Impacts of a Global Climate Cycle on Population Dynamics of a Migratory Songbird

See allHide authors and affiliations

Science  16 Jun 2000:
Vol. 288, Issue 5473, pp. 2040-2042
DOI: 10.1126/science.288.5473.2040


Progress toward understanding factors that limit abundances of migratory birds, including climate change, has been difficult because these species move between diverse locations, often on different continents. For black-throated blue warblers (Dendroica caerulescens), demographic rates in both tropical winter quarters and north temperate breeding grounds varied with fluctuations in the El Niño Southern Oscillation. Adult survival and fecundity were lower in El Niño years and higher in La Niña years. Fecundity, in turn, was positively correlated with subsequent recruitment of new individuals into winter and breeding populations. These findings demonstrate that migratory birds can be affected by shifts in global climate patterns and emphasize the need to know how events throughout the annual cycle interact to determine population size.

The need to understand when and how bird populations are limited is made pressing by recent declines in the abundances of many species, especially migratory songbirds (1). Quantifying the effect and timing of limiting factors for migratory species, however, is difficult because the birds spend different parts of their annual cycle in different locations. Furthermore, events during one stage of the annual cycle are likely to influence populations in subsequent stages (2). Here we show through long-term demographic studies of a migratory songbird that the El Niño Southern Oscillation (ENSO) impacts demographic rates in both the breeding and nonbreeding seasons. Our findings also reveal links in the population dynamics of this species between stages of its annual cycle.

We measured the effect of ENSO on survival, fecundity, and recruitment of the black-throated blue warbler, a migratory songbird that breeds in forested regions of eastern North America and overwinters primarily in the Greater Antilles. This species is territorial, largely insectivorous, and exhibits strong site fidelity in both its breeding and wintering grounds (3). We quantified warbler demography from 1986 to 1998 at two locations during the annual cycle: the overwinter period at Copse Mountain, near Bethel Town in northwestern Jamaica, West Indies, and the breeding season at Hubbard Brook Experimental Forest, West Thornton, New Hampshire, USA. The species' habitat at both sites was mature mesic forest, relatively undisturbed by human activity. Demographic data were collected annually from the overwintering population in late October and from the breeding population in mid-May through August. For all analyses, we used annual mean monthly values of the standardized Southern Oscillation Index (SOI) to represent ENSO conditions for each calendar year (4). High, positive values of SOI indicate La Niña conditions and low, negative values indicate El Niño conditions (5).

Annual survival (6) of black-throated blue warblers in Jamaica (Fig. 1, A to B) was strongly associated with SOI: survival was low in El Niño years and high in La Niña years (Fig. 2). This result is best explained by the impact of ENSO on local climate and a concomitant change in food availability for overwintering birds. In the winter dry season, migrant songbirds are often food-limited (7) and can be in a state of physiological decline (8), especially in late winter before spring migration. During El Niño years in Jamaica, reduced rainfall (9) probably leads to a decreased amount of food available for warblers in the winter dry season and, hence, to lower survival. La Niña years, in contrast, tend to be wetter and thus would result in increased food availability and higher survival.

Figure 1

Chronology of major demographic events in relation to a given ENSO year. The x axis delimits a 24-month period over two concurrent calendar years. “Year 1” is the ENSO year being considered. ENSO phase (open bar) is represented by the mean of Year 1's monthly values of the standardized SOI. Closed bars symbolize overwinter (black) and breeding (gray) periods for warbler populations; thick portions show when birds were present at the respective locations. Letters along the black and gray bars indicate timing of warbler capture and recapture used in survivorship analyses. For example, A to B represents annual (October–October) survivorship in Jamaica. Arrows linking the breeding and overwintering populations signify warbler migration and recruitment (1, autumn migration; 2, spring migration).

Figure 2

Comparison of annual survival estimates (31) for black-throated blue warblers in Jamaica (black line) and New Hampshire (dashed line) to mean monthly values of SOI (gray line). X axis represents capture year for survival curves and ENSO calendar year for SOI curve. Left yaxis gives estimated warbler survival to the next calendar year Righty axis indicates mean monthly values of SOI. Based on capture-mark-recapture analyses (31), annual survival probability in Jamaica from yeari to yeari+1was a linear function of mean monthly SOI in yeari; annual survival probability in New Hampshire was constant among years. Recapture probability was constant among years in both locations (0.95 in Jamaica; 0.86 in New Hampshire).

