Postharvest Management of Grains in Haiti and Gender Roles

: Food security is a continuing challenge in Haiti. The demand for food far exceeds local supply. As a result, Haiti imports nearly 50% of its national food needs. Postharvest management is an often neglected link in the grain value chain that has potential to improve food availability. We interviewed 214 farmers in three departments in July–August 2017 to assess postharvest handling and storage of grains and gender roles in Haiti. Results showed that among the respondents: 64% were male; 55% were over the age of 40 years; and about half had attended secondary school. Maize and beans were the most grown and stored crops. The average production for maize and beans was 288 kg and 88 kg, respectively. About 75% of the respondents stored less than 100 kg of either crop. Rodents and insects were the main causes of loss during storage. Farmers who produced more grain, stored longer, and experienced losses during drying and seed storage were more likely to use insecticides on stored products. Postharvest management practices were gendered at the lower end of the value chain; where women played a key role in marketing the grain. Addressing postharvest management challenges, through targeted interventions, to increase food availability while investing in maize and bean production can improve food security in Haiti.


Introduction
Food insecurity has been a continuing challenge in Haiti [1]. The situation was exacerbated by a 2010 earthquake, longstanding drought, floods, landslides and the devastating 2016 category 4 Hurricane Matthew that severely affected two million people in Haiti [1]. Other drivers of food insecurity are poor performance of the agricultural sector that leads to heavy dependence on food imports [2]. Haiti is making progress, but its global hunger index is at an "alarming" threshold with about half of its population undernourished [2]. Haiti's food and nutrition requirements are ensured mainly through imports for a wide variety of goods including cereals, pulses and edible oils [3]. The majority of Haitian farmers operate small-scale subsistence farms and have less than two hectares of land [3]. Staple crops produced in Haiti are maize, wheat, sorghum, beans, peas, yams, cassava, sweet potatoes, bananas and plantains [3]. In Haiti, crop production is extremely vulnerable to the adverse effects of climatic events and shocks such as insufficient rainfall, hurricanes, flooding, and droughts [4].
Postharvest losses in Haiti are due to several factors including inadequate handling and storage practices. Haitian farmers typically store their crops for three to four months [5]. Farmers suffer The proximity of the Cul-de-Sac Plain to the capital city Porte-au-Prince is significant as a large market for agricultural products.
Sustainability 2020, 12, x 6 of 15 The proximity of the Cul-de-Sac Plain to the capital city Porte-au-Prince is significant as a large market for agricultural products.

Sampling and Data Collection
A sample of 10 respondents was randomly selected from a list of 20 farmers who were members of a farmers' association. We planned to interview 200 farmers (belonging to 20 farmers' associations) over a period of 19 days between July and August 2017; fortunately, we exceeded our target. Responses were collected from 214 farmers: 80 from the Ouest department, 63 from the Centre department, and 71 from the lower Artibonite department. The survey questionnaire was developed in English and translated to Haitian Creole. Then, it was transferred into the Kobo Toolbox platform (Harvard Humanitarian Initiative, Cambridge, MA, USA). The questionnaire was administered through the use of Android tablets via a downloaded data collection application, Kobo Collect. The survey had open and close-ended questions. Data collected included: (i) socio-demographics of the respondents, (ii) types of grain produced, (iii) storage and drying techniques, (iv) pesticide usage, (v) pest problems (insects, rodents, mold), (vi) postharvest losses of maize and beans, (vii) market prices for grains at different times after harvest, and (viii) gender roles associated with crop production, storage, and marketing.

Data Analysis
Field data was downloaded from the Kobo cloud server into Microsoft Office Excel 2011, cleaned and coded. Descriptive statistics were calculated (frequencies, cross tabulation) using the SPSS 24.0 (IBM Corporation, 2016, New York, NY, United States). Gender responses were compared using chi square tests using GraphPad QuickCalcs software (GraphPad Software 2018). The factors affecting decisions to store and use insecticides were assessed using the logistic regression model (GLM) in R v.3.5.3. The analysis aimed at finding statistical significance of variables which increased or decreased

Sampling and Data Collection
A sample of 10 respondents was randomly selected from a list of 20 farmers who were members of a farmers' association. We planned to interview 200 farmers (belonging to 20 farmers' associations) over a period of 19 days between July and August 2017; fortunately, we exceeded our target. Responses were collected from 214 farmers: 80 from the Ouest department, 63 from the Centre department, and 71 from the lower Artibonite department. The survey questionnaire was developed in English and translated to Haitian Creole. Then, it was transferred into the Kobo Toolbox platform (Harvard Humanitarian Initiative, Cambridge, MA, USA). The questionnaire was administered through the use of Android tablets via a downloaded data collection application, Kobo Collect. The survey had open and close-ended questions. Data collected included: (i) socio-demographics of the respondents, (ii) types of grain produced, (iii) storage and drying techniques, (iv) pesticide usage, (v) pest problems (insects, rodents, mold), (vi) postharvest losses of maize and beans, (vii) market prices for grains at different times after harvest, and (viii) gender roles associated with crop production, storage, and marketing.

