Knowledge and practice of farmers on safe use and handling of chemical pesticides: in the case of Tigray, Northern Ethiopia

Abstract

Globally, more than 2 million tons of pesticides are used every year, and many farmers use pesticides in violation of the recommendations, which may lead to adverse health effects. The study aimed to assess the knowledge and practices of farmers regarding the handling of chemical pesticides. A semi-structured questionnaire was developed and cross-sectional data were collected from 306 sample farmers using face-to-face interviews. Descriptive statistics and χ² tests were used to analyze the collected data. Farmers listed the 10 most pesticide-consuming crops and 22 common pesticides used in their farms, including abamectin (“extremely hazardous”) and P,p′-DDT. Among the participants, 53.6% and 58.8% of the farmers received training and had awareness of the negative impact of pesticides, respectively. Of the total sample respondents, 54.9% never used personal protective equipment and 50.3% never read pesticide labels. Moreover, 77.1% and 74% of respondents stored pesticides anywhere in the living room and incorrectly disposed of empty containers, respectively, while 34.7% of the farmers harvested crops within 7 d after spraying. The use of personal protective equipment is significantly associated with farmers’ education (χ² = 6.923; P = 0.031) and experience of pesticide use (χ² = 9.487; P = 0.023). Generally, most of the farmers had poor practice in handling pesticides, and hence, it needs consistent follow-up and recurrent training to minimize the negative impact of pesticides, and to adopt integrated pest management as pest control method.

Introduction

Globally, an average yield of 35% to 45% of food and fiber crops is lost due to agricultural pests including weeds, fungi, bacteria, nematodes, insects, and rodents (Sharma et al. 2019, Gebretsadkan et al. 2020), and farmers have been using pesticides to control such crop pests (Mengistie et al. 2017, Shammi et al. 2020). In 2019, around 2 million tons of pesticides were used globally for pest control (Sharma et al. 2019), and this was increased to 3.69 million tons in 2022 with Brazil was the largest pesticide consumers country worldwide followed by the United States (Statista 2024). In developing countries, including Africa, the use of pesticides is increasing drastically, despite Africa having the lowest (2% to 4%) market share of pesticides (Sharma et al. 2019). Despite the fact that pesticides play a key role in controlling pests, many farmers, particularly in developing countries, are extensively using pesticides in violation of the label for pest control in agriculture (Mengistie et al. 2017, Sai et al. 2019).

The misuse and precarious handling practices of pesticides may lead to higher exposure, resulting in adverse effects on the health of humans, animals, and other beneficial organisms (Gesesew et al. 2016, Jallow et al. 2017, Afata et al. 2021, Mehmood et al. 2021). Pesticide residues may persist in contaminated food, soil, water, and air (Zhang et al. 2015, Mergia et al. 2021, Tudi et al. 2021). Pesticide applicators can be exposed to those chemicals via different routes, including ingestion, inhalation, and dermal contact (Gesesew et al. 2016, Requena et al. 2019, Tessema et al. 2021, Rasool et al. 2022). Such exposures can cause human health problems including carcinogenicity, endocrine disruption, abnormal reproduction, and a change in genetic makeup (Atmaca et al. 2018, Hertz-Picciotto et al. 2018, EFSA et al. 2019, Sharma et al. 2019, Shentema et al. 2022, Tessema et al. 2022). The fungicide mancozeb, for instance, can cause carcinogenicity and affect females’ reproductive performance by blocking ovulation processes (Atmaca et al. 2018). Pesticide poisoning affects around 1 million people (Sai et al. 2019) and hundreds of thousands die each year worldwide (Lekei et al. 2014). Although the consumption of pesticides is low in developing countries when compared to developed countries (Sharma et al. 2019), the health risk of pesticide residue is much higher in developing countries and small-scale farmers compared to large-scale farmers in developed countries (Mequanint et al. 2019, Sai et al. 2019). This is due to the low knowledge of farmers on the behavior of pesticides, lack of use of personal protective equipment (PPE), mishandling and inappropriate storage, and spraying at above-label rates of pesticides in developing countries (Lekei et al. 2014, Jallow et al. 2017, Sai et al. 2019). The misuse and improper handling of pesticides occur primarily at the time of mixing, and during storage and application of pesticides (Sharma et al. 2019). Moreover, the misuse of chemical pesticides can lead to secondary pest outbreaks, the destruction of nontarget organisms, and an increased cost of pest control (Jallow et al. 2017).

