Earthworm impacts on soil organic matter and fertilizer dynamics in tropical hillside agroecosystems of Honduras

Earthworms are important processors of soil organic matter (SOM) and nutrient turnover in terrestrial ecosystems. In agroecosystems, they are often seen as beneficial organisms to crop growth and are actively promoted by farmers and extension agents, yet their contribution to agroecosystem services is uncertain and depends largely on management. The Quesungual slash-and-mulch agroforestry system (QSMAS) of western Honduras has been proposed as a viable alternative to traditional slash-and-burn (SB) practices and has been shown to increase earthworm populations, yet the effect of earthworms on soil fertility and SOM in QSMAS is poorly understood. This study examined the role of Pontoscolex corethrurus in QSMAS by comparing their influence on aggregate-associated SOM and fertilizer dynamics with their effects under SB and secondary forest in a replicated field trial. Both the fertilized QSMAS and SB treatments had plots receiving additions of inorganic 15N and P, as well as plots with no inorganic N additions. Earthworm populations were manipulated in field microcosms at the beginning of the rainy season within each management treatment via additions of P. corethrurus or complete removal of existing earthworm populations. Microcosms were destructively sampled at harvest of Zea mays and soils were wet-sieved (using 53, 250 and 2000 μm mesh sizes) to isolate different aggregate size fractions, which were analyzed for total C, N and 15N. The effects of management system were smaller than expected, likely due to disturbance associated with the microcosm installation. Contrary to our hypothesis that earthworms would stabilize organic matter in soil aggregates, P. corethrurus decreased total soil C by 3% in the surface layer (0–15 cm), predominantly through a decrease in the C concentration of macroaggregates (>250 μm) and a corresponding depletion of C in coarse particulate organic matter occluded within macroaggregates. Earthworms also decreased bulk density by over 4%, but had no effect on aggregate size distribution. Within the two fertilized treatments, the QSMAS appeared to retain slightly more fertilizer derived N in smaller aggregate fractions (<250 μm) than did SB, while earthworms greatly reduced the recovery of fertilizer N (34% decrease) in both systems. Although management system did not appear to influence the impact of P. corethrurus on SOM or nutrient dynamics, we suggest the lack of differences may be due to artificially low inputs of fresh residue C to microcosms within all management treatments. Our findings highlight the potential for P. corethrurus to have deleterious impacts on soil C and fertilizer N dynamics, and emphasize the need to fully consider the activities of soil fauna when evaluating agroecosystem management options.