Reinfestation of Streblidae ectoparasites ( Diptera ) in Carollia perspicillata ( Linnaeus , 1758 ) ( Chiroptera ) Reinfestação de ectoparasitas Streblidae ( Diptera ) em Carollia perspicillata ( Linnaeus , 1758 ) ( Mammalia : Chiroptera )

The mark-recapture method allows analysis on the variation in the abundance of bat ectoparasites at consecutive captures. The objectives of this study were to compare the pattern of Streblidae parasitism between capture and recapture of C. perspicillata; ascertain whether the abundance of Streblidae varied with time after removal of ectoparasites at capture and analyze whether the intensity of parasitism remained the same in each individual at capture and recapture. Using bats netted in the State of Rio de Janeiro, 42 individuals of C. perspicillata parasitized by two Streblidae species, Trichobius joblingi Wenzel, 1966 and Strebla guajiro (Garcia & Casal, 1965), were selected. The pattern of parasitism observed at capture was similar at recapture. No relationship was observed between the capture-recapture time interval and the abundance of ectoparasites. There was no relationship between the abundances of ectoparasites at capture and recapture of each individual.


Introduction
The abundance of ectoparasites in their host can vary depending on environmental factors and factors intrinsic to the host.Differences relating to the host's sex, age and body size are the parameters most analyzed regarding variations in the abundance of ectoparasites on bats (KOMENO; LINHARES, 1999;PRESLEY, 2007;PRESLEY;WILLIG, 2008;PATTERSON et al., 2008a,b).Other biological characteristics such as type of refuge, social behavior, reproductive condition and home range can also influence the abundance of ectoparasites (MOURA et al., 2003;TER HOFSTEDE;FENTON, 2005;RECKARDT;KERTH, 2006;LOURENÇO;PALMEIRIM, 2007;PATTERSON et al., 2007;BORDES et al., 2008;McCOY, 2009).
However, despite these studies, little is known about the pattern of how ectoparasites choose their hosts, especially regarding whether there is any individual predisposition to parasitism.Studies that analyze reinfestation can clarify this pattern of choice.In relation to bats, the studies by Esbérard et al. (2005) and Dick and Dick (2006) demonstrated the behavior of ectoparasites in laboratory experiments.Esbérard et al. (2005) showed that the Streblidae species Megistopoda aranea (Coquillet, 1899), Megistopoda proxima (Séguy, 1926), and Strebla guajiro (Garcia & Casal, 1965) search for their hosts by species: Artibeus lituratus (Olfers, 1818), Sturnira lilium (E.Geoffroy, 1810) and Carollia perspicillata (Linnaeus, 1758), respectively.Dick and Dick (2006) analyzed the behavior of Trichobius joblingi Wenzel, 1966 regarding the choice of host and observed that these flies prefer individuals of C. perspicillata that do not present infestation by other individuals of T. joblingi.However, there was no preference between hosts without ectoparasites and with Speiseria ambigua Kessel, 1925, another ectoparasite found in C. perspicillata.Carollia perspicillata is the species of Neotropical bats that is most often recaptured (MELLO et al., 2004;BIANCONI et al., 2006), and it presents prevalence of Streblidae of around 50% (KOMENO; LINHARES, 1999;BERTOLA et al., 2005).Therefore, C. perspicillata is one of the most appropriate species for studies on reinfestation of ectoparasites.
This study aimed to test the following hypotheses relating to Streblidae parasitism in C. perspicillata: 1) Removal of ectoparasites has an effect on Streblidae parasitism, thus demonstrating differences in the pattern of parasitism between capture and recapture; 2) Individuals recaptured within very short times have lower abundances of ectoparasites, thereby demonstrating that reinfestation is dependent on the time for which the individual is at risk of being parasitized; 3) When individuals from which ectoparasites are removed at the time of capture become reinfested, they reach the same patterns of ectoparasite abundance as seen at the time of capture, with an individual pattern of Streblidae infestation in C. perspicillata.

