Recruitment filtering by a moss layer disadvantages large-seeded grassland species

Abstract

Calcareous grasslands are rich in biodiversity and thus receive much attention in nature conservation. In such grasslands, the formation of moss layers is perceived as a management problem. However, its impacts on the community level are complex, as not only inhibition but also facilitation of vascular plant recruitment occur. Possible filters of recruitment are shading by mosses, isolation from soil resources and the resulting desiccation. To understand how seed size and shape moderate moss effects, a combined glasshouse and field experiment was conducted in southern Germany. Seeds of 14 species from calcareous grasslands were sown either on top or underneath a moss layer, or on bare soil without moss. We determined the total number of emerged and established seedlings, i.e. the ones that survived until the end of the experiment. Both measures were reduced for most species sown on moss, while mortality was slightly enhanced by moss presence. Seed size explained a significant proportion of the moss effects on plant recruitment. Inhibitive effects on recruitment increased with seed size when seeds were placed on top of moss. When germinating underneath moss, the effect on emergence changed from negative in small-seeded plant species to positive in large-seeded species, but this effect was insignificant for establishment. The positive response in large seeds was probably due to their higher moisture requirements for imbibition, and lower dependency on light for germination. However, moss-seed-interactions were not affected by seed shape. Seedling establishment of plants with large seeds strongly depended on their initial position in moss, while small-seeded species showed no effect. Hence, moss layers constitute a selective recruitment filter by modifying seed penetration and hence germination conditions. This effect could increase beta diversity of vascular plant species in calcareous grasslands, and thus conservation should aim at intermediate and patchy moss abundance.

Introduction

During community assembly, environmental filters and biotic interactions drive plant establishment and population persistence (Cornwell and Ackerly, 2009, Kraft et al., 2015). In grasslands, the presence of gaps, litter and moss act as environmental filters for germination and seedling establishment. Interactions between mosses and vascular plants mainly concern seeds and seedlings (During & van Tooren, 1990; Špacková, Kotorová, & Lepš, 1998); a summary of studies on moss effects was compiled by During and van Tooren (1990).

Inhibition of seedling recruitment by moss is caused by (i) a physical barrier leading to desiccation (Drake, Grimshaw-Surette, Heim, & Lundholm, 2018; van Tooren, 1988), (ii) competition for nutrients (Gornall, Woodin, Jónsdóttir, & van der Wal, 2011; Zamfir, 2000), (iii) reduced light supply (Donath & Eckstein, 2010; Keizer, van Tooren, & During, 1985), and (iv) allelopathic effects (Soudzilovskaia et al., 2011, van Tooren, 1990). Facilitation can be due to (i) buffering of soil temperature and moisture (Donath and Eckstein, 2010, Soudzilovskaia et al., 2011), (ii) accumulation of organic debris (Jeschke & Kiehl, 2006), and (ii) protection from predation (van Tooren, 1988). Additionally, mosses trap seeds, causing clumped recruitment (van Tooren, 1988).

Small-scale recruitment filters often interact with other environmental conditions (Cingolani, Cabido, Gurvich, Renison, & Díaz, 2007; Grman, Bassett, Zirbel, & Brudvig, 2015). Moss layer effects vary with shading, soil moisture, nutrients (Otsus and Zobel, 2004, Zamfir, 2000), and layer thickness (Donath and Eckstein, 2010, Gornall et al., 2011). Moss–seedling interactions are also moderated by seed size (Donath and Eckstein, 2010, van Tooren, 1990), which is associated with germination requirements for light (Milberg, Andersson, & Thompson, 2000) and humidity (Boyd & van Acker, 2004), while effects of seed shape have not been documented.

In central Europe, moss layers often hinder regeneration of vascular plant species in semi-natural grasslands that are managed for conservation (Jeschke and Kiehl, 2008, van Tooren, 1990). Moss effects are also significant in restoration ecology, since moss dominance often occurs in restored grasslands (Drake et al., 2018; Eichberg, Storm, Stroh, & Schwabe, 2010; Jeschke & Kiehl, 2008). We suspect that desiccation and the barrier to the ground hamper recruitment in large and aspherical seeds, which tend to be trapped on moss surfaces. Round and/or lightweight seeds may penetrate moss layers and reach the soil, but reduced light intensity may hinder their germination.

This study aimed to find evidence for these mechanisms and to determine their net effects on recruitment of vascular plants with differing seed traits. Donath and Eckstein (2010) showed that seed size interacts with the initial vertical position of seeds in moss, indicating that seed size controls penetration into the moss layer. We adopted their approach and extended it to seed shape. We further investigated how moss affects different aspects of seedling recruitment along gradients of seed size and shape, since Jeschke and Kiehl (2008) showed germination to be more strongly influenced than mortality in three plant species. On those grounds, our hypotheses on recruitment effects of moss layers in calcareous grassland were:

• Seedling establishment is negatively affected by a moss layer.

• Recruitment effects of mosses are driven by germination rather than by seedling mortality.

• Seed size and shape determine how moss presence and the initial seed position within moss control net recruitment.

• Establishment of species with heavy-aspherical seeds is altered by the position of seeds in the moss layer, while species with small-round seeds are unaffected by their initial position as they tend to penetrate to the ground surface in any case.