JPE593.fm Page 378 Wednesday, April 11, 2001 10:28 AM

Journal of Applied Ecology 2001 38, 378 – 389

The effect of dairy farming on barn swallow Hirundo rustica abundance, distribution and reproduction Blackwell Science, Ltd

ANDERS PAPE MØLLER Laboratoire d’Ecologie, CNRS UMR 7625, Université Pierre et Marie Curie, 7 quai St. Bernard, Case 237, F-75252 Paris Cedex 05, France

Summary 1. Recent changes in agricultural practice have been hypothesized to affect the abundance and reproductive success of farmland birds. The influence of dairy farming on barn swallows Hirundo rustica was investigated by comparison of their abundance, phenotype and reproduction on the same farms before and immediately after dairy farming ceased, while a control sample of farms without change in farming practice in the same years was used to check for temporal changes unrelated to farming practice. 2. The abundance of barn swallows decreased significantly when dairy farming ceased, with an average reduction of 48%, while there was no significant difference in the sample of control farms. This was mainly due to a decrease in the abundance of yearling immigrants. The abundance of insect food measured with sweep nets decreased significantly in the absence of cattle, while there was no significant change in the sample of control farms. 3. The mean phenotypes of adult barn swallows breeding on the same farms with and without dairy farming did not differ significantly for any of the 16 variables measured (11 morphological variables, body mass, parasite load, haematocrit, leucocyte counts, and arrival date), nor were there any significant differences recorded in the sample of control farms. 4. The mean phenotypes of barn swallow nestlings on the same farms with and without dairy farming differed for tarsus length, body mass, haematocrit, leucocyte concentration and T-cell mediated immune response, with nestlings being of poorer quality in the absence of cattle. There was no significant differences recorded in the sample of control farms. 5. Start of reproduction was not delayed in the absence of cattle, but size of first clutches was reduced by the absence of cattle. The frequency of second clutches decreased in the absence of cattle, and laying of second clutches was also delayed. Hatching, fledging and breeding success did not differ between the two types of farming practice. This led to an overall reduction in annual reproductive success in the absence of cattle. None of these significant differences was recorded in the sample of control farms. 6. These observations suggest that termination of dairy farming reduces local population size, reproductive success and the quality of offspring produced. There is little evidence of the distribution of phenotypes of adult barn swallows being affected by the presence of cattle. Key-words: agricultural practice, body condition, clutch size, distribution of phenotypes, health, immune response, laying date, population size. Journal of Applied Ecology (2001) 38, 378–389

Introduction Farmland habitats cover most of Europe, yet the quality of farmland as a habitat for reproduction of birds and other animals has only been studied intensively in recent years. It has been hypothesized that changes in © 2001 British Ecological Society

Correspondence: Anders Pape Møller (fax + 33 1 44 27 35 16; e-mail [email protected]).

farming practice have influenced the distribution and abundance of many European birds (Tucker & Heath 1994; Hagemeijer & Blair 1997; Pain & Pienkowski 1997). Long-term studies of changes in crops and the decline in availability of extensively used parts of the agricultural landscape have been associated with dramatic changes in local abundance of birds (Marchant et al. 1990; Petersen & Jacobsen 1997). This has been more noticeable in detailed population studies than in

JPE593.fm Page 379 Wednesday, April 11, 2001 10:28 AM

379 Dairy farming and barn swallows

© 2001 British Ecological Society, Journal of Applied Ecology, 38, 378– 389

standard bird census work (Møller 1989; Petersen & Jacobsen 1997). Comparisons of the abundance of birds on intensively farmed land and organic farms have demonstrated dramatic differences in bird abundance and diversity (Christensen, Jacobsen & Nøhr 1996; Chamberlain, Wilson & Fuller 1999). Because birds feed mainly on seeds and invertebrates, the abundance of these taxa must have been affected equally by recent dramatic changes in farming practice (Borg & Toft 2000). However, most of these hypothesized effects are often based on correlational studies, and the actual mechanisms involved in the association between the quality of farmland habitats and the reproductive success of birds are often poorly known. Many studies of birds, such as the lapwing Vanellus vanellus L. (Matter 1982; Møller 1983b; Baines 1990), curlew Numenius arquata L. (Berg 1992), corncrake Crex crex L. (Stowe et al. 1993), grey partridge Perdix perdix L. (Potts 1978), skylark Alauda arvensis L. (Green 1978; Christensen, Jacobsen & Nøhr 1996; Wilson et al. 1997; Poulsen, Sotherton & Aebischer 1998) and cirl bunting Emberiza cirlus L. (Evans & Smith 1994), have provided detailed information about the effects of farming practice on abundance and reproduction of farmland birds due to effects on habitat choice, foraging and food. However, a better knowledge of the mechanisms contributing to these observed effects is essential for understanding changes in abundance and distribution. The barn swallow Hirundo rustica L. is one of the most widespread and common breeding birds in the Palaearctic, with the total population exceeding 220 million (Moreau 1972). Breeding bird censuses and detailed local population studies have indicated that the population size in many parts of Europe has decreased by more than 50% during recent decades (Møller 1989; Marchant et al. 1990; Tucker & Heath 1994; Møller & Vansteenwegen 1997). Barn swallows in northern Europe mainly breed on farms with cattle (Glutz von Blotzheim 1985; Cramp 1988), and dramatic declines in the number of dairy farms in recent years may have precipitated the decline in numbers in parts of the European range. However, the exact causes are unknown, although changes in farming practice and a decline in the quality of the winter quarters in Africa have both been proposed (Møller & Vansteenwegen 1997). There is currently no evidence available to assess the hypothesis that farming practice is an important determinant of population size. Previous studies of the barn swallow have indicated that colonies on farms tend to be larger on dairy farms than on farms with other domesticated animals or without animals (Hølzinger 1969; Landmann & Landmann 1978), although there have been no attempts to test whether reproductive success differs between farms operating under different farming regimes. An assessment of the impact of farming practice on the abundance and reproduction of animals is difficult by definition because real experimental manipulations

