Plumage polymorphism in a newly colonized black sparrowhawk population: classification, temporal stability and inheritance patterns

bs_bs_bannerjournal of Zoology Plumage polymorphism in a newly colonized black sparrowhawk population: classification, temporal stability and inheritance patterns A. Amar, A. Koeslag & O. Curtis* Journal of Zoology. Print ISSN 0952-8369 Percy FitzPatrick Institute of African Ornithology, DST/NRF Centre of Excellence, University of Cape Town, Rondebosch, South Africa Keywords polymorphism; raptors; inheritance; Mendelian; morphs; pedigree data. Correspondence Arjun Amar, Zoology, Percy FitzPatrick Institute for African Ornithology, University of Cape Town, Private Bag X3, Rondebosch, Cape Town 7701, South Africa. Tel: +021 650 3304 Email: arjun.amar@uct.ac.za *Current address: Overberg Lowlands Conservation Trust, 3 de Kock St., Napier 7270. Editor: Andrew Kitchener Received 7 June 2012; revised 9 August 2012; accepted 14 August 2012 doi:10.1111/j.1469-7998.2012.00963.x Abstract Persistent plumage polymorphism occurs in around 3.5% of bird species, although its occurrence is not distributed equally across bird families or genera. Raptors show a disproportionately high frequency of polymorphism, and among raptors it is particularly frequent among the Accipiter hawks. However, no systematic study of polymorphism in this genus exists. Using a long-term study of the black sparrowhawk (Accipiter melanoleucus), a widespread polymorphic African Accipiter, we first demonstrate that the species shows discrete polymorphism (cf. continuous polymorphism), occurring as either dark or light morph adults, and that morph type and plumage pattern are invariant with age. We then demonstrate that adult morph type follows a typical Mendelian inheritance pattern, suggesting a one-locus, two-allele system within which the allele coding for the light morph is dominant. This inheritance pattern provides further support for classifying polymorphism in this species as discrete. In most of the species range the dark morph is the rarer morph; however, in our study population where the species is a recent colonist, over 75% of birds were dark and this remained fairly constant over the 10 years of our study. This reversal in morph ratio may represent an adaptive response to different environmental conditions or could be a founder effect with colonizing individuals having been mostly dark morph birds simply by chance. The extreme differences in environment conditions (seasonality of rainfall) that occur across the species range in South Africa provide support for an adaptive explanation, but further work is needed to test this hypothesis. Introduction Plumage polymorphism, in which different plumage morphs occur within the same age and sex of a breeding population, occurs in around 3.5% of bird species (Roulin, 2004). Evolutionary ecologists have long been fascinated by this phenomenon because the occurrence of two morphs in the same population runs counter to the notion that selective pressure should favour the optimal form for an environment, and any lesser quality individuals should be quickly eliminated (Huxley, 1955). Various explanations have been postulated for the occurrence and maintenance of polymorphism in birds (Galeotti et al., 2003; Roulin, 2004), and some of the most established hypotheses include: (1) apostatic selection (Fowlie & Krüger, 2003); (2) disruptive selection (Mather, 1955); (3) allopatric evolution (Cooke, Rockwell & Lank, 1995); (4) sexual selection (O Donald, 1983). Underpinning all these theories is the notion that an individual s phenotype is heritable, intransient, and not influenced by environmental variation, and that the colour variant is under selective pressure. However, in many studies these factors often remain untested. Polymorphism is particularly common in raptorial species (Fowlie & Krüger, 2003; Galeotti et al., 2003). Plumage colour in polymorphic raptors can vary continuously or may show two or more discrete morphs, for example, the polymorphic Swainson s hawk (Buteo swainsoni) and common buzzard (Buteo buteo) show continuous polymorphism, although they are often classified as dark, light or intermediate for analyses (Krüger & Lindström, 2001; Briggs, Collopy & Woodbridge, 2011). By contrast, discrete polymorphism exists in ferruginous hawk (Buteo regalis Schmutz & Schmutz, 1981) and Eleonora s falcon (Falco eleonorae Gangoso et al., 2011) with either dark or light morph birds. The type of phenotypic plumage polymorphism is likely to be influenced by the mode of genetic inheritance. Many studies have shown that polymorphic phenotypes are genetically determined in birds and follow a Mendelian mode of segregation (Roulin, 2004). To date, the use of pedigree data to study the genetic pattern of inheritance of plumage morphs Journal of Zoology (2012) 2012 The Authors. Journal of Zoology 2012 The Zoological Society of London 1

