Spatial variation in avian bill size is associated with temperature extremes in a major radiation of Australian passerines

Abstract

Morphology is integral to body temperature regulation. Recent advances in understanding of thermal physiology suggest a role of the avian bill in thermoregulation. To explore the adaptive significance of bill size for thermoregulation we characterized relationships between bill size and climate extremes. Most previous studies focused on climate means, ignoring frequencies of extremes, and do not reflect thermoregulatory costs experienced over shorter time scales. Using 79 species (9847 museum specimens), we explore how bill size variation is associated with temperature extremes in a large and diverse radiation of Australasian birds, Meliphagides, testing a series of predictions. Overall, across the continent, bill size variation was associated with both climate extremes and means and was most strongly associated with winter temperatures; associations at the level of climate zones differed from continent-wide associations and were complex, yet consistent with physiology and a thermoregulatory role for avian bills. Responses to high summer temperatures were nonlinear suggesting they may be difficult to detect in large-scale continental analyses using previous methodologies. We provide strong evidence that climate extremes have contributed to the evolution of bill morphology in relation to thermoregulation and show the importance of including extremes to understand fine-scale trait variation across space.

Keywords: Meliphagides; bill size; climate extremes; thermoregulation.

Figure 1.

Relationships between relative bill size and geographical variables (a) latitude, (b) elevation, (c) minimum distance to coast, and climate variables (d) days < 5°C, (e) mean minimum winter temperature, (f) mean maximum summer temperature, across the continent. Shaded area shows 95% credibility intervals around fitted values, when all other predictors (including their parameter estimates) are held constant for (a,b,c) and for (d,e,f) when posterior mean of non-focal parameters used for predictions. (Electronic supplementary material, figure S6 shows all data points.)

Figure 2.

Associations between relative bill size and frequency of extreme temperatures within climate zones. Shown are relationships between bill size and (a) days <5°C in interaction with winter rainfall (mm) at five quantiles, 5th, 25th, 50th, 75th and 95th across 50 species within the arid zone and (b) corresponding three-dimensional representation; (c) frequency of days ≥35°C in interaction with mean summer vapour pressure (hPa) at five quantiles across 57 species within the temperate zone and (d) corresponding three-dimensional representation; (e) frequency of days ≥35°C in interaction with summer rainfall (mm) at five quantiles across 57 species within the temperate zone and (f) corresponding three-dimensional representation. Red-dash line in (e) shows the amount of rainfall above which bill size shows an increase with increasing hot extremes. (Electronic supplementary material, figure S7 shows all data points.)

Figure 3.

Species variation in the bill size response to extreme temperatures. Shown are bill size associations with (a) frequency of exposure to daily minima < 5°C in arid zone birds at 10 mm (orange) and 50 mm (blue lines) of winter rainfall, (b) frequency of exposure to daily maxima ≥35°C in temperate zone birds at 50 mm (orange lines) and 100 mm (blue lines) of summer rainfall and (c) summer vapour pressure in temperate zone birds when exposure to daily maxima ≥35°C is 10 (blue lines) or 40 days (orange lines) in summer. Each line (orange or blue) presents the model-predicted mean bill size of individual species after controlling for body size, sex, year of capture and IBRA region.