Assessment of Good Agricultural Practices on Cocoa and Coffee Farms in Northern Haiti

Haiti is the poorest country in the Western Hemisphere and presents a unique scenario for the food and agriculture industry, because there is no food safety legislation. The application of Good Agricultural Practices (GAPs) leads to improvements of quality, safety and sustainability of agricultural products. The purpose of the study was to assess the status of Good Agricultural Practices (GAPs) in cocoa and coffee farms in Northern Haiti. A general survey captured information about the farmer and the farm, and an audit checklist was used to assess compliance to GAPs. A total of 11 farms (n = 11) were audited, of which 7 were cocoa farms (64%) and 4 were coffee farms (34%) in the regions of Dondon, Limonade and Milot. Average overall audit scores for coffee farms (73%) were higher than for cocoa farms (55%). Farms affiliated with a cooperative scored higher (78%) than those that were not part of a cooperative (55%). The sections of the survey on “Practices related to premises and production site”, and the “use of agricultural inputs and chemicals” received the lowest scores but were confined to the cocoa farms. “Record keeping” plus “distribution, transportation, and traceability” were cause for concern with both the cocoa and coffee farms. Critical non-conformances included the access of livestock animals and domestic pets to processing and storage areas, the lack of control in the application of agricultural chemicals, a lack of safeguards on equipment and elevated surfaces, and washing of fresh cocoa beans to remove the

by multiple stakeholders to determine the minimum food safety standards, but they are not mandatory. Among the most recognized standards by retailers and manufacturers are the British Retail Consortium (BRC) global standards, Safe Quality Food (SQF), USGAP, ISO19011:2018, GlobalGAP and Canada GAP, all under the umbrella of the Global Food Safety Initiative (GFSI) through its benchmarking process [14]. Though there are benefits in following and monitoring GAPs, the challenges in implementing them must be considered, especially for small farmers of developing countries. The impact of implementing food safety and quality standards on farms is dependent on farm size, agricultural sector, and country [15]. Meeting international standards gives access to international markets, and it also helps improve hygiene and health of workers, modernize production processes, increase yields and quality of product [16] and enhance financial gains in the long run [15]. However, in developing countries food security is a priority and the prices of food need to be kept affordable [17].
The probability that small-scale farmers pursue certification on food safety and quality standards is very low due to the high costs and complex technical knowledge associated with stringent standards [16] [17]. However, not doing so might exclude them from international markets because the standards are gaining more and more importance in trade [16]. Some of the challenges associated with certification are: the lack of funds for initial investment, certification fees and laboratory analyses [16]; the use of traceability and record keeping systems [15]; and the development of technical skills to manage, supervise and maintain the standards [17]. In order to overcome these challenges, it has been suggested that policies are needed allowing small-scale farmers to have access to financial and technical assistance in developing countries [15]. Another issue is the low level of literacy among the Haitian population that may build an additional barrier to comprehending and implementing any record keeping and management system.

Cocoa and Coffee
Cocoa and coffee beans are high-value commodities, driven by the chocolate confectionery industry, baking industry and ice cream industry for cocoa [18], and by the hot beverage industry for coffee [19]. It is expected that the demand for these products will continue to have a stable growth worldwide [20] [21]. In spite of the growing demand over the past century, Haiti has dropped from the third largest exporter of coffee in 1949 to one of the smallest producers and ex- porters. (39) The Canadian Food Inspection Agency (CFIA) [22] has identified the potential hazards in cocoa beans and related products. There are chemical hazards (e.g., pesticide residues, undeclared allergens and contamination by non-food chemicals) and physical hazards (e.g., extraneous material) that can be addressed with the implementation of GAPs.
Cocoa and cocoa products also have Salmonella spp. as a biological hazard [22] and ochratoxin A (OTA) as a chemical hazard [23]. Salmonella spp. present in livestock and domestic pets, are transmitted to food through the fecal-oral route or contaminated water [24]. The source of some outbreaks has been traced back to cocoa beans [25]. Low water activity (aw) and high fat content of cocoa beans make the bacteria viable for prolonged periods. Contamination can occur during pre-processing activities (e.g., harvest, fermentation, drying, storage) on the farm due to poor hygienic conditions. It is best to control this potential contamination with good agricultural practices (GAPs) at the farm level [25], and through thermal processes such as roasting and conching in the food industry [26]. OTA is produced by Aspergillus carbonarius and Aspergillus niger, and is one of the most common mycotoxins found in cocoa beans [27]. Cocoa beans can become contaminated because of damaged pods, poor fermentation, slow drying and poor storage [28], but it is in the latter two stages that OTA is produced at higher levels due to low water activity (aw) and lower microbial competition [27]. It has been reported that the processing steps that remove the most OTA in cocoa beans are shelling and alkalinisation [29]. Even though chocolate products are minor source of OTA, consideration must be given as low levels in large quantities can be consumed by children [28].
Contamination of coffee beans with various microorganisms, particularly filamentous fungi, and the subsequent potential production of ochratoxin A is also food safety concern. It can occur at the farm level and throughout the production chain [30]. Microbial loads are frequently reduced during the roasting process. However, precautionary measures are essential at each stage of the production chain to minimize the risk to consumers.
The purpose of the study was to assess the status of Good Agricultural Practices (GAPs) in cocoa and coffee farms in Northern Haiti, and to identify the most common practices that would compromise the food safety and quality of Haitian cocoa and coffee beans.

