Livestock Research for Rural Development 18 (3) 2006 Guidelines to authors LRRD News

Citation of this paper

Trade offs between economic returns and methane greenhouse gas emissions in dairy production systems in Cajamarca, Peru

W Pradel, D Yanggen and N Polastri*

International Potato Center, Peru. Apartado 1558, Lima 12, Peru
w.pradel@cgiar.org     d.yanggen@cgiar.org

*La Molina Agrarian University, Peru

Abstract

Cajamarca is one of the Departments in Peru with highest poverty rates; however, it is one of the Departments with greatest potential for agricultural development through its comparative advantages on milk production. The purpose of this study is to examine the tradeoffs between poverty alleviation and environmental contamination in the context of dairy production in the Department of Cajamarca. The study is based on data from surveys collected between 2000 and 2001 in La Encañada District in Cajamarca. Data were subsequently analyzed using the Dairy Simulation Model and the Trade Off Analysis Approach to evaluate the different technologies suitable for the region to increase milk production (through feed and cattle improvement, or in other words, land use intensification), comparing the income increase with the methane emissions. Finally, data was scaled up to measure the effects of land use intensification in poverty alleviation as well as global warming generated with the methane emissions produced by cattle.

The results of the study highlight the importance of reducing poverty in the region through the promotion of milk production with improved technology. Methane emissions produced by dairy cows are higher than the total greenhouse effect of cars in Cajamarca. Even though there is an impact of methane emissions on the environment because of such intensification, there are many benefits that make those technologies desirable. When feed and cattle are improved, there can be an increase up to US$ 411 per year for households with just 1 to 4 cattle, enough to bring a substantial proportion up over the poverty line. Around 67 000 farm households in Cajamarca Dairy watershed belong to this group; thus the impact on poverty can be significant. But it is not the only positive effect, it will also be a reduction in malnutrition and there will also be a reduction of the land pressure in rangeland by improving water management. Considering pros and cons of the technologies proposed, methane emissions, given the urgent necessity to alleviate the poverty of farmers in the region, rye grass improved pastures and improved dairy cow breeds offer a promising alternative for the zone, if irrigation infrastructure and/or credit program are established..

Key words: Agricultural intensification, methane emissions, milk production, poverty alleviation, simulation model, trade off analysis


Introduction

Global warming and Peruvian greenhouse gas emissions

The "greenhouse effect" is a natural phenomenon necessary for life on Earth. It is the absorption of part of the solar radiation that is reflected over the earth's surface resulting in an average global temperature of 15ºC. This heat absorption is produced by so-called greenhouse gases, mainly carbon dioxide and methane. However, in the last two hundred years, human activities have increased their concentration in the atmosphere.

After carbon dioxide, the most important greenhouse gas is methane. Although methane doesn't match the volume of carbon dioxide emissions, methane traps over 21 times more heat per molecule compared to carbon dioxide CO2 (EPA 2003). Methane originates from several sources including the burning of fossil fuels, agricultural activities, and the decomposition of residues.

Most methane emissions in Peru are due to non-energetic sources such as agriculture (58%), land use changes (21%) and organic residues (14%). Methane is the most important greenhouse gas in Peruvian agriculture (CONAM 2001). Enteric fermentation constitutes 77% of total Peruvian agricultural emissions highlighting the importance of livestock production in total methane emissions and global warming.

Poverty in Peru

The Peruvian National Statistics Institute (INEI 2003) released alarming poverty findings for 2001. Of a total population of 26.6 million people, 54.8% or 14.6 million live in conditions of poverty, up from 48.4% a year earlier. Those living in extreme poverty constitute 24.4% of the population compared with 15% in 2000 (ENAHO 2000 and 2001). The five departments with the highest levels of poverty are Huancavelica (88%), Huánuco (78.9%), Apurímac (78.0%), Puno (78.0%), and Cajamarca (77.4%). The five departments with the highest levels of extreme poverty are Huancavelica (74.4%), Huánuco (61.9%), Cusco (51.3%), Cajamarca (50.8%) and Apurimac (47.4%). The department of Cajamarca ranks fifth in terms of poverty and fourth in terms of extreme poverty.

In Peru, NGOs, international aid agencies and government institutions have been introducing new production technologies in the context of rural development programs aimed at improving the living conditions of farmers. But there are often consequences to these agricultural technologies in the form of externalities.

Despite Cajamarca's high poverty rate, it has a comparative advantage for milk production, which provides an important source of stable income for farmers in the region. Good soil quality for pastures, extension services, the development of infrastructure and stable markets (milk enterprises like NESTLE provide guaranteed markets) have all been key to the growth of the dairy sector. Recent technological changes, notably improvements in pasture and cow breeds, have shown substantial potential to improve the productivity of the dairy sector. However, many studies document the negative environmental effects of milk production due to methane emissions that contribute to global warming (Hegarty 2001; Joblin 1996; Kurihara et al 1997; Ulyatt et al 1997).

La Encañada district

The research zone is in the La Encañada district, located in the Cajamarca River Watershed, in northern Peru. It is a district characterized by high levels of poverty but substantial milk production. This district was also chosen because it is representative of the overall region in terms of agro-ecological characteristics (CODE 1999). It can therefore be used to scale up results to the level of the entire Andean dairy production region of the Department of Cajamarca.

The La Encañada watershed houses roughly 1300 households, predominantly agricultural producers, and has approximately 8000 cows that produce roughly 42,000 liters or 1.86% of total national milk production. This micro watershed covers approximately 12 000 hectares located between 2,800 and 3,700 meters above sea level. Land in La Encañada is classified into three agro-ecological zones: Jalca (Upper Hillsides) (45%), Hillsides (35%) and Valley (including low hillsides) (20%) (Bernet and Tapia 1999).

The importance of irrigation for milk production systems in La Encañada has been discussed by authors such as Bernet and Tapia (1999) and Valdivia (2001). Irrigation allows farmers to shift feeding from native grass, oats and barley to rye grass. Farmers in La Encañada have stated that in the case of irrigation expansion they would prefer to increase rye grass pasture cultivation rather than crops.

