Oecologia (2000) 125:201–207 DOI 10.1007/s004420000447
T. Szép · A.P. Møller
Exposure to ectoparasites increases within-brood variability in size and body mass in the sand martin Received: 30 November 1999 / Accepted: 8 May 2000 / Published online: 21 June 2000 © Springer-Verlag 2000
Abstract Parasites often have detrimental effects on their hosts, and only host individuals able to cope with parasitism are likely to display induced or genetic resistance. Hosts may respond to parasitism by differential investment in offspring depending on their ability to cope with parasitism, because offspring that perform better than their siblings are themselves likely to have superior induced or genetic resistance. We tested whether nestlings of the highly colonial sand martin Riparia riparia were affected by the haematophagous tick Ixodes lividus by experimentally manipulating parasite loads of nests [nests sprayed with pyrethrum to remove parasites (sprayed), or nests sprayed with water (control)] at three stages of the breeding season. Prevalence and intensity of ticks were significantly affected by treatments. Breeding success was not significantly affected by treatment, although post-fledging survival was twice as high among nestlings from sprayed nests than from controls. Mean phenotypic traits of nestlings generally did not differ significantly among treatments, while within-brood variance in keel length (a skeletal character) and body mass were higher in control treatment broods than sprayed ones. Sedimentation rate, which reflects blood protein and immunoglobulin content, was significantly higher and less variable in sprayed than control broods. These findings are consistent with the suggestion that parasitism effects on host reproductive success act through an increase in the variance of offspring quality. Key words Breeding synchrony · Colony · Ixodes lividus · Phenotypic variance · Riparia riparia T. Szép (✉) Department of Environmental Sciences, College of Nyíregyháza, P.O. Box 166, 4401 Nyíregyháza, Hungary e-mail:
[email protected] Tel.: +36-30-9655836, Fax: +36-42-404092 A.P. Møller Laboratoire d’Ecologie, CNRS URA 258, Université Pierre et Marie Curie, Bât. A, 7ème étage, 7 quai St. Bernard, Case 237, 75252 Paris Cedex 05, France
Introduction Parasites by definition exploit their hosts for limiting resources for some or all of their lifespan. Such exploitation may have more or less severe fitness consequences for hosts depending on resource availability. Parasites may induce fever, immune responses, and behavioural responses (Hart 1997; Wakelin and Apanius 1997), or they may directly interfere with growth and maintenance and hence reduce important fitness components such as timing of reproduction, clutch size, reproductive success and even survival (reviews in Lehmann 1993; Møller 1997). Although parasites have often been shown to increase offspring mortality, the mechanisms involved in parasite effects on host fitness are less clear. Some studies have shown that parasites cause complete brood failure (e.g. Loye and Zuk 1991; Brown and Brown 1996), but in the commonest pattern of mortality one or a few offspring die while the remainder survive (e.g. Møller 1990, 1997). Partial brood loss with the remaining offspring being in good health suggests that some offspring are more susceptible to parasites than others, and that such offspring are at a particularly high risk of death. Christe et al. (1998) hypothesised that parents and offspring may benefit in terms of fitness if parasites were concentrated on a single offspring. Such a “tasty” chick may die during severe parasite attacks, while the remaining part of the brood will survive in prime condition. Although no offspring would “volunteer” for this role, small differences in efficiency of induced or genetic resistance may render particular individuals susceptible to parasitism. Thus, a brood reduction strategy could be based on slight differences in susceptibility as demonstrated by the performance of offspring during their first days of life. Such a strategy would be particularly efficient if offspring differed slightly in size, since size is known to affect access to food (Mock and Parker 1997). A crucial prediction arising from this scenario is that the phenotypes of offspring within a brood should increase in variance when faced with an elevated risk of parasitism.
