«Embargo information No embargo required Item downloaded from Downloaded on 2016-05-23T13:09:50Z Thesis to obtain the ...»
Correlations were calculated between milking practices and the milk hygienic status of raw milk (Table 11). In contrast to expected results, the access to infrastructure and milking equipment found in largescale farms did not mean better milk hygienic status. Indeed, gathering cows in a waiting yard, use of a milking parlor or milking by mechanical means were negatively correlated with values of MBRT. The poor cleaning regime of these buildings (less than once a day), or a lack of a deep cleaning of milking equipment could explain this result. On the contrary, some hygienic practices such as cleaning animal houses at least once per day, using a disinfectant to clean the udder before milking and filtering milk before pouring it in churns, was more effective in improving MBRT. Similarly, values of milk somatic cells showed a positive correlation to low cow dirtiness score. Moreover, a decreased level of SCC was also obtained when the animal house was cleaned at least once per day and when dairy farms had permanent access to a source of potable water.
Processors in the Mantaro Valley collect milk personally at the farm gate, or by buying milk collected by independent collectors. These two logistical options affect milk quality differently according to the component considered. Milk chemical quality parameters were quite similar in both cases (Table 12), as this component is quite stable along the chain and since farmers are not selected by buyers according to their TS performances.
Milk collected by processors themselves showed lower SCC than milk collected by independent collectors. Indeed, dairy processors had more control of the milk collected, since they usually dealt with less than 30 farmers whose milking practices they knew, compared with independent collectors who collected milk from more than 100 farmers.
Table 12 Comparison of milk quality values at processor’s gate according to the plant supplier
MBRT. Indeed, while fresh milk measured at farm level showed an average MBRT time of 460 minutes, it decreased to 240 minutes after almost 12 hours of storage at Mantaro Valley’s ambient temperature (Figure 8).
Methylene blue reduction test (minutes).
Figure 8 Variation in hygienic milk quality at Mantaro Valley. Milk samples were analysed at three times separately: At farm level, during milk collection after almost 12 hours under Mantaro Valley’s conditions, and at plant gate; using methylene blue reduction test (minutes). Black line: minimum recommended value (Suggested values based on Norma Técnica Peruana 202.001 – INDECOPI and R.M. 591 – 2008/MINSA).
The degradation process continued from milk collection at farm gate to its delivery at plant gate, with one hour of extra reduction (180 minutes). These results highlighted the importance for both processors and collectors to better manage logistics between farm and plant in order to reduce degradation of milk. The current logistical organization, based on one daily collection by various vehicles from bicycles to medium-sized trucks, and with plastic containers where milk from various farmers is mixed when individual deliveries are too small, will not improve milk quality.
The first demonstration of the technical innovation received positive reactions from small-scale farmers and dairy processors. Most of them were ready to buy a UMA, since both groups saw the opportunity to better monitor the quality of the deliveries. Processors identified the possibility to change their delivery system in order to individually evaluate each farmer at the plant gate and separate farmers with problems, whereas farmers suggested comparing quality per cow or throughout the year. This result illustrates the capacity of this quite simple but powerful technology to make stakeholders react.
Nevertheless, general lack of knowledge regarding the use of the UMA was observed at Mantaro Valley after the dissemination of the technical innovation.
Differences in understanding milk chemical values provided by UMA varied according to the size of the dairy farm (Figure 9). Indeed, the first reports provided by the equipment were not clearly understood by most of small-scale dairy farmers (more than 80%), but easily interpreted by the majority of largescale farmers (more than 70%). All the large-scale farmers were able to explain in advance possible relationships between milk chemical values and their husbandry practices compared to less than 40% in the case of small-scale farmers. The previous participation of large-scale farmers in a milk quality based payment system played an important role in this result.
Figure 9: Different reactions about the use of UMA according to the scale of the dairy farmers
systems. However, only large farmers reported a clear understanding of advantages and disadvantages of a quality payment system based on UMA results. Sixty percent of small-scale farmers were not able to identify arguments against it, highlighting the risks of failure in implementing a milk quality payment system in the study area if there is not a clear explanation of the rules (premiums and penalties) to all the stakeholders.
From the processor side, the main reason for purchasing the MasterEco was to detect milk adulteration (usually a mix of milk and water or milk and whey) (Figure 10). Although the analyzer provides a range of variables that can be used for improving their manufacturing process or providing support to dairy farmers, dairy processors were mostly focused on evaluating density and water added, both because of their main adulteration concern. In this respect, the lack of training when dairy processors received the device seems to explain the poor interpretation of the milk chemical parameters.
Figure 10: Use of UMA by dairy processors in Mantaro Valley The use of the equipment in the study area enabled some dairy processors to get rid of farmers or collectors delivering poor quality milk. However, a lack of discussion of the results with the rest of stakeholders created conflicts between farmers and processors, such as the discomfort of from some milk suppliers and the movement of some dairy farmers to less demanding dairy processors. These conflicts also led to a reduction in the frequency of use of the UMA. Indeed, at the time of our
two weeks and only 10% daily. Others factors limiting the frequent use of the equipment were the processor’s’ concern about the possible disfunctioning or discalibration of the equipment, and the payment of unnecessary extra cost for the maintenance of the equipment.
The use of UMA in the implementation of milk quality payment systems was suggested to dairy processors. They expressed a contrasting point of view. Most of them agreed that implementing quality controls and milk quality payment systems will allow them to have the security that they are paying a profitable price for the milk they are buying. Nevertheless, their main concern was that it will probably demand trained personnel, time and a better management of the information, three things with which small-scale dairy processors still have problems. In-depth analysis of milk quality results could improve small-scale dairy processors’ performances. However, they do not have adapted tools to help them make the best decisions.
