Resource optimization in crop livestock integrated production system among small-scale cotton farmers in southern Mali

Advances in Agricultural Science 07 (2019), 01: 85-98

Resource Optimization in Crop Livestock Integrated Production System among Small-Scale Cotton Farmers in Southern Mali

Abdoulaye Nientao 1, 2*, Dave Mwangi 1, Oscar I. Ayuya 1, Alpha O. Kergna 2

Department of Agricultural Economics and Agribusiness Management, Faculty of Agriculture, Egerton University, Box 536 –20115, Kenya.
Institut d’Economie Rurale (I. E. R),  B. P : 258, Rue Mohamed  V – Bamako, Mali.


Crop and livestock enterprise combinations among small-scale farmers were studied in southern Mali with the purpose of determining the most profitable combination of crop and livestock enterprises. Primary data was collected from four categories of farmers which are types A, B, C and D. This classification of farmers was done by research and it was based on the level of agricultural equipment (plough and cart) and number of oxen. Therefore, data was collected from 171 randomly selected farmers using a semi-structured questionnaire. This study is based on profit maximization using a linear programming (LP) model through total gross margin of the major crops and livestock enterprises. Results showed that small-scale cotton farmers were do not efficiently utilize their current resources. If present level of resources were efficiently optimized, the results showed that farmers could increase their profit by 104.80%, 54.35%, 23.01% and 19.52% for types A, B, C and D in Southern Mali respectively. Further, the results revealed that resources such as labor, land and number of animals could be the major constraints if the resources were well optimized. The study also showed that the most profitable combination of crop and livestock enterprises is cotton / maize / rice / cattle /sheep / donkey for small-scale cotton farmers in Southern Mali.

Keywords: Small-scale cotton farmers, Resources Optimization, Linear Programming, Southern Mali

How to Cite: Nientao, A., Mwangi, D., Ayuya, O., & Kergna, A. (2018). Resource Optimization in Crop Livestock Integrated Production System among Small-Scale Cotton Farmers in Southern Mali. Advances in Agricultural Science7(1), 85-98.   


Scince 1960s years, crop and livestock integration systems (CLIS) have been promoted in many Sub Saharan African countries included Mali. Franzluebbers, (2007) indicates that CLIS is one of the practices, which have been common approach to agriculture production all over the world before modern industrialization in the 20th century. CLIS reduces soil erosion, strengthen environmental sustainability, increases crop yields and improves profits, thus helping in reduction of poverty and malnutrition (IFAD, 2010). According to the FAO (2009) crop livestock integrated production system improved small-scale farmers’ agricultural productivity by 50% in Ethiopia and income by more than 100% in Zimbabwe. Thornton et al. (2001) argue that, over 50% of meat and 90% of milk in the world are provided by mixed crop-livestock system and it is the most common form of livestock operation in developing countries.

Furthermore, it is projected that CLIS is going to increase (Delgado et al., 1999) over the world in the next thirty years as human population increases, the demand of livestock products mainly meat and milk will follow. According to Hosu and Mushunje (2013), this will give an opportunity for small-scale farmers to benefit from the growing market and raise their income thus contributing to reduced poverty. However, the agricultural production among small-scale farmers is faced with resource constraints that need to be optimally used. Therefore, limited resource use in the farming system requires change in management practices in order to optimally allocate resources among small-scale farmers’ enterprises or activities facing multiple constraints. An efficient management of natural resources is a best way of increasing productivity and farmers’ incomes by combining crop and livestock enterprises.

In Mali, CLIS is an integral component in rural livelihood and particularly in southern Mali area, where cotton crop has a significant impact on socio economic development through many agricultural support programs (Droy et al., 2012). Scoones et al. (2000) in their study in southern and central Mali, argue that integrated of crop and livestock production system interacts in different ways according to the agro-ecological areas. Livestock provides animal draught power to plough and weed crop lands and crop residues are left in the field and grazed by animals or in some cases they are stored as fodders for dry season. In southern and central Mali, livestock represents an important investment and store of wealth for small-scale farmers and procurement of oxen pair is seen as the highest priority livestock investment. Blanchard et al. (2013), state that CLIS management must change in order to adapt and guarantee small-scale farmers’ viability in the face of changing climate, economic, and institutional conditions.

The reality is that small-scale farmers’ production system in southern Mali is faced with multiple constraints that necessitate change in interactions between different productive and limited resources for efficient use. Therefore, resource use optimization in CLIS is the feasible and best way of allocating limited resources in order to meet the new needs of small-scale farmers. This article aims to assess whether the optimization of resources in crops and livestock enterprises combinations among small-scale cotton farmers in southern Mali are attainable and efficiently managed with a view to establishing the most profitable enterprise combinations of crops and livestock. Findings from this study are expected to orientate the policymakers’ decision toward efficient use of resources for a sustainable agriculture by reducing poverty and enhance food security among small-scale farmers in Mali.



