FORESTRY IDEAS, 2017, vol. 23, No 1 (53): 65–76 PRODUCTION AND MASTING BEHAVIOUR IN ORIENTAL (FAGUS ORIENTALIS LIPSKY) FORESTS OF NORTHERN

Vahid Etemad1 and Kiomars Sefidi2* 1Faculty of Natural Resources, University of Tehran, Karaj, Iran. E-mail: [email protected] 2Faculty of Agriculture and Natural resources, University of Mohaghegh Ardabili, Ardabil, Iran. *E-mail: [email protected]

Received: 23 October 2016 Accepted: 30 July 2017

Abstract Mast-seeding behaviour of species has a key role in regeneration of populations. Seed production data were collected from 360 sampling plots to identify environmental factors influencing beech masting in the old-growth beech stands in the north of Iran. The mean seed number per m2 was 4687 ± 164 in the study sites. Seed production significantly correlated with tree diameter, height and crown diameter. In relation to the site properties the highest seeding recorded was close to the parent in the lower altitude and north facing slopes. The canonical correspondence analysis revealed that seed production correlated with first canonical axis, re- flecting the soil nutrition including N, C, Cu, Ca, Fe, and Br. Results revealed that beech masting is affected by the amount of organic C, P, Fe, and Ca content of soil. Seed production varied from tree to tree, and site to site, and quantitative analyses in relation to both environmental factors and seed tree dimension showed a highly complex interplay of factors dictating seed production across the forest stands. Key words: beech nuts, natural regeneration, Fagus orientalis, seed size.

Introduction pact of selection on seed size (Nilsson and Wastljung 1987, Sork 1993). -pro- Masting (mast seeding), defined as the ducing (e.g. ) produce episodic production of large seed crops abundant and highly nutritious , by a plant population at intervals of years which are an important food source for (Kelly 1994), is observed in many tree many forest vertebrates (Jensen and Niel- species. It is expressed synchronously sen 1986, Ouden et al. 2005, Övergaard over the subcontinent (), in et al. 2007). Nuts are also heavy seeds a pattern occurring only once every few that need biotic agents (animals) to be dis- years (Koenig and Knops 2000, Fischer et persed, and, thus, have developed certain al. 2016). Several hypotheses have been traits to attract seed-dispersing animals proposed to explain the evolution of large (Vander Wall 2001). Successful seedling seed crop production. These hypotheses recruitment following mast year is critical include increased pollination efficiency, for some species (Negi et al. 1996, Akashi seed predator satiation and potential im- 1997, Alvarez-Aquino and Williams-Linera 66 V. Etemad and K. Sefidi

