Modeling the influence of existing feeding strategies on performance of grade dairy cattle in Vihiga, Kenya

P M Ongadi, R G Wahome*, J W Wakhungu* and L O Okitoi

Kenya Agricultural Research Institute – Kakamega, Kenya; P.O Box 169-50100, Kakamega
p.mudavadi@cgiar.org   ;   ongadimp@yahoo.com
* Animal Production Department, University of Nairobi, Kenya; P.O. Box 29053, Nairobi

Abstract

A modeling study was carried out on grade dairy cattle in four production systems in Vihiga District. The objective was to evaluate the effects of existing feeding strategies on performance of grade dairy cattle. Data for the model was extracted from results of a survey of 236 grade dairy cattle owning households in Vihiga District. Results showed that feeding strategies for grade dairy cattle in Vihiga District were sub-optimal reflected in low actual and potential milk yields per cow per day. Protein was a major limiting nutrient and the situation was serious during the dry season when low quality forages were available. Further, the costs milk production was higher in intensive production systems as opposed to the extensive production systems.

 

The most optimum existing feeding strategies for Vihiga in terms of economic returns by grade dairy cattle production systems were: a) The basal feed comprising napier grass cut and carry supplemented with dairy meal and other fodder (a mixture of fodder trees and legumes and sweet potato vines) in Stall feeding only and Grazing only production systems, and b) The basal feed comprising  natural pastures and napier grass cut and carry supplemented with dairy meal, other fodder  and crop residue in Mainly stall feeding with some grazing and Mainly grazing with some stall feeding production systems.

 

In conclusion, supplementation of the basal diets with dairy meal and fodder as single supplements or components in compound feeding strategies was necessary in Vihiga for enhanced performance of grade dairy cattle in terms of milk yields, live weight gains, manure production and economic returns.

Key words: Basal and optimal feeding strategies, forages, manure, stall feeding, supplementation, survey

Introduction

Smallholder mixed farming systems in Vihiga, Kenya are characterized by varied agricultural activities including cultivation of food crops and cash crops, as well as milk production (Bebe et al 2002; Salasya 2005). Development of dairy systems on these smallholder farms is limited mainly by land shortage and hence feed supply among other factors resulting into low animal productivity (Odima et al 1994; Omore et al 1999; Staal et al 2001; Waithaka et al 2002). Strategies employed to alleviate the limited feed supply and hence improve animal productivity under different dairy cattle production systems include feeding of crop and agro-industrial by-products, fodder cultivation on roadsides and reliance on purchased fodder (Omore et al 1999; Mwangi and Wambugu 2003).

 

However, feeding strategies and practices adopted by farmers for their dairy cattle are often opportunistic, characterized by intermittent and abrupt changes in the quantity and quality of the feeds offered (Methu et al 2000, Pezo 2001). Consequently, feeding strategies are not related to the expected nutritional requirements of the animals kept limiting performance (Delgado et al 2001; Bebe 2003). This study was carried out to evaluate the influence of existing grade dairy cattle feeding strategies on milk yields, live weight changes, manure production, methane emissions and economic returns under four grade dairy cattle production systems in Vihiga.

 

Materials and methods 

Study area

 

Data for this study was collected in Vihiga, Western Kenya, which is a high agricultural potential area predominantly (95%) in the upper midland one (UM1) agro-ecological zone, with an altitude ranging between 1300 to 1800 metres above sea level, average temperatures of 20.3⁰C and well drained soils that comprise dystric acrisols and humic nitrisols (Jaetzold and Schmidt 1983). The area receives bimodal rainfall that ranges from 1,800 – 2,000 mm/year.

 

Data collection

 

Data for simulation analysis was summarized from a purposive sample of 236 grade dairy cattle owning households using a pre-tested structured questionnaire. Existing feeding strategies were stratified under four grade dairy cattle production systems in the area, namely stall feeding only, mainly stall feeding with some grazing, mainly grazing with some stall feeding and grazing only. Information was collected on feeds offered to grade dairy cattle under each feeding strategy (basal feeds and supplements) and their quantities, their cost per kilogram and milk yield per cow per day.

 

Data analysis

 

Model description

 

The Dairy Simulation Model v3.2, which is part of the Livestock Feeding Simulation “LIFE-SIM” Models group (Quiroz et al 2005) developed by the Natural Resources Department of the International Potato Centre (CIP) was used to model (simulate) the influence of existing feeding strategies on grade dairy cattle performance in terms of milk yield, economic gross margins, manure production and methane emissions under four grade dairy cattle production systems in Vihiga. The model is deterministic and its inputs include specific data for animal description, voluntary intake, nutrient requirements, milk production, manure production, methane emissions, thermal regulation, pasture growth and supplement availability. The model is fully described by Leon-Velarde et al (2005).