202
The objective of this study was to test whether low levels of parasitism affect within-brood variation in size and condition. Nestling survival, and particularly growth, are often affected by exposure to ectoparasites (review in Møller 1997). Even when the mean phenotype of hosts is unaffected by parasites, there might still be negative effects through an increase in within-brood variance in offspring phenotype. In particular, adult hosts may favour creation of a size hierarchy among their offspring when the effects of parasitism are unpredictable. In situations when effects of parasitism are severe, adult hosts may sacrifice certain offspring to parasites while maintaining others as fully viable offspring (Christe et al. 1998; Szép and Møller 1999). Conversely, in situations with little or no effect of parasites on offspring performance early in the breeding cycle, adult hosts may still increase the within-brood variance in offspring quality, if nestlings may potentially suffer from negative effects of parasitism at a later stage of the breeding cycle. We tested this prediction by experimentally manipulating the level of ectoparasitism in nests of the highly colonial sand martin Riparia riparia. Sand martins breed in colonies that range from a few pairs to more than 2,000 nest holes. Holes are dug out in river banks or similar sites by males and females (Cramp 1988), and nest holes and chambers are infested with a number of different ectoparasites including the tick Ixodes lividus (Szép and Møller 1999). Ectoparasites are relatively uncommon in newly formed colonies since they are transported on adult hosts from already infested colonies. However, as nest sites get older, parasite loads may reach very high levels with a prevalence approaching 100% (e.g. Szép and Møller 1999). Ectoparasites generally have a weak effect on brood size at fledging (Alves 1997), with fitness costs of parasitism increasing in old colonies (Szép and Møller 1999). Ticks are known to remove blood from nestlings, but also to induce immune responses that cause a change in growth priorities (Saino et al. 1997), with wing feathers growing relatively faster than other parts of the body of nestlings from heavily infested nests (Szép and Møller 1999). This change in growth allows nestlings to escape parasitism relatively early during their life. We worked in a newly formed colony, where effects of ectoparasites were hypothesised to be weak. This would allow an experimental test of whether exposure to ticks affected offspring growth and development.
week with an endoscope (Olympos JF type gastro-fiberoscope), from arrival of the birds in mid-April until their departure in midAugust. Nests were checked every morning when hatching was expected until the hatching was verified. Hatching synchrony was calculated as the percentage of nestlings hatched on the first and second day of hatching (13 days after the end of egg laying). Start of laying, clutch size, hatching date and number of 15-day-old nestlings were determined for first and second clutches. Nests were classified in three categories, early, peak and late nests, on the basis of their breeding synchrony in the sub-colony. Nests of the early-breeding groups started laying on days 17–20 (where day 1 is 21 April), peak breeders on days 21–25 and late breeders on days 26–30. Nests in which laying started day 31–44 day (called very late breeders) were only considered at the sub-colony level of analysis because of the small number of experimental nests (n=14). A bank collapsed during heavy rains on days 59–62, destroying 87 burrows, among them 25 sprayed and 30 control nests. Manipulation of ectoparasite loads The main ectoparasite in the sand martin colonies is the tick I. lividus. We manipulated parasite loads of nests by randomly allocating nest burrows during the egg laying period to the following two treatments: (1) spraying the nest material with 4 cm3 of a solution of pyrethrin (Bio Kill-Interkémia RT, produced by Jesmond Holding AG, Hungary; 0.25% pyrethrum concentration), hereafter sprayed nests (n=71), or (2) spraying the nest material with 4 cm3 of distilled water during the egg laying period (n=75), hereafter control nests. The efficiency of the treatment was subsequently assessed from the tick loads of large nestlings. Capture and measurement of birds Adult sand martins were captured during the nestling period using a nest trap and mist-nets. Adults were inspected for the presence of ectoparasites such as ticks, mites, fleas and feather lice. A number of measurements were taken: the length of the right and left flattened wing and the right and left outermost tail feathers with a ruler to the nearest 1 mm, the length of the tarsus and the keel with digital callipers to the nearest 0.01 mm, and the body mass with a Pesola spring balance to the nearest 0.1 g. The birds were subsequently sampled for blood, marked with a sex-specific combination of white marks on their wing feathers, provided with an individually numbered aluminium ring and released. Nestlings aged 15–16 days were removed from randomly chosen sprayed and control nests, stratified by breeding time (early, peak, late), using a home-made flexible grabbing device connected to an endoscope with a light. All nestlings were assessed for ectoparasites, measured as described above, blood-sampled and provided with an individually numbered aluminium ring before being returned to their burrow. Intense ringing work was done by mist-netting on days 66–87 with nestlings fledging on days 22–51 (where day 1 is 21 April). Blood sampling and assessment of immune responses
Materials and methods Study site and study population We studied sand martins at a large colony of 1,120 pairs (1,870 burrows) in a sand pit at Ibrány-Nagytanya (48°06′04′′N, 21°41′34′′E), Hungary, during April–July 1999. This is one of many colonies belonging to the sub-population breeding in the upper section of the river Tisza, a population that has been investigated as part of a long-term population study (Szép 1995). We checked 222 nesting burrows in a sub-colony of 196 pairs (one of the first-formed sub-colonies in the Upper Tisza region) twice per
Adult and nestling sand martins had a sample of blood removed from their brachial vein in a standard 100-µl heparinized capillary tube. The capillary was subsequently stored horizontally in a cooling box with frozen cooling blocks for 6–8 h and after returning to our domicile the capillaries were placed vertically for 2 h and the level of sedimentation was measured with a ruler to the nearest 0.1 mm on the same day. The sedimentation rate reflects the amount of proteins in the blood, including immunoglobulins (Sharma et al. 1984). Hence, high values signify good condition. We used parametric and non-parametric tests of SPSS 7.5 depending on the distributions of variables. Values reported are means±SE.