2002, Wagner et al. 2011). Nowadays, management of old-growth In most forest trees the regeneration forest is a crucial challenge for forest potential is determined by the frequency managers in Iran and in other countries.. of mast years (Taylor and Aarssen 1989). The successful establishment of commer- The species of Fagaceae are well known cially valuable beech trees is important for their irregular production of nuts (Mat- their successful long-term forest manage- thews 1955, Nilsson and Wästljung 1987, ment. In forest dominated by shade toler- Hilton and Packham 1997). Masting and ant species much of this regeneration will production of large seed crop is common be recruited in canopy gaps (Drößler and in Fagus grandifolia Ehrh. in North Amer- Von Lüpke 2005, Nagel et al. 2010, Sefidi ica (Canham 1990, Walters and Reich et al. 2011). Knowledge of seed produc- 1996, Alvarez-Aquino and Williams-Lin- tion, dispersal, phenology, and conditions era 2002), F. sylvatica L. in Europe (Nils- needed for successful regeneration is son and Wastljung 1987, Peltier et al. fragmentary or lacking for oriental beech. 1997, Övergaard et al. 2007, Drobyshev Information about masting behaviour and et al. 2010), F. crenata Blume in Japan seed production of mature beech trees (Igarashi and Kamata 1997, Suzuki et al. can be considered as most important for 2005), and F. orientalis Lipsky in Iran (Ete- beech stands in close to nature forestry in mad and Marvi Mohajer 2004). the north of Iran. Fagus orientalis Lipsky is the dominant To address this knowledge gap and species in the Caspian forest region in to advance our understanding of beech Northern Iran and covers about 1,000,000 masting and seed production in a temper- ha (Knapp 2005). Masting behaviour had ate forest of Northern Iran, we been reported previously for this species analysed records of regional masting from (Etemad and Marvi Mohajer 2004). Beech three study areas in beech stands (Fig. 1). forests as productive and commercial We formulated the following objectives: (1) were far from human manipulation in long- to identify whether seed tree variables are term history, so have a high ecological related to seed production; (2) to identify and conservation value. main environmental variables controlling Studies in Oriental beech forests in- seed production of beech. dicated that the structure of such undis- turbed beech stands are more or less ir- regular and uneven-sized (Sagheb-Talebi Material and Methods and Schütz 2002, Sefidi 2012) and that natural regeneration normally occurs in small gaps (Delfan Abazari et al. 2004, Es- mailzadeh et al. 2011, Sefidi et al. 2011). Study area The relic forests it in the north of Iran re- This study was conducted in the forest of main relatively stable and far from large Mazandran province in the north of Iran. scale disturbances since Tertiary (Knapp It is located at latitude from 36°58’ N to 2005). These remaining areas provide a 35°35’ N and longitude from 54°8’ E to valuable reference for studying temperate 51°21’ E. Three study sites were selected deciduous forest ecosystem dynamics, in this region, including: Band, Bon forest and the silvicultural options for beech for- (BS) in the western part of this province, ests in general. Patom forest (PS) in its center, which is Seed Production and Masting Behaviour in Oriental Beech ... 67

Fig. 1. Distribution of Caspian forest in the north of Iran (Modified according to Knapp 2005) and three selected study sites. managed by the University of Tehran as east of the province (Table 1). PS has a an Experimental forest, and the third one long term management history, but BS and was Tolenchal forests (TS) located in the TS are unmanaged and undisturbed area.

Table 1. Location, dominant overstory species, and canopy height of the three F. orientalis – domi- nated forest stands from Northern Iran. Study Site Site code Location Dominant trees Management history Patom PS 36o38’ N, 34o51’ E Acer velutinum – Carpinus betulus Managed Bandbon BS 36o52’ N, 50o37’ E Fagus orientalis – Carpinus betulus Unmanaged Tolanchal TS 36o22’ N, 53o38’ E Fagus orientalis – Carpinus betulus Unmanaged

The climate is sub-Mediterrane- Each location plot was selected to sat- an with a mean annual temperature of isfy the following conditions: (1) the plot 9 °C and total annual precipitation of should be placed in the central part of the 1380 mm. Mature forests are dominated forest to prevent edge effects; (2) the plot by beech, which has an average volume should be located in sufficient distance of 187 m3∙ha-1, representing 74 % of the from former wind throw and managed total stand volume. Structurally, these for- area; and (3) the plot should be homoge- ests have high volumes of coarse woody nous regarding the slope. debris with mature forests averaging 51 m3∙ha-1 (Sefidi et al. 2013). According to the Habibi Kasseb (1992), the most im- Field sampling and sample portant soil types in the Hyrcanian region preparation are Brown, Alluvial, Rendzina, Colluvial, Rankers and Lithosols. Seed collection of the samples was done 68 V. Etemad and K. Sefidi in September 2002 and seedling emer- cording 30 seed trees in each transect. gence and survival was recorded from Intersected trees in a transects were April 2003 to June 2004. considered as a sampling plot. Around In the study area, in order to recognize the sample trees four 1 m2 wide rectan- seed production and related environmen- gular micro plots were laid out in different tal factors, three parallel transects were directions, as illustrated in Figure 2. The established in east-west direction within length of seed traps varied depending on each selected site, starting from random- productive tree’s crown length and area ly chosen points. All Oriental beech seed covered by crown. Micro plots were sep- trees with diameter at breast height (DBH) arated into two equal sections in order to >50 cm that intersected transect were se- record seed production and germination. lected. Transects were separated by 50 m Data collected were from 360 micro plots interval to avoid duplicate sampling. Their in each study site. We also recorded den- length varied depending on the extent on sity of beech nesting in different distances homogeneous forest structure and stand from seed trees and found different pat- density. Measurement continued by re- terns in altitudinal categories.

