P. and Matysik, J. (2001) Effect of proline on of fractions from Portulaca oleracea L. Food Chemistry. 133, the production of singlet oxygen. Amino Acids. 21, 195-200. 775-781. [34] Verma, S. and Dubey, R. (2003) Lead toxicity induces lipid [53] Bagavathy, S. and Xavier, G. (2007) Effects of aqueous extract peroxidation and alters the activities of antioxidant enzymes in of Eucalyptus globulus on germination and seedling growth of growing rice plants. Plant Science. 164, 645-655. sorghum. Allelopathy Journal. 20, 395-402. [35] Fatima, R.A. and Ahmad, M. (2005) Certain antioxidant en- [54] Yang, C.-M., Chang, F., Lin, S.-J. and Chou, C.-H. (2004) Ef- zymes of Allium cepa as biomarkers for the detection of toxic fects of three allelopathic phenolics on chlorophyll accumula- heavy metals in wastewater. Science of the Total Environ- tion of rice (Oryza sativa) seedlings: II. Stimulation of con- ment. 346, 256-273. sumption-orientation. Botanical Bulletin of Academia Sinica. 45, 119-125. [36] He, F., Deng, P., Wu, X., Cheng, S., Gao, Y. and Wu, Z. (2008) Allelopathic effects on Scenedesmus obliquus by two sub- [55] Rice, E.L. (1984) Allelopathy. Academic Press, New York. merged macrophytes Najas minor and Potamogeton malaia- nus. Fresenius Environmental Bulletin. 17, 92-97. [56] Romagni, J.G., Meazza, G., Nanayakkara, N.P.D. and Dayan, F.E. (2000) The phytotoxic lichen metabolite, usnic acid, is a [37] Foyer, C.H., Lelandais, M. and Kunert, K.J. (1994) Photooxi- potent inhibitor of plant p-hydroxyl phenyl pyruvate dioxy- dative stress in plants. Physiologia Plantarum. 92, 696-717. genase. FEBS Letters. 480, 301-305.

[57] El-Rokiek, K. and Eid, R. (2009) Allelopathic effects of Euca- lyptus citriodora on amaryllis and associated grassy weed. Planta Daninha. 27, 887-899. [58] Wu, X., Hu, T., Yang, W., Chen, H., Hu, H., Tu, L., Pan, Y. and Zeng, F. (2012) Effects of Eucalyptus grandis leaf litter decomposition on the growth and photosynthetic characteristics of Cichorium intybus. The Journal of Applied Ecology. 23, 1-8. [59] Gulzar, A. and Siddiqui, M.B. (2014) Evaluation of allelopathic effect of Eclipta alba (L.) Hassk on biochemical activity of Amaranthus spinosus L., Cassia tora L. and Cassia sophera L. African Journal of Environmental Science and Technology.

8, 1-5.

Received: March 12, 2014

Revised: May 05, 2014; May 28, 2014 Accepted: June 11, 2014

CORRESPONDING AUTHOR

Ahmed M. Abd El-Gawad Botany Department Faculty of Science Mansoura University

Mansoura EGYPT

Phone: +201003438980 Fax: +2050224678 E-mail: [email protected]; [email protected]

FEB/ Vol 24/ No 1b/ 2015 – pages 386 - 393

COMBINING SATELLITE AND GIS DATA TO ANALYZE CHANGES IN TROPICAL FORESTS ON CENTRAL HAINAN ISLAND IN RESPONSE TO THE NATIONAL LOGGING BAN AND ECONOMIC DEVELOPMENT

Shudong Wang1,2,*, Yuhe Ji3, Zhiyun Ouyang2, Lifu Zhang1 and Liu Jia1

1State Key Laboratory of Remote Sensing Science, Institute of Remote Sensing and Digital Earth (RADI), CAS, Beijing 100101, China 2State Key Laboratory of Urban and Regional Ecology, Research Center for Eco-Environmental Sciences, CAS, Beijing 100085, China 3State Key Laboratory of Vegetation and Environmental Change, Institute of Botany, CAS, Beijing 100093, China

ABSTRACT KEYWORDS: Forest management, GIS, impact factor, multiple forests transition, object oriented decision tree technique, remote sensing Changes in natural forest cover in tropical areas have attracted international attention. Rubber and pulp plantations threaten Hainan Island’s natural tropical forests. Remote sensing provides a crucial tool for understanding 1. INTRODUCTION how forests change in response to forest protection strategies and economic development. China’s government has Deforestation of natural tropical forest is the main adopted protective measures designed to balance forest cause for loss of biodiversity, and it has attracted interna- protection and economic development on Hainan; however, tional attention [1-4]. In China, a developing country, natural the effect of both management and economic development forests continue to disappear and some forests are widely on natural tropical forest remains unclear. To identify displaced by other land uses because of conflict between changes in forest types, object-oriented decision-tree identi- economic development and forest protection [5]. fication techniques were developed using Landsat TM Tropical forests cover a large area of China’s Hainan images to identify causes of forest change. GIS tech- Island. China’s reforestation policies are currently respon- niques allowed analysis of the forest’s spatial shift using sible for an increase in the total area of tropical forest on elevation, slope, transportation corridors, natural reserves, Hainan Island, but the proportion of natural tropical forest and farmlands linked to three different periods of forestry to economic forest has changed in recent years [6]. Rubber policy and economic development from 1988 to 2008. The and pulp plantations represent a double threat to Hai- analysis shows: (1) Total tropical forest area increased nan’s natural tropical forests [7], and changes in these trop- from 1988 to 2008, while natural tropical forest area in- ical forests have affected the maintenance of biodiversity, creased slightly from 1988 to 1998, but decreased signifi- ecological functioning, and stability of the forest ecosys- cantly from 1998 to 2008, despite implementation of the tem [8-11]. "Natural Forest Protection Project." Meanwhile, economic Previous studies indicate that tropical forests may suffer forests, mainly rubber and pulp plantations, expanded from from the complex effects of population growth, economic 1988 to 2008. (2) Spatial changes occurred. Natural tropical development, national policies, and natural factors (e.g., forest shifted from the lower piedmont to higher mountain- terrain and climate) [6, 12-14]. However, the key factors tops, and economic forests shifted to higher elevations un- causing changes in tropical forest remain unclear, and prov- der the complex effects of multiple factors. (3) The ob- ing which protection measures would be most effective in served changes in forest cover could be related to protec- preventing the destruction of natural tropical forest is dif- tive measures and economic development, with economic ficult. Therefore, managers remain confused with regard to development seemingly having the strongest influence on balancing economic development and forest protection [2- the condition of the forests. Elevation, slope, transportation 4]. corridors, and farmlands also affected the shift of tropical forests. Because the proportion of natural tropical forest to economic forest has changed, two key challenges arise: (1) understanding how and why natural tropical forests are changing, and which factors have led to the changes; (2) * Corresponding author understanding the fate and implications of the succession