often are impossible for logistic and economic reasons. The level of replication needed to test a hypothesis critically would require dramatic investments as individual farms represent the statistically independent observations. A slightly less powerful, but more feasible, method is to use the same farms in the absence and presence of a particular farming practice using a paired statistical design. This approach will control for the effects of topography, soil type, general habitats and the influence of neighbouring farmland on individual and population processes. The present study investigated the effects of dairy farming on the abundance, distribution and reproduction of barn swallows in a sample of 15 farms, before and after the cessation of dairy farming. A control sample of 15 farms without any changes in farming practice from exactly the same years was used to test whether there were any changes in the variables of interest due to temporal changes in an uncontrolled variable. Because the individual breeding barn swallows were the same for all survivors (on average 35% of adults; A.P. Møller & T. Szép, unpublished data), and because the breeding habitats and most foraging habitats close to the farms remained the same before and after the specific change of practice, any change in abundance, phenotype of breeders and their offspring and any change in reproduction was likely to reflect the change in farming practice. Furthermore, as the change in farming practice took place in different years (see the Methods), it was likely that annual fluctuations in abundance, phenotypes and reproductive parameters were evened out across the entire data set. The following predictions were tested based on our knowledge of barn swallow behaviour and biology. (i) If the presence of dairy cattle favours breeding barn swallows, the abundance of breeding birds should decrease after the cessation of dairy farming. If the benefit to barn swallows is due to increased availability of insects, food abundance in the presence of dairy cattle should be greater than in their absence. (ii) If adult barn swallows are attracted to farms with cattle, and if their ability to settle on these farms is related to the phenotypic quality of individuals in some way, we should expect the phenotypes of adults to differ between farms with and without dairy cattle. Specifically, we should expect barn swallows on farms with cattle to have long tails, larger body sizes, greater body masses, fewer parasites, higher haematocrits, lower leucocyte counts, and earlier arrival dates than conspecifics on farms without cattle, as all these variables are associated with high phenotypic quality (Møller 1994). (iii) If reproductive success of barn swallows is enhanced by the presence of cattle, we should expect barn swallows on farms with cattle to start laying earlier, have larger clutch and brood sizes, have greater hatching, fledging and breeding success, and to have a higher frequency of second clutches than barn swallows on farms without cattle. (iv) Finally, if the presence of cattle improves the quality of offspring, we should expect nestlings on farms with cattle to be larger, heavier and in better body

JPE593.fm Page 380 Wednesday, April 11, 2001 10:28 AM

380 A.P. Møller

condition, and to have fewer parasites, higher haematocrits, fewer leucocytes, and greater T-cell mediated immune responses than nestlings on farms without cattle.

Methods

© 2001 British Ecological Society, Journal of Applied Ecology, 38, 378–389

The barn swallow is a small (c. 20 g), socially monogamous, semi-colonial passerine feeding on insects caught on the wing. Sexual size dimorphism is slight, with the exception of the outermost tail feathers which are on average 20% longer in males than in females in a Danish population (Møller 1994). Males arrive at the breeding grounds before females, establish a small territory at the breeding sites, inside buildings, in culverts and under bridges, and attempt to attract a mate. A male and female subsequently build a nest, where the female lays a clutch of three to eight eggs. Incubation is by the female alone, while both pair members feed the offspring (Møller 1994). In the Danish population studied by Møller (1994), most pairs raise two broods per year. Barn swallows feed on large, actively flying, insects, mainly Diptera but also some Aphidoidea, Coleoptera, Lepidoptera and other taxa (Møller 1994). Extensive studies have shown that barn swallows prefer large insects over small ones, as determined from the size of eaten insects and those sampled in the foraging areas, while there is little evidence of preference for particular taxa (Turner 1980; Loske 1994; A.P. Møller, unpublished data). The data in this study were collected at Kraghede (57°12′N, 10°00′E), Denmark, during May–September 1970 – 99. The study site consists of open farmland with scattered plantations, ponds and hedgerows. The main crops on dairy farms are grass, beets and wheat, while other farms that have abandoned dairy farming grow mainly barley, wheat, potatoes, oilseed rape and other crops. The total study area covered was originally c. 15 km2. but was increased three times, in 1982 to 30 km2, in 1987 to 45 km2 and in 1998 to 55 km2, to increase the population size as the population declined. A detailed description of the study site and its breeding population of barn swallows is given in Møller (1994). Among the study farms, a total of 15 farms abandoned dairy farming from one year to the next during the study, and they comprise the database for the tests presented in this paper. These farms abandoned dairy farming in 1974, 1975, 1978, 1992 (two farms), 1993 (three farms), 1994 (three farms) and 1995 (four farms). The last year with dairy farming was the last full summer before dairy farming ceased, while the subsequent year was considered a year without dairy farming. None of these changes in farming practice was associated with buildings being closed for swallows. Hence there was no reason to believe that nest sites became limited as a consequence of changes in farming practice. All farms changed from dairy farming to production of mainly barley, wheat, potatoes and oilseed rape. However, all farms maintained pastures around the farm buildings for beef production, horses and sheep. Thus

there was no reduction in the area of pasture around farm buildings where most foraging by barn swallows took place (Møller 1985). An extensive study of foraging distances from the nest site using adult birds with their ventral sides individually coloured, revealed that more than 98% of all foraging took place within a distance of 500 m and slightly less than 50% within 100 m (Møller 1985). The total number of adult barn swallows recorded during the study was 2607 individuals, and the total number of first clutches was 1198. The 15 farms that abandoned dairy farming included 286 adult individuals the year before and 150 adult individuals the year after dairy farming was abandoned, and 143 and 75 first clutches before and after dairy farming was abandoned. A control sample of 15 farms that did not change farming practice during exactly the same years was used to test whether there was a significant temporal change in the variables of interest independent of farming practice. The sample of 15 control farms included 262 adults in the first year and 236 adults in the second year, while information on reproduction was available for 131 and 118 first clutches in the first and the second year, respectively. The two samples of farms were only comparable if the phenotypic characters of barn swallows breeding on the two kinds of farms did not differ significantly. Unpaired t-tests for all the parameters listed in Tables 1–3 did not reveal any significant differences between the sample of farms that abandoned dairy farming and the sample of control farms, after the significance level was Bonferroni adjusted. Thus there was no indication of this kind of bias. Breeding barn swallows have been censused in the study area since 1970. Adult barn swallows were ringed annually by mist netting, with an annual capture efficiency of more than 97% (A.P. Møller & T. Szép, unpublished data) according to estimates of capture efficiency using capture–recapture estimates (Lebreton et al. 1992). Intensive searches for nest sites were made weekly and all nestlings were ringed when 12–14 days old. The number of breeding pairs was estimated as the number of pairs building nests. The food abundance in farms with and without dairy cattle was estimated from sweep netting. Nine transects of 20 m in each 40° direction from the farms were sampled in the last year of dairy farming and in the following year. Transects were only made close to farms and all transects were conduced on pastures. All transects consisted of a total of five sweeps with the net (circular opening with a diameter of 0·5 m) at a height of 1 m above ground level. Transects were made on a sunny day with less than 10% cloud cover and wind less than 1 Beaufort in the morning, between 10:00 and 12:00 in the period 15–30 June. Insect abundance was recorded as the total number of insects per five sweeps. Previous studies have shown that swallow colony size on farms is positively correlated with this measure of food abundance ( Møller 1987), and that feeding rates at nests are positively correlated with this measure of food abundance (A.P. Møller, unpublished data).

