Technical note: Rapid image-based field methods improve the quantification of termite mound structures and greenhouse-gas fluxes
Abstract. Termite mounds (TMs) mediate biogeochemical processes with global relevance, such as turnover of the important greenhouse gas methane (CH4). However, the complex internal and external morphology of TMs impede an accurate quantitative description. Here we present two novel field methods, photogrammetry (PG) and cross-sectional image analysis, to quantify TM external and internal mound structure of 29 TMs of three termite species. Photogrammetry was used to measure epigeal volume (VE), surface area (AE) and mound basal area (AB) by reconstructing 3-D models from digital photographs, and compared against a water-displacement method and the conventional approach of approximating TMs by simple geometric shapes. To describe TM internal structure, we introduce TM macro- and micro-porosity (θM and θμ), the volume fractions of macroscopic chambers, and microscopic pores in the wall material, respectively. Macro-porosity was estimated using image analysis of single TM cross sections, and compared against full X-ray computer tomography (CT) scans of 17 TMs. For these TMs we present complete pore fractions to assess species-specific differences in internal structure. The PG method yielded VE nearly identical to a water-displacement method, while approximation of TMs by simple geometric shapes led to errors of 4–200 %. Likewise, using PG substantially improved the accuracy of CH4 emission estimates by 10–50 %. Comprehensive CT scanning revealed that investigated TMs have species-specific ranges of θM and θμ, but similar total porosity. Image analysis of single TM cross sections produced good estimates of θM for species with thick walls and evenly distributed chambers. The new image-based methods allow rapid and accurate quantitative characterisation of TMs to answer ecological, physiological and biogeochemical questions. The PG method should be applied when measuring greenhouse-gas emissions from TMs to avoid large errors from inadequate shape approximations.