Sample 1 m tree

50 m Seed germination Seed production record area record area

Fig. 2. Seed traps experiment and sampling design of seed traps around productive trees.

In the laboratory, these samples suffering damage by rodents or birds; were divided into , seeds, , and (5) decayed. The term ‘seed’ is used branches and other material. After identifi- throughout this paper in a broad sense as cation of beech seeds, they were counted being a single dispersal unit usually con- and classified into five categories: (1) vi- taining a single embryo. able, having sound viable cotyledons; (2) In laboratory after seed blowing some immature, having no or undeveloped coty- seed properties were measured including: ledon; (3) insect damaged, suffering dam- seed moisture content, weight of 1000 age by insect larvae; (4) mouse damaged, seeds, physical purity, seed size (length), Seed Production and Masting Behaviour in Oriental Beech ... 69 germination and vigour. Moreover, in sam- Analysis pling plots other important environmental factors influencing seed production such Data followed a normal distribution (Sha- as dominant tree species, crown canopy piro–Wilk W test, P > 0.05) and therefore, coverage, slope steps and aspect were differences in seed production between recorded. Seed germination analysis was the quantitative properties of seed trees conducted in the laboratory. Four hundred such as diameter at breast height (DBH), seeds forming four replications were plant- tree height, crown diameter and height, ed in recommended substratum for germi- and some environmental factors were ex- nation after dormancy breaking using ver- amined by one-way Analysis of Variance nalization treatment. Number of seedlings (ANOVA). emerging daily was counted from the day Spearman’s rank correlation coeffi- of planting. Also to assess seed dispersal, cient was used to test the relationship we sampled seeds in seed traps located between seed production and seed tree in 1 m intervals in different aspects and DBH, tree height and crown height. The slope steps. statistical tests were conducted at a sig- Soil samples were collected in the nificance level of α = 0.05. study sites. The soil texture and propor- To investigate the relationships be- tion of clay silt and loam were measured tween seed production and environmen- by hygrometry method. Large live plant tal parameters, we used a multivariate material (root and shoots) and pebbles approach. Multivariate analyses were in each sample were removed by hand in performed using CANOCO version 4.5 the laboratory and discarded. Then, air- (ter Braak and Smilauer 1998). Principal dried soil samples were sifted with aggre- component analysis (PCA) was used to gates broken up to pass through a 2 mm assess the gradient length of seed crop sieve and roots removed prior to chemical data for each of the environmental factors. analysis. Samples were dried at 608 °C It is mostly used as a tool in exploratory to 708 °C for 18 h, and organic C con- data analysis and for making predictive centration at 0–15, 15–30 and 30–45 cm models. PCA is a mathematical procedure soil depths using a core soil sampler with that uses an orthogonal transformation to an 81 cm2 cross section was determined convert a set of observations of possibly (Rahmani and Zare Maivan 2004). To- correlated variables into a set of values of tal N was determined using the Kjeldahl linear uncorrelated variables called princi- method. The kinetics of nitrogen mineral- pal components. ization for 100 g of each soil sample was measured using a laboratory incubation procedure under controlled conditions Results (Robertson et al. 1999). In order to quan- tify organic P that becomes soluble by ex- The mean of seed production and 1000- traction with 1 M HCl (which is expected seed weight in the study sites were 4688 to remove Al and Fe) we used a second ±164 and 182.6 ±1.3 g, respectively. De- extraction with 0.1 M NaOH to obtain ‘oc- scriptive statistics of seed production is cluded’ P (Condron et al. 1996). shown in Table 2. 70 V. Etemad and K. Sefidi

Table 2. Descriptive statistics of beech seed Figs. 3 and 5). Results of ANOVA indicat- -2 production, n∙m . ed that seed production in beech trees Seed type Mean Median Max Min significantly differs among crown diame- Viable 493.93 ±26 166 6554 0 ter (F = 5.193, P > 0.000, Fig. 4) crown Immature 452.9 ±26 200 7652 0 height (F = 9.025, P > 0.000) and crown Insect 182.3 ±10 75.5 3272 0 area (F = 4.044, P > 0.000) classes. Also predation DBH (F = 6.248, P > 0.000) and the height Other 2081.2 ±105 601 2470 6 predation of seed trees (F = 11.057, P > 0.000) had Total 4687.7 ±164 2395 33472 82 significant impact on seed production.

