Control measures and favorable rainfall anticipated to mitigate impact of desert locust upsurge

• Based on data collected during the 2019 short rains assessment, the Kenya Food Security Steering Group (KFSSG) estimates that 1.3 million Kenyans are facing Crisis (IPC Phase 3) or worse outcomes, a decline of nearly 50 percent compared to the preceding long rains season. Although floods in late 2019 caused crop and livestock losses, the above-average 2019 short rains season led to favorable harvests in most marginal agricultural areas as well as above-average livestock sale values and milk productivity. Many areas of concern have improved to Stressed (IPC Phase 2), but Crisis (IPC Phase 3) persists in Tana Riverine livelihood zone, where 55 percent of planted area was destroyed.[1]

• Desert locusts are present in 21 counties of Kenya, but the impact on food security has been localized and limited to date. In pastoral areas, rangeland resource availability is historically above normal and livestock migration is atypically low. In marginal agricultural areas, net short rains production is approximately 26 percent above the long-term average. Crop losses in late 2019 were primarily affected by heavy rainfall, including post-harvest losses equivalent to more than 10 percent of unimodal production in high and medium potential areas.

• Food availability has improved in early 2020 compared to late 2019, but the delayed unimodal maize harvest, below-average beans harvest, and tight regional supply are still driving high maize and bean prices. In February, the retail maize price in key urban and rural reference markets reached up to 33 percent above the five-year average while wholesale bean prices reached up to 40 percent above the five-year average. However, livestock price increases have outpaced maize price increases and the goat-to-maize terms of trade ranged from 15 to 55 percent above average.

• A forecast of above-average March to May rainfall is expected to consolidate recent gains in livestock production and lead to near-normal crop production, driving Minimal (IPC Phase 1) or Stressed (IPC Phase 2) outcomes through September. The negative impacts of desert locust are expected to be mitigated by current and planned control efforts, headwinds and cooler temperatures that discourage locusts from entering high production areas, and the likelihood that rainfall will regenerate pasture. Areas of greatest concern include Tana Riverine areas, where households are still recovering from the floods, and parts of Northeastern Pastoral livelihood zone, where insecurity will limit control efforts.

  ---

  [1] FEWS NET is a member of the KFSSG. However, FEWS NET’s mapping units differ from the KFSSG, leading to a difference in mapped outcomes.

Current Situation

After the 2018/19 drought, exceptionally above-average rainfall during Kenya’s 2019 short rains season disrupted livelihoods activities and slowed the pace of recovery in many areas, but food security conditions are now gradually improving. Cumulative rainfall was 140-350 percent of the 1981-2010 average across the country, leading to large-scale floods that caused landslides, contaminated the water supply, caused crop and livestock losses, and destroyed critical health, irrigation, road, and social infrastructure (Figure 1). According to final estimates by the Kenya National Disaster Operations Centre, the floods affected approximately 472,000 people, displaced an estimated 24,000 people, and left 266 dead, 82 injured, and 17 missing. In addition, the floods killed an estimated 30,000 heads of livestock. Northwestern Pastoral, Northern Pastoral, Northeastern Pastoral, Southeastern Pastoral, Mandera Riverine, and Tana Riverine livelihood zones were the worst-affected areas, where damage to roads and bridges suspended market functioning and many households endured livestock losses or crop losses. Due to the loss of household food and income sources, Crisis (IPC Phase 3) outcomes persisted through December in many areas. The start of the dry season in January has permitted normal livelihoods activities to resume, which has enabled recovery to Stressed (IPC Phase 2) area-level outcomes in most areas of concern. However, many poor households are still experiencing food gaps or engaging in crisis livelihoods coping strategies indicative of Crisis (IPC Phase 3) or worse.