Polymorphism in black sparrowhawks A. Amar, A. Koeslag and O. Curtis in raptors has been reported in the common buzzard (Krüger & Lindström, 2001), ferruginous hawk (Schmutz & Schmutz, 1981), Swainson s hawk (Briggs, Woodbridge & Collopy, 2010a), Eleonora s falcon (Gangoso et al., 2011) and gyrfalcon (Falco rusticolus Chang, Lejeune & Cheng, 2010). The first four studies suggest a simple one-locus, two-allele autosomal inheritance pattern. For the ferruginous hawk and Eleonora s falcon that show discrete polymorphism, dark alleles are dominant and dark morph birds are thus either homozygous (with the alleles designated DD, capital letters for dominant alleles) or heterozygous (Dl), and light birds are homozygous for the recessive light allele (ll). For common buzzard and Swainson s hawk, dark (d) and light (l) alleles show incomplete dominance and heterozygous (dl) individuals therefore display intermediate plumage between the two homozygous morphs [dark (dd) or light (ll)], and hence give rise to continuous polymorphism along the plumage spectrum. For gyrfalcons, which show a full spectrum from pure white to pure black and many variants in-between, a more complex inheritance pattern is suggested, with colour being controlled by two genes, one controlling pigment production and the other restricting pigment distribution in feathers, with alleles in one gene having dominance and alleles in the other gene being co-dominant (Chang et al., 2010). Although several studies support a genetic basis to plumage variation, relatively few have demonstrated the stability of an individual s morph as it ages (Lowther, 1961; Lank et al., 1995; Brommer, Ahola & Karstinen, 2005). Criticism has been made of some studies which assume that plumage morph patterns remain constant over the course of an individual s life (Roulin, 2004). However, in the only study to examine this question in raptors, Briggs, Woodbridge & Collopy (2010b) found that plumage morph and patterning was invariant over time in 18 Swainson s hawks that were photographed at least 2 years apart. Among raptors, polymorphism occurs frequently in the genus Accipiter, with 11 of the 46 species displaying different colour morphs (Ferguson-Lees & Christie, 2001). However, no empirical research into polymorphism has focused on any species from this genus. Polymorphic species in this genus often show a similar polymorphic adult plumage, with a standard type, for example, common morph (light breast and underwing coverts) and a rarer dark adult morph, which tends to be black on the breast and underwing coverts (Ferguson- Lees & Christie, 2001). The black sparrowhawk (Accipiter melanoleucus) is a widely distributed Accipiter species occurring throughout much of sub-saharan Africa (Ferguson-Lees & Christie, 2001) and is usually described as showing these two morphs, although with varying degrees of white under the chin. The adult dark morph of this species is usually classed as rare (Steyn, 1982; Kemp & Kemp, 1998; Ferguson-Lees & Christie, 2001; Hockey, Dean & Ryan, 2005), although there are no published details on morph frequencies across the species range. The juveniles also occur as two morphs (pale or rufous) although these do not apparently reflect their subsequent adult morphs (Ferguson-Lees & Christie, 2001). The species has recently expanded into South Africa s Western Cape (Sebele, 2012) where it has successfully colonized the Cape Peninsula. The first breeding attempt on the Peninsula was recorded in 1993 (Oettlé, 1994; Curtis, Hockey & Koeslag, 2007) and the current population is estimated to be at least 40 breeding pairs. In this paper, using a long-term study of the black sparrowhawk on the Cape Peninsula, we undertake the first detailed study of polymorphism in an Accipiter species. Using photographs to score plumage characteristics we: (1) describe the type of polymorphism present and establish whether polymorphism in this species is best quantified as discrete or continuous, and (2) determine whether an individual s plumage pattern is invariant over time. Then, using pedigree data from wild, colour-ringed birds with known parental morphs, we explore plumage inheritance patterns to test for a genetic basis to the trait, and whether this follows conventional Mendelian inheritance patterns. Lastly, we examine the morph ratio of this newly colonized population and explore whether (1) it differs between the sexes of breeding adults, and (2) it has changed over time. Methods We monitored the black sparrowhawk population on the Cape Peninsula between 2001 and 2010. The study area features a matrix of habitats including urban gardens, alien pine (Pinus spp.) and Eucalyptus (Eucalyptus spp.) plantations, and small pockets of indigenous Afromontane forest and Fynbos. Altitudes where the birds breed range from sea level to about 300 m, and the climate is temperate, with locally variable winter rainfall (Cowling, MacDonald & Simmons, 1996). Mean annual rainfall is c.1250 mm, with average minimum and maximum monthly temperatures of 12 and 21 C, respectively (South African Weather Service). Monitoring was conducted during the breeding season (March November; Sebele, 2012) each year. Nests were located by surveying suitable stands of trees during the breeding season, searching for calling sparrowhawks, prey