Reading Time: 11 minutes
Cows grazing in field


There has been a strong emphasis in recent years for a reduction in greenhouse gas (GHG) emissions in every sector of business. Countries and large corporations have made commitments for emissions reductions and set targets for the extent to which they are aiming to reduce GHG emissions throughout their supply chains. In 2018, GHG emissions from agriculture contributed 17% of total global emissions. Although it is encouraging that this amount has reduced from 24% in 2000, the actual quantity of emissions from agriculture has increased by 14% from 2000 to 20181. To meet the global demand for reduced GHG emissions there needs to be reductions happening throughout the value chains of every product. This starts at the farm level. To understand how these reductions might happen, it is important to learn from farmers who have been able to reduce their emissions in recent years.

Identifying farmers who have reduced GHG emissions

Trace & Save is an independent sustainable agriculture company that has been working with pasture-based dairy farmers in South Africa for the past 11 years. As part of the work that Trace & Save does, we assess the farm-gate lifecycle GHG emissions of each farm each year. There are 95 farmers who have been participating with Trace & Save for between one and ten years, which gives us a large database. I used this database to find 20 farms that have reduced their GHG emissions per kilogram (kg) of milk over the past five years (2017-2021 for some farms, and 2018-2022 for others). Since the total milk production in South Africa has remained relatively constant over recent years (2.9% increase from 2017 to 20222), any reduction in GHG emissions per kilogram of milk translates to an actual reduction in total GHG emissions for the country. Therefore, lessons can be taken from these 20 farms as to how they have reduced their emissions.

The 20 farms identified are based in different pasture-based regions along the Southern coast of South Africa, therefore they do not include farms from Kwa-Zulu Natal, which is the other large pasture-based dairy production region of South Africa (Trace & Save has not been working in that region as long, therefore there is not yet data for five full years). They represent farmers who have fully irrigated farms, a mixture of irrigation and rain-fed, and farms that are fully dependent on rain-fed pastures. Table 1 provides an overview of some key aspects of the farms, based on their most recent data.

Table 1: Average farm size, milk production and animal numbers for the 20 pasture-based dairy farms included in the study.

Insights from carbon footprint and partial productivity data

Total GHG emissions

The average carbon footprint for the 20 farms was 1.34 (±0.18) kg carbon dioxide equivalents (CO2e) per kg fat- and protein-corrected milk (FPCM) five years ago. This figure reduced steadily to 1.20 (±0.16) kg CO2e/kg FPCM in year five (figure 1). This is a reduction of 11% over five years. The farm with the highest reduction in their carbon footprint lowered it from 1.55 kg CO2e/kg FPCM in year one to 1.14 kg CO2e/kg FPCM in year five. That is a 26% reduction.

Figure 1: Average GHG emissions (with standard deviation bars) on 20 pasture-based dairy farms in South Africa who were able to reduce their GHG emissions over the past five years.

GHG emissions per source

The largest reductions have come from pasture and crop production and purchased fertiliser production, which have reduced by 28% and 29% from year one to year five respectively (Table 2). But the most significant reductions in actual emissions are from pasture and crop production, purchased feed production and enteric fermentation, which have reduced by 0.04, 0.04 and 0.03 kg CO2e/kg FPCM from year one to year five respectively. The only source that has shown an increase in emissions is electricity usage, from 0.07 kg CO2e/kg FPCM in year one to 0.08 kg CO2e/kg FPCM in year five.

Enteric fermentation emissions contribute half (50% of total emissions based on year five data) of the total emissions on these pasture-based dairy farms. Fuel, fertiliser production, pesticide production, transport, and embedded energy contribute less than 2% each to total emissions. Emphasis for reduced emissions should therefore be placed on management practices that will impact on reducing emissions from the sources that have the greatest impact, namely enteric fermentation, manure management, crop & pasture production, electricity usage, and purchased feed.

Table 2: Average GHG emissions by source on 20 pasture-based dairy farms in South Africa who were able to reduce their GHG emissions over the past five years.