203
Results Effect of experimental manipulation on parasite loads The experimental manipulation significantly affected tick infestations of nestlings. The prevalence of ticks on nestlings was 65.5% in control nests (n=29) and 6.5% in sprayed nests (n=31), and this difference was significant (χ2=22.977, df=1, P0.29; Kruskal-Wallis ANOVA). Hatching success was similar to that of first clutches (91%). There was no significant difference in second brood size at 15 days of age (sprayed: 2.62±0.29, n=13; control: 3.12±0.22, n=16; U=78.5, z=–1.198, P=0.254, Mann-Whitney U-test), and there was also no significant difference among breeding groups (P>0.49, Kruskal-Wallis ANOVA). Effects of parasites on nestling phenotypes There was no significant difference between treatments on the mean value of the measured morphological parameters (P>0.08, t-tests or Mann-Whitney U-tests). There was a non-significant difference among treatments in within-brood variance in keel length among nestlings, measured as the coefficient of variation (Fig. 2; t=–2.012, df=42.395, P=0.051, meandiff=0.9151±0.455, unequal variances t-test). There was no significant effect of treatment, breeding time or their interaction (Fig. 2; interaction: F=0.720, df=2,54, P=0.492; treatment: F=3.819, df=1,54, P=0.056; breeding time: F=0.96, df=2,54, P=0.389; main effect: F=1.957, df=3,54, P=0.131, two-way ANOVA). There was a significant difference in within-brood variance in body mass between treatments (Fig. 3; t=–3.007, df=58, P=0.004, meandiff=–2.062±0.686, unequal variances t-test). There was a significant effect of treatment, but no significant effect of breeding time or treatment by breeding time interaction (Fig. 3; interaction: F=0.803, df=2,54, P=0.453; treatment: F=8.568, df=1,54, P=0.005; breeding time: F=0.555, df=2,54, P=0.577; main effect: F=3.22, df=3,54, P=0.03, two-way ANOVA).
204
Fig. 2 Within-brood variance in keel length of nestlings sand martin, measured as the coefficient of variation (CV), in relation to breeding date and parasite treatment. Values are means±SE
Fig. 4 Mean sedimentation rate of blood from nestlings sand martin in relation to breeding date and parasite treatment. Values are means±SE
There was a significant difference in the within-brood variance in sedimentation rate which was larger in control nests (sprayed: 14.709±1.581, control: 20.999±2.344, t=2.224, df=49.677, P=0.031, meandiff=–6.289±2.828)). There was no significant interaction between treatment and breeding time, and only the treatment had a significant effect on within-brood variance in sedimentation rate (interaction: F=0.258, df=2,54, P=0.774; treatment: F=4.846, df=1,54, P=0.032; breeding time: F=0.694, df=2,54, P=0.504; main effect: F=2.098, df=3,54, P=0.111, twoway ANOVA). Effect of parasites on host reproductive success
Fig. 3 Within-brood variance in body mass of nestlings sand martin, measured as the coefficient of variation (CV), in relation to breeding date and parasite treatment. Values are means±SE
Sedimentation rate was significantly larger in sprayed than in control nests (Fig. 4; t=–2.413, df=58, P=0.019, meandiff=–0.041±0.017). There was no significant interaction between treatment and breeding time, although both factors had significant effects on sedimentation rate (Fig. 4; interaction: F=0.022, df=2,54, P=0.978; treatment: F=6.607, df=1,54, P=0.013; breeding time: F=5.778, df=2,54, P=0.005; main effect: F=6.032, df=3,54, P=0.001, two-way ANOVA). However, if we consider the length of the blood in the capillary as a covariate, the effect of breeding time becomes non-significant (covariates: F=7.49, df=1, 53, P=0.008; interaction: F=0.092, df=2,53, P=0.912; treatment: F=6.352, df=1,53, P=0.015; breeding time: F=2.46, df=2,53, P=0.095; main effect: F=3.679, df=3,53, P=0.018, two-way ANOVA).