remains, whitewash and nest structures. Territories were visited regularly (approximately monthly) throughout the season until breeding was detected and then breeding attempts were monitored until conclusion. Where possible, we identified the morphs (dark or light) and sex of both parents attending a nest, which was possible in around 90% of breeding attempts. The species is easy to sex, with males weighing around 40% less than the females (unpublished data, see also Ferguson- Lees & Christie, 2001). In this study, data were only used for pairs where we knew the morph of each pair member. We fitted unique colour-ring combinations to as many breeding adult birds and 2 3-week-old nestlings as possible. Adults were trapped on territories using a bal-chatri baited with live white pigeons (Columba livia Berger & Mueller, 1959). Some of the birds ringed as chicks entered the breeding population during the course of the study, and enabled us to determine their adult morphs, which they began to acquire after their first year. In total, 102 adults (50 males, 52 females) and 131 chicks (76 males, 55 females of which 33 were later resighted in adult plumage) were fitted with colour rings during the course of the study. 2 Journal of Zoology (2012) 2012 The Authors. Journal of Zoology 2012 The Zoological Society of London

A. Amar, A. Koeslag and O. Curtis Polymorphism in black sparrowhawks Resightings of individual birds occurred mainly at breeding territories, but also through occasional observation away from breeding territories, via reporting and photographs taken by the authors and by members of the public. Photographs were taken of birds in adult plumage after being caught for ringing, or by using a 300-mm telephoto lens when birds were perched near the nest. Only photographs that showed the chin, throat, breast and flanks (hereafter front) were used for scoring coloration. Two observers (A. A. s data were used in the final analysis for ease of analysis) scored the percentage of white plumage on the front of birds in each photo. This approach to scoring plumage visually has been used in other studies, both with the birds in the hand (Brommer et al., 2005) and using photographs (Briggs et al., 2010b). Neither observer had knowledge of the other s scores, nor of the identity of the photographed bird. To explore morph frequencies between sexes and years, we used the data from all breeding pairs whose morphs were identified in each year (n = 245 pair years, between 2001 and 2010). However, this led to some level of pseudo-replication, since some birds featured in multiple years. For this reason, we ran a separate analysis, analysing the morph ratio of all pairs in the first year of breeding only (i.e. excluding known or suspected pairs from subsequent years; n = 130 pairs). Pairs and individuals were relatively faithful to nesting territories (unpublished data), and previously established or new pairs were classified based either on the pair s colour rings (n = 43), or in situations where only one bird was colour ringed, the combination of its colour rings and its partner s morph and lack of colour rings (n = 48), or in cases where neither bird was colour ringed (n = 39) we used only morph and sex combinations that were previously present on that territory to determine whether they were new or previously established pairs. We tested whether there were any differences between the sexes in the percentage of white on the front using a general linear model, with arcsine square root transformed percentage data as the response variable and the sex (M or F) as the explanatory variable. Analyses of the two morph types were conducted separately. Changes in frequency of morphs over time were analysed using a generalized linear model, with a binomial distribution and a logit link function. Whether a bird was a dark morph or a white morph (1/0) was specified as the response variable and year (as a continuous variable) was then specified as the explanatory variable. These analyses were carried out for males and females separately. All analyses were carried out in SAS version 9.1 (SAS Inc., 2004). Means are presented 1 standard error. Results Polymorphism classification and individual variance over time We had 135 photographs of individual adult birds which showed their front adequately to score the percentage of white plumage. These photos came from 42 different territories in the study area. Scores from the two observers did not differ significantly (paired t-test t 1,134 =-1.18, P = 0.23) and Frequency 40 35 30 25 20 15 10 5 0 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 % white plumage on front Figure 1 The frequency distribution of birds displaying different amounts of white plumage on their throat, chin, breast, and flanks, and whether they were classified as either dark (solid bars) or light (open bars) in the field. White plumage percentages were visually estimated from 135 photos. The data show there is a clear bimodal pattern and justify the division of birds into either dark or light morphs. there was close agreement between different observers in the scores