Data Analysis
Field data was downloaded from the Kobo cloud server into Microsoft Office Excel 2011, cleaned and coded. Descriptive statistics were calculated (frequencies, cross tabulation) using the SPSS 24.0 (IBM Corporation, 2016, New York, NY, United States). Gender responses were compared using chi square tests using GraphPad QuickCalcs software (GraphPad Software 2018). The factors affecting decisions to store and use insecticides were assessed using the logistic regression model (GLM) in R v.3.5.3. The analysis aimed at finding statistical significance of variables which increased or decreased the probability to use insecticides. The logistic regression models were tested for fitness using the likelihood ratio (LR) test.

Socio-Demographics and Grain Production, Drying and Storage Practices
The sociodemographic data of the respondents are shown in Table 1. Most respondents were male (63.6%). The majority of farmers were over the age of 30 (81.8%) with 48.6% of respondents being middle-aged farmers (31 to 50 y). Secondary school was the highest level of education for 48.6% of the respondents. Nearly half (47.7%) of the interviewed farmers were married. A large majority of the respondents (93.0%) reported agriculture as their main economic activity. A little over half (55.6%) of the farmers reported owning/renting less than 1.3 ha of land. Most farmers reported maize as the main crop (91.6%) along with beans (88.3%). Quantities of maize and beans produced and stored are shown in Table 2. For maize, about 56% and 90% of farmers produced and stored less than 200 kg, respectively; while for beans, about 70.0% and 91% of farmers produced and stored less than 200 kg, respectively. Further analysis into the mean quantities produced, stored, purchased, consumed and sold is shown in Table 3. While farmers produce for consumption, a big fraction of both crops are intended for sale. Data on grain drying practices are shown in Table 4. Nearly all farmers (98.5%) in all departments dry their grain at home in the sun mainly on a mat/tarpaulin (39.9%). The most significant challenge during drying was rain (79.5%). The sources of grain losses during drying were: insects (57.5%), rodents (38.8%), moisture (34.6%), birds and other animals (47.2%), and theft (8.9%). Data on grain storage practices are shown in Table 5. The great majority of farmers (93.0%) stored grain after harvest mainly using sisal sacks (83.4%), and barrels or metal drums (41.2%). The most reported storage location was a room in the house (87.4%). Storage duration was less than 6 months for 68.4% and 85.9% of farmers for maize and beans, respectively. Use of chemical control to manage pests during grain storage was reported by just over half (55.8%) of the farmers. Of the other half, 14.1% used natural products (plant extracts) and 28.1% did nothing to control pests during storage. Of those who did not report using chemical products during storage, 23.8% were concerned about toxic or health problems, while 28.9% reported the products were either not available or there was no information on how to properly use or find them. Most farmers (77.2%) reported rodents as a major source of grain loss during storage, followed by insects (56.9%).

Decision to Use Insecticides
Some factors were found to influence farmers' decisions to use insecticide during storage and the logistic regression model was well fitted (LR test p < 0.001; Table 6). The use of insecticides to protect stored grain was significantly (p < 0.05) influenced by seed loss (farmers who experienced seed loss during storage were more likely to use insecticides), drying losses (farmers who experienced grain loss during drying were more likely to use insecticides), the quantity produced (farmers who produced more were more likely to use insecticides), and the duration of storage (farmers who stored for more than 3 months were more likely to use insecticides).

Return on Investment if Farmers Stored Their Grain in PICS Bags
Price seasonality (price at harvest and during the lean season) varied between crops and sometimes between the different departments (Table 7). Overall, gross margins on maize price showed a near doubling effect between harvest and lean seasons. Beans showed a price increase of between 120 and 138 Haitian gourdes per kg on average. Using PICS hermetic technologies as a possible intervention for storage among Haitian farmers, estimates of the return on investment (ROI) were calculated. The ROI ranged from 39% to 50% for maize and 36% to 41% for beans. Ouest had the highest ROI for maize (50%) while Artibonite had the highest ROI for beans (41%).