Brood size before fledging did not differ significantly between treatments (sprayed: 5.03±0.21, n=31, control: 5.00±0.20, n=29; U=435.5, z=–0.218, P=0.84, MannWhitney U-test). Brood size decreased with breeding time, but the difference was not significant (early: 5.25±0.25, n=20, peak: 5.10±0.28) n=21, late: 4.68±0.20, n=19; χ2=5.121, df=2, P=0.076, Kruskal-Wallis ANOVA). We recaptured at least one fledged nestling from twice as many sprayed nests at the colony than from control nests (sprayed: 53.3%, n=30; control: 26.9%, n=26; χ2=4.014, df=1, P=0.045). We recaptured twice as many fledged nestlings from sprayed nests at the colony than from control nests, but the difference was not significant (sprayed: 10.6%, n=151; control: 5.4%, n=130; χ2=2.525, df=1, P=0.13).
Discussion Parasitism and nest re-use by sand martins Nest treatment had a significant effect on prevalence and intensity of tick infestations in sand martin nestlings,
205
while there was no significant effect of timing of reproduction (Fig. 1). This result is similar to that found in a previous experiment in the same population (Szép and Møller 1999). Ticks arrive in nest holes through migration between holes and through transport by adult sand martins from other colonies. The present study was made in a new colony, so that ticks were absent at the start of the breeding season. While the prevalence was 65.5% for nestlings from control nests, it was 6.5% in nestlings from sprayed nests. This compares with prevalence in nestlings of 93% in control nests and 73% in sprayed nests in the previous study, made in a previously occupied colony (Szép and Møller 1999). For intensity of infestation we found a mean tick load of 0.03 in nestlings from sprayed nests, but 0.36 in control nests (Fig. 1). This almost tenfold difference in intensity was much greater than the twofold difference in tick intensity in nestlings from nests from a previously used colony (sprayed nests: 0.85 ticks per nestling, control nests: 1.60 ticks per nestling; Szép and Møller 1999). Thus, for sprayed nests there was a tenfold difference in prevalence between old and new colonies, while the difference for control nests only amounted to 38%. This difference in prevalence between treatments in the two types of colonies reflects the cost of parasitism due to nest re-use. Sand martins build long tunnels in sandy banks on rivers, lakes and similar places (Cramp 1988). The 13-g adult birds spend up to 14 days, on average 4.4 days, digging a nest hole with a mean length of 65 cm (Sieber 1980; Cramp 1988). Hence, construction of a new nest site is associated with a considerable delay in start of reproduction. The costs of this delay are balanced by a cost of increased parasitism in reused nests, as shown by Barclay (1988) and Møller (1990). Effects of treatment and breeding date on reproductive success and nestling phenotypes We found no significant difference in brood size before fledging between sprayed and control nests. This is similar to findings of a small nest-spraying experiment in Scotland (Alves 1997). However, our own previous studies of sand martins in a heavily tick-infested, old colony revealed a mean reduction in brood size by 21% between sprayed and control nests (Szép and Møller 1999). These differences in the effects of ticks on sand martins suggest that ticks mainly pose a serious problem to their hosts when abundant. However, this interpretation may need to be modified because reproductive success estimated before fledging could be a poor estimator of subsequent success at later stages after post-fledging mortality has taken its toll. An analysis of recapture probability of nestlings at the colony site after fledging suggested that the difference in nestling production between treatments actually increased after fledging, since twice as many nestlings were recaptured from sprayed as compared to control nests. Thus, small effects of parasites on host reproductive success may become en-
larged later when fledglings have to start acquiring their own food. Alternatively, dispersal propensity may differ among treatments, with parasite-free nests producing offspring that disperse less. We consider this hypothesis unlikely given that offspring were recaptured very early after fledging. Mean nestling phenotypes were only weakly influenced by parasite treatment. We found that nestlings from sprayed nests had slightly longer keels than nestlings from control nests, suggesting a depressing effect of parasites on skeletal growth. Alves (1997) found a reduction in nestling body mass of sand martins by 4.4% caused by ectoparasites in a small colony. Szép and Møller (1999) did not find significant differences in body mass or tarsus length of nestling sand martins among treatments in a large colony heavily infested with ticks. Hence, nestling sand martins generally show few signs of negative effects of ectoparasites on nestling size, irrespective of colony size and level of tick infestation. The small effects of ticks on nestling production