Redirection of the carbon cycle by households in a savanna socio ecological system in rural South Africa

Abstract

This study investigated the redirection of the carbon cycle by households in a savanna socio ecological system in rural South Africa. Through land-based activities such as resource harvesting and agriculture, rural households influence the carbon cycle by co-opting a proportion of annual plant biomass production, referred to as Human Appropriation of Net Primary Production (HANPP). Most HANPP data is at global, continental, or national scales, and there is currently a lack of data at the local level, particularly in rural African savanna settings. Such data are important for understanding context-specific human-environment dynamics, especially in significant and rapidly transforming socio-ecological systems, such as African savannas. This study addresses the above gap by 1) quantifying the mass of carbon redirected annually from the local ecosystem by rural households via key land-based livelihood activities, namely fuelwood harvesting and, livestock foraging, and requesting of carbon from the atmosphere through maize crop production, 2) exploring statistical relationships between household carbon redirection and selected household socio-economic and demographic characteristics, and 3) upscaling carbon redirection to village and landscape scales to assess carbon demand relative to carbon stocks in the local environment in a South African rural socio-ecological system. The study was conducted in the Agincourt Health and Socio-Demographic Surveillance System (AHDSS) study site in Bushbuckridge, South Africa, and used longitudinal household socio-demographic and livelihood data from the SUCSES (Sustainability on Communal Socio-Ecological Systems) household panel study, involving a cohort of 590 households that were surveyed annually from 2010 to 2014. The annual mass of fuelwood used, cattle and goats owned, and maize produced, were converted into gross and net mass of carbon redirected per household per livelihood activity per year, after accounting for carbon backflows to the environment from these activities, such as in emissions and animal dung. Trends in carbon redirection by households over a five-year period were assessed. Multiple logistic and linear regression models were used to determine the influence of household socio-economic and demographic characteristics on the likelihood of households engaging in specified land-based livelihood activities, and the mass of carbon redirected through these activities, respectively. Total household carbon demand from the local environment was then upscaled to the village level using two methods; the summed model method which used statistical models to predict carbon demand for each household in v each of nine villages which were then summed, and the multiplied mean method which simply multiplied the mean household level carbon values by the number of households per village. The multiplied mean method was used to calculate annual gross and net carbon demand and carbon backflows to nature for the livelihood activities at the landscape level (420 km2 and 20 villages) from 2010 – 2014. MODIS NPP (MODerate resolution Imaging Spectroradiometer Net Primary Production) data were used to calculate total available carbon within 1.5 km of each of the nine SUCSES villages and across the entire AHDSS landscape in order to quantify carbon redirected as a proportion of available carbon stocks. While fuelwood use was the most prevalent of the land-based activities studied, more carbon was redirected from the environment at a household level by cattle and goat ownership. The highest gross redirection of carbon at household level was by cattle, peaking at 10 972 kgC±1 297 per user-household in 2014, while that from fuelwood harvesting was second highest at 1 183 kgC ±1 995 in 2012. Carbon backflows (emissions and animal dung) at household, village and landscape levels were dominated by fuelwood harvesting, followed by cattle farming, while the least came from goat farming and maize production. Household socio-economic and demographic characteristics such as household size and age of the head of household had significant positive associations with the likelihood of households engaging in all land-based livelihood activities and subsequently the carbon redirection associated with these; while others such as proportion female and number of assets owned were positively associated with some livelihood activities (e.g. fuelwood use and proportion female) and negatively with others (e.g. fuelwood use and assets owned). Humans redirected roughly 11% of the total NPP at the village level, while at the landscape level, they redirected approximately 13%. The proportion of NPP redirected in this lived-in landscape increased steadily over the five-year study period. This study provides new insights regarding the influence of households on the carbon cycle in rural African socio-ecological systems. It fills a gap in the literature concerning HANPP at the household local level and is novel in that it considers carbon backflows to assess net carbon removal from the local environment by rural households. To this end, the data give a new perspective on the sustainability of savanna socio-ecological systems that are important for carbon mitigation policies