In Ethiopia, due to the expansion of modern agricultural farms, consumption of chemical pesticides has shown a 3-fold increase, from 1,440 tons in 2001 to 4,586 tons in 2013 (Negatu et al. 2021). On average, 251 kg of pesticides per hectare was used annually during this period (Negatu et al. 2016). Farm workers in Ethiopia had low knowledge regarding pesticide use, such as inadequate use of PPE, incompatible pesticide mixing, overspray of crops, and disposal in nearby farm fields (Negatu et al. 2021). Studies conducted around Lake Tana (Agmas and Adugna 2020) and Lake Ziway (Mergia et al. 2021) showed that 80.16% and 95% of the farmers had poor pesticide handling practices, respectively. The majority of the respondents (63%) from the Central Rift Valley of Ethiopia did not use PPE during pesticide mixing and spraying (Loha et al. 2022). Negatu et al. (2021) reviewed pesticide residue contamination in surface water, animals, soil organisms, and food items, and they reported a concentration of 0.11 to 138 µg/liter of diazinon, 2,4-D, malathion, and fenpropimorph pesticides particularly in surface water samples collected from around Addis Ababa, the capital city of Ethiopia. Similarly, tissue culture studies of fish and soil sample analysis showed contamination with organochloride pesticides such as DDT and endosulfan. This high quantity of pesticide residues found in drinking water, soil, and food may contribute to the 16% overall prevalence of chronic diseases in Ethiopia (Negatu et al. 2021). Moreover, farmers in the Central Rift Valley of Ethiopia reported health risk symptoms such as nausea, vomiting, headache, skin and eye irritation when applying pesticides (Loha 2022).

To date, apart from the scattered reports in the central areas of Ethiopia, there is little documented information concerning farmers’ level of knowledge and practice on the proper use of pesticides. Based on the previous studies, the types and frequency of pesticides commonly used, the time and methods of pesticide spraying, the major crops on which pesticides are frequently applied, the preharvest re-entry interval, and the pesticide storage and empty container disposal mechanisms are not addressed very well. Furthermore, it remains vague whether the farmers in Ethiopia, particularly in the Tigray region, are using the recommended pesticide application and handling procedures. Thus, the purpose of the current research was to explore the knowledge and practice of farmers on the handling and use of agricultural pesticides, and the result may give a baseline for researchers and decision-makers for future interventions in the proper handling of pesticides and management of pesticide exposures.

Survey Methods and Data Analysis

Study Area Description

The study was conducted between 25 May and 19 June 2021, in Rama-AdiArbaete, Keyh-Tekli, Kilte-awlaelo, and Raya-Azebo Districts of the Tigray region, Northern Ethiopia. The study areas were selected purposively based on their irrigation potential and their extensive growing of horticultural crops such as fruits and vegetables (Fig. 1). From each irrigation area, one village was randomly selected as a study site, and the 4 selected villages were Hadush-Adi, Adiha, Genfel, and Wargba from Rama Adi-Arbaete, Keyh Tekli, Kilte awlaelo, and Raya Azebo districts, respectively. According to the report of the Tigray Region Bureau of Agriculture, Rama Adi-Arbaete is the largest source of fruits such as orange and mango, followed by Keyh-Tekli, while Raya-Azebo is the largest irrigation farm in Tigray known for vegetable crop production. In the first 3 areas, the source of water for irrigation is a dam reservoir, whereas in Raya-Azebo, the source of water is underground water. In all the study areas, farmers manually spray pesticides using a 15 to 20 liter knapsack sprayer, and the mixing activities take place on the farm near the water sources. In Tigray, pesticides are extensively used in the irrigable areas where horticultural crops cultivated.

Sampling Procedures

Farmers who had at least 1 yr experience in irrigation were used as the source of the data. Exactly 306 farmers were selected from 1,500 farmers engaged in irrigation activities in the study areas using a single random sampling method (Adam 2020), which aimed at achieving a 50% population proportion and 5% margin of error. Based on this, 84, 70, 60, and 92 farmers were included in the study from Hadush-Adi, Adiha, Genfel, and Wargba areas, respectively. The farmers were household heads, who were the main actors in the irrigation farms. Most of the farmers in the Mehoni area were commercial farmers who have large investments in both fruit and vegetable crop production; whereas in the other 3 areas, the respondents were mainly small-scale farmers.