of different forest types. Thus, monitoring the dynamics of differences from December to the following January than those changes occurring in natural tropical forest and eco- it does during other seasons [20]. In addition, remotely nomic forest becomes necessary, together with the identifi- sensed images of natural forest, and rubber and pulp plan- cation of the main factors leading to those changes over tations each have obvious characteristics related to their broad spatial and temporal scales. It is established that distribution, patch shape, and texture [4]. For example, those changes are occurring in response to evolving protec- pulp and rubber plantations usually occur in areas where tion strategies, national policies, and economic develop- the slope is less than 25°, and rubber plantations often lie ment. at comparatively lower elevations. Remote sensing provides an efficient technique for the To overcome the above challenges, we used an object- monitoring and managing of tropical forests [15]. Brandt et oriented decision-tree method to deal with the complex al. [16] recognized that mapping forest distribution and suc- change processes of natural forest, and rubber and pulp cession are an essential component of forest biodiversity as- plantations. The overlay technique of GIS was also used to sessment. In addition, a combination of remote sensing and map natural forest, and rubber and pulp plantations during GIS techniques could help scientists discover the intrinsic 1988–2008, and to analyze the main factors driving forest forces driving the dynamics of forests. change and the relationships between the three forest types. However, some challenges remain in solving these Our objectives were to: 1. Map the spatial extent of natural problems with the aid of remote sensing image classifica- forest, and rubber and pulp plantations, and to analyze tem- tion. For example, when using remote sensing data for veg- poral and spatial succession occurring during 1988 and etation classification, there is a problem with mixed pixels 2008. 2. Assess the relationships between the spatial pat- resulting from same objects exhibiting different reflectance terns of multiple forests; identify the main factors driving at varying wavelengths. [17]. In addition, it is difficult to changes. improve the precision of the process of extraction without the support of a priori knowledge, such as the spatial distribution of various forests or patch shapes. 2. MATERIALS AND METHODS In addition, the use of pixel-based methods are prone to involving tiring and labor-intensive work, and misclas- 2.1. Study area sification of pixels is likely to occur because of errors dur- The study area covers 14,000 km2 in the central part of ing spectrum analysis [18]. For example, some natural for- Hainan Island (Fig. 1). The area’s warm and humid tropical ests lying within shaded areas of mountains tend to be re- monsoon climate has an annual average temperature rang- garded as other land use types. An object-oriented method ing from 22 to 26 °C. The coldest and driest months occur could segment remote sensing images into different patch from December to February. Mountainous areas surround sizes based on integrated features of the spectrum, texture, the Five Fingers and Yinggeling Mountains on Hainan Is- shape of patch, and so forth [19]. In addition, the decision- land. The island’s natural tropical forest mainly includes tree method could gradually extract individual land use tropical monsoon forest, tropical rain forest, evergreen types, using remote sensing extraction models and relevant broad-leaved forest, and coniferous forest. Economic for- auxiliary data, such as the distribution of various forest est mainly includes rubber plantations and orchards, which types. usually occur at the foot of the mountains. Since the 1970s, In fact, many auxiliary information could be used. For some national and provincial reserves have been estab- example, when compared with rubber and pulp plantations, lished to protect the biodiversity of these tropical forests. the remotely-sensed spectra of natural forest exhibits more

FIGURE 1 - The location of the study area (the central Hainan Island in China)

2.2. Data sources MNDWI = (Green – MIR) / (Green + SWIR) [21], the Uni- Landsat TM images (path/row 124/47) cover the study versal Pattern Decomposition Method (VIUPD) [22], RVI = area. We used six images captured from December to the NIR/Red [23] and the Grass and Shrub Differing Index GSI following February in 1988, 1998, and 2008; these images = (MIR–NIR) / (MIR + NIR) [6, 23]. The wave bands of were obtained from the University of Maryland website Green, Red, MIR, SWIR, and NIR was used in these mod- (http://glcfapp.glcf.umd.edu:8080/esdi/index.jsp) and sat- els, since they contribute to identify different objects. And ellite ground stations in China. The images were georefer- it is concluded that the MNDVI, VIUPD, RVI, GSI per- enced with a precision better than 0.4 pixels. Finally, the formed well in the information extraction of complex fea- digital data of these images were calibrated to surface re- tures. MNDWI is a modified normalized difference water flectance values using the Fast Line-of-Sight Atmospheric index which is a simple graphical indicator that can be used correction of Spectral Hypercubes (FLAASH) Module in to analyze whether the target being observed in water or ENVI v.4.6. Other sources of information included a digi- not [21]. The vegetation index RVI is very sensitive to veg- tal elevation model (DEM), and social, economic, and field etation canopy chlorophyll content [23]. VIUPD is a vege- survey data. DEM data were obtained from Aster satellite tation index and sensitively reflect the amount of vegeta- data (https://wist.echo.nasa.gov/~wist/ api/imswelcome/). tion and the degree of vegetation vigor [22]. GSI is grass Slope and elevation data were derived from the DEM using and shrub differing index and can be used to separate ARCGIS9.3 software. Some basic data (transportation cor- bushes and grass from other kinds of trees [6, 23]. Then the ridors, population, national and provincial reserves, mete- object-oriented forests information were extracted based orological, social and economic data) were obtained from on decision tree (Fig.4). Multi-source data (including slope the Hainan Provincial Academy of Environmental Sci- from the DEM, pulp plantation planning maps, and other ences. Field surveys were also conducted to collect infor- auxiliary data) were used to extract different land use types. mation on the distribution of forests. According to the stage order, we first separated water-related ground objects (water bodies, paddy fields, and aquaculture 2.3. Methods areas) from images using the MNDWI model, and then 2.3.1. Extraction of multiple forest information datasets identified areas without vegetation coverage, such as ur- We compared the spectral differences of Landsat TM banized areas, using the VIUPD method. Next, the RVI images captured in 1988, 1998, and 2008 of multiple for- model, slope from DEM, and economic forest planning ests in different seasons within a year, and selected images map were used to extract and delineate pulp plantation ar- taken between December and the following January (e.g., eas from other land use types. Finally, we separated natural December 1988 to January 1989). Then we established forest and rubber plantation areas from grassland and or- flow charts and processes for use with a sophisticated ob- chards using the GSI (Grass and Shrub Differing Index) ject-oriented decision-tree method (Fig. 4). First, these im- model and slope data from DEM. To correct some mis- ages were divided into multi-scale segmentations based on takes, we did some field investigation. For example, we in- the texture, spectrum, patch shape, and distribution of land itially were not sure whether some patches were rubber use types using Cognition 7.5 software. There are also plantation or not in Landsat TM image, but we found other parameters need to be set, such as scale, shape, and through field investigation that these patches indeed were compactness which are obtained from the ground compar- rubber plantation if they were near river and residential ative research. Second, a set of indices extracting remote area, so we might correct these mistakes through water sys- sensing information were calculated (Fig. 3), including tem distribution map.

FIGURE 2 - Examples of how phenology was used to discriminate the different forest classes. Representative pixels from the three forest and other land use classes look similar under visual inspection on Landsat images from December to the following February

FIGURE 3 - Using various indices models to identify typical objects (water, paddy field, aquaculture, forest, orchard, grass, pulp woods, rubber woods, natural forest etc.) based on decision tree

2.3.2 Change detection and key driving factor identification Distinguishing water related and land related objects :MNDWI Step 0 Three forest types and other land use types in the study area were extracted: natural forest, rubber plantation, pulp- paper, and other land use types (e.g., paddy fields, dry Distinguishing vegetation Water related lands, orchards, sand, and urbanized areas). Map overlays and non vegetation objects Step 1 of forest type for 1988, 1998, and 2008 were created using areas:VIUPD GIS software to quantify the dynamic transformation between different forest types. Economic activities and pro- Distinguishing between high and Non vegetation tection measures were considered and a comparative anal- low vegetation coverage areas: Step 2 area RVI , DEM and planning map ysis method was used to compare natural rubber demand and price changes, the area of natural forests and the area Distinguishing high and of economic forests before and after the implementation of Identification of low vegetation coverage protection measures. pulp forest Step 3 areas: GSI and DEM To reveal the intrinsic driving factors from 1988 to 2008, in addition to considering economic and policy fac- Shrub and tors, the following factors were also considered: nature Natural Rubber Step 4 forests plantation grassland protection areas, farms, transportation corridors, and the elevation and slope of each site. The data of nature protection

