Song as an indicator of parasitism in the sedge warbler

ANIMAL BEHAVIOUR, 1999, 57, 307–314 Article No. anbe.1998.0969, available online at http://www.idealibrary.com on

Song as an indicator of parasitism in the sedge warbler K. L. BUCHANAN, C. K. CATCHPOLE, J. W. LEWIS & A. LODGE

School of Biological Sciences, Royal Holloway, University of London (Received 20 February 1998; initial acceptance 18 March 1998; final acceptance 28 August 1998; MS. number: 5796R)

We studied female choice and reproductive success in a marked population of sedge warblers, Acrocephalus schoenobaenus, from 1995 to 1996. Three genera of parasitic blood protozoans, namely Haemoproteus sp., Trypanosoma sp. and Plasmodium sp., were identified from blood samples taken from all breeding adults. Relatively high prevalence values of 19.5% in 1995 and 37.5% in 1996 were associated with increased levels of white blood cells relative to the number of red blood cells. Compared with nonparasitized males, parasitized males had significantly lower repertoire sizes in both years of the study; in one year, they also spent less time in song flights and weighed less. They also provisioned their broods at a lower rate. Parasitized females produced the same clutch size as nonparasitized females, although their broods were smaller at 7 days old. We suggest that haematozoan infections may reduce the expression of sexually selected song traits. Furthermore, such infections may influence the standard of parental care provided by males, although further research is needed to determine whether this is mediated through genetic resistance to parasitism or the effects of parasitism upon immediate body condition. 

success. A number of studies have shown that females prefer to mate with nonparasitized individuals (Borgia & Collis 1990; Clayton 1990; Hillgarth 1990; Zuk et al. 1990; Møller 1991a; Saino & Møller 1994), but others have failed to find any such association (Pruett-Jones et al. 1990; Weatherhead 1990; Dale et al. 1996). Furthermore, in order to separate the Hamilton–Zuk ‘good genes’ model from either the transmission avoidance model (Borgia 1986; Hamilton 1990; Able 1996) or the resourceprovisioning model (Hoelzer 1989; Kirkpatrick & Ryan 1991), both resistance and expression of the trait must be shown to be heritable. To date there are few convincing examples of indirect or direct benefits to females from mating with unparasitized males (Møller 1990a, b, 1991a; Hill 1991; Thompson et al. 1997). The role of host endocrinology in mediating a response to parasitic infection has recently attracted much attention, as Folstad & Karter (1992) suggested that testosterone may have a dual effect, increasing investment in sexually selected characteristics, but at the same time compromising the immune system. The association between steroid production and immunosuppression has been interpreted as demonstrating the cost of sexually selected traits, allowing them to reflect male quality accurately. Therefore, only superior males should be able to allocate resources to sexual display or ornamentation, without having detrimental effects upon their own immune system (Wedekind & Folstad 1994).

Although it was Darwin (1871) who first suggested that sexually selected traits could become exaggerated through continual female choice, the exact mechanisms are still poorly understood (Andersson 1994). Zahavi (1975, 1977) suggested that such traits exist because they indicate male fitness by acting as a handicap, thus preventing poor-quality males from producing an exaggerated ornament. Hamilton & Zuk (1982) extended this idea, suggesting that the expression of a sexually selected trait could specifically reflect a trade-off between sexual selection pressures and resistance to pathogens. A number of studies have offered limited support for the Hamilton–Zuk hypothesis (reviewed by Hamilton & Poulin 1997; Hillgarth & Wingfield 1997). There is considerable evidence that parasite infestation affects the standard of both plumage and display characteristics (Zuk et al. 1990; Møller et al. 1993; Sundberg 1995; Thompson et al. 1997). However, recent reviews (Read 1990; Clayton 1991; Hillgarth & Wingfield 1997) have pointed out that full support for the hypothesis requires not only that the parasite should affect the expression of the sexual trait, but also that it should affect host survival or reproductive Correspondence and present address: K. L. Buchanan, Department of Biological and Molecular Sciences, University of Stirling, Stirling FK9 4LA, U.K. (email: [email protected]). C. K. Catchpole, J. W. Lewis and A. Lodge are at the School of Biological Sciences, Royal Holloway University of London, Egham TW20 0EX, U.K. 0003–3472/99/020307+08 $30.00/0