Data Collection Methods

To assess the knowledge and safety practices of farmers on pesticide use and handling household survey method of data collection was used. A semi-structured questionnaire was designed to gather information from farmers in the study areas. Initially, the questionnaire was modified from previous studies conducted in Ethiopia (Negatu et al. 2016, Mengistie et al. 2017), and a pretested interview was conducted with 10 local farmers, 8 farmers’ group leaders, and 6 crop protection experts. The questionnaire, which contained 6 subsections was prepared in English and translated into Tigrigna. Then, a face-to-face interview was conducted with the selected sample respondents.

The first section of the questionnaire focused on the demographic characteristics of respondents, such as age, sex, education level, and farmers’ experience with pesticide use. In the second section, the questionnaire focused on farmers’ knowledge and level of awareness of the handling of pesticides. In all questions in this section, 2 response options were given (1 = yes, 2 = no). The focus of the third section was on the sources and types of pesticides applied to irrigation farms by farmers. For instance, respondents were asked what types of pesticide (fungicides, insecticides, or herbicides) they often use to control common pests, and to list the most important pesticides including insecticides, herbicides, fungicides, bactericides, herbicides, and rodenticides, if any, as it was an open-ended question. In the fourth section, farmers were asked which 5 crops they cultivated were the most pesticide-consuming crops and the top 10 selected crops were considered for ranking analysis.

The fifth section was designed to assess the knowledge and practice of farmers on the application of pesticides, such as time, direction, and frequency of pesticide spray, and modes of action used. The final section was on farmers’ knowledge of preharvest intervals and methods of container disposal after pesticide spray. The time intervals between pesticide application and harvest crops were designed based on Sharaniya and Loganathan (2016).

Data Analysis

The collected data were entered in Microsoft Office Excel 2010 (Microsoft Corporation, Redmond, WA, USA) and imported into SPSS version 21 (IBM, Armonk, New York). Then, data were analyzed using descriptive and inferential statistics. Descriptive statistics such as frequencies and percentages were computed for categorical data (response parameters). Moreover, a Chi-square test (χ²) was used to detect a possible association between the demographic profiles of respondents and the use of PPE at a 5% level of significance. It hypothesized that there is a significance difference between users and nonusers of PPE in association to their demographic profiles.

Results

Demographic Profile of Respondents

As indicated in Table 1, which explains the demographic characteristics of sample respondents, the majority of the sample farmers were male, and nearly all participants (90.2%) were aged between 18 and 57 yr. A large proportion of respondents (54.2%) completed primary school and most of the sample farmers (85%) had at least 5 yr of experience in the use of chemical pesticides.

Knowledge of Farmers on Safe Handling and Negative Impacts of Pesticides

A substantial portion of farmers (58.8%) have awareness regarding the negative impact of pesticides on human, animal, and environmental health (Table 2). However, they buy pesticides without checking the expiration date and do not read or consider the instructions on the labels, as well as never use PPE during handling of pesticides.

Association between Demographic Profiles of Respondents and the Use of PPE

There was a significant association between education level of respondents and use of PPE (χ² = 6.923; P = 0.031) (Table 3). More respondents who completed secondary school or higher reported using PPE, while many of the respondents with primary school and illiterate backgrounds did not use PPE during handling and spraying pesticides. In addition, respondents’ experience with pesticide use and the use of PPE (χ² = 9.487; P = 0.023) were statistically significant at a 5% level of significance. Respondents who had less than 15 yr of experience using pesticides showed a nonsignificant association with the use of PPE. However, respondents’ age and sex had no significant association with the use of PPE.

Sources and Types of Pesticides Used in the Irrigation Farms

There were 2 main sources of pesticides determined for the study area, private shops and farmers’ unions (Fig. 2). The result revealed that most of the farmers, 238 (77.8%), were buying pesticides from their nearby private shops and none of them got pesticides from governmental organizations. While farmers in the study area used insecticides, fungicides, and herbicides, insecticides were the most frequently used pesticides followed by fungicides (Fig. 2).