FIGURE 4 - The steps to extract object-oriented information of for- areas derived from the map of national nature protection ests based on decision tree area; the data of farms and transportation was extracted

from Landsat TM images; the data of elevation and slope charts and processes to develop a sophisticated object-ori- was extracted by from DEM. The study area was divided ented decision-tree method, and segmented the images into into grid cells, and the information in each grid cell was multiple scale patches and separate various forest types extracted by the spatial analysis model of GIS software. By aided by selecting remote sensing models and the related using random permutation tests [24] to choose grid cells in auxiliary data based on the sequence of nodes. each year, the association of the different forest types with the above factors was examined. We used the average Eu- clidean distance between the multiple forest patches and these driving factors to capture differences in conservation value and human activity. For multiple forest conversions, the following method was used to identify the key factors: the patch of the different forest area serves as the basic unit; then, adjacent patches with the same change trends were classified into a uniform block. The average value of a factor in each block with the same trend was used to serve as a sample, and the value of the factor served as a collective value. The ratio of the area covered by a particular land use type in 2008 was compared with the corresponding blocks of the two previous periods (1988 and 1998) to detect changes using the overlay. Then, we took the altitude, slope, and the minimum distance from the farmland and road to the area transformation occurred as variable and took the transformation of forest types as dependent variable. General regression analyses between them were con- FIGURE 5 - The classification of different forest types in 2008 ducted by collectively analyzing the blocks with the same trends [12]. TABLE 1 - Producer (PA) and user (UA) accuracies for each land cover class in 2008 Class PA UA Overall Khat 3. RESULTS AND DISCUSSIONS accuracy Natural forest 95% 90% 91% 0.91 3.1. Detection of changes in natural forest, and rubber and Rubber plantation 87% 92% pulp plantations Pulp forest 93% 89% Khat is an evaluation method for the precision of classification based on We evaluated the precision of the classification (Fig. 5) multivariate statistics by comparing results with high-resolution remote sensing images obtained by the Quick Bird and SPOT satellites, The method could effectively avoid problems related and 300 GPS locations confirmed by local technical staff to the spectral variability of some forested lands in shadow, in the field. Area-adjusted overall accuracy was over 91% caused by strong illumination effects on topography, and for natural forest and for the rubber and pulp plantations. could also effectively reduce uncertainties of classification. For the change in class, producer accuracies (PA) were In addition, the method could effectively identify the spa- lower than the user accuracies (UA), representing a high tial distribution of various vegetation types and improve rate of errors of omission, which means our estimates were classification accuracy by the use of relevant auxiliary data conservative. The lowest accuracy was the PA for the rubber and a priori knowledge. plantations (87%), and the highest (95%) was for natural for- Despite the higher level of accuracy achieved by our est (Table 1). classification, some uncertainties still exist. We set the al- Accurate remote sensing image analysis in our study gorithm parameters of Scale parameter, Shape, and Com- area proved to be challenged by the presence of multiple pactness to 10, 0.2, and 0.5, respectively, and ignored the land use types with similar spectral characteristics, and differences of these parameters, which could improve the strong illumination effects from topography. We employed precision of the classification. In addition, we aggregated an object-oriented decision-tree method to establish five multiple types of natural forests, such as tropical rain forest layers designed to overcome these obstacles in accordance and tropical monsoon forest, and considered them collec- with decision-tree nodes. First, we selected Landsat TM im- tively as a same type of natural forest; therefore, some agery with an acquisition time from late fall to early spring, patches might be identified incorrectly. to accurately identify multiple types of forests. In southern China, the period from December through the following 3.2. The change of different forests during the period of 1988– January is a period with drought and low temperatures; the 2008 fact that different vegetation types respond differently to The total area of tropical forest in the study area in- drought and low temperature was a crucial factor in dis- creased during 1988–2008. The area of tropical forest cov- criminating land use types. Second, we established flow ered about 68.57% of the study area in 1988, and then cov-

ered about 71.89% of the study area in 2008, with the in- to 3.09% of economic forests were converted to other land crease approximately 8% over the period of 1988-2008. uses (Fig. 7). Natural tropical forest always covered the largest area The spatial conversion of forest was also seen because of land use type within the study area, increasing slightly of the transformation of different forest types. The remain- (6.65%) in the first decade (1988–1998), and then decreasing natural tropical forests tended to occur only at higher ing sharply (13.69%) in the second (1998–2008). Generally, elevations and on steeper mountaintops; i.e., the average the area of natural tropical forest tended to decrease, despite distances to farms and transportation corridors increased. obvious fluctuations during the study period. This occurred because some natural tropical forest at lower elevations, and with less steep slopes that were previously The area of economic forest obviously expanded from close to farms and transportation corridors, were trans- 1988 to 2008. Of the economic forest, rubber plantations formed into economic forest during 1988–2008. The in- covered the largest area, but decreased slightly (3.79%) creasing extent of economic forests might result in a loss during the economic recession of 1988–1998; then, it expe- in the value of ecosystem services [25]. Generally, forest rienced a rapid increase in conjunction with the renascent transition theory assumes that increasing forest cover indi- market economy during 1998–2008. A few scattered pulp cates that environmental conditions are improving, but the plantations existed prior to 1988. Pulp wood was planted increasing forest cover in our study area does not neces- widely after 1988, eventually covering about 0.7 × 104 ha. sarily mean that biodiversity and natural ecosystems are re- by 1998 and about 7.09 × 104 ha. by 2008 (Table 2). covering [26].

Landscape-scale transformation of various forest types TABLE 2 - Area of natural forest and economic forest during the pe- obviously occurred. From 1988 to 1998, natural forest in riod of 1988–2008 the piedmont was converted into scattered rubber planta- 4 Year Natural for- Rubber Pulp wood Total forest tion, ultimately covering 9.57 × 10 ha. by 1998 (Fig. 6). est plantation areas area During the next decade (1998–2008), an additional 9.72 × (104 ha.) (104 ha.) (104 ha.) (104 ha.) 104 ha. natural forest was converted into economic forest 1988 73.48 22.57 * 95.91 (rubber and pulp wood plantations), which led directly to a 1998 82.66 17.28 0.70 100.64 2008 63.29 35.91 7.09 107.29 corresponding decrease in the extent of natural tropical for- * Hainan Island had sporadic Eucalyptus in 1988, but Eucalyptus was est. At the same time, up to 6.26% of natural forest and up grown for timber, not for pulp

FIGURE 6 - The transformation of different forest types in 1988–1998

FIGURE 7 - The transformation of different forest types in 1998–2008

3.3. The response of forest transformation to economic devel- in rubber plantations until 1995 [27]. Thus, the conversion opment and protection policies of natural tropical forests into rubber plantations in the Regional market trends contribute to the conversion piedmont had been only sporadic at the period of 1988- from natural forests to economic forests (rubber, pulp for- 1998. However, the market demand for rubber increased ests etc.). Rubber and pulp plantations expanded because sharply after 1999, leading to higher prices (Fig. 8), which of new regulations and increased demand from consumers. encouraged smaller landholders to become deeply in- For example, the demand for rubber kept increasing slowly volved in rubber plantations after 1999. Therefore, the de- and this led to low prices between 1988 and 1998 (Fig. 7). mand for rubber on the open market might be an important In addition, most rubber plantations were planted on gov- factor driving the change of different forests types during ernment-operated farms, typically covering large areas, 1988–1998[28]. whereas small landholders were not involved significantly

