Relative economic value estimates of grasscutter production traits

Department of Animal Science Education, College of Agriculture Education, University of Education, Winneba, P.O. Box 40, Mampong-Ashanti, Ghana. Department of Animal Science, College of Agriculture and Consumer Sciences, University of Ghana, P.O. Box LG 571, Legon, Accra, Ghana. Animal Research Institute, Council for Scientific and Industrial Research, P.O. Box AH 20, Achimota, Accra, Ghana. Department of Animal Science, College of Agriculture and Life Sciences, North Carolina State University, Campus Box 7621/232B Polk Hall, Raleigh NC 27695-7621, North Carolina, USA


INTRODUCTION
The idea of domesticating wild animal (game) species for meat production to improve dietary protein supply in Africa is not new (FAO, 1990;Mensah, 2000;Achana, 2002;Hanotte and Mensah, 2002).Domestication of wild species has been particularly popular in the West Africa sub-region where bush meat is an important dietary item (Asibey, 1966).For many rural people in the West African sub-region, game meat (known as "bushmeat") is a highly valued forest product (FAO, 1990).Game is important source of meat in both rural and urban household diets.Conservationists and advocates of wild animal domestication (Achana, 2002) have argued for the farming of favourite species to increase bushmeat production and supply in the West African sub-region, and to reduce pressure on wild populations.Wild animal species advocated for farming include the grasscutter (Thryonomys swinderianus), giant rat (Cricetomys gambianus) and African giant snail (Achatina achatina) (NRC, 1991).In West Africa where grass provides its main habitat and food, the animal is commonly known as the "grasscutter" or "cutting grass" while in other parts of Africa, particularly Southern Africa, where it is closely associated with cane fields, it is known as "cane rat" (NRC, 1991).
Grasscutters are widespread and abundant across West Africa, and are a common source of food among rural populations.Ghanaians are farming the grasscutter for food.In 2009 the population of domesticated grasscutter was 17,400 (Annor et al., 2009).Ghana is in the process of establishing an Open Nucleus Breeding Scheme for genetic improvement of captive grasscutters (MoFA, 2004).There is a well-established series of logical steps to follow in designing animal breeding improvement programmes (Harris et al., 1984;Harris and Newman, 1994).The first step is to identify planned production, processing and marketing system (s).Using this information, economic merit for individual traits can be defined and subsequently breeding objectives for the species concerned (Morris et al., 1978).Thus, traits which should be improved have to be identified and their relative economic importance established.
Economic value of a trait is defined as the income per animal per unit of improvement in genetic merit for that trait, at constant levels of genetic merit for other traits in the breeding goal (Groen and Van Arendonk, 1996).It expresses the extent to which economic efficiency of production is improved for a given unit of genetic superiority of a trait.It is useful to know economic values of various traits relative to each other; hence the term relative economic value (REV) is commonly used (Blair, 1989).Relative Economic Values are needed to define breeding objectives whereby greater importance may be given to traits of higher economic merits (Morris et al., 1978).Economic values have been calculated for cattle (Annor et al., 2000) and sheep (Annor et al., 2007) in Ghana.However, there is lack of information on REVs of grasscutter production traits in the literature.
The objective of this work was to estimate relative economic values of survival, body weight, growth rate, reproduction, docility and food intake of grasscutter (Thryonomys swinderianus) production as a first step towards defining breeding objective for the grasscutter industry in Ghana.

METHODOLOGY
Data were obtained from records of grasscutters kept at the grasscutter section of the Department of Animal Science Education, University of Education, Winneba, Ghana, from 2006-2010.The derivation of REVs followed the following five steps: 1. Specification of breeding, production and marketing systems 2. Identification of sources of income and expenses 3. Determination of biological traits influencing income and expenses 4. Calculation of economic value (EV) of each trait 5. Calculation of REVs of traits

