Mosquito Species Collected from a Marsh in Western Kenya During the Long Rainsi

SppruMsnn 1991 Moseurro spp. FRoM I Mlnsx tN WnsttnN Knxvl 395

MOSQUITO SPECIES COLLECTED FROM A MARSH IN WESTERN KENYA DURING THE LONG RAINSI

T. M. LOGAN,2 K. J. LINTHICUM,s P. C. THANDE,' J. N. WAGATEH2 AXO C. R. ROBERTS'

ABSTRACT. A total of 475,431mosquitoes representing 8 generaand 43 specieswere collectedfrom a narsh in the western Kenya highlands to determine speciescomposition and successionin relation to the epidemiologyof Rift Valley fever vbus. Culcxpipiens was the most common species,totalling 92.3Vo of the collection,followed by Cx. zombaensis(2.2%), Anopheles coustani (I.I%), An. squamosus(0.8Va), Mansonia uniformis (0.6%), CoquiLlcttidiam.icroannulatus (0.5%), Uranotaenia mashonaensis(0.5%)' Ma. africana (0.4%) and Cq. aurites (0.4%). Aedes quasiuniuittatus was the first floodwater speciest9 emergefrom newly flooded areas and was the most abundant Aedes collected, representing 88% of all Aedes specimens.Culex guiarti and Cx. zombaeruiscolonized newly flooded areas soon after the areas becameflooded.

INTRODUCTION permanent and temporary swamp and marsh- land areas to a greater extent than areas in A serologicalsurvey of cattle in Kenya indi- central Kenya, where earlier observations on cated that Rift Valley fever (RVF) virus is pre- RVF vectorswere made (Linthicum et al. 1983, dominately located in the central and western 1984, 1985). This study was to determine the highlandsofthe country (Davies1975). Studies mosquito speciespresent in a permanent marsh- conducted in the central Kenya highlands sug- land area of the western Kenya highlands after gest that RVF virus is maintained during inter- the onsetof the Iong rainy season. epizootic periods in transovarially infected particularly Ae. floodwater Aedes mosquitoes, MATERIALS AND METIIODS mcintoshi Huang las Aedes lineatopennis Lud- low (Linthicum et al. 1985)1.Epizootics of RVF The study site was Iocated at the Maji Mzuri are thought to occur after periods of prolonged marsh, a permanent marsh located 9 km E of and widespread rainfall (Davies et al. 1985), Moi's Bridge, Uasin Gishu District, Rift Valley when dambo formations (Ackermann 1936, Province, Kenya (0"48'N; 35"12'E: altitude Mackel 1974,Linthicum et al. 1983) containing 1,981 m). The oval marsh was approximately aedine eggs are flooded, allowing development 250 ha and was similar to South African pans of Aedes species transovarially infected with (Garganet al. 1988).The dominant emergent RVF virus. After infection of susceptiblehosts vegetation in the central part of the marsh was by suchAedes vectors, the virus outbreak can be Typha latifolio Linn. and Cyperuslatifolius P oir. amplified by secondary RVF mosquito vectors Predominant amongthe grassesalong the marsh of many different genera (Meegan and Bailey margins and drainage area were speciesof Pan- 1988).The dominant floodwater Aedesbreeding icurnandEchinochiloa. The mean daily high and habitat in the western Kenya highlands appears low temperatureswere 25.4"Cand 11.2"C,re- to be dambos similar to those in the central spectively, with a yearly mean rainfall for the highlands region, as well as swamp and marsh- area of 1,183mm, measured25 km NW of the land margins and drainage areas that become study area at the Kitale meteorological station flooded during the rainy seasons. (Kenya Meteorological Department 1984).Pre- In the Uasin Gishu and Trans Nzoia districts cipitation during the study period was measured of the western highlands, the primary vegetation at Maji Mzuri Estate headquarters,2 km N of features include forests and derived grasslands the marsh. pupae and bushlands (ecological zone II, Pratt et al. Mosquito were sampled daily through- 1966). The land is interspersedwith numerous out the study, from March 31 to May 18, 1989, by sieving 100 dips (0.47-liter dipper) collected along each of 3 discontinuous 100-m sampling lines along the southern 1 km margin of marsh I The views of the authors do not purport to reflect and 3 sampling lines of the same length in the the position of the U.S. Department of the Army, the marsh drainage channel exiting from the north- U.S. Department of Defense or the Government of ern rim of the marsh. Pupae were placed in Kenya. Publishedwith the approval of the Director of 0.027-m3cages at the marsh edge and adult Veterinary Services,Government of Kenya. in 2United States Army Medical Research Unit- specimens collected after emergence the Kenya,Box 401,APO, New York 09675-5000. cages. 3United StatesArmy Medical ResearchInstitute of Collections of resting adults were made daily Infectious Diseases, Fort Detrick, Frederick, MD by walking 10 min along each of the 6 sampling 2t'102-50t1. lines employed in the pupal collections using a 396 JouRNer,oF THEAunnrceu Mosqurro CoNrRor,AssocrerroN VoL. 7, No. 3

