Neglecting non-vascular plants leads to underestimation of grassland plant diversity loss under experimental nutrient addition

Abstract

1 INTRODUCTION

Ecological theory predicts that nutrient addition should cause plant diversity loss (Tilman, 1982), and this has been shown experimentally in many studies of plant communities worldwide (Borer, Seabloom, et al., 2014; Harpole et al., 2016; Seabloom, Batzer, et al., 2021). Furthermore, theoretical and empirical work shows that nutrient-induced losses in plant diversity may be mitigated when herbivores increase light at ground level (Borer, Seabloom, et al., 2014; Olff & Ritchie, 1998). Experimental tests of drivers of plant diversity in terrestrial systems typically focus on vascular plants (VP), while non-vascular plants (NVP; here bryophytes), are usually ignored, even though NVPs can commonly represent 20%–40% of total plant diversity in grassland communities (Dengler et al., 2020; Klaus & Müller, 2014; Löbel et al., 2006; Lyons et al., 2022; Tansley & Adamson, 1925). It is therefore not known whether the response of NVP diversity to nutrient addition or grazing is consistent with that of VPs, in which case VP responses could predict NVP responses. However, NVPs could react differently (e.g. decrease to a greater or lesser degree), in which case their inclusion would refine our understanding of plant diversity loss in terrestrial systems.

In grassland ecosystems, NVPs consist mainly of two clades of land plants (Bryophyta (mosses) and Marchantiophyta (liverworts)) that contribute to many ecosystem functions and properties. Bryophytes often make up a large proportion of above-ground biomass (Boch et al., 2018; Hejcman et al., 2010; Wielgolaski, 1972), regulate the microclimate (Jaroszynska et al., 2023), hydrology (Michel et al., 2013), carbon and nutrient cycling (O'Neill, 2000; Turetsky, 2003), nitrogen fixation (Lindo et al., 2013), and affect seed germination of native and exotic vascular plants (Dollery et al., 2022), soil micro-organism diversity (Xiao et al., 2023) and soil multifunctionality (Xiao et al., 2024). Therefore, the presence and diversity of NVPs in terrestrial systems can have far-reaching ecological repercussions even in VP-dominated ecosystems. Important differences between NVPs and VPs, which may affect their ecological role, include that NVPs are an order-of-magnitude shorter than VPs, lack efficient conducting tissues, roots and stomata, and have relatively low photosynthetic rates (Rydin, 2009). Their small stature makes NVPs poor competitors for light, for which reason NVPs likely experience higher species loss rates than VPs, when increased nutrient supply intensifies competition for light (Rydin, 2009).

A handful of empirical studies over the past few decades provide expectations for NVP responses to increased nutrient availability or grazing pressure. Experimental nutrient addition can reduce the diversity of NVPs in various ecosystems (Bergamini & Pauli, 2001; Jägerbrand et al., 2006; Jäppinen & Hotanen, 1990; Jonasson, 1992; Wilson & Tilman, 2002), often due to increased competition for light with VPs (Aude & Ejrnæs, 2005; Bergamini & Pauli, 2001; Boch et al., 2018; Cusell et al., 2014; Jonasson, 1992; van der Wal et al., 2005; Virtanen et al., 2017). Excessive nutrient loads also may have direct negative effects on NVP survival (Armitage et al., 2012; Pearce et al., 2003; Virtanen et al., 2000). While nitrogen (N) is often assumed to be the most influential nutrient affecting NVPs, joint additions of N and phosphorus (P) can result in stronger negative effects on NVPs than additions of single nutrients (Bergamini & Pauli, 2001; Øien et al., 2018). However, nutrient effects are not entirely consistent among studies, with some finding relatively weak effects on NVPs (Stevens et al., 2004, 2006; Sun et al., 2017) or even positive relationships between nutrient levels and NVP abundance or species diversity (Gordon et al., 2001; Ingerpuu et al., 1998; Slavik et al., 2004). Therefore, the responses of NVPs to nutrient addition appear variable, which suggests that the responses are context dependent.

As with nutrient effects, grazing impacts on NVPs have been found to vary from positive (Bernes et al., 2018; Chollet et al., 2013; Ingerpuu & Sarv, 2015; Takala et al., 2012, 2014; Tansley & Adamson, 1925) to neutral (Austrheim et al., 2007) or negative (Boch et al., 2018; Spitale, 2021; Virtanen & Crawley, 2010). Herbivores can enhance plant diversity by promoting local colonization, stimulating germination from the soil diaspore bank, or altering plant competitive interactions (Olff & Ritchie, 1998). Biomass removal by herbivores can alleviate the competitive effects of VPs on NVPs, compensating for the negative effect of fertilization (Aude & Ejrnæs, 2005). On the other hand, intense grazing also may act as a strong disturbance causing loss of NVP species (Downing, 1992). Based on current, scarce and mixed evidence, it remains uncertain how NVP diversity will respond to nutrient addition and changes in grazing pressure separately and in interaction. Importantly, if NVPs respond differently from VPs, this could alter predictions for loss rates of grassland plant diversity.

Ignoring NVPs may lead to no effect, over- or underestimation of species loss rates under nutrient addition and/or changes in grazing in grasslands. More specifically, these outcomes can arise under the following conditions: (i) No effect on species loss rate estimates. This is the case if NVP and VP respond similarly to nutrient addition and grazing exclusion, or if NVPs are too scarce, and diversity is too low to have any effect. (ii) Overestimation of species loss rates. This could occur if the loss of VPs is compensated for by increased NVP diversity due to direct NVP diversity benefits of nutrient addition or altered level of grazing. (iii) Underestimation of species loss rates. This could occur if NVP diversity declines more than VP diversity due to nutrients directly suppressing NVPs or due to increased competition for light with VPs.

Here, we test the effects of nutrient addition and grazing exclusion on NVP and VP diversities at nine sites that are part of the globally distributed Nutrient Network experiment (Borer, Harpole, et al., 2014). The sites differ in their species composition and environmental context, which enables the quest to find generalities in the responses and mechanisms driving plant diversity. Specifically, we compare the responses of NVP and VP species richness, diversity and evenness to factorial addition of different nutrients (N, P, potassium (K)) and to NPK_+μ and grazing exclusion treatments to uncover whether accounting for NVP responses alters the overall estimates of changes in grassland plant diversity currently based on VPs only.