JPE593.fm Page 381 Wednesday, April 11, 2001 10:28 AM

381 Dairy farming and barn swallows

Insects were preserved in alcohol and subsequently identified to family. Among the 1266 insects sampled, 49·1% were Diptera (mainly Muscidae, Tabanidae and Syrphidae), 48·1% Aphidoidea, 4·8% Coleoptera, 0·6% Lepidoptera and 0·6% other insects. This compares with 58·1% Diptera (mainly Muscidae, Tabanidae and Syrphidae), 35·6% Aphidoidea, 3·8% Coleoptera, 1·7% Lepidoptera and 0·9% other insects in the diet of barn swallow nestlings in the study area (n = 3188 insects collected using the ligature method to obtain food boluses; A.P. Møller, unpublished data).

    

© 2001 British Ecological Society, Journal of Applied Ecology, 38, 378– 389

Adult barn swallows were captured every year in mist nets (and to a small extent in sweep nets on or near the nests), and the following phenotypic traits were recorded: the length of the flattened wing, the length of the two outermost and the central tail feathers (the mean of the right and the left value is hereafter tail length), and the wingspan after stretching the wings to the maximum possible, all measured to the nearest millimetre with a ruler; length, depth and width of the beak and the length of the tarsus and the keel to the nearest 0·01 mm with a digital calliper; and body mass to the nearest 0·1 g with a Pesola spring balance. The circumference of the stretched right wing was traced onto paper when the front of the wing was held at an angle of 90° towards the body axis (Norberg 1990). Wing loading was calculated as body mass times 9·81 divided by the area of the right wing multiplied by 2, plus the area of the square between the two wings calculated from the wingspan and the width of the right wing at its base (Norberg 1990). All morphological characters were measured with a high degree of precision, as determined from very high repeatabilities of morphological characters measured the same and subsequent seasons (Møller 1991a; Møller 1994; Møller, de Lope & Saino 1995). Body condition was estimated as the residuals from a linear regression of body mass on the cube of tarsus length. The abundance of two species of parasites was determined for all adults: chewing louse Hirundoecus malleus L. abundance was assessed from the number of small holes that this species chews in wing and tail feathers of barn swallows (Møller 1991b). Previous studies using extraction of live chewing lice and counts of holes in feathers showed a strongly positive correlation between the two estimates of abundance of chewing lice (Møller 1991b). The abundance of the tropical fowl mite Ornithonyssus bursa Berlese in nests was assessed on the following fledging by placing a hand on the rim of the nest for 10 s. The number was then scored on a logarithmic scale as 0, 10, 100, 1000 or 10 000. This method is reliable as validated by extraction of mites from nests (Møller 1991b). When adult birds were captured, two 75-µl capillary tubes were filled with blood from the brachial vein, and the capillaries were stored in a cooling bag until the

samples were centrifuged for 10 min at 11 500 r.p.m. Two physiological measures of condition were recorded. Low haematocrit values (the percentage of blood that consists of erythrocytes after centrifugation) are characteristic of anaemia but also low physical condition (Coles 1997). Elevated buffy coat levels (which mainly consist of leucocytes and fibrinogen) indicate general health problems with increased production of leucocytes to counter infections (Coles 1997). Haematocrit and the buffy coat were measured with a digital calliper to the nearest 0·01 mm. The age of individual barn swallows was estimated to be the number of years recorded in the study area, with the assumption that the first capture was in the first year of life. This assumption was justified as all 135 local recruits were captured as breeding birds in their first year of life (Møller 1994; A. Møller, unpublished data). The arrival date was estimated as the first capture date during the years of intensive capture, when barn swallows were captured daily during the arrival period.

   Reproduction of barn swallows was recorded during regular visits to nests throughout the breeding season, often daily around the time of laying, hatching and fledging, but at least twice per week. The laying date was recorded as the date when the first egg was laid, assuming that one egg was laid daily, which is usually the case (Møller 1994). Clutch size was defined as the number of eggs laid in succession in a nest, while brood size at hatching was the number of nestlings that hatched and brood size at fledging the number of nestlings that fledged. Hatching success was brood size at hatching divided by clutch size; fledging success was brood size at fledging divided by brood size at hatching; and breeding success was brood size at fledging divided by clutch size. The frequency of second clutches was the proportion of pairs initiating a second clutch. This information was recorded for all clutches during the entire study period.

    When nestlings were 12 days old, their tarsus length and body mass were recorded as described above. An index of body condition was calculated as residuals from a regression of body mass on tarsus length3. All statistical analyses were based on brood means. Two blood samples were obtained from each nestling and used to estimate haematocrit and leucocytes, as described above. At the age of 12 days, all nestlings were injected intradermally in the wing web (the patagium) with 0·2 mg of phytohaemagglutinin-P (PHA; Sigma L8754, Sigma Inc., St Louis, MO) in 0·04 ml isotonic saline (the antigen injection), while the left wing web was injected with the same amount of saline only (a control injection). The thickness of wing webs was

JPE593.fm Page 382 Wednesday, April 11, 2001 10:28 AM

382 A.P. Møller

measured immediately before and 24 h after injection in inoculated sites using a pressure sensitive calliper known as a spessimeter (Alpa S.p.A., Milano, cod. SM112, Teclock, Japan), with an accuracy of 0·01 mm. To express the reaction to PHA while controlling for the effect of injection per se and thickening due to saline injection, we calculated the difference between the change in thickness of the right PHA-injected wing web (thickness 24 h after injection minus thickness before injection) and the corresponding change in thickness of the left wing web, injection with only saline. This procedure followed Saino, Calza & Møller (1997) in their study of barn swallows. The thickness of the wing web was measured three times before and after PHA injection, and the average of these three measurements was used in calculations. The repeatability of the wing web index was high and highly significant (Saino, Calza & Møller 1997).

the presence of cattle. Thus, the statistically independent observation was the individual farm, and mean values were calculated for all variables for each farm before and after dairy farming ceased. Paired t-tests were used to compare the mean values under the null hypothesis that the means were unaffected by the absence of cattle. A large number of statistical tests was made, and the overall level of significance was Bonferroni adjusted for multiple tests to control the type I error rate (Holm 1979; Wright 1992). Strict application of this method severely reduces the power of tests (Wright 1992), but such sacrificial loss of power can be avoided by choosing an experiment-wise error rate higher than the usually accepted 5%. I used 10% as suggested by Wright (1992) and Chandler (1995). Values reported are means (SE) throughout.

Results   A relatively small number of farms with a change in farming practice was studied. Hence the power of statistical tests was relatively low. The power was increased by making pairwise comparisons for the response variables for the same farms with and without

© 2001 British Ecological Society, Journal of Applied Ecology, 38, 378–389

        The number of farms with cattle within the study area decreased dramatically over the study period 1970–99 (Fig. 1a). Barn swallows were closely associated with

Fig. 1. (a) Relative frequency of dairy farms in the Kraghede study area, 1970 – 99. The total number of farms was 47. (b) Relative frequency of barn swallow pairs on dairy farms in the Kraghede study area 1970 – 99. The number of breeding pairs ranged between 21 and 188, with 2194 pairs recorded during the 30 years.

