Fungus-growing termites (subfamily Macrotermitinae) have a distinct geographical distribution, found only in the old-world tropics. Where present, they are considered to be major contributors and regulators of decomposition, with consumption rates often greater than other termite groups. This study sought to understand the relative roles of termite distribution (specifically the presence or absence of fungus-growing termites) and climatic variables (mean annual temperature, mean annual precipitation and mean annual aridity) on global patterns in deadwood decay. To answer this question, we added new salient data to an existing dataset on global wood decay by Zanne et al. (2022) available at https://doi.org/10.6084/m9.figshare.19920416.v1. We filtered the data to only include sites where termites were present and thus analysed a dataset containing 102 sites across 16 countries. We found that termite-driven decay of deadwood increased with aridity but was higher in sites with fungus-growing termites than sites without fungus-growing termites. Our results also showed that the relative role of fungus-growing termites increased with aridity, as rates of wood-discovery by termites increased with aridity but only in sites where fungus-growing termites were present. Our findings indicate that the inclusion of biogeographical differences in termite distribution could potentially alter global estimates of deadwood turnover. This repository contains new datasets on termite-driven deadwood decay of Pinus radiata wood blocks and code used for all data analyses and production of figures.

All experimental data collected on the decay of wood blocks in the datasets 'new_global_wood_decay.csv' and 'pine_shade.csv' followed a standard protocol outlined by Cheesman et al., (2018), https://doi.org/10.1111/aec.12561. Data in 'new_global_wood_decay.csv' was collected from 140 sites across 20 countries by different researchers. A complete description of how data was collected in this dataset is provided in Zanne et al., (2022), https://www.science.org/doi/10.1126/science.abo3856. All researchers followed the same method except for one difference: data from the original Zanne et al., (2022) dataset covered wood blocks with 70% green shade cloth to reduce solar radiation degradation of mesh bags while new data did not use green shade cloth.  Environmental parameters for each site were extracted from global databases, mean annual temperature (MAT) and mean annual precipitation (MAP) was extracted from WorldClim, https://doi.org/10.1002/joc.5086, and mean annual aridity (MAA) from the Global-Aridity Index by Zomer et al. (2022), https://doi.org/10.1038/s41597-022-01493-1. Data in 'pine_shade.csv' was used to analyse if the inclusion of shade cloth had any affect on decay rates of wood blocks. In the dataset 'pine_shade.csv' the decay of Pinus radiata wood blocks is measured in a rainforest site (named DRO) and savanna site (named PNW) in Queensland, Australia, following the methods described in Wijas et al., (2024), https://doi.org/10.1111/1365-2435.14494, but with the inclusion of a 70% green shade cloth. The sites in 'pine_shade.csv' are the same sites named ‘wet rainforest’ and ‘dry savanna’ in Wijas et al. (2024) and site descriptions are also provided in Clement et al., (2021), https://doi.org/10.3389/fevo.2021.657444. 

For analyses, we processed the 'new_global_wood_decay' dataset to only include sites where termites were known to be present (102 of the 140 sites). The presence of fungus-growing termites (subfamily Macrotermitinae) was assigned to sites if sites were within the geographical distribution of fungus-growing termites, i.e. if they were in Afrotropical, Oriental, or Malagasy realms. Additionally, we checked for the presence of fugus-growing termites by reviewing published termite transect surveys at the same sites and through personal communication with researchers based at sites. Termite decomposition of deadwood was considered a two step process: first we looked at termite discovery of deadwood, wood blocks were considered discovered by termites if researchers had noted imported soil on wood blocks by termites, termite related damage to wood blocks or termite presence on wood blocks; second we looked at decay rates of wood blocks discovered by termites. Decay of undiscovered wood blocks was attributed primarily to microbial decay. Decay of discovered wood blocks includes microbial decay but is refered to as termite-driven decay. We compared termite discovery of deadwood and termite-driven decay rates in sites where fungus-growing termites were present and absent. We used R software to run linear regression models to compare differences in termite discovery and termite-driven decay rates with climatic variables (MAT, MAP and MAA) and with the presence or absence of fungus-growing termites.