MARCH 1990 PARfry, FECUNDITYlNo Srzs oF MA. DVAFI

PARITY, FECUNDITYAND BODY SIZE OF MANSONIADYARI IN FLORIDA

L. P. LOUNIBOS.'V. L. LARSON'J AND C. D. MORRIS''4

ABSTRACT. Seasonal trends in reproductive performance and size were examined among female Mansonia dyari. The incidence of parity ranged between 6.9 and 58.1% and. averaged 22.5% during an 8- month interval, with 4 peaks in parous rate believed to follow periods of emergence. Parous rates were positively correlated with wing lengths, indicating size-dependent survivorship. Wing lengths were highest in the winter and early spring, decreased to a minimum at the end of summer, and increased in the autumn, showing negative correlations with ambient temperatures recorded prior to adult capture. Female biomasses and fecundities were positively correlated with wing lengths, and their mean values varied significantly among sample dates. Egg development in Ma. dyari was completed by 5.4 days post blood meal at26'C and 7.6 days at 20'C. No evidence for autogeny was observed.

INTRODUCTION subtropical St. Lucie County (27" 30'N, 80' 30'W) and separatedfrom related speciesby As important vectors of human filariae in the diagnostic keys. This collection site harbors Afrotropical and Orientalregions, mosquitoes of abundant water lettuce (Pistia stratiotes Linn.) genus (Mansonioides) the Mansonia have been growingin abandonedaquaculture ponds. Water objects of comprehensiveentomological studies lettuce occurs year-round at Chinese Farm, al- (e.g, Laurence1960, Wharton 1962).Neotropi- though large plants typically die back during (s.s.), cal Mansonia all of which belong to the winter cold, resulting in a depressionof plant (Knight subgenusMansonia, include 12 species biomassin the late winter and spring (DeWald and Stone 1977;Knight 1978),of which two, and Lounibos 1990).A thermographsituated 1 Ma. dyariBelkin, Heinemannand PageandMa. m aboveground in an oak-palm woodland 6 km (Walker), titillans range into Florida. Although to the north of Chinese Farm provided contin- Ma. dyari has been identified as a maintenance uous records of ambient temperatures for the (SLE) vector of St. Louis encephalitis virus in study. The central tendency of weekly temper- (Gorgas Panama Memorial Laboratory 1978), ature was calculatedas weekly (maximum + and Ma. titillans has been incriminated in the minimum)/2, hereinafter referred to as the transmissionof Venezuelanequine encephalitis weekly temperature. (Gilyard virus to man in the Caribbean 1944), Parity: Parity was scored by examination of relatively little is known of the life histories of tracheolar skeins of freshly killed individuals these2 species. capturedwith dry ice-baitedCDC traps operated Investigations of the habits of Ma. dyari in from dusk to dawn. Four collections for parity (referred Florida to in earlier literature as Ma. determination were made between September indubitans Dyar and Shannon) have, until re- 1986and February 1987,then 18 successivecol- cently, been limited to studies on voltinism, lections at approximately fortnightly intervals preferred (Provost seasonality and host plants from April to December1987. Ovaries of 25 to 1976, Lounibosand Escher 1985, Slaffand Haef- 39 femaleswere examinedfrom eachcollection ner 1985).In spiteofprolific local abundanceof exceptfor December1986 when only 20 individ- Ma. dyari in the vicinity primary of its host uals were scored. Females with mature eggsor plant, water lettuce (Lounibos and Escher, undigestedblood were excluded. 