Barriers to tropical rainforest restoration: Tree seedling regeneration in a degraded tropical pasture

Background/Question/Methods

The current scale of deforestation and continuing decline of tropical forests underscores an important need to restore degraded tropical landscapes. Research here focuses on highly degraded habitats, such as pastures, where forest regeneration may be impeded or no longer possible. The study site, a former cattle-ranching pasture, borders the tropical rainforest of Los Tuxtlas in Veracruz, Mexico. To determine tree recruitment success within pasture habitat, seedling experimental exclosures were built to mitigate herbaceous vegetation and abundant pasture rodent fauna. Exclosures evaluating seed removal by ants and rodents were also built in an adjacent pasture. The fully crossed two-way factorial seedling exclosures tested 1) the degree that herbaceous vegetation inhibited or facilitated recruitment of tree species, 2) seedling mortality by rodent fauna, and 3) the prevalence of life-history traits that correlate with seedling establishment success. Adjacent seed exclosures tested 1) the rate of seed removal by ant, rodents or both on seeds of various seed masses and 2) the effect that forest proximity had on seed removal rates.

Results/Conclusions

Two separate tree species cohorts were evaluated within experimental exclosures, each for a duration of one year. After two years, half of 26 sowed tree species failed to establish within seedling experimental exclosures. In the first experimental year, the presence of grass significantly diminished the recruitment potential of establishing seedlings (p = 0.004). While access to rodents also significantly reduced species seedling establishment (p = 0.018). In the second experimental year, a balanced representation of regional seedling functional types allowed for the assessment of their recruitment potential within the pasture site. Different seedling functional types were found to significantly differ in their responses to grass and rodent treatments revealing a 3-way interaction (p = 0.003). In seed experimental exclosures, small seeds were removed at faster rates compared to large seeds by ants and rodents. (p << 0.005). Large seeds removal by rodents was species specific and not determined by seed mass.Further work will address between and within species physiological variation in a context of spatial distribution within seedling exclosure plots.