Is sexually transmitted fungal infection evidence for size-related mating success in Neotropical guava fruit tlies?

The intluence of wing length on mate preference was exanúned in natural populations of the Neotropical guaya fruit tly, Anastrepha striata Schiner, at two locations in Costa Rica. Based on evidence that the fungi are transmitted during mating, site-specific infection by Laboulbeniales fungi on the body surface was used to assess mating history. At both.sites, males and females that carrled fungi on the legs andlor on the ven­ tral part of the thorax (males), and on both sides of the notum andlor the dorsal base of the abdomen (females), had significantly longer wings than males and females without fungi. This suggests that individuals of both sexes with longer wings (i.e. larger individuals) enjoy higher mating success. Fungus infection is more frequent in the wet than in the seasonally dry forest, possibly because hosts are .available year-roimd in the wet foresto

Sexual selection as an important evolu tionary force, first proposed by Darwin (1871), has attracted increased attention during tbe last two decades (Tbornhill & Alcock 1983, Gwynne & Morris• 1983, Bradbury and Andersson 1987, Fincke 1997).A1though many variables seem to be involved, perhaps tbe single most important factor affecting intrasexual selection among males in insects is body size (Tbornhill & Alcock 1983).
. Previous studies by tbe senior autbor in Costa Rica and Guatemala (Hedstrom 1994 ) have shown tbat fruit flies of tbe species Anastrepha striata Schiner are frequently infected by laboulbenial fungi of tbe genus Stigmatomyces; and it has also been estab lished tbat transmission occurs during copula tion (Fig. 1), either completed or attempted.
Relatively small insect males are often less able to capture and defend a resource tbat promotes reproductive success than are large rivals.Tepbritid fmit flies are no exception.Burk (1 983) demonstrated in the Caribbean fruit fly Anastrepha suspensa (Loew) tbat res ident males, especially larger ones, tend to win fights witb intruders.Hendrichs (1986) found tbat large laboratory-reared A. suspensa males were more successful in competing for and mating with females.Other research has shown that large fruit flies are at ah advantage with respect to increased flight ability (Sharp et al. 1983;B1Oém et al. 1994) and mating fre quency in laboratory tests (Churchill-Stanland et al. 1986).
In general, Anastrepha species, when not visiting fruits, spend a great deal of time on tbe underside of leaves, and des pite elaborately pattemed wings are easily overlooked by the researcher in the field.This fact, along with their movements among individual host trees, makes them difficult to observe in their natur al environment.Fruit fly researchers have tried to .overcome this problem by maintaining close-to-natural conditions in cages over host plants (e.g.Dodson 1982).With few excep tions, studies of mating behavior of tephritids have therefore been conducted in manipulated conditions in the fieId, in the laboratory, or with laboratory reared flies reIeased into caged host trees in the field (Prokopy & Hendrichs 1979).While these methods have allowed detailed descriptions of behavior, they need to be supplemented with observations under completely natural conditions.
Here, we report the effects of body size on mating success in natural populations of the guaya fruit fly, using as probable evidence the occurrence of infection by sexually transmit ted Stigmatomyees aeiurae Thaxter and S. ver rueulosus Thaxter fungi (Laboulbeniales).In particular, we asked the following question: Is the infection more prevalent in certain size classes of (1) males and (2) females compared to other size classes?If so, is this evidence of preference for larger mates in guaya fruit fiies?

DISCUSSION
Insect size and fungus occurrence: Since the fungí presumptuously are transferred between the sexes during copulation, we inter pret the probability of infection as an indica tion•of relative mating success.Iflonger wings characterize larger individuals, our data sug gest that male and female A. striata preferen tially mate with larger individuals.Of course, a stronger test of a size-assortative mating pat tem requires a study of sperm transfer.Prokopy (1984) suggested that for Anastrepha species,a single mating usually provides a sufficient amount of sperm to last for several weeks, if not an entire lifetime, and noted that females usually resist further mat ing attempts following initial copulation.However, Teles da Silva et al. (1985) showed that A. obliqua (Macquart) and A. bistrigata Bezzi are polygamous; maleo and female A. bistrigata were found to copulate several times: one male was observed to mate 15 times, and one female 13 times.The largest number of repeated matings observed in A. obliqua was three for a male and two for a female (Teles da Silva et al. 1985).
Because fungus transmission requires contact between sexes, infected females or males that copulate only once, or females that are mounted by males in an attempt to copu late only once, represent dead-end hosts for the fungus (Hedstrom 1994).Thus, either A .striata females are mounted more than once, andJor males are polygynous.
Habitat type and fungos occurrence: In non-seasonal, tropical wet forests (like the Costa Rican forest considered in this study), a permanent presence of host guava fruits for A. striata allows generations to overlap, which contri bu tes to a high inciden ce of " fungi (Hedstrom 1994).In contrast, the periodic absence or .scarcity of hast fruits in seasonal, tropical moist forests, as well as in tropical dry forest�, probably explains the lower frequency or absence offungus infections in those habi tats.Subsequent collections of detailed data on seasonalpopulation dynamics of A. striata fruit flies (Hedstrom 1993) support the sug gestion that the high rate of fungal infection in wet environments is maintained because host flies are continuously presento The causes and ecology of Laboulbeniales fungus infection are still poor ly understood.Further research should consid et questions such as: Do fungi-infected females carry more spermatophores?Could larger infecttjd flies be preferred for mating because the fungí.liberate a sinomone r'N.Ramírez, pers.com.)?and Doesfungal infec tion affect Anastrepha' fecundity?BaImford, Bo W. Svensson and two anony mous reviewers for helpful comments on the manuscript.The study was partIy financed by the Royal Swedish Academy of Sciences, the Foundation of Zoological Research of Uppsala University, and Mid Sweden University.
Positions of male and feq1ale of the Allaslrepha slriata fruit fly during mating.Transmission of Laboulbeniales fungi of the genus Stigmatomyces occurs between (A) the ventral side and the thorax of the male of the dorsal side of the abdomen of the female, and (B) the front legs of the male and the sides of the female notum.
Adult A. striata fruit fl ies were collected with standard McPhail invaginated glass traps containing 4% arnmonium acetate and borax in water at two locations within a non-seasonal, tropical wet forest environment: an abandoned• guaya (Psidium guajava L.) (Myrtaceae) orchard at Guácímo (l0°l4'N; 83° 59'W) (elev.120 m), Limón Province, Costa Rica, once a week, from 15 March to 1 April 1987, and from 7 to 27 February 1993; and an unsprayed guaya orchard at the Tropical Agronomy Teaching and Research Center (CATIE), Turrialba (09°55'N; 83°41 'W) (elev.600 m), Cartago Province, Costa Rica, on a weekly basis, from 14 September to 12 November 1989.The collected flies in both study sites were pooled, and wíng lengths [as the length from the base of the fírst costal cell to the tip of the wing (e! Foote 1980)] were measured to the nearest 0.05 mm, usíng an eyepiece mícrometer.A light microscope was used in the laboratory to detect perithecia of Stigmatomyees fungi externalIy on legs and/or the ventral part of the thorax (males), and on both sides of the ootum and/or the dorsal base of the abdomen (females).The number of guaya fruit flies with and without fungal ínfec tion was taIlied.Mann Whitney U-Tests were used in the statistical analysis.RESULTSBoth males (n=52) and females (n= 50) car rying Stigmatomyees fungí had significantly longer wings than males (n= 110) and females (n= 102) without flingi (males: U = 3747, P = 0.001; females: U = 3225, P = 0.008) (Hg.2).