Abstract.
Young, immature-plumaged males of species in which plumage maturation is delayed are often unable to obtain mates or pair later in the season to lower-quality females. Their clutches are smaller and the quality of their eggs is lower, leading to fewer nestlings, whose quality and fledging success are reduced. The Saffron Finch (Sicalis flaveola) is an understudied and declining secondary-cavity nester with delayed plumage maturation; yearling males look like females. The aims of this study, of birds nesting in boxes, were to compare the morphology of immature- and mature-plumaged males of the Saffron Finch and their respective females and to compare the breeding success of females mated to each category of male. We expected that females mated to immature-plumaged males would follow the pattern described for other species, having lower reproductive investment and success. Immature- and mature-plumaged males were similar in morphology, but females paired to mature-plumaged males were bigger in bill height and weight. We found no differences, however, between females paired to each category of male in any biologically meaningful variable. Reduced competition among males for nest cavities and enough females willing to pair with immature-plumaged males might be partially responsible for these results.
Resumen.
Con frecuencia, los machos jóvenes con plumaje inmaduro pertenecientes a especies con maduración retrasada del plumaje son incapaces de obtener parejas o se aparean más tarde en la estación reproductiva y con hembras de menor calidad. Sus nidadas son más pequeñas y la calidad de sus huevos menor, lo que conduce a menos pichones cuya calidad y éxito de emplumamiento son reducidos. Sicalis flaveola es una especie poco estudiada que anida secundariamente en cavidades, cuyas poblaciones están disminuyendo y que presenta maduración retrasada del plumaje: los machos de un año tienen apariencia de hembras. Los objetivos de este estudio, basado en aves que estaban anidando en cajas, fueron comparar la morfología de machos de S. flaveola con plumaje inmaduro y con plumaje maduro y la de sus respectivas parejas, y comparar el éxito reproductivo de las hembras apareadas con machos de cada categoría. Esperábamos que las hembras apareadas con machos con plumaje inmaduro siguieran el patrón descrito para otras epecies, presentado una menor inversión en la reproducción y un menor éxito reproductivo. Los machos con plumaje inmaduro y con plumaje maduro fueron similares en morfología, pero las hembras apareadas con machos con plumaje maduro presentaron picos más altos y pesos mayores. Sin embargo, no encontramos diferencias entre las hembras apareadas con las distintas categorías de machos en ninguna variable biológicamente importante. La competencia reducida entre machos por cavidades de anidación y la existencia de suficientes hembras dispuestas a aparearse con machos con plumaje inmaduro podrían explicar parcialmente estos resultados.
Introduction
In many birds reproductive success varies by age, being lower for younger individuals and increasing later in life (Sæther 1990, Forslund and Pärt 1995). Older, usually brighter males (Komdeur et al. 2005, Bitton et al. 2008) have access to higherquality territories and mates (Pärt 2001, Komdeur et al. 2005, Nagy and Holmes 2005) or sire more extra-pair young (Morton et al. 1990, Greene et al. 2000, Schmoll et al. 2007). In addition, females can increase their reproductive investment by selecting mates of higher phenotypic quality (differential allocation hypothesis; Burley 1986, 1988, Sheldon 2000). Females mated to more attractive males may start to breed earlier, may lay more eggs (Balzer and Williams 1998, Johnsen et al. 1998, Rintamaki et al. 1998), may lay larger eggs, or may lay eggs of higher quality (Gil et al. 1999, Cunningham and Russell 2000), leading to more competitive offspring in better condition (Cunningham and Russell 2000, McGraw et al. 2001, Velando et al. 2006).
In many birds, plumage maturation is delayed, sub-adults not acquiring the definitive nuptial plumage until after their first potential breeding season (Rohwer et al. 1980, Hill 1996), making young birds easier to identify. This pattern applies commonly to males (Lyon and Montgomerie 1986, Hill 1988) but only rarely to females (Hussell 1983, Mountjoy and Robertson 1988). Although some immature-plumaged males reproduce (Procter-Gray and Holmes 1981, Hill 1988, Powlesland 2002), more often they are unable to obtain high-quality territories or attract mates (Greene et al. 1996, Pärt 2001, Valle et al. 2006), have lower resource-holding potential (Arcese and Smith 1985), and are socially subordinate to mature-plumaged males (Rohwer et al. 1980, Sundberg 1995, Greene et al. 2000). When immature-plumaged males do reproduce successfully, pairs initiate egg laying later, have smaller clutches, and fledge fewer young (Hakkarainen et al. 1993, Karubian 2002, Komdeur et al. 2005).