1999 The Association for the Study of Animal Behaviour

307



1999 The Association for the Study of Animal Behaviour

308 ANIMAL BEHAVIOUR, 57, 2

Seasonal increases in testosterone are related to the development of sexual activity, but there is less evidence of a causal relationship between ornament development and testosterone levels (Owens & Short 1995). Recent studies have focused mainly on the effects of parasitism upon male plumage characteristics, which have been particularly associated with female choice (Dale et al. 1996; Potti & Merino 1996). Sexual displays are much more likely than morphological traits to be directly affected by immediate hormone levels (Alatalo et al. 1996), as the causal relationship between rising androgen levels and song production has been well established (Arnold 1975; Nottebohm et al. 1987; DeVoogd 1991). Song characteristics are also now recognized as sexually selected traits in a variety of species (Searcy & Andersson 1986; Catchpole 1987; Catchpole & Slater 1995). However, there has been little work examining the potential effects of parasitic infections on sexually selected visual or acoustic displays (e.g. Johnson & Boyce 1991; Ho ¨ glund et al. 1992). Song production, therefore, represents an excellent trait for examination of its role as an indicator in the trade-off between hormone production and the immune response. The only intraspecific study on song and parasitism showed that song rate in the barn swallow, Hirundo rustica, is affected by the presence of parasites (Møller 1991b; Saino et al. 1997), and that both song rate and song structure may be related to testosterone levels (Galeotti et al. 1997). Although there is some evidence that song rate may interact with male tail length to determine male paternity (Møller et al. 1998), there is no direct evidence that either song rate or song structure in the barn swallow is a sexually selected trait. In contrast, repertoire size in the sedge warbler, Acrocephalus schoenobaenus, is known to be a sexually selected male trait (Catchpole 1980; Catchpole et al. 1984). However, song flighting is also important in female choice, as males that spend longer song flighting on the day before pairing pair earlier in the season (Buchanan & Catchpole 1997). Therefore, parasitism should have a detectable effect upon male song in this species. More specifically, we predicted that parasitized males would have smaller repertoires and spend less time song flighting. It is also possible that song is used by females to indicate immediate male condition and therefore the standard of future male parental care. In this case, parasitized males would provide less parental care for their offspring. To test these three predictions, we examined the relationships between parasitic infection, sexually selected song traits and parental care in the sedge warbler. METHODS The study site was at Wraysbury Lakes, a small group of flooded gravel pits in Surrey, U.K. Vegetation consisted of low willow, Salix sp., and bramble, Rubus sp., scrub with access to nearby standing water from all territories. We followed breeding adult sedge warblers as part of a study on breeding success and mate choice (Buchanan & Catchpole 1997). In 1995 and 1996 we caught both male and female sedge warblers as soon as possible

after their arrival at the site and marked them with an individual combination of coloured leg rings. Displaying males were followed from their arrival at the site through to pairing and their subsequent breeding success recorded. Using a Marantz tape recorder and a Sennheiser microphone mounted on a parabolic reflector, we recorded a 15-min sample of song for all males within a few days of their arrival at the breeding site. All recordings were made within the first 5 h of daylight and analysed using a Kay DSP Sonagraph. From this sample we analysed the first 20 songs that were clear and uninterrupted by neighbouring singers. The song of the sedge warbler is enormously variable in structure and length, and is composed of a varying number of different syllable types. The total number of different syllable types within the sample of 20 songs was used as an overall measure of syllable repertoire size. Sampling of syllable types beyond this has shown that the rate of appearance of novel syllable types reaches an asymptote at approximately 10 songs (see Catchpole 1980). Analysis of recordings made on consecutive days for the same male has shown that this measure of repertoire size has more than 98% repeatability (Lessells & Boag 1987). We quantified the song-flighting performance of each male by daily time budget studies. All observations were carried out within the first 4 h of daylight and all males were visited daily between arrival and pairing. Within a 10-min sample we recorded the total amount of time (s) spent song flighting, and the number of song flights. We used these results to calculate the mean number of song flights and the mean time spent song flighting between arrival and pairing, as well as the number of song flights and the amount of time spent song flighting on the day before pairing. Although there is a lot of diurnal variation in individual song-flighting behaviour, observations carried out on the same individual on the same day all have greater than 95% repeatability (Lessells & Boag 1987). We collected blood smears from breeding adults (41 individuals in 1995 and 32 in 1996). One smear was prepared from each individual at the time of capture. We took approximately 100 ìl of blood from each individual from the brachial vein, and used a small drop to prepare a slide. We prepared all slides by touching the end of a 100-ìl capillary tube against the glass slide, to deliver a similar volume each time. We then used a coverslip to push the blood across the slide. Slides were air dried and fixed in 100% methanol for 1 min, within 24 h of sampling, and were subsequently stained with Geimsa (pH 7.2) for 1 h, rinsed, dried, and stored at room temperature until examination. We examined each slide under oil immersion at 1000 magnification, and determined the presence of parasites from the genera Trypanosoma, Haematoproteus and Plasmodium by randomly scanning across the smear for at least 100 fields. Blood counts were carried out at five points on the smear to give a general picture of the blood composition. We estimated the number of cells within a field of view (calculated as 25 ìm2) for red blood cells, lymphocytes,

BUCHANAN ET AL.: SONG AND PARASITISM 309

RESULTS

0.016 White blood cells:red blood cells

monocytes, thrombocytes and macrophages, and then averaged the number of cells of each type for all five counts. All types of white blood cells were pooled for the final analysis. As the volume of blood used in a smear is difficult to standardize, we expressed counts of white blood cells as the ratio of white cells to the number of red blood cells. This protocol has now been used in a number of studies and has given repeatablity estimates of over 90% for both parasite prevalence and blood composition (Ratti et al. 1993; Sundberg 1995). As the intensity of blood parasite infection can vary with time postinfection and also the stage of the breeding season, we classified individual birds as parasitized and nonparasitized hosts. We used MINITAB and Statview 4.01 for all statistical analyses.