allocated to individual birds (R 2 = 0.96, P < 0.0001). A histogram of the scores clearly showed a discrete bimodal distribution in plumage coloration (Fig. 1). Birds described as dark morphs or light morphs had on average 8% (range 0 33%) and 84% (68 92%) white plumage on their front, respectively. For a subset of these data, where the sex was known (n = 125), we explored, for each morph type, whether the percentage of white plumage on the front differed between the sexes. For light morphs, although sample size was small (n = 28), there appeared to be no difference in the percentage of white plumage between the sexes (males: 85.5 1.7%, females: 84.3 1.1%; c 2 1 = 0.54; P = 0.46). For dark morphs (n = 97), although not quite statistically significant, there was a tendency for males to have less white than females (males: 7.4 0.8%, females: 10.4 1.1%; c 2 1 = 2.90; P = 0.08). For six colour-ringed individuals (three males and three females; three dark morphs and three light morphs), we had photos taken at least 4 years apart (range 4 11 years). From the scores given to these birds, it was apparent that the morphs did not change over the years, with the small difference between years not showing any directional change over time and was most likely attributable to observer error (Table 1). Inheritance patterns of morphs We observed 33 birds (17 males, 16 females) in adult plumage that were colour ringed as chicks at nests where we also knew the morphs of both attending parents. These birds were produced in 18 different territories and from the parental colouring combinations they came from 18 different pairings. Thirteen dark x dark pairings produced 13 dark morph offspring (8 males, 5 females) and no light morphs, whereas from 19 dark x light pairings, 13 dark morph offspring (6 Journal of Zoology (2012) 2012 The Authors. Journal of Zoology 2012 The Zoological Society of London 3

Polymorphism in black sparrowhawks A. Amar, A. Koeslag and O. Curtis Table 1 Percentage of white plumage estimated from photographs of breeding black sparrowhawks, with photos spanning 4 or more years. Percentage of white on the front (chin, throat, breast and flanks) Sex Morph Mean ( SD) 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 Male L 78.2 ( 2.3) 75 80 78 80 Male L 84.4 ( 1.4) 84 86 84 82 84 85 86 Male D 5.3 ( 1.2) 6 5 5 7 7 4 4 5 Female L 88.8 ( 3.0) 92 90 90 84 88 Female D 2.6 ( 0.6) 3 2 3 Female D 4.6 ( 0.6) 5 4 5 SD, standard deviation. Table 2 Inheritance patterns according to the morph of the male and female parents and the morph (dark/light) of the offspring for all offspring and for male and female offspring separately. Male parental morph Dark Light All offspring Female parental morph Dark 13/0 5/0 Light 8/6 1/0 Male offspring Female parental morph Dark 8/0 3/0 Light 3/3 NA Female offspring Female parental morph Dark 5/0 2/0 Light 5/3 1/0 Birds were ringed as chicks with unique colour-ring combination. Inheritance patterns appear to follow a simple one-locus, two-allele system whereby the light allele is dominant. Light birds are therefore heterozygous with the genotype LL, or Ld, and dark birds are homozygous and have the genotype dd. Percentage of dark/light morphs 120 100 80 60 40 20 0 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 Year Figure 2 Percentage ( 1 standard error) of dark morph male (solid line) and female (dashed line) black sparrowhawks in the population on the Cape Peninsula between 2001 and 2010, from the 244 pairs for which the morphs of both birds were known. There were significantly more dark morph males than females (P < 0.001), but the frequency of dark morphs did not differ over time for either sex (P > 0.60). males, 7 female) and 6 light morphs (3 males, 3 females) were produced. We had only one record of a light x light pairing and this produced one dark offspring (Table 2). Inheritance was apparently autosomal rather than sex linked because, on several occasions, birds produced offspring of the same sex that differed from their own morph (Table 2). However, it was interesting to note that light morph males produced no light morph offspring (from six cases), whereas light females produced light offspring at c. the 1:1 ratio predicted, although admittedly sample size for this finding is small. From these data, inheritance of morph type appears to follow a simple Mendelian one-locus, two-allele system, whereby the light allele is dominant. Light birds are therefore homozygous or heterozygous (LL or Ld) and dark birds are invariably homozygous (dd). Seventy-six per cent of birds in our population were dark morphs (see later). Assuming these dark morphs are homozygous (dd based on the inheritance patterns observed in this investigation), we can calculate the expected frequencies of L and d alleles in our study population under Hardy Weinberg equilibrium conditions. We calculated an allelic frequency of 0.87 for the d allele and consequently 0.13 for the L allele. We therefore expect that a minimum of 22% of birds are heterozygous light morphs (Ld) and only around 2% of birds are homozygous light morphs (LL). Stated differently, of