Smallholder farmers ’ perceptions on groundnut ( Arachis hypogaea L . )-based cropping systems : A case study of Chisamba District , Zambia

The inclusion of legume crops in cropping systems has been shown to improve soil fertility and productivity, but the adoption rate is low among many smallholder farmers. A study to determine the perceptions of the smallholder farmers on groundnut production, use of inputs and cropping systems was conducted in Chisamba District, Zambia. Using a multi-stage sampling technique, 164 farmers from 20 villages within 11 agricultural camps in two agricultural blocks were randomly chosen for the study. Questionnaires were administered to only smallholder farmers who cultivated groundnuts. Results showed that the cropping systems used by smallholder farmers in Chisamba District were crop rotation (90%), mixed-intercropping (5%) and sole-cropping (5%). Groundnut production was mainly undertaken by female smallholder farmers (71%) on land averaging 1.68 ha. However, there was non-significant relationship between gender of the household head and adoption of cropping system (χ 2 (2) = 1.726; p = 0.414). Fifty-seven percent of the smallholder farmers perceived the groundnut variety Natal Common to be most suitable for the study area (χ 2 (4) = 9.745; p = 0.045). It was concluded that particular traits of varieties affected the perceptions and hence adoption decisions of smallholder farmers in Chisamba District.


INTRODUCTION
Groundnuts are produced virtually throughout Zambia (MAFF, 2000;Mukuka and Chisanga, 2014) but mainly in agro-ecological regions II and III (MAL, 2012) under rainfed conditions (Sichoongwe et al., 2014).It is ranked second to maize both in terms of production and area cultivated in Zambia (Ross and Klerk, 2012;Mukuka and Shipekesa, 2013).However, there has been a decline in production with the national average yields as low as 642 kg kernels / ha being common (FAOSTAT, 2014).Central Province, Chisamba District inclusive, has seen yields as low as 340 kg / ha (CSO and MACO, 2011).Such low yields deprive smallholder farmers in the District who entirely depend on maize and groundnut for livelihood of their much needed protein and income (Sitko et al., 2011).The low kernel yields have partly been attributed to poor soil fertility caused by inappropriate cropping systems (GART, 2011), and use of low yielding varieties (Mukuka and Chisanga, 2014) which are susceptible to rosette disease and pests (Ross and Klerk, 2012).This has been exacerbated by the collapse of parastatal markets whereby exportation of groundnuts has dramatically reduced in Zambia (Ross and Klerk, 2012).
The yields of groundnuts can be improved by adopting sustainable cropping systems such as crop rotation (Ogunleti et al., 2014) and use of high yielding varieties (MAFF, 2000).The benefit of groundnut-based crop rotation is that the crop can fix up to 35 kg N / ha for the subsequent crops (Bado et al., 2013).Cereal-legume crop rotations improves soil fertility (Thierfelder et al., 2012;Bellwood-Howard, 2014;Ojiema et al., 2014) and consequently increases yields of the succeeding crops (Bonsu and Asibuo, 2013;Khaitov and Allanov, 2014).High yielding varieties such as Musekara Groundnut Variety 4 (MGV4) and Musekara Groundnut Variety 5 (MGV5) have been released in Zambia, but still the adoption levels are low (Ross and Klerk, 2012).
There are a lot of factors that influence decisions of smallholder farmers in adopting agricultural innovations such as crop rotation and high yielding varieties.Some of these factors include lack of credit facilities, nonavailability of inputs, risk aversion, limited access to information, farm size, labour constraints and nonavailability of market networks (Feder et al., 1985;Asfaw et al., 2012;Jerneck and Olsson, 2013;Franke et al., 2014;Tanellari et al., 2014;Meijer et al., 2015).The adoption process of an agricultural innovation involves learning, perceiving and then based on the perception developed, the farmer either adopts or rejects the innovation (Feder et al., 1985;Schroeder et al., 2013;Meijer et al., 2015).This study focused on the perceptions of smallholder farmers in Chisamba District on groundnut production systems.Understanding why smallholder farmers had certain preferences in their groundnut cropping systems can result in improving yields.It was imperative, therefore, that a study was commissioned to understand the perceptions of smallholder farmers in Chisamba District on groundnut production, use of inputs and cropping systems.