0.012

0.008

0.004

0

Parasitized

Nonparasitized

Parasite Infection

Figure 1. The ratio of white blood cells to red blood cells (X+SE) in parasitized and nonparasitized sedge warblers from 1995 and 1996 combined. Mann–Whitney U test: U=256, N1 =21, N2 =53, P<0.001.

In 1995, 19.5% (eight of 41 birds) of individuals were infected with one or more of the three haematozoans (and so classed as parasitized), with prevalence values of 17.1, 9.5 and 2.4% being recorded for Plasmodium sp., Trypanosoma sp. and Haemoproteus sp., respectively. In 1996 a higher overall prevalence of 37.5% (12 of 32 birds) of individuals was observed, but Plasmodium sp. (12.5%) was no longer the most abundant parasite in this year as prevalence values of 15.6% were recorded for both Haemoproteus sp. and Trypanosoma sp. The prevalences of haematozoans in males and females were similar, with 16.7% of males and 23.5% of females in 1995 and 35% of males and 41.7% of females in 1996 being parasitized. There was no significant difference between the red blood cell counts of parasitized and nonparasitized birds in either 1995 or 1996 (Mann–Whitney U test: 1995: U=72.5, N1 =33, N2 =8, P=0.133; 1996: U=117, N1 =21, N2 =12, P=0.907). However, parasitized birds had significantly elevated numbers of white blood cells, relative to the number of red cells in the sample (Fig. 1). This was also true when considering 1995 and 1996 separately, as in both years parasitized individuals had a relatively higher ratio of white cells to red cells (Mann–Whitney U test: 1995: U=66.0, N1 =33, N2 =8, P=0.029; 1996: U=29.0, N1 =21, N2 =12, P<0.001). Furthermore, this was also true when pooling both years, but considering the sexes separately (Mann–Whitney U test: males: U=66.0, N1 =28, N2 =10, P=0.014; females: U=61.0, N1 =25, N2 =10, P=0.019). There was a significant difference in the relative number of white blood cells between individuals carrying no parasites, or one, two or three types of infection in both 1995 and in 1996 (ANOVA: 1995: F3,40 =3.6, P=0.022; 1996: F2,31 =7.18, P=0.003). Whilst this effect could be due to the extra load of carrying multiple infections, it is also possible that certain parasites are particularly detrimental to the host. For this reason, we also considered the effects of the three types of parasite, pooling samples from 1995 and 1996 because of small sample sizes. A general linear model was run on square

root-transformed data, with the number of infection types a continuous independent variable, the ratio of white blood cells as the dependent variable and with the infection intensities (number of parasites in 100 fields) of the three parasite types as covariates. The model showed that the intensity of infection by any one parasite type had no significant effect on the relative number of white blood cells (Plasmodium sp.: F1,72 =0.7, P=0.407; Haemoproteus sp.: F1,72 =0.93, P=0.338; Trypanosoma sp.: F1,72 =0.33, P=0.569). However, the number of different infections did have a significant effect (F1,72 =6.37, P=0.014) when nonparasitized individuals were included, but not when they were excluded (F1,19 =0.76, P=0.397). Including sex and year in this model showed that neither factor had a significant effect on the relative number of white blood cells (sex: F1,16 =0.21, P=0.659; year: F1,16 =0.03, P=0.873). It seems likely, therefore, that within this sample, the only significant differences in host physiology are observed between nonparasitized and parasitized individuals and, for this reason, we continued this approach in the analysis. In 1996, the return rate of individual birds from 1995 was 16.6%. Parasitized and nonparasitized individuals were equally likely to return to breed the next year (÷21 =2.56, P=0.15). Nest predation rates were greater than 60% in both 1995 and 1996. In 1995 only 47% of males fledged any young during the season, whilst in 1996 the figure was 20%. Owing to the high rate of nest predation, we could not compare the reproductive success of parasitized and nonparasitized individuals, as sample sizes were much reduced. There was no difference in age between parasitized and nonparasitized individuals in either 1995 or 1996 (Mann–Whitney U test: 1995: U=58.0, N1 =5, N2 =24, P=0.86; 1996: U=39.0, N1 =9, N2 =11, P=0.386), and no significant differences between parasitized and nonparasitized males in either arrival date (Mann– Whitney U test: 1995: U=17.0, N1 =3, N2 =12, P=0.885; 1996: U=23.0, N1 =7, N2 =8, P=0.558), or pairing date (1995: U=10.0, P=0.246; 1996: U=22.0, P=0.794).

