We restrict our review to the tropics, where devising appropriate interventions to manage trees and tree genetic resources is important to meet international development goals of poverty alleviation and community

resilience (FAO, 2010 and Garrity, 2004). We NVP-BGJ398 mw also restrict our consideration to three production categories: non-timber forest product (NTFP) harvesting (from natural, incipiently- and/or semi-domesticated forests and woodlands); agroforestry tree products (AFTPs) and services (provided by a wide range of mostly semi-domesticated local and exotic trees in smallholder-farm landscapes); and woody perennial commodity crops (which are often completely domesticated, exotic in major production centres, and grown in both smallholdings and larger plantations, though our concern here is only with the former). The boundaries between these production categories are not always easy to define, as evidenced, for example, by often subtle transitions in landscapes between forests this website and agroforests in a gradient of transformation and intensification (Balée, 2013, Michon, 2005 and Wiersum, 1997). In fact, one category often depends upon another for supporting sustainability, as, for example, many AFTPs

and tree commodity crops were once NTFPs, and often also still are (thus, the continued improvement of AFTP and tree commodity crop production may depend to a greater or lesser degree on accessing genetic resources GABA Receptor maintained in natural stands; Hein and Gatzweiler, 2006, Mohan Jain and Priyadarshan, 2009 and Simons and Leakey, 2004). Our three production categories have received considerable attention for their roles in meeting development targets for small-scale harvesters and smallholder farmers in the tropics, both of which groups are the subject of our attention here (Belcher et al., 2005, Garrity, 2004 and Millard, 2011). Our categories are, however, not fully exhaustive of the benefits received by tropical rural communities from trees, as we do not, for example, consider the value of commercial

forest timber harvesting by local people (e.g., Menton et al., 2009). Nonetheless, the division into our three categories provides a useful way to structure the different benefits of trees to communities, to illustrate the issues faced in describing value and to determine appropriate interventions for improved management. Considering these different categories also demonstrates the importance of taking a wide view in determining where best to intervene for maximum impacts on livelihoods, for example, in minimising unintended consequences due to potentially negative interactions between different production systems (the same attention to interactions is important when promoting appropriate tree conservation interventions among a range of options, see Dawson et al., 2013).