Two-step cluster analysis suggested four clusters, and the flexibility of this method allowed us to explore cluster solutions from three up to seven. We decided to retain four-cluster solution because cluster solutions above four were not excellent, as they had overall silhouette measures of cluster cohesion and separation values of ≤ 0.7 . The sample size, a limitation of this study, was small because we focused on active land reform farms in a context where their numbers are relatively small , n.d.; Netshipale et al., 2017, and participation was voluntary and based on the ability of a respondent to complete the survey. The power of multivariate analysis techniques is positively related to the number of the observations , hence the sample size was a limitation. We could not use archetypal analysis in this study because the focus was on identifying universal patterns , and this method tends to have limitations regarding ‘validity and boundaries of the archetypes, and selection of appropriate attributes’ . This study provides a snapshot of farming systems which existed in land reform farms during the study period, 2013–2014, as farming systems are dynamic by nature . In this study, we observed that the variables land size, herd size, land use activities and the contributions of activities to farm income and farming costs explained the diversity of farming system types, in line with literature . However, hired labour and farm gross margin did not, though these were reported to be among key variables in farming systems research . Hired labour ratio was similar among farming system types because we analysed labour to assess whether the contribution of hired labour costs to the total labour costs differed between the types. This lack of differences was attributed to the fact that in type H, family labour was permanent and hired labour was casual, whereas in type M, hired labour was not required because the production scale was relatively small. Hence, we observed similarities in hired labour ratios between the types in this study.

We deduce that our findings could hold true even if hired labour was expressed quantitively, as hired labour was permanent in resource-extensive types and was casual and periodic in resource-intensive types . Further, hydroponic bucket the use of both family and hired labour observed in this study is in line with literature about typologies of either farmers or farms or farming systems, though in literature the use of hired labour contributes to the observed diversity among the types . In this study, we chose to use farm gross margin because farm income and farming costs differed between the types and within types . The issues discussed in this section show that relevant variables were used to develop a typology of farming systems in land reform farms of the study area. Hence, we concluded, given that this study focused on previously disadvantaged farmers, that each of the four identified farming system types is efficient in its own way. The diversity of farming systems was reported to arise from, among others, biophysical, economic and socio-institutional variation which were often beyond the control of farming households . The farming system types observed in this study resulted from interactions among land reform policy models, capital endowments of farmers, physical capital endowments of farms and the market type for produce. Farms of type CR and R were present mostly under Rest, LRAD1 and PLAS policy models. Two attributes of these models influence the emergence of CR and R farm types. Firstly, the models transferred farms of relatively large sizes which suited resource-extensive activities . Secondly, capital endowed farmers capable of owning ruminants benefited under these models. The ability of farmers in type CR and R to sell crops and ruminants in formal markets, meant cashflow was certain, a contribution to the emergence of these types as the demand for crop and ruminant products in informal markets was low. The large farm size required for CR and R meant these types of farms could only exist in peri-urban and rural locations, and their economic viability depended on produce being sold in formal markets, which were observed to be reliable than informal markets . We attributed the presence of farming system type with mixed farming , observed in farms of relatively large size, to two factors. Firstly, land reform policies gave farmers of type CR land which was previously used for crop plus ruminants. Secondly, farmers of type CR had interests on crop plus ruminants, hence they continued with these activities.

Our findings suggest that mixed farming emerged in farms of large size and was intended to spread risk through diversification and was in line with literature . Where land reform policy models transferred farms of relatively small sizes either with potential for horticulture farming , 2014:85 or for monogastric farming, under intensive resource use, farming system type H and M emerged. In addition, capital endowed farmers with access to capital required for intensive land uses benefited the most from land reform policies and post-settlement financial support was provided , 2014:15. Literature also acknowledges the need for financial support, especially where resource-intensive activities are practiced by smallholders or in small farms . The small farm size required for type H and M meant farms of these farming system types could exist in urban, peri-urban and rural locations. Farms of type H prevailed the most in peri-urban location because markets for farm inputs were often in urban centres and farmers sold produce in formal markets to ensure that farms were economically viable. Type M farms were in peri-urban and urban location because farm physical capital limited the scale of production, which meant production levels could not meet the demand in formal markets. Hence, type M farms could only exist next to locations where relatively large populations live as produce were intended for informal markets. In line with our findings, Nesamvuni et al. reported that horticulture production is dependent on suitable natural capital , and availability of physical and financial capital. We used literature on the biophysical conditions in the study area to make deductions regarding the influence of the interactions between these conditions and land reform policy models on the presence and emergence of farming system types. Literature declared the study area to be semi-arid with poor water sources , 2014:85. Land reform policy models were cognisant of the biophysical conditions as farms of relatively large sizes, which suited resource-extensive activities , were transferred under Rest, SLAG, LRAD1 and PLAS . It was this cognisance of the biophysical conditions by the policy models which led to irrigation being a key land use activity in fewer farms compared to farms where it was not . Ruminant farming is the most suited agricultural activity for semi-arid conditions , 2014; hence this was the key activity in 48% of the investigated farms. We attributed the low prevalence of type H farms to targeting of the economically disadvantaged section of the society by land reform policies, in addition to the biophysical conditions, but that of type M farms was solely due to the former ,1997, 2006; MALA , 2001.Agriculture, or farming as it is commonly known, is the practice of growing crops and raising cattle. It contributes greatly to a country’s economy. Many raw materials and food products are produced by agriculture. Raw materials such as cotton, jute is used by industries for manufacturing various products that is used in day-to-day life. Agriculture not only helps for food production but also produces resources needed for creating commercial products.Conventional or traditional farming is mostly practiced all over the world. It involves techniques suggested by experienced farmers. These techniques are not precise hence results in hard labor and time consumption.