JPE593.fm Page 383 Wednesday, April 11, 2001 10:28 AM

383 Dairy farming and barn swallows

dairy farms in the Kraghede study area, where 83·7% of all 2194 recorded breeding pairs were found on dairy farms in any single year (Fig. 1b). The observed and expected number of dairy farms with barn swallows differed significantly from a random distribution across farms (χ2-test, χ2 > 5·88, P < 0·05 in all years). Hence, the observed number of barn swallows on dairy farms was greater than expected based on the number of farms available. The local abundance of breeding barn swallows decreased significantly when dairy farming ceased. The average colony size decreased by 48% when cattle disappeared [colony size in the presence of cattle: 9·53 (1·95); colony size in the absence of cattle: 5·00 (1·46), paired t-test, t = 6·13, d.f. = 14, P < 0·001]. In the control farms colony size did not decrease significantly during the same period [colony size in the first year: 8·73 (1·16); colony size in the second year: 7·87 (1·06), paired t-test, t = 0·94, d.f. = 14, P = 0·36]. Insect abundance on farms was significantly affected by the presence of dairy cattle. The mean abundance of insects was 52% fewer in the absence of dairy cattle than in their presence [dairy cattle present: 6·34 (0·45) insects per five sweeps; dairy cattle absent: 3·04 (0·26) insects per five sweeps; n = 15 farms, paired t-test, t = 8·76, d.f. = 14, P < 0·001]. There was no significant change in insect abundance on the control farms during the same years [first year: 6·37 (0·47) insects per five sweeps; second year: 6·33 (0·44) insects per five sweeps; n = 15 farms, paired t-test, t = 0·23, d.f. = 14, P = 0·82].

     A very large number of morphological and other phenotypic characters was recorded for male and female barn swallows, and comparisons were made between adult barn swallows breeding in the presence and absence of dairy cattle. Almost all of these statistical tests were not statistically significant (Table 1). Hence there was little evidence that dairy farming affected the distribution of individuals of different phenotypes, with just three variables indicating significant differences for males and two for females. However, the percentage of yearlings recruited to colonies was reduced in the absence of cattle, as indicated by a lower frequency of yearlings after the cessation of dairy farming, and this effect was significant for females even after Bonferroni correction for the number of statistical tests (Table 1). For the control farms with no change in farming practice there were no significant differences in phenotypes of adults between years after Bonferroni correction (Table 1).

    © 2001 British Ecological Society, Journal of Applied Ecology, 38, 378– 389

Reproduction of barn swallows in the presence and the absence of cattle was recorded at a number of different stages of the reproductive cycle (Table 2). Most reproductive parameters did not differ significantly between

the two categories. However, the size of the first clutch decreased significantly on farms that abandoned dairying, and this difference was not found in the control sample of farms. Similarly, the frequency of second clutches fell on farms abandoning dairying, but did not in the sample of control farms. Hence annual reproductive success measured as the number of offspring produced in first and second clutches combined decreased significantly when farmers ceased dairy farming ( Table 2). Because the frequency of second clutches changed after dairy farming ceased, this implied that any differences in reproductive parameters of the second clutch could have arisen from effects of dairy farming and/or a different sample of adults contributing to the sample.

     The effects of the presence of cattle on the quality of nestling barn swallows were assessed in a pairwise comparison. The phenotypes of nestlings with and without cattle differed in several respects (Table 3). Nestlings from dairy farms tended to have longer tarsi (on average 3·7%) and a significantly larger body mass (on average 4·6%) and body condition index before the abandonment of dairying. There was no similar change in the sample of control farms (Table 3). The abundance of tropical fowl mites did not differ between farms. Haematocrit was reduced significantly (on average by 9·6%), while leucocyte abundance increased significantly (on average by 64%), after the abandonment of dairying. No significant difference was found in the control farms ( Table 3). The measure of T-cell mediated immune response (PHA response) fell significantly, with a mean reduction of 32% after cessation of dairy farming ( Table 3). There was no significant change in the sample of control farms. Hence, the presence of cattle affected the phenotype of barn swallow nestlings produced on those farms.

Discussion        The consequences of farming practice on barn swallow abundance and reproductive success were investigated using longitudinal information for two kinds of farms: farms that abandoned dairy farming, and a control group of farms that did not change farming practice. The consequences of farming practice on barn swallows was assessed immediately following the sudden cessation of dairy farming. As many of the same birds were breeding on a farm under these two different conditions, the study can be considered to assess the influence of farming practice on performance of barn swallows under different environmental conditions. This approach is semi-experimental in the sense that many factors potentially affecting performance of barn swallows will be

JPE593.fm Page 384 Wednesday, April 11, 2001 10:28 AM

384 A.P. Møller

Table 1. Phenotypes of adult barn swallows in relation to the presence of dairy cattle. For control farms, farming practice did not change from one year to the next. Values are means per farm (SE). Sample size is 15 farms, except for haematocrit and leucocytes, where it is 13 farms Variable

© 2001 British Ecological Society, Journal of Applied Ecology, 38, 378–389

Dairy cattle present

Dairy cattle absent

t

P

Sample of farms that abandoned dairy farming Males Wing length (mm) 127·05 (0·22) Tail length (mm) 108·88 (1·06) Central tail length (mm) 44·86 (0·16) Wingspan (mm) 330·80 (1·04) Tarsus length (mm) 11·31 (0·06) Keel length (mm) 21·61 (0·04) Beak length (mm) 7·66 (0·03) Beak depth (mm) 2·85 (0·07) Beak width (mm) 12·02 (0·04) Body mass (g) 19·13 (0·12) Wing loading 1·24 × 105 (0·02 × 105) Chewing lice 9·77 (1·04) Haematocrit (%) 49·64 (0·20) Leucocyte count (%) 0·57 (0·01) Age (% yearlings) 66·3 (1·6) Arrival date (1 = 1 May) 14·86 (1·01)

127·50 (0·35) 108·21 (1·14) 43·86 (0·20) 331·83 (1·01) 11·20 (0·07) 21·52 (0·09) 7·50 (0·08) 2·69 (0·07) 12·02 (0·08) 19·45 (0·15) 1·23 × 105 (0·01 × 105) 10·14 (1·12) 50·94 (0·43) 0·59 (0·03) 60·6 (1·1) 18·32 (1·46)

1·44 0·44 1·48 0·63 1·53 0·90 1·42 2·10 0·09 1·34 0·45 0·26 2·71 0·49 2·52 2·50

0·18 0·67 0·16 0·54 0·15 0·38 0·18 0·054 0·93 0·20 0·66 0·80 0·019 0·64 0·025 0·026