Description of the study area
Chisamba District is located about 50 km north of Lusaka, in the CHIRWA et al. 299 Central Province of Zambia.The District covers 2 978.5 km 2 and is located between latitude 14° 30' and 15° 00' S and longitudes 28°00' and 28° 30' E. It is 1 138 m above sea level.The District is in agro-ecological zone II a (AEZ II a) which receives annual rainfall of 800 to 1 000 mm.The temperature ranges from 14.31 to 27.31°C (GART, 2011).

Study design and approach
The study used primary data which was collected directly from smallholder farmers using a structured questionnaire as an instrument of data collection.The study was conducted in two stages.The first stage involved a reconnaissance survey, conducted between 22 nd January and 20 th April, 2014.Information on smallholder groundnut farmers was obtained with the help of the District Agricultural Officer, Extension Officers and traditional zone leaders.Therefore, the reconnaissance survey helped to refine the questionnaire.The second stage of the study involved using the multi-stage sampling technique for the selection of respondents in three steps.The sampling stratum was composed of number of villages in the camp, total number of households and ideal number of villages selected per camp (Table 1).In the first step, two agricultural blocks (Chisamba and Muswishi) were purposively selected.Then subsequently, proportionate stratification depending on the respective weighting of the households was done.This was followed by random selection of 20 villages (Table 1).The face to face interviews, which included follow ups, were conducted between 26 th April and 20 th June, 2014.Despite follow ups, only 164 respondents were available in the randomly selected stratum.Socio-demographic characterisation of the households (hh) and human capital was used to determine the perceptions of smallholder farmers.Questions to capture perceptions of farmers on cropping systems, benefits of crop rotations and information on the traits of varieties such as drought and disease resistance, yield and oil content were asked.The questionnaire also captured information on the perception of smallholder farmers on new varieties, quality of seed and use of fertilizers.New and old groundnut seed varieties packed in plastic bags were shown to farmers to help them remember the varieties they had been planting in the three previous agricultural seasons, namely 2011/12, 2012/13 and 2013/14.The label from inside the plastic bag was concealed and removed systematically before the variety could be shown to the farmer.Statistical Package for Social Scientists (IBM SPSS, windows version 16.0) was used to obtain descriptive and inferential statistics.The Chi-Square (χ 2 ) was used to determine the test of independence between relationships at α = 0.05 (Kothari, 2009).

Household characterisation of smallholder groundnut farmers
Table 2 shows the characteristics of households (hh) in Chisamba District that produced groundnuts.It was observed that 77% of the hh were male headed while 23% were female headed.There were no child headed households in the study area.The age of 34% male household heads ranged between 41 and 55 years.The mean age of male household heads was 42 years old.On the contrary, the age of the majority (35%) of the female household heads ranged between 25 and 40 years old.The mean age of the female household heads was 35 years old.The younger the age of smallholder farmer the higher the likelihood that the farmer would adopt new agricultural innovations (Sichoongwe et al., 2014).There was a non-significant relationship in this study between age of the household head and adoption of new groundnut varieties or crop rotation or incurporating groundnuts in crop rotation.Age was not a factor in influencing the perceptions and hence adoption decisions made by the farmers in the study area.
The average size of the families in Chisamba District was observed to be 5 members per hh.This was below the average hh size of 5.5 for Central Province and national average of 5.2 members per hh (CSO, 2011).Groundnut production is labour intensive (MAFF 2000) but because the family size provided it, very few farmers (7.6%) perceived labour pressure as a major constraint to groundnut production in the study area (Figure 1).For example, the majority (93%) of the smallholder farmers in the surveyed area took only three days to plant groundnuts on fields averaging 1.68 ha.This example exemplified that family size of the rural poor was critical for hh labour in the District.
In this study, the literacy level of the household head (hhh) was measured by attainment of 12 years of schooling.It was observed that 34% male and 21% female household heads had attained 12 years of schooling.When segregated by block, literacy levels in Muswishi (53%) were higher than Chisamba block (37%).Educated people may easily follow agricultural instructions (Sichoongwe et al., 2014;Ghimire et al., 2015), apply innovations from extension officers (Ojo et al., 2013) and have the ability to understand profits associated with use of improved varieties (Ghimire et al., 2015).It was therefore, expected to find more adopters of legume technology in Muswishi than Chisamba block.However, this was not the case because there was a non-significant relationship between level of education of the hhh and adoption of new groundnut varieties or crop rotation or incorporating groundnuts in the cropping system in both Muswishi and Chisamba blocks.This means that education was not a factor in influencing perceptions of farmers on adoption decisions.