A key issue facing La Encañada is increasing agricultural land scarcity due to rural population pressures. Between 1972 and 1994, rural population increased 44% in La Encañada while agricultural land increased just six percent (CODE 1999). Agricultural land used for pastures increased 43% in the same period of time. This land scarcity causes pressure on fallow lands used for grazing. While the need to intensify production is clear, the question remains as how to achieve this in a sustainable manner so that both economic and environmental concerns are taken into consideration.

Objectives of the Study

The purpose of this study was to examine the tradeoffs between poverty alleviation and environmental contamination in the context of dairy production in the Department of Cajamarca. Given the dual goals of reducing poverty and greenhouse gas emissions, there is a need for a policy analysis that systematically documents these two outcomes of dairy production. This information can help support decision makers in the process of formulating policy concerning development strategies for the agricultural sector of the region.


Methodology

Tradeoff Analysis Methodology

This study quantifies the tradeoffs between the economic and environmental sustainability of agriculture. The tradeoff curve represents the joint distribution of economic-environmental outcomes as one moves progressively from low to higher levels of technology adoption. Generally, gains in one area cause losses in the other one, but in some cases  "win-win" outcomes can be achieved (Yanggen et al 2000).

There is a growing recognition that economic activities such as agriculture are having important consequences for the environment. However, in rural areas of countries such as Peru, there is an urgent need to increase farmers' incomes as a means of reducing poverty. Tradeoff analysis makes explicit how a change in an economic indicator affects environmental outcomes and vice versa. This type of analysis allows policy makers to make informed decisions about the development of the agricultural sector.

Dairy Simulation Model

The specific tools used generate the tradeoff relationships between methane emissions and economic returns are a combination of the Dairy Simulation Model and an econometric model. The Dairy Simulation Model v3.2 is part of the LIFE-SIM model group and was developed by the Natural Resources Department of the International Potato Center (CIP) and is available for any user. It was used to evaluate the effects of the different feeding strategies on animal performance. This model is dynamic and its components include specific subroutines for voluntary intake, nutrient requirements, milk production, manure production, methane emissions, thermal regulation, pasture growth, and supplement availability. The model is fully described in Leon-Velarde et al (2005).

Classification

Samples were stratified by cow breeds and agro-ecological zone. There were two cow groups according to breeding: Improved cows and Creole cows. Improved cows include two main breeds: Holstein and Brown Swiss with high milk production potential; the other breeds are negligible, Creole cows have been adapted to local condition for centuries sacrificing milk production in order to get better resistance to the environment.

The other factor considered in the classification was the agro-ecological zone: Hillside and Jalca or Upper Hillsides. They differ from each other because of the altitude, slope, water access and food availability. In data analysis, cows from Hillside were fed mainly with supplements such as oat and barley hay, and natural pasture of low nutritional value; while in Jalca, there were an extensive use of cultivated pastures such as rye-grass-clover.

The total number of categories wer four: improved cows living in Hillside areas, improved cows living in Jalca areas, Creole cows living in Hillside areas and Creole cows living in Jalca areas.

Data Collection

Most of the data were collected by the research project "Tradeoffs in Sustainable Agriculture and the Environment in the Andes: A Decision Support System for Policy Makers" from 1997 to February 2000. The total sample was 36 farmers chosen randomly within the La Encañada watershed (Valdivia 2001). Valdivia evaluated 14 farmers from this group for milk production between 1999 and 2000 in the Hillside and Jalca area. From the total number of cows registered in the survey, there were 2 Creole cows from Hillside areas, 4 Creole cows from the Jalca, 18 Improved breed cows from Hillside areas, and 12 Improved breed cows from the Jalca.

A second survey of the same group of farmers was conducted by Pradel (unpublished data) to complete missing data concerning feed intake and input costs in March 2004. Data from the food intake survey were cross-referenced and combined with Valdivia's data to get more realistic information about livestock food intake in La Encañada. The data were completed with secondary information from Bernet and Tapia (1999). Information about native pasture production, rye grass productivity, milk production potential, and weather in La Encañada was calculated based on Bernet and Tapia's findings.

The model is sensitive to forage and native grass price, and results can be influenced by those prices. We considered a calculated price for the hillside areas of 0.02 Nuevos Soles per kg of rye grass, and 0.012 Nuevos Soles per kg of native grass; and in Jalca, we considered a price of 0.018 Nuevos Soles per kg of rye grass, and 0.012 Nuevos Soles per kg of native grass (Exchange rate at the time of analysis was 3.3 Nuevos Soles per US dollar).


Results

Classification of categories

Irrigation is necessary for cultivated pastures such as rye grass. Because this study analyzes the tradeoffs between methane and gross margins when native grass is progressively replaced by rye grass, it is relevant to present the actual figures on irrigated land in order to understand consumption patterns in La Encañada. Table 1 shows the access to irrigation for each group in La Encañada.

Table 1.  Average total area, irrigation and rye-grass hectares per farm household per category

Cluster

Creole-Hillside

Creole-Jalca

Improved-Hillside

Improved-Jalca

Total Area, ha

7.39

16.51

10.43

16.51

Irrigated Land, ha

1.14

5.40

2.84

5.40

Irrigated Land, %

12%

33%

21%

33%

Rye Grass-Clover, ha

0.99

5.40

1.25

5.40

Source: Valdivia`s survey

Table 1 shows that all groups in La Encañada have access to irrigated land and rye grass-clover mixtures. Groups in Hillside areas have less hectares of rye grass than those in the Jalca. Groups from the Jalca not only consume more rye grass, they also suffer less shortage of pasture in dry season as shown in Table 2. To complement livestock feeding, cows from Hillside areas are fed with barley hay and oats. These supplements are also fed to cows from Jalca, but their use is much more restricted due to their low food quality.