1985),there is no published information on the Wing lengthsand biornasses:Wing length was reproductive ecologyofthis species.Such infor- measuredfrom the baseof the costato the wing mation would be important if Ma. dyari should margin,exclusive of fringes,under 20x. From 3 play a role in the maintenanceof SLE virus, a separateCDC light trap collections in June, human health concern in Florida. August and Octoberof 1987,a wing was removed and measured,and the remaining corpse dried MATERIALS AND METIIODS at 80'C for 24 h. Dry weights were measuredto Collectionsites: Adult Mo. dyariwerecollected the nearest0.01 mg on a Cahn electrobalance. by methodsdescribed below at ChineseFarm in Egg deueloprnenttime: During August 1987, females were collected in 'lard-can' traps in which were allowed to blood-feed from a 'Florida gth they Medical EntomologyLaboratory, 200 chicken.On the morningof the day of collection, StreetS.E.. Vero Beach.Florida 32962. 2 femaleswith blood were placed in incubators at Polk County Environmental Services,P.O. Box + + 39.Bartow. FL 33830. L2L:I2D and 26 1'C or 20 0.5'C. Mosquitoes 3Current addressofV. L. Larson: Marine Resources wereheld in 0.045m3 cages with accessto a l0% Council, 2519Vasser Street, Melbourne, FL 32901. sucrose solution. At successive24-h intervals a Current addressof C. D. Morris: Florida Medical beginning 62 h after blood-feedingat 26"C and Entomology Laboratory. 86 h at 20'C, 10 femaleswere dissected to assess JounNar,oF THEAunnrclt Moseurro Corrlrnor,AssocrnrroN VoL. 6, No. ovariandevelopment. Christophers' (191 1 ) ovar- ian stages (I-V) were transcribed into arabic numerals. Intermediate stages of 2.5, 3.5 and 4.5,corresponding respectively to ovarianstages IIb, IIIb and IVb of Clements(1963), were rec- ognized.Follicles in which all but the anterior o= o tube of the mature egg (Linley et al. 1986) were r! filled with yolk were scoredas 4.5. o- Fecundity: On 8 dates betweenApril and No- I vember \987, Ma. dyari femaleswere captured in dry ice-baitedCDC or chicken-baited'lard can' traps and allowedto blood-feedon chickens. Females with blood were held at 26 + l"C, t2L:t2D with access to l0% sucrose but no 198&1987 oviposition site. Six to 10 days after the blood Fig. 1. Frequency of parity amongMansonia d.yari meal, the length of a wing was measured,the captured by CDC traps at ChineseFarm. The first 4 femalewas killed and the number of mature eggs sampleswere not separatedby regular intervals, but (stage 5.0 follicles) counted under a dissecting betweenApril and December1987; sampling was done mrcroscope. fortnightly; n : 25-39/point exceptfor December1986 Autogeny: Femaleswere captured with emer- wheren = 20. gencetraps (Slaff et al. 1984)placed over water lettuce, then maintained with males for 7-10 days in 0.045 m3 cages at 26'C in a I2L:L2D incubator or outside in the shade in ambient I t "".ou" summerconditions. Subsequent examination of t'ulli'arous spermathecaerevealed that insemination, which E ,tTl I E is necessaryfor the expression of autogeny in {t{$tJr -, somemosquito species (O'Meara 1985), was not tt) tr occurring under these conditions. Therefore, fe- o 1'LI4ff J , r, *,T{hl malesfrom 2 autumnal collectionswere exposed Et I x x Lrt -Iflltl' 't to malesin 2.5 1 1 m outdoorcages, where = inseminationrates of 34.3%(n:35) and73.9% 4il{,iltTl r Tl,l- (n: 29 were recorded.The ovariesof all dis- o #l sected females were staged by Christophers' = {- (1911)method.