the white morphs the vast majority (92%) are heterozygous (Ld) while only 8% are homozygous (LL). Morph frequencies over time and between sexes Over the entire study period, 76% of birds were dark morph, with the frequency being significantly higher for males than females (males: 83 2%, females 68 3%; c 2 1 = 14.85; P = 0.0001; Fig. 2). Between 2001 and 2010, the frequency of the two morphs did not change, either for the population as a whole (c 2 1 = 0.16, P = 0.68), or for either sex (males: c 2 1 = 0.12, P = 0.72; females: c 2 1 = 0.06, P = 0.80; Fig. 2). Using data only from the first year pairs were found breeding (to avoid issues of pseudoreplication), the same patterns emerged. There was no change 4 Journal of Zoology (2012) 2012 The Authors. Journal of Zoology 2012 The Zoological Society of London

A. Amar, A. Koeslag and O. Curtis Polymorphism in black sparrowhawks over time detected for either sex (males: c 2 1 = 0.02, P = 0.87; females: c 2 1 = 0.70, P = 0.40) and the frequency of dark morphs was significantly higher for males than females (males: 85 3%, females 73 4%; c 2 1 = 5.41; P = 0.02). Discussion Our study showed that (1) black sparrowhawks in our study area display discrete polymorphism in plumage coloration; (2) morph type and plumage patterns of individual birds change little, if at all, with age; (3) the observed inheritance pattern suggests that morph type is under genetic control consistent with a single Mendelian locus at which the light allele is dominant; (4) the Cape Peninsula population shows a reversal of the morph frequencies generally encountered in the rest of this species range, with over 75% of birds being dark morphs, and with significantly more dark morph males than females. From our plumage scores, it was clear that although there was considerable variation in the percentage of white plumage, black sparrowhawks show two distinct morphs, revealed as two clear modes on the histogram of percentage white plumage (Fig. 1). White morphs usually had around 85 90% and never less than 70% white plumage on their front whereas dark morphs usually had between 0 10% and never more than 35% white frontal plumage. There was clear separation of these modes, with no birds of intermediate plumage. Although not statistically significant, the percentage of white on the front of dark morph birds tended to be greater for females than for males, potentially indicating that darker males may enjoy a selective advantage in this population. Similar sex differences have been observed for polymorphic barn owls (Tyto alba), with females on average more reddish brown with more prominent black spotting than males (Roulin, 1999). It is also interesting that there were significantly more light morph females in the population than males (see later). However, whether the frequency of white morphs in the population and the percentage of white on the front of dark morph birds are in any way linked through similar selective pressures is unknown. Observation of known colour-ringed birds over time (4 11 years) provided strong evidence that plumage coloration in this species is fixed and does not vary with age (within this study s time frame). The species average life expectancy is around 10 years (based on average annual adult survival rates of 89% unpublished data). Data from males and females of both plumage morphs suggested that the proportion of white plumage on their front varied little with age. Indeed, the variation between years is probably attributable to observer error. Closer examination of the photographs of known birds suggests that plumage variability in this species might be successfully used to identify individuals over time. For example, obvious features of the plumage, such as the location of dark or light streaks, also appear to remain constant over time. This work therefore adds to that of Briggs et al. (2010b) in suggesting that plumage patterns in polymorphic raptors persist over an individual s lifetime, and provide further support for other studies that identify individuals in highly polymorphic populations by plumage alone (e.g. Krüger & Lindström, 2001; Curtis et al., 2005). Furthermore, because it is unlikely that these individuals would have experienced the same environmental conditions in the different years of their lives, it does suggest that environmental variables, such as weather, or body condition is relatively unimportant in determining the plumage patterns in this species, as Briggs et al. (2010b) also concluded. This result, together with our other findings on inheritance patterns (see later), further strengthens the argument that these plumage patterns may well be under genetic control. Observations of birds in adult plumage which were colour ringed as chicks with known parental morphs revealed that morph inheritance showed a classic Mendelian pattern. Dark x dark crosses produced only dark offspring, dark x light crosses produced both dark and light morph offspring, and in one instance a light x light cross produced a dark offspring. Hartley (1976) also reports an instance of light x light black sparrowhawk pairing