Tree invasion in secondary grasslands diminishes herbaceous biomass and diversity: A study of mechanisms behind the process

Abstract

Woody plant invasion in grasslands is a global problem widely reported during the last decade (Browning et al. 2008; Van Auken 2009; Simberloff et al. 2010; Vilà et al. 2011; Archer et al. 2017). This process involves the successful establishment of shrubs and/or trees in grasslands with strong consequences on its structure (e.g. biomass, diversity and composition) (Zalba & Villamil 2002; Archer 2009), inducing change in cover and diversity of grass species (Vilà et al. 2011; Chaneton et al. 2013). Although woody plant invasion in grasslands occurs worldwide, the mechanisms by which woody plants affect grass biomass, composition and diversity remains under discussion (Hejda et al. 2009; Eldridge et al. 2011). Resource competition has been invoked as one of the main mechanisms in altering the structure and functioning of invaded grassland communities (Levine et al. 2003; Strayer et al. 2006; Simberloff 2011; Vilà et al. 2011), and in the case of woody plant invasion, adult trees appear as superior competitors for light than grasses (Wilson, 1998). However, changes in light conditions promoted by trees may have positive, negative or neutral effects depending on grass photosynthetic pathway (C3 or C4), tree characteristics (deciduous or evergreen) and climate conditions (Mazía et al. 2016). In the case of deciduous tree invaders, environmental conditions beneath the tree canopy could favor C3 winter grasses (Clavijo et al. 2005; Ansley et al. 2019). However, during spring and summer tree shading of deciduous species could exclude light demanding C4 grasses (Belsky 1994; Scholes & Archer 1997; Clavijo et al. 2005). This effect could be counterbalanced by a decrease in atmospheric water demand below the tree canopy, that decreases water stress of grasses, favoring their establishment and growth, even in the case of no-shade-tolerant C4 grasses (Rossi & Villagra 2003; Ansley et al. 2019). In addition to changes in light conditions beneath trees, woody canopy may indirectly modify herbaceous communities by changing soil conditions. Woody plants can alter soil conditions by leaving a soil legacy in invaded patches affecting the plant community composition in the future (Gioria & Pyšek 2015; Kostenko & Bezemer 2020). Soil legacies after tree invasions include changes in soil seed bank, litter quality and quantity, soil microbiota, and nutrients and organic matter, among others. The “soil memory” can maintain the legacy of past plant species composition, even after their removal (Elgersma et al. 2011; in ‘t Zandt et al. 2020). When trees invade grasslands, the type of tree (deciduous, evergreen, N- fixers and the magnitude of mycorrhizal association, among others) may influence the soil legacy, mainly through altering nutrient availabilities, as compared with soil grassland patches where trees were historically absent. Until now, there are few manipulative experiments that have evaluated separately the mechanisms by which the grasslands communities are affected by tree invasions, unravelling possible soil legacy effects vs direct resource competition (Levine et al. 2003; Vilà et al. 2011). To disentangle the relative importance of such mechanisms in explaining changes in the composition, biomass and diversity of a grassland community, it would be necessary to make reciprocal transplants of grasslands species between original habitats (woody or grassland patches) characterized by different light conditions and historical soil legacy. Temperate grasslands represent one of the most modified biomes by cattle husbandry, agriculture and the introduction of exotic woody species (White et al. 2000; Baldi & Paruelo 2008). Argentinian Pampas region was historically characterized by the lack of trees, which were introduced during the colonial period as a source of shade, firewood and as ornamental species (Chaneton et al. 2013). Nowadays several tree species are invasive in the region, mainly Morus alba, Melia azedarach, Gleditsia triacanthos, Ligustrum lucidum and Pinus sp. (Zalba & Villamil 2002). In particular, G. triacanthos had been introduced in the mid-nineteenth century and is currently the most aggressive exotic woody plant species in the region (Ghersa et al. 2002; Fernandez et al. 2017). Because of the expansion of trees into grasslands of the region, there is an increasing interest in studying the effects of woody plant invasions on resident communities and their impact over time (Strayer et al. 2006). In particular, during the last decade several authors have warned about the magnitude of the ongoing tree invasion and their potential impacts on the diversity and composition of the Pampas grasslands (Cuevas & Zalba 2010; Simberloff et al. 2010; Chaneton et al. 2013). The objective of this work was to evaluate the relative importance of the changes in light conditions generated by the invader G. triacanthos L. tree (honey locust), and the historical soil legacy of vegetation (grassland or woody patch origins) in modulating aboveground and belowground biomass, species composition and diversity of a secondary successional grassland invaded by G. triacanthos trees. We established a long-term manipulative reciprocal transplant (soil monoliths) coming from two origins: grasslands (G) and invaded woody patches (W) into three different habitats: open grasslands (OG), woody patches (WP), and artificial shaded grasslands (SG). The artificial shaded treatment (see methods for details) was employed to simulate light conditions beneath a tree canopy, to evaluate if the observed effects on the grassland´s community were based only on modifications in the light conditions or in other additional mechanisms. We postulated the following working hypothesis: 1- Light condition of a habitat, independently of the soil monolith origin (grassland or woody), is a main driver of differences in grassland biomass (aboveground and belowground), diversity and plant functional types. We predict that, independently of the origin, soil monoliths placed in open grasslands will have higher biomass and species diversity than soil monoliths placed beneath trees or artificial shade, and that C4 and C3 species will be dominant in OG and WP, respectively. Alternatively: 2- The soil monolith origin acts as a secondary driver of differences in community biomass, diversity and plant functional types due to differences in soil legacy, according to the soil monolith origin (grassland or woody habitat). We predict that woody soil monoliths placed in open grasslands, beneath the tree canopy or under artificial shade, will always have lower biomass, lower species diversity and more dominated by C3 species than soil monoliths coming from grasslands.