The application of digital technologies which includes robots, electronic devices, sensor and automation technologies is associated with Precision Agriculture. This technology aims to reduce workloads, increase profitability and decision management. Precision agriculture additionally referred to as precision farming is a farming control system that provides a comprehensive approach to deal with the spatial and temporal crop and soil variability to maximize profitability, optimize yield, improve quality of production. Precision Agriculture is an efficient way to improvise the yield. On discussing about the adoption rate of precision agriculture, the high value enterprise farms adoptionrate was more compared to low value enterprise farms. The adoption rate of precision agriculture also depends on the country and the geological locations. The adoption rate of Precision Agriculture in mountain zone is less compared to farmers in the valley. The variation of adoption rate is due to the high investments needed. Hence there needs a way to reduce cost on machines hence all farm-size can adopt precision agriculture. Precision agriculture is aided by advanced technologies such as IoT, Data Mining, Artificial Intelligence, and Data Science. The Internet of Things is a network of interconnected computational things like sensors and smart gadgets that can communicate with one another and share data. In agronomic applications, wireless sensor networks are being used to remotely monitor ambient and soil characteristics in order to predict crop health. Using WSN as a forecasting approach, the watering schedule of agricultural fields can be predicted. Wireless Sensor Networks acquire data from external variables such as pressure, humidity, and temperature, as well as soil moisture, salinity, and conductivity. Machine learning makes agricultural applications incredibly efficient and simple. Data acquisition, model building, and generalization are the three stages of the machine learning process. The majority of cases, machine learning algorithms are used to deal with complex problems when human competence is insufficient. Machine learning may be used in agriculture to forecast soil parameters like organic carbon and moisture content, as well as crop yield prediction, disease and weed identification in crops, stackable planters and species detection. Traditional machine learning is improved by Deep Learning by adding additional complexity to the model and changing the input with various functions that allow data representation in a hierarchical manner, through multiple levels of abstraction, depending on the network architecture employed. A significant benefit of Deep Learning is feature learning, or indeed the automatic extraction of features from original data. The ability to identify unknown things such as anomalies rather than just a collection of existing items is a key aspect of the Deep learning model, which uses the homogeneous properties of an agricultural field to discover faraway, badly obstructed, and unknown objects. Blockchain has swiftly become a key technology in a variety of precision agriculture applications. The requirement for smart peer-topeer systems capable of verifying, securing, monitoring, and analyzing agricultural data has prompted researchers to consider developing block chain-based IoT systems in precision agriculture.

Block chain plays a critical role in transforming traditional methods of storing, sorting, and distributing agricultural data into a digital format a way that is more dependable, immutable, transparent, and decentralized. The combination of the Internet of Things and block chain in precision farming results in a network of smart farms. More autonomy and flexibility are attained as a result of this pairing. The above-mentioned technologies such as IOT, Data Science, Machine Learning, Deep Learning, Block chain deals mostly with data which are very useful for understanding and providing great insights of data. Hence, these advanced technologies are used in various agriculture practices such as identifying the best crop for a particular location, identifying factors that would destroy the crops such as weeds, insects and crop diseases to obtain insights about the crop growth and help in decision making. Agriculture can be divided into 7 important steps that includes Land Management, Soil Preparation, Water Monitoring, Identifying the weeds, Pesticides Recommendation, identifying diseased crops, and cost estimation. Land Management refers to the monitoring physical features that includes weather conditions, geological characteristics. This is important since there are variations in climatic conditions across the globe which would affect the crops. Rainfall is an important aspect of the earth’s climate, and its unpredictability has a direct impact on agriculture, water management systems, and biological systems. As a result, tools that assist in predicting rainfall in advance are required so that crop management can be simplified. Soil is an essential component of agriculture. Rooting, moisture and nutrient storage, mineral reserve, anchoring, and a variety of other variables that affect plant growth are all determined by soil depth.