Table 1. Numbers of mosquitoescollected from a marsh in Kenya by 3 methods. Number of mosquitoescollected Total Reared number from Resting Light-CO, as Voof total Species pupae adults traps Total collection Cx. pipiens 72 8,347 430,4451 438,804 92.3 Cx. zombocnsis 250 163 9,862 r0,275 2.2 Cx. antennatus 23 76 1,012 1,111 0.2 Cx. uniuittatus 424 798 826 0.2 Cx. uentrilloni IZD 725 0.2 Cx. guinrti 540 Cl 546 0.1 Cx. bitaeniarhynchus 359 359 <0.1 Cx. uansomerini 93 220 Jto <0.1 Cx. mirificus 7 130 r37 <0.1 Cx. aurantapex 65 65 <0.1 Cx. theileri A A <0.1 poicilipes Cx. 58 o.J <0.1 Cx. jinjaensis <0.1 Cx. annulioris 15 15 <0.1 Cx. sitiens 4 8 L2 <0.1 Cx. tigripes I I <0.1 Cx. toroensis o q <0.1 Cx. hnpkinsi 1 1 <0.1 Ma. uniformis 2,9t7' 2,9r7 0.6 Mo africana 1,669 1,669 0.4 An. coustani 36 5,068 5,106 1.1 An. squamosus 3,905 3,958 0.8 gambiae An. ID lo <0.1 An. dthali 1 1 <0.1 An. snithii 1 1 <0.1 Cq. microannulattn 2,246 2,246 0.5 Cq. aurites 496 1,413 1,909 0.4 Cq. annetti 13 t23 136 <0.1 Cq. uersicolar 69 69 <0.1 Cq. pseu.doconopos 45 45 <0.1 Cq. aurea 2 <0.1 Ur. mashonaensis , t11 2,308 0.5 Ur. bilineata 222 95 3t7 <0.1 Ur. balfouri 5 5 <0.1 Ae. quasiuniuittatus 52r 50 1 901 0.2 Ae. d,entatus 95 2 238 335 <0.1 Ae. mcintoshi 2 'D 78 <0.1 Ae. circurnluteolus + 4 <0.1 Ae. unidentatus I o <0.1 Ae. calignostts 2 <0.1 Ae. subdentatus 1 1 <0.1 Ml. farquharsoni 128 <0.1 ML plurnosa 6 6 <0.1 I Populationpeaked at endof Marchand mid-April. 'Populationpeaked in mid-April.

battery powered aspirator (Davis and Gould and Harbach (1988), and voucher specimens 1973). Solid-state, army miniature light traps retained. (John W. Hock Co., Gainesville,FL) baitedwith 1 kg of dry ice pellets as a sourceof COz (10-15 RESULTS traps/sampling night), were set biweekly, from March 24 to May 16, 1989,to collect mosquitoes A total of 475,431mosquitoes belonging to 8 from areas along 1 km of marsh margin and 1 genera and 43 specieswere collected as shown km of marsh drainagearea. Mosquito specimens in Table 1. Of the 43 speciescollected, 38 species were identified according to reference keys of were present in light trap collections, the re- Edwards (1941),Gillies and de Meillon (1968) maining 5 species: Malaya farquharsoni (Ed.- Seprnvsrn 1991 Mosqurro spp. FRoM l Mensn rN Wnstnnx KnNve 397 wards), Culcx aurantaper Edwards, Cx. jinjoen- soon afber flooding, as were Ur. mashonaensis srsEdwards. Cx. toroensisEdwards and Gibbins Theobald(n : 6), Cx. antennatrzs(Becker) (n: and Cr. hopkinsi Edwards were collected with 9), Cx. pipiens Linn. (n : 5) and An. coustani the aspirator only. Light trap collections made Lavaran (n : 2). Pupal collections made along along the marsh margin were similar to those the marsh margin demonstrated that Ae. quas- made along the drainage area with the excep- iuniuattatus was the first floodwater speciesto tions that Anophelesdthali Patton (n = l), An. emergeafter the water level expandedover pre- smithii Theobald (n = l), Ae. calignnsus(Gra- viously unflooded banks of the marsh; however, ham) (n :2), Ae. subdentatusEdwards (n = l) the successionof specieswas unclear due to an and Cr. poicilipes (Theobald) (n : 5) were found influx and mixing of speciesalready colonizing only in collections made along the marsh mar- the marsh water with emerging floodwater spe- gin. Also Ae. unidentatus Mclntosh (n : 2) and cies and opportunistic speciesbreeding in the Urarntacnia balfouri Theobald (n : 5) were in newly flooded margin areas.The pupaecollected collections only along the drainage area. from the marsh margins, beginning from first The 3 sampling lines in the drainage area day pupae were seen after the margin was became flooded on different days, as rainfall floodedon April 13 until May 16, 1989,consisted produced overflow water that moved down the of: Cx. guiarti (n : 249), Cx. zombaensis(n: watercourse,inundating each sampling line in 108),Ae. dentatus(n: 40),Ae. quasiuniuittatus succession.The data collected from the first (n:23), Cx. antennatus(n: 74), Ur. rnashon- sampling line that was flooded were represent- aensis(n : 14), Cx. pipiens (n : 7) and Cr. ative of the 2 other sampling Iines in the drain- uniuittatus Theobald (n = 4). agearea, and these data can be usedto illustrate species successionin the drainage area from DISCUSSION March 31 to May 18, 1989 (Fig. 1). Aedesquas- iuniuittatus (Theobald) was both the first flood- Although many specieswere found to be well water speciesto emergefrom the newly flooded establishedaround the permanent marsh habi- areasas well as the most abundant, representing tat, Cx. pipiens was the predominant species. 88% (4991566)of the Aedes spp. collected as Different collection methods may have revealed pupae from the drainage area. Other Aedes spe- additional mosquito species.Light trap collec- cies present werc Ae. dentatus (Theobald) (n = tions yielded 38 species,aspirator collections of 55) and Ae. mcintoshi (n : 2). Culex guiarti resting adults yielded 23 species (including 5 Blanchard (n = 292\ and Cr. zornbaensisTheo- speciesthat were not present in light trap col- bald (n : 135) were collected from the areas lections), and pupal dip collections yielded 10