FIGURE 8 - Chinese natural rubber demand and price changes

The Chinese Government has strongly encouraged the established nature reserves, and protection policies sup- planting of pulp plantations with the goal of developing a pressed the transformation process. Furthermore, the de- domestic wood pulp industry that could meet growing do- velopment of economic forests, such as rubber and pulp mestic demand. The government provided substantial cap- plantations, propelled the transformation between different ital subsidies in the form of low interest rate loans, dis- forest types, even in provincial reserves formally under the counted loans, and extended repayment periods for loans protection policies during 1998–2008. Generally, the spa- from state banks [29, 30]. tial extent of natural forests decreased significantly from Thus, new plantations of economic forests greatly re- 1998 to 2008. Although some protection efforts have been duced the available space for natural tropical forests; there- attempted, such as the "Natural Forest Protection Project," the fore, the area of natural tropical forest decreased by 2008. establishment of nature reserves, and the closing of hillsides Protection policies and the establishment of reserves de- to facilitate afforestation [31], our results suggest that some layed the transformation of some forests. In the early of these protection measures were implemented poorly. From 1950s, policies encouraged the establishment of rubber another viewpoint, economic development exerted an enor- plantations. Several nature reserves were established on mous influence on the tropical forests of Hainan Island and Hainan Island starting in 1976. Currently, nine national na- reduced the effects of these protection efforts. Extensive ar- ture reserves cover 11.20 × 104 ha., while provincial nature eas of natural tropical forest on Hainan Island have been con- reserves cover 13.84 × 104 ha.. Generally, both national na- verted into economic forest. Local politicians are evaluated ture reserves and provincial nature reserves limited the ex- strongly on their support of economic development. This pansion of rubber plantations. The transformation of rainfor- encourages them to support economic development at the ests to plantation forests was suppressed by these protection expense of environmental protection. Changing policies re- policies, and especially by the existence of natural reserves lated to local government officials could increase protec- during 1988–1998 (Table 3). tion for natural landscapes. To resolve the conflict between development and pro- TABLE 3 - The percentage of plantation forests in different natural tection, a balance between economic development and forest reserves in 1988–2008 protection needs to be found based on monitoring data [32].

Year The establishment of ecological compensation mechanisms 1988 1998 2008 might contribute to providing local communities with some National reserves 16.31% 10.96% 7.22% economic benefits in exchange for forest protection [33]. Provincial reserves 14.24% 8.80% 16.77% 3.4. Response of forest transformation to environmental factors Different types of reserves provide different levels of protection. The national nature reserves play important Elevation, slope, farmlands, and transportation corri- roles in protecting natural forest. During 1988–2008, in the dors were selected for use in the analysis of their spatial national reserves, the area of natural forest increased stead- relationships with each other and with forest transfor- ily while the area of economic forest decreased slightly. mation. Economic forests (rubber forest, pulp forest, and However, provincial natural reserves play an ineffective non-forest) were observed to occur mainly near farms and role in protecting natural forest, because local and provin- transportation corridors usually in the 1980s, but natural cial government agencies support economic growth over forests occurred further away from farmland and those cor- the conservation of natural resources. Therefore, some nat- ridors after the 1980s. Thus, because of the spatial transfor- ural forests in provincial reserves were replaced by eco- mation of forests during 1988–2008, on average, natural for- nomic forests during the same period (Table 3). est shifted to higher elevations with greater slopes during this period. Similarly, at the same time, the average locations Generally, the main forces driving the transformation of economic forest also shifted to higher elevations (Table of forest types seem to be protection policies and economic 4) (Fig. 9). In general, the locations of trans- development. Some natural forests were protected by the

TABLE 4 - Areas of forest transformation and the location (elevation and slope) where forest transformation occurred

1988–1998 1998–2008 Area Area Transformation of forests Mean slope Mean elevation (m) Mean slope Mean elevation (m) (ha. 104) (ha. 104) Natural to Non-natural forest 3.07 15.70 345.57 22.79 17.42 358.12 Non-natural to Natural forest 12.17 17.33 375.33 3.35 14.65 372.13 Rubber to Non-rubber forest 6.56 11.26 200.64 3.07 10.31 199.46 Non-rubber to Rubber forest 1.26 8.55 187.31 22.73 11.77 245.91 Non-pulp to Pulp forest 0.7 13.71 303.30 6.38 16.77 397.32 Non-forest to Forest 11.06 16.52 382.29 17.3 10.01 263.78 Forest to Non-forest 6.56 11.20 242.25 10.7 16.66 370.67

Natural forest Rubber woods Pulp woods SLope(degree)

Years

Rubber woods Pulp woods (km)

line

triffic

form

Distance

Years

FIGURE 9 - The response of forest transformation to slope and ele- FIGURE 10 - The response of economic forests transformation to dis- vation tance of farmland and transportation corridors

Table 5 Stepwise regression analysis between environmental factors and transformation of different forests. The relationship was showed between the area transformation occurred and the altitude, slope, and the minimum distance from the locations transformation occurred to farmland and road in 1988-1998

The minimum distance from a cell to land use type in 1988–1998 Transformation of forest type Farmland Road Altitude Slope Natural to rubber plantation – – – – Natural to pulp plantation – – – – Natural to non-forest – – – – Rubber to natural forest – – – – Rubber to pulp plantation – – – – Rubber to non-forest – – – – Non-forest to natural forest – – 0.62 – Non-forest to rubber plantation – – – – Non-forest to pulp plantation – – 0.59 0.60 Note: the number is the coefficient of stepwise regression, which gains significant at 5%

TABLE 6 - Stepwise regression analysis between environmental factors and transformation of different forests. The relationship was showed between the area transformation occurred and the altitude, slope, and the minimum distance from the locations transformation occurred to farmland and road in 1998-2008

The minimum distance from a cell to land use type in 1998–2008 Transformation of forest type Farmland Road Altitude Slope Natural to rubber plantation 0.55 0.49 0.52 – Natural to pulp plantation 0.57 0.47 0.53 0.50 Natural to non-forest – – – – Rubber to natural forest – – – – Rubber to pulp plantation – – – – Rubber to non-forest – – – – Non-forest to natural forest – – – – Non-forest to rubber plantation 0.52 0.53 – – Non-forest to pulp plantation – 0.51 – 0.54 Note: the number is coefficient of stepwise regression, which gains prominence at 5%. The symbol “-” not significant at 5%

portation corridors and farmland, together with slope and el- nomic forests, resulting in an expansion of economic for- evation, were the most important intrinsic dynamic factors ests (rubber and pulp plantations). Forests also shifted spa- affecting the transformation of forest types (Fig. 10). tially during 1998–2008. Elimination of some natural tropical forest in the lower piedmont, due to economic devel- Approximately 300 sites were samples in the study opment, caused the average location of the remaining trop- area to collect the transformation area and its distance to ical forest to become located at higher altitudes. The trans- environmental factors (farmland, road), altitude and slope. formation of forests was driven mainly by protection poli- Stepwise regression analysis confirmed that terrain factors cies and economic development, but economic develop- (elevation and slope) affect forest transformation. Trans- ment exerted a much stronger influence and minimized the formations between non-forest land and pulp plantations protection efforts. Transportation corridors, farmland, were observed to be associated with elevation and slope, slope, and elevation were also important intrinsic dynamics and transformations between non-forest and natural forest that affected the transformation of tropical forests. were associated with elevation in 1988–1998 (Table 5). Similar discoveries suggest that transformations between natural forest and rubber plantations were associated with ACKNOWLEDGMENTS elevation, and transformations between natural forest and pulp plantations were associated with elevation and slope Grants from the National High Technology Research in 1998–2008 (Table 6). Therefore, it could suggest that the Development Plan (Grant No. 2012AA12A309), the Na- transformations between different forests types might be tional Natural Science Foundation of China (Grant No. affected by altitude, slope, road and farmland. 41371359) and the Education Department of Shaanxi Prov- ince Special Research Project（）12JK0484 supported this work. In addition, the authors thank the anonymous re- 4. CONCLUSIONS viewers for their helpful and constructive feedback.