The Saffron Finch (Sicalis flaveola) is a granivorous secondary-cavity nester whose breeding biology remains almost unstudied (see Mason [1985] for his findings during irregular visits to 23 nests and De la Peña [1995] for the content of 11 nests). This species does not hold territories year round and in winter flocks with other passerines (Palmerio, pers. obs). It nests in abandoned nests of the Rufous Hornero (Furnarius rufus), natural cavities in trees, bridges, and buildings, and nest boxes (Mason 1985, this study). Subspecies pelzelni ranges from southern Brazil and eastern Bolivia to central Argentina, south to La Pampa Province (Ridgely and Tudor 1989). Although it is relatively abundant in Argentina (Narosky and Yzurieta 1987), inhabiting rural zones and the ecotone between forests and grasslands, its numbers are diminishing because of its capture as a cage bird, making it vulnerable in Buenos Aires Province (Narosky and Di Giácomo 1993). In males, plumage maturation is delayed; from our data on males banded as nestlings, we know that in this subspecies they do not reproduce in the same breeding season they are hatched, retain the immature plumage through the entire year of hatching, are able to reproduce at an age of one year while still wearing the immature plumage (this study), and all acquire the golden-yellow color with olive-streaked black upperparts after their first potential breeding season is over. Yellow (mature-plumaged) males develop an orange crown only during the breeding season. During their first potential breeding season, males look like females, grayish, streaked with black above, with breast and flanks whitish, streaked brownish (Narosky and Yzurieta 1987).
The purpose of this study was to compare the morphology of the Saffron Finch by age and sex class, and to compare the breeding biology of females mated to immature- and mature-plumaged males. We predicted that immature-plumaged males and their females should be smaller than mature birds and that females mated to immature-plumaged males should breed later in the season, should lay smaller clutches and smaller eggs, and should have a lower ratio of egg to body weight and lower fledging success and/or lower-quality nestlings.
Methods
We studied Saffron Finches nesting in boxes in flat farmland at the Instituto Tecnológico de Chascomús (InTeCh), Buenos Aires Province, Argentina (35° 34′ S, 58° 01′ W), from October to March from 2004–2005 to 2006–2007. In the study area monthly mean daily temperatures ranged from 18 °C in October to 23 °C in February and 20 °C in March. The nest boxes (96, 117, and 123 in the 2004–2005, 2005–2006, and 2006–2007 breeding seasons, respectively) were placed 1.3–1.7 m above the ground on fence posts delimiting paddocks, approximately 30 m from each other, and distributed in an area of 70 ha. The paddocks are edged with tala (Celtis tala) and espinillo (Acacia bonariensis); the buildings are surrounded by a mixed forest of eucalyptus (Eucalyptus sp.), pine (Pinus sp.), honey locust (Gledtisia triacanthos), and tala, and within the study area are several small patches of eucalyptus.
The boxes measured 25 × 17 × 13 cm, had an entrance hole 4 cm in diameter, a lateral door that allowed monitoring of the nest's progress, and a “wig-wag” trap for capturing the breeding adults when they enter the box. The trap is activated by pulling on a long piece of monofilament fishing line attached to a swiveling piece of wood when a bird enters the box. Females were trapped late in incubation, males on the day of hatching, as earlier captures prompted nest abandonment. The birds were given a numbered aluminum band and a unique combination of plastic color bands. Using a digital caliper (±0.01mm), we measured bill height, bill width, and tarsus (from the joint with the tibiotarsus to the distal end of the last tarsal scale; NABC 2003). We measured the flattened wing span with a tapered rule and body mass with a spring balance (±0.25 g). We sexed the mature-plumaged males by color and distinguished immature-plumaged males from females by their singing (only males sing), absence of brood patch (only females incubate, AGP, pers. obs.), and cloacal protuberances.
We visited the nest boxes every other day, except during the presumed egg-laying period, hatching, and fledging, when we visited the nests daily (Martin and Geupel 1993, Martin et al. 1996). For each pair, we recorded the date the female laid the first egg of the season as the initiation of breeding. We marked eggs with water-proof ink at the lesser pole (foot, sensu Gosier et al. 2005), measured their length and width to the nearest 0.01 mm with a digital caliper, and weighed them in the early morning to ascertain their freshness (Acculab PP 62, ±0.005 g). Egg volume was calculated as length × width2 × 0.5071 (Manning 1979). Initial clutch size was recorded when no additional eggs were laid for two consecutive days. Incubation duration was calculated as the time from the onset of incubation, detected by early morning visits, to the hatching of the first egg.