Seed production significant- ly varies within three study sites (P < 0.000, F = 58.86). The high- est number of seeds was counted in TS (914.2 m-2), the lowest in PS (416.6 m-2) and intermediate one – in BS (424.28 m-2). Figure 3 il- lustrates the number of seeds per square meter among study sites. The mean of DBH and height of seed trees was 81.53 cm and 33.6 m, respectively. Tree diameter Fig. 3. Seed production in the study sites. (r = 0.488, P < 0.000) and height (r = 0.468, P < 0.000) were signifi- cantly correlated with the seed pro- duction. Also, the crown diameter as third variable was significantly correlated with seed production (r = 0.378, P < 0.000). In relation to the site properties the highest number of seed produc- tion was recorded in the lower alti- tude and north facing slopes. The mean weight (of 1000 seeds) also significantly varied among study sites (F = 26.281, P < 0.000). The highest seed weight was recorded Fig. 4. Relationship between seed production in TS (195.45 ±2.57) and the low- and tree crown diameter. est one – in PS (172.12 ±1.94). BS had middle value of seed weight (178 Seed quality and dispersal ±1.95). The mean length of Oriental beech seeds was 14.5 ± 0.05 mm. The results The quality of seed was analysed in three showed that the seed size significantly dif- study sites. The mean number of viable fered among sites (F = 19.25, P < 0.000, seeds in study sites was 493.93 ±26 per Seed Production and Masting Behaviour in Oriental Beech ... 71

We also recorded density of beach nesting in different distanc- es from seed trees; according to our results we have different pat- terns in altitude categories. Beech seeds in uplands are found close to the seed trees. Increasing the distance from seed trees caused decrease of the amount of beech seeds in all study sites (Fig. 6).

Soil properties Seed production moderately cor- related with the soil C (r = 0.519, P > 0.021), N (r = 0.686, P > 0.021), Fig. 5. Box plot of seed size in three study sites. Cu (r = 0.321, P > 0.061) and Fe square m. Table 3 show results of ANO- (r = 0.121, P > 0.012). VA for the three study sites. The highest Multivariate PCA showed that the first amount of viable seeds was recorded in axis (which accounted for 24.0 % of the BS. total variation) was represented in hori- Table 3. Results of one way ANOVA of study zontal structure and heterogeneity of veg- sites. etation. The second component (43.4 %) Source of variation df F P represented increasing vertical heteroge- Viable 2 69.85 < 0.000 neity. Immature 2 45.48 < 0.000 Of micro- and macroelements studied, Insect predation 2 40.35 < 0.000 percent content of K in the soil showed Other predation 2 13.55 < 0.000 different impact on the seed production Total 2 58.82 < 0.000 in comparison to N, C, Cu, Ca, Fe, and Br. Results revealed that beech masting Results of seed germination analysis was affected by the amount of C in soil, revealed significant difference between soil depth and P, Fe, Br, and Ca content study sites. The highest amount of ger- (Fig. 7). mination was recorded for seeds from the middle elevation (around 1500 m). The results of ANOVA showed that elevation categories had a consistent significant effect on seed germination (F = 164.31). In BS 36 % of the seeds germi- nated. In PS and TS this amount was 40 and 41 %, respectively. Mean number of seedlings in TS -2 (9.8 m ) was significantly higher Fig. 6. Change in the seed frequency by distance than in the other two sites. from seed trees in different elevation. 72 V. Etemad and K. Sefidi

Fig. 7. Seed production-soil properties biplot diagram from PCA with soil related variables including nutrients (left) and other soil properties (right).