Gender Roles in Postharvest Management Practices
Gender roles in postharvest practices are shown in Table 8. Overall, 66.4% of the farmers reported that the male in the household owns/rents the land (x 2 = 49.64; df = 2; p < 0.0001). Grain drying activities were reportedly female dominated (53.7%) (x 2 = 22.71; df = 2; p < 0.0001). For grain storage activities, 49.2% said that the women were responsible, while 31.2% reported both men and women were responsible (x 2 = 12.86; df = 2; p = 0.0016). For marketing maize and beans, 88.2% (x 2 = 134.54; df = 2; p < 0.0001) and 96% (x 2 = 149.66; df = 2; p < 0.0001) reported that the women were responsible for taking the grain to the market, respectively. It is also important to note that women play an important role in saving and buying seed (43%), though it seems that both men and women are nearly equally responsible for this task (x 2 = 5.420; df = 2; p = 0.0665) ( Table 8). Overall farmers source seed from the market (76.6%), and own seed saving (62.6%) and agroshops (31.3%) (data not shown). Some of the most important saved seed included maize, beans, rice and vegetables. Among 134 farmers who save their own seed for planting, 86.6% said that they incur losses during storage (data not shown).

Grain Production, Drying and Storage
This study showed that smallholder farmers in Haiti rely heavily on agriculture for their livelihoods. Prior work in northwestern Haiti found that agriculture accounted for nearly 90% of farmers' livelihoods [11]. Therefore investment in Haiti's smallholder farmers and the agricultural sector in general is vital to building food security. Land available to each farmer for agricultural activities is scarce and this study shows 55% own less than 1.3 ha. In 2003, the average plot was 2.7 hectares [16,17]; while at the time of this study, the land holdings were less than 1 ha. These data may be compared to other countries such as Malawi, where 90% of total agriculture comes from smallholder farmers who own less than 1 ha [18,19]. Grain production was dominated by small scale farmers producing and storing less than 200 kg of either maize or beans. On average, farmers reported storing 87 kg of maize and 80 kg of beans each season. Most of the produced grain is sold right after harvest to earn cash to meet other household needs [9]. This leaves the households' food insecure and vulnerable to shocks (e.g., climate). The small amount of grain being stored may be kept for consumption and/or seed for the next planting. It is critical to increase productivity for Haitian smallholder farmers in the plain to meet household and national food needs.
The majority of the respondents reported drying their grain at home in the sun on mats or tarpaulins. Some of the associations interviewed had received tarpaulins from the UMCOR project in 2013/2014. The project had distributed 800 tarpaulins to 25 farmers' associations in the Cul-de-Sac Plain [14]. Some farmers pre-dry their grain in the field or dry them on the side of the roads. Drying challenges reported include rain, insects, rodents, birds and animals. Since drying grain is done out in the open and often on the ground, the grain is predisposed to moisture and humidity leading to mold growth, decay and aflatoxin contamination that persists during storage [20]. Presence of aflatoxin contamination in peanuts and maize has previously been reported in Haiti [21,22]. Both of the studies found aflatoxin levels well above the American and European standards [21]. We observed that farmers typically leave maize in the field after maturity thus exposing the grain to weather events such as tropical downpours, which often occur on a daily basis. In a comparable study on the effects of delayed harvest in Uganda by Kaaya et al. 2005 [23], aflatoxin levels increased as much as 7-fold when the maize harvest was delayed by three or four weeks after crop maturity. Training farmers on proper postharvest handling and drying technologies would help improve grain quality.
Sisal sacks were the most widely used storage containers, probably due to their low cost, availability, ability to further grain drying during storage, and ease of transport when selling grain at market [24]. Often second-hand cleaned oil drums were used as hermetic containers for protecting grain from insects and rodents [25]. Bulk storage systems, such as silos, were not commonly used due to the high cost of construction, maintenance (repair), and the large quantities of grain needed to fill them to capacity [19]. However, in this study about 17% of farmers reported using silos for storage. This may be due to the 2013/2014 UMCOR project intervention that distributed 95 silos to 25 agricultural farmers' associations [14]. Introducing cost-effective on-farm storage methods, such hermetic storage bags that have been widely disseminated in Africa and other regions of the world [26], would diversify postharvest management options for farmers in Haiti.
Duration of storage was less than 6 months with farmers storing beans for slightly shorter times than maize. The duration of storage depends on quantity stored and household needs such as cash and food/consumption. Because most maize and beans were sold at harvest, most farmers did not have much grain to store for long. Maize and beans are often sold in the market giving them a "cash crop" status while other crops are kept and consumed in the home [11]. During storage, insects and rodents were the two most important pests. Studies in several countries in Africa have found that rodents are major pests of stored products-coming before or after insects [27]. Higher rodent damage in Haiti may be attributed to the inadequate storage facilities, close proximity of the storage space to surrounding fields and poor storage hygiene [28]. In addition, most farmers used sisal sacks to store their grains, which often draw mice and rats. Rodents pose a threat to grain quality and consumer health because they are disease-vectors [29]. Insect infestations that start during harvesting and drying often continue during storage. Storage insects cause 20-30% loss for cereal crops like maize [7]. Prior work has documented important insect pests in Haiti such as the rice weevil (Sitophilus oryzae L. (Coleoptera: Curculionidae)) and the granary weevil (Sitophilus granarius L. (Coleoptera: Curculionidae)) in stored maize [30]. An extensive checklist was compiled for the Island of Hispanola (Haiti and Dominican Republic) and it was reported that the bean weevil (Acanthoscelides obtectus Say (Coleoptera: Bruchidae)) was found in stored beans [31]. Rodent and insect control measures can be put in place at a household level with integrated pest management training through extension services.

