AbstractAimTo assess the largely neglected patterns and causes of latitudinal diversity gradients in soil protists, using South American testate amoebae as model organisms.LocationSouth‐western South America (SSA).MethodsWe assessed the relationship among richness, latitude and predictions of eight ecological hypotheses in SSA, a region characterized by harsh abiotic conditions northward and southward from mid‐latitudes. We also examined the patterns of latitudinal variation in species range sizes, species composition and phylogenetic relatedness among species to assess whether the responses of species to current climate rely on evolutionary constraints.ResultsRichness followed a unimodal trend with abrupt declines northward and southward from mid‐latitudes. This trend was mostly explained by water–energy balance. To our knowledge, this mechanism has never been reported for soil microorganisms before. Species range sizes were larger northward and southward from mid‐latitudes, a pattern which we have called the ‘two‐way Rapoport effect’. Species composition was due to turnover at mid‐latitudes and to nestedness towards low and high latitudes. Species at mid‐latitudes tended to be less related to each other than those at low and high latitudes, supporting the existence of evolutionary constraints to water–energy availability.Main conclusionsThe unimodal response of species diversity to water and energy availability stems from a combination of the ancestral adaptation of testate amoebae to warm, humid climates and past climate changes that set up permanent harsh abiotic conditions northward and southward from mid‐latitudes. Given that other soil microorganisms also depend on water and energy to survive and reproduce, we predict that the unimodal response to water–energy balance observed will also be valid for a large number of microbial groups in SSA. While this diversity trend may not hold true for testate amoebae from other regions, we predict that their evolutionary constraints to water–energy availability will always produce strong relationships among richness, water and energy at large spatial‐scales.