Discussion es few and heavy seeds that disperse lo- cally by gravity. This can be advantageous This research highlighted key parameters for seedling growth in vicinity of the parent influencing masting of F. orientalis in the trees. Cutini et al. (2013) reported signifi- unmanaged and managed natural for- cantly lower seed masting in beech than ests in the north of Iran. Our analysis of in the other tree species like Oak the most regional dataset of beech seed (Quercus cerris) and Chestnut (Castanea production suggested that tree dimension, sativa) in Italian broadleaved forests. The including height and diameter, crown di- widely accepted evolutionary advantages ameter and some environmental variables of masting behaviour are associated with such as altitude and nutrient availability in seed-predator satiation and improved the soil were the most important factors pollination (e.g. Kelly 1994, Shibata et influencing beech seed production. al. 1998, Koenig and Knops 2000). High Quantitative analyses of seed pro- immature seed record in the study area duction showed that 10.5 % of nuts were could be due to unsuccessful pollination viable and it seems this percent cannot in the spring. Late spring frost observed provide suitable seed bank for successful in the study area affected pollination and regeneration. Total 4687.7 ±164 seed per damaged seed quality. Male and female unit of ground area were counted in the flowers of beech are vulnerable to spring study sites, but 20,263 seeds were pre- frosts (Young and Young 1992). dated that equals to 48.2 % of total seed Piovesan and Adams (2001) reported production and is higher than the values that sufficient seed production in three reported for Fagus crenata stands (Kon et beech species depends on climatic con- al 2005; 48 %). However, in some stud- ditions during two consecutive years be- ies 70–100 seeds per m2 were reported fore masting year. The best predictors as sufficient seed bank for the sustaina- were the unusually moist and cold sum- ble regeneration of beech (Osward 1983). mer two years before masting, favouring Beech, as shade tolerant species produc- carbohydrate build-up, and the early sum- Seed Production and Masting Behaviour in Oriental Beech ... 73 mer drought in the next year. According to was recorded in the lower altitude, and the authors (Piovesan and Adams 2001) north facing slopes. F. orientalis dominates there is a close relationship between mas- in the northern aspects of Elborz moun- ting (mast year) and preceding growing tain, so high number of seeds was record- season climate events in Eastern North ed in the north facing slopes. The highest America and Europe. The high temper- amount of germination was recorded for ature seems to be a climatic factor con- seeds from the middle elevation (around trolling the seed crop in Oriental beech 1500 m). Moreover, seed germination sig- forests – this fact was emphasized also nificantly varied among different elevation by similar study in Japan (Masaki et al. classes. In the same study within mixed 2008). They demonstrated that mainly two beech stands seed production correlated types of climatic signals cause mast seed- with soil nutrition, northern aspects, and ing of F. crenata in the subsequent year: an altitude of around 1500 m. Seeds also higher temperature in summer and lower had greater size and weight in stands of temperature in spring. 750 to 1500 m (Etemad and Marvi Moha- As we expected, in the study area tree jer 2004). Several studies have reported dimensions, including diameter, height a relationship between mast occurrence and crown diameter were significantly and physiographical characteristics of correlated with the seed production. Am- site in other types of forests (Masaki et mer et al. (2002) reported that seeding of al. 2008, Ayari et al. 2011). Topographic beech was directly influenced by crown gradient, particularly in mountain ecosys- canopy. Forest managers often know tems, is an important factor influencing about tree age effect on the seeding that plant growth. Site condition and resource is caused by some physiological changes. availability varies across topographic gra- Seed production occurs during the mature dients, so forest sites with different topo- phase in which flowering occurs. Matura- graphical gradients support different life tion or adult phase first appears at the top history characteristics for individual tree of crown as tree gets older and increases species. Moreover, moisture availability crown size (Barnes et al. 1998), so exist- and daily insolation are hypothesized to ence of positive and significant relation be major factors affecting plant distribution among tree crown and seed production is and growth. These variables are affected expectable. In the same way, F. sylvatica