Motivation
In 08. A New Walking Shoe: Modern Portfolio Theory , Malkiel introduced me to Modern Portfolio Theory, a technique for reducing portfolio risk, despite risky investments. It made me wonder how similar principles are applied in agriculture. This page is the result of that investigation.
~16% of Kenya’s land is sufficiently fertile and has good rainfall. In 2006, ~75% of working Kenyans are farmers (compared to 80% in 1980). (en.wikipedia.org)
Adaptation Strategies in Kenya [Hoang, 2013]
Adaptation strategies: crop diversification, mixed cropping, tree planting, water-conservation, irrigation, cattle rearing, and proper land use along landscape gradients.
Drought and rainfall prediction hinge on social networking and indigenous knowledge. Access to water is regarded as the main vulnerability factor.
Coping Better with Current Climatic Variability [Cooper, 2008]
Rain-fed agriculture is the dominant source of food production and income for most of the rural poor. Climate change may make matters worse by increasing variability of rainfall predictions. Even the intra-season variability can have a major effect on crop productivity. To make it worse, semi-arid regions have more variability in seasonal rainfall totals.
Coping Strategies in Sub-Saharan Africa (SSA)
The coping strategies are designed to mitigate risk. They fail to exploit the average & better than average seasons. Farmers tend to over-estimate the negatives and under-estimate the positives.
Many small-scale rural farmers don’t have the ability to diversify their livelihood. Promising areas are drought-resistant varieties and better water management.
The Importance of Climate Analytical Tools
As climate analytical tools get better, we gain greater understanding of the variability. This can help rain-fed agriculture on two fronts.
Seasonal Weather Forecasting
Farmers are constrained by the timing, scale and format of conventional forecasts. They do not trust nor comprehend the forecasts. With competent guidance, seasonal forecasts would go a long way.
Furthermore, such forecasts could be used for disaster preparedness by feeding them into crop growth simulation models that provide probabilistic crop yield and production estimates in advance of the harvest.
Characterizing and Mapping the Agricultural Implications of Climatic Variability
Crops principally respond to sequences of daily rainfall. Daily data (of which there’s a ton) can help predict a wide range of parameters, e.g. start of growing season, frequency of intra-season dry spells, high intensity erosive rainfall, length of growing season, success of crop, water & soil management strategies, etc.
Case Study: Nitrogen Fertilizers in Semi-Arid Zimbabwe for Maize
Agricultural Productions Systems Simulator, APSIM, was used. N-fertilizer is recommended at \(52 kg/ha\). Farmers consider the recommendation too risky and expensive - but they could do \(17 kg/ha\).
Using 46 years of daily climactic data, APSIM simulated maize yields with \(0 kg/ha\), \(17 kg/ha\) and \(52 kg/ha\).
Except in very bad years, rates of return from the \(17 kg/ha\) were substantially better. This provided quantified risk and opportunities of N-fertilizer use. The \(17 kg/ha\) was applied and despite below-average rains, micro-dosing increased maize yields by 30% to 50%.
Providing ‘on the spot’ answers to farmers' climate risk management concerns aroused enormous interest amongst farmer groups. This may directly help farmers in their decision making.
Limitations of (Daily) Climatic Data
The climate data is usually collected at specific weather stations, making interpolation of outputs between weather stations problematic.
Indigenous Knowledge (IK) Related to Climate Variability and Change [Speranza, 2009]
Field Interviews
On the causes of drought: God’s wish to punish man (69%), deforestation (21%), changes in weather patterns & conditions (12%), lack of hills, water masses & mountains (4%), lack of traditional sacrifice (4%), witchcraft (2%) and misuse of sacred areas (2%).
On sources of predictions: IK plus Kenyan Meteorology Department forecasts (29%), IK only (28%), IK but season is God-given thus no need to look further (16%), no signs available (13%), no position (10%), IK plus diviners (5%), IK plus traditional sacrifices (1%).
IK featured indicators from: fauna (37%), local weather conditions (36%), rainfall amount and patterns (22%), flora (20%), astrological constellations (11%) and the local physical environment (4%).
Of the 29% that knew of the likelihood of drought, 22% adapted (drought resistant crops, early maturing crops, ceasing to sell stored grains, saving money). Within this 22%, 16% believe IK to be reliable.