RESULTS 1987 Trends in parity and wing length: The inci- Fig. 2. The wing lengths of Mansonia dyari females dissectedfor parity on 19 occasionsduring 1987.Ver- dence of parity varied time, ranging over be- tical bars represent + I SD. Means separated by tween 6.9% in late February to 58.t% in late asteriskswere significantly different as determinedby May (Fig. 1). Overall, 21.9%of all femalesdis- t-tests(two-tailed). sected(n: 668) betweenSeptember 1986 and December1987 were parous. Between February and December 1987, representing most of one correlation coefficients (Sokal and Rohlf 1981), annual cycle,22.57oof females(n : 583) were revealed that the rs could have come from the parous. During this period, 4 peaks in parity same population (12: 5.696 < X2.osftt : 14.067). were recognized in late May, mid-July, early The mean + SD wing length of all parous Septemberand mid-December. females (3.23 + 0.25 mm, n : 131) was not Mean wing lengths were greatest in early significantly different (t: 1.38, P: 0.16) from spring, declined,on average,during the summer that of all nulliparous females (3.20 I 0.24 mm, and were shortestin early September(Fig. 2). n : 452\. Comparisons were performed sepa- Wing lengths increased again in the autumn, rately for each collection date to test the null but averagevalues in November and December hypothesis (two-tailed) that the mean wing were Iessthan in the preceding spring. Average length of parous females was equal to that of wing lengths from 19 collection dates were neg- nullipars. Once in early May and once in late atively coilelated with weekly ambient temper- July, mean wing lengths of parous mosquitoes atures,which rangedfrom 13.5to 26'C. Signif- were significantly longer (May: t : 2.74, P : icant correlationswere obtainedwith tempera- 0.01; July: t : 2.24, P : 0.03) (Fig. 2). For a tures which laggedadult capture by 3-15 weeks single collection in mid-June, the mean wing (Table 1). Testing for homogeneity among all length of nulliparous individuals was signifi- MARCH 1990 Plurv, FrcuNorty erqoSrzr or Me. ovent

Table 1. Product-moment correlations between weekly ambient temperatures [(maximum + minimum)/2] and mean wing lengths of Mansonia dyari. Time lag

(weeks) 1l IJ 15 r' -0.18 -0.46 -0.68 -0.70 -0.76 -0.61 -0.61 t Levelsof significancefor 17 df: Por = 0.58;P,os = 0.46.

Table 2. Regressionand ANCOVA results for biomassregressed upon wing length and a posteriori comparisonof adjustedmeans. Regressionr ANCOVA'

Sample Mean Adjusted date R2 mass (mg, mean S.E. June 67 0.006 -1.586 0.581 0.569 0.417a 0.013 Ang. 80 0.005 -0.975 0.583 0.401 0.436a 0.008 Oct. 80 0.006 -1.116 0.448 0.560 0.625b 0.010 t Coefficientsin the equationy : bx + a; all R'valuessignificant at P < 0.001. 'Adjusted means followed by different letters are significantly different at P < 0.05 [approximate interval methodof Gabriel(1978)1.