A multistage sampling technique was used to get the study sample where Southern Mali and four villages were purposively selected. It focused on all small-scale farmers combining crops and livestock enterprises. The study takes cognizance of the main crops and livestock enterprises (which are cotton, millet, maize, rice, sorghum, groundnut, oxen, other cattle, sheep and donkey) that the farmers practised and the family farm was the sampling unit. At last, linear systematic sampling method were used to select the respondents in each village proportionate to size from the available list of farmers.

Both primary and secondary data were used in this study and cross sectional data were collected on the inputs and outputs of crop and livestock production and farm characteristics. The inputs were used as resources under the analysis to know that if they are excessing or lacking compare to the required resources in maximizing small-scale farmers’ profit function. Primary data were collected for the period of 2016/2017 agricultural campaign through interview using semi-structured questionnaire, which was based on the enterprises of crops and livestock practised by farmers. The primary data of crops and livestock enterprises were the available and the real inputs used and the man day used for each enterprise with the price per unit of each input to determine the cost. Also, the outputs were the yields for crops and number of animals for livestock that maximized small-scale farmers’ profit at the present level of resources. Secondary data on the required quantity of each input for both crops and livestock such fertilizers, man day, antibiotic and vaccine, were obtained from Malian Company of Textile Development (CMDT), National Research Institution (IER), journals, and other written literature. They were also used in the model LP as requirement data to cross check the primary data obtained from farmers. Therefore, some extreme values reported by farmers were eliminated from the data analysed. Also, the tools such as SPSS, STATA and Excel were used for data entries and analysis.



The restrictions for the Linear Programming (LP) models for each type of farmers were specified by using the average values of available resources for the main crops and livestock enterprises identified. Labour resources were captured for each operation of crops and livestock keeping including feeding animals. Inputs such as fertilizers, herbicides and insecticides were measured in kilogramme and litre for crops while livestock inputs such as vaccine, dewormers and veterinary services were captured in number. Land and livestock unit were obtained from the average land in hectares and livestock in Tropical Livestock Unit (TLU) owned by each type of farmers.

Respondents’ socioeconomic characteristics were described in STATA using descriptive statistics such as frequencies tables, means, cumulative frequencies and percentages. Gross margin was computed using the total cost of inputs used for crops and livestock enterprises subtracted from the total production value of crops and revenue of selling livestock and milk. This was done to determine farmers’ objective function of profit maximization constrained by the average available resources such as land, labour, inputs and number of livestock owned. The LP model was fitted using the Excel Solver to determine farmers’ maximized profit under crop-livestock integrated system. In spite of, this LP model

The idea behind LP is to maximize objective function, which is in our case either “Profit” or “Gross Margin” subject to some constraints, which are restrictions that show what we are allowed to do. This LP model or simplex algorithm was formulated by George Dantzig (1947) and it has been a veritable tool for decision making and has also been widely applied in agriculture area. In this study, the model was adapted as applied by Hosu and Mushunje (2013). The problem to solve is which enterprise combination in crop livestock integrated production system gives the higher gross margin with the limited resources available production factors. Despite its usefulness of optimizing small-scale farmers’ profit, some the limitations of LP model are that it does not consider the changes and the evolution of variable as time goes by. Indeed, the LP model in this study does not take into account the variability observed among farmers, the changes in inputs and outputs prices. One of the solutions to minimize the bias in the variability of farmers is to classify farmers into different types which are the base of the study’s analysis. Another limitation arises in the formulation process which is the values must be known with certainty and this should be taken into account when programming the model.

The combination of crops and livestock enterprises consider the optimal use of available resources such as fertilizers, land, man day, vaccine and crops residues for feeding animals in order to obtain a maximized profit. The LP model were maximized an objective function (1), which is gross margin by allocating optimally the resources or factors of production to the most productive crop-livestock enterprise combination. Equation (2) requires that the quantity of resources used ( ) should not exceed the available resources ( ) and equation (3) shows the non-negativity of the unit of livestock and the amount for crop. Therefore, the optimized problem of small-scale farmer was specified as follow:



Results and Discussion

Most farmers are small-scale farmers in nature with a land sizes in hectares ranging from 0.7 to 54 and an average of about 11.6 hectares mainly under food crops such as maize, millet, sorghum, rice and groundnut. The results from table 1 showed that only 9% of the respondents had gone to primary school, 1% to secondary school, 41% were not educated and the rest of the respondents 49% have received rural literacy education in national language. The study also observed that 32% of the respondents were in the age bracket of 41-50 years and 26% in 51-60 years which means farmers are in the productive age, which is good for farming activities. About 63% of the respondents were household heads and 99% of the household activities were based on agriculture as the main occupation while 71% of household participated in off-farm activities. This implied that farmers had diversified their activities to reduce the climatic risks. Similarly, 99% of the respondents were male whereas 98% were married with the majority (57%) being polygamous households about 57%.

