Soil carbon stocks in conservation agriculture systems of Southern Africa

Highlights

• Increases in soil C stocks in Southern Africa with CA were lower than anticipated.

• Low C input, as a results of limited biomass production, limited soil C increase.

• In most study sites C concentrations was below critical levels after 2–7 years.

• Carbon sequestration should therefore not be oversold in CA promotion.

Abstract

In view of the importance of soil carbon (C) and the scarce data on how conservation agriculture might influence its accumulation in Southern Africa this study presents data from 125 on-farm validation trials across 23 sites in Malawi, Mozambique, Zambia and Zimbabwe. These validation trials are paired plot comparisons of conventional agricultural practice and conservation agriculture that had been established between 2004 and 2009. Traditional cropping systems vary across the study area although they all are tillage based and maize is the main crop grown. The treatments proposed on the validations trials reflect this variability in conventional practice and propose an adapted conservation agriculture option. The sites are thus grouped into four specific treatment comparisons. Bulk density and soil C concentrations were measured from samples collected at four depth layers (0–10 cm, 10–20 cm, 20–30 cm and 30–60 cm), thereafter C stocks were calculated. On the basis of the stover biomass harvest C inputs were assessed. No consistent differences in bulk density and soil C concentrations were found. Carbon stocks were found to be positively influenced by conservation agriculture only when a mouldboard ploughed maize-legume rotation was compared to a direct seeded maize legume rotation (with residue retention). Even when increases were significantly greater under conservation agriculture the order of magnitude was small (∼2 Mg ha⁻¹). Limited C inputs, ranging between 0.1 and 1 g C kg⁻¹ soil yr⁻¹, are likely to be the major bottleneck for C increase. These results, based on on-farm validation trials indicate that there is a limited potential for conservation agriculture to significantly increase soil C stocks after up to 7 years of conservation agriculture practices, in the studied systems.

Introduction

Conservation Agriculture (CA) is defined as a system based on (a) minimum soil disturbance, (b) permanent soil cover (e.g. through crop residue retention) and (c) crop rotations or association (FAO, 2002). It has been widely promoted in Southern Africa in the past decade to address accelerating soil degradation and declining productivity as well as to adapt to climate variability and change (Thierfelder et al., 2014). Although biophysical benefits at the field level have been well documented (Hobbs, 2007, Hobbs et al., 2008, Kassam et al., 2009, Mazvimavi et al., 2008, Mupangwa et al., 2008, Mupangwa et al., 2012, Mupangwa et al., 2013, Mupangwa et al., 2007, Thierfelder et al., 2012a, Thierfelder et al., 2012b, Thierfelder et al., 2013a, Thierfelder et al., 2013b, Thierfelder and Wall, 2009, Thierfelder and Wall, 2010, Wall, 2007) the system has also sparked debate with claims that CA has not been sufficiently tested and scientific data is missing to warrant widespread promotion (Andersson and Giller, 2012, Andersson et al., 2014, Giller et al., 2011, Giller et al., 2009). Among the knowledge gaps highlighted by Giller et al. (2009) was uncertainties of CA systems to increase soil carbon stocks.

In CA systems major inputs in carbon can be expected through the retention of crop residues, crop rotation and, depending on the soil type, the reduction in soil tillage (Chivenge et al., 2007, Thierfelder and Wall, 2010, Thierfelder and Wall, 2012). Results from different authors on long-term stabilization of SOM showed a general increase in soil carbon stocks under no-tillage (the first principle of CA) compared with conventional tillage for both tropical and temperate soils (Six et al., 2002). A meta-analysis of soil carbon stocks done to at least 30 cm depth after at least 5 years of continuous practice, found that in 7 out of 78 cases the soil carbon stock was lower in no-tillage compared with conventional tillage, in 40 it was higher, and in 31 there was no significant difference, thus showing variability in carbon accumulation in the soil with no-till practices (Govaerts et al., 2009). Govaerts et al. (2009) also highlighted that sufficient information and research evidence on the influence of tillage, residue management and crop rotation on carbon storage was lacking especially for tropical and subtropical areas. More recently scientists have published data in an attempt to fill this research gap for Southern Africa. Experimental data from Henderson Research Station in Zimbabwe showed significantly greater soil carbon stocks in the 0–10 cm and 10–20 cm depth layers after only four years of different CA practices (e.g. ripline seeded maize-legume intercropping as compared with conventionally ploughed continuous maize) (Thierfelder et al., 2012b). Similarly, at Monze Farmer Training Centre in Southern Zambia organic carbon concentrations (Thierfelder and Wall, 2010) and carbon stocks (Thierfelder et al., 2012a) were greater under CA than conventional practice for the 0–10 cm depth layer as well as the 0–30 cm depth layer after only four years of treatment. At Matopos Research Station in southern Zimbabwe significant increases in soil organic carbon concentrations were observed over time for ploughed, ripped and basin planted fields that received annual applications of cattle manure and maize residues, with the greatest increases recorded in the planting basins (Mupangwa et al., 2013). However, under on-farm conditions, where residue management is less controlled, results on soil organic carbon changes in response to CA practices were inconclusive. Nyamangara et al. (2013) for example found no significant difference in organic carbon concentration (0–20 cm depth) between CA (up to nine years) and conventional animal drawn tillage fields across fifteens districts in Zimbabwe. In another on-farm study Nyamangara et al. (2014) subdivided soils sampled to 15 cm depth into two categories based on clay content (low: 12–18%; high: >18–46%). They found significantly greater organic carbon concentrations under CA (up to five years) as compared with conventional animal drawn tillage for soils with low clay content, but no differences for soils with a high clay content and no differences either in carbon stocks (0–15 cm depth) of both low and high clay soils (Nyamangara et al., 2014). In Malawi significantly greater organic carbon concentrations for 0–20 cm depth were found under no-tillage (with residue retention) as compared with conventional ridge tillage for fields that had been under no-tillage for 4 and 5 years consecutively (Ngwira et al., 2012). Mloza-Banda et al. (2014) found increased soil organic carbon concentrations under CA (two and four years of no-tillage and residue retention) as compared with conventional ridge tillage in a sample of six on-farm fields in Central Malawi. In all these studies minimum soil disturbance and crop residue retention were parts of the CA systems investigated, while rotations were not always implemented.

In the previous noted studies not all scientists measured bulk densities and reported changes in carbon stocks, making it impossible to assess the potential of CA systems to sequester carbon. In the present study, the focus was on changes in carbon stocks in response to different CA practices in smallholder farmers' fields of Southern Africa. Data were collected from 125 on-farm validation trials in 23 sites across southern Africa. The results presented here test the hypothesis that carbon stocks in fields planted under CA for 2–7 years are greater than in conventionally tilled fields.