by several topographic characteristics studies showed seed production relates to such as slope aspect, position, and steep- tree diameter (Wagner 1999). Seed pro- ness. North-facing slopes, lower-slope duction and flowering begins at about the positions, and low degree of slope often age of 40 to 50 for F. sylvtica (Young and characterize mesic site conditions, while Young 1992, Kon et al. 2005) and about at southern aspects, upper-slope or ridge- the age of 60 in F. orientalis with masting top positions and steep slopes often sup- every 3 to 18 years (Mirbadin and Namira- port xeric sites. nian 2005). In the same study in the north Soil analysis indicates that beech mas- of Iran, some evidence was reported ting is affected by the amount of organ- about strong dependence of flowering ic C in soil, soil depth, P, Fe, Br, and Ca and fruiting on climate and site conditions content. Soil nutrition effect on seed pro- (Etemad and Marvi Mohajer 2004). duction and density of regeneration was The highest number of seed production shown in the European and Hyrcanian 74 V. Etemad and K. Sefidi beech forests (Modry et al. 2004, Etemad Alvarez-Aquino C., Williams-Linera. G. 2002. and Marvi Mohajer 2004). In terms of dry Seedling bank dynamics of Fagus grandi- weight, most seeds consist of more than folia var. mexicana before and after a mast 90 % storage reserves. The nutrient re- year in a Mexican cloud forest. Journal of Vegetation Science 13: 179–184. serves contain carbohydrates, lipids (usu- Ammer Ch., Mosandl R., El Kateb H. 2002. Di- ally triacyl glycerides) and specialized rect seeding of beech ( L.) storage proteins. These storage reserves in Norway spruce (Picea abies [L.] Karst.) are of critical importance for germination stands – effects of canopy density and fine and especially, when combined with differ- root biomass on seed germination. Forest ent micro- and macronutrient. In the pine Ecology and Management 159: 59–72. forest results showed that the density was Ayari A., Moya D., Rejeb M.N., Ben Mansoura the highest whereas soil clay content and A., Albouchi A., De Las Heras J., Fezzani T., pH was lowest (Puhlick et al. 2012). Ef- Henchi B. 2011. Geographical variation on cone and seed production of natural Pinus fects of soil P on pollen production was halepensis Mill. forests in Tunisia. Journal demonstrated previously (Lau and Ste- of Arid Environments 75: 403–410. phenson 1994). This study showed that Canham C.D. 1990. Suppression and release growing conditions such as soil P avail- during canopy recruitment in Fagus gran- ability can influence the size of a pollen difolia. Bulletin of the Torrey Botanical Club grain and its chemical composition, which, 117: 1–7. in turn, can affect its ability for successful Condron L.M., Davis M.R., Newman R.H., pollination and mature seed production. Cornforth I.S. 1996. Influence of conifers Agnes and Partick (1997) showed that on the forms of phosphorus in selected New Zealand grassland soils. Biology and deficiency of Fe and Br cause decrease Fertility of Soils 21: 37–42. of the transmission of hydrocarbons and Cutini A., Chianucci F., Chirichella R., Donag- delay in flowering, which resulted in poor gio E., Mattioli L.M., Apollonio M. 2013. seed crop. Furthermore, ability of seed Mast seeding in deciduous forests of the production is affected by some physiolog- northern Apennines (Italy) and its influence ical processes, which are dependent on on wild boar population dynamics. Annals soil nutrient availability. We can conclude of Forest Science 70: 493–502. that seed production varies from tree to Drobyshev I., Övergaard R., Saygin I., Niklas- tree and site to site. Quantitative analyses son M., Hickler T., Karlsson M., Sykes M.T. 2010. Masting behaviour and dendrochro- of seed production in beech trees showed nology of European beech (Fagus sylvati- relation to both environmental factors and ca L.) in southern Sweden. Forest Ecology seed tree dimension, which reflects a and Management 259: 2160–2171. highly complex interplay of factors shap- Drössler L., Von Lüpke B. 2005. Canopy ing the seed production in the studied for- gaps in two virgin beech forest reserves est stands. in Slovakia. Journal of Forest Science 51: 446–457. Delfan Abazari B., Sagheb-Talebi Kh., Namiran- ian M. 2004. Regeneration gaps and quan- References titative characteristics of seedlings in dif- ferent development stages of undisturbed Akashi N. 1997. Dispersion pattern and mortal- beech stands (Kelardasht, Northern Iran). ity of seed and seedlings of Fagus crenata Iranian Journal of Forest and Poplar Re- Blume in a cool temperate forest in western search 12(2): 302–306 (Farsi with English Japan. Ecological Research 12: 159–165. abstract). Seed Production and Masting Behaviour in Oriental Beech ... 75