In the study area, small-scale farmers have been classified into four categories which are types A, B, C and D by the national institute of research (IER) and the Malian Company of Textile Development (CMDT). This classification was based on the level of agricultural equipment (plough and cart) and number of oxen (Scoones et al., 2000). Class A farms were well equipped with more than two pairs of oxen and a set of ploughing and weeding tools. Class B farms were partially equipped with at least one pair of oxen and a plough. Class C farms were under equipped with either oxen without plough, or plough without oxen and class D farms were those without experiences in animal traction. This classification provides the foundation of the results interpretation in the optimization model (Linear Programming) since the analyses were focused on the four types of farms. The major enterprises of crop and livestock and their profit by type of farms are shown in table 2. The average gross margin per unit differs by type of farms and by enterprise. The differences are because of differing expenses that is, labor and other inputs such as fertilizers, seeds and veterinary services incurred by enterprise and by type of farms to produce. These expenses are incurred in one year for crops and several years for livestock depending on the reformed ages (Bosma et al., 1996) of the type of livestock. Only sheep and goat have a maturity age of less than two years (one and half year). Crops enterprises such as rice, cotton and groundnut had the highest average gross margin per hectare followed by ox and other cattle enterprises per Tropical Livestock Unit. These gross margins for crops and livestock enterprises were determined how farmers could maximize their profit

Table 1. Socioeconomic characteristics of integrated crop-livestock farmers in Southern Mali

 Socioeconomic characteristics Frequency Percentage Cumulative
Age (in years) 21-30 11 6.3 6.3
31-40 34 19.3 25.6
41-50 56 31.8 57.4
51-60 45 25.6 83.0
>60 30 17.0 100.0
Household head Male 176 99.4 99.4
Female 1 .6 100.0
Marital Status Never married 2 1.1 1.1
Monogamously married 72 40.7 41.8
Polygamously married 102 57.6 99.4
Widowed 1  0.6 100.0
Education Level None 72 40.7 40.7
Primary 16 9.0 49.7
Secondary 2 1.1 50.8
Alphabetised 87 49.2 100.0
Household size (number of person) 1-15 95 53.7 53.7
16-30 64 36.2 89.8
31-45 12 6.8 96.6
46-60 5 2.8 99.4
>60 1 .6 100.0
LandSize (ha)





0.1-1 3 1.7 1.7
1.1-10 100 56.5 58.2
10.1-20 50 28.2 86.4
20.1-30 16 9.0 95.5
30.1-40 5 2.8 98.3
40.1-50 2 1.1 99.4
>50 1 .6 100.0

Source: Field Surveys, 2017.



Table 2. Gross margin of major Crop and livestock enterprises identified in the study area.

Variable Enterprises Average gross margin (in Fcfa)/enterprise
Type A Type B Type C Type D
X1 Cotton 166,026.20 145,992.30 137,033.60  137,970.70
X2 Millet   67,267.13   64,873.38    61,582.92    12,526.67
X3 Sorghum   45,094.58    37,832.58    40,410.46    45,389.33
X4 Maize   79,762.52    58,483.99    99,710.64    83,826.86
X5 Rice    345,269.00    243,243.00  118,663.80     62,684.00
X6 Groundnut    220,173.00    183,215.20    77,644.70       –
X7 Oxen    92,517.18    137,845.00    103,293.00  103,336.67
X8 Other cattle    124,690.00    95,405.60    102,100.00   77,783.70
X9 Sheep    25,242.50    38,815.00      37,300.00     32,755.00
X10 Goat      15,489.00    20,004.00    14,950.50    11,814.98
X11 Donkey      43,520.00    27,450.00    31,460.00     10,660.00

Source: Field surveys, 2017.


functions with limited resources. The summary of Linear Programming (LP) results that maximized profit function subject to the available resources for an integrated crop-livestock system by each type of farms is shown in table 3. The LP results showed that in Southern Mali small-scale cotton farmers could be more efficient working in small sized land because of input efficiency than working in large sized lands. It also showed the importance of crops-livestock enterprise mixes in small-scale farmers’ diversification strategies to maximize their profit and enhance food security because livestock enterprises system in Southern Mali freely grazed which is less costly in comparison to crop-based enterprises. To convert livestock, we used the conversion factors of Tropical Livestock Unit (TLU) for each type of livestock owned by farmers in the final solution and 1 ox=0.8 TLU, 1 other cattle=0.7 TLU, 1 sheep=0.10 TLU, 1 goat=0.08 TLU, 1 donkey=0.50 TLU (storck, et al., 1991).