and mean nestling phenotype are interesting per se, because they can be used to infer that hosts do not suffer mortality directly from the effects of parasitism. Any effects on host phenotype thus must be sub-lethal. Effects of treatment and breeding date on within-brood variance in nestling phenotypes Although nestlings showed little evidence of parasite effects on mean condition or size, there was clear evidence of an effect of ticks on within-brood variance in nestling phenotype. Within-brood variance in nestling skeletal size, as determined by keel length, and in nestling body mass, increased in the presence of ticks (Figs. 2, 3). Hence, variance in nestling size and condition changed in response to tick treatment, even though there were no significant effects on average nestling phenotype. Variation in nestling size and condition affects the position within the hierarchy of a brood, and this in turn will affect access to food and hence prospects for survival (review in Mock and Parker 1997). Christe et al. (1998) hypothesised that parent birds might sacrifice one or a small number of nestlings to parasites, when faced with the costs of parasitism. These so-called tasty chicks would suffer disproportionately from the negative effects of parasitism and thereby enhance the survival prospects of the remaining nestlings. Seen from the perspective of the tasty chick, sacrificial costs of parasitism would be selected if they were balanced by greater indirect fitness benefits through the survival of siblings. Which specific chick is the tasty subject for parasites may depend upon the relative level of induced and genetic resistance to parasites, or simply depend on which chick hatches first and hence becomes the target of parasitism. The present study demonstrated that such differences in phenotype within broods, which could have important consequences for the performance of nestlings, were present well
206
before any negative effects of parasites on reproductive success. The effects of parasite treatment on subsequent survival after fledging were tested using recapture information from the breeding colony. While there was only a weak and non-significant effect of ticks on offspring production, this effect was subsequently enhanced during the post-fledging stage. While this effect was not formally statistically significant, a difference in recapture probability after fledging compared to the small difference in brood size just before fledging suggests that small phenotypic effects before fledging may subsequently become magnified. Thus, while the relationship between size and reproductive success may be weak at the stage of fledging, it may become considerably stronger as soon as nestlings fledge. The peak of breeding activity in birds has been hypothesised to be timed so that maximum food demand of offspring coincides with peak food availability (Lack 1954, 1968; Martin 1987). This hypothesis has been supported by descriptive studies and numerous food supplementation experiments (Martin 1987). However, this seasonal decline in food availability is exacerbated by populations of ectoparasites increasing during the course of the breeding season (Burtt et al. 1991; de Lope and Møller 1993; Merino and Potti 1995). Since parasites exploit host resources, offspring will face increasing difficulties in terms of growth late during the breeding season. Parasites may thus have particularly severe, negative effects on their hosts late during the breeding season, as suggested by some studies (de Lope and Møller 1993; de Lope et al. 1993; Merino and Potti 1995). We investigated seasonal changes in the effects of tick infestations on reproductive performance in the sand martin, but found little evidence of such changes. In particular, there were no significant seasonal effects on mean or coefficient of variation in morphology. However, the sedimentation rate was significantly lower among late breeders. Sedimentation rate is an indirect estimate of protein content in the plasma, with immunoglobulins constituting an important component (Sharma et al. 1984). Hence, a high sedimentation rate reflects superior health status. A previous study of sand martins with experimentally manipulated levels of tick infestations has shown that concentrations of immunoglobulins in hosts are increased by ticks (Szép and Møller 1999). Thus, decreased sedimentation rate among late breeders may be a consequence of elevated concentrations of immunoglobulins. The apparent lack of seasonal effects on morphology may seem surprising. However, sand martins have highly synchronised reproduction within sub-colonies (Cramp 1988), and such synchronisation may reduce the negative effects of parasites late during the breeding season. In this context, it is perhaps interesting to note that highly colonial birds generally have highly synchronised reproduction (e.g. Gochfeld 1980; Brown and Brown 1996), and it is such species that particularly suffer from the negative effects of parasitism (Møller and Erritzøe 1996).
