Earthworm densities have been regarded as reliable indicators of soil
health. A long-term field experiment was conducted in two sites (15% and 25%
slope) to compare the effects of manual and chemical weeding (using paraquat
and glyphosate) and determine other factors that may affect earthworm
populations in a banana plantation. Based on four years of field observation, no
significant difference in earthworm count between manual and chemical plots
(15% slope: F-ratio: 0.96, p = 0.43; 25% slope: F-ratio: 14.18, p = 0.06) were
observed. The earthworm species composition was found to differ between the
two sites. The 15% site tends to have a higher earthworm population compared
to the 25%-slope site, likely because of the former’s higher soil organic matter
content. Earthworm populations were on a declining trend in both treatments
for both sites, but regression analyses show these trends to be insignificant.
Rainfall, organic mulch, and weed cover were not significantly correlated with
the earthworm counts. However, the declining pH in both sites could help
explain the decline in earthworm populations. Pontocolex corethrurus showed
significant avoidance response to normal glyphosate concentrations (8.055 ×
10–3 mL per 350 g soil) (p = 0.03), but not to paraquat (1.5 × 10–3 g per 350 g
soil) (p = 0.55). Experiments suggest that both weed management treatments
do not pose a significant threat to earthworms under the conditions studied.
The negative effect of declining pH needs further study.