The production activities that maximized profit are 2.44 ha of cotton (X1), 3.26 ha of maize (X4), 1.34 ha of rice (X5) and 1.6 TLU of ox (X7) for type A farmers. These mixed activities of crops and livestock (cotton/maize/rice/ox) give a value of 1,263,180.72 Fcfa for the maximized total gross margin of type A farmers. This means that to maximize their profit with limited resources, types A farmers should opt for cotton/maize/rice/oxen enterprises mix. It does not mean that type A farmers should have no interest in growing other crops and rearing other type of livestock; only that using this combination would likely maximize their profits. The results further reveal that the maximized profit of major crops and livestock is higher than the real total average gross margin of 616,783.91 Fcfa currently gotten by type A. Therefore, type A farmers using the present level of resources can optimized their productivity to yield a 104.80% increase in their present profit.

The production activities for type B farmers that could maximize their profit are mixed enterprises of cotton/maize/rice/ox/other cattle. Type B farmers can have a maximized total gross margin of 700,003.16 Fcfa with the available resources. These mixed enterprises of crop and livestock for type B farmers could also be more efficient with small sized lands of 1.12 ha for cotton (X1), 1.74 ha for maize (X4), 1.16 ha for rice (X5), 0.8 TLU for ox (X7) and 0.7 TLU for other cattle (X8). The maximized total gross margin is higher compared to the current total average gross margin of 453,526.63 Fcfa. Further, type B farmers can optimize their profit to yield a 54.35% increase in the total gross margin at the present level of available resources if they efficiently used them. This means that type B farms should choose cotton/maize/rice/ox/other cattle enterprises mixes to maximize their profit with a total land size of 4.02 ha for crops (cotton, maize and rice) and a total number of 1.5 TLU (ox and other cattle) for livestock.

Crops and livestock enterprises mixed that maximize type C farmers profit were cotton/millet/maize/rice/ox/other cattle/sheep with a maximized total gross margin of 376,212.10 Fcfa for the optimal solution. The production activities that could maximize type C farmers profit function were 0.57 ha of cotton (X1), 0.36 ha of millet (X2), 0.41 ha of maize (X4), 1.05 ha of rice (X5), 0.8 TLU of ox (X7), 0.7 TLU of other cattle (X8) and 0.1 TLU of sheep (X9). This maximized profit is greater than the current total average gross margin of 305,831.07 Fcfa and type C farmers can maximized their profit to yield a 23.01% increase in the total gross margin by efficiently using the present level of available resources.

The production activities that could maximize type D farmers profit are 0.82 ha of cotton (X1), 1.06 ha of maize (X4), 0.8 TLU of ox (X7), 0.1 TLU of sheep (X9) and 0.5 TLU of donkey (X11). The value of the maximized profit function for type D farms plan was 362,677.69 Fcfa with the highest contribution of 137,970.7 Fcfa from the production of cotton (X1) activity. Type D farmers’ real total average gross margin of 303,431.97 Fcfa is smaller than the maximized profit of 362,677.69 Fcfa. Therefore, type D farmers can optimized their profit to yield a 19.52% increase in the total gross margin

Table 3. Summary of optimal solution by type of farm.





Variable amount in the final solution
Type A Type B Type C Type D
X1 Cotton  2.44 Ha 1.12 Ha 0.57 Ha 0.82 Ha
X2 Millet         0         0 0.36 Ha        0
X4 Maize  3.26 Ha 1.74 Ha 0.41 Ha 1.06 Ha
X5 Rice  1.34 Ha 1.16 Ha 1.05 Ha        0
X7 Oxen   1.6 TLU   0.8 TLU   0.8 TLU   0.8 TLU
X8 Other cattle         0   0.7 TLU   0.7 TLU        0
X9 Sheep         0         0   0.1 TLU   0.1 TLU
X11 Donkey         0         0         0   0.5 TLU
Solution Maximized (Fcfa)  1,263,180.72 700,003.16 376,212.10 362,677.69

Source: Field surveys, 2017.


at the present level of resources. They should choose cotton/maize/rice/ox/sheep/donkey enterprises mixed for maximizing their profit subject to the current available resources. The results showed that type D farmers had experiences with animal traction such as oxen to plough their land contrary to what had been mentioned in literature review about farmers’ classification.