The heavy rains, coupled with tropical cyclones in the Indian Ocean, created conducive conditions for the spread of desert locust in Kenya, but the negative impacts on crop and livestock production have not been large scale to date. As of late February, desert locusts had spread to 21 counties in Kenya (Figure 2). Crops were broadly in the maturation stages or had already been harvested, while the rains and positive soil moisture anomalies facilitated rangeland regeneration and offset pasture losses. The Kenya Food Security Steering Group (KFSSG) and County Steering Groups estimate that pasture and crop losses from desert locust are 1-5 percent in southeastern Kenya and 5-15 percent in northern Kenya. At the national level, national and county governments, NGOs, and international organizations have funded about USD 18.5 million out of the 21.4 million allocated to aerial and ground control operations. To date, only 35,000 hectares of infested areas have been sprayed, which represents a small proportion of affected areas. According to FAO’s Desert Locust Watch and confirmed by the KFSSG’s short rains assessmen, there is ongoing hatching of locust nymphs and formation of immature swarms in northern and central counties.

Crop and livestock production: Food availability has broadly improved in early 2020, though crop losses in unimodal, high and medium potential agricultural areas due to the atypically heavy rains has resulted in tighter-than-normal supply. According to FEWS NET’s estimates, total maize production in 2019 – including national long and short rains production – is three percent below average and eight percent below 2018. The easing of rainfall during the January to February dry season permitted the completion of both the delayed, unimodal long rains harvest in high and medium agricultural areas in western Kenya and the bimodal short rains harvest in marginal agricultural areas in central and southeastern Kenya. Based on the Ministry of Agriculture (MOA) estimates, post-harvest losses are estimated to be upwards of 10 percent since the heavy, sustained, heavy rainfall constrained maize drying and storage causing losses from rotting.

In rain-fed marginal agricultural areas, specifically, short rains production of maize, cowpea, and green grams generally performed better than previously expected. Beans, which are less moisture tolerant, performed poorly. According to the findings of the KFSSG Short Rains Assessment, total maize production is approximately 26 percent above the five-year average, attributed to above-average area planted and high yields in areas not affected by floods. However, below-average area planted and crop losses from flooding, leaching, and water logging, as well as minimal losses from Fall Army Worm and desert locust infestations, resulted in below-average maize production in Kilifi, Kwale, Embu (Mbeere), and Tharaka Nithi (Tharaka), where production ranged from 10 to 18 percent below average on the county level. Total cowpea production was 68 percent above average and green grams production ranged from 15 to 33 percent above average on the county level, primarily attributed to an increase in area planted and county government subsidy programs for seeds, fertilizers, and chemicals. In contrast, beans production ranged from 25 to 38 percent below average on the county level.

The 2019 floods caused livestock losses in some areas, but livestock production is now largely benefitting from enhanced rangeland conditions as a result of the above-average rains. Based on field information and supported by the satellite-derived Normalized Difference Vegetation Index (NDVI), pasture and browse conditions are good and above-normal. In late February, NDVI data showed vegetation greenness was upwards of 140 percent of normal across most of the country and 105-120 percent of normal in western and northeastern Kenya (Figure 3). Similarly, surface water availability ranges from normal to above normal, though some water sources were destroyed by floods. As a result, return trekking distances for livestock from grazing areas to water sources were considerably below average in February, ranging from 2 to 8 kilometers (km) compared to 4-15 km normally. Consequently, livestock are being watered twice as regularly across all pastoral areas except in Wajir and Isiolo, where the frequency remains normal at once every two days. Most livestock have remained close to homesteads in the wet season grazing areas and have not migrated to typical dry season grazing areas.

Good rangeland resource availability and lower return trekking distances have helped to maintain good livestock body conditions, which has in turn supported healthy livestock births and conceptions, boosted milk production, and improved livestock value on the market. In February, the National Drought Management Authority (NDMA) observed that all livestock species exhibit good body conditions due to good pasture, browse, forage, and water availability. Birth and conception levels have generally returned to normal among small stock, but births among large stock remain below normal as they have only recently fully recovered.  However, field assessment information collected by the KFSSG in February shows that poor households’ livestock holdings in terms of tropical livestock units (TLUs)[1] remain below the 10-year average due to losses during recent droughts and floods as well as distressed sales. On the county level, poor households’ livestock holdings range from 2 to 6 TLUs, compared to the 10-year averages of 3-10 TLUs. Despite lower livestock holdings, milk production has rebounded from the drought, reflecting good birth levels and body conditions. Daily net production levels are 2-7 liters per household per day compared to 2-6 liters normally. However, some exceptions exist, such as in Isiolo, where livestock holdings are lowest (2 TLUs) and milk production is one liter per household per day on average. Veterinary department interventions have also proven instrumental to limiting livestock disease outbreaks and maintaining livestock health. Despite widespread flooding, no livestock disease outbreaks have been reported since October except in Marsabit, where FMD, CCPP, and PPR cases have resulted in a mortality rate of around six percent of small stock.

