Agropastoral areas in and around Bay face a risk of Famine from June onward

FEWS NET assesses a credible risk of Famine (IPC Phase 5) in crop-dependent agropastoral areas of Bay, Bakool, and Gedo between June and September. If the below-average April-June gu rains end early or dry spells occur during critical crop development stages in May and June, there is high concern that all three technical thresholds for Famine (IPC Phase 5) may be crossed in two to three months. In these areas, Emergency (IPC Phase 4) outcomes are already ongoing, and some households are likely in Catastrophe (IPC Phase 5). The failed October to December 2025 deyr season and subsequently harsh jilaal dry season impeded households’ ongoing recovery from the historic 2020-2023 drought. Multiple seasons of crop, livestock, and labor income losses; record-high food, fuel, and water prices; and escalating conflict have severely eroded livelihoods and coping capacity. Acute malnutrition has risen sharply (Figure 1) amid high morbidity, reduced access to health and nutrition services, and historically low and declining levels of humanitarian food assistance.

The most likely scenario assumes the gu rains continue as forecasted through the end of the season in June, supporting partial crop recovery and up to two months of cereal stocks. The gu rains have been markedly poor to date, delaying critical improvements in cropping and livestock conditions. Land preparation, wet planting, and seed germination were notably delayed in Bay Region. Although rainfall has somewhat improved over the past two weeks, deep deficits persist. However, short-cycle sorghum crops, which are more resilient to dry conditions than maize, are likely to rebound if rainfall accumulation continues to improve over the next two to three weeks. Conditions will remain dire for poor households until the harvest, as the green and main harvests are still six to eight weeks away and agricultural labor income will remain below average. During this time, households will rely heavily on bush products, wild foods, informal labor, food sharing during the Hajj in May, and zakat during Eid al-Adha festivities in June.   

However, the potential for an alternative scenario is credible. If these notoriously volatile rains underperform during the flowering and grain-filling stages, a second consecutive season of crop failure would likely occur. FEWS NET has been closely monitoring the potential for this scenario since December, and several pieces of evidence now point to a moderate likelihood that this alternative scenario may materialize, including below-average short- and longer-term rainfall forecasts through June and typically low June rainfall totals. A second failed harvest would eliminate critical food and income from poor households’ own-produced crops and agricultural labor wages. As increasing numbers of households seek alternative work, competition for limited informal opportunities would intensify, and wages would fall. Wild food and bush product availability would likely plummet due to overexploitation. At the same time, food prices would remain at historic highs, leading to a substantial decline in household purchasing capacity. Poor households do not have the capacity to cope with yet another season of losing their most important sources of food and income; it is likely that many have already liquidated their assets. Already weakened social support networks would become overwhelmed by the scale of need.  

Acute malnutrition levels have already worsened substantially in recent months, driven by the compounding impacts of poor food consumption, high disease incidence, and reduced access to nutrition and health services. During the peak of the deyr lean season in late November/early December 2025, the prevalence of global acute malnutrition (GAM) as measured by weight-for-height z-score (WHZ) stood at 19.5 percent (16.7-22.7 95% CI) in Bay agropastoral areas, with estimated severe acute malnutrition (SAM) by WHZ and morbidity rates of 4.4 percent and 30.5 percent, respectively. By mid-April, GAM WHZ in Bay agropastoral areas had risen to 25.2 percent (21.6-29.2 95% CI), a statistically significant deterioration, while SAM WHZ and morbidity rose to 5.8 percent and 36.4 percent, respectively. In Buurhakaba District, specifically, GAM WHZ reached 37.1 percent, surpassing the 30 percent GAM WHZ threshold for Famine (IPC Phase 5)¹, while the SAM WHZ reached 10.2 percent. Buurhakaba has experienced a comparatively sharp reduction in health and nutrition services, with the number of treatment sites falling from 16 in 2024 to three by April 2026.  

The trajectory toward crossing the Famine (IPC Phase 5) thresholds could be rapid. In the absence of green harvests in June, GAM WHZ prevalence would likely surpass 30 percent across a wider number of areas, and SAM WHZ rates – which are associated with a higher risk of death – would likely also increase. While mortality is still relatively low with a crude death rate (CDR) of 0.40 in these agropastoral areas, the convergence of extremely critical acute malnutrition, high morbidity, and reduced services raises concern that mortality would rise rapidly if hunger were not alleviated by the forthcoming gu harvest.  

The difference between partial crop recovery and another failed harvest is likely to determine whether outcomes improve modestly or deteriorate rapidly toward Famine (IPC Phase 5). Consecutive below-average to failed seasons previously led to Famine (IPC Phase 5) in southern agropastoral areas in 2011 and likely would have resulted in Famine (IPC Phase 5) in 2016/17 and 2022/23 if not for the large and sustained scale-up of humanitarian food assistance. In contrast to 2016/17 and 2022/23, however, humanitarian food assistance and health and nutrition service coverage are already historically low and are expected to be cut back further after June. This significantly heightens the risk that another failed harvest would drive rapid deterioration before food assistance can expand quickly, and the likelihood of the large-scale loss of lives and livelihoods would be high.