Sample Selection
The study was carried out in Northern Haiti (Figure 1) with the assistance of the Haiti Food Hub (HFH). HFH is a local enterprise run by Haitians whose mission is to promote a social agri-business model that is financially self-sustainable by helping small-scale farmers supply quality products, while improving social and environmental conditions. The project was reviewed by the University of Guelph Research Ethics Boards (REB) and granted approval (REB no. 15AP010). Selection criteria mandated that participants must do so on a strictly voluntary basis, be of legal age, and have been actively involved with the Haiti Food Hub prior to this study. There were also language stipulations (Creole, English, or Spanish), due to the availability of interpreters. As such, French-speaking farmers were excluded from the survey pool. A letter was sent to the HFH agronomist inviting interested parties from the 70 cocoa farmers and 70 coffee producers affiliated with the HFH to participate in this study. Based upon these criteria, a sample of 7 cocoa farms and 4 coffee farms was obtained after recruiting participants.
Farmers who expressed interest were given a more detailed explanation of their involvement in the project. After giving their oral consent, a consent form in Creole was distributed. While this was a somewhat low number of participants, it was considered appropriate in providing an initial baseline in identifying current practices employed on typical small-scale farming operations for cocoa and coffee. There were also time and budgetary constraints that made it impractical to survey a larger sample size. It was considered that the results would provide directionality for future studies and preferably actions to improve the agricultural practices and economic outlook for Haitian agricultural producers.

General Survey
A general survey was adapted from Ganpat et al. [31]  open-ended whenever specific information was required. A code was assigned to each participant based on the crop of interest, followed by a two-digit number to indicate the number of the farm belonging to that category. "CC" was assigned to cocoa producers and "CF" to coffee producers. The codes CC01 to CC07 were assigned to the seven participating cocoa farms, and CF01 to CF04 designated the four coffee farms involved. Appendix A shows the form that was used to collect the data described above.

Audit Checklist
An audit checklist was developed considering FAO guiding principles of good agricultural practices [13]. The audit checklist was developed integrating information from Ganpat, et al. [ Each section contained several items that described the specific practices to be assessed and a severity value associated with it, depending on the potential impact on food safety. The number of items per section varied, but the weight of total points was evenly distributed throughout the sections, ranging from 9% to 13% ( Figure 2). The audit checklist had a total of 57 items.
Each practice was assessed on-site through observation and/or objective evidence provided by the farmer (e.g., records, labels) and, in some cases, photographs were taken to document the objective evidence.
Items were classified and assigned a numeric value using the criteria outlined in Table 1.
Each item was pre-assigned a numeric value based on the severity of not following the practices as described: 1-Low (minor); 2-Medium (serious); 3-High (hazardous). This was done in order to give more points to high severity practices that were compliant. The rating of each item was obtained by multiplying the severity value times the numeric score assigned for the status of compliance (value = 3) or non-conformance (value = 2, 1 or 0) of the practice. If the specific item was not applicable to the farm, "N/A" was marked in the rating column, a rationale was provided in the "Comments" column and it was not considered in   All practices were followed as described and objective evidence was observed. 3 Minor Non-Conformance All practices were followed as described but no objective evidence was observed; some of the practices were not followed as described.

