University of Zürich Institute for Evolutionary Biology and Environmental Studies; Oregon State University Department of Botany and Plant Pathology, United States
Human land-use is causing biodiversity to be lost and re-shuffled and climate change is altering environmental conditions. Yet, how these factors interact to determine disease risk remains poorly understood. Biodiversity loss is often coupled with predictable changes in host trait distributions. Specifically, the hosts best able to spread disease often share similar traits with hosts most resistant to biodiversity loss (fast pace-of-life, high host competence), resulting in more disease as host communities disassemble. A trait-based approach might therefore be a useful framework for addressing this question. However, the relationship between biodiversity loss, host traits, and disease could be limited in situations where existing biotic and environmental filters alter the abundance of fast-paced, more competent host species or alter how particular traits relate to disease risk.
To disentangle how human land-use, environmental change, biodiversity loss, and shifts in host trait distributions jointly contribute to disease risk, I carried out a survey of host specific leaf area (SLA; a trait commonly associated with host competence and pace-of-life) and foliar pathogen infection across a land-use and environmental gradient in semi-natural and cultivated grasslands throughout Switzerland. The survey was carried out in summer 2022 in coordination with the Swiss farmland biodiversity monitoring program (www.allema.ch), and included 2730 plants, embedded in 167 host communities across a 2182-meter elevation, 11.4C temperature, and 141mm precipitation gradient. Botanical surveys of these communities were conducted in 2017 and 2022.
Structural equation modelling revealed that SLA, biodiversity-loss, and disease risk were sensitive to environmental conditions and land-use (Fisher's C=5.0; p=0.76; Disease R2=0.26; SLA R2=0.69; Biodiversity-loss R2=0.89; Land-use R2=0.31). Consistent with trait-based predictions, increasing temperature and biodiversity-loss indirectly increased disease mainly via changes in host community trait distributions: communities located in warmer environments and that lost more host species over time had higher community-weighted mean SLA, which was in turn associated with more severe disease. However, these relationships were restricted to semi-natural grasslands. More intensively used grasslands, which were often located at lower-elevation, warmer sites, were often dominated by high SLA species. But communities with higher community-weighted mean SLA counterintuitively experienced less disease in intensively managed grasslands. Disease risk in intensively managed grasslands also increased strongly with increasing temperature, independent of changing host community structure. Together these results support the hypothesis that biodiversity can regulate disease through predictable changes in host functional traits but suggest that these relationships can become uncoupled by climate and land-use.