Table 2.  Average daily fresh feed intake per food category by cluster group. Dry matter intake is presented between brackets

Cluster

Creole-Hillside

Creole-Jalca

Improved-Hillside

Improved-Jalca

Fresh Matter: Supplements
(Oat and barley hay), kg/cow/day

6.21
(2.46)

3.67
(1.40)

4.16
 (2.33)

4.01
 (1.66)

Fresh Matter: Rye-Grass – Clover, 
kg/cow/day

13.42
(2.68)

27.58
 (5.52)

26.72
(5.34)

34.08
(6.82)

Fresh Matter: Native Grass,
kg/cow/day

7.14
(1.93)

7.06
 (1.96)

8.63
 (2.30)

7.31
(1.93)

Total feed intake (including mineral salts),
kg/cow/day

26.87
 (7.17)

38.40
 (8.97)

39.61
(10.07)

45.50
(10.50)

Potential Milk Production , kg/cow/day

4.89

4.89

9.83

9.83

Actual Milk Production, kg/cow/day

3.25

4.63

5.00

7.16

Source: Valdivia`s Survey and Bernet and Tapia (1999)

Bernet and Tapia (1999) reported that in hillside areas a lack of irrigation access prevented farmers from growing rye grass, but Valdivia's data and Pradel's subsequent field visits revealed changes in the current situation. There is now a presence of irrigation that permits the cultivation of rye grass-clover mixtures in small areas. The information given by Table 1 is used to understand the actual use of forage and supplements per cluster category. Total feed intake per food group is shown in Table 2.

Table 2 shows the importance of rye-grass consumption for most of the groups with the exception of Creole cows in Hillside areas. Farmers from this group have less irrigated land and, as a consequence, fewer hectares of rye grass. These cows are fed with more barley hay, oats and native grass. These grasses have lower food quality, which affects milk productivity. Even though improved breed cows in Hillside lands have a higher proportion of rye grass intake, they also need to be supplemented with a greater percentage of native grass and barley hay and oats. Creole cows and improved cows in the Jalca are fed with a higher percentage of rye grass and less supplements than hillside cows due to the greater quantity of rye grass hectares farmers cultivate there. The feed intake reflects the milk production pattern per cow. The lowest milk production belongs to the Creole group fed with lower quality feed and the highest milk production is associated with improved breed cows fed with higher quality feed.

Analysis of the Current Situation

The Dairy Simulation Model generates economic and environmental information concerning the dairy production system in terms of gross margins and methane emissions. Milk production varies by breed more than by agro-ecological zone. The agricultural census of 1994 (INEI web page 1994) shows that in la Encañada there were 2053 Holstein, 1045 Brown Swiss and 17167 Creole cows. Creole cows resist better harsh weather conditions compared to improved breeds. But in La Encañada, improved breed cows show relatively good performance and they can be recommended when farmers have enough capital to afford Holstein or Brown Swiss cows.

The Dairy Simulation Model calculates gross margin from milk production. Gross margins are calculated by subtracting variable costs from the gross revenues of milk sales. It is important to consider that not all revenues and all costs are included in gross margin calculation. Revenues not included are male calf sales, meat sales, manure sales, and other outputs. On the other hand, the costs included in the model are all variable costs: food costs, labor costs (based on cattle unit (AU) in man equivalent labor (MEL), which is defined as MEL = 0.432 + 0.024 AU (Leon-Velarde 1991), and the other variable costs. Fixed costs such as depreciation, rent, mortgage payments, taxes, are not included in the calculation of costs.

An important factor that affects economic returns is distribution of food availability throughout the year. Hillside areas have less food availability, especially in the dry season, leading farmers to supplement with barley hay which is costly and low in nutritional quality. Figure 1 shows the gross margin for each category

Figure 1. Average gross margins per agro-ecological zone and breed in La Encañada
Source: Own calculations using Dairy model

Gross margin per cow and per liter of milk are low for Creole cows. The result is explained by the information in Table 1. Creole cows from hillside areas don't get enough food to reach their milk production potential and most of their food consists of native pasture, barley hay and oats with less rye-grass - clover mixture than the other categories. Creole cows from the Jalca have good food quality and quantity, but they are limited by their own genetic potential. Table 1 indicates that their actual milk production is close to its potential limit.

Improved breed cows, on the other hand, generate more than double the income of Creole cows because of their higher milk production. In contrast with Creole cows on hillside areas, improved breed cows in hillside areas receive a greater percentage of rye grass in their diet. This fact, along with greater genetic potential, explains their better performance, but nevertheless there is still some scarcity of food in the dry season. Improved cows in the Jalca have the highest income due to a more constant supply of food throughout the year as well as a higher genetic potential for milk production.

Figure 2 estimates the current levels of methane production for each category using conversion coefficients from Ulyatt et al 1997):

Figure 2.  Average current methane emissions per agro-ecological zone and breed in La Encañada.
Source: Own calculations using Dairy model

Methane is a by-product of milk production. It is commonly believed that higher pasture quality will lead to a reduction in methane emissions. That is certainly true when consumption is fixed. However, pasture quality and feed intake are innately linked in a positive manner. When pasture digestibility increases, consumption also increases. There are feed-factors (e.g.. bulk and resistance to digestion), which limit intake as well as animal factors (e.g. weight and maturity), and factors that involve both (e.g., palatability and anti-nutritive factors). When more milk is produced, more methane will be emitted through fermentation. The adoption of purportedly methane-reducing feeding practices may in fact increase rather than decrease the total annual emissions from a farm due to an overall increase in the productivity of the dairy system.

The results indicate that methane emissions per cow are higher for improved breed cows compared to Creole cows. However, emissions are lower when calculated per liter of milk produced by improved breed livestock. Methane emissions for Creole cows are about 72 kg per cow per year while improved cows have around 20% more total emissions. Conversely, Creole cows produce 25% more methane per liter of milk than improved cows, but their lower production translates to lower total methane emissions.

Trade Off Analysis

Tradeoff analysis involves the estimation of the relationship between economic and environmental impacts under a series of different scenarios. These scenarios are run for the four categories of production systems based on breed and agro-ecosystem.