Non-Conformance
There was evidence that most practices were not followed as described. 1 Critical Non-Conformance Controls were not in place and the practices followed posed a public health risk if no corrective actions were implemented. 0 final score. To calculate scores, total points and maximum points were used [Score (%) = (Points obtained/Maximum points) × 100]. Maximum points were calculated by assuming that all practices were compliant, excluding those that were not applicable (N/A). Each section score (%) was calculated dividing the points obtained per section by the maximum points per section. The overall audit score of the farm (%) was calculated dividing the sum of points obtained in all sections by the sum of maximum points of all sections. Appendix 1 contains the complete audit checklist used to assess the farms.

Farmer Information
Information gathered in the survey that is specific to the farmers is summarized in Table 2. A total of 11 farms (n = 11) were audited, out of which 7 were cocoa   characterized by coffee producers, and the region of Milot by cocoa producers.

Farm Information
( Figure 1) The farm size of all participants was less than 5 hectares, and 71% of cocoa farms were between one and five hectares, whereas 75% of coffee farms were less than 1 hectare. More than 90% of the farmers owned their land and had tree crops (e.g., mangoes, plantain, sour sop, breadfruit) and root crops (e.g., yams) growing on the farm. None of the cocoa producers that participated in the survey was affiliated with a cooperative nor was exporting their products. However, 75% of coffee producers (CF02, CF03 & CF04) were members of a cooperative and 50% of them (CF02 & CF03) exported their "Certified Organic" product to the United States (Table 3). Livestock animals were not observed on any of the 4 coffee farms visited. Pigs, chickens, goats, and cattle were seen in 4 of the 7 cocoa farms, regardless of farm size.

Audit Scores
The audit results are reported in Table 4(a) and To aid in comparison of the data, Figure 3 provides a graphical representation of the overall average audit scores from Table 4(a) and Table 4(b) for the cocoa and coffee farms.
As can be seen, the scores for the coffee farms were consistently higher than those for the cocoa farms, with one exception. It was only in Section 6 (Equipment and Tools) that the cocoa farms received a score greater than the coffee farms.  Survey sections with lower scores appear closer to the center of the circular graph in Figure 3. On average, cocoa farms showed the lowest overall scores in four sections of the survey. These were, in numerical order: Section 1-Premises and production site (33%); Section 3-Use of agricultural inputs and chemicals Typically, the coffee farms were compliant or showed only minor non-compliance in all sections except for Section 8-Human welfare, health, and safety; and Section 9-Record keeping. This is in sharp contrast to the cocoa farms which showed minor to major levels of non-conformance in all but Section 4-Water Quality; Section 5-Harvest and post-harvest operations; and Section 6-Tools and equipment.
In Figure 4, the survey results for the eleven farms have been separated based on their affiliation or lack of affiliation with a cooperative. Three of the four coffee farms were part of a cooperative, while only one of the seven cocoa farms was a co-op member. Figure 4 shows that overall audit scores of farms affiliated with a cooperative were higher (78%) than those that were not co-op affiliated (55%). The only exception to this trend was Section 6-Equipment and tools, where the cocoa farms showed very high scores even though only one of them was affiliated with a cooperative.

