Direct seeding of late-successional trees to restore tropical montane forest

Despite continued forest conversion and degradation, forest cover is increasing in countries across the globe. New forests are regenerating on former agricultural land, and forest plantations are being established for commercial and restoration purposes. Plantations and restored forests can improve ecosystem services and enhance biodiversity conservation, but will not match the composition and structure of the original forest cover. Approaches to restoring forest ecosystems depend strongly on levels of forest and soil degradation, residual vegetation, and desired restoration outcomes. Opportunities abound to combine ambitious forest restoration and regeneration goals with sustainable rural livelihoods and community participation. New forests will require adaptive management as dynamic, resilient systems that can withstand stresses of climate change, habitat fragmentation, and other anthropogenic effects.

Our meta-analysis of 126 nitrogen addition experiments evaluated nitrogen (N) limitation of net primary production (NPP) in terrestrial ecosystems. We tested the hypothesis that N limitation is widespread among biomes and influenced by geography and climate. We used the response ratio (R approximately equal ANPP(N)/ANPP(ctrl)) of aboveground plant growth in fertilized to control plots and found that most ecosystems are nitrogen limited with an average 29% growth response to nitrogen (i.e., R = 1.29). The response ratio was significant within temperate forests (R = 1.19), tropical forests (R = 1.60), temperate grasslands (R = 1.53), tropical grasslands (R = 1.26), wetlands (R = 1.16), and tundra (R = 1.35), but not deserts. Eight tropical forest studies had been conducted on very young volcanic soils in Hawaii, and this subgroup was strongly N limited (R = 2.13), which resulted in a negative correlation between forest R and latitude. The degree of N limitation in the remainder of the tropical forest studies (R = 1.20) was comparable to that of temperate forests, and when the young Hawaiian subgroup was excluded, forest R did not vary with latitude. Grassland response increased with latitude, but was independent of temperature and precipitation. These results suggest that the global N and C cycles interact strongly and that geography can mediate ecosystem response to N within certain biome types.