Methodology
Sampling was done from August 1999 to September 2009, in 38 areas at elevations ranging from sea level to 640 m in the state of Rio de Janeiro.The samples were collected by a team from the Bat Diversity Laboratory of the Federal Rural University of Rio de Janeiro.The bats were captured and recaptured in mist nets (7 × 3 m or 9 × 3 m), weighed on a digital balance (accurate to the nearest 1 g), tagged with plastic collars for individual identification (ESBÉRARD; DAEMON, 1999) and released at the same place no more than three hours after capture.Ectoparasites were removed from the bats with the aid of fine-point tweezers and were placed in vials containing a batch of 95 °GL alcohol, for each bat at each collection time.The ectoparasites collected were deposited in the reference collection of the Bat Diversity Laboratory of the Federal Rural University of Rio de Janeiro.For analysis purposes, the first recapture of each individual of C. perspicillata was used, and any recaptures that occurred on the same day as the first capture were ignored.Only the capture-recapture sets in which Streblidae were present, either at capture or at recapture, were taken into consideration.The dates of capture and recapture, body mass and Streblidae ectoparasites found at the times of capture and recapture were analyzed.Ectoparasites were identified with the aid of a stereomicroscope, using identification keys (WENZEL, 1976;GRACIOLLI;CARVALHO, 2001).For each species of ectoparasite, the prevalence, mean abundance of infestation and mean intensity of infestation at the times of capture and recapture were calculated, following the definitions of Bush et al. (1997).Since body size is a factor that can influence the intensity of infestation (PRESLEY;WILLIG, 2008;PATTERSON et al., 2008a), possible differences in body mass between individuals at the times of capture and recapture were checked using the chi-square test.This test was also used to analyze the abundances of Streblidae found in individuals at the times of capture and recapture, and between the proportions of individuals that presented more than one species of Streblidae concurrently, at the times of capture and recapture.
The data were converted into logarithmic form [ln (x + 1)] to carry out the analyses.Pearson's correlation was used between the abundances of pairs of ectoparasite species, in order to check for possible interference of the abundance of one species on another, at capture and recapture.It was also used between the abundance of Streblidae at recapture and the time that elapsed between capture and recapture.Because of wide variations in time between capture and recapture, it was not possible to carry out further analysis on this parameter.Pearson's correlation was performed between the abundances of each species of Streblidae at capture and recapture, to ascertain patterns of abundance per individual.Student's t-test was performed between the total abundance of each species of Streblidae at capture and the total for Streblidae at recapture, and analysis of variance was used between the intensities of ectoparasites at the times of capture and recapture.

Results
Forty-two individuals of C. perspicillata were found to present Streblidae at capture and/or recapture, thereby totaling 16.0% of 263 recapture events.Of the 38 areas analyzed, only 23 showed recaptures of C. perspicillata and only 12 showed C. perspicillata with Streblidae at capture and/or recapture (Table 1).The total sample comprised 21 males and 21 females of C. perspicillata, and all of them were adults.The interval in days between capture and recapture ranged from 1 to 1411 days (mean of 198.0 ± 262.9; median of 151 days).Body mass ranged from 9.0 to 27.0 g for captures (mean: 16.3 ± 2.9) and from 12.0 to 19.0 g for recaptures (mean: 15.6 ± 1.9).There was no statistical difference in individuals' body mass between capture and recapture (c 2 < 3.841; df = 1; p > 0.05).
There was no correlation between the abundance of T. joblingi (n = 42, r = 0.038, F = 0.058, p = 0.811) or of S. guajiro (n = 42, r = 0.000, F = 0.302, p = 0.586) at the time of recapture and the time interval in days that elapsed between capture and recapture.
The abundances of T. joblingi (n = 42, r = 0.199, F = 1.650, p = 0.206) and S. guajiro (n = 42, r = 0.096, F = 0.369, p = 0.547) did not show any relationship between the time of capture and the time of recapture.Individuals that had more ectoparasites at the time of capture did not have more Streblidae at the time of recapture.
Two individuals presented significant differences in abundance of Streblidae between capture and recapture.One of them had a high intensity of S. guajiro at the time of capture, but no presence of Streblidae at recapture (c 2 = 9.0, df = 1, p = 0.003), which occurred three days after capture.The other had a high intensity of T. joblingi on recapture, 342 days after the initial capture, but had not shown any Streblidae at capture (c 2 = 6.0, df = 1, p = 0.014) (Table 2).