Annual survival of warblers breeding in New Hampshire (Fig. 1, C to D) was relatively constant and did not consistently fluctuate with changes in ENSO (Fig. 2). Breeding populations of black-throated blue warblers mix extensively on their Caribbean winter quarters (10), and these islands vary in the extent to which they are affected by ENSO (9, 11). Thus, the lack of association between annual survival in New Hampshire and SOI is probably due to many individuals in the breeding population overwintering on other islands, particularly Cuba (12), where the climatic effects of ENSO can be less severe compared with Jamaica (9, 13).

ENSO had significant effects on warbler fecundity in New Hampshire. We measured two components of fecundity (14): (i) number of offspring fledged and (ii) mass at fledging, a general predictor of offspring survival (15). Mean number of young fledged per warbler pair was weakly correlated (16) with SOI (r = 0.39,P = 0.18). However, mass at fledging was highly correlated with SOI: fledglings weighed less in El Niño years relative to La Niña years (Fig. 3A). After accounting for variation in fledgling mass and hence its expected effect on survival (17), annual fecundity was correlated with SOI (Fig. 3B).

Figure 3

Reproductive parameters of black-throated blue warblers were correlated with ENSO phase. X axis, mean monthly SOI. Numbers by data points indicate year; data were lacking for 1987. (A) Mean mass of warbler chicks at fledging, a predictor of offspring survival (15). (B) Warbler fecundity, i.e., mean number of offspring fledged per warbler pair per year after controlling for nestling mass (17).

Because black-throated blue warbler fecundity is limited by food availability (18), we tested if variation in availability of lepidopteran larvae, the warbler's primary prey in summer, was related to ENSO. Food is most limited from mid-June to mid-July when adults are feeding nestlings and dependent juveniles. At the New Hampshire site, total larval biomass recorded annually on mid-June through mid-July censuses (19) was positively correlated with annual fecundity (r = 0.58, P < 0.05) and with SOI (r = 0.58, P < 0.04). Prey biomass was low during El Niño years and high during La Niña years. These results suggest that ENSO influences fecundity of black-throated blue warblers by affecting their food supply.

The effects of ENSO on warbler fecundity had consequences for demography in subsequent seasons (20). First, the number of juveniles at the Jamaica site each October (i.e., recruitment to the overwintering population; see arrow “1” in Fig. 1) was positively correlated with SOI (r = 0.71, P < 0.007) and with fecundity from the preceding summer in New Hampshire (Fig. 4A). Second, the number of yearling breeders each May in New Hampshire (i.e., recruitment of the preceding year's fledglings into the breeding population; see arrow “2” inFig. 1) was positively correlated with both warbler fecundity (Fig. 4B) and mean monthly SOI (r = 0.59, P < 0.04) from the previous year. In both Jamaica and New Hampshire, low annual recruitment of both juveniles and yearlings was associated with El Niño conditions, whereas high recruitment was associated with La Niña events.

Figure 4

Recruitment into winter and breeding populations was correlated with warbler fecundity. X axis, mean warbler fecundity in yeari (17). Only males were included in these analyses due to uncertainty in determining the age of female black-throated blue warblers. Numbers by data points indicate yeari; data were lacking for 1987. (A) Number of juvenile males at the Jamaica site each October in yeari. (B) Number of yearling males at the New Hampshire site each May in yeari+1.

ENSO has been shown to impact demographic rates and food resources of many animal taxa, including seabirds (21), raptors (22), Pacific island passerines (23), primates and rodents (24), and arthropods (25). The results presented here provide evidence that ENSO, through its effect on food supply, limits survival, fecundity, and recruitment of a migratory songbird. Additionally, they illustrate an important interaction between summer and winter population dynamics that operates through a common link to the ENSO cycle. Populations of migratory birds are therefore susceptible over a range of spatial and temporal scales to shifts in global climate patterns.

Evidence is accumulating that bird populations are being affected by global warming associated with long-term climate change (26). Global warming could also be increasing the severity of ENSO events (27). If this is true, we predict that variance in demographic rates of migratory bird populations will become amplified, leading to elevated extinction risk, especially for small populations (28). Because many migratory bird species are declining in abundance and therefore of conservation concern, field research and demographic modeling efforts should focus on understanding how events throughout the annual cycle are interconnected and on how multiple limiting factors, both natural and human-related, determine population size.

  • * To whom correspondence should be addressed. E-mail: t.scott.sillett{at}


View Abstract

Navigate This Article