Cold is a major challenge to growth and reproduction for alpine life. Growing season temperatures above timberline frequently drop below freezing, posing special challenges for ectothermic organisms. We tested the impact of nest site selection on the capacity for thermoregulation in alpine populations of the ant, Formica neorufibarbus gelida. The four populations surveyed occur along an elevation gradient from sub-alpine meadow (3474 m) to high alpine tundra (3886 m) in the central Colorado Rocky Mountains (USA).  F. n. gelida nests in the soil beneath rocks, carrying larvae to the soil surface to warm during the day and moving them back to lower depths at night.  On average, ant colonies established nests under rocks with 20% greater surface area per unit volume than expected at random (P<0.041). We measured excess temperature before sunrise and at mid-day beneath nest rocks and nearby randomly chosen unoccupied rocks.    

Results/Conclusions Rock dimensions had little impact on soil warming before dawn, but strongly affected soil temperature at mid-day, with surface area increasing and volume decreasing the capacity for soil warming (P<0.0005 and P<0.015, respectively).  Moreover, the mid-day soil temperature beneath nest rocks averaged more than twice that beneath unoccupied rocks (P<0.0072).  Ants of F. n. gelida exhibit a developmental polymorphism, giving rise to workers of red and black body color.  Red ants are larger and potentially have broader diet breadth than black ants.  By choosing rocks with warm soil microclimate, colonies appear to manipulate relative production of the two color morphs. Surveys of nests at the upper reach of timberline (3530 m) showed that as soil temperature increased the proportion of red ants increased accordingly (R² = 0.80, P<0.0066, n = 7 nests).  This result suggests that non-random nest site selection represents a form of niche construction with indirect benefits for colony productivity that complement the direct effects of warming on larval viability and development rate. Taken together these advantages may explain why colonies of F. n. gelida that persist through the summer at 3530 m are located under significantly larger rocks than others that die or abandon nests (P<0.034). More generally, our results underscore the contribution of developmental plasticity to niche breadth and ecological range limits in nature.