Resources status of all type of farmers is shown in table 4. The constraining resources are similar for all the four types of farmers. The common constraints are labor for crops and livestock, urea fertilizer and land. These resources are called active resources because any unit change of them will affect the optimal solution of maximized profit. For instance, any additional unit of labor (man/day) for harvesting operation will increase the profit of all the four type of farmers by 3,983.96 Fcfa, 3,614.22 Fcfa, 914.39 Fcfa, and 3,143.67 Fcfa for types A, B, C and D respectively. The constraining resource, which gives the highest increase in farmers’ profit plan, is land under rice for types A, B and C with 204,466 Fcfa, 124,740.88 Fcfa and 32,073.49 Fcfa respectively. This means that the profit will increase by these respective amounts as result of an extra unit of land (1ha) growing rice. As regards to type D farmers, they will have the highest increase of 18,851.32 Fcfa in their profit as result of one additional unit of sheep. Further, resources not constraining farmers’ profit maximization plan, are called inactive. This is to say that any unit change of these resources would have no effect on the optimal solution and do not affect the maximized profit.

Results from appendices 1, 2, 3 and 4 showed the slack variables. Slacks with the value zero means that those resources have been fully used by farmers to maximize their profit. When maximizing the profit, type C farmers face more constraints compared to other types which might also explain the importance of mixing seven enterprises in their diversification strategies and risk management for increasing their profit. The maximized solution of gross margins under the existing available resources for crops and livestock enterprises were increased about 104.80, 54.35, 23.01, 19.52 percent for types A, B, C and D farmers respectively. This implies that the available resources were not efficiently used by all the four types of farmers. The results underscored that crop and livestock are well integrated in the study area and all the types of farmers are practicing both activities of crops and livestock. It also revealed that farmers in southern Mali are labor constrained for both crops and livestock activities as shown by the resource status.

Table 4. Summary result of the LP model on resources status by type of farmers.

Resource variable Type A Type B Type C Type D
Shadow price (Fcfa) Resource status Shadow price (Fcfa) Resource status Shadow price (Fcfa) Resource status Shadow price (Fcfa) Resource status
Harvesting labor 3,983.96 Active 3,983.96 Active 914.39 Active 3,143.67 Active
Urea Quantity 133.35 Active 133.35 Active 573.30 Active 244.48 Active
Insecticide Quantity 0 Inactive 0 Inactive 17,948.28 Active 0 Inactive
Average_Land_Rice 204,446.80 Active 204,446.80 Active 32,073.49 Active 0 Inactive
Labour Livestock 0 Inactive 0 Inactive 289.47 Active 463.38 Active
Crops_Residus_Labor 8,410.65 Active 8,410.65 Active 0 Inactive
Dewormers 0 Inactive 0 Inactive 6,781.58 Active 0 Inactive
Antibiotic 0 Inactive 0 Inactive 0 Inactive 5,562.78 Active
Straw _Labor 0 Inactive 0 Inactive 3,476.61 Active 0 Inactive
Average Sheep 0 Inactive 0 Inactive 0 Inactive 18,851.32 Active

Source: Field surveys, 2017.


These results are similar to past studies, which found that farmers’ resources were not efficiently used in the existing plan compared to the optimum plan. Hosu and Mushunje (2013) using LP to model crop and livestock enterprises mixes that would maximize profit in South Africa, discovered that farmers could yield a 122% increase in their present profit margin if they had optimized the available resources use. Igwe and Onyenweaku (2013) in their study applying LP to food crops and livestock enterprises in Nigeria, found that optimizing and reallocating available resources could bring significant increases in farmers’ existing gross margin up to about 61.35%.

Resources such as land and livestock were not limiting factors to optimize small-scale farmers’ gross margins in Southern Mali and only land available for rice was limiting. However, resources are not used efficiently by farmers and there is an opportunity for increasing farm gross margin by using small sized land and keeping small unit of livestock. This corroborates with Igwe et al. (2013), using LP to combine crop, monogastric farm animal, and fish enterprises in Abia state, Nigeria, concluded that farm resources were not optimally allocated and optimization of crop and livestock enterprises combination could improve the gross returns about by72.90% to the farmers with the present resources optimally used. It also supported Sanni et al. (2003) who found that resources were unused or inefficiently used among smallholder farmers in integrated crop livestock farming systems in Katsina State, Nigeria.