Specification of breeding, production and marketing system
Computer models were developed for the life cycle production of a family of grasscutter and the growth performance of their offspring.The family consisted of 4 breeding does and 1 breeding buck.Models were developed in Microsoft Excel spreadsheet of Windows 2007.Model calculations were based on complete life cycle of the breeding doe, which lasted for 60 months (Figure 1).Input and output parameters (Table 1) used in models were estimated from records of 502 kids born between 2006 and 2010.Model output consisted of profit derived from sale of mature bucks, does, and culled does (Figure 1).
Five classes of animals were considered in the model flow diagram: neonates (pre-weaners), young does, young bucks, breeding does and breeding bucks.Surplus does and all bucks were sold for consumption at 12 months old.Although most farmers sell whole carcass (live or slaughtered) to consumers without weighing, average sale price of a unit live weight was computed for the purposes of this study (Table 1).
Culled does and bucks were disposed off at 90% of the expected weight of marketed does and bucks, respectively to compensate for wastage resulting from old age (Morris, 1980).Average life cycle of a breeding doe ended at 60 months and age at first parturition was 11.9 months.Young does were mated for the first time at 6.9 months.Mortality was considered at the pre-and post-weaning stages, and in the breeding doe herd.Doe replacement rate was 16.7%.This covered culling of old and unproductive does (14.6%) and also accounted for 2.1% mortality rate in breeding does.Breeding bucks were replaced at about 18 months of age when their female offspring were mature for mating.Replacement bucks were bought from outside and culled bucks were sold off.This means that mortality in breeding bucks was assumed to be zero.
Feed was provided in the form of elephant grass (Pennisetum purpureum), harvested freely from the rangeland by one labourer.It was assumed that the labourer spent only 2 h per day on the animals.Supplementary feed was provided in the form of concentrate (mixture of wheat bran, maize, oyster shell, common salt and vitamin-mineral-premix). Supplementary feed was provided to breeding does from birth to weaning and in a two weeks flushing period.Breeding bucks were given supplementary feed during a two week flushing period before mating.Feed (grass) costs were assumed to be zero (except compounded feed supplements) as grass was harvested free from the rangeland using costed labour.Costed labour also included cleaning of cages every morning.Treatment of sick animals against injuries and diseases was carried out by a veterinary officer.There was no cost associated with the marketing of animals because animals were sold at the farm gate.

Identification of sources of income and expenses
Profit was defined as the difference between income and expense.Income was derived from sale of surplus bucks and does, and culled does.In each class of animals, income was calculated as the sum of the product of number of animals sold and price per  individual.Expenses were derived from supplementary feed, veterinary service charges and drugs, and fixed costs (breeding stock, cages, labour, equipment and stationery).

Calculation of economic value of each trait
A partial budgeting approach was used to calculate economic values of traits (Barwick, 1992;Barwick and Fuscs, 1992).It involved combining income and expense as a function of profit.Economic value of a trait was defined as marginal profit per doe per year resulting from a one unit increase or decreases in the value of each trait, under the condition that performance levels of all other traits were held constant, at their mean values (Upton et al., 1988;Ponzoni and Newman, 1989).For example, if P is profit of breeding does in a herd and P 1 is the profit after increment of average value of a trait by one unit, then economic value of that particular trait is the difference between P and P 1 .Profit (P in Ghana Cedis, GH¢) was expressed as a function of traits in the breeding objectives as follows: Where, m is the number of animal classes in the profit function, n is the number of expressions for a trait in the i th class of animal, V is the revenue per unit, C the cost per unit for trait X, and K is fixed cost (Bekman and Van Arendonk, 1993).At the time of preparing the manuscript 1.45 GH¢ was equivalent to 1.00 US$ (BOG, 2010).
Combined life cycle profit was scaled down from a life time profit to a per doe per year basis by dividing total profit by total number of years in which a doe completed her entire life cycle.All costs and returns were discounted, taking into account the time of expression of traits (Smith, 1978) at a discount rate of 13.5% (BOG, 2010).
Economic value of docility was derived without using the above procedure (partial budgeting approach) because under the prevailing grasscutter production system, docility is neither connected to cost of production nor returns from production.Docility was defined as the ability of an animal to accept human presence.
The capacity of the animal to accept human presence was scored on a scale of 1 to 4 as follows: 1. Score 1: Docile -means that the animal is friendly and accepts to be touched and caressed.2. Score 2: Flighty -means that the animal accepts to be touched but moves away slowly.
3. Score 3: Restless -means the animal does not allow to be touched, stays away and move around in the cage.4. Score 4: Aggressive -means that the animal jumps when it sees somebody in an attempt to escape.It hits itself on the sides of the cage.
A docility test was carried out on 321 animals (162 females and 159 males) that were selected at random.Economic value of docility was derived by using regression analysis (Schroeder et al., 1992;Amer, 1994).An assumption was made that farmers will have to pay a premium price for buying docile animals in future.Farmers paid additional GH¢3.00,GH¢2.00,GH¢1.00 and GH¢0.00 to current average selling price if they are buying an animal with docility score of 1, 2, 3 and 4, respectively.Prices of animals were regressed on docility scores to compute economic value of docility (SAS, 2008).