Etemad V., Marvi Mohajer M.R. 2004. Investi- Koenig W.D., Knops J.M.H. 2000. Patterns of gation on quality and quantity of seed pro- annual seed production by Northern Hemi- duction of beech (Fagus orientalis Lipsky) sphere trees. The American Naturalist 155: in Mazandaran Forests. In: Sagheb-Talebi 59–69. K., Madsen P., Tearzawa, K. (Eds.), Pro- Lau T., Stephenson A.G. 1994. Effects of soil ceedings from the 7th International Beech phosphorus on pollen production, pollen Symposium. Tehran, Iran, 10–20 May: size, pollen phosphorus content, and the 57–60. ability to sire seeds in Cucurbita pepo (Cu- Esmailzadeh O., Hosseini S.M., Tabari Kouch- curbitaceae). Sexual Plant Reproduction aksaraei M., Baskin C.C., Asadi H. 2011. 7(4): 215–220. Persistent soil seed banks and floristic di- Masaki T., Oka T., Osumi K., Suzuki W. 2008. versity in Fagus orientalis forest communi- Geographical variation in climatic cues for ties in the Hyrcanian vegetation region of mast seeding of Fagus crenata. Populula- Iran. Flora 206(4): 365–372. tion Ecology 50: 357–366. Fischer H., Huth F., Hagemann U., Wagner S. Matthews J.D. 1955. The influence of weather 2016. Developing restoration strategies for on the frequency of beech mast years in temperate forests using natural regenera- England. Forestry 28: 107–115. tion processes. In: Stanturf, J. A. (Ed.) Res- Mirbadin A., Namiranian M. 2005. Determina- toration of boreal and temperate forests, tion of seeding cycle by stem analysis of second edition. CRC Press, Boca Raton: three beech stands. Iranian Journal of For- 103–164. est and Poplar Research 13(3): 353–378 Habibi Kasseb H. 1992. Fundaments of forest (Farsi with English abstract). soil science. Tehran University Press, Teh- Modry M., Hubeny D., Rejsek K. 2004. Differ- ran, No 2118, 428 p (In Persian). ential response of naturally regenerated Hilton G.M., Packham J.R. 1997. Sixteen-year European shade tolerant tree species to record of regional and temporal variation in soil type and light availability. Forest Ecolo- the fruiting of beech (Fagus sylvatica L.) in gy and Management 188: 185–195. England (1980–1995). Forestry 70: 7–16. Nagel T.A., Svoboda M., Rugani T., Diaci J. Igarashi Y., Kamata N. 1997. Insect predation 2010. Gap regeneration and replacement and seasonal seedfall of the Japanese patterns in an old-growth Fagus-Abies for- beech, Fagus crenata Blume, in northern est of Bosnia-Herzegovina. Plant Ecology Japan. Journal of Applied Entomology 121: 208: 307–318. 65–69. Negi A.S., Negi G.C., Singh S.P. 1996. Estab- Jensen T.S., Nielsen O.F. 1986. Rodents as lishment and growth of Quercus floribunda seed dispersers in a heath oak wood suc- seedlings after a mast year. Journal of Veg- cession. Oecologia 70: 214–221. etation Science 7: 559–564. Kelly D. 1994. The evolutionary ecology of Nilsson S.G., Wästljung U. 1987. Seed preda- mast seeding. Trends in Ecology & Evolu- tion and crosspollination in mast-seeding tion 9: 465–470. beech (Fagus sylvatica) patches. Ecology Kon H., Noda T., Terazawa K., Koyama H., Ya- 68: 260–265. saka M. 2005. Evolutionary advantages of Ouden J., Jansen P.A., Smit R., Forget P.M., mast seeding in Fagus crenata. Journal of Lambert J.E., Hulme P.E., Vander Wall Ecology 93: 1148–1155. S.B. 2005. Jays, mice and oaks: predation Knapp H.D. 2005. Die globale Bedeutung der and dispersal of Quercus robur and Q. pe- Kaspischen Wälder. In: Nosrati K., Mo- traea in North-western Europe. In: Forget hadjer R.M., Bode W., Knapp H.D. (Eds.), P.M., Lambert J., Vander Wall S.B. (Eds.), Schutz der Biologischen Vielfalt und in- Seed Fate: Predation, Dispersal and Seed- tegriertes Management der Kaspischen ling Establishment. CABI Publishing, Wall- Wälder (Nordiran). Naturschutz und Biolo- ingford: 223–240. gische Vielfalt, BFN 12: 45–63. Osward H. 1983. Densies de semis nrcessaire 76 V. Etemad and K. Sefidi