Current feed strategies for each cluster category were used to set up the scenarios. Those scenarios were constructed based on the response surface methodology to select the optimum range of technological alternatives of feed intake scenarios (Leon-Velarde and Quiroz 1999; Myers et al 1989). Three out of the four feed categories (rye-grass, oat and barley hay) were correlated to the different scenarios (Appendix 1). The fourth feed category, native pasture, changed indirectly, because native grass is consumed when there is no availability of other pasture or supplements at any specific time of year. The Dairy model estimated the effects of consuming different proportions of pastures and supplements in cows' diets (21 combinations per group) on gross margins and methane emissions.

After the dependent and independent variables were obtained, a production function was used to get the coefficients for the next part of the study which was to analyze the effects of shifting from native grass to rye grass. Farmers use supplements because there is no more cultivated grass to feed cows. The assumption is that if there is enough cultivated grass, farmers will avoid to provide supplements (oat and/or barley) to cows and will feed them exclusively with grass. Therefore the supplement was subsequently cancelled from the equation to see the effect from feeding animals with native grass to progressively feed them with rye grass.

The chosen functional form is as follows:

y =  b0 + S bi* Xi

Where:     y = outputs (Gross Margin and Methane),
                X = inputs (ryegrass, supplements and native grasses) and
               
b are the parameters to be estimated.

This function permits us to assign different percentages of rye grass and native grasses in the diet in order to see the effect of these changes on methane emissions and gross margins and their joint distribution. Feed consumption was shifted from pure native grass intake to pure rye grass intake, including combinations of both pastures between these extremes. Figure 3 shows the results of the percentage change of rye grass and native grass in the diet.

Figure 3. Tradeoff relationships between gross margins and methane emissions
changing the proportion of rye grass and native grass in La Encañada

The tradeoff estimations between gross margins and methane emissions show that Creole cows in both Hillsides and the Jalca range from 20.9 to 92.1 US dollars in economic returns and from 51 to 78 kg of methane per year. On the other hand, improved cows range from 50.6 to 191 US dollars and from 52 to 90 kg of methane per year. The increase in economic returns for Creole cows are notably limited in comparison to improved breeds.

This narrower range of results for the Creole cows is related to their maximum potential milk production. Creole cows have just half of improved cows' potential, so income and methane emission are limited by this constraint. Improved breed livestock can potentially generate double the income per cow compared to Creole cows, but they can also produce 25% more methane.

The impact can also be appreciated on a per hectare basis. This is of critical importance because increasing land scarcity will force farmers into a downward spiral of poverty and land degradation if they are unable to intensify production sustainably by producing more on the limited remaining available land. When farmers have no access to irrigated land with rye grass, they have to graze animals on native grasses with an average stocking rates on hillsides of 0.4 cows per hectare (Tapia 1996). Irrigated rye grass, on the other hand, has a stocking rate of 1.0 cows/ha. When stocking rates are taken into account and cows are fed exclusively with native grass, economic returns are just 18.9 US dollars per hectare per year in hillside areas and 25.4 US dollars per hectare per year in the Jalca. Methane emissions are around 23 kg per hectare in both cases (Figure 4). The intensification of the production to exclusively rye grass increases gross margin per hectare by 830% in Hillsides and 660% in the Jalca. These higher rates of gross margin per hectare are the combination of the increased stocking rate and increased efficiency associated with irrigated rye grass. Given the accelerating demographic pressures combined with land scarcity, this type of intensification of land use is all but inevitable if farmers are to remain in the agricultural sector. Methane emissions also increase by 360% per hectare under this scenario. However, it bears reiteration that to achieve the same economic results for reducing poverty using the currently dominant technologies much greater levels of methane emissions would result not to mention the need to clear substantially more ecologically fragile land for grazing.

Figure 4.   Economic returns and methane emissions/ha: native vs. rye grass.
Scaling up the analysis to the Cajamarca dairy watershed

La Encañada can be considered representative of the entire Cajamarca dairy region due to its similar resource endowments and the importance of milk production as a means for subsistence. Most farmers from other provinces in Cajamarca, however, haven't benefited from the same level of technical assistance as La Encañada. La Encañada has higher levels of improved breeds and improved pasture. Market forces along with population pressure and land scarcity are likely to lead to an intensification of production in the rest of the region similar to the processes under way in La Encañada. The dairy processing companies operating in the region, Nestle and Gloria, are increasing the demand for milk and are planning to expand their processing capacity (Indacochea et al 1998).

As we have seen, changes in both the livestock production systems related to feeding strategies and breeds can improve not only the economic welfare of farm households but also increase the impact of milk production on global warming. While the influence of La Encañada itself on global warming may not be significant, when the analysis is scaled up to the level of the Andean dairy region of Cajamarca, the numbers increase substantially in magnitude.

This section estimates the tradeoffs between economic returns and methane emissions for the entire Cajamarca Andean dairy watershed based on the results of the tradeoff study of milk production in La Encañada. The scaling up consists of summing up the impact of the total number of cows (Creole and improved breeds) in both the Hillside and Jalca zones to evaluate changes in income and methane emissions.

Characteristics of La Encañada and the Cajamarca dairy watershed

La Encañada, with its roughly 12,000 hectares of arable land and 40 000 hectares of native grass, is located in the Cajamarca River watershed. The Highlands or Sierra of Cajamarca have approximately 1,200,000 hectares of total land or 80% of all Cajamarca, 420,000 hectares of arable land and 540,000 hectares of pastures land (of which 480,000 are native).

The amount of natural pasture rangeland per improved cow in la Encañada is very similar to the amounts in the Hillsides and Jalca in the Dairy Watershed. The main difference between La Encañada the other two areas is availability of improved pasture. La Encañada has 223% more rye-grass per cow compared to the Dairy Watershed (see Figure 5). Given the large amount of natural rangeland, there is an opportunity to substantially expand the amount of irrigated improved pastures in these areas as a strategy for increasing production and alleviating poverty.



Figure 5.  Average number of hectares of ryegrass and native grass per cow in La Encañada,
the Cajamarca River watershed and the Dairy watershed of Cajamarca

Why scale up results to the Cajamarca dairy watershed?