Items by Category and Section
The scores reflect the number of compliant items and non-conformance items (minor, major, and critical). Non-applicable (N/A) items have been removed from the scoring. A summary of the number of items falling into each level of compliance is presented for each farm in Table 5. Percentage scores have been included as a means of comparing their scores with the levels of conformance.  On average, coffee farms had more compliant items, with an average of 18.0 compliant items per farm compared to only 7.9 for the cocoa farms. The average number of minor non-conformance items was higher for the coffee farms (13.5) than for the cocoa farms (9.0). There was an average of 11.5 items in the major non-conformance category per coffee farm compared to 19.3 items per cocoa farm. Two of the four coffee farms had one critical non-compliant item which gave the coffee farms an average of 0.5 items in this category. Six out of the seven cocoa farms had at least one critical non-conformance item, for an average of 1.9 critical items per farm. Table 6 provides a break-down of the levels of compliance by survey section. Compliant Items: The section with most compliant items for all farms was Section 5 (Harvest and post-harvest operations) and the one with the least compliant items was Section 9 (Record keeping. It is noteworthy that in the case of the seven cocoa farms, there were no compliant items in Section 7-Distribution, transportation and traceability, and Section 9-Record keeping. Individual responses to each survey item have not been included here, in the interest of brevity. Minor non-conformance items: The proportion of minor non-conformance items in each section was not more than 33% for all farms. Minor non-compliance items were evenly distributed throughout all sections for coffee farms, whereas cocoa farms had a high proportion of minor non-compliant items in Section 4 (Water quality) and Section 5 (Harvest and post-harvest operations). No minor non-conformances were observed in Section 5 for coffee farms.
Major non-conformance items: There was a great disparity in the proportion of major items in the sections of all farms. They were very low (i.e., lower than 10% in Sections 3, 4, 5 and 6) or very high (i.e., higher than 50% in Sections 1, 7 and 8). In the case of cocoa farms, most sections had a higher proportion of  major items than coffee farms. Also, major items in cocoa farms averaged 71% for Section 7 (Distribution, transportation and traceability), 60% for Section 1 (Premises and production site) and 57% for Section 8 (Human welfare, health and safety). No major items were observed in coffee farms for Section 3 (Use of agricultural inputs and chemicals) or Section 5 (Harvest and post-harvest operations). Critical non-conformance items: Critical non-conformance items had the largest impact on scores because no points were assigned to the specific item. On cocoa farms, critical non-conformance items were noted in Sections 1, 2, 3, 4 and 8. Section 1 (Premises and production site) had the largest proportion (14%) of critical items, followed by Section 3 (Use of agricultural inputs and chemicals) and Section 4 (Water quality) with 7% critical items in each section. All critical non-conformances in coffee farms were identified in Section 8 (Human welfare, health and safety) concerning the lack of safeguards on equipment and around elevated surfaces (item 8.2), leading to a high risk of accidents in the processing area of the cooperative.
Non-Applicable Items: Table 6 shows that 30% of all items of the audit checklist were not applicable (N/A) to the farms audited. Section 3 (Use of agricultural inputs and chemicals) and Section 6 (Equipment and tools) had more than 60% of their items classified as N/A. In general, cocoa farms had more N/A items than coffee farms.

Frequency of Items
Each section was divided into specific items, having a total of 57 items in each  Table 7.
Among the more frequent compliant items for all farms were the inspection of product before and during harvest (item 5.3) and the removal of extraneous materials and damaged product (item 5.4). Most farms were also compliant in inspecting the site prior to harvest (item 5.1). The most common minor issue in all and lack of training records (item 9.1).
Major non-conformances were more frequent in the identification of areas of the production site and premises and the lack of designated areas for storage, equipment, chemicals and product (item 1.1), and the lack of cleaning instructions and cleaning practices (item 2.7). Cocoa farms frequently had more major issues in important areas such as traceability (items 7.3, 7.4 and 7.5), record keeping (items 9.3, 9.5 and 9.9) and lack of pest control programs (item 9.10).
Coffee farms, however, had major issues in items related to human welfare, health and safety, specifically in the availability of first aid kits and personal protective equipment (item 8.1), and awareness of transferrable diseases to food (8.5).
Critical items were more frequent in cocoa farms than in coffee farms. The risk assessments of production site and premises (item 1.4) and adjacent areas (item 1.2) of cocoa farms were critical because no fences were observed and livestock animals and domestic pets had access to the production site plus processing and storage areas, and garbage was all over the farm. Other critical items in cocoa farms included washing of fresh cocoa beans to remove the mucilage with water that had not been treated or tested for potability (item 4.5), and the lack of control in the application of agricultural chemicals and harvest intervals (item 3.2) ( Table 7). The only critical item found in coffee farms (item 8.2) has been addressed previously. Table 7 reports that the most frequent non-applicable items for all farms were the cleaning of agricultural chemical application equipment (item 6.4), disposal of agricultural chemical containers (item 2.3), condition and cleanliness of scales (item 6.5) and calibration records (item 9.11).