At optimal levels, small-scale farmers in Southern Mali should devote their present resources for small land size of 7.04 ha, 4.02 ha, 2.39 ha, and 1.88 ha for types A, B, C, and D farmers respectively with a small unit of livestock of 1.4 to 1.6 TLU. The implication for small-scale farmers in Southern Mali to cultivate large land size is that they pursue the goals of ensured their households’ food security rather than optimizing profit. Hosu and Mushumaximization plan and smallholder farmers in Eastern Cape Province of South Africa were more efficient with small sized land because of input efficiency than working on a large sized land. Kumari et al. (2014), found out that when optimizing allocation of agricultural land to the vegetable crops in India, allocation of land to various crops with limited resources becomes the major challenges to fetch higher profits and therefore there is need for proper land utilization and proper cropping pattern at farm level. It is also noted that from this study results, in the optimal solution some crops such as millet, sorghum and groundnut do not appear. This may be because farmers in southern Mali do not apply inputs such as fertilizers and herbicides into sorghum or millet plots and then incur less cost compared to cotton, maize and rice.

Therefore, when optimizing the resources use these crops do not appear in farming rotation because the required quantity in terms of inputs is high than what is available (or applied). Also, the average gross margin of cotton, rice and maize are greater than those for millet, sorghum and groundnut. It is instructive to note that a limitation of LP is that the prices of output and input were assumed to remain constants. Therefore, if they change so that would bring about different combinations of crops and livestock enterprises in farmers’ optimal solution.



Efficient resource use for profit maximization is one of the major challenges among small-scale farmers in most of the developing countries. This study had showed that resource use in crop-livestock integration system among small-scale cotton farmers can be economically profitable when resources are efficiently used. Small-scale cotton farmers’ resource optimization for maximizing their profit is equivalent to reinforcement of rural livelihood sustainability by increased farm income. The LP model used in this study pointed out that small-scale cotton farmers are not efficiently using their current available resources. Therefore, farmers’ managerial capacity should be improved by working out appropriate policies on farmers advisory and refresher training on farm management tailored according to the present level of resources. The results revealed that revenue from livestock are low compared to the one from crops implying there should be incentive actions to encourage farmers to practice intensive or semi-intensive livestock system instead of free grazing system adopted by small-scale cotton farmers in southern Mali. Resources, which are mainly constraining such as labor and inputs for crops and livestock, can be used efficiently by reducing the land size and practicing intensive livestock system. Further, the LP model depicted that farmers held some resources in excess, which is an indication of inefficiency in actual resource use by small-scale cotton farmers in CLIS. Therefore, the study concluded that at optimal resource use in crop-livestock integration system, small-scale cotton farmers could be more efficient when working in small sized land with small unit of livestock from 7.04 ha to 1.88 ha for crops and 1.4 to 1.6 TLU for livestock. Based on the findings arising from the study, it is further recommended that an effective advice to farmers on the efficient allocation of farm resources, would be building into programs promoting increased agricultural productivity and income among small-scale farmers cotton-based system. Though the study attempts to determine the most profitable crop and livestock enterprises mixed in Southern part of Mali, further investigations need to be conducted to establish the optimal combination enterprises prototype for all cotton zones.



This material is based upon work supported by the United States Agency for International Development, as part of the Feed the Future initiative, under the CGIAR fund, award number BFS-G-11-00002, and the predecessor fund the Food Security and Crisis Mitigation II grant, award number EEM-G-00-04-00013.



Bamiro, O. M., Adedeji, I. A., Otunnaiya, A. O., Soluade, W., and Ogunjobi, J. O. (2015). Enterprise Combination in Livestock Sector in Southwestern, Nigeria. World Economic Research, 4 (2):38-44.

Blanchard, M., Vayssieres, J., Dugué, P., and Vall, E. (2013). Local Technical Knowledge and Efficiency of Organic Fertilizer Production in South Mali: Diversity of Practices. Agroecology and Sustainable Food Systems, 37 (6): 672-699.

Bosma, R., Bengaly, K., Traoré, M., Roeleveld, A. (1996). L’élevage en voie d’intensification: Synthèse de la recherche sur les ruminants dans les exploitations agricoles mixtes au Mali-Sud. Amsterdam: (KIT) Royal Institute of the tropics, Pays-Bas; Institut d’Economie Rurale, Bamako, Mali, 202p.

Christopher, C. O. (2008). Comparative Analysis of Enterprise Combination Costs and Returns in Cassava-Based Food Crop Farming Systems in Delta State, NIGERIA. Agricultural and Biological Science, 3 (4):27-32.

Debertin, D. L. (2012). Agricultural Production Economics. Second Edition, University of Kentucky, USA.