Our results emphasize that forest monitoring could in- The authors have declared no conflict of interest. corporate remote sensing imagery of specific periods and an object-based decision-tree method to identify forest change. The object-based information extraction based on REFERENCES the decision-tree method proved applicable and effective in [1] Situ S. J. (1987). The history of Hainan Island land develop- identifying the main forests with complex surface features ment research. Haikou: Hainan people press. in a tropical evergreen ecosystem. Based on the results of [2] Jodi S, Kuemmerle T and Li H. M. (2012). Using Landsat im- forest information extraction, it is determined that the trop- agery to map forest change in southwest China in response to ical forests of the central Hainan Island suffered an obvious the national logging ban and ecotourism development. Remote change during 1988–2008. Initially, the spatial extent of Sensing of Environment, 121, 358–369. natural tropical forest increased slightly, and then experi- [3] Rohan F. (2012). Tropical forest monitoring, combining satel- enced an obvious decrease during 1998–2008. Larger lite and social data, to inform management and livelihood im- amounts of natural tropical forest were replaced by eco- plications: Case studies from Indonesian West Timor. Interna- tional Journal of Applied Earth Observation and Geoinfor- mation, 16, 77–84.

[4] Zhai D. L, Charles H and Cannon J.W. (2012). Rubber and [21] Xu, H.Q.（ 2006） . Modification of normalized difference wa- pulp plantations represent a double threat to Hainan’s natural ter index (NDWI) to enhance open water features in remotely tropical forests. Journal of Environmental Management, 96, sensed imagery. Internal Journal of Remote Sensing, 64-73. 27（）14 , 3025 - 3033. [5] Oberling, D.F., La Rovere, E.L. de Oliveira Silva, H.V. (2013) [22] Zhang L. F, Furumi S, Murumatsu K, Fujiwara N, Daigo M, SEA making inroads in land-use planning in Brazil: The case and Zhang L. P (2007). A new vegetation index based on the of the Extreme South of Bahia with forestry and biofuels. Land universal pattern decomposition method. International Journal Use Policy, 35, 341– 358 of Remote Sensing, 28(1), 107-124. [6] Wang, S. D. (2011). Ecosystem patterns and service changes [23] Zhao Y.S. (2006).The principle and method of analysis of re- of Hainan Island, China. Beijing: Research Center for Eco- mote sensing application. Science Press. Beijing. Environmental Sciences, Chinese Academy of Sciences. [24] Webb C.O. (2000). Exploring the phylogenetic structure of [7] Zhang M.X, Fellowes J.R and Jiang X.L. (2010). Degradation ecological communities: an example for rain forest trees. Am. of tropical forest in Hainan, China, 1991–2008: Conservation Nat. 156, 145-155. implications for Hainan Gibbon. Biological Conservation, [25] Jose S. (2012). Agroforestry for conserving and enhancing bi- 143, 1397–1404. odiversity. Agroforest System, 85, 1–8. [8] Ouyang Z. Y, Jin Y and Zhao T. Q. (2011). Ecosystem regu- [26] A.S. Mather, J. Fairbairn and C.L. Needle (1999). The course lating services and their valuation of Hainan Island. Journal of and drivers of the forest transition: The case of France. Jour- Resources and Ecology, 2(2), 132-140. nal of Rural Studies 15: 650–90. [9] Huang B. R and Ouyang Z. Y. (2009). Assessment of eco-en- [27] Umezaki M and Jiang H.W.(2009). Changing adaptive strate- vironmental vulnerability of Hainan Island, China. Chinese gies of two Liethnic minority villages in a mountainous region Journal of Applied Ecology, 20(3), 639-646. of Hainan Island, China. Southeast Asia Stud. 47, 348-362. [10] Li W.H and Zhang B. (2009). Research on ecosystem services [28] Jin Y. (2006). Ecological Assessment and Ecological Function in China: Progress and perspectives. Journal of Natural Re- Regionalization of Hainan Province. Beijing: Research Center sources, 24(1), 1-10. for Eco-Environmental Sciences, Chinese Academy of Sci- ences. [11] Huang J. C. (2007). Study on Sustainable Management of [29] Barr C and Cossalter C. (2004). China’s development of a Tropical Forest in Hainan Island, China. Beijing: Chinese plantation-based wood pulp industry: government policies, Academy of Forest. ﬁnancial incentives, and investment trends. Int. Forest. Rev. 6, [12] Yang S, Feng X. M and Chen L. D. (2009). Spatial temporal 267-281. differentiation and mechanism of land use / cover change: a [30] Barr C and Cossalter C. (2006). How Cost Competitive is case study of Haidian and Yanqing Districts, Beijing. Acta Wood Pulp Production in South China? Presentation Prepared Ecologica Sinica, 29(8), 4501-4511. for Seminar on Forest Trade and Livelihoods, Beijing, 20-21. [13] Zuo L.J and Xu J.Y. (2011). Spatial-temporal land use change [31] Wu S.H. (2000). Protection and development of Hainan tropi- and landscape response in Bohai Sea coastal zone area. Jour- cal forest. Tropical Forestry, 28(2): 41-44. nal of Remote Sensing, 15(3), 604-611. [33] Ying Z, Gao M.X and Liu J. C. (2011). Green accounting for [14] Wang S. D, Ouyang Z. Y, Zhang C.P, Xu W.H and Xiao Y. forest and green policies in China-A pilot national assesment. (2012). The dynamics of spatial and temporal changes to for- Forest Policy and Economics, 13(7), 513-519. ested land and key factors land driving change on Hainan Is- land. Acta Ecologica Sinica, 32(23), 7364-7374. [32] Zhao X, Zhou D and Fang J. Y. (2012). Satellite-based stuies on large-scale vegetation changes in China. Journal of Inte- [15] Sieber A, Kuemmerle T and Prishchepov A.V. (2013). Land- grative Plant Biology, 54(10), 713-728. sat-based mapping of post-Soviet land-use change to assess

the effectiveness of the Oksky and Mordovsky protected areas in European Russia. Remote Sensing of Environment, 133, 38–51.

[16] Brandt G.S, Kuemmerle T and Li H.M. (2012) Using Landsat imagery to map forest change in southwest China in response Received: January 23, 2014 to the national logging ban and ecotourism development. Re- Revised: June 05, 2014; June 26, 2014 mote Sensing of Environment, 121, 358–369 Accepted: July 02, 2014 [17] Vyas D, Krishnayya N.S.R and Manjunath K.R. (2011). Eval- uation of classifiers for processing Hyperion (EO-1) data of tropical vegetation. International Journal of Applied Earth Ob- servation and Geoinformation, 13, 228–235. CORRESPONDING AUTHOR [18] Pan J.H and Li T.Y. (2013). Extracting desertification from Shudong Wang Landsat TM imagery based on spectral mixture analysis and Albedo-Vegetation feature space. Natural Hazards, 68, 915– State Key Laboratory of Remote Sensing Science 927. Institute of Remote Sensing and Digital Earth (RADI) CAS [19] Li X.J, Meng Q.Y and Gu X.F. (2013). A hybrid method combining pixel-based and object-oriented methods and its appli- Beijing 100101 cation in Hungary using Chinese HJ-1 satellite images. Inter- P.R. CHINA national Journal of Remote Sensing, 34(13), 4655–4668. [20] Chen Y.Z, Lin Q.H, Hua Y.G, Luo W and Lin Z.M. (2010). E-mail: [email protected] The analysis of hyperspectral characteristics of rubber leaves. Journal of Tropical and Subtropical Botany, 18(3), 293-297. FEB/ Vol 24/ No 1b/ 2015 – pages 394 – 404

A REVIEW OF THE PROGRESS IN CHINA’S RENEWABLE ENERGY GENERATION, AND ITS LINKS WITH ENERGY CONSERVATION, ENVIRONMENTAL DEGRADATION AND ECONOMIC GROWTH

Qianyu Dong* and Tohru Futawatari

Faculty of Environmental Engineering, The University of Kitakyushu, Fukuoka, Japan