310 ANIMAL BEHAVIOUR, 57, 2

Parasites and Song

80

Combining the 2 years, parasitized males had significantly smaller repertoires than nonparasitized males (Mann–Whitney U test: U=47.5, parasitized N=9, nonparasitized N=20, P=0.045). This relationship remained significant when considering the 2 years separately (Fig. 2), offering strong support for our prediction that parasitized males should have smaller repertoires. There was no significant difference between parasitized and nonparasitized males in the mean time spent song flighting between arrival and pairing (Mann–Whitney U test: 1995: U=8.0, N1 =3, N2 =11, P=0.186; 1996: U=8.0, N1 =7, N2 =8, P=0.817). However, in 1996, parasitized males spent significantly less time song flighting on the day before pairing than nonparasitized males (Fig. 3a) and also weighed less (Fig. 3b), although this was not quite significant when we used residual body mass, controlling for wing length (Mann–Whitney U test: U=12.0. N1 =7, N2 =8, P=0.064). Neither relationship was significant in 1995 (Mann–Whitney U test: song flighting: U=14.0. N1 =3, N2 =11, P=0.697; male weight: U=12.0, N1 =3, N2 =12, P=0.885). When we pooled data from males from both 1995 and 1996, there was a significant correlation between the relative number of white blood cells and residual body weight (Spearman rank correlation: rS =0.394, N=38, P=0.016).

75

(a)

70 65 60 55

Repertoire size

50 45 40 80

(b)

75 70 65 60 55 50

Parasites and Parental Effort Parental provisioning rates were available for 11 males from 1995 and for seven males from 1996, enabling us to test the prediction that parasitized males should provide a significantly lower standard of parental care. When we combined the 2 years, there was a nonsignificant tendency for parasitized males to provide a smaller share of the parental feeding visits than nonparasitized males (Mann–Whitney U test: U=19.5, N1 =N2 =9, P=0.063). However, there was a significant difference in absolute male provisioning rate when we controlled for brood size (Fig. 4). Considering 1995 and 1996 together, there was no significant difference between the weights of parasitized and nonparasitized females (Mann–Whitney U test: U=45.0, N1 =5, N2 =19, P=0.859). However, many females were weighed during the egg-laying period when variation in weight could have masked any effects of overall body condition. There was no significant difference in the first clutch size of parasitized and nonparasitized females from 1995 and 1996 (Mann–Whitney U test: U=2.5, N1 =2, N2 =8, P=0.1). However, parasitized females had significantly smaller broods at 7 days old than nonparasitized females (Fig. 5).

DISCUSSION In general, our results are consistent with predictions from the Hamilton–Zuk hypothesis. The higher ratio of

45 40

Parasitized

Nonparasitized

Figure 2. Repertoire sizes (X+SE) of parasitized and nonparasitized males (a) 1995; Mann–Whitney U test: U=3.0, N1 =3, N2 =12, P<0.05. (b) 1996; Mann–Whitney U test: U=8.0, N1 =6, N2 =8, P<0.05.

white blood cells to red cells in parasitized individuals shows that the parasitic infections observed are associated with an immune response, at some cost to the individual. The use of cell ratios can present some problems if both red and white cell counts are altered by infection. However, it seems in this case that the parasitic infections considered had little effect on individual red cell counts. Considering ratios of white:red blood cells would also present problems in interpretation if dramatically different volumes of blood were used in smear preparation. However, all smears were prepared according to a standardized protocol. The dramatic difference between the blood picture of parasitized and nonparasitized individuals is therefore more likely to be due to an associated immune response than sampling error. More specifically, the incidence of infection covaried with the expression of two sexually selected song traits, which have been shown to be important in female choice (Buchanan & Catchpole 1997). There is also some evidence that female choice for such traits might be linked to parental care, as parasitized males provisioned their offspring at a lower rate.

BUCHANAN ET AL.: SONG AND PARASITISM 311

7

(a)

25

6

20

5 Brood size

Song flighting (s)

30

15 10

2

0

1 0 (b)

11.5

11

10.5

Parasitized

Nonparasitized

Figure 3. (a) The amount of time spent song flighting (X+SE) on the day before pairing in 1996 for parasitized and nonparasitized males. Mann–Whitney U test: U=11.0, N1 =7, N2 =8, P<0.05. (b) Weight in 1996 for parasitized and nonparasitized males. Mann–Whitney U test: U=9.5, N1 =7, N2 =8, P<0.05.

3

2.5

2

1.5

1

0.5

0

Parasitized

Nonparasitized

Figure 5. Brood sizes (X+SE) of parasitized and nonparasitized females, pooled from 1995 and 1996. Mann–Whitney U test: U=4.0, N1 =3, N2 =11, P<0.05.

12 Male weight (g)

3

5

12.5

Number of feeds by male/chick/h

4

Parasitized

Nonparasitized

Figure 4. Feeding rate (mean number of feeds to the nest/chick/ h+SE) for parasitized and nonparasitized males from 1995 and 1996 combined. Mann–Whitney U test: U=18.0, N1 =N2 =9, P<0.05.