Droy, I., Bélières, J.F., and Bidou, J. E. (2012). Between Crisis and Surplus: Questions Regarding the Durability of Cotton Production Systems in Mali. Development Research, 24 (3): 491-508.

FAO. (2009). Food Security and Agricultural Mitigation in Developing Countries: option for capturing synergies. FAO, Rome, Italy.

Franzluebbers, A. J. (2007). Integrated crop-livestock systems in the southeastern USA. Agronomy Journal, 99: 361-372.

Hosu, S., and Mushunje, A. (2013). Optimizing Resource Use and Economics of Crop-Livestock Integration Among Small Farmers in Semiarid Regions of South Africa. Agroecology and Sustainable Food Systems, 37: 985–1000.

Igwe, K.C., and Onyenweaku, C.E. (2013). A Linear Programming Approach to Food Crops and Livestock Enterprises Planning in Aba Agricultural Zone of Abia State, Nigeria. American Journal of Experimental Agriculture, 3 (2): 412-431

Igwe, K.C., Onyenweaku, C.E., and Tanko, L. (2013). A Linear Programming Approach to Combination of Crop, Monogastric Farm Animal, and Fish Enterprises in Ohafia Agricultural Zone of Abia State, Nigeria. Global Journal of Science Frontier Research Agriculture and Veterinary Sciences, 13:3

International Fund for Agriculture Development. (2010). Integrated crop-livestock farming systems. IFAD, Livestock Thematic Papers, 1:8.

Kumari, P.L., Reddy, G.K., and Krishna, T.G. (2013). Optimum Allocation of Agricultural Land to the Vegetables Crops Under Uncertain Profits Using Fuzzy Multiobjective Linear Programming. IOSR Journal of Agriculture and Veterinary Science, 7 (12): 19-28.

Scoones, I., and Wolmer, W. (2000). Pathways of change in Africa: crops, livestock and livelihoods in Mali, Ethiopia and Zimbabwe from .

Thornton, P. K., and Herrero, M. (2001). Integrated crop – livestock simulation models for scenario analysis and impact assessment. Agricultural Systems, 70: 581–602.

Tiku, N.E.,  Ele, I.A., and  Igbodor, F.O. (2015). Gross margin analysis of sweet potatoes production in the in- land valleys of Obubra Local Government Area of Cross River State, Nigeria. Agricultural Science, 3(1): 1-6.

Zoundi, J. S., Butare, I., Ndikumana, J., and Adomefa, K. (2006). INTEGRATION AGRICULTURE-ELEVAGE: une alternative pour une gestion durable des ressources naturelles et une amélioration de l’économie familiale en Afrique de l’Ouest et du Centre. Ouagadougou: INERA, Naïrobi: ILRI, Dakar: CORAF/WECARD. 374p.



Appendix 1. Result of the LP model on resources status for type A farms.

Resource variable Shadow Price (Fcfa) Resource status Slack
Land 0 inactive 8.84
Ploughing labor 0 inactive 7.04
Sowing labor 0 inactive 12.53
Weeding labor 0 inactive 12.78
Earthing labor 0 inactive 9.88
Harvesting labor 3,983.96 active 0
NPK Quantity 0 inactive 862.81
Urea Quantity 133.35 active 0
Herbicide Quantity 0 inactive 11.6
Insecticide Quantity 0 inactive 5.12
Average_Land_cotton 0 inactive 5.75
Average_Land_Millet 0 inactive 3.55
Average_Land_Sorghum 0 inactive 2.43
Average_Land_Maize 0 inactive 2.01
Average_Land_Rice 204,446.80 active 0
Average_Land_Groundnut 0 inactive 1.3
Labour Livestock 0 inactive 285.79
Vet-services 0 inactive 35.58
Vaccins 0 inactive 39.19
Crab 0 inactive 295.35
Salt 0 inactive 165.88
CottonCake 0 inactive 454.29
Bran from Cereals 0 inactive 325.19
Crops_Residus_Labor 8,410.65 active 0
Dewormers 0 inactive 53.61
Antibiotic 0 inactive 53.78
Ligneous Leaves Labor 0 inactive 7.33
Straw _Labor 0 inactive 60.58
Average ox 0 inactive 4.54
Average other cattle 0 inactive 35
Average Sheep 0 inactive 12
Average Goat 0 inactive 10
Average_Donkey 0 inactive 1

Source: Field Surveys, 2017.


Appendix 2. Result of the LP model on resources status for type B farms.