producing a dark morph bird. From these findings, it therefore appears that polymorphism in this species is under genetic control with a one-locus, two-allele inheritance pattern, and that the light allele is dominant. Thus, light birds can be either homozygous (LL) or heterozygous (dl), whereas dark morph birds are always homozygous (dd). In this study, we have assumed no extra-pair paternity, which although unlikely to be true, occurs at a relatively low rate among raptors (Mougeot, 2004). Our finding of no impossible offspring from any pairings, given our predictions, also suggests that extra-pair paternity was not particularly high or not high enough to disrupt these results. Several Accipiter spp. occurring across the globe show similar polymorphic patterns to the black sparrowhawk (i.e. with dark or light morphs Ferguson-Lees & Christie, 2001): it would be interesting to know whether the same inheritance patterns also operate in these species. We could not use our data on the morphs of offspring from known parents to test accurately whether the phenotypes of these offspring were as predicted from our assumed genetic inheritance patterns (i.e. one locus, two alleles, with the light allele being dominant), because the offspring s morph was only assigned once a bird was recruited into the adult population rather than as a chick. Thus, any morph-linked differences in dispersal or juvenile survival would produce biased results. However, if we do assume equal juvenile survival rates between morphs, and given our estimated distribution of genotypes [i.e. 92% of light birds being heterozygous (dl)], we would predict an approximate dark : light morph ratio of 1:1.17 from our mixed-pair crosses. However, from the 19 offspring recruited from mixed pairs, we found a ratio of 2.2:1 in favour of dark morphs. Thus, more than twice as many dark birds as expected were subsequently found breeding from these crosses. This in turn suggests that either survival of white morphs prior to recruitment is lower than that of dark morphs, or that light morphs disperse further away from our study areas, or that we have incorrectly identified the inheritance mechanism. Although inheritance in our study was apparently not sex linked, no light offspring were ever recruited from light fathers, and this result largely explains the lower than predicted occurrence of light birds recruited from mixed pairs; however, the sample Journal of Zoology (2012) 2012 The Authors. Journal of Zoology 2012 The Zoological Society of London 5

Polymorphism in black sparrowhawks A. Amar, A. Koeslag and O. Curtis size (n = 6) for this finding was small. To investigate this further will require the use of genetic markers to test the genotypes of the chicks produced. Many studies have now shown the importance of the melanocortin-1 receptor gene in determining similar plumage polymorphisms in other bird species (Mundy, 2005) including for some (Gangoso et al., 2011) but not all raptor species (Hull et al., 2010). In future studies, we hope that the use of molecular markers will allow us to determine the morph genotype of the young chicks and to examine whether juvenile and sub-adult survival differs between morphs. Our Hardy Weinberg estimates suggested that 92% of the light morphs in our population were heterozygous. However, while this statistic is useful, we acknowledge that the accuracy of Hardy Weinberg equation relies on a number of assumptions, such as no selection, no non-random genetic drift and no gene flow, which are unlikely to be strictly true in our study population. For the current study, we have focused on polymorphism of adult black sparrowhawks. However, juvenile black sparrowhawks are also polymorphic, displaying a pale morph and a rufous morph (Ferguson-Lees & Christie, 2001). Initially, juvenile morphs were thought to be sex linked; however, pale and rufous juveniles of both sexes have been found in the same nest (Ferguson-Lees & Christie, 2001). Juvenile morph is also apparently not directly linked to adult phenotype, because of two rufous juveniles that were followed into adulthood, one became a light morph and the other a dark morph adult (Ferguson-Lees & Christie, 2001). Rather, we hypothesize that juvenile morphs could be linked to the adult morph genotype in the following way: birds that carry the recessive dark allele would be rufous juveniles and those birds that are homozygous light morphs (LL) would be pale morphs. Unfortunately, we do not yet have the data to test this hypothesis (e.g. from enough chicks with known morphs). The higher frequency of dark morph males as compared with females was an interesting result. One possible explanation for this could be that dark morph males survive better than dark morph females, and thus more dark morph males are recruited into the population. Again, an alternative explanation could be that females disperse further than males and that immigrant females come from population with a lower frequency of dark morphs. However, of the six light birds recruited into the population, three were male and three were female. Alternatively, this difference