Belowground impacts of regenerative agriculture in a South African seed production system

Nel, Gemma

Belowground impacts of regenerative agriculture in a South African seed production system

Thesis / Dissertation

2023

Abstract

Intensive agriculture, that is environmentally detrimental, is a serious global problem. Widespread deforestation as a means to expand agricultural areas, and the overuse of agrochemicals causes rapid biodiversity loss, large scale soil erosion and agroecosystems to become stuck in negative feedback cycles. Agricultural management thus plays an important role in the stability and resilience of agroecosystems. Shifts towards more sustainable practices have been encouraged, with regenerative agriculture gaining a lot of traction over the years. It is crucial to understand the impact of various management strategies on various aspects of agroecosystems to gain a greater, more holistic understanding. A newly established long-term trial provided an ideal opportunity to explore and compare the impacts of regenerative agriculture with conventional agriculture within a cauliflower seed production system in the Western Cape of South Africa. Specifically, this study investigated the effects of different management strategies on soil fauna and the important soil processes they govern. The treatments included “conventional agriculture” (C) which was the use of agrochemicals and tillage, “mulching combined with compost” (MC), “cover cropping” (CC), and crop rotation. To determine the changes in soil fauna community assemblages and species richness, two components of the belowground fauna were investigated. Firstly, arthropod diversity indices were compared between conventional agriculture and regenerative agriculture management strategies (Chapter 2). Four arthropod taxa were selected for study, providing information on the macrofauna and mesofauna of the area. These included ants, beetles, spiders, and springtails. Soil fauna were sampled every two months for a year, using pitfall traps. Differences in species richness and abundance of the four study taxa were assessed in relation to agricultural treatment, sites, crop lines, and sampling times. ANOSIM (Analysis of Similarities) was conducted to determine sample differences between the study factors of site, treatment, crop rotation and sampling time. The presence of particular species was assessed for associations between factors, to determine whether any species show influential characteristics towards particular agricultural treatments, sites or seasons. It was found that total arthropod abundance and species richness was generally greatest during the pre-trial sampling event that took place in February 2021. Arthropod abundance and species richness responses were taxa specific and differed between field sites and sampling times. Arthropod communities within each site, treatment, crop line and season, were not more similar to the communities of different sites, treatments, crop lines and seasons. Several species of interest were determined based on their presence within the study factors. Particularly, the spider Pardosa crassipalpis, displayed agrobiont behaviour within this agroecosystem and will be a key species of interest for further monitoring within the trial. The presence of this highly abundant species was particularly noted at the site Jondu, within mulching and compost treatments, and typically during the final (February 2022) sampling period. Furthermore, the abundance of species of springtails recorded within this agroecosystem is remarkable. Native Entomobryidae made up more than 50% of all abundances and more than half of all species recorded. These species fall within a group of collembolans that display remarkably high thermal tolerances and desiccation resistance, within such an arid agroecosystem. Secondly, decomposition rate (k), remaining litter mass (g), carbon to nutrient ratios and soil mesofauna community compositions were compared between agricultural treatments, by means of a litter bag experiment (Chapter 3). Multivariate analyses based on decomposition rate with respect to temporal changes showed significant site, treatment and temporal effects on decomposition rate. On average, decomposition rates were highest for litter bags placed at the site Jondu. Conventional agricultural treatment showed significantly lower decomposition rates and mass loss than the other two treatments. Site and sampling times were the most important factors explaining the variation in carbon to nutrient ratios. Carbon to nutrient ratios showed overall decreases over time, with particular site-specific responses observed. Particularly interestingly, C/N ratios showed considerable site-specific responses, with litterbags at Jondu recording decreased C/N over time, compared to increased C/N at the other sites. This study provided essential baseline information for the system. Treatment effects were not as evident as expected. However, due to the study being conducted within the first established year of the trial, this may be expected. Site effects were far more prominent than treatment or crop rotation effects. Differences in soil conditions between sites may be accounting for a great deal of the variation in arthropod community assemblages and litter decomposition. Continued assessment of the impact of agricultural management strategies and crop rotation as the trial progresses and becomes more established will prove useful in determining effects of regenerative agriculture on soil fauna and litter decomposition.

Keywords

Biological Sciences

Reference:

Collections

Masters

Full item page