Poverty strongly determines capabilities of preventive measures. Less than a third could take any adaptive measures against the foreseen impacts of the 1999/2000 drought. Therefore, many agro-pastoralists' decisions regarding adverse rainfall variability are of a contingent and reactive nature.
On the Future of IK
People well-versed in IK are dying out. IK is no longer systematically passed to the next generations. IK is not incorporated into current curriculum, and even then as second-hand knowledge because people do not experience it themselves
As climate changes, plants adapt. The future use of biotic objects for climate change monitoring is limited.
IK derived from climatic variables like temperature, wind direction, rainfall amounts and patterns can still provide valuable information on climate change. But these indicators require continuous monitoring as they also change.
Resilience in Agriculture through Crop Diversification [Lin, 2011]
| Type of diversification | Nature of diversification | Benefit | Examples |
|---|---|---|---|
| Increased structural diversity | Makes crops within the field more structurally diverse | Pest suppression | Strip-cutting alfalfa during harvest allows natural enemies to emigrate from harvested strips to adjacent nonharvested ones (Hossain et al. 2001) |
| Genetic diversity in monoculture | Growing mixed varieties of a species in a monoculture | Disease suppression | Genetic diversity of rice varieties reduces fungal blast occurrence (Zhu et al. 2000) |
| Increased production stability | Increased genetic diversity was positively related to mean income and stability of income (Di Falco and Perrings 2003) | ||
| Diversify field with noncrop vegetation | Growing weed strips or vegetation banks in and alongside crops | Pest suppression | Grassland or refugia planted at field margins (beetle banks) were used as overwintering habitat for natural enemies (Thomas et al. 1991) |
| Pest suppression | Using white and black mustard on the field margins of sweet corn crops trapped pests and prevented them from entering the cornfield (Rea et al. 2002) | ||
| Crop rotations | Temporal diversity through crop rotations | Disease suppression | Alternating cereal crops with broadleaf crops and changing stand densities disrupts the disease cycles (Krupinsky et al. 2002) |
| Increased production | Manipulating diversity through crop rotations of greater cover crop and nitrogen-fixing crops increased the yield of the primary crop (Smith et al. 2008) | ||
| Polycultures | Growing two or more crop species and wild varieties within the field; spatial and temporal diversity of crops | Disease suppression | Grassland fields planted with multiple species to decrease disease transmission (Mitchell et al. 2002) |
| Climate change buffering | More ecologically complex systems with wild varieties and temporal and spatial diversity of crops were able to grow under climate stress (Tengö and Belfrage 2004) | ||
| Increased production | Grassland plots with greater in-field species diversity led to more stable feed and fodder production (Tilman et al. 2006) | ||
| Increased production | Grassland plots with greater in-field species diversity led to increased production (Picasso et al. 2008) | ||
| Agroforestry | Growing crops and trees together; spatial and temporal diversity | Pest suppression | Willow trees grown in natural willow habitats experience lower rates of pest outbreak of the leaf beetle (Dalin et al. 2009) |
| Pest suppression | Greater shade diversity increased bird natural enemy abundance for larval control on crop plant (Perfecto et al. 2004) | ||
| Pest suppression | Coffee berry borer control increased with greater ant diversity and abundance in shade systems (Armbrecht and Gallego 2007) | ||
| Climate change buffering | Greater shade cover led to increased buffering of crop to temperature and precipitation variation (Lin 2007) | ||
| Climate change buffering | Greater shade tree cover led to increased buffering from storm events and decreased storm damage (Philpott et al. 2008) | ||
| Mixed landscapes | Development of larger-scale diversified landscapes with multiple ecosystems | Pest suppression | Complex landscapes that have areas of woodland and hedgerows interspersed within fields had higher rates of larval parasitism (Marino and Landis 1996) |
| Pest suppression | Oilseed rape crops adjacent to complex, structurally rich, and large old fallows had higher rates of parasitism by the rape pollen beetle (Thies and Tscharntke 1999) | ||
| Increased production | Mixed land use of organic cropland, crop rotations, and intensive managed grazing led to optimal diversity and profitability strategies (Boody et al. 2009) |
Conclusion
It looks like a decent amount of work has been done to treat agriculture as a portfolio problem. However, the application of the proposed strategies is lacking in Kenya. The main hindrance is majority of the farmers being small scale and low income. Even when they do have access to the information, the ability to act on it is out of reach.
Why do these numbers add up to 102%?