For parasitic infections to affect the expression of sexually selected traits, they must impose costs on the host. The relatively higher levels of white blood cells in parasitized individuals suggests a cell-mediated response to infection. Although little is known about the costs of mounting such a response (Sheldon & Verhulst 1996; Hillgarth & Wingfield 1997), the transfer of resources into the immune system presumably occurs at the expense of other metabolic activities (Festa-Bianchet 1989; Gustafsson et al. 1994; Norris et al. 1994; Richner et al. 1995). In 1996, parasitized males were lighter than their nonparasitized conspecifics, which may indicate their relatively poor condition. Covariance of male parasitism and condition does not provide conclusive proof for a direct cost of infection. It is possible that only poor quality individuals are parasitized and that the detrimental physiological conditions observed are not a direct result of infection. Experimental manipulation of an immune response involving new antigens would confirm the direct effects of such an infection. However, the covariance of a relatively enhanced immune response with parasitic infection is a strong indication that our first prediction is upheld: parasitized individuals incur a physiological cost of infection. In support of the more specific predictions of this study, both repertoire size and song flighting differed significantly between parasitized and nonparasitized males. Our results are therefore consistent with predictions from the Hamilton–Zuk hypothesis, that male resistance and sexually selected traits should covary. Repertoire size correlates consistently with male pairing success in the sedge warbler (Catchpole 1980; Buchanan & Catchpole 1997). By choosing a male with a large repertoire, females are also choosing males who are better able to mount an immune response. However, full support for the Hamilton–Zuk hypothesis requires not only that the selected trait and resistance to parasites are linked, but also that they are heritable (Clayton 1991; Hillgarth &

312 ANIMAL BEHAVIOUR, 57, 2

Wingfield 1997). Owing to the low return rates to the breeding area, this study can provide no evidence for the heritability of resistance to blood parasites. Cross-fostering experiments with barn swallows suggest that the genetic quality of the parent may well affect the resistance of the offspring to parasitism (Møller 1990a). Saino & Møller (1996) showed that in swallows with artificially lengthened tails the immune response was reduced compared with those with shortened or unmanipulated tails. Although this relationship falls in the direction predicted by the immunocompetence handicap hypothesis (Folstad & Karter 1992), it does not demonstrate that the cost of the trait directly affects the immune system. Further research, using reliable tests of individual immunocompetence (Lochmiller 1995), is now needed to reveal whether there are genetic associations between sexually selected traits and the vigour of the immune system. Our study also identified a difference in the amount of time spent song flighting between parasitized and nonparasitized males, although this effect appeared in only 1 of the 2 years. Song flighting has been described as one of the most energetically expensive sexual displays (Hedenstro ¨ m & Møller 1992). We also know that song flighting in the sedge warbler is correlated with female choice (Buchanan & Catchpole 1997) and may well be used as an indicator of male condition. Without evidence for heritable fitness (Hillgarth 1990; Møller 1990a), it is always possible that male traits merely indicate high phenotypic quality. In support of our third prediction we found that nest-provisioning rates were lower for parasitized male sedge warblers than for nonparasitized males. This supports the possibility that female choice for male sexual traits is based on selection of direct benefits because such males provide better parental care or are better able to defend territorial resources (Hoelzer 1989; Milinski & Bakker 1990). Møller (1994) found that male barn swallows with nests inoculated with haematophagous mites fed their offspring less than males with unparasitized nests. For sedge warbler chicks, a lower male provisioning rate is likely to affect fledging weight and the viability of offspring, as has been found for other species (Bart & Tornes 1989). Furthermore, as females may increase their provisioning rate in response to a compromised partner (Wright & Cuthill 1989), there may be an additional cost to females. For repertoire size to be an honest indicator of male fitness, there must be some cost associated with developing or producing a large repertoire. Within the genus Acrocephalus, there is a correlation between the size of the song nuclei (HVC) in the brain and repertoire size (Sze´kely et al. 1996). As yet, the costs of the production and maintenance of the song nuclei are unknown, although HVC size may reflect the cost of producing a large repertoire (Catchpole 1996). There is now some evidence that a parasitic infection can directly affect certain cognitive functions (Kavaliers et al. 1995) although there has been no investigation into whether this is due to the poor development of associated brain nuclei.

