This dataset contains the results from a metabarcoding study of terrestrial leech blood meals to detect differences in the diets of two leech species, Haemadipsa picta and Haemadipsa sumatrana. Mammal taxa were identified using metabarcoding of 16s rRNA and comparisons of operational taxonomic units (OTUs) to a curated reference database from NCBI (National Center for Biotechnology Information) GenBank. All leeches were collected from the Stability of Altered Forest Ecosystems project (SAFE; www.SAFEproject.net) as part of the NERC Human Modified tropical Forest Programme and the LOMBOK consortia (Land-use Options for Maintaining BiOdiversity & eKosystem functions). Leech samples were collected at different sites across a habitat gradient, to assess these invertebrates as molecular sampling tools for mammals. Individuals were pooled before amplicon sequencing with Illumina MiSeq 150-200bp x2. The resultant raw sequences were filtered and clustered at 97%, curated and then assigned to the reference database using BLAST and MEGAN programmes. Full details about this dataset can be found at https://doi.org/10.5285/3affed0d-fe6f-4916-89e3-e672639191e5

57 soil samples were collected from 12 sites across Patagonia, the sub-Antarctic and the Maritime Antarctic between January and February 2018. The dataset presented here comprises two tables. The first shows the numbers of DNA reads of 105 genera of plant or animal pathogenic fungi in the soil samples. In the second table, for each of the 57 soil samples, seven bioclimatic variable predictors, six measures of soil chemistry, the total number of fungi reads and plant and animal pathogen richness and abundance are given, as well as the relative abundance of each fungal genus in each sample. The data in this second table were inputted into LASSO regression models and species indicator analyses. This research was jointly funded by the NERC-CONICYT award (NE/P003079/1 and PII20150126, respectively) and Danish National Research Foundation award (VOLT, DNRF 168).