1. Global warming will likely cause more ecological change by altering how species interact with each other than by directly affecting individual species. Field heating experiments are essential to test how warming will change species interactions. However, such experiments pose many logistical challenges, including heater construction and accuracy and accessing necessary infrastructure.
2. To facilitate these experiments, we developed portable active heaters suitable for heating microhabitats and sites of species interactions. We validated heater performance using two different target microhabitats: rolled leaf refugia and aquatic microcosms.
3. Using plastic mesh, resistive heating wire, and an Arduino UNO microcontroller system with a custom shield, we built adjustable heaters that can dynamically heat small targets to at least 5°C above ambient temperature. Leaf target systems were heated with mean absolute errors (MAE) of 0.40‐1.06°C. Water target systems were heated with mean absolute errors of 0.02‐0.04°C.
4. These heaters can be customized to accurately heat a wide range of target sites (leaves, flowers, nests, small pools of water, etc.), many of which cannot be easily heated with traditional heaters. They can be easily constructed and require less field site infrastructure than many active heaters. These heaters’ adjustability and portability mean that they can help test the effects of global warming in many situations, particularly in remote and lightly developed areas.

The data were recorded automatically by the heater control system in 1-minute intervals in a text file. We removed the file header and converted it to a spreadsheet format. For details of how the heaters were constructed and deployed, please see the Methods in Ecology and Evolution paper.

These are pilot data for dynamic portable microhabitat heaters described in the associated Methods in Ecology and Evolution paper. In this test, rolled Pleiostachya pruinosa (Heliconiaceae) leaves were heated to 5°C above the ambient temperature. This heater system operated an ambient temperature sensor and four heaters (labeled A, B, C, and D) and recorded mean temperature and power usage every minute. All temperature data are in degrees Celsius. Power usage is expressed as a percentage of maximum power. In this case, "maximum power" is 44 W. There are no missing values.