Avocado cultivation is displacing traditional coffee production in Veracruz, Mexico. This change modifies the floristic composition and agronomic management (AM), both of which are affected by socioeconomic factors and generate variation in edaphic conditions. The objective of this research was to identify the socioeconomic variables that modify the characteristics of the AM of avocado and coffee systems and have repercussions on the carbon and nitrogen pools. Four case studies were analyzed: a renovated coffee polyculture (SRC), severely cleaned coffee polyculture (IPC), avocado–coffee system (ACS), and conventional avocado orchard (CAO). A socioeconomic analysis was carried out with a 30-year economic projection and 56 variables. Three plots were installed where vegetal biomass carbon (VBC) was evaluated; soil samples were taken in theses plots at three depths (in triplicate) to determine soil organic carbon (SOC) and total soil nitrogen (TSN) in each case. A principal component analysis (PCA) was performed. Four economic variables explain 50.9% of the variation in agronomic management and five social variables 50.2%, in both cases where the component is considered. In all systems, the variation in VBC is explained by the agronomic management, while the reported SOC presents an inverse relationship with AM. The ACS presents the higher values in total C in the system, and SOC and TSN pools, while the CAO shows negative impacts concerning soil organic matter quality. The ACS is sensitive to investment, market diversity, education, and external organic residue addition, explaining the variations in AM and the TSN pools.

The production of resin from Bursera bipinnata, a species native to the tropical deciduous forest (TDF), known as copal, represents a crucial source of income for rural Mexican communities. This study evaluated associations between ecofunctional gradients and resin yield using principal component analysis (PCA), hierarchical cluster analysis, and a multiple linear model. Ecological, edaphic, dendrometric, and management variables were evaluated in 27 trees established in agroforestry systems (AFS) in the municipality of Tepalcingo, Morelos, Mexico. Horn’s parallel analysis supported the retention of three principal components, which explained 63.66% of the total variance. The multiple linear model was significant $\left(F\right(\mathrm{5,21})=3.23, p=0.0256; R^2=0.435; adjusted R^2=0.300)$. PC1 was positively associated with log-transformed resin yield $(\beta =0.400, p=0.0245)$ and was characterized by negative loadings for total tree height (TH), total fungi (TF), soil organic matter (OM), actinobacteria (AC), available phosphorus (P), and total nitrogen (TOTN), and a positive loading for elevation (ALT). In contrast, PC3 was negatively associated with log-transformed resin yield $(\beta =-0.408, p=0.0269)$ and was characterized by negative loadings for soil bulk density (BD) and magnesium (MG), and a positive loading for soil moisture content (MC). Trees from site $S_2$ also showed lower log-transformed resin yield than those from the reference site $S_1$ $(\beta =-1.786, p=0.0067)$, whereas no significant difference was detected between $S_3$ and $S_1$. Although the clusters differed in their observed mean resin yield, these differences were not statistically significant. Overall, the findings indicate exploratory associations between resin yield and multivariate gradients integrating ecological, edaphic, dendrometric, and management attributes within the sampled trees and study sites.