Changes in fire regimes can drastically alter local plant communities. In the upland forests of Hawaiʻi, increased dominance of non-native grasses has increased fire frequency, leading to native tree mortality. Work to reestablish forest plant communities in recently burned areas has been stymied by limited understanding of what factors constrain seedling success. We conducted field and greenhouse studies to address how populations of two native, nitrogen-fixing trees Koa (Acacia koa) and Māmane (Sophora chrysophylla) establish in recently burned areas. We asked three questions: (Q1) Which factors are most important in determining seedling success in field conditions? (Q2) Does a recent history of fire alter seedling growth rates, in response to the soil microbial community? (Q3) Does the soil microbial community around a germinating seedling alter that seedling’s growth rate? We sampled field plots that were seeded with Māmane and quantified native seedling emergence and a suite of factors that may affect seedling success (including non-native grass cover) in habitats previously dominated by Māmane, Koa, and/or a native non-nitrogen-fixing shrub ʻAʻaliʻi (Dodonaea viscosa). Our greenhouse project examined both Koa and Māmane seedling growth rates in sterile soil inoculated with field soil from Koa, Māmane, and ʻAʻaliʻi, sourced from inside and outside the recent fire. Inoculation allowed us to examine the microbial community without other soil-based confounding factors. We used structural equation modelling (SEM) to disentangle the effects on seedling success of biotic (Q1-3) and abiotic variables (Q1).
Results/Conclusions
(Q1) Increased exotic grass cover led to fewer seedlings of our focal native plant species (p=0.02). (Q2) Fire does not directly affect soil microbial communities in a way that impacts plant growth rates (p>0.05). (Q3) The soil in which a seedling germinates does impact its growth rates (p<0.01). For both nitrogen-fixing species, growth rates were higher in soil inoculated from conspecifics than either heterospecific soil source. This response was mediated by the number of nodules that each plant produced (p<0.001). We conclude that both our focal species support their own regeneration, through increasing nodule formation when grown near conspecifics; this may lead to reinforcement of species dominance where these trees occur. Additionally, fire does not alter soil microbial communities in a way that changes these patterns. Finally, reducing the amount of grass cover will likely allow increased survival of native seedling species if that is a management goal.