Female Hematozoan Infection Reduces Hatching Success but not Fledging Success in Pied Flycatchers Ficedula hypoleuca

—We report association between female blood parasite prevalence (percentage of infected birds) just after egg laying and reproductive success in two successive breeding seasons, in a breeding population of Pied Flycatcher Ficedula hypoleuca in central Spain. Females infected with Trypanosoma spp. had a higher probability of deserting their clutches during the incubation period than noninfected females. Females infected with Haemoproteus balmorali hatched proportionally fewer eggs than noninfected females. Female infected with H. balmorali during the incubation period may have a decreased ability to thermoregulate which may affect their incubation capacity. Fledging success, breeding success, fledgling mass, and tarsus length were not associated with infection of the mother by blood par1 E-mail: [email protected] asites during the incubation period, suggesting that females and their mates may compensate during the nestling period for the negative effect of blood parasites during the incubation period. Most parasites exert an intense selective pressure on their hosts by reducing their condition, survival prospects, mating, or reproductive performances (Loye and Zuk 1991, Møller 1997). That detrimental effect may vary in relation to parasite virulence. In order to understand evolutionary interactions occurring between parasites and their hosts, it is necessary to know the extent to which parasites have a detrimental effect on their hosts’ reproduction and survival in wild populations. Reproductive success has been the target of many field studies of birds, because it is expected to relate closely to fitness (Stearns 1992). Blood parasites have been shown to negatively affect reproductive success of their hosts July 2001] 751 Short Communications because they compete with the host for resources (e.g. review in Møller 1997). If blood parasites have a detrimental effect on their hosts, heavy parasite infections will correlate with low host reproductive success. Most studies in wild populations have shown a negative effect of parasites on some measure of host reproduction (e.g. clutch size, fledging success; reviewed by Møller et al. 1990, Møller 1997). In particular, many studies have focussed on nestling survival. A problem of those correlative studies is that most birds were sampled at the end of the nestling period, after the main reproductive investment was performed by parents. In birds, high reproductive effort is positively correlated with blood parasite load both for natural clutch sizes (Korpimäki et al. 1993, Allander and Bennett 1995, Dufva 1996, Oppliger et al. 1997, Ilmonen et al. 1999) and for experimentally manipulated clutch sizes (Norris et al. 1994, Richner et al. 1995, Ots and Hõrak 1996, Allander 1997, Siikamäki et al. 1997, Wiehn and Korpimäki 1998, Fargallo and Merino 1999, Wiehn et al. 1999; but see Merino et al. 1996). If host reproductive effort increases susceptibility to blood parasite infection or the likelihood that latent, chronic infections may relapse, parasitism may be a consequence of reproductive effort, not its cause (Oppliger et al. 1997). Therefore, studies looking for effects of blood parasites on reproductive success in free living individuals should sample birds before their main reproductive investment is performed. Moreover, because reproduction and immunity are likely to compete for resources (Gustafsson et al. 1994, Sheldon and Verhulst 1996, Lochmiller and Deerenberg 2000, Norris and Evans 2000) and the total amount of energy a bird is able to put into both may depend on food availability (Appleby et al. 1999), studies looking for effects of blood parasites on reproductive success should sample birds under different ecological conditions. We examined association between blood parasites (Haemoproteus balmorali, Trypanosoma spp.) and reproductive success of female Pied Flycatchers (Ficedula hypoleuca) breeding in central Spain. Blood parasites of females were sampled just after egg laying in two successive breeding seasons. We tested the hypothesis that blood parasites may negatively affect reproductive success of females by examining observed relationship between maternal blood parasite infection and hatching, fledging, and breeding success of Pied Flycatchers. Methods. The Pied Flycatcher is a small (12–13 g), migratory, sexually dichromatic, and hole-nesting passerine of European woodlands (Lundberg and Alatalo 1992). Egg laying in our study population typically begins in late May, and clutch size ranges from 2 to 8 eggs with a mode of 6 (Sanz 1997). The female incubates alone and during that period receives part of her food from her mate. Both sexes feed the young, which fledge within 15–18 days of hatching (Lundberg and Alatalo 1992). The study was conducted during the springs of 1998 and 1999 in a deciduous forest of Pyrennean oak (Quercus pyrenaica) at 1,200 m above sea level in the vicinity of La Granja, central Spain (408549N, 048019W). Frequent checks of nest-boxes provided data on dates of clutch initiation and clutch size for all breeding pairs. On the day after clutch completion, females were captured and banded with numbered rings. Length of incubation was defined as number of days between completion of clutch and first signs of hatching. Nestlings were measured and ringed on day 13 after hatching (hatching date 5 day 0). Nestlings were weighed to the nearest 0.1 g and their tarsus length was measured to the nearest 0.01 mm with a digital caliper (Svensson 1992). Unmanipulated broods were visited daily from day 16 onwards to establish numbers of fledged and dead young. Hatching success (proportion of eggs hatched), fledging success (proportion of hatchlings that resulted in fledged young), breeding success (proportion of eggs that resulted in fledged young), and offspring size (fledgling mass and tarsus length) were considered as partial measures of reproductive success. Blood sampling. We chose 47 females among 123 Pied Flycatcher pairs trapped in 1998 and 57 females among 142 Pied Flycatcher pairs trapped in 1999. A drop of blood from the left brachial vein was collected in a microcapillary tube and transferred to a glass slide, smeared, air-dried, and fixed in absolute ethanol some hours later. Smears were Giemsastained for 45 min to assess hematozoan infections (Merino and Potti 1995, Merino et al. 1997). Parasite analysis was performed by E. Arriero. Smears were microscopically scanned at 2003 in search of large parasites such as Trypanosoma spp. To prevent the possibility that the symmetry of the blood smear might cause a nonrandom distribution of parasites (Godfray et al. 1987), one-half of each smear was entirely scanned (;300 fields scanned, one-half being chosen at random; Merino and Potti 1995). Subsequently, intraerythrocytic parasites such as Haemoproteus balmorali were detected with oil immersion at 1,0003 in the other half of the smear. Blood-parasite prevalence was defined as proportion of infected females in the sample. Statistical procedures. For females that had been sampled in both years, we randomly picked one observation to be included in the analyses to avoid pseudoreplication. Thus, all birds were entered in the analysis only once. Breeding attempts were classified in two categories, deserted (1) and nondeserted (0), during the incubation period. Logistic regression analysis was used to explore variables that may influence the female’s decision to desert the clutch, that is, year, blood parasite prevalence. Effect of blood parasites on reproductive performance was analysed 752 [Auk, Vol. 118 Short Communications TABLE 1. Results of logistic regression analysis of factors determining clutch categories (deserted vs. nondeserted) of female Pied Flycatchers during incubation period. Variable (n 5 104) Estimate –2logLR df P Study year Female parasite status by Trypanosoma Female parasite status by Haemoproteus Interaction (year 3 Trypanosoma) Interaction (year 3 Haemoproteus) 0.7590 1.2183 0.2440 20.5020 20.5143 5.674 25.219 0.914 2.620 4.139 1 1 1 1 1 0.0172 ,0.001 0.3389 0.1055 0.0419 FIG 1. (A) Probability of clutch desertion according to study year and female infection by Trypanosoma spp. (open bars 5 uninfected females; hatched bars 5 infected females) during incubation period. (B) Mean hatching success according to study year and female infection by Haemoproteus balmorali (open bars 5 uninfected females; hatched bars 5 infected females). The error bars represent the standard error of the mean. Numbers above bars are sample sizes. using ANOVA and entering respective breeding parameters as dependent variables. All statistical analyses were performed using SPSS for Windowst (Norusis 1993) and were two-tailed. Results. In the present population, 35 and 16% of females sampled during incubation period were infected with Haemoproteus balmorali and Trypanosoma spp., respectively (n 5 167; Sanz et al. 2001). To study effect of blood parasites on decision of females to desert their clutches, a logistic regression analysis was performed of clutch categories (deserted vs. nondeserted) against Haemoproteus balmorali infection status (infected vs uninfected) of females, Trypanosoma spp. infection status of females, and study year. Probability of clutch desertion was dependent on study year (Table 1), but was independent of female H. balmorali infection status (Table 1). A higher number of clutches were deserted during the incubation period in 1999 than in 1998 (Fig. 1a). Probability of clutch desertion was significantly dependent on female Trypanosoma spp. parasite status (Table 1). More clutches were deserted when females were infected by Trypanosoma spp. than when females were uninfected (Fig. 1a). Interaction term between female Trypanosoma spp. parasite status and study year was nonsignificant (Table 1), showing that differences in probability of clutch desertion between study years for infected and uninfected females by Trypanosoma spp. were in the same direction (Fig. 1a). There were no significant differences in laying date and clutch size between Haemoproteus balmorali infected and uninfected females (Table 2), and between females with or without Trypanosoma spp. infection when the effect of study year was controlled (Table 2). Haemoproteus balmorali infected females had lower hatching success than uninfected females (Table 2; Fig. 1b), and there were no differences in hatching success between females with or without Trypanosoma spp. infection when controlling for the effect of study year (Table 2). There were no differences in incubation period, fledging success, breeding success, fledgling body mass, and tarsus length (mean values per brood) between H. balmorali infected and uninfected females (Table 2), and between females with or without Trypanosoma spp. infection when the effect of study year was controlled (Table 2). Discussion. We have shown that female Pied Flycatchers infected with Trypanosoma spp. had a higher probability of clutch desertion during incubation period than uninfected females. Moreover, females infected with Haemoproteus balmorali had a lower hatching success than uninfected females. In that southern breeding population, H. balmorali was the most common blood parasite for females during incubation period. Species of H. balmorali are the most common hemosporidians encountered in free-living birds and are believed to be the least pathogenic of blood parasites (Atkinson and van Riper 1991). However, negative effects may not be easy to document, especially July 2001] 753 Short Communications TABLE 2. Analyses of variance of reproductive parameters of female Pied Flycatchers in relation to blood parasites prevalences (Haemoproteus balmorali, Trypanosoma spp.) and study years.