Females Wing length (mm) Tail length (mm) Central tail length (mm) Wingspan (mm) Tarsus length (mm) Keel length (mm) Beak length (mm) Beak depth (mm) Beak width (mm) Body mass (g) Wing loading Chewing lice Haematocrit (%) Leucocyte count (%) Age (% yearlings) Arrival date (1 = 1 May)

123·66 (0·32) 90·80 (0·63) 44·42 (0·34) 326·92 (0·79) 11·15 (0·09) 20·75 (0·11) 7·56 (0·09) 2·66 (0·07) 11·82 (0·06) 20·40 (0·42) 1·29 × 105 (0·02 × 105) 11·02 (1·45) 40·20 (0·52) 0·58 (0·04) 59·3 (1·3) 16·66 (0·88)

1·93 1·40 0·66 2·23 1·53 0·90 1·35 1·94 0·68 0·88 1·07 0·09 1·15 0·30 4·04 1·07

0·07 0·18 0·52 0·043 0·15 0·38 0·20 0·07 0·51 0·39 0·30 0·93 0·27 0·77 0·0012* 0·30

Sample of control farm in which dairy farming was not abandoned Males Wing length (mm) 126·61 (0·44) 125·60 (0·44) Tail length (mm) 108·16 (1·20) 105·67 (0·77) Central tail length (mm) 43·92 (0·25) 44·20 (0·28) Wingspan (mm) 332·96 (1·09) 331·64 (1·08) Tarsus length (mm) 11·33 (0·10) 11·37 (0·15) Keel length (mm) 21·62 (0·07) 21·69 (0·13) Beak length (mm) 7·55 (0·06) 7·50 (0·11) Beak depth (mm) 2·84 (0·08) 2·88 (0·11) Beak width (mm) 11·93 (0·06) 12·02 (0·10) Body mass (g) 18·96 (0·15) 18·50 (0·15) 1·24 × 105 (0·01 × 105) Wing loading 1·19 × 105 (0·02 × 105) Chewing 11·50 (1·26) 12·16 (1·94) Haematocrit (%) 50·03 (0·46) 51·72 (0·68) Leucocyte count (%) 0·57 (0·02) 0·59 (0·02) Age (% yearlings) 62·6 (1·8) 66·9 (1·6) Arrival date (1 = 1 May) 14·51 (1·04) 14·32 (1·19)

2·01 1·75 0·63 0·84 0·40 0·63 0·41 0·40 0·69 2·37 1·83 0·39 3·01 0·48 0·45 0·28

0·06 0·10 0·54 0·42 0·69 0·54 0·69 0·69 0·50 0·03 0·09 0·70 0·01 0·66 0·66 0·78

Females Wing length (mm) Tail length (mm) Central tail length (mm) Wingspan (mm) Tarsus length (mm) Keel length (mm) Beak length (mm) Beak depth (mm)

0·27 0·48 0·62 0·04 1·08 2·19 0·05 0·84

0·79 0·64 0·55 0·97 0·30 0·05 0·96 0·42

124·40 (0·26) 89·81 (0·45) 44·74 (0·19) 324·80 (0·58) 11·40 (0·06) 20·93 (0·06) 7·70 (0·04) 2·81 (0·06) 11·87 (0·04) 20·00 (0·19) 1·32 × 105 (0·02 × 105) 10·86 (1·21) 48·73 (0·36) 0·57 (0·01) 64·8 (1·3) 17·71 (0·56)

123·77 (0·32) 89·48 (0·40) 44·75 (0·25) 325·23 (1·17) 11·33 (0·08) 20·84 (0·07) 7·66 (0·06) 2·84 (0·07)

123·67 (0·39) 89·00 (0·86) 44·40 (0·35) 325·28 (2·04) 11·24 (0·10) 20·99 (0·10) 7·66 (0·09) 2·77 (0·10)

JPE593.fm Page 385 Wednesday, April 11, 2001 10:28 AM

385 Dairy farming and barn swallows

Table 1. continued Variable Beak width (mm) Body mass (g) Wing loading Chewing lice Haematocrit (%) Leucocyte count (%) Age (% yearlings) Arrival date (1 = 1 May)

Dairy cattle present

Dairy cattle absent

11·79 (0·10) 19·82 (0·25) 1·27 × 105 (0·02 × 105) 10·88 (2·34) 49·46 (0·79) 0·57 (0·01) 68·8 (5·0) 17·56 (0·52)

11·85 (0·09) 20·06 (0·27) 1·28 × 105 (0·02 × 105) 8·02 (1·09) 49·03 (0·56) 0·57 (0·03) 65·3 (6·1) 16·65 (0·87)

t

P

0·56 0·73 0·52 1·14 0·50 0·29 0·70 0·90

0·58 0·48 0·61 0·28 0·63 0·77 0·50 0·38

*P < 0·05 after sequential Bonferroni adjustment.

Table 2. The effects of presence of dairy cattle on reproductive variables of barn swallows. For control farms, farming practice did not change from one year to the next. Values are means per farm (SE). Sample size is 15 farms Variable

Cattle present

Cattle absent

t

P

Sample of farms that abandoned dairy farming First clutch Laying date 33·12 (0·80) Clutch size 4·94 (0·07) Brood size at hatching 4·48 (0·05) Brood size at fledging 4·32 (0·10) Hatching success (%) 90·20 (2·09) Fledging success (%) 95·20 (1·12) Breeding success (%) 85·77 (2·20)

34·11 (1·34) 4·70 (0·05) 4·32 (0·10) 4·25 (0·12) 92·80 (1·43) 86·96 (6·20) 80·73 (6·07)

0·73 4·45 1·30 1·61 1·17 1·40 0·82

0·48 0·0006* 0·21 0·13 0·26 0·18 0·43

Second clutch Second clutches (%) Laying date Clutch size Brood size at hatching Brood size at fledging Hatching success (%) Fledging success (%) Breeding success (%) Annual reproductive success

45·71 (3·33) 83·00 (0·77) 4·31 (0·06) 4·01 (0·09) 3·69 (0·13) 92·91 (1·24) 92·33 (2·72) 85·61 (2·88) 6·48 (0·43)

5·00 3·58 1·18 1·87 1·50 2·32 0·11 0·67 4·03

0·0002* 0·003* 0·26 0·08 0·16 0·036 0·91 0·51 0·0012*

Sample of control farm in which dairy farming was not abandoned First clutch Laying date 33·76 (1·54) Clutch size 4·75 (0·10) Brood size at hatching 4·44 (0·13) Brood size at fledging 4·20 (0·15) Hatching success (%) 93·60 (2·63) Fledging success (%) 94·60 (2·05) Breeding success (%) 88·64 (3·25)

36·57 (1·72) 4·70 (0·10) 4·47 (0·12) 4·41 (0·14) 95·44 (2·60) 98·56 (1·04) 94·13 (2·83)