Discussion
Poor scores in Section 1 (Premises and production site) are the reflection of the lack of fences installed, poor construction materials, garbage observed on the production site, and presence of livestock animals and domestic pets in farm and around the house. Cocoa farms had more of these major and critical non-conformances, where the main risk was that livestock was not contained in a pen, and dogs were running around where beans were processed or stored. More than 50% of cocoa farmers had livestock in their farms (Table 3) but lacked good husbandry. It is of particular concern because animals can carry Salmonella spp., and a lack of segregated areas increases the risk of cross-contamination [25]. Permanent structures (e.g., farmers' houses) tended to be of poor construction and were not ideal to store finished product. Unfortunately, in the case most farmers in this  (Table 3). They brought their harvest to the cooperatives, which were cleaner and had decent permanent structures with fences and walls to keep animals away. This suggests that being part of a cooperative could help reduce risks related to premises, considering that they are in good conditions.
Sanitation and waste management practices varied widely (33% to 81% conformance) among the cocoa and coffee farms, which was cause for concern.
Items related to the disposal of chemical containers were more frequently "N/A" The section that showed to be less applicable to the farms assessed was Section 3 (Use of agricultural inputs and chemicals) because not all of the farmers used fertilizers or pest control products. All items of Section 3 were not applicable to 6 farms (Table 4(a) and Table 4(b)), deeming the whole section "N/A". The farms that used agricultural chemicals either scored low (i.e., cocoa farms) because of the lack of control in the application, insufficient training and purchase from questionable suppliers, or scored high (i.e., coffee farms) because they exported "Certified Organic" product and the cooperative provided adequate training. Pests and diseases have a direct impact on yields and profits [35].
Therefore, farmers should consider the use of agricultural chemicals, focus on resistant varieties, or follow stringent practices to prevent, control and eliminate pest and diseases [36]. Similar results were reported in a study among cocoa farmers in Ghana, where most farmers did not spray their farm with pesticides because it was an expensive practice [35]. Other challenges of using agricultural chemicals are the use of banned or unapproved chemicals, inefficient application, and lack of training and monitoring activities [35] [37]. On the other hand, not using agricultural chemicals on the farms assessed could be seen as reducing the risk of pesticide residues in the finished product and lowering risks to the health and safety of workers. In addition, there is a niche market for "sustainable" products with a growing trend in cocoa and coffee products [18] and sev-A. Navarro et al.
Regarding water quality, it was found that farmers relied on rain for their crops, as there were no irrigation systems on any farm visited. They also accessed water from rivers, natural springs, or private wells. None of the rivers were accessed by the farmers in this survey because farmers mentioned that the rivers were far from their farms. However, downstream contamination was observed as people bathed, washed their clothes and refreshed their animals. Private wells were out in the open, few of them were fenced, and all had manual pumps to bring the water up. As stated previously [5], it is estimated that less than 70% of the population in Haiti has access to potable water sources. This was difficult to verify because neither the farmers nor the cooperatives conducted potability tests. In most cases, such items were classified as minor non-conformances.
Item 4.5, concerning the use of potable water, was one of the two most frequent non-conformances in cocoa farms because some farmers rinsed the cocoa beans after breaking the pod to remove the mucilage, speed up the drying process, and sell the beans sooner. By doing so, the fermentation of beans would not take place and, most important for this section, the finished product would have been rinsed with water of questionable potability. Cocoa is considered a cash crop because it has no "subsistence value" to the household [38] and when farmers have income problems, they tend to compromise the quality of the product to get cash quicker [37]. It certainly was the case in this study, where 91% of participant's income came from sources other than farming (Table 1).
Section 5 (Harvest and post-harvest operations) received the highest scores in all farms. Scores ranged from 70% to 100%, and all coffee farms scored 100%. High scores in Section 6 (Equipment and tools) were a combination of compliant items and "N/A" items (Table 6). Also, this section was the only one in which cocoa farms scored higher than coffee (Figure 3). This could be explained because cocoa farmers only used their machetes to harvest and process, whereas coffee farmers used a wooden mortar and pestle or diesel-operated grinders.
Distribution, transportation and traceability practices faced several challenges.
Interestingly enough, farms that were affiliated with a cooperative averaged 74% against 41% that were not affiliated ( Figure 4). Farmers not affiliated with a cooperative stored their product at home and either waited for a cooperative to come and buy it from them, or they would send the product to a market or association. Three types of transportation were observed: by animal, usually a mule those means of transportation exposed the finished product to environmental elements such as dust, rain, or extreme heat. In general, vehicles were not well maintained. Farmers did not identify their fields and harvested product, nor did they keep records of their farming activities, which makes it challenging for traceability purposes. Traceability systems are important for maintaining low levels of pesticide residues, heavy metals, and mycotoxins [36]. For the chocolate industry, it has been suggested that traceability systems could be useful beyond legal implications of food safety, such as flavour profile and better selection of cultivars [39]. Being a member of a cooperative helped having better records and proper lot identification of the product, as Figure 4 shows high scores in Section 7 (Distribution, transportation and traceability).
There was not much awareness among participants related to health and safety practices. For instance, there was no first aid kit or personal protective equipment available. There was no awareness of transferrable diseases to food, problems associated with the lack of running water, and the implications of poor hygiene practices. Often, toilets and hand washing stations were not fully equipped. Similar attitudes and practices among street vendors in Haiti were reported by Samapundo et al. [9]. The only critical item among coffee producers was in Section 8 (Human welfare, health and safety) because of a lack of safeguards on equipment and elevated surfaces within processing areas. When the coffee cherries were received, they were put in a basin where the good ones will sink and the bad ones will float and be discarded. The good coffee cherries were then transported in buckets by hand to a higher basin. No railings were observed here, even though the height was more than 3 m, and the edge of the basin was no wider than 90 cm. The operator that controls the feed of cherries to the grinder would stand in front of it on a very unstable surface, and the equipment had no safeguards. Awareness is needed to minimize risks of accidents and job injuries. One of the benefits of following GAPs is the improvement of hygiene, health and safety of workers [16].
It was expected that Section 9 (Record keeping) would have one of the lowest scores of the audit (47%). Most farmers did not record training, cleaning or harvesting activities, and water treatments. Most had no visitors' policies nor pest control programs in place. The two coffee farms that exported green coffee beans (CF02 & CF03) had a visitors' policy and pest control program in place because they had been audited before. Record keeping was definitely a challenge for small-scale farmers [15].