La Encañada, with its proximity to the city of Cajamarca and higher levels of technical assistance due to its selection as a pilot intervention site of various institutions, appears to represent the future development path for the rest of the Cajamarca dairy watershed. With adequate support for the adoption of improved cows and pasture technologies, farmers in other areas of the dairy watershed will be able improve their living standards and poverty will decrease in the region. The national market has a deficit with respect to domestic milk supply. Of the 2 million liters consumed annually, around 800 thousand liters are produced in Peru. The difference is covered by imports. Cajamarca produces 17% (208 thousand liters) of commercialized national milk production.

The demand for milk products is likely to increase because, at 50 liters of annual per capita milk consumption, Peru is substantially below the consumption level of 115 liters of milk recommended by the FAO for developing countries. This situation constitutes an opportunity for dairy farmers to increase their milk production thereby increasing incomes and reducing poverty. In fact, Cajamarca has been experiencing a consistent increase in milk production with an annual average growth of 11.4% (Figure 6).

Figure 6.  Cajamarca: Trends in milk production, Thousand of tonnes

But, again, it is important to estimate the environmental costs of increased production in order to make informed policy decisions concerning development strategies for the dairy sector.

Thirty percent of the land in Cajamarca has potential for pasture crops such as alfalfa, rye grass and sorghum (Indacochea et al 1998). To date, livestock activities are based overwhelmingly on grazing natural pastures and few efforts have been made to improve pastures. The proportion of cultivated pasture is extremely small in relation to the natural grassland.

Method of aggregation

Information regarding the number of irrigated and rain-fed hectares and the number of cows (classified as improved breed or Creole) in the Dairy watershed was obtained from the Third National Agricultural Census of 1994. According to Tapia (1996), Cajamarca land area is 25% valley, 45% hillsides and 30% Jalca. This distribution is used for the scaling up analysis that follows. Excluding valley areas (because the cows from this agro-ecological zone are already using pastures for dairy cow feeding) the total area is 60% Hillside areas and 40% Jalca. The distribution of cows is 60% in Hillside areas and 40% in the Jalca.

The total area includes the sum of the land area of the six Andean dairy watershed provinces: Cajabamba, Cajamarca, Celendín, Chota, Cutervo, Hualgayoc, San Marcos, San Miguel, and San Pablo. These contain approximately 353,000 ha of total land, 432,000 ha of rangeland, 8,300 ha of rye grass, and 446,277 cattle, of which 46.5% are cows and heifers (INEI, III Agricultural Census). The number of people dedicated to milk production in these areas is 97,967 (Table 3).

Table 3.  Distribution of cow holdings in the dairy watershed of Cajamarca

Highlands of Cajamarca

Farmers with cattle

Total Cattle

Cattle/ Farmer

Farmers with improved cows

Improved cow and heifers

Cow and heifers/ Farmer

Total Highlands

97967

446277

4.56

11611

57699

4.97

 1-4 Heads

67819

161512

2.38

7845

13789

1.76

 5- 9 Heads

21833

138059

6.32

2326

12720

5.47

10-19 Heads

6619

82483

12.46

1015

11054

10.89

20-49 Heads

1510

39864

26.4

357

8529

23.89

> 50 Heads

166

24359

146.74

67

11607

173.24

Source: INEI, III agricultural census data

The analysis takes in account the technological improvements of rye grass pasture as well as breed improvement as the percentage of Creole cows is very high (87%). 47% of cattle in Cajamarca are cows. The estimates obtained from the Dairy model per category of cow are summed up according to the distribution of cows per breed and agro-ecological zone (excluding Valley)

Results of the aggregation

Data from the 1994 national census for the Dairy watershed of Cajamarca indicate that the total number of dairy cows is approximately 207,500 including both improved and Creole breeds. From this total, approximately 68,000 cows are located in Valley and the rest are located in Hillsides area and the Jalca (139,500 cows). Cows from Valley areas were excluded from the analysis because they are generally already using improved pasture. Cows from the other two zones are fed mainly with natural pasture and have a higher percentage of Creole cows. Development programs should therefore focus on these areas where there is the greatest potential for growth.

The tradeoffs between economic returns and methane emissions is initially calculated using the actual percentages of Creole and Improved cows (87% and 13%, respectively) and the amount of land in each agro-ecological zone to get an aggregate estimate for the region (Figure 7).

Figure 7.  Tradeoff relationships between gross margin and methane emissions
by increasing rye grass feed proportion (race is breed)

A shift from native grassland to rye-grass without changing the current proportion of improved cows and Creole cows can have a high effect on economic returns (see Figure 7). Gross margins go from 4.01 million dollars per year from milk produced on native pasture to 12.99 million dollars ith the full use of rye-grass. The effect on methane emissions is also important, going from 7.27 to 11.00 million kg.

However, if we tried to reach the same income level with diets of 100% native pastures, it would be necessary to increment the number of cows by 324% (to 451,980 cows) and emissions would roughly triple from their base level of 7.27. So while intensification with rye-grass does increase emissions, achieving the same livelihood impact for farmers without this technological change would be much more damaging to the environment.

Besides land intensification we can also estimate the impact of improved cow breeds. With 100% adoption of improved breeds, the extra benefits to the region would be much higher, increasing from approximately 8.3 to 24.6 million dollars. This increase is almost double the income of the scenario where the current breed mix is fed with rye-grass. In order to generate the same level of income (24.6 million) using the current mix of breeds and natural pasture, it would be necessary to increase the number of cows to 857,130 a 614% increase in the total population.

To calculate the impact on farmer income, we note that the extra income generated per cow using quality forage (from 20% of rye-grass on diet to 80%) is $US39. Table 3 shows the distribution of cows that can be used as an indicator of the distribution of the additional economic returns to farmers. Farmers with more than 20 cows are less than 0.01% of the total farmers, therefore they are not significant and they probably have already installed improved pastures and raise improved cows.

From Table 3, we can predict that income increases would be concentrated on the farmers with households between 1 and 9 cows. They make up 92% of total farmers in the Dairy watershed of Cajamarca and due to their numbers and poverty they should logically be the target for development initiatives. The increment of US$ 39 per cow will increase the income of farmers with 1 to 4 cows on average by US$ 93. For farmers with between 5 and 9 cows, the average increase is US$ 246.5.