could have come about through mate choice, for example, if males actively select light females or if females actively select dark males, such similar mate choice has been reported for other species (Knapton & Falls, 1983; O Donald, 1983), and therefore remains a potential explanation for this finding. Dark morph adults were the more common (> 75%) morph in our population, which contrasts with most other sources that suggest that dark morphs are the rarer morph for this species (Steyn, 1982; Kemp & Kemp, 1998; Ferguson-Lees & Christie, 2001; Hockey et al., 2005). We are unaware of any published data on the morph frequency from other populations. However, unpublished data from a population studied in eastern Mpumalanga (formerly known as the Transvaal), South Africa (for details, see Tarboton & Allan, 1984), supported the idea that dark morphs are usually the rare morph for this species, with only around 22% (n = 36) of birds being dark morphs (W. Tarboton, unpublished data). What might explain the almost complete reversal in the frequency of the different morphs in these two study populations? There are two possible explanations for this pattern. Firstly, that it is simply a founder effect, with the first colonizing birds arriving on the Cape Peninsula being dark morphs, purely by chance. Alternatively, that this is an adaptive response to different environmental conditions (e.g. the different rainfall seasonality of the two regions). Both populations breed over the winter months (Sebele, 2012): this period coincides with the dry season in the north and east of South Africa, but with the wet season in the south west of the country. Thus, individuals breeding in our study population are exposed to far higher rainfall levels during the breeding period. Recent reviews on the causes and functions of polymorphism (Galeotti et al., 2003; Roulin, 2004) have both suggested a strong link between light conditions and variations in plumage, with crypsis (background matching) likely to play a key role. Thus, it may be that dark morph birds in our study population are at a selective advantage because they benefit from improved hunting efficiency in the poorer light conditions that would be associated with higher rainfall. As with most Accipiters, males provide most of the food during the breeding season, feeding the female during the incubation and early nestling stage, and this could explain the different morph frequencies between sexes if there is greater selection pressure for males to be dark than females. Similar relationships between habitat background and colour-morph ratio have been identified for other bird species (e.g. Rohwer, 1990). Alternatively, pressure from parasites may be greater in these wet conditions which may favour dark morphs, as different raptor morphs can show different immune responses (Gangoso et al., 2011) and subsequent parasite loads (Chakarov, Boerner & Kruger, 2008). It is also possible that dark morph birds have a thermal advantage in these colder wet conditions or that darker birds are able to withstand the feather degrading bacteria which may be more abundant or virulent in such conditions (Burtt & Ichida, 2004). To test further whether the unique morph ratios found in this population are likely to be an adaptive trait, future research will focus on whether dark morphs have any selective advantage by comparing reproductive output and survival with light morph birds: the findings presented here lay the foundations for this future research. Acknowledgements We are very grateful to Sharon Yodaiken and Gerry Meihuizen for all their help in the field. We are also grateful to Jacqueline Bishop and Gareth Tate who helped improve the manuscript immensely. We thank Warwick Tarboton for providing unpublished information on the morphs of his study population. Thanks are also due to Lovelater Sebele, Fitzum Baldi and Rowan Martin for useful discussions and advice. We thank M. Boerner and an anonymous reviewer for their comments which greatly improved the manuscript. We are 6 Journal of Zoology (2012) 2012 The Authors. Journal of Zoology 2012 The Zoological Society of London

A. Amar, A. Koeslag and O. Curtis Polymorphism in black sparrowhawks also grateful to all the land owners who have granted us access to carry out fieldwork on their land, and we are particularly grateful to South African National Parks for access to the Table Mountain National Park. References Berger, D.D. & Mueller, H.C. (1959). The bal-chatri: a trap for the birds of prey. Bird-Banding 30, 18 26. Briggs, C.W., Collopy, M.W. & Woodbridge, B. (2011). Plumage polymorphism and fitness in Swainson s hawks. J. Evol. Biol. 24, 2258 2268. Briggs, C.W., Woodbridge, B. & Collopy, M.W. (2010a). Inheritance patterns of plumage morph in Swainson s hawks. J. Raptor Res. 44, 232 235. Briggs, C.W., Woodbridge, B. & Collopy, M.W. (2010b). Temporal morph invariance of Swainson s hawks. J. Raptor Res. 44, 70 73. Brommer, J.E., Ahola, K. & Karstinen, T. (2005). The colour of fitness: plumage coloration and lifetime reproductive success in the tawny owl. Proc. Roy. Soc. Lond. B 272, 935 