Twenty-two types of pesticides were identified as the common pesticides in the major irrigation farming areas of Tigray (Table 4). According to the 2019 WHO pesticide classification by hazard, most of the pesticides used by farmers belonged to the group of “moderately hazardous (class II)”, while one insecticide, abamectin, was categorized as class Ib (extremely hazardous) pesticide (WHO 2020). Insecticides such as dimethoate, malathion, lambda-cyhalotrin, profenofos, and diazinon, and herbicides such as 2,4-D and pyroxsulam were reported to be used by most of the farmers. Among the listed pesticides, mancozeb, metelaxyl, and propiconazole fungicides were used by 100%, 98.7%, and 62.1% of the farmers, respectively, to control disease in horticultural crops. A few farmers (10.5%) were still reported to use the globally banned pesticide, P,p′-DDT, as a rodenticide (Table 4).

Major Crops that More Frequently Receive Pesticide Treatments

Based on descriptive analysis, onion, tomato, pepper, maize, and teff were the 5 most prevalent annual crops on which pesticides were frequently applied, as ranked by 98%, 94.8%, 73.9%, 63.4%, and 52.9% of the farmers, respectively. Cabbage and 4 other crops were included in the 10 farmers’ preference crops with regard to pesticide usage (Fig. 3).

Knowledge and Practice of Farmers on Safe Storage and Application of Pesticides

Most of the sample farmers (77.1%) store pesticides anywhere in the house, while a few farmers (15%) used separate places for pesticides storage (Table 5). Many (64.1%) of the farmers reported that they used large amounts of pesticides during the irrigation season when compared with the rainy season. About 72.5% and 19.6% of the respondents perceived that the appropriate time for spraying pesticides is early in the morning and late in the afternoon, respectively. Moreover, the majority (74.5%) of the respondents spray pesticides along with the wind direction (Table 5).

With regard to the spray frequency, 47.1% and 30.1% of the respondents perceived that 1 to 2 and 3 to 4 sprays per season, respectively, are enough to control pests in cereal crops (Fig. 4). On the contrary, the majority of the farmers (64.8%) perceived that 7 or more sprays of pesticide are required to control pests in vegetables and reduce yield losses. In fruit crops, however, farmers show divergence in their responses and mention a variety of options with no significant difference from the lowest to the highest frequencies of spray per season (Fig. 4).

Farmers’ Knowledge of Preharvest Interval and Disposal of Pesticide Wastes

Of the farmers surveyed, 212 (69.3%) were throwing away or disposing of the pesticide wastes (residual pesticide solution, empty containers) inside and around their farms where children and animals may gain easy access (Fig. 5A). In this regard, 24.8% of sample farmers reported that they burnt the container and yet, a very few of them (5.9%) used the empty bottles as food containers.

Most (53%) of the farmers harvest fruits or vegetables for food 1 wk after spraying and any farmers reported that they still harvest, and eat vegetables and fruits within a week of spraying (Fig. 6). A similar report was mentioned by most of the respondents that harvesting for animal feed and market takes place within 2 wk after pesticide spray.

