147
or two short periods (e.g., Gusha et al. 2014; Mataka et al. 2007; Sarwatt et al.
2002). Despite most data being on a relatively small number of practices, significant
data are available for practices of high interest to the development community. For
example, 28% of the data is on practices that diversify production systems such as
rotations, intercropping and agroforestry (e.g., Myaka et al. 2006; Munisse et al.
2012; Thierfelder et al. 2013; Nyamadzawo et al. 2008; Chamshama et al. 1998).
Therefore, some information exists to reduce the uncertainty about implementing
such interventions. Other commonly studied practices include mulching, organic
fertilizers and reduced tillage.
Common recommendations for CSA interventions include packages of technologies, such as conservation agriculture or systems to intensify rice production.
When multiple practices are adopted together, they can have synergistic or antagonistic effects on CSA outcomes. A significant majority (72%) of our data is from
practices done in combination with at least one other CSA practice (e.g., agroforestry + mulching, intercropping + manure). This provides insights into how practices operate alone or in combination, which helps in making decisions and
recommendations on best practices under specific conditions.
Lastly, we analysed the distributions of outcomes. The first striking pattern is that
82% of data are related to the productivity pillar – yields, incomes, etc. (Fig. 12.4b).
Contrastingly, resilience outcomes make up only 17.5% of the data, which is primarily related to soil quality (11.4%) and input-use efficiencies (4.5%). This means
that there is scant evidence on many other indicators, especially those that are
believed to impart some level of resilience. It is also indicative of the difficulty in
defining resilience indicators in the literature. Finally, only 0.5% of the data set is
related directly to mitigation outcomes, such as greenhouse gas emissions or total
carbon stocks. Thus, there are major gaps in our understanding of how potential
CSA practices affect resilience and mitigation outcomes across various contexts in
East and Southern Africa. There is almost a complete lack of data on mitigation,
which requires urgent action to calibrate low emission trajectories.
One of the fundamental goals of CSA is to produce win-win or win-win-win
outcomes across productivity, resilience and mitigation. However, our data set suggests that it is only possible to analyse win-win outcomes, given the dearth of information on mitigation. That is because most studies only examine a single pillar,
about 32% study two pillars and less than 1% study all three (Fig. 12.4a). This is a
critical insight into the evidence base of CSA because it shows the lack of co-located
(in the same study) research across pillars. It is often not possible to extrapolate
results on the same practice between sites because outcomes can be significantly
influenced by local context (e.g., Pittelkow et al. 2015a, b; Bayala et al. 2012).
Given the general lack of co-located research across CSA outcomes, aggregation
techniques such as the Compendium and meta-analyses, can be used to gain insights
into multiple outcomes from practices, including looking into potential trade-offs
between different objectives.
It was not a surprise that most studies on potential CSA practices examine yields
and soil health, as they are the basis of agronomic research. Perhaps the biggest
surprise in the data set is that there is a significant amount of economic information
12 What Is the Evidence Base for Climate-Smart Agriculture in East and Southern…
or two short periods (e.g., Gusha et al. 2014; Mataka et al. 2007; Sarwatt et al.
2002). Despite most data being on a relatively small number of practices, significant
data are available for practices of high interest to the development community. For
example, 28% of the data is on practices that diversify production systems such as
rotations, intercropping and agroforestry (e.g., Myaka et al. 2006; Munisse et al.
2012; Thierfelder et al. 2013; Nyamadzawo et al. 2008; Chamshama et al. 1998).
Therefore, some information exists to reduce the uncertainty about implementing
such interventions. Other commonly studied practices include mulching, organic
fertilizers and reduced tillage.
Common recommendations for CSA interventions include packages of technologies, such as conservation agriculture or systems to intensify rice production.
When multiple practices are adopted together, they can have synergistic or antagonistic effects on CSA outcomes. A significant majority (72%) of our data is from
practices done in combination with at least one other CSA practice (e.g., agroforestry + mulching, intercropping + manure). This provides insights into how practices operate alone or in combination, which helps in making decisions and
recommendations on best practices under specific conditions.
Lastly, we analysed the distributions of outcomes. The first striking pattern is that
82% of data are related to the productivity pillar – yields, incomes, etc. (Fig. 12.4b).
Contrastingly, resilience outcomes make up only 17.5% of the data, which is primarily related to soil quality (11.4%) and input-use efficiencies (4.5%). This means
that there is scant evidence on many other indicators, especially those that are
believed to impart some level of resilience. It is also indicative of the difficulty in
defining resilience indicators in the literature. Finally, only 0.5% of the data set is
related directly to mitigation outcomes, such as greenhouse gas emissions or total
carbon stocks. Thus, there are major gaps in our understanding of how potential
CSA practices affect resilience and mitigation outcomes across various contexts in
East and Southern Africa. There is almost a complete lack of data on mitigation,
which requires urgent action to calibrate low emission trajectories.
One of the fundamental goals of CSA is to produce win-win or win-win-win
outcomes across productivity, resilience and mitigation. However, our data set suggests that it is only possible to analyse win-win outcomes, given the dearth of information on mitigation. That is because most studies only examine a single pillar,
about 32% study two pillars and less than 1% study all three (Fig. 12.4a). This is a
critical insight into the evidence base of CSA because it shows the lack of co-located
(in the same study) research across pillars. It is often not possible to extrapolate
results on the same practice between sites because outcomes can be significantly
influenced by local context (e.g., Pittelkow et al. 2015a, b; Bayala et al. 2012).
Given the general lack of co-located research across CSA outcomes, aggregation
techniques such as the Compendium and meta-analyses, can be used to gain insights
into multiple outcomes from practices, including looking into potential trade-offs
between different objectives.
It was not a surprise that most studies on potential CSA practices examine yields
and soil health, as they are the basis of agronomic research. Perhaps the biggest
surprise in the data set is that there is a significant amount of economic information
12 What Is the Evidence Base for Climate-Smart Agriculture in East and Southern…