Chicks were marked on the legs with nontoxic indelible markers and weighed every other day with spring balances of 10 g (± 0.1) and 20 g (± 0.25). At day 8 of life (day of hatching = day 0) they were given a numbered aluminum band. We measured the mass gain of nestlings and adjusted the growth of each nestling to a Gompertz sigmoid function y(t) = aex, where x = eKt, a = asymptotic weight, and K is the growth rate. We averaged the parameters of the nestlings belonging to each nest and compared a and K for nests of females mated to immature-plumaged males to those for nests of females mated to mature-plumaged males.
We estimated the nestling period as the time from the hatching of the first egg the departure of the last nestling and defined fledging success as the number of nestlings fledged per number of nestlings hatched. We considered nests to have fledged young successfully if one or more fledglings were observed near the nest on the presumed day of fledging and defined nesting success as the number of nests fledging at least one young per number of nests initiated.
Statistical Analyses
The morphology of immature- and mature-plumaged males and of females paired to them was analyzed by Mann—Whitney U-tests, as the variables were not normally distributed and could not be normalized. The potential variation by year of morphological variables was analyzed with Kruskal—Wallis tests. We analyzed the onset of breeding attempts, number of clutches, clutch and egg sizes, incubation length, growth of nestlings, nestling period, fledging and nesting success of females paired to immature- and mature-plumaged males, and their interactions with generalized linear mixed models, with stepwise backward elimination, from GenStat release 11 (VSN International, Hemel Hempstead, UK). We used male color as the fixed factor in all cases and, to account for multiple broods within a breeding season and across years (Shall 1991), we included the identity of the females, the number of the brood nested in female's identity, the year, and egg-laying date as random factors. To compare the date a pair's first clutch was laid and the number of clutches laid by females mated to immature- and mature-plumaged males, we used the female's identity and year as random factors. We analyzed the following variables with a Poisson error distribution and logarithm-link function: date of laying of first clutch, number of clutches, egg and clutch size, number of fledglings, and incubation and nestling period. We used a binomial error distribution and a logit-link function to analyze hatching, fledging, and nesting success. Finally, we analyzed the asymptotic weight and the growth rate K with a normal error distribution and a logarithm-link function. We used the Wald statistic, with F distribution, to examine the significance of the fixed effect (P < 0.05) and calculated the power associated with nonsignificant results according to Seavy et al. (2005). We compared the number of nests lost during the egg and nestling periods by male color with a χ2 test for differences of proportions. Results are reported as means ± SE.
Results
We followed 167 nesting attempts (64 of immature-plumaged and 103 of mature-plumaged males); the percentage of nest boxes occupied was 26%, 50%, and 55% during the 2004–2005, 2005–2006, and 2006–2007 breeding seasons, respectively. During that period we detected 11 nests built in natural cavities; six of them belonged to immature-plumaged males, five to mature-plumaged males, but most of them were out of reach and not followed. Both sexes contributed to building the nest, at the center of the box on a mat of dry grasses lined with horse hair and thin grasses. The breeding season lasted approximately 5 months; the earliest nesting attempt was made on 6 October, the latest on 9 March. The mating system was socially monogamous, the pair staying together throughout the breeding season, rearing up to three broods. Twenty-four males were monitored only during one breeding season, 13 for two seasons, and three for three seasons. Nineteen females were present only one breeding season, 10 during two, and 16 during three breeding seasons.
We found no significant difference by year in bill height or width, tarsus or wing length, or body mass of immatureplumaged males (n = 26), mature-plumaged males (n = 33), or females mated to each category of male (Kruskal—Wallis test, P > 0.05 in all cases). Similarly, we found no significant difference between immature-plumaged (n = 26) and matureplumaged males (n = 33) in any of these variables (Mann—Whitney U-test, P > 0.05 in all cases). Females mated to mature-plumaged males, however, were significantly bigger in bill height and body weight than those mated to immatureplumaged males (Table 1).
Morphology of female Saffron Finches (mean ± SE). Differences between females mated to immature- and mature-plumaged males compared with the Mann—Whitney U-test.
Morphology of female Saffron Finches (mean ± SE). Differences between females mated to immature- and mature-plumaged males compared with the Mann—Whitney U-test.