Discussion

Pesticides are economically important in minimizing yield loss caused by insects, diseases, weeds, and other pests, while misuse of those chemicals may negatively affect human, animal, and environmental health (Sharma et al. 2019). This study focused on assessing the perceptions, knowledge, and practices of farmers regarding pesticide use and safety issues. Most of the respondents were young male farmers who completed primary school and had at least 5 yr of experience using pesticides. However, most of the farmers were practically poor in terms of the safe handling of pesticides. More than half of the respondents bought and used pesticides without reading the pesticide label, which suggests the respondents’ inability to read and recognize the presence of toxicity symbols as well as the expiration date. This result is in line with the findings of Damalas and Khan (2016) and Mengistie et al. (2017), who reported that 73% and 67% of farmers never read the instructions on the container, respectively. This could be due to low awareness and a lack of recurrent training on the proper use of pesticides (Negatu et al. 2016, Jallow et al. 2017, Sapbamrer and Thammachai 2020, Staudacher et al. 2020). The results of the current study revealed that many of the farmers had never taken any training related to the safe handling and application of pesticides, which may have contributed to their low level of understanding about the safe use and handling of pesticides. In line with the results of the current study, 85% of farmers in the central and eastern parts of Ethiopia had never taken any formal training on the safe use and hazardous effects of pesticides (Negatu et al. 2016). Moreover, lack of training is the major cause of low awareness and inappropriate handling of pesticides (De Bon et al. 2014, Agmas and Adugna 2020). Most of the farmers in the current study have some theoretical awareness of the negative impact of pesticides on human, animal, and environmental health. However, the majority of them had never implemented the proper use of PPE, which is similar to the findings in Tanzania (Lekei et al. 2014), Southwest Ethiopia (Gesesew et al. 2016), Kuwait (Jallow et al. 2017), India (Sai et al. 2019), and North West Ethiopia (Mequanint et al. 2019, Agmas and Adugna 2020). The failure to use protective equipment indicates that farmers are at high risk of exposure to pesticides. The main reasons for farmers not using PPE in the study area could be attributed to a lack of awareness creation, insufficient training for change in attitude, and a lack of technical support at the grass-roots level. In addition, the current study revealed that the use of PPE is associated with farmers’ experiences of pesticide use and education level, which indicates that educational background and experience can positively influence safety in pesticide handling, including the use of PPE. In this study, farmers with 16 or more years of experience with pesticide use and those who attended secondary school or higher had a strong association with the use of PPE, as similarly reported by Jallow et al. (2017). Farmers in developing countries are poorly practicing the use of PPE (Sharma et al. 2019) and highly exposed to risk associated with pesticides (Negatu et al. 2021). Farmers in the 4 major irrigable areas of Tigray in which this study was carried out mentioned 22 major pesticides, including dimethoate, malathion, lambda-cyhalotrin, profenofos, diazinon, 2,4-D, pyroxsulam, mancozeb, metelaxyl, and propiconazole as the most commonly used. Most of the listed pesticides were used to control pests of onion, tomato, pepper, maize, and teff, which are the top 5 crops most commonly affected by a number of pests in the study area. This could be associated with the occurrence of major pests such as thrips, basal rot, purple blotch, and white rot in onions; fruitworms and late blight in tomatoes; bacterial blight and root rot disease in pepper; stalk borer in maize and shoot fly in teff. Most of the pesticides used by the farmers of the study areas were insecticides followed by fungicides as similarly reported by De Bon et al. (2014) and Mengiste et al. (2017). This could be related to the fact that farmers may easily detect insect pests compared to plant diseases and apply insecticides more frequently. The less frequent use of herbicides can be explained by the fact that farmers in the area commonly use labor-intensive hand weeding instead of herbicides to control weeds. According to the WHO toxicity classification (WHO 2020), most of the listed pesticides belong to the group of “moderately hazardous (Class II)” pesticides; while one insecticide, namely Abamectin, is categorized as Class Ib, which is extremely hazardous to human and animal health. Moreover, some farmers reported the use of P,p′-DDT as a rodenticide in the field (Negatu et al. 2016, Mengistie et al. 2017), while it was banned in 1996 under the Stockholm Convention and recommended for indoor use in 2006 (EPA 2023). The continued use of such highly hazardous pesticides for human health and the environment calls for strengthening the regulatory system in the region. In support of the current study, Merga et al. (2021) reported that malathion, dimethoate, metalaxyl, diazinon, chlorpyrifos, fenitrothion, and endosulfan were detected in water samples collected from Lake Ziway, Central Ethiopia.

Consistent with the reports by Negatu et al. (2016), Loha et al. (2022), and Miyittah et al. (2022), the current study showed that most of the respondents use private shops as their main source of pesticides, and they store it anywhere in the living room or the kitchen. Storing pesticides in residential areas, where farmers prepare food and sleep, could pose a potential risk for high exposure (Miyittah et al. 2022). Such risky practices by farmers may be attributed to a lack of knowledge and training on safe pesticide use and handling. Farmers who get access to training on pesticide use tend to store pesticides in separate stores allocated for pesticides (Jallow et al. 2017). Respondents indicated that they spray pesticides 1 or 2 times per season to control cereal pests, while they spray more than 7 times to control vegetable pests. Most farmers in Sub-Saharan Africa apply pesticides 5 to 19 times per season for a single vegetable crop, and the range depends on the type and infestation rate of the pest (De Bon et al. 2014, Miranda et al. 2021).