3.3. Discussion 3.3.1. Dealing with a large milk quality variability Results regarding milk quality analysis showed the large variation in milk composition and hygienic values that exist in Mantaro Valley. This variability also emphasizes the importance of having adequate milk quality controls and sufficient knowledge for understanding the impact of milk quality parameters on the manufacturing process. Hygienic status of raw milk at farm level was often good compared to other contexts in developing countries where hygienic contamination is a critical issue (Sraïri et al., 2006; Grimaud et al., 2007). But the considerable deterioration of hygienic milk quality during milk collection indicated the poor hygienic management and lack of interest in the hygienic status of raw milk (Koussou and Grimaud, 2007; Srairi et al., 2006). Poor processors’ logistic aspects such as the use of unclean containers (Kivaria et al., 2006), mixing fresh milk and milk from the previous afternoon (Millogo et al., 2008); a longer time between fresh milk collection at farm gate and plant delivery (Gran et al., 2002), and lack of cooling facilities during the rainy season when the temperature is warmer (Grimaud et al., 2009) increased milk deterioration of the whole batch. This
products. Thus, they are obliged to improve milk collection logistics in order to ensure the safety of their dairy products, reduce the degradation of milk batches and justify the effort to push dairy farmers to improve hygienic milk quality.
Chemical quality variability is partly explained by the effect of climatic conditions on farmers’ feeding strategies (Larsen et al., 2010), which influence farmer’s capacity to effectively react to the imbalance between stocking rates and forage production. Indeed, low availability of good quality fodder, especially during the dry season, plus poor animal genetics (Bartl et al., 2009) and lack of capital, limit smallholders’ capacity to permanently invest in animals, milking machines, purchase of land, or new technologies (Solano et al., 2000). In this respect, small-scale farmers could improve milk chemical quality through the implementation of collective actions between famers i.e. creating farmers’ associations or cooperatives (Dulcire et al.,, 2012), which may support the common investment in land and technology for the benefit of all the members.
3.3.2. Constraints of introducing technical innovations Literature on agriculture highlights two major drivers of successful technology adoption in developing countries: (i) the availability and affordability of technologies; and (ii) stakeholders expectations that adoption will remain profitable (Kasirye, 2013). Other factors affecting technology use are related to characteristics of the technology and objectives of the stakeholders (Doss, 2006). Despite the fact that the UMA proved to be very easy to handle, its introduction at Mantaro Valley was not supported adequately with sufficient information about milk quality issues and management practices. Moreover, implementing milk quality controls based on UMA results were not considered attractive to stakeholders, since farmers did not receive incentives for improving their current status. Indeed, dairy processors’ objectives were oriented mainly towards a strict control of milk quality adulteration, instead of building milk quality payment systems to attract new dairy suppliers, motivate the rest of milk producers to focus their efforts on farm management practices (Botaro et al., 2013) and improve the general milk quality status at plant gate (Nightingale et al., 2008).
Results showed poor knowledge about milk quality in the study area, similar to other dairy supply chains in developing countries where the informal sector is predominant (Srairi et al., 2006). Indeed, quality was not a concept considered important by stakeholders when this study started. But, measuring quality directly on farms and dairy plants helped both farmers and processors to clarify their ideas about this concept. Dairy processors started to be more conscious about the need to individually identify each farmer at plant gate and to demand better milk quality, but still without realizing the impact that improving milk quality standards can provide to their business. For these reasons, processors were not ready to pay more for better milk quality even when asked by farmers and no main changes were observed during the 12-month monitoring of milk quality in the area. In that respect, the design of a simulation tool could provide the support to show dairy processors the benefits of implementing quality controls and quality based payment systems.
and milk payment systems.
The present chapter shows the usage of a decision support system which deal with strategic issues that dairy processors face in interaction with their suppliers and buyers, i.e. selecting their product portfolio according to markets opportunities and designing milk payment systems encouraging dairy farmers to supply good quantity and quality milk throughout the year. The approach tested with two small-scale dairy plants in the Mantaro Valley (Peru) showed that (i) they could increase their total profits by modifying their current portfolio towards higher value products, assuming milk delivered to the plant attains a given quality; (ii) they do not pay correctly farmers who deliver good quality milk and overpay some bad quality milk; (iii) their profits would not be affected by adopting a payment system based on milk quality. More detail information is available in volume 2 of the present document.
4.1 Materials and Methods This study was conducted in two steps. The first one consisted in designing and developing a simulation tool called DairyPlant able to calculate the processor’s profits and farmers’ gross products corresponding to a given configuration of dairy plant and of milk payment system. The second one focused on testing DairyPlant with two small-scale dairies, by supporting them in better selecting their market orientations and evaluating the impacts of implementing new milk payment systems in their current supply and processing systems.
DairyPlant was designed based on a participatory research conducted with five small-scale dairy processors in the Mantaro Valley. They were monitored weekly from May to July 2013 in order to estimate production functions from raw milk to dairy product. Then, two of them were selected, based on their predisposition to adopt innovative incentives, to carry out the support process and to discuss the feasibility to implement payment systems including milk quality components.
Quantitative data were collected such as volume of milk collected per day, dairy product produced, price of dairy products, cost of processing dairy products, as well as qualitative ones, such as ways of