Ian Hartley, Louise Rowe and David Leech, Biological Sciences, Lancaster University, Lancaster, UK.

Field methods. We used a population of blue tits breeding in nestboxes near Lancaster, UK. When chicks were two days old, we carried out a cross-fostering experiment by swapping nestlings between nests of a similar stage. The swaps incorporated a brood size manipulation so that broods were either increased or decreased by two chicks, or stayed the same. We cross fostered nests in groups of three or five so that all nestlings were raised by foster parents. Chicks were marked with indelible marker pens until they were large enough to be fitted with a numbered ring. Every two days, up to day 14, we measured the chicks' mass and head-bill length and collected faecal samples whenever possible. Chicks were ringed at 6 days and a blood sample taken under Home Office licence.

Testosterone analysis. Testosterone (T) titres from the air dried faecal samples (pg per gram of dry faeces) were measured using kits (Amersham) which measured T using tritiated dihydro-testosterone and a testosterone binding antibody. Other studies have shown that faecal T measurements correlate well with blood plasma T in a range of different bird species (Bishop & Hall 1991, Cockrem & Rounce 1994, Schwabl 1996a). The short term fluctuations in T found in blood samples make them hard to interpret as measurements have no repeatability within individuals, but short term fluctuations are buffered over time in measurements from faeces. Our samples were not hydrolysed so some T might have been undetected in a conjugated form. Developing a hydrolysis method is a major task (K. Hirschenhauser pers. comm.) and was outside the scope of this project. However, 'free' and conjugated testosterone are strongly, positively correlated, so measurements of unextracted faeces provide a good indication of the total T levels in individual chicks (Schwabl 1996a). We carried out a test for repeatability of our faecal T measures for chicks which had provided more than one faecal sample. Faecal T could be measured repeatably for individual chicks.

Growth analysis. The measurements for each nestling were fitted to a logistic growth model (Hartley et al, 2000), using the SPSS statistics package. For each chick, we could estimate the growth rate constant (K), asymptote (final size) and predicted measures from the fitted curve for the ages of 2, 7 and 14 days old.

Paternity and nestling sex. The blood samples were analysed at the NERC Molecular Genetics Facility at Sheffield University. Using established methods, we used microsatellite analysis and a sex marker to identify the sex and parentage of 449 individual chicks (Hartley et al 1999, Leech et al in press)

General. During the 2000 breeding season, we were able to cross-foster and trace the development, parentage and sexes of 54 broods of blue tits. Sample size was constrained by brood availability, weather and nest synchrony as to which nests were available for the experiment.

1. Associations between T and offspring development. We were able to measure T from 173 nestlings from 50 different broods (Mean titre = 786.4 pg/g faeces, SD = 714.1). The assay kits allow for the measurement of T as low as 2 pg/g. Not all nestlings provided faecal samples during handling and the titres showed that T was present in very low concentrations, nevertheless our data demonstrate several important points for the first time. Faecal T could be measured repeatably for individual chicks (P<0.00001). There was no difference between male and female chicks for their T titres, neither was there a significant difference between chicks fathered through extra-pair copulations versus those fathered through within-pair matings. There were, however, only 8 individuals which were fathered through extra-pair matings and for which we also had T titres, so this non-significant effect should be treated with some caution. One hypothesis suggested in the grant proposal was that T titres might differ between offspring of different potential reproductive values, but this appears not to be the case with respect to sex or paternity, although T titres do not show significant relationships with mass or body size at fledging. T was generally measured from faecal samples collected between days 6 to 12 after hatching as small chicks often failed to produce samples when handled. In this analysis, therefore, we make the assumption that T measured during this time reflects the T titres of the chicks throughout their growth period. This may not be a realistic assumption but more longitudinal data is needed. From the analyses it appears that T titres are mostly positively related to body sizes during the first half of chick growth but become less important during the second half (after day 7).

2. Potential reproductive value in relation to offspring development. Offspring which are potentially more valuable to parents may fledge at heavier weights or a larger size if they have been given preferential parental investment during rearing, or they have the genetic propensity for faster/larger development. One such group of chicks which might be those fathered through extra-pair copulations, although different fledging biometrics are often difficult to separate from parental preferential treatment. In this study we cross-fostered all nestlings, so were able to compare the biometrics of extra-pair (EP) chicks (i.e. those fathered through extra-pair copulations) versus within-pair (WP) chicks raised in an environment removed from any direct potential parental preference. Analyses compares WP and EP offspring, when both types of chick have a common nest of origin and rearing nest. As in our previous study (Leech et al in press), we found no relationship between paternity type and offspring sex (this data: Chi² test, p = 0.52) but our data show that blue tits become sexually size dimorphic from at the latest day 7, so we also controlled for offspring sex in the analyses. We found that skeletal size was generally larger in EP chicks compared to WP chicks, but that masses did not differ significantly. Like the T titre effects, the relationships were strongest when the chicks were in the first week after hatching, when any maternally derived effects would be expected to be greatest.

3. Contribution of parental and environmental effects on offspring growth and testosterone titres. Our cross-fostering experiment allows us to test for variation in chick traits which is due to common origin (i.e. genetic or maternal effects) and which is due to the rearing (foster) environment. Chicks fathered through extra-pair copulation were excluded from the analyses. The results clearly show that both the origin and rearing environment of nestlings contributed significant variance to their biometrics and T titre levels. Note that the chicks were swapped between nests on day 2, so we should expect no significant variation due to foster box at this age, and that is exactly what we found for weight and skeletal size but not T titre. This was probably due to the fact that T was mostly measured from faecal samples taken after day 6. Importantly though, T has a highly significant genetic/maternal effect component.

4. Sex differences in response to variation in environmental quality. There is sexual size dimorphism in blue tits so we expected, and generally found, a significant effect of sex on growth. The data show that there were no significant interactions between sex and manipulation for their relationship between growth measures, suggesting that male and female blue tit chicks react similarly to environmental stress. There were also no significant effects of the manipulation, suggesting that the blue tit growth patterns were either not plastic in response to environmental variation or the manipulations did not stress the chicks enough to show an effect. Other studies have found that a brood manipulation change of two chicks is enough to elicit a response to the change (e.g. Kunz & Ekman 2000) but further experiments, with greater brood manipulations would help to answer the question.

This was a pilot study to investigate the effects of nestling T on growth and identify the sources of the variation in growth patterns and T in blue tit chicks. Our data clearly show that variances in growth and T titres in nestlings have both a genetic-parent component (either a genetic or maternal effect) and a rearing environment component. Additionally, we have shown that T is significantly related to body size in the early stages of chick development, so it could potentially be used by females to influence the growth and competitive ability of nestlings within broods and could, therefore, provide a mechanism for parental control of sibling rivalry. Further work is needed to look at the effects of chick T and size on begging activity and success within the nest, using nest cameras to monitor behaviour. A grant application has already been submitted to develop an experimental system using captive zebra finches to independently manipulate female, egg and chick testosterone levels, and quantify the effects on chick fitness. We have also shown that male and female chicks develop in similar ways, although to different final sizes, even across an environmental quality gradient. A lack of effect of our brood size manipulations on chick growth is surprising given previous studies which have shown a plastic response to growth patterns (Gebhard-henrich & van Noordwijk 1994, Kunz & Ekman 2000), whether this result is population specific or due to the effects of the study year, it warrants further investigation.

Back