Farmers' perceptions on groundnut production and risks associated with it
Land allocated to groundnuts was used to determine production in this study.There was a non-significant relationship between gender of the hhh and landownership in both Chisamba (p = 0.124) and Muswishi (p = 0.446) blocks.The majority (92%) of the hh in the study area owned untitled land.This concurs with Hichaambwa and Jayne (2014) who reported that smallholder farmers (70%) in Zambia owned less than 2 ha of untitled land.The mean land holding size of the hh in the study area was observed to be 4.41 ha.Smallholder farmers in Chisamba District (46%) produced groundnuts on fields that were less than half a hectare and 40% on land between 0.5 and 1 ha.A larger proportion of land (78%), 1 to 5 ha, was disproportionately allocated to maize production.It can therefore, be concluded that because of this land disproportion, groundnut production in Chisamba District was done at a small scale level.Comparatively, this is in agreement with Girei et al. (2013) who reported that farmers in Nigeria produced groundnuts at subsistence level on farms which were between 0.1 and 1 ha.The results were also in agreement with Simtowe et al. (2012) who also reported that groundnut production in Malawi was mainly on a small scale on land which averaged 1.32 ha.
Experience is measured by length of time the farmer has used an innovation (Feder et al., 1985).In this study, an experienced farmer in groundnut production was determined by five or more years of consecutively growing the crop.Therefore, the majority (68%) of smallholder farmers in the surveyed area were experienced groundnut producers.It was observed that female smallholder farmers were the major (71%) groundnut producers.On the contrary, Girei et al. (2013) reported that groundnut production in Nigeria was dominated by male farmers (57%).The adoption of mainly groundnut production by male smallholder farmers in Nigeria was attributed to high profits.While in Chisamba, male smallholder farmers perceived maize production to be more profitable.They also perceived maize production as a man's crop because it was the main food crop.In Malawi, men mainly undertook maize production because of traditional predilection and the prominence associated with the crop (Franke et al., 2014).The majority (85%) of the smallholder farmers in the District produced groundnuts for both cash and home consumption.This is in agreement with Ross and Klerk (2012) who reported that smallholder farmers in Eastern province of Zambia produced groundnuts for both consumption and cash.Similarly, Ojo et al. (2013) reported that smallholder farmers in Nigeria produced groundnuts mainly for food and cash.The study revealed that 13% of the respondents in Chisamba perceived groundnuts as just a food crop and therefore produced it exclusively for consumption.These farmers disclosed that groundnuts were eaten raw, boiled or roasted.The thirteen percent farmers also said that they ground kernels to powder and added it to meat or vegetables as a sauce or mixed the powder with maize flour to prepare porridge for both infants and adults.It was also observed that 2% female farmers produced groundnuts solely for sale.There were no male farmers who produced groundnuts merely for sale probably because they focused mainly on production and sale of maize.
It was observed that only 2.4% of the male household heads and 3.7% of the female household heads failed to harvest groundnuts in the 2011/12 to 2013/14 agricultural seasons.There was, however, a non-significant relationship between gender of the hhh and failure to harvest groundnuts in the three agricultural seasons (p = 0.714).The farmers attributed the failure to harvest to late planting (6%), weed pressure (5.5%) and the ground being too hard (3%).The smallholder farmers disclosed that the major bottle necks in groundnut production in the study area were erratic rainfall (24%), poor soil fertility (19%) and weed pressure (15%) (Figure 1).They revealed that lack of access to inputs (12%) such as quality seeds and fertilisers impeded crop production.The findings were in agreement with Orr et al. (2014) and Ross and Klerk (2012) who reported similar constraints associated with groundnut production in Eastern province of Zambia.