 

Model inputs

 

The animal

 

The average grade dairy animal was a 3.5 year old Ayrshire cross cow (the most common breed type in Vihiga) averaging 300 kg BW with a potential lactation yield of between 2000-2500 kg over a 305 day lactation period. The expected calf birth weight was 24kg and a lactation length of 10 months (301 days). The chemical composition of the cow’s milk was 4.0%, 3.3%, 8% fat, protein content and solids not fat (S.N.F) respectively. The loss of weight during the first three month of lactation was estimated at 6%, within the range between 5-7% for crossbred cattle allowed by the model specifications.

 

Adjustment values

 

The energy (expressed in metabolisable energy, ME) and protein (expressed in terms of total protein, N*6.25) were adjusted by 6.5 Mcal/kg body weight (BW) and 20% based on the energy and protein concentration required to gain 1 kg of live weight respectively.

 

Potential dry matter intake (PDMI)

 

The potential dry matter intake was 3.12 kg/cow/day, determined from the reference table of live weight (LW) and metabolic weight (MW) provided in the model specifications. A stochastic variability of 5% was added to cater for the animal’s inherent variable attitudes over a period of days. The correction factor for the influence of dry matter intake on milk production was 0.1, and this ranged between 0.1-0.15 in the model.

 

Potential milk yield

 

Potential milk yield was determined based on the four grade dairy cattle production systems in relation to the cow’s body weight (i.e. 300kg BW for Ayrshire crosses), parameters for the milk production (lactation) curve derived from the Wood’s equation (1967) quoted by Leon-Velarde et al (2005) and actual milk yield/cow/day. Therefore, the parameters for the lactation curve were: a = Actual milk yield, kg/cow/day for Vihiga (5.443, 5.801, 5.041, 5.40 for stall feeding only, mainly stall feeding with some grazing, mainly grazing with some stall feeding and grazing only production systems respectively, Ongadi et al 2007), b = 0.2582, and c = 0.00715. Once a, b and c were specified, the model automatically generated over 305 days lactation period, the yield at peak lactation, days at peak lactation and milk production per lactation.

 

Basal feeds

 

Natural pastures

 

The availability of natural pasture, the basal feed resource in extensive grade dairy cattle production systems (grazing only and mainly grazing with some stall feeding) in Vihiga was 350–700 kg DM/ha per year depending on the rainy season. The wet season was from March-July and October-November, while the dry season was from December to February and August-September. Natural pastures had a digestibility of 50-60% and a protein content of 4.5–7%. The energy cost of harvesting feed (grazing correction factor) was 5-30% of the maintenance requirements, accounting for locomotion. This value was lower for stall feeding only production system (5%) and higher for grazing only production system (30%). The stocking rate was 1.2A.U/ha (1A.U = 300kg of B.W).

 

Cut and carry

 

Fresh napier grass (Pennisetum purpureum Schum.), the basal feed resource, was offered at between 35–65kg per cow per day and had a dry matter content of between 17–22%.  Fresh napier grass offered depended on grade dairy cattle production systems and rain seasons (wet and dry). More was offered under stall feeding only production systems as opposed to the other grade dairy cattle production systems. Digestibility of napier grass ranged between 50–65% depending on the season with a protein content of 7–10% (Schreuder et al 1993).

 

Supplementation

 

Supplements were classified as (a) concentrate (dairy meal), (b) protein rich fodder that was a mixture of fodder legumes/fodder trees and sweet potato (Ipomea batatus) vines in the ratio of 0.3 (25%) and 0.7 (75%) respectively and (c) crop residue (mainly maize (Zea mays) stover). Nutrient contents of the supplements and basal feeds were specified into the model before formulating the different feeding strategies (rations) as indicated in Tables 1 and 2 of description of existing feeding strategies.

Cost

 

The calculated cost of natural pastures and napier grass was KES 0.67 and KES 1.25 respectively based on their estimated yields per ha. Napier grass yield was between 10 to 40 tonnes DM/ha (Schreuder et al 1993) depending on soil fertility, climate and management. The yield of tropical natural pastures was 500kg DM/ha (Boonman 1997). From figures obtained from the Ministry of Livestock and Fisheries Development, Vihiga district annual reports (Anonymous 2004), the average yield of napier grass in the district was 20 tonnes DM/ha. The unit of trade was a wheelbarrow of napier grass weighing about 25kg and costing KES 50.00 on average. Therefore, one tonne of napier grass gave 40 wheelbarrows and 20 tonnes DM/ha gave 500 wheelbarrows costing about KES 25000.00 (500 x 50.00), which when divided by yield/ha in kilograms (i.e. 20000 kg DM/ha), gave a napier grass cost of KES 1.25.