Discussion
The two Streblidae species considered in this analysis, T. joblingi and S. guajiro, are species commonly found in C. perspicillata (WENZEL, 1976;KOMENO;LINHARES, 1999;BERTOLA et al., 2005;GRACIOLLI et al., 2006;DICK;GETTINGER, 2005;DICK et al., 2007), and they may coexist in the same host (BERTOLA et al., 2005;SANTOS et al., 2009).Removal of ectoparasites may result in different proportions of ectoparasite species when reinfestation occurs, since individuals without ectoparasites may attract infestation by new species (WENZEL; TIPTON, 1966;DICK;DICK, 2006;TELLO et al., 2008).However, this result is consistent with the study by Tello et al. (2008), who studied the relationship between these two species and showed that there was no interference between the abundances of one species and the other.The speed of the infestation (or reinfestation) process depends on the duration of contact between infested and uninfested individuals, the kind of shelter, the number of individuals in the colony, the parasite population size and the average duration of grooming, among other things (JOHNSON et al., 2004;PRESLEY;WILLIG, 2008;McCOY, 2009).Higher bat densities in their shelters result in increased contact between colony members and thus may increase the chance of infestation (MOURA et al., 2003;JOHNSON et al., 2004;TER HOFSTEDE;FENTON, 2005;PATTERSON et al., 2007).Studies have shown that the shelters used by C. perspicillata, which mainly consist of caves and hollows (TUTTLE, 1976), may have greater abundance of ectoparasites than is observed in open shelters (MOURA et al., 2003;TER HOFSTEDE;FENTON, 2005;PATTERSON et al., 2007).The high loyalty to the shelter and the social stability shown by C. perspicillata (FLEMING, 1988) may contribute towards a population of parasites with constant intensity of parasitism, since the parasite load present in a shelter tends to be balanced between individuals (JOHNSON et al., 2004).The occurrences of reinfestation of individuals in which there was only one day between the removal of ectoparasites and recapture demonstrate that reinfestation can be rapid and can occur within the first hour of contact.A host or social group in a shelter can be considered to be a habitat for Streblidae (TELLO et al., 2008).If we consider each host to be a habitat, it might be expected that some hosts would be better than others, but this could not be demonstrated in the present study.In the case of the individual that had a high intensity of Streblidae at the time of capture but did not show a high intensity at the time of recapture, this recapture occurred three days after the capture, which could explain the difference.Many of the individuals had similar intensities between capture and recapture.Even individuals in the same group may have different intensities of parasitism without any pattern recognition regarding body mass, reproductive condition, sex or age (OVERAL, 1980;PRESLEY;WILLIG, 2008;BERTOLA et al., 2005;DICK;DICK, 2006).Since infestation of bats by Streblidae may occur in shelters through direct contact between hosts or between the shelter and the host (DICK; PATTERSON, 2006), it is possible that the characteristics of the bats' positions during the daytime are more important in relation to infestation and intensity of parasitism.Streblidae can stay in the shelter and use its walls and substrate for larviposition and formation of the puparium, the adult will seek its host after eclosion (DICK; PATTERSON, 2006).The factors that contributed towards the differences in abundance found for these two species of Streblidae are unclear, as are the negative effects of high intensity of parasitism in bats.The low number of samples found at a single collection point, even with a large number of possible hosts, and the large variation in time between capture and recapture demonstrate the difficulties in dealing with reinfestation.Other studies using mark-recapture of hosts and experimental reinfestation should be conducted to obtain a better understanding of the behavior of these ectoparasites and their relationships with specific individuals.
In the present study, it could not be shown that host capture and ectoparasite removal resulted in greater reinfestation in the individual, but the patterns of parasitism by Streblidae at capture were similar to the patterns at recapture.The most highly infested individuals at capture were not necessarily the most highly infested individuals at recapture, thus indicating that there was no individual predisposition or no greater or lesser resistance to parasites.

Table 1 .
Total number of individuals of Carollia perspicillata recaptured and the capture-recapture sets with presence of Streblidae in the State of Rio de Janeiro.

Table 2 .
Abundance of Streblidae among 42 individuals of Carollia perspicillata at captures and recaptures, prevalence, mean intensity, mean abundance and total abundance of Streblidae.Interval at 95% confidence between parentheses.