Carbon storage capacity in coffee and secondary forest
Carbon stocks varied according to the type of vegetation cover (Table 3). The total carbon stock in the different vegetation cover types ranged from 31.1 t C/ha in coffee without shade trees to 132.2 t C/ha in the secondary forest (Table 3). The overall mean of carbon stocks in the different types of vegetation cover was 89.5 t C/ha. The highest carbon content was found in the secondary forest and coffee with Inga sp. shade trees with a total of 132.2 and 118.2 t/ha (p < 0.05), respectively.
Aboveground carbon stocks varied from 1.6 t C/ha in coffee without shade trees to 75.8 t/ha in the secondary forest (Table 3). The total carbon storage capacity (t C/ha) in the different cover types decreased in the following order: aboveground carbon (AGC) > soil organic carbon (SOC) > belowground carbon (BGC) > litter carbon (LC) > herbaceous carbon (HC). Vegetation cover influenced carbon stocks in the different components with the exception of SOC (p < 0.05). The AGC compartment had the highest C content in all the sampled cover types. The secondary forest and the coffee system with Inga sp. shade trees had the highest AGC (75.8 and 55.8 t C/ha) and BGC (10.7 t C/ha) carbon stocks (Table 3). Likewise, coffee without shade trees had the lowest LC and SOC carbon stocks, with 0.2 and 0.6 t C/ha, respectively. With respect to SOC, we found values ranging from 28 t C/ha in coffee without shade trees to 45.7 t C/ha in the coffee with Inga sp. shade trees (Table 3). Likewise, SOC showed a higher coefficient of variation (CV = 43.9%), followed by LC (CV = 26.3%), AGC (24.6%), BGC (23.8%), and HC (19.1%), indicating a greater spatial variability of carbon stocks in soils than in vegetation.
The proportions of carbon stored in the different vegetation and soil compartments were affected by the different vegetation cover types (p < 0.05). The secondary forest sequestered 57.4% of carbon as AGC, 8.1% as BGC, 27.0% as SOC, and the rest as LC and HC (Fig. 1). The coffee with Inga sp. shade trees sequestered 46.7% of carbon as AGC, 39.2% as SOC, and 14.1% as LC, HC, and BGC. Coffee with polyculture farming sequestered 54.2% of carbon as SOC, 27.7% as AGC, 8.3% as LC, and 9.8% as HC and BGC; whereas the coffee without shade trees sequestered 89.4% of carbon as SOC, followed by AGC, HC, BGC, and LC with 5.5%, 2.5%, 2.0%, and 0.6%, respectively (Fig. 2).
Glomalin and spore content of arbuscular mycorrhizal fungi (AMF)
The lowest number of AMF spores was found in the secondary forest with 102 spores on average, showing significant differences (Fig. 2). The highest GRSP content was found in coffee without shade trees with an average of 18.5 mg/g, and the lowest content was found in the secondary forest with 7.1 mg/g on average, showing significant differences for both covers (p < 0.05). Whereas coffee cover with Inga sp. shade trees and coffee with polyculture farming did not show significant differences for GRSP content among them (Fig. 3).
Carbon sequestration
Coffee plants also contributed to the total carbon stock in the different types of vegetation cover. The carbon sequestered by coffee plants was 2.97, 2.82, and 2.17 t C/ ha in coffee with Inga sp. shade trees, coffee with polyculture farming, and coffee without shade trees, respectively (Table 4). Coffee plants contributed about 7.46%, 3.89%, and 2.54% of the carbon sequestered in coffee without shade trees, coffee with polyculture farming, and coffee with Inga sp. shade trees, respectively (Table 4). The overall average carbon sequestered by coffee plants in the different types of vegetation cover was 2.65 t C/ha (average of 4.63% of total sequestered carbon). The mean values of carbon sequestered by coffee plants with Inga sp. shade trees and polycultures were significantly different from those of coffee without shade trees, but did not differ significantly from each other; while the secondary forest was not taken into account due to the absence of coffee plants (Table 4).
In the study area, coffee agroforestry systems sequestered a large amount of carbon in vegetation, including coffee plants and soils proportional to the area of each stratum, approximately 118.20, 76.48, and 2.17 t C/ha were stored in coffee cover with Inga trees, polyculture, and coffee without shade trees, respectively (Table 5). This shows that a total of 196.85 t C/ha was stored in all coffee agroforestry systems in the study area. Particularly, coffee plants added a total of 7.96 t C/ha in the coffee agroforestry systems (Table 5). All coffee agroforestry systems in the current study area trapped 722.44 t of CO2 from the atmosphere, while coffee plants captured 29.23 t of CO2 from the atmosphere and stored it as carbon in the agroforestry systems.
GRSP and mycorrhizae-mediated carbon
The variation of carbon stocks, number of AMF spores, and GRSP content in the different vegetation covers was explained by 75.9% in the first two principal components of a PCA (Fig. 4). Coffee without shade trees was characterized by the highest values of GRSP and number of AMF spores, variables that were positively associated (Fig. 3). Likewise, the secondary forest was characterized by the highest values for AGC, BGC, and LC, variables that showed positive associations. That is, carbon in aboveground biomass increases as underground biomass and litter biomass increase in the secondary forest cover.
The variable GRSP ranged between 6.22 and 24.36 (mg/g) and the number of AMF spores between 99 and 145 units for 25 g of soils (Fig. 5), variables that presented significant and positive correlation with a value of 0.766. GRSP presented significant and negative correlation with AGC, LC, and BGC with values of -0.657, -0.703, and -0.591, respectively. AMF spore number only presented significant and negative correlation with litter carbon with a value of -0.788.