Table 3. ANCOVA results for fecundity regressed upon wing length and a posteriori comparisonof adiustedmeans. tt, c g Mean Adjusted tt Sample date n no. eggs meant S.E. (E April 13-14 16 152.8 136.9ab 8.53 o J April27-28 10 154.6 137.2ab 10.34 o (! May 20-21 I 122.7 1O7.7a 10.38 o- June 8-9 6 159.5 143.2ab 12.44 I July 22-23 7 158.1 167.4b 10.90 August4-5 33 136.7 144.4ab 5.40 Sept.29-30 25 145.3 158.3b 6.87 Nov.18-19 17 154.3 155.4b 6.76 l Adjustedmeans which do not sharea common MedianWing Length (mm) letter aresignificantly different at P < 0.05[approxi- Fig. 3. The relationshipbetween parous rate and mateinterval method of Gabriel(1978)1. medianwing length of 7 sizeclasses of Mansonindyari. Numbersabove each point aresample sizes. The ver- tical scaleconverts rates into the angulartransfor- for 45-58% of the explained variance (Table 2). mation. The three regressioncoefficients did not differ significantlyfrom one another (Fz.zzr:2.738,P cantly longer than their parous counterparts (t : 0.067),indicating homogeneity of slopes. : 2.62,P : 0.01).From the 16 remainingcom- Mean dry weights from the 3 sample dates parisons there was no evidenceto reject the null were comparedfollowing analysis of covariance hypothesisof equal wing lengths (0.21 < P < (ANCOVA) which adjusted for differences in 0.92). wing lengthsamong the samples.The ANCOVA Data from the above 19 collection dates were detectedsignifi cant differencesamong adjusted pooledand femalessegregated into ? wing length mean biomasses(Fz,zzr : 165.836,P < 0.001). classes.The sizeclasses were 0.15mm in width An a posteriorl comparison indicated that Oc- exceptfor the smallest,which rangedfrom 2.59- tober's adjusted mean dry weight was signifi- 2.79mm and the largest,ranging between 3.56 - cantly greater than June's or August's (Table 3.92 mm. Parousrates, subjectedto the arcsin 2). transform, were positively correlated with me- Fecundity: Variation in number of eggs was dian wing lengthsof the 7 sizeclasses (Fig. 3). high, ranging from 66 to 209.Three femaleswith No parous individuals were recorded amongst no eggswere discounted from further analyses. the smallest Ma. dyari. Mean numbersof eggsper samplevaried from a Wing lengths and body nloss: Plots of dry Iow of 122.7in May to a high of 159.5in June, weight versus wing length for 3 sample dates but showed no obvious seasonaltrends (Table yielded significant linear regressionsaccounting 3). t24 JouRNel oF THE AltapRIceNMosQutto CourRol AssocllrroN VoL.6, No. 1