Farmers' perceptions on groundnut varieties and inputs
The farmers in the study area were more familiar with the old groundnut varieties, Chalimbana (12%), Makulu Red (16%) and Natal Common (57%) (Table 3).The farmers' familiarity with growing old groundnut varieties was a major contributing factor to adoption.The farmers have been traditionally growing these three varieties since 1950 (Chalimbana), 1964 (Makulu Red) and1976 (Natal Common) when they were introduced.The study therefore, concluded that there was low level acceptability of the new groundnut varieties amongst the smallholder farmers in Chisamba District.
The smallholder farmers disclosed that lack of market contracts with seed companies limited the adoption of new groundnut varieties.They explained that the price of a 10 kg bag of MGV 5 and MGV 4 currently is K 188 ($24.74) which was too expensive for them.
The decision to grow Natal Common (p = 0.149), Makulu Red (p = 0.373), Chalimbana (p = 0.535), MGV 5 (p = 0.414) and MGS 2 (p = 0.492) in the 2011/12 to 2013/14 agricultural seasons did not depend on the hhh (Table 4).However, there was a highly significant relationship between the hhh and production of MGV 4 (p = 0.014) (Table 4).The smallholder farmers' attitude towards MGV 4 was influenced by the existence of a seed growers' club in Muswishi block which was propagating the seed variety.On follow ups to verify the result, the leaders of the seed growers' association explained that they assisted their members to market the surplus kernels.This motivated the adoption of MGV 4.
Therefore, level of education and gender of the hhh may not be the only factors influencing adoption decisions of certain varieties (Table 4).For instance, in this study, the smallholder farmers planted Chalimbana because it was high yielding (p = 0.006).The adopters of Makulu Red planted it because the variety was both high yielding and contained higher oil content (p = 0.008).The perception of farmers who adopted Natal Common were that the variety was high yielding, drought resistant and early maturing (p = 0.045).The smallholder farmers explained that because it was early maturing, Natal Common was planted two times in one agricultural season.The first harvest of January and February was sold as fresh unshelled kernels.While the second harvest of April and May was stored for food, and part of it was sold later in the year as dry shelled kernels.The smallholder farmers interviewed indicated that other varieties were susceptible to pests and drought due to erratic rainfall.They claimed that Chisamba was prone to intra-seasonal dry spells, and varieties like MGV 5 did not survive the dry spells.It can be concluded, therefore, that particular traits of these varieties were very important in affecting the perceptions and hence adoption decisions of smallholder farmers in Chisamba District.This is consistent with Asiedu-Darko (2014) who reported that farmers in Ghana produced traditional varieties based on taste characteristics.
The smallholder farmers in Chisamba did not use certified seeds.The seed was either sourced from previous seasons (77%) or bought from fellow farmers (14%).Only a few (9%) smallholder farmers in the District managed to purchase certified groundnut seeds.The smallholder farmers reported that they purchased certified seeds from local traders who were agents of local commercial seed companies.There was a highly significant relationship between gender of hhh and use of certified groundnut seeds in Chisamba District (p = 0.008).Of the 15 that managed to purchase certified seeds, nine were male smallholder farmers.This means that only 6 out of the 38 female headed households in the study area managed to buy certified seeds.The high cost of certified groundnut seeds limited most (91%) smallholder farmers to recycled seeds from the previous seasons in Chisamba District.This is in agreement with Nzima (2014) who reported that most (67.8%) of the smallholder farmers in Malawi recycled groundnut seeds.The smallholder groundnut producers in Chisamba District revealed that they did not apply manure or inorganic fertiliser to groundnuts.Farmers disclosed that they applied fertiliser to maize only.They said it was too expensive to apply fertiliser to both crops.None of the 164 respondents interviewed applied lime or sprayed herbicides and insecticides to the groundnut field and crop.Therefore, smallholder groundnut farmers in Chisamba District practised low external input agriculture.

