A twin-plot approach to determine nutrient limitation and potential productivity in Eucalyptus plantations at landscape scales in Brazil

摘要

Many forest experiments test for nutrient limitation in replicated fertilization trials at one or several locations, providing a poor basis for extrapolation across large landscapes. We present an experiment that paired control and treated plots at 127 locations (with similar initial wood biomass, "twin-plot"), providing no experimental replication within sites, but strong replication within the population of stands across the landscape. This twin-plot approach, with 6-month period re-measurements, showed that forest productivity increased from 19.6 to 24.4 Mg ha~(-1) yr~(-1), representing an average response to heavy application of fertilization of 4.8 Mg ha~(-1) yr~(-1) (10 m~3 ha~(-1) yr~(-1)) for 2 years. The response in stands derived from clonal plantlets and seedlings was the same, although the fertilized clonal stands were 48% more productive than stands derived from seedlings. Older stands of both types were more responsive than younger stands, with a 0.6 Mg ha~(-1) yr~(-1) increase in growth for each additional year of age. Coarse-textured Entisols showed twice the response that developed in more-productive, fine-textured Entisols. The growth response was much smaller during the dry season (4.5 Mg ha~(-1) yr~(-1)) than in the rainy season (9.4 Mg ha~(-1) yr~(-1)). Analysis of growth by treatment and season allowed us to estimate the potential productivity of the sites, which averaged 29.1 Mg ha~(-1) yr~(-1) (62 m~3 ha~(-1) yr~(-1)). The twin-plot design is relative simple and inexpensive, and can be used in conjunction with routine inventory designs to guide decisions about overall stand silviculture.