dynamic modelling, montane cloud forest, recruitment dynamics, seedling survival, species coexistence, ungulate herbivory
Introduction
Most studies on recruitment dynamics have tested the Janzen-Connell hypothesis:
specialist natural enemies (usually belonging to pathogens or insects) can suppress abundant species through conspecific density-dependent mortality and thus benefit rare species (Dyer et al., 2010; Bagchi et al., 2014; Comita et al., 2014; Norghauer and Newbery, 2014). However, still few researches address mammalian herbivores, especially ungulates, which can effectively mediate recruitment dynamics by herbivory in the understory, and affect species coexistence of tree seedlings (Asquith et al., 1997;
Royo and Carson, 2005; Smit and Vandenberghe, 2007; Royo et al., 2010; Paine et al., 2016). Empirical studies on how ungulate herbivory affect seedling species coexistence are rare for two reasons. First, ungulates are more often generalists (mixed feeders with flexible diets by browsing or by grazing) and purely dietary specialization are very rare for their nutritional requirements (Dearing et al. 2000). Secondly, ungulate herbivory is usually underestimated because compound causes of death or “vanishing deads” of seedlings due to grazing by ungulates (Hulme 1996) are not easy to identify or distinguish in field surveys.
Because ungulates are more often generalist herbivores to tree seedlings, knowing how ungulate herbivory affects on seedling survival i.e., either causing species-specific deaths or just random deaths irrespective of species, is prerequisite to explore consequent herbivory effects on species coexistence. In particular, unevenly distributed species-specific deaths can greatly change species assembly in bio-communities (Wright 2002).
Usually, non-random, species-specific deaths occur at early stage of seedling establishment (Green et al. 2014); however, over ontogeny, mortality of seedling species is towards random while species richness of the survivors become higher. These imply
that a positive consequence for species coexistence can arise from species-specific seedling deaths at early stage. Species-specific deaths of new recruits or younger seedlings may be caused by some selective browsers (Royo et al., 2010; Curran and Webb, 2011); however, whether generalists (e.g. mixed feeders or grazers) can cause species-specific deaths for new recruits remains inconclusive. These are in contrast to chance deaths (and also chance births) as suggested in the “community drift model” (Hubbell 1979), which assumes ecologically equivalence in species and theoretically predicts random processes can drive species coexistence in community dynamics.
The potential of grazer effects of generalists to generate unevenly distributed species-specific deaths of seedlings may be underestimated because their major diets are herbs, but not woody seedlings. However in reality, grazers can reduce exploitative competition from understory plants to seedlings (Hulme 1996), and thus indirectly benefit shade-tolerant seedling species. In addition, for resisting against grazers and browsers at an early stage of tree seedlings, some species-specific attributes help seedlings to survive after herbivorous damages. For example, tree seedling species with large seeds, hypogeal cotyledons (which will not be raised aboveground after germination), usually with greater initial height, are not easily eaten completely by herbivores. Most importantly, large-seeded seedlings are able to resprout from hypogeal cotyledons that facilitate establishment (Baraloto and Forget, 2007). In contrast, epigeal, small-seeded tree seedlings, usually light-demanding species, are easily become “vanishing deads” when suffering herbivory. Furthermore, shade-tolerant seedling species can benefit from neighboring understory plants or shrubs that provide nurse-plant effects, which may create shelter for smaller seedlings to grow underneath them and reduce ungulate herbivory (George and Bazzaz, 1999; Royo and Carson, 2008).
Species-specific deaths by ungulate herbivory can be the consequences through bio-interactions (e.g., nurse-plant effects) or species attributes (e.g., shade-tolerance, seed size) that help to resist ungulate herbivory. However, whether this species-specific effect on seedlings can be beneficial for rare seedling species remains unclear. For this argument, the well-proven mechanism of density-dependent deaths (Comita et al. 2014) is still worth rethinking even if considering that ungulates are non-specialists to tree seedlings.
Actually for insect-herbivores, a generalists-removal-experiment resulted in increased seedling species richness, as well as increased abundances of rare species (Dyer et al.
2010). For generalist enemies to seedlings, apparent competition further hinders species coexistence. Apparent competition means that prey species may indirectly depress each other under increment of shared natural enemies (Holt 1977); for tree seedlings, apparent competition is more likely to happen when sizeable proportion of recruitments of different species simultaneously burst in time (hereafter, synchronized recruitments). Under this circumstance, rare seedling species may be impacted by apparent competition resulted from nearby abundant seedling species that attract ungulates’ browsing. A possible mechanism that can prevent rare seedling species from apparent competition is a lottery model through recruitment fluctuations (Hulme 1996). Aperiodic recruitment fluctuations by long-lived tree species increase temporal heterogeneity, thus reduce interspecific competition, and facilitate chance survivors in a long-term process, so called “storage effects” (Usinowicz et al. 2012). To explore seedling species coexistence, the mechanisms that can effectively interact with ungulate herbivory from this long-term perspective should be addressed (Lischke and Löffler, 2006; Usinowicz et al., 2012;
Green et al., 2014; Paine et al., 2016).
In this study, our objective is to explore which factors are important at maintaining a diversifying seedling bank in a forest with frequent herbivory by ungulates. We expected
that intense ungulate herbivory would greatly affect seedling survival and growth; as such, species coexistence in seedling bank is likely mediated by ungulate herbivory. We hypothesize that ungulate herbivory can cause unevenly distributed species-specific deaths, and thus change recruitment dynamics and subsequent community assembly. To examine our expectation, we conducted an experiment with fence control to determine whether most of the herbivorous damages were caused by ungulates. We monitored recruitment dynamics for three years and recorded each seedling’s surviving state and causes of death. We evaluated “parameters” concerning recruitment and death of each seedling species based on our field surveys, and used a dynamic model to simulate community assembly “reconstructing” from random shuffling bursts of recruitment and death events. By controlling intensity of herbivory pressure and incorporating bio-processes such as density-dependent effects and recruitment fluctuations, our dynamic model is designated to explore possible mechanisms that ungulate herbivory affects seedling species coexistence in a long-term regeneration process.