1·17 0·38 0·21 1·23 0·49 1·57 1·35

0·26 0·71 0·83 0·24 0·63 0·14 0·20

Second clutch Second clutches (%) Laying date Clutch size Brood size at hatching Brood size at fledging Hatching success (%) Fledging success (%) Breeding success (%) Annual reproductive success

60·95 (6·65) 82·98 (0·78) 4·53 (0·10) 4·22 (0·14) 3·94 (0·19) 93·07 (1·57) 93·85 (4·03) 87·37 (4·03) 6·94 (0·46)

0·31 0·64 0·01 1·27 1·82 1·78 0·91 1·81 0·46

0·76 0·53 0·99 0·22 0·09 0·10 0·38 0·09 0·65

65·79 (3·04) 80·38 (0·99) 4·43 (0·09) 4·23 (0·08) 3·93 (0·13) 95·55 (0·83) 92·71 (2·24) 88·06 (2·62) 7·39 (0·25)

57·48 (7·01) 81·86 (1·62) 4·53 (0·10) 4·37 (0·12) 4·32 (0·15) 96·59 (1·53) 98·66 (1·65) 95·22 (1·90) 6·64 (0·41)

*P < 0·05 after sequential Bonferroni adjustment.

© 2001 British Ecological Society, Journal of Applied Ecology, 38, 378– 389

kept constant. Therefore, the power of the statistical tests is considerably enhanced. Farming practice may affect the abundance, distribution and reproductive success of barn swallows. The

present study provides a statistical test of this hypothesis. Barn swallows are disproportionately associated with dairy farming in the Kraghede study area (Fig. 1), and presumably also in other parts of northern Europe

JPE593.fm Page 386 Wednesday, April 11, 2001 10:28 AM

386 A.P. Møller

Table 3. The effects of presence of dairy cattle on the phenotypes of barn swallow nestlings. For control farms, farming practice did not change from one year to the next. Nestlings were measured when 12 days old. Values are means per farm of brood means (SE). Sample size is 15 farms, except for haematocrit, leucocytes and PHA response, where it is 13 farms Variable

Cattle present

Cattle absent

t

P

11·20 (0·10) 21·45 (0·10) – 0·26 (0·07) 0·36 (0·12) 42·28 (0·60) 0·42 (0·03) 1·19 (0·10)

7·22 6·02 6·17 1·74 4·07 3·85 4·73

0·0001* 0·0001* 0·0001* 0·10 0·0001* 0·002* 0·0005*

Sample of control farm in which dairy farming was not abandoned Tarsus length (mm) 11·60 (0·06) 11·54 (0·09) Body mass (g) 22·32 (0·11) 22·32 (0·14) Body condition – 0·03 (0·10) 0·02 (0·10) Mite abundance 0·49 (0·10) 0·49 (0·12) Haematocrit (%) 46·24 (0·54) 46·57 (0·61) Leucocyte count (%) 0·26 (0·05) 0·24 (0·04) PHA response (mm) 1·55 (0·08) 1·56 (0·08)

0·65 0·37 0·29 0·08 1·09 2·05 0·23

0·52 0·71 0·78 0·94 0·30 0·06 0·82

Sample of farms that abandoned dairy farming Tarsus length (mm) 11·61 (0·07) Body mass (g) 22·43 (0·12) Body condition 0·31 (0·06) Mite abundance 0·50 (0·10) Haematocrit (%) 46·35 (0·51) Leucocyte count (%) 0·26 (0·05) PHA response (mm) 1·57 (0·09)

*P < 0·05 after sequential Bonferroni adjustment.

© 2001 British Ecological Society, Journal of Applied Ecology, 38, 378–389

(Glutz von Blotzheim 1985; Cramp 1988). Populations fluctuate dramatically between years (Møller 1989), making it difficult to identify those factors that contribute to annual fluctuations at a local scale. However, when local population size was estimated as colony size it was found to decrease significantly when dairy farming ceased (on average by 48%). As breeding philopatry is extremely high, with almost all surviving birds breeding at the same site year after year (Møller 1994), this result implies that local recruitment was reduced at farms when dairy farming ceased. Consistent with this suggestion, the age distribution of barn swallows changed towards older birds when dairy farming ceased (Table 1). The benefits of dairy farming for barn swallows include high food abundance (mainly Diptera), barns with constant high ambient temperatures that reduce the energy costs of incubation, brooding and thermoregulation in offspring and adults, and availability of nest sites. Food abundance is likely to be of particular importance because it will directly affect the condition of barn swallows, particularly during poor weather conditions. Food abundance measured from sweep netting was more than 52% lower on farms without cattle compared with those where cattle were present. Barn swallows usually feed at a height of 0–5 m above ground level (Turner 1980), and most foraging takes place within a distance of a few hundred metres from the colony and almost all within 500 m (Møller 1987). Hence, the measure of food abundance recorded in the present study is likely to be biologically meaningful for barn swallows. This assumption is supported by a strong positive correlation between colony size and food abundance measured with sweep netting (Møller 1987). The hypothesis that ambient temperature may affect reproductive success has been investigated in hole nesting birds, and high temperatures may affect metabolism

considerably (Mertens 1987). However, there is little evidence of barn swallow reproduction differing between barns with different temperature regimes (Löhrl & Gutscher 1973). The availability of nest sites seems relatively unimportant, as the barn swallow population in the study area was several times larger in the early 1970s than today without any significant change in the number of nest sites. Furthermore, none of the farm buildings was closed for barn swallows after dairy farming ceased. Differences in the abundance of barn swallows on farms with and without cattle may be due to effects on local recruitment or differential emigration. The phenotypes of male and female adult barn swallows on farms with and without cattle generally showed little evidence of significant differences (Table 1). Adult barn swallows on dairy farms were not larger nor in better condition than barn swallows on these farms after cessation of dairy farming. Similarly, male tail length, which is a secondary sexual character, did not differ between the two situations. Parasite load and health parameters, such as haematocrit and leucocyte counts, were also similar on farms with and without cattle. Finally, arrival date was not significantly different on farms with and without cattle (Table 1). The very large number of tests provided little evidence of phenotypes of adult barn swallows changing when dairy farming ceased. This conclusion suggests that the distribution of adults of different phenotypes is unaffected by farming practice.