When the effect of intensification is included in the analysis, the impact of shifting from a low quality to a high quality pasture is more dramatic. One cow needs 2.5 hectares of native pasture rangeland to feed itself, whereas one hectare of rye grass is enough to feed one cow (Tapia 1996). This implies that, by changing the 2.5 hectares of rangeland to rye-grass through irrigation, farmers can increase the number of cows by a factor of 2.5. This further implies that, for the category of farmers with 1 to 4 cows, the average number of cows could increase to 5.95/farmer leading to an increase in income of US$ 232. Also, if we add the effect of cow improvement (they have 87% of Creole cows and 1 Creole produces half of what an improved cow produces), income can increase by US$ 411 for this group.

If we wish to generate the same income as improved pastures and cow breed, using only Creole cows and native pasture, it is not only necessary to increase the both the number of Creole cows and the total hectares of pasture to feed those animals. For example, to feed 139,500 cows with rye-grass (and the current mix of breeds), it is necessary to have 139,500 ha of this pasture (one hectare of rye grass can feed one cow). However, in order to generate the same economic return (with the current proportion of Creole and Improved cows) it would be necessary to use 1,129,950 hectares of native pastureland to feed 451 980 cows given the stocking rate of one cow/2.5 ha. Moreover, if we wish to reach the same income generated with 139,500 cows, all of improved breed and consuming rye grass, it would be necessary to expand grazing to 2,142 825 hectares of natural pasture. This magnitude of increased pressure on available land would certainly lead to substantial overgrazing, deforestation, erosion and loss of biodiversity.

The effect on methane emissions is not as important as the increase in farmer income. The former increase from 8,019 tonnes to 10,288 tonnes of methane when rye-grass intake increases from 20% to 80%. To give a perspective on the magnitude of these emissions, an average car produces about 226 g (0.226 kg) of carbon dioxide (CO2) per kilometer and an average car accumulates 19,800 kilometers/yr (CACAQ 2002). The result is a production of 4480 kg of CO2 per year. By transforming the amount of CH4 emission into CO2 equivalent (21 times), current emissions from cows in Cajamarca are equal to those of 37,600 cars. Feeding cows with 75% of rye-grass, the amount of contamination is comparable to 48,225 cars or approximately 10,500 additional cars. In other words, cows produce more greenhouse gas emissions than cars in Cajamarca as Cajamarca has around 20,000 vehicles according to the IX National Census.

Methane emissions have been analyzed up to this point only considering technological change. In order to achieve the same increase in income due to the adoption of 100% rye-grass, but limiting the development of the dairy sector to the use of natural pastures, would result in the generation of 213% more methane emissions. Similarly, in order to achieve the same level of income produced with 100% rye-grass and 100% improved breed cows, but limiting the technologies to native pastures and the current mix of livestock breeds, would result in an increase of 406% in methane emissions. In other words, it would be like increasing the number of cars from 51,600 to 110,390 cars and 209,260, respectively.


Conclusions


References

Bernet T, Herve D, Lehmann B and Walker T 2002. Improving land use of slope farmers in the Andes: An economic assessment of establishing small-scale sprinkler irrigation to shift toward milk production. Mountain Research and Development (USA). ISSN 0276-4741. 2002. 22(4): 375-382.

Bernet T and Tapia M 1999 Análisis de los Sistemas de Producción en la Microcuenca de la Encañada: Documento Base para Investigaciones y Acciones Futuras en la Sierra Norte del Perú. Departamento de Ciencias Sociales, Documento de Trabajo Number. 1999-1. Centro Internacional de la Papa (CIP), Lima, Perú.

CACAQ (Citizens' Advisory Committee on Air Quality) 2002 Transportation component. http://pirg.uwaterloo.ca/cacaq/g9lp/pdfs/transportation.pdf

CODE (Comité de Desarrollo de Encañada) 1999 Plan de Desarrollo de la Encañada 1999- 2008. Intermediate Technology Development Group, ITDG-Perú, Lima, Peru

CONAM (Consejo Nacional del Ambiente) 2001 Comunicación Nacional del Perú a la Convención de las Naciones Unidas sobre Cambio Climático. Primera Comunicación. Consejo Nacional del Ambiente, Lima, Perú. 118 p.

ENAHO (Encuesta Nacional de Hogares) 2000 Condiciones de vida y Pobreza 2000. Web Page INEI: http://www.inei.gob.pe 

ENAHO (Encuesta Nacional de Hogares) 2001 Condiciones de vida y pobreza 2001. Web Page INEI: http://www.inei.gob.pe

EPA (Environmental Protection Agency) 2003 Global Warming-Emissions. U.S. Environmental Protection Agency. http://yosemite.epa.gov/OAR/globalwarming.nsf/content/Emissions.html

Hegarty R 2001 Overview of sources of livestock emissions. In: The Australian Greenhouse Office. Chapter 1. http://www.greenhouse.gov.au/agriculture/livestock/index.html#chapter1

Indacochea. A, Abolió B, Bedoya L, Carrillo J, Negrón G, Sánchez L, Santillana M 1998 Chapter 10: Sector Pecuario, In: Cajamarca competitiva. Saywa ediciones s.r.l., Lima, Peru

INEI (Instituto Nacional de Estadística e Informática) 1994 III Agricultural Census. http://www.inei.gob.pe

INEI (Instituto Nacional de Estadística e Informática) 2003 Condiciones de vida en los departamentos del Perú - ENAHO 2001- IV "La Pobreza en el Perú 2002-2003 - ENAHO IV Trimestre"

Joblin K N 1996 Options for reducing methane emissions from ruminants in New Zealand and Australia. In Coping with Climate Change. (Editors W J Bouma, G I Pearman, M R Manning) pp 437-449. CSIRO Publishing, Collingwood, Australia.

Kurihara M, Shibata M, Nishida T, Purnomoadi A, Terada F 1997 Methane production and its dietary manipulation in ruminants. In Ruminal Microbes and Digestive Physiology in Ruminants (Editors R. Onodera et al) pp 199-208. Japan Scientific Societies Press, Tokyo.

