Human land use alters soil microbial composition and function in a variety of systems, although few comparable studies have been done in tropical forests and tropical agricultural production areas. Logging and the expansion of oil palm agriculture are two of the most significant drivers of tropical deforestation, and the latter is most prevalent in Southeast Asia. The aim of this study was to compare soil fungal communities from three sites in Malaysia that represent three of the most dominant land-use types in the Southeast Asia tropics: a primary forest, a regenerating forest that had been selectively logged 50 years previously, and a 25-year-old oil palm plantation. Soil cores were collected from three replicate plots at each site, and fungal communities were sequenced using the Illumina platform. Extracellular enzyme assays were assessed as a proxy for soil microbial function. We found that fungal communities were distinct across all sites, although fungal composition in the regenerating forest was more similar to the primary forest than either forest community was to the oil palm site. Ectomycorrhizal fungi, which are important associates of the dominant Dipterocarpaceae tree family in this region, were compositionally distinct across forests, but were nearly absent from oil palm soils. Extracellular enzyme assays indicated that the soil ecosystem in oil palm plantations experienced altered nutrient cycling dynamics, but there were few differences between regenerating and primary forest soils. Together, these results show that logging and the replacement of primary forest with oil palm plantations alter fungal community and function, although forests regenerating from logging had more similarities with primary forests in terms of fungal composition and nutrient cycling potential. Since oil palm agriculture is currently the mostly rapidly expanding equatorial crop and logging is pervasive across tropical ecosystems, these findings may have broad applicability.

Tropical forest conversion to agriculture is a major global change process. Understanding of the ecological consequences of this conversion are limited by poor knowledge of how soil microorganisms respond. We analyzed the response of soil bacteria to conversion from primary rain forest to oil palm plantation and regenerating logged forest in Malaysia. Bacterial diversity increased by approximately 20% with conversion to oil palm because of higher pH due to liming by plantation managers. Phylogenetic clustering indicated that bacterial communities were determined by environmental filtering. Regenerating logged forests did not have significantly different soil chemistry, which did not correspond with significant differences in bacterial richness, diversity, or the relative abundances of particular taxa. However, there were significant differences in the structure of bacterial community networks between regenerating logged forests and primary forests, highlighting previously unobserved effects of these two land uses. Network analysis highlighted taxa that are potentially central to bacterial networks, but have low relative abundances, suggesting that these rare taxa could play an ecological role and therefore warrant further research.

In this study, the following hypotheses were tested: (i) microbial communities would be distinct for each land-use type (primary forest, regenerating forest, and oil palm plantation); (ii) microbial communities from forest sites (primary forest and regenerating forest) would be more similar to each other than those found in the oil palm plantation site; (iii) microbial communities would be distinct at different soil horizon depths; and (iv) soil microbial biomass (SMB) would be greater in forest sites than in the oil palm plantation sites. To test these hypotheses, soil samples collected from an intact tropical rain forest, a tropical rain forest recovering from logging 50 years previously and an oil palm plantation in active production located in and around the Pasoh Forest Reserve in Peninsular Malaysia, were analysed. Total SMB estimated by summing PLFA compunds from each soil depth revealed the SMB was significantly higher in the forest sites than that in the oil palm plantation (P<0.001). Across soil depths (0-2, 2-10 and 10-20 cm), the greatest magnitude of difference in SMB across sites was observed within the 10-20 cm profile. At this depth, both forest sites had greater SMB than the oil palm plantation and between the forest sites, the regenerating forest had greater SMB than the primary forest (P<0.05). In the 0-2 cm depth, both forest sites had greater SMB than the oil palm plantation (P<0.05). Last, in the 2-10 cm depth, there was no significant difference between the primary forest and the oil palm plantation (P =0.06) but the regenerating forest had greater SMB than both the primary forest and the oil palm plantation (P<0.001). At all sites, bacterial-to-fungal ratios declined with sampling depth. Bacterial-to-fungal ratios in the oil palm plantation were significantly lower than in the forest sites and this was most notable in the lower soil horizons (P<0.01). ANOSIM of lipid composition across samples (all horizons summed for each sample) showed that the microbial community composition was distinct across all three sites (P<0.01) for all contrasts. Finally, the results that oil palm ecosystems have lower SMB and altered microbial communities compared to forest sites indicate that plantations may experience altered biogeochemical cycling compared to forest sites.