940. Burtt, E.H. & Ichida, J.M. (2004). Gloger s rule, featherdegrading bacteria, and colour variation among song sparrows. Condor 106, 681 686. Chakarov, N., Boerner, M. & Kruger, O. (2008). Fitness in common buzzards at the cross-point of opposite melanin-parasite interactions. Functional Ecology 22, 1062 1069. Chang, V.S., Lejeune, J. & Cheng, K.M. (2010). The pattern of inheritance of melanin-based plumage color variants in the gyrfalcon (Falco rusticolus). J. Raptor Res. 44, 224 232. Cooke, F., Rockwell, R.F. & Lank, D.B. (1995). The snow geese of La Pérouse Bay, natural selection in the wild. Oxford: Oxford University Press. Cowling, R.M., MacDonald, I.A.W. & Simmons, M.T. (1996). The Cape Peninsula, South Africa: physiographical, biological and historical background to an extraordinary hot-spot of biodiversity. Biodivers. Conserv. 5, 527 550. Curtis, O., Malan, G., Jenkins, A. & Myburgh, N. (2005). Multiple brooding in birds of prey: South African black sparrowhawks Accipiter melanoleucus extend the boundaries. Ibis 147, 11 16. Curtis, O.E., Hockey, P.A.R. & Koeslag, A. (2007). Competition with Egyptian geese Alopochen aegyptiaca overrides environmental factors in determining productivity of black sparrowhawks Accipiter melanoleucus. Ibis 149, 502 508. Ferguson-Lees, J. & Christie, D. (2001). Raptors of the world. London: Christopher Helm. Fowlie, M.K. & Krüger, O. (2003). The evolution of plumage polymorphism in birds of prey and owls: the apostatic selection hypothesis revisited. J. Evol. Biol. 16, 577 583. Galeotti, P., Rubolini, D., Dunn, P.O. & Fasola, M. (2003). Colour polymorphism in birds: causes and functions. J. Evol. Biol. 16, 635 646. Gangoso, L., Grande, J.M., Ducrest, A.L., Figuerola, J., Bortolotti, G.R., Andrés, J.A. & Roulin, A. (2011). MC1R-dependent, melanin-based colour polymorphism is associated with cell-mediated response in the Eleonora s falcon. J. Evol. Biol. 24, 2055 2063. Hartley, R. (1976). Some notes on the plumages of the black sparrowhawk. Bokmakierie 28, 61 63. Hockey, P.A.R., Dean, W.R.J. & Ryan, P.G. (2005). Roberts birds of southern Africa. 7th edn. Cape Town: Trustees of the John Voelcker Bird Book Fund. Hull, J.M., Mindell, D.P., Talbot, S.L., Kay, E.H., Hoekstra, H.E. & Ernest, H.B. (2010). Population structure and plumage polymorphism: the intraspecific evolutionary relationships of a polymorphic raptor, Buteo jamaicensis harlani. BMC Evol. Biol. 10, 224. Huxley, J.S. (1955). Morphism in birds. Acta 6th International Ornithological Congress, Basel, 1954, pp. 309 328. Kemp, A. & Kemp, M. (1998). Birds of prey of Africa and its islands. Cape Town: New Holland. Knapton, R.W. & Falls, J.B. (1983). Differences in parental contribution among pair types in the polymorphic whitethroated sparrow. Can. J. Zool. 61, 1288 1292. Krüger, O. & Lindström, J. (2001). Lifetime reproductive success in common buzzard, Buteo buteo: from individual variation to population demography. Oikos 93, 260 273. Lank, D.B., Smith, C.M., Hanotte, O., Burke, T. & Cooke, F. (1995). Genetic polymorphism for alternative mating behaviour in lekking male Ruff Philomachus pugnax. Nature 378, 59 62. Lowther, J.K. (1961). Polymorphism in the white-throated sparrow, Zonotrichia albicollis (Gmelin). Can. J. Zool. 39, 281 292. Mather, K. (1955). Polymorphism as an outcome of disruptive selection. Evolution 9, 52 61. Mougeot, F.S. (2004). Breeding density, cuckoldry risk and copulation behaviour during the fertile period in raptors: a comparative analysis. Anim. Behav 67, 1067 1076. Mundy, N.I. (2005). A window on the genetics of evolution: MC1R and plumage colouration in birds. Proc. Roy. Soc. Lond. B 272, 1633 1640. O Donald, P. (1983). The arctic skua: a study of the ecology and evolution of a seabird. Cambridge: Cambridge University Press. Oettlé, E. (1994). Black sparrowhawk breeds on the Cape Peninsula. Promerops 212, 7 7. Rohwer, S. (1990). Foraging differences between white and dark morphs of the Pacific Reef Heron Egretta sacra. Ibis 132, 21 26. Roulin, A. (1999). Nonrandom pairing by male barn owls Tyto alba with respect to a female plumage trait. Behav. Ecol. 10, 688 695. Journal of Zoology (2012) 2012 The Authors. Journal of Zoology 2012 The Zoological Society of London 7

Polymorphism in black sparrowhawks A. Amar, A. Koeslag and O. Curtis Roulin, A. (2004). The evolution, maintenance and adaptive function of genetic colour polymorphism in birds. Biol. Rev. 79, 815 848. SAS Institute Inc. (2004). SAS/STAT 9.1 User s Guide. North Carolina: SAS Institute Inc. Schmutz, S.M. & Schmutz, J.K. (1981). Inheritance of color phases of ferruginous hawks. Condor 83, 187 189. Sebele, L. (2012). Factors influencing the timing of breeding in a range expanding raptor at two spatial scales. Unpublished MSc thesis, University of Cape Town. Steyn, P. (1982). Birds of prey of southern Africa: their identification and life histories. Cape Town: David Philip. Tarboton, W.R. & Allan, D.G. (1984). The status of birds of prey in the Transvaal. Pretoria: Transvaal Museum. 8 Journal of Zoology (2012) 2012 The Authors. Journal of Zoology 2012 The Zoological Society of London