Leon-Velarde C U 1991 A simulation model to analyze the bio-economic function of cows in intensive dairy farms using a systems approach. Thesis Ph.D. University of Guelph. p.256.

Leon-Velarde C U and Quiroz R 1999 Selecting optimum ranges of technological alternatives by using response surface designs in system analysis. In: Impact on a Changing World, CIP Program Report 1997-1998. International Potato Center, Lima, Peru. P. 387-394.

Leon-Velarde CU, Quiroz R, Cañas R, Osorio J, Guerrero J 2005 LIFE - SIM: Livestock feeding strategies Simulation models. Natural Resources Division Working Paper Number 7. International Potato Center (CIP), Lima Peru.

Myers R H, Khuri A I, Carter Jr W A 1989 Response surface methodology: 1966-1988. American Statistical Association and the American Society for Quality Control. Technometrics, May 1989, Volume 31, Number 2. pp 137-157.

Tapia M 1996 Slope conditions in the North Sub region of the Highlands of Peru. In Ecodesarrollo en los Andes Altos. Fundación Friedrich Ebert, Lima, Peru.

Ulyatt M J, Lassey K R, Martin R J, Walker C F and Shelton I D 1997 Methane emission from grazing sheep and cattle. Proceedings of the New Zealand Society of Animal Production, 57, 130-133.

Valdivia R O 2001 Farm data collection in La Encañada, Cajamarca, Peru to support analysis of tradeoffs in sustainable agriculture. Montana State University, U.S. 38p.

Yanggen D, Antle J, Crissman C, Stoorvogel J, Bowen W and Barrera V 2000 Trade off analysis: support for policy making. International Potato Center, Lima, Peru.





Appendix 1.  Proportion of fresh matter feed intake, milk production, methane emission, total variable costs, and gross margin per scenario and per cluster category.

 

Cluster

Salt

Ryegrass

Oat hay

Barley hay

Native grass

Milk Production, kg milk/year

Cost, US$/year

Benefit, US$/year)

