T he majority of overland movements by duck broods occur in the first week after hatching, as they travel between nesting and brood-rearing areas.1,2In this study, we use newly hatched broods falling into a canal when making overland movements to compare the breeding phenology of four duck species in Mediterranean Spain: Marbled Teal Marmaron- etta angustirostris, Mallard Anas platyrhynchos, Red-crested Pochard Netta rufina and Pochard Aythya ferina. This is one of very few studies of
hatching. However, a small number of older *Correspondence author. Email: firstname.lastname@example.org
broods were included and some loss of duck- lings, or mixing of broods, may have occurred prior to rescue.
Marbled Teal broods observed in 1995 were ascribed to three duckling size-classes: ‘small’ (less than half adult size, 1—15 days old);
‘medium’ (at least half but less than three- quarters adult size, 16—27 days); ‘large’ (at least three-quarters adult size, 28—49 days).7
Estimated hatching date was backdated from duckling size, using the following approximate ages for size-classes: ‘small’ as four days,
‘medium’ as 22 days, ‘large’ as 39 days. Marbled Teal broods seen in the field (i.e. not in the canal) in 1995 during regular surveys were ascribed to the same size-classes, and age was estimated in the same way except that ‘small’ broods were estimated as eight days old, owing to less skew towards newly hatched broods amongst field observations.
Differences between species in observation date of broods in the canal were compared with analysis of variance (ANOVA) and t tests, using transformed data to remove differences in heterogeneity between groups. The timing of observations in 1995 differed significantly between species (Table 1; Fig. 1), with Red- crested Pochard having the earliest and Marbled Teal the latest mean date. Tukey HSD tests revealed significant differences in date of observation between Red-crested Pochard and both Pochard (P < 0.04), and Marbled Teal (P < 0.002). In 1994, Marbled Teal broods (n = 9, range 2 Jun—11 July, mean ± sd = 76.2 ± 16.8,
1 = 1 April) were found in the canal signifi- cantly later than Pochard broods (n = 19, range
22 May—15 June, mean = 62.6 ± 7.1; two-tailed, two-sample t test on log-transformed data, t26=
MarbledTeal151—188.3±5.214May—15Jul71.6±19.9380178711.8793 FM, mean female body mass (g); EMexp(%), mean egg mass observed as a percentage of egg mass expected (e) from the equation e = 0.47 b0.72where b is female body mass (equation calculated for all Anatidae11); CS, mean clutch size; FM (%), total clutch mass as a percentage of FM. Source of data for FM, egg mass and CS is Cramp
& Simmons16unless otherwise indicated. Means are given ± sd. a1 = 1 April. bExcludes clutches in the UK which are larger (probably because of domestic genes in the population).16
Apr-II May-I May-II Jun-I Jun-II Jul-I Jul-II Half-monthly periods
Figure 1. Breeding phenology of Marbled Teal, Mallard, Red-crested Pochard and Pochard in Alicante, 1995. Marbled Teal II refers to estimated hatching dates backdated from duckling size for broods observed in the canal (n = 13) or elsewhere (n
= 20). Other data are for observations of broods in the canal (see Table 1 for sample sizes). Differences in date between species were highly significant (one- way ANOVA on square-root-transformed data, F3,78=
6.47, P < 0.001).
Hatching date estimated from Marbled Teal broods ascribed to size-classes in 1995 (n = 33) had a similar distribution to observation dates of Marbled Teal broods in the canal (Fig. 1), although the average date was slightly earlier (range 10 May—11 July, mean 64.9 ± 18.0). Hence the timing of observations of broods in the canal in Alicante closely reflects the timing of emergence from the nest.
Marbled Teal broods were considerably larg- er than those of other species, as expected given recorded differences in clutch size (Table 1). Large broods of more than 14 ducklings observed in Marbled Teal, Red-crested Pochard and Pochard were probably a consequence of the nest parasitism common in all three species.7,8However, post-hatch brood amalga- mation9also occurs in these species, as illustrated by the field observation of a brood of
17 small and one large Marbled Teal ducklings in El Hondo on 17 June 1995.
A similar pattern for the timing of breeding of these four species has been found in the Marismas del Guadalquivir, 500 km southwest of El Hondo, the one difference being that Mallard nested earlier than Red-crested Pochard in the Marismas.3,7The differences observed between species are compatible with several hypotheses to explain the timing of nesting initiation in ducks,10such as the ‘laying- delay hypothesis’ that nesting may be delayed by the time required by females to pay the energetic costs of egg production. Mallard and Red-crested Pochard breed early and produce relatively small eggs and relatively light
clutches (Table 1). Pochard produce relatively large eggs and have an intermediate clutch mass (Table 1). The Marbled Teal breeds last, and its relative clutch mass is much higher, whereas its relative egg size is intermediate (Table 1). Marbled Teal invest more resources in egg production than the majority of Anatidae.7
According to Rohwer,11only four of 151 Anatid species produce a heavier clutch as a percent- age of female mass.
There is an inverse correlation across species between female mass and date of hatching (Table 1), which may also indicate a causal relationship. Larger ducks may have a greater tendency to store fat reserves prior to their arrival on breeding grounds, thus allowing earlier nesting.12In other duck communities, there is a trend for larger species to nest sooner.13—15The observed differences between species in phenology could also be explained if each species timed brood emergence to coincide with peak abundance of different food items, or responded to differing functions of changing predation risk over time.10
C. AgustI, J.C. Aranda, G. Ballesteros, M. Campderros, J.L. EchevarrIa, J. Falco, M. Ferrández, L. Fidel, J. Huertas, E. MartInez, N. Ramon and A.J. Ramos helped to collect data. J.A. Amat, B. Ebbinge, A.D. Fox, L. Hillstrom, B. Hughes, C.W. Jeske, P. Nummi, H. Poysa, R. Pettifor and M.C. Woodin commented on drafts of this article. WWT Slimbridge provided data on captive duckling growth rates (collected by A. Chisholm, J. Hunter and S. Leigh-Hunt). The Sociedad Ornitologica Marmaronetta was fund- ed by the ConsellerIa de Medi Ambient, Generalitat Valenciana. We are also grateful to Ambiental S.L.
1. Rotella, J.J. & Ratti, J.T. (1992) Mallard brood movements and wetland selection in southwest- ern Manitoba. J. Wildl. Manage., 56, 508—515.
2. Poysa, H. & Virtanen, J. (1994) Habitat selection and survival of Common Goldeneye (Bucephala clangula) broods — preliminary results. Hydrobiolo- gia, 279/280, 289—296.
3. Amat, J.A. (1982) The nesting biology of ducks in