Summary
Predicting ecological consequences of climate change will be improved by understanding how species are affected by contemporary climate variation, particularly if analyses involve more than single ecological variables and focus on large-scale climate phenomena. I used 18 years of data from red-winged blackbirds (Agelaius phoeniceus) studied over a 25-year period in eastern Ontario to explore chronological and climate-related patterns of reproduction. Although blackbirds started nesting earlier in years with warmer springs, associated with low winter values of the North Atlantic Oscillation Index (NAOI), there was no advance in laying dates over the study. Nesting ended progressively later and the breeding season lasted longer over the study, however, associated with higher spring values of NAOI. As the length of the nesting season increased, offspring sex ratios became more female biased, apparently as a result of females adjusting the sex of the eggs they laid, rather than from sex-biased nestling mortality. Clutch size did not vary systematically over the study or with climate. Opposing trends of declining nest success and increasing productivity of successful nests over the study resulted in no chronological change in productivity per female. Higher productivity of successful nests was associated with higher winter NAOI values, possibly because synchrony between nesting and food availability was higher in years with high NAOI values. Other than the association between the start of nesting and spring temperatures, local weather (e.g., temperature, rainfall) patterns that linked NAOI with reproduction were not identified, suggesting that weather patterns may be complex. Because climate affected most aspects of red-winged blackbird reproduction examined, focusing on associations between climate and single variables (e.g., first-egg dates) will have limited value in predicting how future climates will affect populations.
Methodology
Field methods: Data presented here come from studies of red-winged blackbirds conducted at the Queen's University Biological Station in eastern Ontario (45.36'N, 76.13'W) between 1974 and 2000. The individual studies had a number of field methods in common and were conducted using the same group of cattail marshes. Although every site was not used in every study, every study used multiple sites and each study had multiple sites in common with other studies. Data were collected for 18 years between 1974 and 2000. In most years, every nest at a site was found and its fate determined. In 1981-1983, however, eggs were collected late in incubation, precluding determination of nest success, and in 1999-2000, reduced nest searching later in the season precluded determination of last nest dates. Detailed methods for specific years are available from the following: 1974-1975 (Weatherhead and Robertson 1977), 1981 (Weatherhead 1983), 1982-1983 (Weatherhead 1985), and 1985-1995 (Weatherhead and Sommerer 2001). Methods for 1999-2000 are unpublished, but with the exception noted above, were similar to those in Weatherhead and Robertson (1977) for aspects relevant here. Unless otherwise indicated, in each analysis I include all years for which data were available.