Agricultural land-use legacies affect soil bacterial communities following restoration in a global biodiversity hotspot

Ecosystem restoration in post-agricultural landscapes is a critical response to agricultural land abandonment, climate change, and the escalating biodiversity crisis. However, effective restoration of these landscapes can be hampered by land-use legacies that create biotic and abiotic barriers to ecosystem recovery, particularly in ancient Tertiary landscapes where vegetation is adapted to nutrient deficient soils. While our understanding of how to overcome these barriers when restoring plant communities is improving, there is limited knowledge of how these legacies impact on recovery of soil microbiota - the biodiverse and functionally-important communities of soil microbes. Here, we used amplicon sequencing of the bacterial 16S rRNA gene extracted from soils across a restoration project in southwest Western Australia, a global biodiversity hotspot, to examine recovery of soil microbiota following post-agricultural restoration. We sampled soils at six sites under four land conditions - degraded post-agriculture, actively revegetated post-agriculture, passively regenerated, and remnant bushland - generating 1,609,618 sequences corresponding to 15,009 unique bacterial taxa. We show that soil bacterial communities in revegetated and degraded samples were similar across sites but strongly dissimilar to adjoining remnant samples. We show that limited recovery of bacterial communities was linked to elevated soil phosphorus levels. Together, our results indicate soil microbiota have not recovered despite revegetation taking place up to 17 years ago, and this lack of recovery is likely driven by soil nutrient legacies from past agricultural practices. Our study highlights a key challenge faced by conservation practitioners when integrating soil microbiota into ecosystem restoration in post-agricultural landscapes in ancient, nutrient-poor landscapes.