Socio-ecological analysis of multiple-use forest management in the Bolivian Amazon

Community families throughout tropical regions derive an important share of their income from multiple forest products, with generally positive outcomes on their livelihoods. The production of these products in a multiple-use forest management scheme (MFM, the production of multiple forest products within a single management unit) encompasses many (yet) unknown socioeconomic and ecological feedbacks. In particular, MFM entailing timber and non-timber production may be affecting the future availability of valuable timber and non-timber tree species due to the extraction of vital plant components, which may have undesired outcomes on the income that community families derive from forests. In this thesis, I evaluated the social, economic, and ecological viability of an important MFM scheme widely practiced by community households in the Bolivian Amazon: the production of Amazon or Brazil nut (<em>Bertholletia excelsa</em>) and timber from other tree species. Data was obtained from a two-year (2014 and 2015) survey questionnaires of 24 community households in six <em>campesino</em> communities with community forest management plans (CFMPs) and from ecological surveys of 72 2 ha permanent research transects (three transects per household forest) harvested at varying Amazon nut and logging intensities. A CFMP entails the planning and execution of logging activities in compliance with formal rules intended to secure the long-term provision of timber at community-owned forest. Household-level decisions to harvest Amazon nut and to log timber allowed us to account for household forest as our sampling unit. We used multi-model inference and structural equation modelling techniques to determine the impact of socio-ecological factors on the income that community families derived from Amazon nut and timber (chapter 2), and regression and matrix modelling techniques to determine the impact of Amazon nut harvest and logging intensity on <em>Bertholletia </em>(chapter 3) and commercial timber species (chapter 4). In general, we found that few socioeconomic and biophysical factors of community households, together with a general positive response of studied species to timber logging and customary silvicultural intervention, make the production of Amazon nut and timber production of other tree species viable in a MFM scheme. In chapter 2, we found that community households could reduce their dependency on forest resources by increasing income opportunities from other existing livelihood activities. Amazon nut represented the largest source of household income (44% of the total household net income); and off-farm (salary, business and gifts; 21%), husbandry (generally subsistence agriculture, animal rising, and agroforestry; 21%), and timber (9%) incomes were complementary to their livelihood. Increased skills and ecological knowledge of community households enhanced household income derived from forest products. For example, an increase in the number of management practices reduced the need for timber income by increasing Amazon nut production; decreasing further pressure on timber of other tree species. In chapter 3, logging intensity was found to increase <em>Bertholletia</em>’s seedlings and saplings growth rate, and liana cutting was found to increase Amazon nut production rate. Both, logging and liana cutting intensities played a key role on <em>Bertholletia</em> population growth rate. Increased logging and liana cutting intensities counteracted the negative impact of Amazon nut harvesting intensity on the number of new recruits (i.e., due to nut harvest), indicating a trade-off between logging, liana cutting and Amazon nut harvesting intensities. Considering the overall stem density of commercial timber species (chapter 4), we found that 17% of the species present at unlogged sites (3 species out of 17: <em>Swietenia macrophylla</em>, <em>Tabebuia impetiginosa</em> and <em>Terminalia</em> sp.) were not present at sites six years after logging; and a larger percentage (71%) of the species present at unlogged sites in the harvestable size (trees>minimum diameter cutting – MDC) were not present at sites six years after logging, e.g., <em>Cedrela</em> spp. Stem density and timber volume of five of the eight most abundant commercial timber species under study differed among community-owned forests, after accounting for the effects of logging intensity and time since logging as indicated by our best models; whereas, potentially harvestable and harvestable timber volume differed between communities for only two and three species, respectively. Best models indicated that logging intensity increased either stem density or timber volume of <em>Apuleia leiocarpa</em>, <em>Cedrela odorata</em>, <em>Dipteryx micrantha </em>and <em>Hymenaea parvifolia</em>, decreased potentially harvestable timber volume of <em>T. serratifolia</em>, and had no effect on the other three species investigated. We also investigated the impact of logging intensity on congeneric species given that lumping congeneric species for logging is a common simplification during forest inventories and censuses, and is accepted in CFMPs assuming that closely related species respond to timber logging in a similar way. However, logging intensity had a differentiated effect on congeneric species. Logging intensity favoured growth rate of <em>C. odorata</em> trees >10 cm DBH and had no effects on <em>Cedrela fissilis</em>. Regarding <em>Hymenaea</em> congeneric species, logging intensity favoured <em>H. parvifolia</em> survival of individuals <10 cm DBH, but decreased growth rates of <em>H. courbaril</em> trees >10 cm DBH. In conclusion, Amazon nut harvest and timber logging of other tree species are compatible under certain socioeconomic and biophysical conditions, and as long as commercial timber species differential response to harvesting are accounted for in managing these species in a MFM scheme. This compatibility is due to existing socioeconomic complementarity of both activities and to the positive impact of logging intensity levels as practiced in the region on Amazon nut production and on most commercial timber species. Community families’ better negotiation skills to obtain better prices for Amazon nut, and increased implementation of management practices to increase Amazon nut production (e.g., liana cutting) helped families to increase their income and also decrease pressure on timber. These results highlight the need to look at both socioeconomic and ecological aspects when assessing the long-term sustainability of MFM schemes. Results of this research have important implications for policy to support the sustainable development of community forestry in the Bolivian Amazon. The compatibility found between Amazon nut and timber production calls for the investigation of the compatibility of timber production with other valuable NTFPs commonly harvested by community families throughout the tropics. We argue that management needs to be done at species-specific level, rather than at the level of products or at the level of species groups. This may result prohibitively expensive for communities and smallholders. Thus, we urge governments and the international community to revalorize local ecological knowledge of community people to manage their forests, while supporting the development of technologies, such as the ones based on hyperspectral LiDAR technology, to develop tools that could help reduce management costs of tropical forests at the required level. Such policies need to be accompanied by capacity building programs on different management tasks and negotiation skills to enhance the income obtained from MFM schemes. The research approaches used here could be used in other contexts and scales involving natural resources management to get a better understanding of the systems.