 

The unit of trade of natural pastures in Vihiga was a sack-load of natural grass weighing about 15 kg and costing KES 10.00. Farmers in Vihiga gave away natural grass for free when available or sold for as little as KES 10.00 per sack-load. Therefore, a natural grass availability of 500 kg DM/ha gave about 33.33 sack-loads costing KES 333.33, which when divided by 500 kg DM/ha gave a cost of KES 0.67.

 

Feeding costs were 75-80% of the total milk production costs per year based on the level of intensification (grade dairy cattle production and feeding systems) and were higher for intensive production systems (stall feeding only and mainly stall feeding with some grazing) as opposed to extensive production systems (grazing only and mainly grazing with some stall feeding production system). The average cost of milk/litre in Vihiga was KES 30.00.

 

Description of existing feeding strategies

 

The main feeds for grade dairy cattle, summarized from the data were entered into the data base of feeds provided in the model. Their nutrient contents in terms of Dry matter (DM), Crude protein (CP), Digestibility (Dig) and Metabolisable energy, ME (Dig*3.64) as obtained from literature (Quiroz et al 2005: Leon-Velarde et al 2005; Abdulrazak et al 1996; Muinga et al 1992, 1993, 1995; Kariuki 1998; Muia 2000; Anindo et al, and Porter 1986) and cost per kilogram of feed in KES were then specified. Once these were specified, the model automatically calculated the cost per ME (KES/Mcal) and CP (KES/kg) as indicated in Table 1.

 

The feeds were categorized in the model as a) basal feeds (napier grass and natural pastures), b) supplement 1 which was the concentrate (dairy meal), c) supplement 2 which was protein rich fodder (a mixture of sweet potato vines and fodder legumes/trees in the ratio of 0.7 (75%) to 0.3 (25%) respectively) and d) crop residue which was mainly maize stover. Using the average quantities summarized from the data (Table 2), these feeds were then balanced and formulated to make the different feeding strategies or scenarios for grade dairy cattle production systems as indicated in Table 3. The model automatically generated the nutrient values of the formulated rations (feeding strategies) as indicated in Table .3.

Simulation

 

Scenarios were generated based on based the model inputs described above for every existing feeding strategy in each of the four grade dairy cattle production systems in Vihiga. The outputs of the dairy model included the expected milk yield during the lactation period, the changes in body weight during the same period, the amount of manure produced and an estimate of methane emissions.

 

Model validation

 

Average fresh feed intakes of the different grade dairy cattle feeds and actual milk yield per cow per day summarized from the data collected from grade dairy cattle owning households by grade dairy cattle production systems (Table 2) were fitted in the model to determine validity and accuracy of the model in assessing influence of existing feeding strategies on performance of grade dairy cattle in Vihiga.

 

Critique of the model

 

• The model adequately estimated and reflected reality on milk production per lactation for Vihiga, but tended to overestimate growth and live weight gains by mature grade dairy cows over the lactation period.

  
  

• The range of solids not fat (S.N.F) in milk specified in the model (i.e. 7.5-12%) was high and resulted in overestimation of total solids in milk.

  
  

• Calf birth weight was set default at 28.0kg, though Ayrshire crosses in Vihiga had a lower birth weight for calves (24.0kg).

Results and discussion 

Basic and optimal feeding strategies

 

Basic feeding for each grade dairy cattle production system comprised napier grass alone, natural pastures alone or a combination of napier grass and natural pastures (Table 4).

Dry matter intakes were higher for the optimal feeding strategies that is, when basic feeding strategies were supplemented with dairy meal and protein rich fodder (a mixture of sweet potato vines and fodder legumes/trees) in all the four grade dairy cattle production systems.

 

However, in all the four grade dairy cattle production systems in general, supplementation levels and hence dry matter intakes were low and this was reflected in performance (Table 4) for both the optimal and basic feeding strategies. Quantities of high protein forages were not adequate for supplementing lactating cows as similarly observed by Mwangi and Wambugu (2003). In addition, supplementation using commercial concentrates was at minimal levels mainly because of the high costs in relation to milk prices (Abate and Abate 1991; Abdulrazak et al 1996).

 

As indicated in Figures 1a and b, dry matter intake of basic and optimal feeding strategies varied over the lactation period, mainly due to the seasons (wet and dry) that influenced feed availability.

Dry matter intake was higher during the wet season than the dry season. It was also lower in extensive production systems (grazing only and mainly grazing with some stall feeding) compared to intensive production systems (Table 4). Deficiencies in energy and protein supply to grade dairy cows were greater with the basic feeding strategies in all production systems, affecting dry matter intakes, milk production and live weights (Figure 1). Difficulties in bridging these gaps in energy and protein supply were as a result of inadequate forage both in quantity and quality for the grade dairy cattle. This was mainly because of diminishing land sizes and seasonality in forage production.    
 