Farmers' perceptions on cropping systems in Chisamba District
The three main cropping systems used for groundnut

Does respondent know benefits of crop rotation
Groundnut incorporation vs. fixing N 0.011 Groundnut incorporation vs. increase crop residues 0.439 Groundnut incorporation vs. disease and pests cycles 0.705 Groundnut incorporation vs. weed and labour pressure 0.697 production in Chisamba District were crop rotation (90%), sole-cropping (5%) and mixed-intercropping (5%).The adoption of the cropping system and incorporation of groundnuts in the cropping system did not depend on either the age or level of education of the household head (Table 4).The experience in groundnut production was also not a factor (p = 0.414) in deciding on adopting the type of cropping system the hh used (Table 4).Furthermore, the decision to incorporate groundnuts in the cropping system did not depend on the number of years of experience the smallholder farmers had acquire (p = 0.438).
The smallholder farmers incorporated groundnuts in the cropping system because they perceived it to fix nitrogen and improve yields of the subsequent crops (p = 0.011).This symbiotic ability of the crop to fix N (Bado et al., 2013) was implicitly understood by most (51%) of the farmers in the study area (Figure 2).Only a few (16%) smallholder farmers associated increase of crop residues to integrating groundnuts in crop rotation (Figure 2).These few farmers indicated that crop residues were important not only for their livestock but also for improving the soil structure in their fields.Crop residues are critical in providing substrates for microbes and consequently improving soil structure (Kamkar and Akbari, 2014).
The study observed that only a small (16%) number of smallholder farmers knew that crop rotation helped in the reduction of weed and consequently labour pressure (Figure 2).Only 17% of the smallholder farmers interviewed associated crop rotation with reduction in disease and pest cycles (Figure 2).These few smallholder farmers in the surveyed area disclosed that when they grew maize on the portion previously allocated to groundnuts, diseases and pests were fewer in both the subsequent maize and groundnuts fields.The study therefore concluded that the acceptability of crop rotation as one of the perceived sustainable technologies in improving both soil fertility and groundnut and maize yields in Chisamba District was based mainly on the ability of the groundnuts to fix N.

Measures to facilitate the adoption of new innovation
The study observed that perceptions were critical in influencing the smallholder farmers' attitudes towards an innovation and consequently adoption decisions.For example, inputs in this study were perceived to be very expensive and as a result most farmers (91%) recycled their seeds.Another example is how smallholder farmers (57%) perceived Natal Common to be the most suitable variety to harsh ecological conditions of the study area.Therefore, it is recommended that before new innovations are implemented, the external environment, costs and benefits of such innovations, and socio-economic factors are considered together with the perceptions of smallholder farmers.A holistic approach towards implementation of an innovation is recommended.

Conclusion
Groundnut production in Chisamba District is conducted on small parcels of land and at subsistence level.It was observed that groundnut production is dominated by female farmers.The acceptability of new groundnut varieties in the surveyed area is low.The groundnut variety Natal Common was perceived as suitable for the study area as it could be planted two times in one agricultural season.The smallholder farmers in Chisamba District perceived new varieties to be very expensive, and therefore, recycled seeds.They also perceived fertiliser to be very expensive and therefore, applied it to maize only.

Figure 1 .
Figure 1.Risk associated with groundnut production in Chisamba District.

Figure 2 .
Figure 2. Percentage of groundnut producers who know the benefits of crop rotation in Chisamba District.

Table 1 .
Sample design and description, Chisamba District.

Table 4 .
Summary of chi-square tests.