     Because food abundance was affected by the presence of dairy cattle, this may translate to differences in fledging

JPE593.fm Page 387 Wednesday, April 11, 2001 10:28 AM

387 Dairy farming and barn swallows

© 2001 British Ecological Society, Journal of Applied Ecology, 38, 378– 389

success. Thus variables relating to reproduction could potentially be influenced by farming practice. Laying date may be affected by food availability (Lack 1966; Rowe, Ludwig & Schluter 1995; Martin 1987); however, barn swallows did not breed later when cattle were absent (Table 2). Similarly, clutch size may be affected by food availability (Lack 1947–48, 1966; Martin 1987); the barn swallows were found to have larger clutches in the presence of dairy cattle (Table 2). The reproductive variable that was most strongly affected by dairy farming was the frequency of second clutches, which was reduced by 30% in the absence of cattle compared with when cattle were present (Table 2). Thus, mean annual reproductive success was reduced by the absence of cattle. There was no such effect among control farms where farming practice did not change. Differences in local abundance may have caused density-dependent effects on reproductive success, and this would invalidate the paired comparison between the situation before and after cessation of dairy farming. This seems unlikely because performance generally was better before abandonment of dairy farming. Furthermore, there is little evidence of density-dependent reproduction in this barn swallow population (Møller 1989). Not only the quantity but also the quality of offspring was affected by farming practice. Nestlings tended to be smaller and had significantly lower body mass when cattle disappeared from farms (Table 3). Nestling body mass has been shown to correlate positively with probability of local recruitment in birds in general (Lack 1966) and in barn swallows in particular (Møller 1994). The change in offspring condition affected by cessation of dairy farming also influenced measures of health status, such as haematocrit, leucocyte counts and T-cell mediated immune response as assessed by the PHA test (Table 3). The latter injection is known as a stimulant of T-cell inducing mitogenic activity (Roitt, Brostoff & Male 1996) and it provides a reliable measure of the cellular immune response in vivo (Goto et al. 1978; McCorkle, Olah & Glick 1980). Subcutaneous injection with PHA produces a complex series of physiological reactions, externally viewed as an inflammatory local response (Goto et al. 1978). Low haematocrit values are characteristic of anaemia, but also low physical condition (Coles 1997). Elevated buffy coat levels indicate general health problems with increased production of leucocytes to counter infections (Coles 1997). Differences in these measures were not found on farms where farming practice did not change. Hence, nestlings produced in the absence of dairy farming showed several general indications of poor condition. Extensive studies of barn swallows in the Kraghede study area, and in other populations in Italy and Spain, have shown that surviving nestlings have higher haematocrit and lower concentrations of leucocytes than nestlings that died in the nest (A.P. Møller , N. Saino & F. de Lope, unpublished data). Thus, it seems likely that the physiological differences reported here will be associated with differences in viability.

Host immune function has been shown to be positively affected by body condition in general, particularly in birds (Chandra & Newberne 1977; Møller et al. 1998). As barn swallow nestlings with strong T-cell dependent immune responses are generally in prime condition, as determined by the positive relationship between parental feeding effort and immune response (Saino, Calza & Møller 1997), it seems likely that nestlings raised on farms without dairy cattle will have a reduced likelihood of recruiting into the breeding population. The presence of dairy cows has a strong effect on the abundance, distribution and reproductive success of barn swallows. This effect acts mainly through the abundance of food (large Diptera). All changes in farming practice were to arable farming producing a mixture of grain, potatoes and oilseed rape, so the importance of other types of farming for swallows could not be assessed directly. However, previous studies have indicated that dairy farming rather than pig farming or other kinds of animal husbandry is the single most important predictor of the abundance of breeding barn swallows (Hölzinger 1969; Löhrl & Gutscher 1973; Landmann & Landmann 1978; Møller 1983a). Thus dairy farming is necessary to sustain a large population of barn swallows in the study area. Changes in farming practice during the last 30 years have consisted mainly of a reduction in the number of dairy farms and concentration of cows in a few, large, farms (Møller 1983b; Fig. 1). These changes are a direct consequence of farming subsidies within the European Community, which have encouraged cessation of dairy farming, thereby having dramatic effects on farming practice and land use in most parts of western Europe. In conclusion, barn swallows were disproportionately common on farms with dairy cows compared with farms without cows, and cessation of dairy farming was followed by a reduction in the size of the swallow breeding population. There was an association between an elevated abundance of insects on dairy farms and the number of swallows. However, different phenotypes of adult barn swallows were distributed randomly across farms with different farming practices. Barn swallow reproduction was negatively affected by the absence of cows, and barn swallows on such farms seemed to produce fewer, low quality, offspring, thus potentially reducing recruitment rate.

Acknowledgements Many different farmers have tolerated my activity during three decades of fieldwork. N. Cadée and E. Flensted-Jensen kindly helped with fieldwork. My research was supported by a grant from the European Community (METABIRD, contract no. EVK2-CT1999–0017 METABIRD).

References Baines, D. (1990) The roles of predation, food and agricultural practice in determining the breeding success of the lapwing

JPE593.fm Page 388 Wednesday, April 11, 2001 10:28 AM

388 A.P. Møller

© 2001 British Ecological Society, Journal of Applied Ecology, 38, 378–389

(Vanellus vanellus) on upland grassland. Journal of Animal Ecology, 59, 915 – 929. Berg, Å. (1992) Factors affecting nest-site choice and reproductive success of curlew Numenius arquata on farmland. Ibis, 134, 44 – 51. Borg, C. & Toft, S. (2000) Importance of insect prey quality for grey partridge chicks Perdix perdix: a self-selection experiment. Journal of Applied Ecology, 37, 557 – 563. Chamberlain, D.E., Wilson, J.D. & Fuller, R.J. (1999) A comparison of bird populations on organic and conventional farm systems in southern Britain. Biological Conservation, 88, 307 – 320. Chandler, C.R. (1995) Practical considerations in the use of simultaneous inference for multiple tests. Animal Behaviour, 49, 524 – 527. Chandra, R.K. & Newberne, P.M. (1977) Nutrition, Immunity, and Infection. Plenum Press, New York, NY. Christensen, K.D., Jacobsen, E.M. & Nøhr, H. (1996) A comparative study of bird faunas in conventionally and organically farmed areas. Dansk Ornithologisk Forenings Tidsskrift, 90, 21– 28. Coles, B.H. (1997) Avian Medicine and Surgery, 2nd edn. Blackwell, Oxford, UK. Cramp, S. (1988) The Birds of the Western Palaearctic, Vol. V. Oxford University Press, Oxford, UK. Evans, A.D. & Smith, K.W. (1994) Habitat selection of cirl buntings Emberiza cirlus wintering in Britain. Bird Study, 41, 81– 87. Glutz von Blotzheim, U.N. (1985) Handbuch der Vögel Mitteleuropas, Vol. 10. AULA-Verlag, Wiesbaden, Germany. Goto, N., Kodama, H., Okada, K. & Fujimoto, Y. (1978) Suppression of phytohaemagglutinin skin response in thymectomized chickens. Poultry Science, 57, 246 – 250. Green, R.E. (1978) Factors affecting the diet of farmland skylarks. Journal of Animal Ecology, 47, 913 – 928. Hagemeijer, E.J.M. & Blair, M.J. (1997) The EBCC Atlas of European Breeding Birds: Their Distribution and Abundance. T. & A. D. Poyser, London, UK. Holm, S. (1979) A simple sequentially rejective multiple test procedure. Scandinavian Journal of Statistics, 6, 65 – 70. Hölzinger, J. (1969) Fünfjährige Untersuchungen über den Brutbestand der Mehl- und Rauchschwalbe (Delicon urbica et Hirundo rustica) in der Umgebung von Ulm. Anzeiger der Ornithologischen Gesellschafts Bayern, 8, 610 – 624. Lack, D. (1947 – 48) The significance of clutch size. Ibis, 89, 302 – 352, 90, 25 – 45. Lack, D. (1966) Population Studies of Birds. Clarendon Press, Oxford, UK. Landmann, A. & Landmann, C. (1978) Zur Siedlungsbiologie der Rauchschwalbe Hirundo rustica and Mehlschwalbe Delichon urbica in der Unteren Schranne, Nordtirol. Anzeiger der Ornithologischen Gesellschafts Bayern, 17, 247 – 266. Lebreton, J.-D., Burnham, K.P., Clobert, J. & Anderson, D.R. (1992) Modelling survival and testing biological hypotheses using marked animals: a unified approach with case studies. Ecological Monographs, 62, 67 –118. Löhrl, H. & Gutscher, H. (1973) Zur Brutbiologie der Rauchschwalbe (Hirundo rustica) in einem südwestdeutschen Dorf. Journal für Ornithologie, 114, 399 – 416. Loske, K.-H. (1994) Untersuchungen zu Überlebensstrategien der Rauchschwalbe (Hirundo rustica) im Brutgebiet. CuvillerVerlag, Göttingen, Germany. McCorkle, F., Olah, I. & Glick, B. (1980) The morphology of the phytohaemagglutinin induced cell response in the chicken’s wattle. Poultry Science, 59, 616 – 623. Marchant, J.H., Hudson, S.P., Carter, S.P. & Whittington, P. (1990) Population Trends in British Breeding Birds. BTO, Tring, UK.

