Tuesday, August 4, 2009 - 4:20 PM

COS 43-9: Experimental restoration of tropical connectivity: context and preconditions

Henry F. Howe1, Yuliana Urincho-Pantaleon2, Marines de la Peña-Domene3, and Cristina Martinez- Garza2. (1) University of Illinois-Chicago, (2) Universidad Autónoma del Estado de Morelos, (3) Universidad Nacional Autónoma de México

Background/Question/Methods

Ecological restoration re-establishes some native species and encourages immigration of others from surrounding landscape. We ask what landscape context and conditions pre-exist for an experimental restoration near Los Tuxtlas, Veracruz, Mexico. Twenty-four 30x30 m plots were fenced in active pasture in 2006 and planted with 10 animal-dispersed tree species (8 plots), 10 wind-dispersed tree species (8 plots) or left as unplanted controls (8 plots) in 2006-7 as a test of dispersal limitation from nearby oldgrowth forest and isolated pasture trees. Reported here are seed rain from the landscape and seedling emergence from seed rain and pre-existing seed banks through June 2007, before appreciable growth or any reproduction of experimental plantings could directly add recruits or attract dispersal agents that might transport them.

Results/Conclusions

Early results indicate different patterns of seed fall and seedling emergence of trees dispersed by animals and wind.  Multiple regression of dispersal category against plot distance from old growth forest (300 – 1090 m) and pasture trees near the experimental grid (20-650 m) shows that seed rain and seedling establishment of ubiquitous wind-dispersed pioneers are independent of both distance factors, while seed rain and seedling establishment of old growth trees dispersed by wind decline with distance from old growth forest (p < 0.005 and p < 0.025).  Seed rain of animal-dispersed pioneers more closely reflects distance from pasture trees (p < 0.05) than distance from forest.  Sampled seed rain of later-successional trees dispersed by birds, bats, and ground mammals is not strongly associated with either distance measure (p > 0.1), but seedling emergence of these species is more closely associated with proximity to pasture trees just outside the grid than distance from old growth forest (p < 0.02), underscoring the multimodal nature of seed-dispersal effects by animals.  In all multiple regressions, Pillai’s trace is significant (p < 0.05). Seedling analyses exclude (Bursera simarouba) near sites where fruiting adults had been cleared before planting.  This species is absent from the current seed rain, but seedling emergence where fruiting individuals were removed represents a pre-condition from dispersal history.

Three-dimensional graphs of seed fall and emergence illustrate the interplay of ongoing dispersal processes and contextual preconditions. These are likely to be overwhelmed by experimental plantings, now beginning to seed, that will disperse their own seeds and in replicates of animal-dispersed stands induce seed rain and emergence of forest species carried to the plots from the surrounding landscape.