ABSTRACT RE is generally defined as energy that comes from resources which is naturally replenished on a human time- China is under immense pressure of economic growth, scale. RE generation mainly includes hydropower, solar energy consumption and environmental crisis. Meanwhile, (PV) power, wind power, biomass energy and geothermal renewable energy is regarded as a key solution for tackling energy. In the year of 2012, China’s RE generation capacity those issues. However, is the renewable energy really ef- reached 1010 TWh, which ranked the second of the world fective to break through the dilemma as expected? The an- [1], and represented 20.2% of China’s annual total electric- swer is not clear yet. Therefore, in this paper, we elaborated ity generations [2]. the current status and trends of renewable energy genera- China’s aspiring RE development has attracted in- tion in China both from the installed capacity and genera- creasing attentions, a mass of studies have conducted re- tion by sources. Then, based on a correlation analysis, we lated studies, including conducting renewable case studies found the development of renewable energy generation has in a region or province [3]; analyzing the interactions be- a greater correlation with GDP growth while being less re- tween RE development and environment [4]; reviewing the lated with fossil fuel electricity consumption decrease and barriers and opportunities of RE [5, 6]; circumstantiating CO2 emissions reduction. Accordingly, we considered that the development situation for a certain kind of RE. With so to improve energy efficacy of fossil fuel power will be more many studies focused on a certain technology aspect of RE helpful for China’s sustainable development rather than pro- in China, only few of them have invested the progress and moting renewable energy generation in the short term. At problems from the aspects of installed capacity and gener- last, some suggestions for China’s renewable energy generation. Moreover, rare researches discussed about that ation and sustainable development were presented. whether the advance of RE generation really correlated with the country’s sustainable development as expected or not. As

a result, this study aims to 1) analyze the reasons and prob- KEYWORDS: Renewable energy; generation; GDP; CO2 emis- lems for development of RE generation; 2) discuss the cor- sions; sustainable development relations of RE generation with CO2 emissions reduction, fossil fuels electricity consumption and economic growth; 3) give recommendations for the future development of RE generation industry. 1. INTRODUCTION

As the largest developing country in the world, China not 2. STATUS QUO ANALYSIS only faces the problem of maintaining a fast and steady eco- OF REGENERATION IN CHINA nomic growth, but also a number of energy-related challenges such as shortage of electricity supply and environmen- 2.1. The Distribution of RE Plants tal pollution. Under this circumstance, renewable energy In contrast to the fossil fuel plants, RE power plants (RE) as a kind of alternative energy, which does not harm the are located in the resourceful areas since RE resource is not environment and could meet the power demand, has been able to be transported. Figure 1 presents the top 10 prov- placed high hopes for Chinese sustainable development. inces with RE installed capacities in 2012. Clearly, most of these power plants are centralized in the central developing * Corresponding author regions of China. However, these resource-rich regions only

consume less electricity. Meanwhile, as shown in Fig. 2, the ecological influence of hydropower exploitation. Even some developed regions (such as northeast and east coastal though because of the heavy pressure of reducing CO2 areas) with higher power consumptions, rely to a greater emissions, they restarted the approval again in 2010, but extent on fossil fuels electricity. Therefore, there is a large the growth speed is slower than before. And the new con- spatial disparity between RE resources and load centers. structions are focused on small size hydropower stations. Furthermore, the lack of greater interconnection among re- Wind power, as shown in Table 1, replaced hydro- gional and sub-regional grids imposes constraints on opti- power to own the biggest newly installed capacity from mal use and delivery of energy. 2010, and it enjoyed the highest growth speed till 2009. We considered the reason can be attributed to meet the rising 2.2. Installed capacity and electricity production of RE power demands in north-east industrial areas and devel- Figure 3 describes the trends of installed capacity of oped eastern coastal areas (they are wind resource-rich RE by sources from 2000 to 2012. Table 1 lists the growth zones, as shown in Fig. 1). China turned to focus on enlarg- rate of installed capacity of each renewable, respectively. ing the scale of wind power plants in these regions since Clearly, hydropower has the whip hand of accumulative in- 2000. And a plenty of market-oriented policies has been set stalled capacity, but compared with other renewables, its out. For example, the concession bidding of wind power growth rate is the lowest. The root reason should be that was implemented between 2003 and 2007. It aimed at pro- the growth of hydropower is from a higher baseline. More- moting the competitive mechanism and a reasonable feed- over, China had stopped the approval of new large-scale in tariff of wind power, and urged the multiply increase in hydropower plants in 2009, for the heated controversy about installed capacity of wind power.

Hydro Wind PV Biomass Rank Province Total power Power power power ① Sichuan 39640 20 - 116 39780 ② Hubei 35950 170 12 406 36540 ③ Yunnan 33060 1310 30 122 34520 Inner Mongo- ④ 1080 16930 210 151 18370 lia ⑤ Guizhou 17280 960 - - 18240 ⑥ Guangxi 15360 100 - 60 15520 ⑦ Guangdong 13060 1390 8 439 14900 ⑧ Hunan 13720 190 - 218 14130 ⑨ Gansu 7300 5970 380 1 13650

⑩ Fujian 11400 1130 1 208 12740

Source:2012 China Renewable Energy Outline by China National Renewable Energy Center (CNREC)

FIGURE 1 - The provinces with top 10 RE power installed capacity in 2012 (MW).

Source: Ten cities and thousand vehicles program FIGURE 2 - Generation mix of the six interprovincial power grids.

From 2010, PV has displaced the wind power as the Sun’ was launched from 2009 to 2013, which aimed to in- fastest growing RE (Table 1). This amazing growth could crease 500 MW for PV installed capacity. be attributed to three reasons. Firstly, unlike hydropower, In the field of electricity production, Fig. 4 depicts the the development of PV is from a pretty low baseline. Con- proportion of RE in total power generation and renewables sequently, it is easier to achieve a high growth rate. Sec- generation by sources from 2001 to 2012. Table 2 tabulates ondly, the growth is mainly pushed by China’s manufac- the annual growth rate of RE generation during this period. turing industry of solar cells. Actually, since China has be- Similar to the status in installed capacity, hydropower has come the biggest PV cell producer of the world in 2007, its the highest generation but lowest growth rate. Besides, heavy dependency on oversea market has given rise to the since hydropower is susceptible to water resources, which appeal of expanding the domestic PV market. Meanwhile, not only made the hydropower generation fluctuating, but the declining demands of oversea market also forced China also greatly affected the ratio of RE power in China’s gross to explore the domestic market to solve excess capacity. generation. Thirdly, the national PV support project such as ‘Golden

Source:2012 China Renewable Energy Outline by China National Renewable Energy Center (CNREC)

FIGURE 3 - Installed capacity of RE in China from 2000 to 2012(GW).

TABLE 1 - Growth rate of installed capacity of RE in China (%).

2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Hydro 3.1 4.6 3.7 25.8 8.4 11.0 13.8 16.4 13.7 10.1 7.8 6.8 Wind 12.1 17.6 21.9 36.1 69.6 106.3 61.6 99.8 109.8 68.1 56.3 35.5 PV ------766.7 753.5 47.8 Biomass 0.0 0.0 50.0 0.0 33.3 25.0 20.0 9.0 40.7 19.6 27.3 14.3

Source:2012 China Renewable Energy Outline by China National Renewable Energy Center (CNREC) FIGURE 4 - Electricity generation of RE in China from 2000 to2012 (TWh).

TABLE 2 - Growth rate of RE generation in China (%).

