Direction of plant–soil feedback determines plant responses to drought

Abstract

Droughts exarcerbate Plant–soil feedbacks (PSFs) making positive PSFs more positive and negative PSFs more negative. Alterations in PSFs that droughts induce could relate to the rooting depth of the tested plants. We present some rare evidence on how a driver of global change will alter a biotic interaction. A large body of literature addresses how ecosystems currently function and how this might change in the near future in response to components of ongoing climate change, such as higher in frequency droughts (e.g. IPCC, 2021). Even though existing models capture well how the fitness of organisms scales with changing environmental factors, they often downplay the role that biotic interactions might play in structuring future ecosystems (Blois et al., 2013). It has been proposed that climate change could among others intensify interspecific competition, foster the release of some species from their predators (i.e. ecological release), but also bring about novel interactions (Gilman et al., 2010). Most of these hypotheses, however, remain untested because it is difficult to find suitable datasets to analyse (Blois et al., 2013). There is also an apparent possibility that either through novel interactions or changes in existing interactions, climate change alters the fitness of organisms selectively, leading on the longer term to competitive exclusion of some species and eventually extinctions (Blois et al., 2013; Gilman et al., 2010). Any group of organisms, but particularly plants, that experience community shifts should evoke alterations in the functioning of the ecosystems they populate (e.g. Singh et al., 2018). A type of biotic interactions that is particularly pertinent for the functioning of ecosystems is the biotic constituent of plant–soil feedbacks. Plant–soil feedbacks (PSFs) are described as ‘changes to soil properties that are caused by plants, which in turn influence the performance of plants’ (van der Putten et al., 2013) and are classified as either positive, when a plant grows better in ‘trained’ soil (i.e. the plant grew the previous growth season in that soil), or negative, when it grows worse in trained soil. Plant–soil feedbacks have both abiotic and biotic constituents. For example, a well-known form of negative biotic plant–soil feedbacks arises from the gradual accumulation of pathogens and herbivores, which led to the ‘Janzen-Connell’ hypothesis (e.g. Wills et al., 2006). A likely reason why terrestrial plant communities are so diverse and resilient to environmental variability is the prevalence of negative plant–soil feedbacks (Forero et al., 2021). The climatic driver that has been studied most in relation to plant–soil feedbacks is drought (Beals et al., 2020). Droughts are predicted to increase in frequency and intensity (IPCC, 2021) and so the body of work available presents an important opportunity for synthesis on the way the biotic constituent of PSFs changes with this key driver of global change. Most existing studies report less negative PSFs following drought (summarized by a meta-analysis, comprising only seven studies addressing PSF and drought, by Beals et al., 2020). To the best of our knowledge, however, the existing literature completely overlooks the possibility that drought-induced alterations in PSF deterministically vary across plant species or key traits related to drought. We summarize here some key results from a quantitative synthesis on how drought alters PSFs using 16 studies (see Appendix S1). We asked (1) whether there is a relationship between the magnitude of PSF and the extent to which droughts induce PSF (Question One), and (2) whether it is possible to link drought-induced alterations in PSF to belowground plant traits (Question Two). To address these questions, we synthesized studies where PSFs, in the form of a log-response ratio of either plant biomass or seed biomass in trained soil relative to a sterilized control (i.e. often called PSF index), had been assessed at two or more levels of moisture availability. We used as an effect size the difference in those log-response ratios (ΔPSF): a positive ΔPSF thus suggests that drought favours plants grown in trained soil compared to control conditions, whereas a negative value suggests drought favours plants in sterilized soil. We additionally report on some quality control considerations on the meta-analysis such as adherence to the PRISMA guidelines, a power analysis and a sensitivity analysis (Appendix S2) and present the raw data and the list of studies (Appendices S3 and S4). The take home message of our meta-analysis is that plants experiencing positive PSFs were more likely to increase their PSFs following drought and vice versa (Figure 1a; Appendix S1). This finding suggests that there is a qualifier, relating to the original intensity of PSFs, that determines how PSFs change following droughts. Many of the determinants of PSFs, such as adaptations of plants or microbes to the specific environments (Remke et al., 2020), microbial community structure and diversity and seasonality can have an overwhelmingly large impact on PSF observations compared to the influence of the plant species used. Our observation could thus have been exclusively due to systematic differences in ΔPSF because of these factors. To address this possibility, we further explored relationships between ΔPSF and four plant traits: a pertinent trait for drought tolerance, rooting depth; a representative of the Leaf's Economic Spectrum, specific leaf area; a trait related to dispersal potential and thus an avoidance strategy, seed mass; and a trait describing life history, plant longevity (i.e. whether plants were annuals or perennials; Figure S1). In the two scenarios we examined (i.e. a direct relationship with ΔPSF or to correct for the likely influence of growth settings an indirect relationship with the residuals from Figure 1a), a belowground trait, root depth, showed strong trends which despite the limited statistical power were significant in the case of the relationship with the residuals (ρ = 0.74, p = .046; Figure 1b,c; Appendix S3). Because of the small sample size, we remain cautious in interpreting the likely relationship between rooting depth and ΔPSF. What becomes apparent, however, is that drought-induced alterations in PSF differ for different species. This raises the question whether there may be plant traits that systematically drive such alterations in PSF, which could be based on the limited evidence we present here, as well as be rooting depth. Although numerous PSF experiments have been undertaken (Beals et al., 2020), it has not always been easy to synthesize across studies. Kulmatiski et al. (2008) carried out a synthesis of 45 studies to observe that the main driver that induces systematic differences in PSFs was the longevity of plants, with annuals (but also early successional plants) experiencing more negative PSFs than perennials. Our observations (Figure 1a) could therefore imply that the expected increases in the intensity and frequency of droughts might speed up succession, but also select against annual plant species (Proposition One). Annuals often maintain shallower roots than perennials, which agrees with our observations (Figure 1b) that plants with shallow roots might experience stronger negative PSFs than deep-rooted plants (Proposition Two). Narrowing the focus into two traits, longevity of plants and rooting depth (i.e. through formulating those two propositions) might be the first step in designing targeted experiments where systematic changes in PSFs can be tested. Additionally, these two parameters represent plausible drivers of PSFs under drought settings. As an example, deep-rooted plants tolerate droughts better (Chaves et al., 2003), but also tend to maintain a larger volume of root in the soil that influences soil biotic properties to generate PSF. It is thus likely that root depth also determines how benefits from PSF will increase or decrease from the anticipated change in climatic conditions. Obviously, PSFs also have an abiotic constituent; however, the way we combined estimates through the ΔPSF index corrected for abiotic changes that the plants had induced in trained soil, and so our observations likely captured exclusively biotic constituents of PSFs. We present here expectations on how a key component of climate change might alter PSFs, which is an underexplored process relative to biotic interactions. We partially support our expectations with a meta-analysis of 16 studies. PSFs are a critical mechanism by which plants interact with soil biotic and abiotic conditions that leads to substantial effects on plant fitness, community composition and ecosystem processes. We show that plants experiencing positive PSFs were more likely to increase their PSFs following drought and vice versa, and this was related to species identity and rooting depth. Thus, we expect that climate change may exaggerate plant–soil interactions and the resulting impacts on individuals, communities and ecosystems. The authors declare that there is no conflict of interest. Raw data are made available in the form of Appendices Three and Four. Data are also made available through zenodo with the DOI 10.5281/zenodo.6465805 (the full link is: https://github.com/StavrosVeresoglou/Direction-of-plant-soil-feedback-determines-plant-responses-to-drought/releases/tag/v1.0.0). DataS1 Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.

Topics & Concepts