Wednesday, August 5, 2009 - 2:50 PM

COS 78-5: Growth and development of seedlings planted in plantations and tree islands to facilitate tropical forest recovery

Rakan A. Zahawi, Organization for Tropical Studies, Karen D. Holl, University of California, Santa Cruz, Rebecca J. Cole, University of California, Rebecca Ostertag, University of Hawaii at Hilo, and Susan Cordell, USDA Forest Service.

Background/Question/Methods

Planting tree seedlings is the predominant restoration strategy to facilitate tropical forest recovery after agricultural use.  Planting seedlings in patches or “islands” rather than plantations simulates the nucleation process of succession and is less costly, but the biological effectiveness of this strategy has rarely been tested.  To evaluate these two restoration strategies and compare growth rates over a large spatial scale, we compared tree seedling survival and growth over 3-yr in plantations and tree islands at 12 sites spread across 100 km2 in southern Costa Rica.  In 2004 and 2005 we planted four tree species (Terminalia amazonia [Combretaceae], Vochysia guatemalensis [Vochysiaceae], Erythrina poeppigiana, and Inga edulis [both N-fixing Fabaceae]) in two mixed-species treatments using the same planting density but varying the area planted: plantation (full 50 × 50 m area planted) and island (six patches of three different sizes planted within a 50 × 50 m area). 
Results/Conclusions

Overall seedling survival was 81% after 3-yr; survival varied strongly among sites (66-94%) but did not differ by treatment.  Height varied considerably by site (range 1.3-4.4 m after 3-yr averaged across species) and surprisingly was higher in plantations than in islands at seven of twelve sites (four were equivalent and one had the opposite result).  This result appears to be both an artifact of the restoration design – there are more N-fixing trees in plantations, which may enhance growth of neighboring trees – and that islands have more ‘edge’ seedlings where 3-yr height was ~0.25 m less than in plantation interiors.  Moreover, island seedlings appear more susceptible to accidental cutting by machete during periodic clearing of ruderal vegetation around seedlings; the irregular planting layout of islands made seedlings harder to find in tall grass.  Because sites had strong differences in bulk density and soil nutrients, we expected that these factors would affect seedling growth rates.  However, we found no consistent relationship between soil properties and seedling growth rates, although tissue N concentrations in the two hardwood species (Terminalia and Vochysia) were higher in sites where the two N-fixing species had higher growth rates.  Results suggest that there are tradeoffs to planting in islands as compared to more resource intensive plantations.  We also emphasize the degree of heterogeneity found among sites in terms of physical properties and plant responses, and the importance of broad spatial replication in studies to account for high variability and make generalizable restoration recommendations.