Soil microbial community responses to active and passive restoration of selectively logged Bornean tropical forest

Samuel J B Robinson^1,2*Dafydd Elias¹Tim Goodall¹Niall McNamara¹Robert Griffiths³Noreen Majalap⁴Yap Sau Wai⁵Nick Ostle²

• ¹UK Centre for Ecology and Hydrology (UKCEH), Wallingford, United Kingdom
• ²Lancaster Environment Centre, Faculty of Science and Technology, Lancaster University, Lancaster, England, United Kingdom
• ³School of Natural Sciences, College of Environmental Sciences and Engineering, Bangor University, Bangor, United Kingdom
• ⁴Forest Research Centre (Malaysia), Sandakan, Malaysia
• ⁵Faculty of Tropical Forestry, Universiti Malaysia Sabah, Kota Kinabalu, Malaysia

Tropical rainforests support critical biogeochemical cycles regulated by complex plant-soil microbial interactions but are threatened by global change. Much of the uniquely biodiverse and carbon rich forest on Borneo has been lost through extensive conversion to monoculture plantation, and a significant proportion of the remaining forest has been heavily modified by selective logging. Ecological restoration of tropical forest aims to return forests to a near pristine state, but restoration initiatives are hindered by limited understanding of the underpinning plant-soil feedbacks, and impacts on soil microbial communities are unresolved. We characterised soil properties and soil bacterial and fungal communities using amplicon sequencing across adjacent old-growth and selectively logged lowland dipterocarp forest in Borneo undergoing either natural regeneration or restoration by enrichment planting. While many soil properties were similar across forest types, we found contrasting responses of different soil microbial groups to active and passive restoration. Bacterial and fungal community composition were generally distinct in old-growth forest and more similar in logged forest. Bacterial alpha diversity and rate of spatial turnover appeared to recover towards old-growth forest with active restoration, while fungal alpha diversity showed slower signs of recovery. The composition and rate of spatial turnover in mycorrhizal communities was most different between old-growth and actively restored forest, possibly resulting from mycorrhizal associations of tree species planted during restoration. Surprisingly, old-growth forest shared fewer microbial taxa with actively restored forest than with naturally regenerating forest, suggesting current restoration practices (removal of lianas and understorey vegetation) may be selecting for different microbial communities. Taken together, our findings show that certain attributes of key soil microbial groups remain distinct from old-growth forest almost two decades after logging disturbance, and some may diverge with active restoration. Changes in enrichment planting practices to promote rehabilitation of belowground communities may be required for successful biodiversity conservation and recovery of vital ecosystem functions.