Agricultural land-use diversity and forest regeneration potential in human- modified tropical landscapes

A major challenge in tropical human-modified landscapes (HMLs) is meeting the ever-growing demand for agricultural products while conserving biodiversity and ecosystem services provided by forest ecosystems. Within this challenge, a major issue is the understanding of the forest potential to naturally regenerate in abandoned agricultural fields. To assess such potential, it is necessary to know the diversity of agricultural land uses in the landscape, quantify the ecological disturbance inflicted by such uses, and evaluate forest regeneration as a function of disturbance both at the field and landscape levels. Our previous work has shown that in abandoned fields the abundance and species diversity of regenerating rain forest trees decline as disturbance level increases. Here we aimed to achieve the following: 1) to quantify the diversity of agricultural land uses in HMLs; 2) to assess ecological disturbance regimes caused by different agricultural land uses, at the field and landscape scales; and 3) to identify groups of agricultural land uses with contrasting effects for forest regeneration at the landscape level. We approach these issues by using a case study of HMLs in a southeastern region of Mexico, which are representative of landscapes in the agricultural frontier in the Neotropics. We interviewed 68 landowners to gather information on agricultural land uses and management of 156 fields. Based on this information, we quantified an ecological disturbance regime associated with each field considering the following: field size (in hectares), duration of agricultural use (in years), and land-use disturbance severity (i.e. frequency or magnitude of fire, agrochemicals, machinery, grazing or removal of tree cover). By integrating disturbance inflicted by different land uses and the proportion of the landscape covered by each land use, we constructed a landscape ecological disturbance index. Finally, by using this index and data gathered from nine landscapes (3 × 3 km each), we tested the hypothesis that structural attributes (abundance, biomass, and species diversity of trees) of regenerating forests decrease as agriculture disturbances increase in the landscape. There was a high inequality in the proportion of land allocated to the 13 recorded agricultural land uses, with cattle pastures representing ca. 90% of total agricultural land. There was a wide disturbance gradient, ranging from land uses with high (e.g. cattle pastures) to low disturbance (e.g. coffee and cocoa plantations). Three major groups of land uses with contrasting disturbance regimes were detected: 1) agroforestry systems, characterized by small size, low to intermediate duration, and low disturbance severity; 2) monocultures, typically small size, long duration, and medium to high disturbance severity; and 3) extensive farming, large size, short to intermediate duration, and high disturbance severity. Biomass and species diversity of regenerating forests consistently reduced with increasing levels of agriculture disturbance in the landscape. We conclude that positive balances between biodiversity conservation and agricultural production in HMLs will depend on establishing agricultural land uses that inflict low disturbance regimes (such as agroforestry systems) embedded in a matrix of old-growth forest and long-lasting second-growth forests. Our results may inform farmers, policy makers and land managers about HMLs where agricultural production and conservation of biodiversity and ecosystem services can be conciliated.