Successional Generation of Functional Multidiversity

Biodiversity is of indispensable value for ecosystem functioning and stability and provides multiple direct- and indirect-use values crucial for human wellbeing. The different ecological roles of plants, animals and microorganisms in ecosystems as well as their manifold interactions demand comprehensive inventories of all organisms present in a habitat. Such inventories allow assessing the multidiversity of ecosytems, i.e. the cumulative diversity of all organisms in an ecosystem. Causes and consequences of multidiversity declines are increasingly studied, but similar large-scale endeavors investigating the origin of and the increase in multidiversity are missing. Austrian glacier forefields provide an excellent opportunity to track the concerted and interdependent increase in plant, arthropod, and microbial diversity from virtual absence of life to vibrant ecosystems. In these landscapes colonizable substrate age is well documented (time since deglaciation) and can be correlated to the diversity of plants, arthropods, and microorganisms. Field surveys will be complemented with microcosm experiments designed to investigate the mechanisms underlying the establishment of multidiverse communities. The dataset gathered during the project will be analyzed either with state-of-the-art statistical methods, or with methods that will be specifically developed. These analyses will reveal successional transformations of ecological communities and interdependencies between organisms. These approaches allow the detection of ecological processes that remain hidden using established and available statistical methods. The major aims of the project are a) to develop statistical approaches to analyze complex multidiversity data and to answer fundamental ecological questions that will also be widely applicable to other disciplines apart from ecology and b) to gain a comprehensive understanding of ecosystem processes involved in the establishment of multidiversity. These findings will be essential for future conservation and restoration efforts of natural and anthropogenic altered ecosystems.

Biodiversity loss is major threat for ecosystem functioning and human wellbeing. While the rate of biodiversity loss of different taxa and ecosystem functions are often the focus of scientific research, the successional increase of multiple interacting taxa is understudied. In summer 2019 and 2020, we established 140 permanent plots along the successional gradient of the glacier forefield of the Ödenwinkelkees (in the Land Salzburg, Austria) and performed a full inventory of vascular plants, bryophytes, soil inhabiting, and plant associated bacteria and fungi, as well as below and above ground arthropods. Additionally, environmental parameters were assessed. Based on these data, we tracked the increase in diversity as well as changes in community composition and tested the prevalence of assembly processes in different taxa as well as at different successional stages. Our results imply that stochastic, likely dispersal-dominated, processes are replaced by rather deterministic processes such as environmental filtering and biotic interactions after around 60 years of succession. These shifts in ecosystem states along successional gradients occurred abruptly once abiotic and biotic factors dominate over dispersal as main driver. To validate correlational findings resulting from field data, we designed lab experiments to specifically test hypotheses generated from field data. These lab-experiments supported the threshold-induced change in successional processes and the mutual dependence between taxa in community assembly. Additionally, our project also introduced statistical approaches capable of detecting non-monotonic or non-functional dependencies. We developed qad (short for quantification of asymmetric dependence), a non-parametric statistical method to quantify directed and asymmetric dependence of bivariate samples and demonstrated its applicability on ecological data. Using qad, we showed that up to 59% of all significant associations in the studied glacier forefield are non-monotonic. Further, we showed that pairwise associations between plants, bacteria and fungi are specifically characterized by their strength and degree of monotonicity, e.g., plant-microbe associations are clearly separated from microbe-microbe associations. To conclude, our results advance the understanding of ecological successions, provide novel statistical methods for ecology and beyond, and demonstrate the importance of ecological non-monotonicity.