In conclusion, tick infestation in the highly colonial sand martin increased within-brood variance in size and condition without affecting the mean values. Since size differences among nestlings of a single brood affect access to food, parents may create slight differences in size to facilitate brood reduction if parasite infestations reach high levels. Acknowledgements We thank Zsolt Nagy for valuable help with field work, the participants of the AKCIO RIPARIA ringing camp for help with ringing, and the Nyíregyháza Local Chapter of the Hungarian Ornithological Society for providing tools and facilities for field and laboratory work. The study was carried out with the support provided by Magyary Zoltán posdoctorate and OTKA T29853 grants to T.S.
References Alves MAS (1997) Effects of ectoparasites on sand martin Riparia riparia nestlings. Ibis 139:494–496 Barclay RMR (1988) Variation in the costs, benefits, and frequency of nest reuse by barn swallows (Hirundo rustica). Auk 105:53–60 Brown CR, Brown MB (1996) Coloniality in the cliff swallow. University of Chicago Press, Chicago Burtt EH Jr, Chow W, Babbitt GA (1991) Occurrence and demography of mites of tree swallow, house wren and eastern bluebird nests. In: Loye EJ, Zuk M (eds) Bird-parasite interactions. Ecology, evolution and behaviour. Oxford University Press, Oxford, pp 104–122 Christe P, Møller AP, Lope F de (1998) Immunocompetence and nestling survival in the house martin: the tasty chick hypothesis. Oikos 83:175–179 Cramp S (ed) (1988) The birds of the Western Palearctic, vol 5. Oxford University Press, Oxford Gochfeld M (1980) Mechanisms and adaptive value of reproductive synchrony in colonial seabirds. In: Burger J, Olla BL, Winn HE (eds) Behavior of marine animals, vol 4. Marine birds. Plenum, New York, pp 207–270 Hart BL (1997) Behavioural defence. In: Clayton DH, Moore J (eds) Host-parasite evolution. General principles and avian models. Oxford University Press, Oxford, pp 59–77 Lack D (1954) The natural regulation of animal numbers. Oxford University Press, Oxford Lack D (1968) Ecological adaptations for breeding in birds. Methuen, London Lehmann T (1993) Ectoparasites: direct impact on host fitness. Parasitol Today 9:8–13 Lope F de, Møller AP (1993) Effects of ectoparasites on reproduction of their swallow hosts: a cost of being multi-brooded. Oikos 67:557–562 Lope F de, Møller AP, González G, Peréz JJ (1993) Increased detrimental effects of ectoparasites on their bird hosts during adverse environmental conditions. Oecologia 95:234–240 Loye JE, Zuk M (eds) (1991) Bird-parasite interactions. Oxford University Press, New York Martin TE (1987) Food as a limit on breeding birds: a life-history perspective. Annu Rev Ecol Syst 18:453–487 Merino S, Potti J (1995) Mites and blowflies decrease growth and survival in nestling pied flycatchers. Oikos 73:95–103 Mock DW, Parker GA (1997) The evolution of sibling rivalry. Oxford University Press, Oxford Møller AP (1990) Effects of parasitism by a haematophagous mite on reproduction in the barn swallow. Ecology 71:2345–2357 Møller, AP (1997) Parasitism and the evolution of host life history. In: Clayton DH, Moore J (eds) Host-parasite evolution. General principles and avian models. Oxford University Press, Oxford, pp 105–127
207 Møller AP, Erritzøe J (1996) Parasite virulence and host immune defense: host immune response is related to nest reuse in birds. Evolution 50:2066–2072 Saino N, Calza S, Møller AP (1997) Effects of a dipteran ectoparasite on immune response and growth trade-offs in barn swallow (Hirundo rustica) nestlings. Oikos 81:217–228 Sharma SC, Bhatia BB, Pachauri SP (1984) Blood cellular and biochemical studies in chicken experimentally infected with Toxoxara canis. Int J Parasitol 8:159–163 Sieber O (1980) Kausale und funktionelle Aspekte der Verteilung von Uferschwalbenbruten. Z Tierpsychol 52:19–56
Szép T (1995) Relationship between west African rainfall and the survival of the central European adult sand martin (Riparia riparia) population. Ibis 137:162–168 Szép T, Møller AP (1999) Cost of parasitism and host immune defence in the sand martin Riparia riparia: a role for parentoffspring conflict? Oecologia 119:9–15 Wakelin D, Apanius V (1997) Parasitism and the evolution of host life history. In: Clayton DH, Moore J (eds) Host-parasite evolution. General principles and avian models. Oxford University Press, Oxford, pp 30–58