A cubx Euiarli ' aaedeidentatus D culex zombrensis o- a Aedes quasiunivinatus Q o.t o ooe z uJ = ssE 0.6 6 ^^J IIJ ru< o- lt an zs4 zoE

ApRrL ------_---- l- MAy + I Fig. 1. Mean numbers of Aedes quasiuniuittatus, Ae. dentatus, Culex guiarti and. Cx. zombaensis ptpae collected per dip after the Maji Mzuri marsh drainage area became flooded, March 31 to May l-8, 1989. JouRrler,oF THE Arr,rnntclx Mosgurro Cou.rRolAssocra,rror.r Vor..7, No. 3

species.Emphasis was placed on those species neededto document the occurrenceand abun- found near or along the margins of the marsh. dance of Ae. mcintoshi and other floodwater In temporarily flooded grassland and forest Aedesin the western Kenya highlands, particu- dambos, only 8 and 6 different species,respec- larly in dambo formations. The paucity of sig- tively, were found in collections of immatures nificant numbers of Aedes specimens (Linthicum [particu- et al. 1983,1984). In the grassland larly Aedes (Neomelaniconjon)speciesl and the dambo, Cx. pipiens was the most commonlv tremendous numbers of Cx. pipiens and other found species(Linthicum et al. 1984). Cule.r species in the collections indicate that Aedes quasiuniuittatus was the first floodwa- permanent marsh areasin western Kenva could tet Aedes to emerge from newly flooded areas serve as an important source of RVF epizootic along both the margins and drainage area ofthe vectors, but would be unlikely to play a role in marsh, and it was also the most abundant Aedes the maintenance of RVF virus. Culex pipiens collected. Further studies to document whether and Cr. zombaensishave been reported as epi. Ae. qurniuniuittatus is also the dominant flood- zootic vectors in Egypt and South Africa, re- water speciesin temporarily flooded dambo for- spectively (Hoogstraal et al. 1979, Mclntosh et mations in the western Kenya highlands should al. 1983). be conducted.Linthicum et al. (1983, 1984) showedAe. mcintoshi to be the dominant Aedes ACKNOWLEDGMENTS in dambo formations studied in the central Kenya highlands. A small number of Cul,expu- The authors thank J. N. Mwaura, Director. Maji Mzuri Estate, pae, collected in newly flooded areas, began for accommodatingus at the freld site. C. provided emerging2 days after the first emergenceof Ae. W. Kamau much technical assistanceand quasiuniuittarfus,indicating either that these M. J. Turell and S. W. Gordon kindly reviewedthe manuscript. species(Cx. guiarti, Cx. zombaensisand Cx. antennatus) quickly colonizednewly flooded areas or that they were carried into the sampling areas REFERENCES CITED with the overflowing marsh water. Ackermann,E. 1936.Dambos in Nordrhodesien.Wiss. Figure 1 illustrates that a second emergence Veroff.Dt. Mus.Landerke.Leipzig, NF 4:147-1b2. of Aedesoccurred during the first week of May, Davis,E. W. andD. J. Gould.19?3. A portablesuction as some of the areas in the sampling line that apparatus for collecting mosquitoes. Mosq. News had dried became reflooded during the rainfal^ 33:246-247. Davies, F. G. 1975. Observations in the last week in April. The small numbers of on the epidemiology of Rift Valley fever in Kenya. J. Hyg. (Lond.) Cx. pipiens and.Anopheles spp. in pupal collec- 75:219-230. tions suggest that these species,even though Davies,F. G., K. J. Linthicum and A. D. James.lg8b. predominate in adult collections, are either Rainfall and epizootic Rift Valley fever. BulI. slower in colonizing newly flooded areas than W.H.O. 63:941-943. Cx. guiarti, Cx. zombaenslsand Cr. antennattts. Edwards, F. W. 1941. Mosquitoes of the Ethiopian or breedin areasof the marsh other than alone region. III. Culicine adults and pupae. British Mu- the margins. seum (Natural History), London. Gargan, T. P., P. G. Jupp and The number of Aedes specimens collected R. J. Novak. 