Although little is known about the cost of developing and maintaining a complex repertoire, the high cost of song flighting is obvious. It is strange, therefore, that the more robust result from this study should be between repertoire size and parasitic infection. However, songflighting effort is also correlated with environmental conditions (unpublished data) and in 1995 (in contrast to 1994 and 1996) song flighting did not correlate with female choice (Buchanan & Catchpole 1997). It is possible, therefore, that in this year it was not representative of male condition. However, it is also possible that repertoire size represents male condition over the longer term, whilst song flighting indicates more immediate physiological condition (Buchanan & Catchpole 1997). If this is so, other factors, including environmental conditions, may mask the short-term effects of parasitic infections upon male condition, making the effects of parasitism upon song-flighting effort less detectable. In our study, parasitized females reared smaller broods than nonparasitized females. This may reflect a poorer body condition causing lower investment in egg production or rearing their young. Alternatively, parasitized females may be low-quality individuals and may themselves be paired with low-quality males. Without experimental manipulation, the causal relationship between parasitism and reduced fledging success of offspring is unclear. Siika¨maki et al. (1997) found that female pied flycatchers, Ficedula hypoleuca, showed no reduction in clutch size when suffering from haematozoan infections. However, they did find that infected parents showed an increase in infection intensity when their clutch size was enlarged, implying a trade-off between current reproductive effort and investment in immune functions. Within certain populations parasitism rates may increase with age (Andersson & Gordon 1982). Furthermore, as older individuals may have greater expression of secondary sexual traits, it could be argued that the relationship between the expression of the trait and parasite prevalence occurs as a result of an age effect (Thomas et al. 1995). In the sedge warbler there is a relationship between age and repertoire size (Birkhead et al. 1997); however, in our study there was no direct relationship between age and the prevalence of haematozoan parasites. A recent study of the great reed warbler, Acrocephalus arundinaceus, has shown that offspring of males with large repertoires have higher survival rates than offspring of males with smaller repertoires (Hasselquist et al. 1996). However, there is as yet little evidence for the heritability of repertoire size, which would be necessary for the coevolution of resistance to parasitism and the trait. It is interesting to speculate whether a parasitic infection might affect song repertoire learning, song production, or both. However, despite the lack of evidence for the heritability of repertoire size, our study has shown that haematozoan infections covary with the expression of sexually selected song traits and individual investment in reproduction. The close link between song production and hormonal and neural changes makes bird song a prime candidate for a trait that may signal male quality by an immunocompetence handicap, and help to

BUCHANAN ET AL.: SONG AND PARASITISM 313

explain why repertoire size can honestly reflect male quality. Acknowledgments We thank the Runnymede Ringing Group for assistance in the field, Professor Frank Cox for assistance in the identification of the haematozoan parasites and also two anonymous referees whose comments greatly improved the manuscript. This research was supported by a grant to C.K.C. from the Natural Environmental Research Council. References Able, D. J. 1996. The contagion indicator hypothesis for parasitemediated sexual selection. Proceedings of the National Academy of Sciences U.S.A., 93, 2229–2233. Alatalo, R., Ho ¨ glund, J., Lundberg, A., Rintama¨ki, P. T. & Silverin, B. 1996. Testosterone and male mating success on the black grouse leks. Proceedings of the Royal Society of London, Series B, 263, 1697–1702. Andersson, M. B. 1994. Sexual Selection. Princeton, New Jersey: Princeton University Press. Andersson, R. M. & Gordon, D. M. 1982. Processes influencing the distribution of parasite numbers within host populations with special emphasis on parasite-induced host mortalities. Parasitology, 85, 373–398. Arnold, A. P. 1975. The effects of castration and androgen replacement of song, courtship, and aggression in zebra finches (Poephila guttata). Journal of Experimental Zoology, 191, 309–326. Bart, J. & Tornes, A. 1989. Importance of mongamous male birds in determining reproductive success. Behavioral Ecology and Sociobiology, 24, 109–116. Birkhead, T. R., Buchanan, K. L., DeVoogd, T., Pellatt, E. J., Sze´kely, T. & Catchpole, C. K. 1997. Song, sperm quality and testes asymmetry in the sedge warbler, Acrocephalus schoenobaenus. Animal Behaviour, 53, 965–971. Borgia, G. 1986. Satin bowerbird parasites: a test of the bright male hypothesis. Behavioral Ecology and Sociobiology, 19, 335–358. Borgia, G. & Collis, K. 1990. Parasites and bright male plumage in the satin bowerbird (Ptilorhynchus violaceus). American Zoologist, 30, 279–286. Buchanan, K. L. & Catchpole, C. K. 1997. Female choice in the sedge warbler Acrocephalus schoenobaenus: multiple cues from song and territory quality. Proceedings of the Royal Society of London, Series B, 264, 521–526. Catchpole, C. K. 1980. Sexual selection and the evolution of complex songs among warblers of the genus Acrocephalus. Behaviour, 74, 149–166. Catchpole, C. K. 1987. Bird song, sexual selection and female choice. Trends in Ecology and Evolution, 2, 94–97. Catchpole, C. K. 1996. Song and female choice: good genes and big brains. Trends in Ecology and Evolution, 11, 358–360. Catchpole, C. K. & Slater, P. J. B. 1995. Bird Song. Biological Themes and Variations. Cambridge: Cambridge University Press. Catchpole, C. K., Dittami, J. & Leisler, B. 1984. Differential responses to male song repertoires in female songbirds implanted with oestrodiol. Nature, 312, 563–564. Clayton, D. H. 1990. Mate choice in parasitized rock doves: lousy males lose. American Zoologist, 30, 251–262. Clayton, D. H. 1991. The influence of parasites on host sexual selection. Parasitology Today, 7, 329–334.