Methane, l/year

Scen1

Creole Jalca

1.1%

39.6%

3.9%

5.8%

49.5%

1041

355

520

89

Scen2

Creole Jalca

0.8%

90.2%

1.6%

1.2%

6.2%

1474

504

737

112

Scen3

Creole Jalca

1.0%

38.4%

3.7%

16.2%

40.7%

1085

357

542

93

Scen4

Creole Jalca

0.8%

89.7%

1.7%

1.4%

6.4%

1466

527

733

111

Scen5

Creole Jalca

1.1%

38.8%

10.9%

5.4%

43.8%

1158

404

579

91

Scen6

Creole Jalca

0.8%

89.2%

3.3%

0.3%

6.4%

1473

571

736

108

Scen7

Creole Jalca

1.0%

37.9%

11.3%

11.4%

38.4%

1155

408

577

96

Scen8

Creole Jalca

0.8%

89.7%

3.2%

0.2%

6.1%

1467

592

733

101

Scen9

Creole Jalca

1.1%

13.4%

8.2%

13.0%

64.2%

909

346

454

86

Scen10

Creole Jalca

0.7%

96.7%

0.5%

0.0%

2.1%

1481

648

740

105

Scen11

Creole Jalca

0.9%

71.4%

5.6%

1.3%

20.8%

1311

459

655

100

Scen12

Creole Jalca

0.9%

69.1%

5.3%

10.9%

13.9%

1308

454

654

106

Scen13

Creole Jalca

0.9%

71.6%

1.1%

10.5%

15.9%

1278

395

639

100

Scen14

Creole Jalca

0.7%

51.4%

31.9%

5.3%

10.7%

1310

485

655

103

Scen15a

Creole Jalca

0.9%

69.8%

5.6%

8.6%

15.2%

1300

437

650

102

Scen15b

Creole Jalca

0.9%

70.0%

5.3%

9.3%

14.5%

1309

438

654

102

Scen15c

Creole Jalca

0.9%

69.7%

5.4%

8.4%

15.6%

1304

439

652

103

Scen15d

Creole Jalca

0.9%

69.0%

5.4%

9.4%

15.3%

1318

440

659

103

Scen15e

Creole Jalca

0.9%

69.5%

5.4%

8.7%

15.5%

1314

439

657

103

Scen15f

Creole Jalca

0.9%

68.7%

5.5%

9.5%

15.4%

1306

440

653

103

Scen16

Creole Jalca

2.2%

0.0%

25.6%

34.6%

37.6%

933

356

466

88

Scen1

Creole hillside

1.3%

17.6%

6.6%

14.4%

60.2%

994

385

497

87

Scen2

Creole hillside

1.1%

52.2%

5.9%

13.0%

27.8%

1291

429

645

97

Scen3

Creole hillside

1.2%

17.9%

5.8%

35.4%

39.8%

960

373

480

91

Scen4

Creole hillside

1.0%

50.4%

5.3%

28.4%

14.9%

1281

430

640

103

Scen5

Creole hillside

1.2%

18.6%

20.6%

12.0%

47.7%

1094

447

547

90

Scen6

Creole hillside

1.0%

49.2%

20.4%

11.8%

17.7%

1312

504

656

100

Scen7

Creole hillside

1.1%

17.7%

19.9%

28.9%

32.3%

1110

454

555

97

Scen8

Creole hillside

1.0%

47.5%

19.6%

22.5%

9.4%

1312

508

656

107

Scen9

Creole hillside

1.3%

5.1%

14.0%

25.9%

53.8%

906

394

453

87

Scen10

Creole hillside

1.0%

62.0%

11.1%

15.6%

10.2%

1334

484

667

104

Scen11

Creole hillside

1.2%

33.3%

13.2%

4.5%

47.7%

1205

449

602

90

Scen12

Creole hillside

1.1%

29.9%

11.7%

37.6%

19.8%

1232

441

616

102

Scen13

Creole hillside

1.2%

31.4%

2.2%

26.2%

39.1%

1145

380

572

93

Scen14

Creole hillside

1.1%

28.8%

25.5%

24.3%

20.3%

1237

487

618

98

Scen15a

Creole hillside

1.1%

31.0%

11.9%

25.5%

30.6%

1216

435

608

95

Scen15b

Creole hillside

1.1%

31.2%

12.0%

25.8%

29.9%

1209

432

604

94

Scen15c

Creole hillside

1.1%

31.5%

12.1%

26.0%

29.3%

1195

430

597

94

Scen15d

Creole hillside

1.1%

30.8%

12.2%

26.3%

29.6%

1209

432

604

94

Scen15e

Creole hillside

1.1%

30.7%

12.0%

25.7%

30.5%

1224

436

612

95

Scen15f

Creole hillside

1.1%

31.2%

12.1%

26.0%

29.7%

1201

434

600

95

Scen16

Creole hillside

1.2%

0.0%

22.2%

37.3%

39.3%

1019

418

509

92

Scen1

Improved cows Jalca

0.9%

37.8%

3.9%

6.8%

50.6%

1560

461

780

109

Scen2

Improved cows Jalca

0.7%

95.3%

1.7%

0.8%

1.6%

2573

635

1286

124

Scen3

Improved cows Jalca

0.9%

36.1%

3.6%

18.7%

40.7%

1567

460

783

113

Scen4

Improved cows Jalca

0.7%

95.1%

1.8%

1.2%

1.2%

2580

669

1290

117

Scen5

Improved cows Jalca

0.9%

37.5%

11.3%

6.8%

43.6%

1627

502

813

110

Scen6

Improved cows Jalca

0.7%

95.1%

2.7%

0.0%

1.5%

2629

683

1314

119

Scen7

Improved cows Jalca

0.8%

35.7%

10.3%

17.4%

35.8%

1645

500

822

114

Scen8

Improved cows Jalca

0.7%

95.6%

2.3%

0.0%

1.4%

2642

738

1321

103

Scen9

Improved cows Jalca

0.9%

12.5%

7.3%

13.3%

66.0%

1409

441

704

106

Scen10

Improved cows Jalca

0.6%

99.1%

0.0%

0.0%

0.2%

2954

832

1477

97

Scen11

Improved cows Jalca

0.8%

73.8%

6.4%

1.8%

17.2%

2015

533

1007

118

Scen12

Improved cows Jalca

0.7%

69.9%

6.4%

10.8%

12.2%

2076

577

1038

126

Scen13

Improved cows Jalca

0.8%

72.8%

0.9%

10.9%

14.6%

1995

509

997

120

Scen14

Improved cows Jalca

0.7%

69.9%

11.2%

6.5%

11.7%

2072

595

1036

124

Scen15a

Improved cows Jalca

0.8%

71.4%

6.2%

8.9%

12.9%

2073

552

1036

123

Scen15b

Improved cows Jalca

0.8%

71.4%

6.5%

9.4%

12.0%

2064

551

1032

123

Scen15c

Improved cows Jalca

0.7%

70.9%

6.3%

9.4%

12.6%

2071

551

1035

123

Scen15d

Improved cows Jalca

0.8%

70.8%

6.3%

9.0%

13.2%

2067

552

1033

123

Scen15e

Improved cows Jalca

0.8%

71.7%

6.4%

8.7%

12.4%

2065

550

1032

122

Scen15f

Improved cows Jalca

0.7%

71.4%

6.3%

9.0%

12.6%

2061

550

1030

123

Scen16

Improved cows Jalca

0.9%

0.0%

11.4%

20.7%

67.0%

1332

448

666

105

Scen1

Improved cows hillside

0.9%

27.7%

1.9%

13.3%

56.1%

1432

467

716

107

Scen2

Improved cows hillside

0.7%

83.7%

1.6%

9.7%

4.3%

2118

568

1059

123

Scen3

Improved cows hillside

0.9%

18.6%

1.9%

40.0%

38.6%

1376

446

688

112

Scen4

Improved cows hillside

0.7%

80.7%

1.5%

14.0%

3.0%

2156

628

1078

130

Scen5

Improved cows hillside

0.9%

27.8%

5.8%

13.7%

51.8%

1465

497

732

108

Scen6

Improved cows hillside

0.7%

82.9%

3.9%

8.4%

4.1%

2202

616

1101

124

Scen7

Improved cows hillside

0.8%

25.5%

5.1%

36.4%

32.2%

1452

488

726

114

Scen8

Improved cows hillside

0.7%

80.4%

4.0%

13.1%

1.8%

2232

664

1116

130

Scen9

Improved cows hillside

0.9%

8.1%

3.8%

27.4%

59.8%

1387

441

693

107

Scen10

Improved cows hillside

0.7%

95.6%

0.4%

3.2%

0.0%

2528

707

1264

125

Scen11

Improved cows hillside

0.7%

48.2%

2.7%

19.5%

28.9%

1879

512

939

114

Scen12

Improved cows hillside

0.8%

59.4%

3.6%

25.6%

10.5%

1943

550

971

129

Scen13

Improved cows hillside

0.8%

54.9%

0.7%

23.5%

20.1%

1817

503

908

119

Scen14

Improved cows hillside

0.8%

53.9%

5.9%

23.1%

16.3%

1898

554

949

120

Scen15a

Improved cows hillside

0.8%

54.6%

3.3%

23.4%

17.9%

1871

531

935

120

Scen15b

Improved cows hillside

0.8%

54.7%

3.3%

23.7%

17.5%

1828

526

914

119

Scen15c

Improved cows hillside

0.8%

53.6%

3.3%

24.0%

18.2%

1840

527

920

119

Scen15d

Improved cows hillside

0.8%

54.1%

3.2%

23.1%

18.8%

1866

531

933

120

Scen15e

Improved cows hillside

0.8%

55.4%

3.3%

23.8%

16.7%

1856

527

928

119

Scen15f

Improved cows hillside

0.8%

54.6%

3.3%

24.0%

17.2%

1833

525

916

118

Scen16

Improved cows hillside

0.9%

0.0%

5.5%

39.8%

53.8%

1275

436

637

108


Received 30 November 2005; Accepted 16 January 2006; Published 10 March 2006

Go to top