1988. Panveld oviposition sites of floodwater Aedes from this marsh habitat mos- over b6 days was only quitoes and attempts to detect transovarial trans- 1,324mosquitoes, with a mean of 1 Aedesspec- mission of Rift Valley fever virus in South Africa. imen collected per dip recorded during the Med Vet. Entomol. 2:231-236. period when Aedes pupae were most abundant. Gillies, M. T. and B. deMeillon. 1968.The Anopheli- Collections made at a temporarily flooded forest nae of Africa south of the Sahara (Ethiopian zoo- edgedambo in central Kenya (Linthicum et al. geographicalregion), 2nd ed.,S. Afr. Inst. Med. Res. 1983)contained more than twice as many Aedes Publ. 54. Harbach,R. E. 1988.The mosquitoesofthe pupaeper dip as comparedwith collectionsmade subgenus Cul.ex in southwestern Asia and Egypt (Diptera: during this study, and collections made in a Culicidae). Contrib. Am. Entomol. Inst. (Ann Ar- flooded grasslanddambo in central Kenya (Lin- bor) 24(1):1-240. thicum et al. 1984)yielded almost 5 Aedespupae Hoogstraal,H., J. M. Meegan,G. M. Khalil and F. K. per dip. Adam. 1979.The Rift Valley fever epizooticin Eg5pt The prevalence of RVF antibodies in many 1977-78. 2. Ecological and entomological studies. cattle (Davies 1975) in the western Kenya high- Trans. R. Soc. Trop. Med. Hyg. 73:624-629. lands may be related to the regular emergence Kenya MeteorologicalDepartment. 1984.Climatolog- ical statistics for of floodwater Aedes vectors. Kenya. Kenya Meteorological De- In view of the in- partment, Nairobi. crimination of Ae. mcinto.shias a maintenance Linthicum, K. J., F. G. Davies,C. L. Bailey and A. host of RVF in the central Kenya highlands Kairo. 1983.Mosquito speciessuccession in a dambo (Linthicum et al. 1985), further studies are in an East African forest. Mosq. News 43:464-470. Sspr:nl{esn 1991 Moseurro spp. FRoMA MARsH In WnstsnN KnNYn 399

Linthicum, K. J., F. G. Davies,C. L. Bailey and A. Mclntosh, B. M., P. G. Jupp, I. dos Santosand A. C. Kairo. 1984. Mosquito species encountered in a Rowe. 1983.Field and laboratory evidenceimplicat- flooded grassland dambo in Kenya. Mosq. News ing Culex zombaensisand Aedes circumluteolus as 44:228-232. vectors of Rift Valley fever virus in coastal South Linthicum, K. J., F. G. Davies, A. Kairo and C. L. Africa. S. Afr. J. Sci. 79(2):61-64. Bailey. 1985.Rift Valley fever virus (family Buny- Meegan, J. M. and C. L. Bailey. 1988. Rift Valley aviridae, gen:usPhlebouirus). Isolations from Dip- fever,pp. 5!-76. In: T. P. Monath (ed.).The arbo- tera collected during an inter-epizootic period in viruses: epidemiology and ecology, Vol. IV., CRC Kenya.J. Hyg. (Camb.)95:L97-209. Press,Inc., Boca Raton,FL. Mackel, R. 1974.Dambos: a study in morphodynamic Pratt, D. J., P. J. Greenwayand M. D. Gwynne.1966. activity on the plateau regions of Zambia. Catena A classification of East African rangeland,with an I:327-365. appendixon terminology.J. Appl. Ecol. 3:369-382.