Adequate water availability has similarly benefitted domestic consumption and food utilization. In February, the cost of water is near normal or up to 50 percent below normal, ranging from KES 2 to 5 for a 20-liter jerrican across most pastoral areas and KES 10 in Turkana. Return trekking distances from domestic water sources are similarly normal to below normal, ranging from 2 to 5 km compared to 4 to 8 km normally. Reductions in trekking distances have resulted in improved access to clean water for cooking, cleaning, and drinking, which has reduced household vulnerability to waterborne disease. Households are also spending less time trekking to fetch water, leaving more time to engage in livelihood activities to obtain food and income. However, an exception to this trend is Tana Riverine livelihood zone, where severe flooding contaminated water sources and led to increased waterborne disease incidence.

In addition to improved productivity, adequate availability of rangeland resources has led to generally lower levels of resource-based conflict in pastoral areas. Nonetheless, localized insecurity, cattle rustling, and sporadic intercommunal conflict periodically affects local livelihoods and market access. In Marsabit, for example, retaliatory attacks in response to cattle rustling in parts of North Horr sub-county recently led to temporary closures of markets and primary schools and displaced about 150 households. Sporadic cattle raids also affected market operations at the border of West Pokot and Turkana at Kainuk, but the situation has since returned to normal. In Samburu, recurring communal and resource-based conflicts also periodically limit access to forage and water for livestock and impede market operations in Baragoi.

Markets and trade: The delayed unimodal maize harvest, below-average beans harvest, and tight regional supply are driving high maize and bean prices in most urban and rural key reference markets, which is affecting food access among poor households who lost crops due to the floods, have few saleable animals, and primarily rely on labor income. Although FEWS NET cross-border monitoring data indicate that maize imports into Kenya during the fourth quarter of 2019 accounted for 12 percent of total regional imports, imports have not filled the national cereal gap. NDMA price data in February showed that the retail price of maize was 14-33 percent above the five-year average in most reference markets; exceptions included markets in Kwale county, where the price of a kg of maize was near average, and markets in Turkana county, where the price was 16 percent below average due to sufficient supply flows from source markets in Trans Nzoia county. Similarly, the wholesale price of dry beans ranged from 18 to 40 percent above the five-year average in urban reference markets and from near average to 14 percent above average in rural markets.

Assumptions

The February to September 2020 most likely food security outcomes are based on the following national-level assumptions:

• According to the convergence of NOAA, ECMWF, and GHACOF forecasts, probabilistic models indicate the March to May bimodal long rains season and February to August unimodal long rains season are most likely to be above average. The forecast is driven by a likelihood of slightly warmer-than-normal Indian Ocean sea surface temperatures and neutral ENSO and IOD conditions through mid-2020.
• Ongoing desert locust breeding is expected to result in the formation of new swarms in March/April in grasslands and croplands in central, northern, and eastern Kenya, which coincides with the vegetative stage in bimodal cropping areas. A second wave of breeding is expected in June. Based on recent funds committed by donor governments and the Kenyan government, aerial and ground control measures are expected to continue to mitigate the impact of desert locust on crop and livestock production, though efforts will likely be limited by insecurity in areas along the Kenya-Somali border. In addition, relatively colder temperatures and a shift in seasonal wind patterns are most likely to mitigate the spread of desert locust to western high potential areas, while forecast above-average rainfall is expected to help offset pasture losses during the long rains season. As a result, desert locusts are most likely to cause localized cash and food crop losses in southeastern marginal agricultural, western agropastoral, and central highland high potential areas. In pastoral areas, atypical pasture deterioration in localized areas, particularly in the east, is likely during the June to September dry season.  
• Based on the above-average long rains forecast, planned desert locust control operations, and high likelihood that better-off households will seek to maximize crop production prospects during a favorable rainfall season, area planted for the unimodal and bimodal long rains seasons is expected to be above average. However, the bimodal long rains harvest in July/August is most likely to be near average due to the likelihood of crop losses in marginal agricultural areas from desert locust.
• Agricultural labor demand is generally expected to be above normal from February through September as a result of increased area planted and heightened demand for weeding during above-average rainfall seasons, especially in high and medium potential areas. Labor demand may also be driven by farm-level control efforts to spray or scare away desert locusts. However, crop losses from desert locust may lead to declines in agricultural labor demand during the weeding stages and during the July/August bimodal harvest in localized areas. 
• Based on FEWS NET estimates, the national maize balance is expected to be approximately 3.06 million metric tons (MMT) in June 2020, prior to the start of the long rains harvest. The anticipated balance is near the five-year average of 3.147 MMT, but due to population growth and increased household consumption levels that is eight percent above the five-year average, the ending stocks in June are projected to be 0.1 MMT, 52 percent below the five-year average and indicative of a deficit. The maize import gap is anticipated to be about 0.69 MMT, or 10 percent above the five-year average. Although imports from Tanzania, Uganda, and Ethiopia are expected to fill the gap, above-average demand and high marketing costs are expected to contribute to above-average maize prices.
• Based on FEWS NET’s technical price projections in reference markets in Nairobi, the retail price of a kg of maize is expected to range from 34 to 40 percent above the five-year average throughout the scenario period. Prices are most likely to peak at around KES 50 between April and August. The wholesale price of a kg of beans is expected to be approximately 55-60 percent above average throughout the scenario period, peaking at about 120 KES.
• Based on current above-normal vegetation, the above-average rainfall forecast, and anticipated desert locust control operations, rangeland resources are expected to remain broadly above normal throughout the scenario period. As a result, livestock migration is broadly expected to occur at below-normal levels during the June to September dry season. However, atypical pasture losses and associated migration is likely in localized areas of Mandera, Wajir, and Garissa, where insecurity will limit desert locust control operations.
• Outbreaks of Rift Valley Fever (RVF) are likely to occur in Marsabit, Isiolo, Wajir, Mandera, Garissa, Tana River and Kilifi counties in February and from late May. The long rains will be the second above-average rainfall season and, based on current above-normal soil saturation levels, there is an increased likelihood of standing water that creates habitats for RVF mosquito vectors. However, the outbreaks are anticipated to be less severe than those of 2019 due to heightened awareness, surveillance systems, and improved control measures both at the national and county level. 
• Given that favorable rangeland resources will most likely support good livestock body conditions, most households are expected to limit livestock sales in order to restock their herds, and quarantines may be imposed in response to RVF outbreaks which will further restrict market supply, livestock prices are expected to be sustained at moderately to significantly above average prices throughout the scenario period.
• Humanitarian assistance is expected to continue across the country as vulnerable and food insecure households are supported by a combination of the national and county governments and humanitarian agencies. Interannual assistance in the form of cash transfers amounting to KES 2,700 per month by NDMA’s Hunger Safety Net Programme (HSNP), funded by DFID, is expected to reach about 100,000 households (600,000 beneficiaries) in Mandera, Turkana, Marsabit, and Wajir counties. Cash transfers, supplementary feeding and in-kind distribution of food commodities by WFP to over 800,000 beneficiaries will likely continue throughout the scenario period.
• The refugee populations in both Dadaab and Kakuma refugee settlements, continue to receive food assistance rations reaching more than 25 percent of their population. However, as anticipated in October due to funding constraints, WFP cut rations down to 70 percent of daily kilocalorie needs and this will persist throughout the scenario period. 