Insight into farm systems and practices which result in reduced GHG emissions

Partial productivity indicators provide insight into farm systems and practices. These indicators can be used to identify what changes farmers have made to result in the decrease in GHG emissions over the past five years. The most important improved management practices to focus on are those which have the greatest impact on reducing GHG emissions, and over which the farmers have the most control to change. These include:

  • Improved milk production efficiency, as indicated by milk production per 100 kilograms of liveweight (l/100kg LW) and milk solids per 100 kg of liveweight (kg solids/100kg LW). There has not been a large improvement, on average, on these farms over the five years. In year one 1 205 (±147) l/100kg LW and 97.6 (±2) kg solids/100kg LW were produced over a standardised lactation. In year five 1 211 (±145) l/100kg LW and 99.4 (±15.3) kg solids/100 kg LW were produced (figure 2). This is associated with the lack of significant reduction in enteric fermentation GHG emissions, as improved performance in production per kilograms of liveweight can reduce enteric fermentation emissions per kg FPCM.

Figure 2: Average milk production efficiency, measured in litres of milk produced per 100 kg liveweight over a standardised lactation

  • Improved feed conversion efficiency, as indicated by bought feed fed relative to milk production. The average concentrates fed per milk production have decreased from 420 (±108) g/l in year one to 369 (±77) g/l in year five (figure 3). The average bought roughage fed per litre of milk produced has decreased from 208 (±124) g/l in year one to 132 (±94) g/l in year five (figure 4). These indicate a significant improvement in feed conversion efficiency on these farms, which has contributed to the decrease in GHG emissions from bought feed production.

Figure 3: Average concentrates fed relative to milk production.

Figure 4: Average bought roughage fed relative to milk production.

  • Increased proportion of pasture intake, as indicated by an increase in the percentage of pasture contributing to the overall farm diet. The proportion of pasture has increased from 53 (±6) % in year one, to 57 (±11) % in year five (figure 5). Although this increase does not seem significant, it is associated with the decrease in bought feed, since in order to maintain milk production efficiency with reduced bought feed, farmers need to ensure an increase in pasture in the diet.

Figure 5: Average proportion of pasture in the overall farm diet.

  • Improved fertiliser efficiency, as indicated by a reduction in the amount of nitrogen (N) fertiliser applied per hectare. The N applied has reduced on these 20 farms from 195 (±115) kg N/ha in year one to 113 (±51) kg N/ha in year five (figure 6). This is associated with the decrease in crop & pasture production and fertiliser production emissions.

 Figure 6: Average nitrogen fertiliser application rate per hectare.

  • Reduced number of heifers being raised as replacement heifers, as indicated by the reduced heifer replacement rate. Although this figure of heifers being raised as a percentage of adult animals on the farm has not reduced significantly, 27 (±12) % in year one to 24 (±9) % in year five (figure 7). Raising fewer heifers results in a higher proportion of productive animals on the farm; therefore, it is associated with lower GHG emissions from enteric fermentation and manure management relative to milk production. This association is indicated more strongly by the decrease in heifer replacement rate on the farm with the highest decrease in enteric fermentation emissions (as referred to earlier) which had a decrease in heifer replacement from 58% in year one to 15% in year five. This is not necessarily related to inefficiency, since farms can often raise excess heifers strategically to grow their milking herd. But it does show the significant impact of raising excess heifers on a farms carbon footprint.

Figure 7: Average number of heifers being raised as a percentage of the adult animals on the farm.

  • Electricity and fuel emissions are directly associated with usage relative to milk production. There have not been any significant reductions in these emissions on the 20 farms included in this study over the past five years. Although this does not mean that there is not opportunity for farms to reduce these emissions – reduced fuel and electricity use are both important aspects of lowering a farm’s carbon footprint.


The reductions in emissions on the 20 pasture-based dairy farms over the past five years are an encouragement to any farm that would like to reduce their environmental impact. The most significant improvements have come from increased feed conversion efficiency, a higher proportion of pasture in the diet, and lower N fertiliser application rates. There are complex farm management practices associated with achieving these improvements, but this article is not the space to explore these. The same improvements that have led to reduced emissions on these farms are associated with increased farm profitability, since they are associated with decreased input costs and maintained milk production efficiency.

The encouragement to farmers is that there is opportunity to improve the efficiency of farm management, leading to reductions in bought feed and fertiliser, which will have an associated decrease in GHG emissions. This is a win-win scenario for farmers, industry stakeholders and consumers alike.


  1. 2020. Emissions due to agriculture. Global, regional and country trends 2000–2018. FAOSTAT Analytical Brief Series No 18. Rome. ( – accessed 28 August 2023)
  2. Milk SA. 2023. Lacto Data, Vol 26, June 2023. ( – accessed 28 August 2023)
Craig Galloway