Simulated live weight change

 

Simulated live weight at the end of lactation (301 days), live weight at the end of the year (365 days) and live weight after calving were lower when grade dairy cows were fed basal diets alone without supplementation in all production systems (Table 5).

Similarly, average daily weight change (gain or loss) after end of lactation and per year was lower when cows were offered basal diets without supplementation. Inclusion of crop residue in the feeding strategies resulted into higher live weight at the end of lactation, at the end of year and after calving in all the production systems except grazing only (Table 5).

 

Live weight change during the lactation period in all the four grade dairy cattle production systems highly depended on the quantity and quality of dry matter intake from all the existing feeding strategies. Live weight change by lactating cows from existing feeding strategies, was influenced more by protein than energy supply during the lactation period. The influence varied with the rainy season (wet and dry) as indicated in Figures 2a and b.

Initially cows lost weight, during the first few months of lactation but eventually regained weight as lactation progressed (Figures 2a and b).

 

However, cows lost weight towards the end of lactation when their basal diets comprised natural pastures or a combination of napier grass and natural pastures without supplementation as similarly observed by Kariuki (1998) and Muia et al (1999). The reported lower milk yields and greater weight losses in cows offered basal diets only, compared to those offered basal diets supplemented with concentrates (Anindo and Porter 1986; Muinga et al 1993) or forage legumes (Muinga et al 1995; Abdulrazak et al 1996) were consistent with our simulated results.

 

Simulated milk production

 

Lactation and daily milk production for grade dairy cattle was lower than potential milk production in all production systems in Vihiga (Tables 4 and 6). The production was lower when cows were offered basal diets without supplementation. Inclusion of dairy meal and protein rich fodder in the feeding strategies resulted in increased milk production in all production systems. Including dairy meal alone or with crop residue in feeding strategies for all production systems resulted into low lactation milk production in all production systems.

 

Lactation milk production was more innately related to the protein supply than energy supply from existing feeding strategies under all production systems (Figures 3a and b).

This was similar to experimental findings that higher milk yields could be obtained when basal diets were supplemented with high energy and protein content feed resources (Combellas and Martinez 1982; Anindo and Porter 1986; Van Bruchem et al 1989; Muinga et al 1992, 1995; and Mukisira et al 1994). In general, potential and lactation yield with basic and optimal feeding strategies was far below the genetic potential of the grade dairy cattle (Figures 3a and b) and was attributed to inadequate levels of feeding with existing feeding strategies and low quantity and quality of basic diets, especially during the dry season, as similarly observed by Valk et al (1990), Reynolds et al (1996) and Bebe (2003).

 

Simulated waste production

 

Feeding basal feeds comprising napier grass alone or napier grass and natural pastures resulted into more manure production/cow/year than when animals were fed natural pastures alone in grazing only production system (Table 6).

Similar to observations by Lekasi et al (1998), manure production was higher in all production systems where napier grass was included as a basal feed in feeding strategies. Generally, natural pastures based feeding strategies in grazing only production system resulted into the lowest manure production/cow/year.

 

Methane, a by-product of milk production, was low when basal diets were offered without supplementation in all production systems. Further, natural pastures based feeding strategies in grazing only production system resulted into the lowest methane emissions/cow/year. Ulyatt et al 1997 and Pradel et al 2006 supports our findings that pasture and fodder quality and feed intake were innately positively linked and thus when pasture digestibility increased, consumption also increased leading to more methane emissions

 

Simulated economic assessment

Lower total production costs, gross incomes, gross margins/cow/year, daily gross incomes/kg milk and income-cost ratios were realized from feeding strategies that comprised basal feeds alone without supplementation in all production systems (Table 6). In fact, heavy losses were realized in all production systems except grazing only when basal diets were offered without supplementation. Production costs per kilogram of milk were, however, higher when basal feeds were offered without supplementation. Gross incomes from milk were higher from exiting feeding strategies utilized in intensive production systems (stall feeding only and mainly stall feeding with some grazing) as opposed to extensive production systems (mainly grazing with some stall feeding and grazing only). Similarly, feeding napier grass supplemented with protein rich fodders in Stall feeding only production system resulted in higher income from milk (Table 6).

 

Feeding natural pastures supplemented with dairy meal and natural pastures supplemented with dairy meal and protein rich fodders in grazing only production system, though resulting in lower gross incomes than in the other production systems, had the highest gross margins, daily gross income/kg milk and income-cost ratios as a result of lower production costs per kilogram of milk. Feeding napier grass alone or in combination with natural pastures without supplementation in all production systems except grazing only resulted in loss of revenue because of the high costs of napier grass production as similarly observed by Muia (2000). Generally, supplementing basal diets for existing feeding strategies in all production systems resulted into increased returns. The high costs of milk production with existing feeding strategies under intensive production systems as opposed to extensive systems reflected high cost of concentrate feed used (Staal et al 2003).