Reproduction and mortality must be balanced in a population whose size remains constant. But why does this balance vary from species to species, from population to population, and from location to location? What modifies it? And what causes the general shift in this balance observed with latitude or altitude? The most famous shift is the increase in clutch size with increasing latitude, illustrated by comparisons across and within species (Martin et al. 2000; Cardillo 2002; Tieleman et al. 2006). This variation in clutch size already puzzled early animal ecologists in the 1940s: David Lack, Alexander Skutch and Reginald Moreau each explained the phenomenon emphasizing different ecological factors like food availability or nest predation (Ricklefs 2000). The focus on a single life-history attribute such as clutch size left out of consideration many of the other factors that affect the balance between reproduction and survival. Comprehensive studies – reporting all life-history attributes – are scant, partly because measuring such variables as survival and recruitment rates, number of (re)nesting attempts, and offspring quality is difficult. In addition, understanding variation in life-history attributes among populations is an inherently intricate problem. Multiple correlations between life-history traits and independent variables complicate comparisons and interfere with the design of experiments that control all but a single variable (Ricklefs 2000). Bears et al. (this issue) have bravely and successfully tackled the first obstacle by providing a comprehensive overview of the life histories of two groups of dark-eyed juncos, living at high (2000 m) and low (1000 m) elevation.
Dark-eyed juncos are small passerine birds that breed in most of North America, from Alaska to the southern USA, and depending on the population, migrate south for the winter; their wintering range extends from southern Canada to northern Mexico and Florida. Adults feed on a mixture of arthropods and seeds during the summer, and switch predominantly to seeds and berries in winter, whereas offspring are provided solely with arthropods. Bears et al. (this issue) studied the life histories of a migratory population of juncos on the slopes of the Canadian Rocky Mountains in four sites at each elevation, i.e. 1000 m and 2000 m above sea level, during the years 2000–05. At high elevation, the reproductive season started later and was more compressed, spanning less than half the time used by low-elevation birds. High-elevation birds had the same brood size (2·7 vs. 2·9 fledglings per brood) but about 50% fewer broods per season (0·75 vs. 1·55) compared with low-elevation birds. High-elevation fledglings were heavier, carried more fat and survived better during the first 25 days of their life (survival 94 vs. 80% for low-elevation fledglings). Annual apparent survival of adult males was also higher by 15–20% at high elevation (for females the sample size was insufficient to make reliable estimates). Taken together, the life-history strategy of high-elevation juncos showed a shift towards a ‘slower pace’: relatively smaller investments in reproduction and larger investments in survival. The authors discount the possible explanation that birds might be differentially distributed over the mountain slopes as a result of competition that might displace inferior-quality birds to breed at higher (or lower) elevation. In contrast, they argue that the juncos show high fidelity to both breeding site and natal site.
Putting the findings of Bears et al. (this issue) in the traditional framework that explains life-history variation along a latitudinal gradient points out a number of inconsistencies with current thinking about proximate and ultimate factors that cause variation in life histories. These inconsistencies provide exciting directions for follow-up studies to advance our understanding at the evolutionary, ecological and physiological levels; the junco study system would be well suited to address these. One puzzling issue is the connection between brood size, parental workload and adult survival. When compared with tropical birds, higher-latitude birds with larger clutch sizes are thought to have a higher metabolic rate, both basal metabolic rate and field metabolic rate while raising nestlings, and a lower survival rate (Ricklefs & Wikelski 2002; Tieleman, Williams & Visser 2004; Wiersma et al. 2007). The proposed proximate mechanism to connect these traits is that higher energy expenditure results in damage to the body, expediting mortality (Drent & Daan 1980; Daan, Deerenberg & Dijkstra 1996; Speakman et al. 2002; Tieleman et al. 2008). The work level considered is normally the per-day level required to raise the number of nestlings in the nest (Drent & Daan 1980; Daan et al. 1996), not the duration of the breeding season, i.e. the number of broods raised (but see Martin 1995). For the junco system, it would be now interesting to measure parental energy expenditure during the nestling period, and cellular and molecular damage to the parental bodies at the end of the breeding season. The prediction would be that field metabolic rates are the same between elevations because brood sizes are too – if not higher at the higher elevations because of the lower temperatures – but that body condition at the end of the season is worse in the low-elevation birds because their effort lasts longer.
A second puzzling phenomenon relates to the micro-evolutionary background of the study system and the potential selection pressures driving the combination of life-history traits. Juncos occur on the entire mountain slope, and one would imagine that genes freely flow up and down between 1000 m and 2000 m above sea level, through conspecifics living at intermediate altitudes. Yet, Bears et al. have brought birds from both elevations to a common garden set up in captivity, where some morphological and behavioural differences are maintained (Bears, Drever & Martin 2008; Bears et al., this issue). Although maternal effects, developmental effects or other carry-over effects from the field cannot presently be excluded, the authors hint at genetic differences between low- and high-elevation juncos. Confirmation of such micro-evolutionary patterns with translocation experiments or a cross-breeding programme in captivity would raise the question what the selection pressures are that shaped and maintain the variation in life-history strategies between elevations.
From the latitudinal variation in life-history strategies, three factors jump out as potentially important: nest predation, food availability, and adult mortality risk. Nest predation is often invoked to explain clutch size (Skutch 1949; Ghalambor & Martin 2001), which does not apply to the juncos, whose brood sizes are the same at both elevations. Likewise, variation in food availability does not apply as commonly used: to explain differences in brood size (Lack 1947; Martin 1987; Martin 1995). Consistent with ideas about food availability and brood size, however, the similarity of brood sizes at low and high elevation combine neatly with the almost identical heights of the arthropod food peak at both elevations, as documented by Bears et al. (this issue) during a single year. Yet, more interesting is the indirect selective potential of the duration of the food peak: the short duration at high elevation is likely connected with the brief breeding season, and could only indirectly select for higher annual survival if the trade-offs between hard work and survival also apply to the duration of the hard work period and not just the level of it as suggested above. The third factor put forward to explain life-history variation with latitude is the risk of adult mortality. The tropics and other low-latitude environments are thought to provide a relatively stable year-round environment, lacking seasons with particularly high risks of mortality, such as the winter at higher latitudes (Ashmole 1963; Martin 2004; Tieleman 2007). Environments with more seasonal variation are thought to select for fast paces of life with high reproductive output because mortality is high in such environments. The junco results suggest the opposite: the high-elevation environment is most likely characterized by larger annual fluctuations in temperature, primary productivity and hence food availability, yet the high-elevation juncos have a higher annual survival rate. It is not difficult to imagine why the annual seasonality explanation may not apply to the juncos: they migrate away from their breeding areas. Are the annual cycles outside the breeding season similar for low- and high-elevation juncos? Understanding the factors that lead to a higher survival rate in high-elevation populations will require a year-round view.
The variation in life-history strategies of dark-eyed juncos from high- and low-elevation populations, fitting with earlier results along altitudinal gradients, cannot be explained by the same processes applied to life-history variation along latitudinal gradients. The emerging challenge is how to understand the similarity in life-history strategies of high-altitude and low-latitude birds. Or do we perhaps miss the boat completely, trying to find a general rule to explain identical strategies in such different ecological settings?