Dale, S., Kruszewicz, A. & Slagsvold, T. 1996. Effects of blood parasites on sexual and natural selection in the pied flycatcher. Journal of Zoology, 238, 373–393. Darwin, C. 1871. The Descent of Man and Selection in Relation to Sex. London: Murray. DeVoogd, T. J. 1991. Endocrine modulation of the development and adult function of the avian song system. Psychoneuroendocrinology, 16, 41–66. Festa-Bianchet, M. 1989. Individual differences, parasites and the cost of reproduction for bighorn ewes (Ovis canadensis). Journal of Animal Ecology, 58, 785–795. Folstad, I. & Karter, A. J. 1992. Parasites, bright males and the immunocompetence handicap. American Naturalist, 139, 603– 622. Galeotti, P., Saino, N., Sacchi, R. & Møller, A. P. 1997. Song correlates with social context, testosterone and body condition in male barn swallows. Animal Behaviour, 53, 687–700. Gustafsson, L., Nordling, D., Andersson, M. S., Sheldon, B. C. & Qvarnstrom, A. 1994. Infectious diseases, reproductive effort and the cost of reproduction in birds. Philosophical Transactions of the Royal Society, Series B, 346, 323–331. Hamilton, W. D. 1990. Mate choice: near or far. American Zoologist, 30, 341–352. Hamilton, W. D. & Poulin, R. 1997. The Hamilton and Zuk hypothesis revisited: a meta-analytical approach. Behaviour, 134, 299–320. Hamilton, W. D. & Zuk, M. 1982. Heritable true fitness and bright birds: a role for parasites. Science, 218, 384–387. Hasselquist, D., Bensch, S. & von Schantz, T. 1996. Correlation between male song repertoire, extra-pair paternity and offspring survival in the great reed warbler. Nature, 381, 229–232. Hedenstro ¨ m, A. & Møller, A. P. 1992. Morphological adaptations to song flight in passerine birds: a comparative study. Proceedings of the Royal Society of London, Series B, 247, 183–187. Hill, G. 1991. Plumage coloration is a sexually selected indicator of male quality. Nature, 350, 337–339. Hillgarth, N. 1990. Parasites and female choice in the ring-necked pheasant. American Zoologist, 30, 227–233. Hillgarth, N. & Wingfield, J. C. 1997. Parasite-mediated sexual selection: endocrine aspects. In: Host–Parasite Evolution. General Principles and Avian Models (Ed. by D. H. Clayton & J. Moore), pp. 78–104. Oxford: Oxford University Press. Hoelzer, G. A. 1989. The good parent process of sexual selection. Animal Behaviour, 38, 1067–1078. Ho ¨ glund, J., Alatalo, R. V. & Lundberg, A. 1992. The effects of parasites on male ornaments and female choice in the lekbreeding black grouse (Tetrao tetrix). Behavioral Ecology and Sociobiology, 30, 71–76. Johnson, L. L. & Boyce, M. S. 1991. Female choice of males with low parasite loads in the sage grouse. In: Bird–Parasite Interactions. Ecology, Evolution and Behaviour (Ed. by J. E. Loye & M. Zuk), pp. 377–388 Oxford: Oxford University Press. Kavaliers, M., Colwell, D. D. & Galea, L. A. M. 1995. Parasitic infection impairs spatial learning in mice. Animal Behaviour, 50, 223–229. Kirkpatrick, M. & Ryan, M. 1991. The evolution of mating preferences and the paradox of the lek. Nature, 350, 33–39. Lessells, C. M. & Boag, P. T. 1987. Unrepeatable repeatabilities: a common mistake. Auk, 104, 116–121. Lochmiller, R. L. 1995. Testing the immunocompetence handicap theory. Trends in Ecology and Evolution, 10, 372–373. Milinski, M. & Bakker, T. C. M. 1990. Female sticklebacks use male colouration in mate choice and hence avoid parasitized males. Nature, 344, 330–333.