DISCUSSION c) CD T -.r' (E _ J."- The 4 peaks in parous rates during 1987pre- o L.-'-"4----*- sumably followed periods of Ma. dyari emer- .g gence. mosquitospecies, (E Among most temperate o the frequency of parity is Iow while adult num- ^o {i''lvl bers increase,then rises after the emergence o peak (Morris and DeFoliart 1971).In Florida E CL the number and tim- o I Ma. dyari is multivoltine, o ing of generationsdepending on winter temper- E I that affect water lettuce abundance o atures (Provost1976, Lounibos and Escher 1985).Hi- 62 78 94 110 126 144 150 176 bernal diebackofP. stratioteswas not severein HoursAfter Blood-Meal 1986-87(DeWald and Lounibos1990), allowing of Ma. dyari, a high propor- Fig. 4. Box plot of Christophers' stage of ovarian a spring emergence developmentas a function of time. Broken Iine con- tion of which blood-fed and oviposited in May nects median (50th percentile)values at 26'C, and (Fig.1). solid line connectsmedians at20"C. Upper and lower Among Mansonia uniformis (Theobald) in boundariesof boxesare 75th and 25th percentiles,and Ceylon, a parous rate of 50-60% was found to upper and lower bounds of stems are 90th and 10th, be relatively constant from month to month respectively. Where absent, these percentiles were (Samarawickrema1968). In equitable,equato- identical to the more central values.Each point based rial temperatures the gonotrophic cycle of Ma. (n: on 10 dissectedfemales except at 62 h 9) and 86 uniformis remains close to 4 days, and 1 blood h (n:7\ postblood meal. meal is sufficient to mature eggs (Wharton 1962).Under such conditions daily survival may The regression of fecundity on wing length be realistically estimated by the parous rate was significant (Fr,rzr: 6.408, P : 0.013), but formula of Davidson(1954). explained only 5% of thevariance. When regres- Steady-stateconditions do not apply to Ma. sion coefficients were compared among sample dyori populations in Florida where the parous dates, no significant differences were detected rate underwent frequent large fluctuations (Fig. (Fz,rra: 0.451, P : 0.867). An ANCOVA with 1), and adult mosquitoesmay experienceaverage wing length as the covariate showed significant temperature differences of >10' between sum- variation among the adjusted mean fecundities mer and autumn, which will affect gonotrophic (Fz,rra: 2.900, P < 0.001). Computation of the cyclelength (Fig. a). Nevertheless,if the perioc 95% confidence intervals of adjusted means in- of sampling between February and December dicated that significantly fewer eggs were devel- representsa completebreeding season,then the oped in May than in samples from July, Septem- overall parous rate (0.225) is the annual proba- ber or November (Table 3). bility of completing one gonotrophic cycle (Haw- Gonotrophic cycle lengths.' Dissections after ley 1985). blood meals revealed that egg development was Size variation within natural mosquito popu- complete (stage 5.0) for all females at 183 h (7.6 lations has been hypothesized to be related to days) post blood meal at 20"C and 129 h (5.4 larval resource limitations or predation (Fish days) at 26'C (Fig. 4). Although the median 1985).Fish calculatedcoefficients of variation female at 20'C had developed ovaries to stage (CVs) to comparethe variability in size of 10 4.0 by ca. 4 days post blood meal, completion of speciesof northern mosquitoescaptured in light the final stage by all individuals required an- traps.Using the samecommon regression equa- other 3 days at the lower temperature. Detection tion for converting wing lengths to biomasses, of a few resting stage (2.0) follicles between 62 we havecalculated a CV for Ma. dyari of 23.59' and 139 tr (Fig. 4) indicated that blood had not This placesMa. dyari in Fish's categoryof mod- promptly triggered egg development in all fe- erate sizevariability, composedmainly of mos- males. quito inhabitants of permanent marshes and Absence of autogeny: Dissected ovaries of all Iake marginswhere predation is consideredthe but 1 of 169 females held without blood for 9- most important determinant of invertebrate 15 days after emergence were in Christophers' community structure (Zaret 1980).Although at- stage 1.0 or 2.0. The lone exception had one- tachmentto roots may protect Mansonialawae third of her follicles developed to stage 3.0. and pupaefrom somepredators (Van den Assem These observations indicated that neither mated 1958). fish and Odonata larvae consumeMa. nor unmated Ma. dyari are able to complete egg dyari immatures on water lettuce in Florida (L. development autogenously, P. Lounibos,unpublished data). Mlncu 1990 PlRtrv. Fnculrorrv ANDSrzE oF MA. DvARI r25