Decision to Use Insecticides
Farmers relied heavily on pest control methods such as chemicals and natural plant extracts to control insects. Elsewhere, in Zambia and Malawi, over 70% of smallholder farmers used synthetic pesticides to control insect pests of maize and beans [32]. The decision to use insecticide was mainly determined by seed loss, drying losses, and storage duration. Farmers who did not report using chemical products during storage were concerned about their toxicity or health risks. Pesticide use in developing countries poses many challenges due to lack of strict regulations and farmers' limited knowledge on their proper use [33]. Cases of pesticide poisoning are known and can be attributed to using inappropriate chemical products, incorrect dosage, and timing and targeting of application [34,35]. Given their reliance on chemicals, Haitian farmers may be receptive to non-chemical methods (e.g., hermetic storage technologies) for protecting their grain.

Return on Investment
The Return on Investment (ROI) using PICS hermetic technology varied by crop and by department. Overall, Haitian farmers would increase their income should they store their maize and beans in hermetic PICS bags for about 6 months. ROI for maize was highest in Ouest (50%) while ROI for beans was highest in Artibonite (41%). The data suggests that it is more profitable to store maize for sale than any other crop in Haiti. Maize is an important food crop in Haiti and is grown across 10 departments [5]. Recently, the departments of Northeast, Northwest, Nippes, South, and Center were affected by a drought between 2018 and 2019 [5]. These intermittent droughts mean that there is always demand for grain on the market. Farmers who are able to store grain in PICS for several months can increase their income [15,26,36].

Gender Implications
Gendered results in this study mainly apply to land ownership and grain sales. Men tend to own the land on which grain crops are grown. Although women are also entitled to own and manage land, they generally end up with smaller plots, due to informal and unfavorable customary land laws that undermine women inheritance rights [37]. In some cases, customary laws radically forbid women from owning land [38]. In Haiti, approximately 70% of Haitian women do not own land titles, 20% own a property jointly with relatives, and only 9% own property on their own [39]. Women are mainly responsible for taking the grain to the market for sale. Sales of produces are considered "women-only activities". The informal markets in which sales occur are usually female dominated [40].
A gender and food security assessment on PICS bags conducted in the Grand'Anse department of Haiti revealed that women heads of households were too busy, hence did not gain from the "konbit"-a cost-saving form of traditional organization of labor through reciprocal assistance among farmers [13]. These women end up paying for additional labor to work their land. Haitian female-headed households account for 81.4% of the agricultural workforce and are most vulnerable to food insecurity and poverty [39,41]. The assessment also highlighted ways in which postharvest management efforts using hermetic PICS bags can positively affect the welfare of women farmers. Women who used PICS bags to store maize can save considerable time during grain storage. They reported using the extra time to engage in other income-generating activities. Such an asset (time) is certainly beneficial for women heads of households working in the grain value chain. Moreover, with grains being available all year round, women were able to keep grains in various unprocessed and processed forms, which contribute to diversifying households' diets and improving their household food and nutrition security [42].

Conclusions
This study revealed that there is need for preharvest and postharvest intervention to improve production and food security in Haitian. Increasing productivity of farms in the Cul-de-Sac Plain of Haiti will significantly increase the quantity of grain produced and eventually stored, thus cushioning Haitians from shocks of food insecurity. Farmers are likely to store more if they produce more. Strengthening postharvest extension and education activities can help Haitian farmers improve grain handling and storage to reduce postharvest losses. Introducing drying technologies adapted to smallholder farms would improve grain quality and reduce aflatoxin contamination. Scaling-up chemical-free and cost-effective storage technologies (e.g., hermetic bags) to supplement silos may help mitigate storage losses and reduce chemical use. Adoption of hermetic bags will improve overall livelihoods of the farmers through the additional income. Targeted interventions that are gender sensitive would be needed to empower women since they play such a pivotal role in the grain value chain (e.g., producers and traders). Funding: This research was funded by an anonymous donor who made a donation to Purdue University, the D. Thomas Woods Fund, and the PICS3 project (Grant number OPP1038622) funded by the Bill and Melinda Gates Foundation to Purdue University.