Conclusions
There was a low level of compliance of good agricultural practices (GAPs) among cocoa farmers, and a moderate level among coffee farmers. The root cause of the non-conformances, regardless of the commodity, was either related to poor physical and organizational infrastructures, or a lack of technical training. A factor that seemed to improve the level of compliance of GAPs was affiliation with a cooperative which tended to provide training to its members. Cooperative members had premises that were in better conditions than the farmer's house, and kept more records than the individual farmers who were not associated with a cooperative.
While some of the issues addressed in this study could be corrected by a Despite the limitations mentioned above, this research project has provided the first step to understand and effect positive development for the cocoa and coffee producers of Haiti, and the people of that nation in general. These results have been communicated to the Haiti Food Hub agronomist and will be used as a metric from which to gauge future progress in the identified areas of concern and non-compliance.

Author Contributions
The three authors of this paper worked as a close-knit group with a high level of involvement through all aspects of the research project. The following is a summary of their contributions. Abraham Navarro was a graduate student in the

Conflicts of Interest
The authors declare no conflict of interest. The Guelph Food Technology Centre d) Water sources are tested for total coliforms and E. coli at least once prior use and once during season from all sources (except municipal water). Strongly suggested for agricultural water (i.e., used for irrigation and preparing solutions of fertilizers and chemicals), and mandatory for water that is in contact with product (i.e., cleaning, fluming, cooling, drenching). e) Water used for cooling, drenching, fluming or washing product, and misting, cleaning operations comes from a potable source and is either kept potable or the product gets a final rinse with potable water. If water is treated for potability (i.e., with chlorine or any other alternative method), it is done following a validated procedure, pH and chlorine concentration is monitored.
f) Backflow prevention devices are in place where water is used for agricultural chemical application.