Resource variable Shadow price (Fcfa) Resource status Slack
Land 0 inactive 5.48
Ploughing labor 0 inactive 7.32
Sowing labor 0 inactive 12.44
Weeding labor 0 inactive 10.71
Earthing labor 0 inactive 6.05
Harvesting labor 3,614.22 active 0
NPK Quantity 0 inactive 490.51
Urea Quantity 28.47 active 0
Herbicide Quantity 0 inactive 6.13
Insecticide Quantity 0 inactive 3.44
Average_Land_cotton 0 inactive 2.76
Average_Land_Millet 0 inactive 2.93
Average_Land_Sorghum 0 inactive 1.9
Average_Land_Maize 0 inactive 1.17
Average_Land_Rice 124,740.88 active 0
Average_Land_Groundnut 0 inactive 1.13
Labour Livestock 0 inactive 83.18
Vet-services 0 inactive 7.95
Vaccins 0 inactive 14.66
Crab 0 inactive 20.8
Salt 0 inactive 41.47
CottonCake 0 inactive 28.43
Bran from Cereals 0 inactive 112.3
Crops_Residus_Labor 6,354.05 active 0
Dewormers 0 inactive 1.61
Antibiotic 0 inactive 13.51
Ligneous Leaves Labor 8,487.88 active 0
Straw _Labor 0 inactive 3.92
Average ox 0 inactive 2.56
Average other cattle 0 inactive 3.04
Average Sheep 0 inactive 4
Average Goat 0 inactive 5
Average_Donkey 0 inactive 2

Source: Field Surveys, 2017.


Appendix 3. Result of the LP model on resources status for type C farms.

Resource variable Shadow Price (Fcfa) Resource status Slack
Land 0 inactive 4.33
Ploughing labor 0 inactive 7.59
Sowing labor 0 inactive 14.24
Weeding labor 0 inactive 21.71
Earthing labor 0 inactive 5.41
Harvesting labor 914.39 active 0
NPK Quantity 0 inactive 204.96
Urea Quantity 573.3 active 0
Herbicide Quantity 0 inactive 4.07
Insecticide Quantity 17,948.28 active 0
Average_Land_cotton 0 inactive 2.89
Average_Land_Millet 0 active 2.63
Average_Land_Sorghum 0 inactive 2.35
Average_Land_Maize 0 inactive 2.04
Average_Land_Rice 32,073.49 active 1.05
Average_Land_Groundnut 0 inactive 1.04
Labour Livestock 289.47 active 0
Vet-services 0 inactive 5.76
Vaccins 0 inactive 0.16
Crab 0 inactive 16.68
Salt 0 inactive 32.22
CottonCake 0 inactive 13.65
Bran from Cereals 0 inactive 72.42
Crops_Residus_Labor 0 inactive 1.66
Dewormers 6,781.58 active 0
Antibiotic 0 inactive 3.76
Ligneous Leaves Labor 0 inactive 8.69
Straw _Labor 3,476.61 active 0
Average ox 0 inactive 2.6
Average other cattle 0 inactive 1.39
Average Sheep 0 inactive 3.83
Average Goat 0 inactive 2
Average_Donkey 0 inactive 1

Source: Field Surveys, 2017.


Appendix 4. Result of the LP model on resources status for type D farms.

Resource variable Shadow Price (Fcfa) Resource status Slack
Land 0 inactive 2.82
Ploughing labor 0 inactive 3.37
Sowing labor 0 inactive 5.58
Weeding labor 0 Inactive 0.79
Earthing labor 0 inactive 2.81
Harvesting labor 3,143.67 active 0
NPK Quantity 0 inactive 166.33
Urea Quantity 244.48 active 0
Herbicide Quantity 0 inactive 4.03
Insecticide Quantity 0 inactive 0.43
Average_Land_cotton 0 inactive 0.88
Average_Land_Millet 0 inactive 0.6
Average_Land_Sorghum 0 inactive 0.5
Average_Land_Maize 0 inactive 0.64
Average_Land_Rice 0 inactive 0.1
Labour Livestock 463.38 active 0
Vet-services 0 inactive 14
Vaccins 0 inactive 7
Crab 0 inactive 35.75
Salt 0 inactive 14.75
CottonCake 0 inactive 23.25
Bran from Cereals 0 inactive 32.35
Crops_Residues_Labor 0 inactive 1.38
Dewormers 0 inactive 7.25
Antibiotic 5,562.78 active 0
Ligneous Leaves Labor 0 inactive 0.79
Straw _Labor 0 inactive 2
Average ox 0 inactive 0.79
Average other cattle 0 inactive 2
Average Sheep 18,851.32 active 0
Average Goat 0 inactive 2
Average_Donkey 0 inactive 0.71

Source: Field Surveys, 2017.