Temperature and rainfall track pollination specialization more closely than latitude does

The tropical expectation had a sensible basis. Tropical communities hold more interacting plant and animal species, so many ecologists assumed specialist pairings would be commoner there. The idea had gone largely untested because suitable data did not exist ^[8].

Lead author Sailee Sakhalkar, who recently completed her doctorate at Charles University, described what a specialist gives up ^[15]. "Specialization can be understood as an evolutionary bargain. A close match can make feeding and pollination very efficient, but it also increases dependence. If one partner disappears or becomes active at a different time, a specialist has fewer alternatives," she said ^[10].

The study is a synthesis. A network here is a set of links between plant and pollinator species, and the team pooled them from 162 studies, about 21 networks per study ^[2][1]. Spread across all 3,415, the links average a little over 32 per network ^[3][2]. "No single team could study pollination on this scale," Sakhalkar said ^[11].

Robert Tropek, head of the Insect Community Ecology Research Group, jointly led the work at Charles University and the Biology Centre of the Czech Academy of Sciences ^[15][1]. He argued that latitude was never the variable organisms respond to. "We tend to draw geographic patterns in pollination as lines running from the poles to the equator, but plants and pollinators do not experience latitude itself. They experience heat, cold, rain and changing seasons, each in their own way," he said ^[12].

Tropek's point is about proxies, and the climate result fits it ^[6]. It is still a correlation across sites. The thing this doesn't tell you is how wide the margin between the climate and latitude models was. The public summary does not report the specialization index, the effect sizes, or how the analysis handled differences in sampling effort among the 162 studies ^[2]. I think the descriptive result about the tropics is on firm ground, since it rests on thousands of networks ^[5]. How much temperature and rainfall add over latitude is a figure that has to come from the paper's own tables.

The split among pollinator groups is the part with consequences for climate forecasting. Rising temperatures and altered rainfall can change where species live, when plants flower and when pollinators are active ^[13]. "A bee, hoverfly, moth and hummingbird differ substantially in their bodies, behavior and ways of using flowers. Their relationships with plants may therefore respond very differently to changes in temperature and rainfall," Sakhalkar said ^[14].