Results of the generalized linear mixed model used to analyze the breeding success of females paired to immature- and mature-plumaged males. Fixed factor: male color; random factors: female, female × number of brood, year, date of laying. Reported are means ± SE, n = sample size for each variable; numbers in parentheses refer to the power of the test (see Methods).
Results of the generalized linear mixed model used to analyze the breeding success of females paired to immature- and mature-plumaged males. Fixed factor: male color; random factors: female, female × number of brood, year, date of laying. Reported are means ± SE, n = sample size for each variable; numbers in parentheses refer to the power of the test (see Methods).
Breeding Biology of Females Mated to Immature- and Mature-Plumaged Males
Females mated to immature- and mature-plumaged males did not differ in either date of laying of the first clutch or in the number of broods raised in the season (Table 2). The categories of females did not differ significantly in clutch size, egg size, egg mass relative to the female's body weight, or incubation period (Table 2).
At hatching, the body weight of nestlings was 2.11 ± 0.04 g, and we found no differences in the growth of nestlings in nests of immature- and mature-plumaged males, either in the growth rate K or in asymptotic weight, which was attained around day 8 after hatching (Table 2, Fig. 1). Both sexes contributed to feeding the nestlings and removing fecal sacs (Palmerio, pers. obs.). Nestlings belonging to immatureplumaged males nests tended to leave the nest half a day later (at 14.6 ± 0.2 days vs. 14.1 ± 0.2 days for nestlings of matureplumaged males), but this result did not reach significance (P = 0.06, Table 2). Finally, the total number of fledglings per female and per season, the fledging success, and the nesting success of the two categories of nests were not significantly different (Table 2).
Nest failures. During the egg stage Saffron Finches lost 62 nests, nearly half of them because of competition for the box with Southern House Wrens (Troglodytes musculus) (n = 14), Bay-winged Cowbirds (Molothrus or Agelaioides badius) (n = 12), and White-rumped Swallows (Tachycineta leucorrhoa) (n = 4); the remaining causes were nest predation and destruction by unidentified species. We found no differences between immature- and mature-plumaged males in the percentage of nests lost during this period (12 % and 19 %, respectively, nimmature = 64, nmature = 103, P = 0.24). Saffron Finches lost 14 nests during the nestling period, likely to rats (Rattus rattus and R. norvegicus) and small opossums (Didelphis albiventris). Again, nests attended by immature- and mature-plumaged males did not differ in the percentage of nests lost during this period (3% and 9%, respectively, nimmature = 64, nmature = 103, P = 0.13).
Discussion
This is the first study to describe the breeding biology of the Saffron Finch in the wild. We found that immature-plumaged males of this species often nest successfully in boxes; moreover, the reproductive biology of females paired to them did not differ from that of females paired to mature-plumaged males in any relevant variable.
Immature- and mature-plumaged males, clearly distinguishable by color, were not different in body size. Several studies of other birds, however, have found that young and dull individuals are smaller and lighter (Järvi et al. 1987, Wagner and Morton 1997, Mitrus 2007). In contrast, in the Saffron Finch females mated to mature-plumaged males were larger than females mated to immature-plumaged males, as Wagner and Morton (1997) found in the Purple Martin (Progne subis); in particular, Saffron Finches females mated to matureplumaged males had deeper bills and were heavier. If females' differences in size correspond to differences in age, it suggests that Saffron Finches mate assortatively by age. Assortative mating by age and/or quality has been found in several species (Mountjoy and Robertson 1988, Potti 2000); if both males and females are exerting mate choice, mature-plumaged males may have had the opportunity to select females in better condition and/or more experienced (Weggier 2001), which should produce more young. Unfortunately, not knowing the age of the banded females we cannot address this subject, but even if there is age-assortative mating in this species, it did not result in differences in breeding success.
Brood averages of chick weight ± SE as a function of age (days). Numbers above and below the curves represent the sample size of broods weighed that belong to mature-plumaged (white circles) and immature plumaged males (black circles), respectively.
Several studies have found that immature-plumaged males arrive on breeding grounds later (Balbontin et al. 2007), are unable to attract mates, or do so later in the season (Ficken and Ficken 1967, Greene et al. 2000, Kingma et al. 2008). Females mated to immature-plumaged males start to lay eggs later than do those mated to mature-plumaged males (Lanyon and Thompson 1986, Sæther 1990, Komdeur et al. 2005) and lay smaller clutches (Lyon and Montgomerie 1986, Sæther 1990, King et al. 2001). Similarly, the number of fledglings and fledging success is normally lower in nests of immatureplumaged males (Grant 1990, Sæther 1990, Weggier 2001, but see Flood 1984, King et al. 2001).