2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Hydro 24.7 3.8 -1.5 24.6 12.1 4.6 13.6 20.0 1.1 20.1 -2.7 29.3 Wind 12.1 17.6 21.9 36.1 69.6 106.3 125.3 129.8 110.8 78.9 50.0 36.0 PV ------600.0 400.0 Biomass 0.0 0.0 50.0 0.0 57.6 34.6 39.1 51.1 40.7 19.6 27.3 20.6

Wind power keeps a stable increase. PV power has an On account of the limited data about RE generation in explosive growth based on a low baseline. Particularly, ap- China, we collected yearly data from 2000 to 2012 as the proximately 72% of PV power came from centralized PV samples. The data mainly come from China National Re- stations [2]. This is different with PV power well-devel- newable Energy Center (CNREC), the U.S. Energy Infor- oped countries, where distributed PV system is the main- mation Administration (EIA), International Energy stream. However, a distributed power market requires to Agency (IEA), China Energy Statistics Yearbook, World break China’s electricity monopoly, and it is pretty difficult Development Indicators (WDI), and the World Bank Date- to be finished in a short time. book etc. In sum, China’s progress in RE generation is largely Firstly, we used scatter plot to investigate the possible led by hydropower and wind power, and government plays relationship between two variables. If they are related, an important role. Meanwhile, the domestic market-driven then, we adopt Spearman's rank-order correlation to assess mechanism is forming, and the links between RE genera- how well the relationship can be described by a correlation tion and external conditions are stronger. But, electricity coefficient (rs). Spearman's rank correlation coefficient (rs) monopoly has blocked the further growth of RE genera- always gives an answer between −1 and 1. The numbers are tion. like a scale, where −1 is a negative correlation, 0 is no link, and 1 is a positive correlation; if |rs| is closer to 1, the cor- 2.3 Problems for the development of RE generation relation is stronger. The p-value is a measure of the proba- Based on Figs. 3 and 4, we observed that the share of bility of obtaining a result at least as extreme and as the one RE is close to one third of the total installed capacity, but that is actually observed; so the lower the value (usually the share of RE in gross generation is only one fifth. This below 0.05 or 0.01) the more significant the result [9]. problem might in part because RE is susceptible to weather conditions. But, the figure revealed a more serious problem 3.1 RE electricity and CO2 emissions reduction that is ‘Blindly expand’ of RE generation. It means that the According with the scatter plot of RE generation and installed capacity of RE is already considerably larger than CO2 emissions from electricity production shown in Fig. 5 the output capacity of power grid. Take wind power as an (a), we found that the two variables are positively corre- example, we found a phenomenon called ‘wind curtail- lated. Then, based on correlation coefficient (rs) in Table 3, ment’ which has already happened in Inner Mongolia. The it further proved the RE generation is strongly positive cor- abandon rate of China’s wind power was 17% in 2011, that related with CO2 emissions from generation (rs＞0.9, p- means more than 120 GW of wind power was wasted [7]. value＜0.05). This result revealed that CO2 emissions in- The reasons could not only be the low utilization level, but crease while the RE generation grows, which also is sug- also the unplanned exploration. If the grid is not capable gesting that the development of RE generation is not sig- enough to accept the increase of RE electricity in a time, nificantly correlated with CO2 emissions reduction. More- instead of accelerating the installed capacity, they should over, Fig. 5 (b) demonstrated that the ratio of RE in gross focus on updating the grid and related management sys- generation is uncorrelated with CO2 emissions from elec- tems. tricity production; this is suggesting that, no matter the proportion of RE increase or decrease, there is no certain change to the CO2 emissions from electricity production. 3. CORRELATION ANALYSIS OF REGENERA- In other words, even if the proportion of RE in total power TION AND SUSTAINABLE DEVELOPMENT IN generation increases, the proportion of CO2 emissions from CHINA electricity production might not decrease.

Without doubt, the national energy system is compli- 3.2 RE electricity and fossil fuels electricity cated and interlocking. However, to some extent, the cor- As shown in Fig. 5 (c), RE electricity and fossil fuel relations among the indicators and phenomena are not hard electricity are strongly positively correlated. However, as to find. Hence, in this paper, we focus on exploring the rea- shown in Fig. 5(d) and Table 3, the growth rate of RE and sons behind the links, and find realistic solutions for the fossil fuel electricity is significantly and negatively corre- future development. lated (rs＜0, p-value＜0.05), suggesting that, as the RE

(a) RE generation and CO2 emissions from electricity production (b) Ratio of RE in total generation and CO2 emissions from electricity production

(c) RE generation and fossil fuel electricity (d) Growth rate of RE generation and growth rate of fossil fuel electricity

(e) RE generation with GDP (f) RE generation with GDP growth rate

FIGURE 5 - Scatter plot of RE generation with CO2 emissions, fossil fuel electricity and economic growth.

TABLE 3 - Correlations between RE generation and variables of CO2 emissions, fossil fuel electricity and GDP.

generation accelerated, the growth rate of fossil fuel elec- well whereas Fig. 5(f) indicates that RE generation is un- tricity decreased. correlated with real GDP growth rate.

3.3 RE generation and economic growth In brief, RE generation has a greater correlation with Based on Fig. 5(e) and Table 3, RE generation is sig- GDP growth while being less related with fossil fuel elec- nificantly positive correlated with GDP (rs ＞0.9, p＜0.01), tricity consumption decrease and CO2 emissions reduction. suggesting that, as RE generations increase, GDP grows as

4. DISCUSSION new peak of RE generation was reached in 2012, its share is not stable and big enough to contend with coal power. China is on the developing period of industrialization, Thus, to reduce CO2 emissions in China, the more im- and its economic growth is driven by large-scale produc- portant should be increasing energy efficiency. tion. But scale economy not only comes with large sum of Based on the problems we explored, some recommen- energy consumption but also the by-products, such as en- dations are given as follows: Firstly, China should focus on vironmental damages. And RE is used to be regarded as the increasing energy efficiency of coal power. Because of big- key to address this dilemma. However, in this study, we ger base number, the efficiency of thermal power plants found that RE generation has a greater correlation with will be more significant on CO2 emissions reduction. Sec- GDP growth while being less related with fossil fuel elec- ondly, transmission capacity should be enlarged prior to tricity consumption decrease and CO2 emissions reduction. explore installed capacity of RE. If only with a huge in- We discussed this as follows: stalled capacity, but a laggard power transmission system, Firstly, the enormous investment on RE has become an it means more RE generation will be wasted. Thirdly, the important driving force of China’s GDP. China’s invest- policy planning and decision making mechanism should be ment at RE has reached 66.6 billion dollars in 2012 [8], completed. Fourthly, the concession bidding system in all which occupied 0.81% of GDP. The huge investment not kinds of RE generation industry should be expanded be- only greatly promoted RE generation, but also stimulated cause this system is proved by the wind power department related industries (such as the raw material supply and which could act as a buffer between competitive mecha- manufacturing industries of RE plants, the technology nism and policy-oriented industry. Fifthly, the distributed companies). For another perspective, just like we discussed RE power market must be opened up to break the electric- previously in section 2, the over fast growth of RE installed ity monopoly. In addition, a comprehensive and objective capacity might be used as a way to boost economic growth understanding about the influence of hydropower should of China. Besides, in order to achieve the target of ‘RE ac- be conducted both for the public and policy makers. Lastly, counts for 30% of total installed capacity in 2020’ set by the loss in GDP for decelerating new construction of RE National Development and Reform Commission, we be- could be shifted to the promotion of power system and en- lieve the installed capacity of RE will keep growing. But, ergy efficiency. So, even the growth in installed capacity the gap between RE installed capacity and generation of RE is slowed down, the influence on economy can be should be attached more importance. After all, the electric- minimized. ity production by RE is more important than inapplicable huge installed capacity. Secondly, the ever-increasing power demand has im- 5. CONCLUSION proved the growth of both RE electricity and fossil fuels electricity, but the growth of fossil fuels electricity is by a At present, RE generation is hard to drive a fundamen- faster speed and greater capacity. RE electricity is hard to tal transition in China’s energy system because its limited challenge the fossil fuel electricity in a short term since fos- effects on CO2 emissions reduction and fossil fuel electricity decrease. If China expects to expand the RE’s influence, sil fuel electricity has dominated China’s energy system for it will require sharply limiting the use of fossil fuels on decades. Besides, RE generation is easy to be affected by which more than 80 % of today’s Chinese energy system weather conditions generation (especially hydropower), depends. And that, in turn, would set off a sweeping tran- whereas the fossil fuel electricity could be transported and sition of one of the most extensive technological com- change with the insufficient any time. Moreover, the cost plexes and deeply embedded elements of China’s physical of coal power is much cheaper than RE generation, and for infrastructure: the national energy system, economic growth the grid company, to invest on thermal power plants is and industrial structure. But these transformations are in- more cost-effective. As a result, RE’s effect on reducing credibly difficult to be finished in a decade. Therefore, fossil fuels electricity is not significant. China will necessarily continue to rely on fossil fuels in the Thirdly, the CO2 emission is positively correlated with coming decades. And during this period, compared with a the fossil fuel electricity consumption. China’s RE genera- bigger ratio of RE in generation, to raise the energy effi- tion has experienced a rapid expansion since 2000, but that ciency of fossil fuels is more useful and practical for is from a lower baseline. Meanwhile, the fossil fuels elec- China’s sustainable development. tricity climbs up as well by a faster speed and greater capacity than the RE electricity. Consequently, even though The authors have declared no conflict of interest. RE consumption and generation keeps fast growing, this growth in RE generation was more than offset by a 6.4% increase in coal consumption [1], which has a higher base- REFERENCES line while most of CO2 emission is attributed to the low- efficient coal power plants. Moreover, the fluctuating [1]. PBL Netherlands Environmental Assessment Agency, (2014) growth rate of RE generation leads to a floating ratio of RE Trends in Global CO2 Emissions 2013 Report. The Nether- generation in gross electricity generation. So, although a lands.

