One of the best-known biogeographic rules for ectotherms is the temperature-size rule, which asserts that ectotherms produce smaller adults at warmer temperatures. Although this is often true, it has become clear that there is no single process behind the pattern and many exceptions to the rule. To disentangle such complex temperature-size relationships, individual clades must be examined at ecological and evolutionary scales.

We examined temperature-size relationships for 2106 individuals from 64 populations and 40 species of Cephaloleia rolled-leaf beetles (Chrysomelidae; Cassidinae) occurring along two tropical elevational gradientss: Barva and the Talamanca Cordillera in Costa Rica, Central America. We tested whether the temperature-size rule applied to interspecific elevational assemblages, intraspecific elevational populations, or different rearing temperatures for individual populations.

At the interspecific scale, evolutionary history, rather than elevation, explains body size. At the intraspecific scale, only one of seven species followed the temperature-size rule across elevations. When larvae were reared at different temperatures, only one of five populations followed the temperature-size rule. Most populations grew to a fixed size regardless of temperature.

Size in Cephaloleia beetles is constrained by their evolutionary history and responds to factors that rarely correlate with temperature. As temperature increases, ectotherms will not universally shrink, but determining if and why their size will change will require further investigation.

Here we provide the following datasets used to determine effects of temperature on insect body size at population and community levels: Supplement S1. Body length for 2106 individuals representing 64 populations from 40 species of Cephaloleia rolled-leaf beetle along the Barva and Talamanca elevational gradients, Costa Rica. Supplement S2. Intraspecific differences in length for 798 individuals from seven Cephaloleia species present at multiple life zone. Supplement S3. Effects of developmental temperatures on Cephaloleia adult size in 968 individuals from four species and five populations reared in the laboratory at temperatures between 10 and 35 °C.

Record of Cephaloleia distributions and size

We recorded elevational distributions of Cephaloleia species from 2005 to 2016. Beetles were collected on the Barva transect (eastern slope of the Central Cordillera 10° 0′–10° 26′ N, 83° 59′–84° 07′ W) and the western slope of the Talamanca Cordillera (8° 38′–8° 58′ N, 82° 50′–83° 05′ W). On the Barva transect, beetles were collected at La Selva Biological Station (50 m a.s.l.) and transect shelters in Braulio Carrillo National Park (700, 1070, 1500 and 2100 m a.s.l.). Collections on Talamanca were made between Ciudad Neily and Fila de Cal (60-400 m a.s.l.), as well as at Las Cruces Biological Station (1200 m a.s.l.) and in La Amistad National Park (1500-2100 m a.s.l.). Beetles from these sites were assigned to three wet forest life zones based on elevation: lowlands (0-499 m), premontane (500-1499 m), and montane (1500-2500 m). Cephaloleia beetles were identified using a combination of traditional taxonomy and DNA barcoding. The specimens were measured individually using Moticam software (Motic, Xiamen, China) and a calibrated Fisher digital dissecting microscope.

Experiments testing the effects of temperature on body size

We performed these experiments at La Selva Biological Station, Costa Rica from September 2017 to November 2018. To test if temperature can directly influence adult size, we reared in the laboratory at multiple temperatures individuals of C. belti from lowland forests and C. belti from montane forests, as well as C. aff. dilaticollis, C. aff. dorsalis, and C. placida from lowland forests. For each population, wild males and females were collected, paired, fed ad libitum with young leaf tissue from each species’ most commonly used host plant, and given squares of mature leaf tissue for oviposition substrate. Every 48 h, eggs were removed from the leaf tissue, individually placed in containers lined with filter paper, and distributed among environmental chambers. At the beginning of the experiment, eggs were assigned to six temperatures: 10°C, 15°C, 20°C, 25°C, 30°C, and 35°C. These temperatures range from the minimum temperature experienced in the montane forests to the predicted lowland maximum temperature in 2100.

We measured adults the day after eclosion by photographing them with a digital camera (Diagnostic Instruments Inc. Model 3.2.0, Sterling Heights, MI, USA) attached to a stereoscope (Leica MZ 12s). We then estimated adult lengths from the photographs to the nearest 10^-2 mm using the program SPOT V.3.5.8 (Diagnostic Instruments, Inc.).