Our results show that farmers were throwing away the empty containers around the farm after the use of pesticides, while a few farmers are still using the empty pesticide containers for storing food and other home materials, as similarly reported by Mequanint et al. (2019) and Damte and Tabor (2015). It was also reported that many farmers dispose of the leftover pesticide solution in the field and release sprayer-rinsed water to the soil or into a flowing river. This may lead to the accumulation of harmful pesticide residue in agricultural produce, soil, and water, posing a threat to human and environmental health (Negatu et al. 2021, Shentema et al. 2022). Most of the farmers reported that they harvest fruits or vegetables for market, food, or animal feed within 2 wk after the time of spraying, and a few of the farmers even harvest and eat within a week. Similar studies in Sri Lanka (Sharaniya and Loganathan 2016) and Nepal (Rijal et al. 2018) indicated that 60% and 62% of the farmers harvested their products within a week of the pesticide spray, respectively. Studies conducted in Ethiopia and Kuwait reported that high pesticide residue was found in one-third of the fruit and vegetable samples collected (Gelaye and Negash 2024), which indicates that harvesting crops without following the preharvest interval requirements may lead to pesticide exposure and consequential health effects (Rijal et al. 2018, WHO 2020, Mehmood et al. 2021). According to Prodhan et al. (2018), the recommended time interval between application of malathion, diazinon, and cypermethrin pesticides, and harvesting of vegetable crops is 10, 7, and 5 d, respectively. Moreover, UC-IPM (2024) reported that the preharvest interval for abamectin, dimethoate, and malathion is 7, 15, and 7 d after spray, respectively. This implies that the preharvest interval differs from crop to crop and pesticide to pesticide, and 7 d after spray is the most acceptable time for most of the pesticides used in the study areas.

Conclusion and Future Directions

Generally, pesticides are agrochemicals applied by farmers to control agricultural pests and reduce yield losses. Many farmers, particularly in developing countries, wide range of pesticides in an unsafe way. The current study revealed that farmers had poor knowledge and practices in the use of PPE, understanding the proper frequency of pesticide application, disposing of pesticide waste, and using appropriate preharvest interval time despite having experience with pesticides and awareness of the negative impact of them. Inappropriate handling of pesticides can cause human and animal health disorders and negatively affect soil, water, and air quality. Therefore, the current study may give a baseline for researchers, decision-makers and other relevant stakeholders for future interventions in the proper handling of pesticides and management of pesticide exposures. Farmers should get recurrent training on integrated pest management options and handling pesticides to ensure effective pest control and reduce the negative impact of pesticides on human, animal, and environmental health.

Acknowledgments

The authors would like to thank the respondents and all others who participated in the data collection.

Author contributions

Abraha Gebretsadkan (Conceptualization [equal], Formal analysis [equal], Investigation [equal], Project administration [lead], Supervision [lead], Writing—original draft [equal]), Filmon Tquabo (Conceptualization [equal], Investigation [equal], Methodology [equal], Validation [equal], Writing—review & editing [equal]), Alemu Araya (Conceptualization [equal], Investigation [equal], Methodology [equal], Validation [equal], Writing—review & editing [equal]), Gebremedhn Beyene (Conceptualization [equal], Investigation [equal], Methodology [equal], Validation [equal], Writing—review & editing [equal]), Haftay Berhane (Investigation [equal], Methodology [equal], Validation [equal], Writing—review & editing [equal]), Leake Gebreslassie (Conceptualization [equal], Formal analysis [equal], Investigation [equal], Methodology [lead], Software [lead], Writing—review & editing [equal]), Semira Taju (Investigation [equal], Methodology [equal], Validation [equal], Writing—review & editing [equal]), Teklebrhan Welday (Conceptualization [equal], Investigation [equal], Methodology [equal], Validation [equal], Writing—review & editing [equal]), Hadush Tsehaye (Supervision [equal], Validation [equal], Writing—review & editing [lead]), and Ibrahim Fitwy (Supervision [equal], Validation [equal], Writing—review & editing [lead])

Ethical statement

The Research Evaluation and Monitoring Committee (REMC) of Mekelle University approved the study. Consent to collect and publish data was obtained from the participants, and those participants were informed of the purpose of the study.

Funding

This research project was financially supported by Mekelle University under grant number of CRP/MU/Medium/CoDANR/15/2019-2021.

Conflicts of interest: The authors declare that they have no competing interests.

Data availability

The data that support the findings of this study are available from the corresponding author, upon reasonable request.

References