Seasonalsamples often show an inverse rela- (Ludlow) and probably also Aedes triseriatus tionship between correlates of body size and (Say), variations in adult size are important environmental temperature (Landry et al. 1988, population regulating mechanisms (Hawley Bock and Milby 1981).Wing lengths of Culex 1985).Size variability is Iessamong Ma. dyari tarsalis Coquillett in central California showed and, hence, probably less important in popula- a high negative correlation with ambient tem- tion regulation. Secondly, in contrast to perature one week before trapping (Bock and Livdahl's and Hawley's work, our study con- Milby 1981).For Ma. dyari temperature lags of trolled neither female agenor blood meal size, 3 or more weeks before capture yielded higher which factors among other culicid speciescon- correlationswith wing lengths (Table 1). These tribute to the large deviations from comparable different temporal relationshipsto ambient tem- regressions(Packer and Corbet 1989). perature may reflect both the slower larval The average number of eggs developed by growth of Mansonia, requiring almost 1 month field-caught Ma. dyari ranged between 722 and from hatch to emergenceunder optimal condi- 159 (Table 3), which is more than the mean tions for Ma. uniformis (Laurence 1960),and numbersof 101and 111matured by laboratory- the slower warming of the deeper,shaded larval rcared Ma. uniformis or Ma. africana (Theo- habitats of Ma. dyari. bald), respectively (Laurence 1960).The differ- Although a few studies have indicated that ences between mean fecundities adjusted for large mosquitoesof some speciesor populations wing length (Table 3) might be explainedby may be more likely than smaller counterparts to discrepanciesin mosquito ages. The signifi- be parous (Haramis 1983, Packer and Corbet cantly lower fecundity value occurred in Iate 1989),other investigations have revealedno re- May when the parous rate was highest (Fig. 1). lationship between body size and survival, as Since mosquito clutch size tends to decreasein measuredby frequency of parity or recapture successivegonotrophic cycles (Clements 1963), (Landry et al. 1988,Walker et al. 1987).Our the older age structure of the May Ma. dyari data indicate that larger Ma. dyari are more sample may have contributed to its decreased likely to be parous, the smallest females rarely averagefecundity. completing a gonotrophic cycle (Fig. 3). This Knowing the annual probability (P) of a fe- evidencesupports the presenceof size-depend- male'scompleting a gonotrophiccycle, then the ent survivorship among Ma. dyari. number of egg clutches laid per female is given Femalesfrom October sampleshad a signifi- bv P/t - P (Hawley 1985). For Ma. dyari this cantly higher adjustedmean biomassthan sam- value is 0.29. Based on the range of fecundities ples from June or August (Table 2). Van den of 122-159, the averagefemale will contribute Heuvel (1963)demonstrated that the relation- 35-46 eggsto the next generation. ship between dry body weight and wing length amolagAedesaegypti (Linn.) dependedupon the temperature experiencedby developing larvae. ACKNOWLEDGMENTS The seasonaldifferences in Ma. dyorf biomass We appreciatethe help of L. B. DeWald and may be related to temperature changes in the in the field. Commentsfrom J. F. larval habitat. R. L. Escher Day, F. O'Meara and, particularly, W. A. Autogeny has never been reported in the ge- G. manuscript preparation. The re- nus Mansonio (O'Meara 1985), although prob- Hawley aided supported,in part, by contracts from ably few specieshave been examined.Ma. dyari searchwas (DAMD 17-85-C-5182)and FIor- apparently require blood to developeggs. There the U.S.Army of Health and Rehabilitative is no information on blood meal sourcesof Ma. ida Department is of Florida IFAS dyari in Florida, but this speciesfeeds on mam- Services. This University journal seriesno. R-00147. mals and birds in Panama (Gorgas Memorial Laboratory L978),where cormorants and herons are suspectedas reservoirsof SLE virus (J. L. REFERENCES CITED Petersen, personal communication). Birds which frequent water lettuce habitats may also Bock, M. E. and M. M. Milby. 1981. Seasonalvariation provide blood sourcesfor Ma. d,yari in Florida. of wing length and egg raft size in Culex tarsalis. Although fecundity was positively related to Proc. Calif. Mosq. Vector Control Assoc. 49:64-66. wing length, the regressionexplained a small Christophers, S. R. 1911. The development of the egg proportion follicle in Anophelines. Paludism 2:73-88. of variance compared to similar fe- Clements, A. N. 1963. The physiology of mosquitoes. cundity:size relationships calculated for tree Pergamon, Oxford. hole Aedes (Livdahl 1984, Hawley 1985). Both Davidson, G. f954. Estimation of the survival rate of biological and methodologicalfactors may con- anopheline mosquitoes in nature. Nature 174:792- tribute to this difference. ln Aedes sierrensis 793. 126 JoURNAL oF THE ArranRrc.qNMoSQUITo CoNtRor- AssocrnrroN Vor.. 6, No. 1

