Traits response to herbivory
Overall, there was a high proportion of studies which found non-significant trait responses to herbivory (Figure 6). At the study level, plant traits might not change with herbivory because the traits measured may not be indicators of specific plant-herbivore interactions studied in their focal system. It may also be that some traits which confer grazing tolerance/avoidance are also helpful, or of neutral advantage under ungrazed conditions. This is particularly relevant for vertebrate focussed studies which often compared plant traits values between grazed areas and exclusion plots, which usually had grazers excluded for between 5-10 years. The dominant plants that exist in these exclosures likely possess traits which confer grazing avoidance/tolerance, but these traits may not be a fitness disadvantage to the plant in ungrazed conditions and remain similarly expressed.
Trait response to vertebrate herbivory was most frequently recorded for SLA, LDMC, leaf nitrogen, plant height and leaf area; wherefor invertebrate herbivory it volatile organic compounds (VOCs), leaf nitrogen, aboveground biomass, total phenols and total tannins were most frequent. Vertebrate herbivory was mostly positively correlated with leaf nitrogen and SLA, and negatively with LDMC (Figure 6). As SLA and leaf nitrogen are positively correlated with high relative growth and LDMC positively correlated with leaf carbon content (Perez-Harguindeguy et al., 2016), these results suggest that many of the plant species and communities examined in these grassland studies, invest more in growth to tolerate herbivory, and less in carbon based defensive structures to avoid herbivory. This also aligns with the response of photosynthetic capacity and leaf phosphorous content, which mostly responded positively to vertebrate herbivory (Figure 6). Indirect influences from grazing such as nitrogen deposition through urine and faeces, may also increase the availability of nutrients in grazed environments and facilitate growth (de Mazancourt et al., 1998). Tolerating herbivory through rapid growth, is a common strategy for species from the Poaceae family, which dominate grasslands. This is particularly true for grass species growing in productive environments (Briske, 1996; Briske, 1999; Diaz et al., 2007; Díaz et al., 2001), where their dominance is often mediated via herbivory (Lunt et al., 2007). When grazers are removed in these environments, dominant grasses can sometimes reach a point of maximum growth, where restrictions space and/or nutrients limits further growth, and their growth rates may decline. In this scenario, grasses may invest more carbon in their stems and less in their leaves as structural support to grow tall under competitive light conditions (Irving, 2015).
In contrast, some studies found negative relationships between vertebrate herbivory and leaf area and plant height (Figure 6), which are typical mechanisms of grazing avoidance, documented in many other studies (Landsberg et al., 1999; McNaughton & Sabuni, 1988; Noy-Meir et al., 1989; Sala et al., 1986). A negative relationship between herbivory, leaf area and plant height may also be due to increased light availability in grazed areas, potentially reducing the need for plants to grow tall to access the light (Borer et al., 2014).
VOCs were mostly examined in relation to invertebrate herbivory and responded mostly positively, with only a handful of studies reporting a negative relationship with herbivory (Figure 6). This suggests that VOCs are mostly expressed or up regulated in response to herbivory. VOCs may also increase in response to a variety of other stimuli (Vivaldo et al., 2017) and play an important role in plant-plant and cross-trophic signalling (Baldwin et al., 2002; Mäntylä et al., 2008). Kigathi et al. (2009) who found the expression of some VOCs to decrease with invertebrate herbivory in the field, suggest that under field conditions the plant is responding to multiple stimuli at once and this may also affect the type and abundance of VOC emissions. They also highlight that reduced VOC production in response to herbivory may be due to the allocation of carbon and nitrogen being prioritised for growth or for production of others defensive compounds over VOCs.
Where a significant response was found, leaf nitrogen was usually positively correlated with invertebrate herbivory, while leaf carbon to nitrogen ratio was usually negatively correlated (Figure 6). Similar to the response to vertebrate herbivory this result suggests that for these plant species, herbivory potentially results in a greater investment in growth over carbon-based defence as indicated by higher leaf nitrogen for every part carbon. Because many invertebrate focussed studies compared leaf traits from leaves with and without visual evidence of herbivory, greater nitrogen in grazed leaves may reflect herbivore preference for leaves with higher nitrogen, rather than the leaf increasing nitrogen in response to herbivory (Loranger et al., 2012).
Aboveground biomass (measured for individual species) responded mostly negatively, but also showed little evidence of a change with invertebrate herbivory. The non-significant responses mostly came from one paper however (Ladygina et al., 2010) which looked at the effect of belowground herbivores (nematodes and wire worms) on aboveground biomass. Because aboveground biomass was not being directly removed via herbivory the effect of herbivory may be more nuanced. Aboveground biomass was also mostly negatively associated with vertebrate herbivory. One study was the exception, finding perennial forb biomass was higher in deer grazed pastures in comparison to ungrazed exclosures, likely due to reduced competition from grass species (Paige, 1992). In fact, in some grasslands the abundance and richness of forbs is thought to dependent on niche construction by large herbivores, due to the legacy of large and mostly extinct vertebrates, such as mammoths (Bråthen et al., 2021). This points to an important distinction between direct herbivore effects and indirect herbivore effects on plant traits.
Phenols are often attributed to herbivore defence and are usually hypothesised to increase in response to herbivory (Salminen & Karonen, 2011). Phenols are an extremely diverse group of compounds, however, and play several other roles in plant metabolism (Salminen & Karonen, 2011). The production of phenols is also reliant on adequate photosynthetic capacity to allow for the accumulation of carbon molecules (Frier et al., 2012). Total phenols responded mostly positively to vertebrate and invertebrate herbivory, although one study found phenols to decline in response to herbivory across multiple sites (Knappová et al., 2018) (Figure 6). Reported high rates of folivory in this study may have compromised the photosynthetic capacity of plants and limited phenol production (Knappová et al., 2018).
Leaf silica was relatively frequently measured (31 measures) in response to both vertebrate and invertebrate herbivory and where a significant effect was found was mostly positively associated with herbivory (Figure 6). Most studies examining the response of silica were on grass species, which are known for their relatively high silica content in comparison to forbs and their ability to accumulate silica in response to herbivory, but also in response to increases in soil silica content (Hall et al., 2019). Within the studies from this review, silica responded positively to herbivory in eight different grass species (Appendix B). One study from the tundra grasslands in Norway, found that for some grass species, silica content declined with summer reindeer herbivory (Petit Bon et al., 2022). Here they suggest that silica accumulates with age in these grass species, and perhaps herbivory is keeping the phenological age of the leaves young and thus maintaining lower leaf silica levels than un-grazed plants (Bañuelos & Obeso, 2000). Understanding how plant defence traits change with stages in plant phenology is an interesting area of future research and may help to untangle variable responses of plant traits to herbivory.