We did not record the pattern of males' settlement, but nest boxes were abundant and available from the end of September. If mature-plumaged males settled before immature-plumaged males, their females did not begin to lay earlier in the season, a behavior that should be advantageous in a multiple-brooded species like the Saffron Finch (this study). In addition, females mated to immature-plumaged males laid the same number of clutches in a season as those mated to mature-plumaged males, and their clutch size, egg size, and egg mass relative to body weight were similar, suggesting their quality in this respect was not different. Apparently, females did not invest differentially according to their mate's color in the variables analyzed; the descriptive nature of the data, however, does not allow us to reject the differential-allocation hypothesis (Burley 1988).
The power of the generalized linear mixed model, that is, the probability that the test rejects a false null hypothesis, was rather low for some variables and acceptable for others, according to the definitions of Nur et al. (1999). The desirable power of statistical tests mentioned by them, however, refers to the power required in designing studies, which is usually high (Nur et al. 1999). Indeed, the usefulness of retrospective techniques in power analysis is controversial (Colegrave and Ruxton 2003). Recent literature on the use of post-hoc statistical power states that (a) computing the observed power after observing the P value should cause nothing in our interpretation of the P value to change and (b) higher observed power does not imply stronger evidence for a null hypothesis that is not rejected (Hoenig and Heisey 2001). Nonetheless, as we were unable to reject the null hypothesis of lack of differences in the breeding biology of females mated to immature- or matureplumaged males, the results of our study should be interpreted with caution.
We found no differences between females mated to males of different colors in either in the total number of fledglings per season or in the fledging success or quality of their nestlings, as measured through their growth and asymptotic weight. What we do not know yet is the relative contribution to feeding rates of immature- and mature-plumaged males and their respective females. Females mated to mature-plumaged males might have been willing to cooperate more in feeding their offspring (differential-allocation hypothesis; Burley 1988, Sheldon 2000), or, conversely, they might have worked less to take advantage of the good paternal care provided by their mates (good-parent hypothesis; Hoelzer 1989, Price et al. 1993, Préault et al. 2005); the opposite pattern could be expected for females mated to immature-plumaged males. Alternatively, males and females could invest irrespectively of the male's color. We focused on females' reproductive biology variables because, unless there is intraspecific brood parasitism, females are genetically related to all offspring in the nest. Young immature-plumaged males might be losing paternity to older mature-plumaged males, as reported in several species (Morton et al. 1990, Bitton et al. 2007, Shmoll et al. 2007), including some with delayed plumage maturation (Morton et al. 1990, Greene et al. 2000, Webster et al. 2008).
Early reproduction could be costly if the fecundity or the survival of adults or offspring is reduced (Stearns 1992). We did not detect reduced survival of offspring at fledging, but long-term data on the immature-plumaged males' survival and fecundity and the survival of their offspring after leaving the nest are lacking, and these costs might have gone undetected. By providing nest boxes, we have probably decreased competition among males for secondary cavities, making it easier for immature-plumaged males to settle and reducing the costs of male—male aggression. Nonetheless, it is remarkable that immature-plumaged males are fully capable of attracting mates and breeding as well as mature-plumaged males under these conditions, and we found that young birds are also able to breed in natural cavities. Our data suggest that there are enough females willing to pair to immature-plumaged males and that the reproductive biology of these females is the same as that of those mated to mature-plumaged males. Whether the former pay a cost of lower paternal investment from immature-plumaged males or whether they cuckold the immature-plumaged males at higher rates to favor mature-plumaged males is the subject of continuing research. In the meantime, the provision of nest boxes in the Saffron Finch's typical breeding habitats favors the species' reproduction and is a valuable tool for its recovery in populated areas where its numbers are diminishing.
Acknowledgments
We are indebted to G. Somoza, L. Miranda, and C. Frasch, who provided logistic support at the InTeCh-CONICET. We are grateful to F. Bulit, G. Herdman, G. Herdman, Jr., and S. Herdman for help in the field and to G. J. Fernández, C. Bosque, and an anonymous reviewer for their comments on earlier versions of the manuscript. This study was made possible by an Ubacyt X-140 grant and a CONICET PIP-5875 to Massoni.
Literature Cited
