The ability of plant communities in natural ecosystems to modify temperature has become increasingly important due to the profound impacts of global climate change, particularly in arid regions. However, previous studies have provided limited information on the long-term temperature feedback of these plant communities and the biotic drivers behind these changes. This study aimed to determine the functional traits and types of plant communities as biotic drivers of land surface temperature (LST) at the plant community scale, with a focus on identifying co-functioning communities in the Sirjan region of Kerman Province. To achieve this, we utilized the MODIS-LST 8-day composite product at the plant community scale and measured functional traits of dominant species through field operations. The results revealed that leaf dry matter content (LDMC), maximum height (MH), and leaf width (LW) traits significantly reduce LST. Additionally, cluster analysis indicated that the plant communities in the study area can be classified into five functional groups, which fall into two co-function categories. The S-strategized co-function (e.g., 26 communities), characterized by high LDMC values and a combination of abrupt and trend feedback in LST, was found to be more effective than the R-strategized co-function (e.g., 13 communities), which exhibited only trend feedback. Therefore, it can be argued that extreme temperatures, as a global concern, can be mitigated through careful selection of vegetation based on functional traits and strategies. This approach, particularly through rangeland improvement practices using species such as Astragalus spachianus, Cornulaca monacantha, and Launaea acanthodes, could play a significant role in addressing this challenge.