314 ANIMAL BEHAVIOUR, 57, 2

Møller, A. P. 1990a. Effects of a haematophagous mite on the barn swallow (Hirundo rustica): a test of the Hamilton and Zuk hypothesis. Evolution, 44, 771–784. Møller, A. P. 1990b. Effects of parasitism by the haematophagous mite (Ornithonyssus bursa) on reproduction in the barn swallow (Hirundo rustica). Ecology, 71, 2345–2357. Møller, A. P. 1991a. Parasites, sexual ornaments and mate choice in the barn swallow. In: Bird–Parasite Interactions: Ecology, Evolution and Behaviour (Ed. by J. E. Loye & M. Zuk), pp. 328–348. Oxford: Oxford University Press. Møller, A. P. 1991b. Parasite load reduces song output in a passerine bird. Animal Behaviour, 41, 723–730. Møller, A. P. 1994. Parasite infestation and parental care in the barn swallow Hirundo rustica: a test of the resource-provisioning model of parasite-mediated sexual selection. Ethology, 97, 215– 225. Møller, A. P., Dufva, R. & Allander, K. 1993. Parasites and the evolution of host social behaviour. Advances in the Study of Behavior, 22, 65–102. Møller, A.P., Saino, N., Taramino, G., Galeotti, P. & Ferrario, S. 1998. Paternity and multiple signaling: effects of a secondary sexual character and song on paternity in the barn swallow. American Naturalist, 151, 236–242. Norris, K., Anwar, A. & Read, A. F. 1994. Reproductive effort influences the prevalence of haematozoan parasites in great tits. Journal of Animal Ecology, 63, 601–610. Nottebohm, F., Nottebohm, M. E., Crane, L. A & Wingfield, J. C. 1987. Seasonal changes in gonadal hormone levels of adult male canaries and their relation to song. Behavioral and Neural Biology, 47, 197–211. Owens, I. P. F. & Short, R. V. 1995. Hormonal basis of sexual dimorphism in birds: implications for new theories of sexual selection. Trends in Ecology and Evolution, 10, 44–47. Potti, J. & Merino, S. 1996. Decreased levels of blood trypanosome infection correlate with female expression of a male secondary sexual trait: implications for sexual selection. Proceedings of the Royal Society of London, Series B, 263, 1199–1204. Pruett-Jones, S. G., Pruett-Jones, M. A. & Jones, H. I. 1990. Parasites and sexual selection in birds of paradise. American Zoologist, 30, 287–298. Ratti, O., Dufva, R. & Alatalo, R. V. 1993. Blood parasites and male fitness in the pied flycatcher. Oecologia, 96, 410–414. Read, A. F. 1990. Parasites and the evolution of host sexual behaviour. In: Parasitism and Host Behaviour (Ed. by C. J. Barnard & J. M. Behnke). pp. 117–157. London: Taylor & Francis. Richner, H., Christe, P. & Oppliger, A. 1995. Paternal investment affects prevalence of malaria. Proceedings of the National Academy of Sciences U.S.A., 92, 1192–1194.

Saino, N. & Møller, A. 1994. Secondary sexual characters, parasites and testosterone in the barn swallow, Hirundo rustica. Animal Behaviour, 48, 1325–1333. Saino, N. & Møller, A. P. 1996. Sexual ornamentation and immunocompetence in the barn swallow. Behavioral Ecology, 7, 227– 232. Saino, N., Galeotti, P., Sacchi, R. & Møller, A. P. 1997. Song and immunological condition in male barn swallows (Hirundo rustica). Behavioral Ecology, 8, 364–371. Searcy, W. A. & Andersson, M. 1986. Sexual selection and the evolution of song. Annual Review of Ecology and Systematics, 17, 507–533. Sheldon, B. C. & Verhulst, S. 1996. Ecological immunology: costly parasite defences and trade-offs in evolutionary ecology. Trends in Ecology and Evolution, 11, 317–321. Siika¨maki, P., Ra¨tti, O., Hovi, M. & Bennett, G. F. 1997. Association between haematozoan infection and reproduction in the pied flycatcher. Functional Ecology, 11, 176–183. Sundberg, J. 1995. Parasites, plumage coloration and reproductive success in the yellowhammer, Emberiza citrinella. Oikos, 74, 331– 339. Sze´kely, T., Catchpole, C. K., DeVoogd, A., Marchl, Z. & DeVoogd, T. J. 1996. Evolutionary changes in a song control area of the brain (HVC) are associated with evolutionary changes in song repertoire among European warblers (Sylviidae). Proceedings of the Royal Society of London, Series B, 263, 607–610. Thomas, F., Renaud, F., Meeus, T. De & Ce´zilly, F. 1995. Parasites, age and the Hamilton-Zuk hypothesis: inferential fallacy? Oikos, 74, 305–309. Thompson, C. W., Hillgarth, N., Leu, M. & McClure, H. E. 1997. High parasite load in house finches (Carpodacus mexicanus) is correlated with reduced expression of a sexually selected trait. American Naturalist, 149, 270–294. Weatherhead, P. J. 1990. Secondary sexual traits, parasites and polygyny in red-winged blackbirds, Agelaius phoeniceus. Behavioral Ecology, 1, 125–130. Wedekind, C. & Folstad, I. 1994. Adaptive or nonadaptive immunosuppression by sex hormones? American Naturalist, 143, 936– 938. Wright, J. & Cuthill, I. 1989. Manipulation of sex differences in parental care. Behavioral Ecology and Sociobiology, 25, 171–181. Zahavi, A. 1975. Mate selection: selection for a handicap. Journal of Theoretical Biology, 53, 205–214. Zahavi, A. 1977. The cost of honesty (further remarks on the handicap principle). Journal of Theoretical Biology, 67, 603–605. Zuk, M., Thornhill, R., Ligon, J. D. & Johnson, K. 1990. Parasites and mate choice in red jungle fowl. American Zoologist, 30, 235–244.