Martin, T.E. (1987) Food as a limit on breeding birds: a lifehistory perspective. Annual Review of Ecology and Systematics, 18, 453 – 487. Matter, H. (1982) Einfluss intensiver Feldbewirtschaft auf der Bruterfolg des Kiebitzes Vanellus vanellus in Mitteleuropa. Der Ornithologische Beobachter, 79, 1–24. Mertens, J.A.L. (1987) The influence of temperature on the energy reserves of female great tits during the breeding season. Ardea, 75, 73 – 80. Møller, A.P. (1983a) Breeding habitat selection in the swallow Hirundo rustica. Bird Study, 30, 134–142. Møller, A.P. (1983b) Changes in Danish farmland habitats and their populations of breeding birds. Holarctic Ecology, 6, 95 –100. Møller, A.P. (1985) The social behaviour of swallows Hirundo rustica during the breeding season. PhD Thesis. Institute of Zoology and Zoophysiology, University of Århus, Århus, Denmark. Møller, A.P. (1987) Advantages and disadvantages of coloniality in the swallow Hirundo rustica. Animal Behaviour, 35, 819 – 832. Møller, A.P. (1989) Population dynamics of a declining swallow Hirundo rustica L. population. Journal of Animal Ecology, 58, 1051–1063. Møller, A.P. (1991a) Sexual selection in the monogamous barn swallow (Hirundo rustica). I. Determinants of tail ornament size. Evolution, 45, 1823–1836. Møller, A.P. (1991b) Parasites, sexual ornaments and mate choice in the barn swallow Hirundo rustica. Ecology, Behaviour, and Evolution of Bird–Parasite Interactions (eds J.E. Loye & M. Zuk), pp. 328–343. Oxford University Press, Oxford, UK. Møller, A.P. (1994) Sexual Selection and the Barn Swallow. Oxford University Press, Oxford, UK. Møller, A.P. & Vansteenwegen, C. (1997) Barn swallow. The EBCC Atlas of European Breeding Birds (eds W.J.M. Hagemeijer & M.J. Blair), pp. 478 –479. T. & A. D. Poyser, London, UK. Møller, A.P., Christe, P., Erritzøe, J. & Mavarez, J. (1998) Condition, disease and immune defence. Oikos, 83, 301– 306. Møller, A.P., de Lope, F. & Saino, N. (1995) Sexual selection in the barn swallow Hirundo rustica. VI. Aerodynamic adaptations. Journal of Evolutionary Biology, 8, 671– 687. Moreau, R.E. (1972) The Palaearctic–African Bird Migration Systems. Academic Press, London and New York. Norberg, U.M. (1990) Vertebrate Flight. Springer-Verlag, Berlin, Germany. Pain, D.J. & Pienkowski, M.W. (1997) Farming and Birds in Europe. Academic Press, London, UK. Petersen, B.S. & Jacobsen, E.M. (1997) Population trends in Danish farmland birds. Bekœmpelsesmiddelforskning fra Miljøstyrelsen Nr. 37, 1997. Miljø-vog Energiministeriet, Miljøstyrelsen, Copenhagen, Denmark. Potts, G.R. (1978) The Partridge. Collins, London, UK. Poulsen, J.G., Sotherton, J.G. & Aebischer, N.J. (1998) Comparative nesting and feeding ecology of skylarks Alauda arvensis on arable farmland in southern England with special reference to set aside. Journal of Applied Ecology, 35, 131–147. Roitt, I.M., Brostoff, J. & Male, D.K. (1996) Immunology, 4th edn. Mosby, London. Rowe, L., Ludwig, D. & Schluter, D. (1995) Time, condition, and the seasonal decline of avian clutch size. American Naturalist, 143, 698 – 722. Saino, N., Calza, S. & Møller, A.P. (1997) Immunocompetence of nestling barn swallows in relation to brood size and parental effort. Journal of Animal Ecology, 66, 827– 836. Stowe, C., Newton, A.W., Freen, R.E. & Mayes, E. (1993) The

JPE593.fm Page 389 Wednesday, April 11, 2001 10:28 AM

389 Dairy farming and barn swallows

© 2001 British Ecological Society, Journal of Applied Ecology, 38, 378– 389

decline of the corncrake in Britain and Ireland in relation to habitat. Journal of Applied Ecology, 30, 53 – 62. Tucker, G.M. & Heath, M.F. (1994) Birds in Europe: Their Conservation Status. BirdLife International, Cambridge, UK. Turner, A.K. (1980) The use of time and energy by aerialfeeding birds. PhD Thesis. University of Stirling, Stirling, UK.

Wilson, J.D., Evans, J., Browne, S.J. & King, J.R. (1997) Territory distribution and breeding success of skylarks Alauda arvensis on organic and intensive farmland in southern England. Journal of Applied Ecology, 34, 1462–1478. Wright, S.P. (1992) Adjusted P-values for simultaneous inference. Biometrics, 48, 1005–1013. Received 7 September 1999; revision received 31 August 2000