Most Likely Food Security Outcomes

Pastoral area outcomes: Stressed (IPC Phase 2) outcomes are expected to be sustained in most pastoral livelihood zones throughout the scenario period, and many poor households who are currently in Crisis (IPC Phase 3) in Northwestern, Northern, and Northeastern Pastoral areas are expected to improve to Stressed (IPC Phase 2). During the March to May long rains season, household income from livestock production and household milk consumption is expected to range from normal to above-normal levels, based on above-normal rangeland resources, anticipated livestock births, high milk production, and above-average terms of trade. This trend is likely to broadly continue through the June to September dry season, given the positive effect of two consecutive above-average rainfall seasons on household income from livestock production. Improved livestock productivity and high terms of trade are most likely to prevent most households from experiencing food consumption gaps throughout the lean season. Although the ongoing desert locust upsurge is of high concern, above-average rainfall is anticipated to continue to regenerate pasture, forage, and browse from March to May, offsetting losses. Further, current and planned control measures are most likely to control the intensity of breeding and mitigate the severity of damage to rangelands. However, given the likelihood that RVF outbreaks will result in quarantines during the rainy season and suspend livestock sales, poor households in affected areas will be likely to periodically engage in stressed or crisis coping strategies to meet their minimum food needs. In addition, persistently low livestock holdings will continue to be a limiting factor to their ability to purchase non-food needs.

Although Stressed (IPC Phase 2) outcomes are expected in most livelihood zones at the area level, inaccessible, insecure areas in Northern Pastoral livelihood zone are most likely to deteriorate to Crisis (IPC Phase 3) during the lean season. In these areas, household access to food and income sources is increasingly likely to be disrupted by the impact of insecurity on market functioning, while the additional, new hazards presented by desert locust and RVF outbreaks are likely to be more severe given more limited control measures. Households that are currently in Emergency (IPC Phase 4) due to the impact of the floods are expected to regain access to labor income or gifts in the context of more widespread community recovery, but given low to no livestock holdings, these households will likely continue to experience food consumption gaps through the lean season and will be in Crisis (IPC Phase 3). Throughout the scenario period, acute malnutrition levels are expected to remain stable or marginally improve across all pastoral areas, given improvements in milk and food intake. For example, Turkana (Turkana West) is expected to improve from ‘Critical’ to ‘Serious’. However, chronic food and non-food security factors are likely to sustain high but typical ‘Critical’ levels of acute malnutrition in Mandera, Turkana, and Wajir counties and in parts of Baringo (East Pokot sub-county) and Marsabit (Laisamis and North Horr sub-counties).

Marginal agricultural area outcomes: Minimal (IPC Phase 1) or Stressed (IPC Phase 2) outcomes are most likely throughout the scenario period. Poor households’ own-produced food stocks from the short rains harvest are expected to last through April in Kwale, Lamu, Kilifi, Makueni and Tharaka Nithi (Tharaka) counties and through July in areas that had better production. Due to anticipated above-average labor demand, income earned from land preparation activities for the long rains cropping season is most likely to normalize food access beginning in March, despite high food prices. Although localized staple and cash crop losses from desert locust are expected, high agricultural labor demand for replanting and locust control is expected to provide income that partially offsets poor households’ own-produced crop losses. However, given the early depletion of own-produced food stocks, poor households in Kwale, Lamu, Kilifi, Makueni, and Tharaka counties will likely continue to face difficulty meeting non-food needs given increasing dependence on higher priced staple food purchases from May until the short rains harvest in July. By July, the near-average long rains harvest, which comprises only about 30-40 percent of the total annual harvest in these areas, is expected to provide an additional food source. However, given localized crop production shortfalls due to desert locust, likely reductions to income from harvesting labor that a large proportion of poor households depend on, and high staple food prices, additional households in areas like Meru (Meru North), Nyeri (Kieni), Embu (Mbeere), Kitui, Makueni, and Taita Taveta will deteriorate to Stressed (IPC Phase 2). Based on historical trends and given that milk availability will remain high, ‘Acceptable’ levels will likely be sustained throughout the scenario period.

Events that Might Change the Outlook

Possible events over the next eight months that could change the most-likely scenario.

For more information on the outlook for specific areas of concern, please click the download button at the top of the page for the full report.