, 1998). Given the importance of SC to the economy and food security of BN-extractive communities, as well as the species’ light-gap dependence and its ability to resprout from consecutive slash-and-burn events, we evaluated whether the high BN regeneration density observed in fallows near nut-producing areas could be explained by the (i) number of SC cycles, (ii) past agricultural use, (iii) resprouting capability, and (iv) distance to parent trees. Finally, we asked if the spontaneous enrichment of fallows

influences landholders’ decisions to protect them from further conversion into crop or pasture sites. The study took place in the Reserva Extrativista do Rio Cajari, Amapá, Eastern Amazon, Brazil. The region contains a dense and open submontane rainforest with an Am Köppen climate (Peel et al., 2007). The annual average temperature is 25 °C with 2300 mm of average Apoptosis antagonist rainfall concentrated between December and June (Souza and Cunha, 2010). The relief is very hilly, and the predominant soil type is deep oxisols of Tertiary origin Lonafarnib (RADAMBRASIL, 1974). Our fieldwork was conducted from June to December, 2008, in the vicinity of two communities, Martins (52°17′30″W; 0°34′36″S) and Marinho (52°13′25″W; 0°34′40″S), both with a long BN-extractive tradition.

These settlements followed the 19th- and 20th-century rubber-tapper migrations (tappers of Hevea brasiliensis). Following the decline in latex prices, these communities have subsisted chiefly on BN extraction and small-scale agriculture. For local dwellers, SC is more than a complementary activity to the seasonality of BN production. In those years when the market prices offered for the nuts do not even pay the costs of harvesting, agriculture guarantees a minimum income and food security. Currently, the landscape surrounding the two villages

is a mosaic of mature forest with or without BN trees, active crops, pastures, and secondary forests in multiple seral stages. For the purposes of our study, BN regeneration refers to the individuals (seeders and resprouts) that we found colonizing agricultural sites following disturbances by cultivations. others We related the BN regeneration density to a series of seven biotic and abiotic environmental variables measured at 40 sites with known agricultural past use and established near parent BN trees. For each site, we interviewed the responsible landholder about (1) past agricultural use and (2) the number of cultivation cycles, which were later confirmed by remote sensing techniques. We also recorded (3) current agricultural use, (4) fallow age, (5) site area, (6) distance to the nearest parent trees, and (7) landholder’s decisions to preserve BN enriched fallows.