7) Section 7 -Distribution, Transportation and Traceability
a) The cargo area of the vehicle that will transport the product is inspected for cleanliness, is free of pests (or evidence of infestation) and well maintained. The vehicle is constructed in a way that environment elements will not contaminate the load, or that a protective covering (e.g., tarp, plastic wrap) secures the load.
b) Outgoing product is inspected before loading it into the vehicle. When product is loaded, it is placed in a manner that it does not come in contact with other materials that have the potential risk of cross-contamination (e.g., aller- f) Toilets and hand washing facilities are available, clean, easily accessible and fully equipped (toilet paper, soap, paper towels, hand wipes, hand sanitizer, garbage). There are signs that remind users to wash hands after using the toilet.

Appendix B
Audit Questions-Assessment of Good Agricultural Practices in Haiti A scoring matrix was constructed based on the following nine sections. A severity ranking for each item was assigned to each item.

1) Section 1-Production Site and Premises
a) There is a layout of the production site and premises, which identifies: Equipment, operations, storage, personal hygiene facilities, waste disposal containers, water sources hand washing stations, pest control devices.
b) Adjacent areas to production sites have no risk of cross-contamination from crop inputs (e.g., fertilizer and/or agricultural chemicals) or other contaminants (e.g., air, water or soil), have proper drainage and no standing water was observed. NOTE: Follow assessment form.
c) Permanent structures are constructed with doors and windows that fit, and without crevices, holes or leaks in walls or ceiling; the surroundings areas are kept clean to discourage harboring of pests (garbage, long grass, unused equipment). d) Interior buildings are clean, without access to animals (domestic or wild) and no signs of pests are observed; not used for slaughter; lighting is adequate and protected where product is handled; has adequate drainage, pipes are not leaking and there is no risk of condensation. e) An area is dedicated for breaks and lunch, and is separate from areas where product is handled, and another area for storing personal belongings.
2) Section 2-Sanitation and Waste Management a) Sufficient waste disposal containers are located throughout the production site and interior buildings. They do not pose a risk for finished product, water sources, packaging materials or compost. They are covered where pest or animals can have access to them, and identified according to the type of waste (i.e., recycling, garbage or compost) b) Garbage is taken out when containers are full; they are cleaned at least once a month in a separate area that would not pose a risk of cross contamination. c) Before discarding agricultural chemical containers, they are rinsed at least 3 times; containers are not reused and are disposed according to local regulations. d) Toilet waste and waste from hand washing stations are not disposed in production site but into a septic system, municipal sewage or through a service and non-porous materials are cleaned before used.
6) Section 6-Equipment and Tools a) Equipment is not used for slaughter operations b) Equipment and tools that have contact with food are easy to clean, made out of non-porous materials, and are not a potential source of contamination.
They are located or installed where there is sufficient space between floors and walls that allows easy cleaning, and stored in a manner that avoids cross-contamination of product or water sources. c) Equipment is inspected before its use. Equipment (agricultural chemical d) Agricultural chemical application equipment is flushed or rinsed in between applications and in a way that water sources or production site is not contaminated.
e) Cross contamination is avoided by dedicating cloths for wiping product only, and cleaning scales in between use of weighing agricultural chemicals and product.

7) Section 7-Distribution, Transportation and Traceability
a) The cargo area of the vehicle that will transport the product is inspected for cleanliness, is free of pests (or evidence of infestation) and well maintained. The vehicle is constructed in a way that environment elements will not contaminate the load, or that a protective covering (e.g., tarp, plastic wrap) secures the load.
b) Outgoing product is inspected before loading it into the vehicle. When product is loaded, it is placed in a manner that it does not come in contact with other materials that have the potential risk of cross-contamination (e.g., allergens, livestock, and agricultural chemicals).
c) Shipments are recorded (e.g., date, product description and ID, quantity, vehicle inspection, carrier, destination, customer and person responsible). Training is done by a competent/certified person. c) Training is provided to employees before the season, to new employees, when good practices need to be reinforced. Training is provided in a language that the employee understands. Agricultural Sciences tention as soon as possible, and is removed from activities that may contaminate the product (e.g., handling fresh product). Bandaged wounds are covered with waterproof materials (e.g., rubber gloves).
f) Toilets and hand washing facilities are available, clean, easily accessible and fully equipped (toilet paper, soap, paper towels, hand wipes, hand sanitizer, garbage). There are signs that remind users to wash hands after using the toilet.