[2]. China National Renewable Energy Centre (CNREC) (2013) China Renewables and Non-fossil Energy Utilization. Beijing, China,

[3]. Hong L, Lund H, Mathiesen BV, Möller B. (2013) 2050 path way to an active renewable energy scenario for Jiangsu province. Energy Policy, Vol. 53, 267-278

[4]. Liu T, Xu G, Cai P, Tian L, Huang Q. (2011) Development forecast of renewable energy power generation in China and its influence on the GHG control strategy of the country. Re- newable Energy, Vol. 36, 1284-1292

[5]. Wang Q, Chen Y. (2010) Barriers and opportunities of using the clean development mechanism to advance renewable energy development in China. Renewable Sustainable Energy Reviews, Vol. 34, 1989-1998 [6]. Ma HY, Oxley L, Gibson J, et al. (2010) A survey of China′s renewable energy economy. Renewable and Sustainable En- ergy Reviews, Vol.14, 438-445 [7]. State Electricity Regulatory Commission (SERC), (2012) Re- search Report on the Wind Power Development in China. Bei- jing, China. [8]. China National Renewable Energy Center (CNREC), (2012) China Renewable Energy Outline. Beijing, China. [9]. Field, A.. (2010) Discovering Statistics using SPSS, 3rd ed.: Sage Publications: London, England.

Received: May 12, 2014 Revised: June 30, 2014 Accepted: July 04, 2014

CORRESPONDING AUTHOR

Qianyu Dong

Faculty of Environmental Engineering

The University of Kitakyushu

Fukuoka

JAPAN

E-mail: [email protected]

FEB/ Vol 24/ No 1b/ 2015 – pages 405 - 411

SUBJECT INDEX

A I adsorption kinetics 302 intermittent aeration 371 adsorption 291 Izmir 317 algae 324 allelopathy 386 L Alnus glutinosa 335 larvicide 311 lipid peroxidation 386 B bioleaching 379 M biomass 317 maize 278 biosorption 302 MDA 365 branch 335 mercury (II) 324 meristic 263 C microbial community structure 371 cadmium removal 302 microfungus 317 CO2 emissions 405 morphometric 263 constructed wetland substrates 291 mosquito 311 Cucurbita pepo 386 multiple forests transition 394 cysteine 379 N D North-eastern Mediterranean 263 DNA fingerprinting 343 Notonecta 311 dust retention effect 285 O E oasis cities 285 ecotoxicology 343 object oriented decision tree technique 394 enzyme activities 386 enzyme activity 371 P estrone 291 photoreduction 324 photosynthetic bacteria 355 F photosynthetic pigment 278 fiber diameter 335 phytoplankton 355 fiber length 335 picocyanobacteria 355 fiber lumen diameter 335 picoeukaryotes 355 Forest management 394 population structure 263 Fourier transform infrared spectroscopy 379 Populus alba var. pyramidalis (Bunge) 285 fungi 317 Portulaca oleracea 386 fungus 317 proline 278 protein profile 343 G proteomics 343 GDP 405 pyriproxyfen 278 generation 405 GIS 394 R gold ores 379 Rainbow trout 365 Greece 269 Red barracuda 263 green-belt conservation 269 remote sensing 269 remote sensing 394 H renewable energy 405 heavy metal concentrations 285 responses 371 hospital 317 root 335

I S impact factor 394 Scanning electron microscope/ 379 insecticide 278 Energy-dispersive X-ray spectroscopy integrated vertical-flow constructed wetland 371 (SEM/EDS)

AUTHOR INDEX

S A stem 335 Abaci Gunyar, Ozlem 317 stomatal index 278 Abd El-Gawad, Ahmed M. 386 suspensions 324 Abliz, Abdulla 285 sustainable development 405 Aishan, Tayierjiang 285 Alatar, Abdulrahman A. 343 T Alaydin, Elif 302 thorium 343 Alemdar, Sonay 302 toxicity 311 Altin, Ahmet 302 trans-boundary river 269 Altin, Sureyya 302 Turkey 269 Ates, Mustafa 317

U B uranium 343 Baidourela, Aliya 285

V C

vertical distribution 355 Cetin, Huseyin 311 vitamin A 365 Cevher Özeren, Saniye 269 vitamin C 365 Çilingir Yeltekin, Asli 365 vitamin E 365 Coskun, Yasemin 278

X D X-ray diffraction (XRD ) 379 Dai, Bao-Lin 291 Demir, Halit 365 Deng, Lin 324 Deng, Nansheng 324

Dimitriou, Elias 269

Dong, Qianyu 405

El-Shora, Hamed M. 386 Elyas, Aygvl 285 Emam, Mona H. 343 Ezgi Duran, Ragbet 278

F Faisal, Mohammad 343 Futawatari, Tohru 405

G Gao, Yun-Chen 291 Goker, Erdem 317 Gurlek, Mevlut 263

H Hakkı Yörük, İbrahim 365 Halik, Ümüt 285 Haliki-Uztan, Alev 317 He, Feng 371 Hegazy, Ahmad K. 343 Huang, Shi-Bei 291

J Jakubowska, Natalia 355 Ji, Yuhe 394

J Z Jia, Liu 394 Zhang, Chen 379 Zhang, Lifu 394 K Zheng, Hongai 379 Kabiel, Hanan F. 343 Zheng, Hongai 379 Kahraman, Bekir Fatih 302 Zogaris, Stamatis 269 Kaźmierska, Agnieszka 355 Kiaei, Majid 335 Kilic, Semra 278

L Li, Bingyue 379 Li, Sheng Fang 371 Liang, Lu 291 Liu, Jianshe 379

M Markogianni, Vassiliki 269 Moya, Roger 335

O Ouyang, Zhiyun 394

S Ser, Onder 311 Shi, Jincheng 324 Shi, Penghui 379 Song, Hai-Liang 291 Szeląg-Wasielewska, Elżbieta 355

T Tao, Min 371 Turan, Cemal 263

W Wang, Qi Shuo 371 Wang, Shudong 394 Wang, Xuejuan 379 Wang, Zhonghong 379 Wu, Yang 379 Wu, Zhen-Bin 371

X Xie, Xuehui 379 Xuejuan Wang 379

Y Yaglioglu, Deniz 263 Yang, Caiqian 324 Yang, Xiao-Li 291 Yoltas, Aysegul 317

Z Zhang, Chen 379

414