a la reussite d une regeneration natuelle. ran. 178 p. Lehetre page: 238–239. Sefidi K., Marvie Mohadjer M.R., Mosandl R., Övergaard R., Gemmel P., Karlsson M. 2007. Copenheaver C.A. 2011. Canopy gaps and Effects of weather conditions on mast year regeneration in old-growth Oriental beech frequency in beech (Fagus sylvatica L.) in (Fagus orientalis Lipsky) stands, northern Sweden. Forestry 80: 553–563. Iran. Forest Ecology and Management Peltier A., Touzet M.A., Armengaud C., Ponge 262: 1094–1099. J.F. 1997. Establishment of Fagus sylvati- Sefidi K, Marvie Mohadjer M.R., Mosandl R., ca and Fraxinus excelsior in an old-growth Copenheaver C.A. 2013. Coarse and Fine beech forest. Journal of Vegetation Sci- Woody Debris in Mature Oriental Beech ence 8: 13–20. (Fagus orientalis Lipsky) Forests of North- Piovesan G., Adams J.M. 2001. Masting behav- ern Iran. Natural Areas Journal 33(3): 248– iour in beech: linking reproduction and cli- 255. matic variation. Canadian Journal of Forest Sagheb-Talebi Kh., Schutz J.P. 2002. The Research 79(9): 1039–1047. structure of natural Oriental beech (Fagus Rahmani R., Zare Maivan H. 2004. Investiga- orientalis) forests in the Caspian region tion diversity and structure of soil inverte- of Iran and potential for the application of brate in relation to beech, hornbeam and the group selection system. Forestry 75: oak-hornbeam forest types. Iranian Journal 465–472. of Natural Resources 56: 425–437 (in Per- Taylor K.M., Aarssen L.W. 1989. Neighbor ef- sian). fects in mast year seedlings of Acer sac- Robertson G.P., Sollins P., Ellis B.G., Laj- charum. American Journal of Botany 76: tha K. 1999. Exchangeable ions, pH, and 546–554. cation exchange capacity. In: Robertson Vander Wall S.B. 2001. The evolutionary ecol- G.P., Coleman D.C., Bledsoe C.S., Sollins ogy of nut dispersal. Botanical Review 67: P. (Eds.), Standard Soil Methods for Long- 74–117. term Ecological Research. Oxford Univer- Wagner S. 1999. The initial phase of natural sity Press, Oxford: 106–114. regeneration in mixed ash-beech stands – Shibata M., Tanaka H., Nazashizuka T. 1998. ecological aspects. Schr. Forstl. Fak. Univ. Causes and consequences of mast seed Gött. Niedersächs. forstl. Vers., Sauer- production of four co-occuring Carpinus länder’s Verlag. 262 p. (in German). species in Japan. Ecology 79: 54–64. Walters M.B., Reich P.B. 1996. Are shade tol- Sork V.L. 1993. Evolutionary ecology of mast erance, survival and growth linked? Low seedling in temperate and tropical oaks light and nitrogen effects on hardwood (Quercus spp.) Vegetation 108: 133–147. seedlings. Ecology 77: 841–853. Sefidi K. 2012. Late successional stage dy- Young J.A., Young C.G. 1992. Seeds of namics of Oriental beech stands in the Woody in North America. Diosco- north of Iran, PhD-thesis, University of Teh- rides Press, Portland, OR. 407 p.