The steering role of plant-soil interactions in natural community dynamics and nature restoration

Abstract

Biodiversity is declining worldwide and many ecosystems have been degraded due to human actions. There have been many attempts to restore degraded ecosystems, but restoration success varies. Past human management has left important abiotic and biotic legacies and active intervention is needed to overcome these legacies. Legacy effects include altered abiotic conditions and limited availability of appropriate seeds. However, plants also have many interactions with the myriad organisms that inhabit the soil. Soil biota include e.g. bacteria, fungi, nematodes, collembolan, and mites. Restoring plant-soil interactions may be key to successful ecological restoration, because studies on natural succession in ecosystems show that both plant and soil communities develop in concert. In addition, late-successional soil communities promote the performance of late-succession plant species that are often the target species for restoration. The aims of my thesis were to 1) test whether inoculation of living soil can improve restoration of species-rich grasslands and dry heathlands, and 2) understand how plant-soil interactions affect plant composition and diversity. In a large-scale field experiment, called “Reijerscamp-experiment”, I tested the potential of soil inoculation to speed up ecosystem restoration. On a former arable field large areas of on average 0.5 ha were inoculated with a thin layer of I also tested how mixtures of inocula from different donor systems affect restoration success. In a greenhouse experiment I made replacement series of soil inocula sourced from arable fields, mid-succession grasslands and dry heathlands and monitored the responses of target and ruderal plant species. The target species all responded positively to higher proportions of heathland material in the inoculum, while the responses of the ruderal species were variable. Interestingly, a 50:50 mixture of arable and heathland inoculum strongly reduced the growth of the ruderal species. Soil inoculation may be considered as a way of microbiome engineering, which is a newly emerging field mainly used to improve human health and agricultural production. My results show that conceptually similar techniques can be applied to improve inocula for the restoration of ecological communities. In a second field experiment I tested the long-term consequences of soil inoculation with and without sowing mid-successional plant species for plant and soil community composition. I found that sowing strongly altered plant community composition for over two decades. Soil inoculation, on the other hand, substantially altered the composition of the soil nematode community and that these effects persisted for at least 15 years. However, in contrast to the Reijerscamp experiment, the effect of soil inoculation on vegetation composition was transient. I propose that in this case the presence of an intact arable top soil, as well as perhaps a too minimal difference between the composition of the donor and recipient soil communities may have limited the impact of the soil inocula. In general, the restoration of plant cover and a number of common (‘matrix’) plant species can be achieved using standard approaches, e.g. reducing site fertility and providing seed material, but creating conditions that allow for coexistence of both locally dominant and rare subordinate species proves much more elusive. Fundamental knowledge on how biodiversity is regulated is needed to restore diverse plant communities including the rare species. Testing plant-soil feedback provides a way to directly study the net consequences of the myriad interactions between plants and soil biota for plant performance and community composition. However, while both plants and soil communities are strongly heterogeneous in space and time, spatiotemporally explicit tests of plant-soil feedback are rare. In a greenhouse experiment I studied how spatial heterogeneity in plant-soil feedbacks influence plant communities. I found that when multiple species conditioned the soil, plant performance was reduced compared to mono-specific soil conditioning. This reduction in competitive ability led to a higher plant diversity in the experimental communities. The plant responses were not related to differences in abiotic conditions, but soil conditioning induced clear changes in fungal community composition. Recent meta-analyses and experiments have shown that spatial heterogeneity in abiotic conditions only promotes plant diversity when the grain of the heterogeneity is larger than the size of individual plants. When it is smaller, heterogeneity simply selects for those species that have the highest root plasticity and this leads to lower plant diversity. Together, these results suggest that spatial heterogeneity in abiotic conditions only promotes plant beta diversity, while interaction with the soil community, primarily soil-borne antagonists, maintains plant alpha diversity. Finally, I used repeated soil conditioning by conspecific and heterospecific species to show that soil feedbacks may carry over across soil conditioning periods. In contrast to what is commonly assumed my data show that heterospecific soil-conditioning can result in equally negative PSF as repeated conspecific soil-conditioning and repeated conspecific soil-conditioning does not always lead to stronger negative feedback. Instead, the particular sequence of plant species that successively condition the soil strongly determines the sign and magnitude of PSF. These results highlight the need to incorporate sequential soil-conditioning in models of plant communities and effective crop-rotations. In conclusion, plant-soil interactions are a key aspect in the natural dynamics of plant communities and can be used to improve restoration of semi-natural ecosystems. Abiotic conditions and dispersal ability determine which species may occur in a given site. However, at small spatial scales plant-soil feedbacks and particularly interactions with soil borne antagonists can enhance plant species diversity. Manipulation of the soil community, through inoculation of soil from well-developed donor sites can speed up natural succession and even steer its direction in the field. However, soil inoculation success will not be universal and depends on the match in abiotic conditions of donor and recipient sites, as well as the community composition of the inoculum and the resident communities. Future studies are needed to test the success of introducing soil communities across environmental gradients.