DeWald, L. B. and L. P. Lounibos. 1990. Seasonal associatedwith water lettuce (Prstio stratiotesl in growth of Pjstla stratiotesin south Florida. Aquatic southeasternFlorida. FIa. Entomol.68:169-178. Bot. (in press). Morris, C. D. and G. R. DeFoliart.1971. Parous rates Fish, D. 1985.Analysis of adult sizevariation within in Wisconsinmosquito populations. J. Med. Ento- natural mosquitopopulations. p. 419-430.In:Loun- mol.8:209-212. ibos,L. P., J. R. Rey and J. H. Frank (eds.),Ecology O'Meara,G. F. 1985.Ecology of autogenyin mosqui- of mosquitoes:proceedings of a workshop.Florida toes.p. 459-471.1n: Lounibos,L. P., J. R. Rey and MedicalEntomology Laboratory, Vero Beach. J. H. Frank (eds.),Ecology of mosquitoes:proceed- Gabriel,K. R. 19?8.A simple method of comparison ings of a workshop.Florida Medical Entomology of means.J. Am. Stat. Assoc.73:.724-729. Laboratory, Vero Beach. Gilyard, R. T. 1944.Mosquito transmission of Vene- Packer,M. J. and P. S. Corbet. 1989.Size variation zuelan virus encephalomyelitis in Trinidad. Bull. and reproductive successof female Aed.espunctor U.S. Army Med. Dept.75:96-107. (Diptera: Culicidae).Ecol. Entomol. 14:297-309. Gorgas Memorial Laboratory Annual Report. 1977. Provost, M. W. 1976. Mawonia mosquitoes:genera- (1978).U.S. GovernmentPrinting Office,Washing- tions per year in Florida. Proc. Fla. Anti-Mosq. ton, DC. Assoc.47th Meet., p. 24-27. Haramis, L. D. 1983.Increased adult size correlated Samarawickrema,W. A. 1968.Biting cyclesand parity with parity in Aedestriseriatus. Mosq. News43:77- of the mosquitoMansonia (Mansonioides)uniformis (Theo.)in Ceylon.Bull. Entomol. Res.59:299-314. ffu*i"v, W. A. 1985. Population dynamics of Aedes Slaff,M. and J. D. Haefner.1985. Seasonal and spatial sietensis.p. 167-184.1n: Lounibos,L. P., J. R. Rey distribution of Mansonia dyari, Mansonia titillans, and J. H. Frank (eds.),Ecology of mosquitoes:pro- and Coquillettidiaperturbans (Diptera: Culicidae)in ceedingsof a workshop. Florida Medical Entomol- the central Florida, USA, phosphateregion. J. Med. ogy Laboratory, Vero Beach. Entomol. 22:624-629. Knight, K. L. 1978.Supplement to a catalogof the Slafl M., J. D. Haefner,R. E. Parsonsand F. Wilson. mosquitoes of the world (Diptera: Culicidae). 1984.A modifiedpyramidal emergence trap for col- Thomas Say Found. Vol. 6, Suppl. Entomol. Soc. lectingmosquitoes. Mosq. News. 44:197-199. Am., CollegePark, MD. Sokal,R. R. and F. J. Rohlf. 1981.Biometry. 2nd ed. Knight, K. L. and A. Stone. 1977.A catalogof the W. H. Freeman,San Francisco,CA. mosquitoes of the world (Diptera: Culicidae). Van den Assem,J. 1958.Some experimental evidence Thomas Say Found. Vol. 6, Entomol. Soc. Am., for the survival value of the rootpiercing habits of CollegePark, MD. Marxonia larvae to predators. Entomol. Exp. Appl. Landry, S. V., G. R. DeFoliart and D. B. Hogg. 1988. l:125-729. Adult body size and suwivorship in a field popula- Van den Heuvel, M. J. 1963. The effect of rearing tion of Aedes triseriatus. J. Am. Mosq. Control temperature on the wing length, thorax length, Ieg Assoc.4:121-128. length and ovariole number of the adult mosquito, Laurence, B. R. 1960. The biology of two speciesof Aedesaegypti (L.). Trans. R. Entomol. Soc.,Lon- mosquito,Mansonia africana (Theobald) and.Man- don.115:197-216. sonia uniformis (Theobald) belonging to the Walker, E. D., R. S. Copeland,S. L. Paulson and subgenusMansonioides (Diptera, Culicidae).Bull. L. E. Munstermann. 1987.Adult survivorship, pop- Entomol.Res. 51:491-517. ulation density, and body size in sympatric popula- Linley, J. R., P. A. Linley and L. P. Lounibos.1986. tions of Aedes triseriatrzs and Aedes hendersoni Light and scanningelectron microscopy of the egg (Diptera:Culicidae). J. Med. Entomol. 24:485-493. of Mansonia titillans (Diptera: Culicidae). J. Med. Wharton, R. H. 1962.The biology of Mansonia mos- Entomol. 23:99-104. quitoes in relation to the transmission of filariasis Livdahl, T. P. 1984.Interspecific interactions and the in Malaya. Bull. Inst. Med. Research,Fed. Malaya, r-K continuum: a laboratory comparison of geo- 11. graphic strains of Aedes triseriatus. Oikos 42:L93- Zaret,T. M. 1980.Predation and freshwatercommu- 202. nities. Yale Univ. Press,New Haven, CT. Lounibos, L. P. and R. L. Escher. 1985. Mosquitoes