Calculation of relative economic value of each trait
Relative Economic Value for each trait expressed as a percentage of total selection emphasis was calculated by multiplying discounted economic value by genetic standard deviation of the trait and then dividing each individual value by the sum of the absolute values of all traits (Wilder and van Vleck, 1988;Jagannatha et al., 1998;Wolfava et al., 2005;Eady and Garreau, 2008).The result was converted into percentage by multiplying by 100.A sensitivity analysis was carried out on relative economic values to find the effect of removing feed cost from the model.This was done because most grasscutter farmers in Ghana feed only grass to their animals.Grass is harvested free of charge from the rangeland.Another sensitivity analysis was carried out to find out the effect of using genetic coefficient of variation (CV g ) in computing REVs, instead of genetic standard deviation.Data on genetic standard deviation co-efficient of variation of traits were obtained from Annor et al. (2012a).

REVs when feed cost was considered
Results obtained for REVs of traits are presented in Table 3. REV of pre-weaning survival (PRS) was higher than that of post-weaning survival (POWS).This means that PRS was more important than POWS.Among reproductive traits, parturition interval (PI) was the most important trait, whilst age at first parturition (AFP) was the least important.Mature body weight of males (bMWT) and females (dMWT) were the most important body weight traits and birth weights (blBWT and dlBWT) were the least important.REVs of growth from birth to weaning (blPWDG and dlPWDG) were small compared with those for the post-weaning (bPODG and dPODG) period.Young buck feed intake (YBFI) and young doe feed intake (YDFI) were more important than buck feed intake (BLFI) and doe feed intake (DLFI).
Relative to all traits, mature body weights (bMWT and dMWT) had the highest REVs.These were followed by weaning weight and survival traits.Traits with the lowest REVs were BLFI and DLFI.The combined body weight traits had the highest REV (81.97%), followed by survival (11.16%), reproduction (3.04%), growth rate (2.91%) and feed intake (0.52%).Docility was the least important trait with REV of 0.40%.

REVs when feed cost was set to zero
When feed intake was set to zero, discounted EVs of all survival and reproduction traits increased slightly, with exception of POWS which remained constant (Table 4).
Discounted EV of docility also did not change.However, discounted EVs of all body weight traits regressed towards zero, whilst those of growth rates did not change.The new REVs of all traits increased whilst those of mature body weights decreased.Weaning weight and PRS ranked higher than mature body weight.
Relative to all other traits, the order of importance was blWWT, dlWWT, PRS, bMWT, POWS and dMWT, with AFP being the least important traits.Body weight traits still remained the most important traits, despite removing FI from the objective.When similar traits were combined, the order of importance of traits still remained the same as when FI was included in the objective.

Using coefficient of variation to standardize EVs
When the CV g was used to calculate the REVs, there were changes in magnitude and ranking of REVs (Table 5).Post-weaning daily gain and docility emerged as the most important traits (Table 5).These were followed by LSB, PRS and PI.The least important traits were preweaning food intake of males and females, respectively.Based on the combined REVs it was concluded that growth rate was the most important trait with REV of 42.94%, followed by reproduction (19.09%), docility (15.01%), survival (12.44%) and body weight (10.04%).
Feed intake was the least important trait with combined REV of 0.47%.

REVs when feed intake was considered
Relative Economic Value of PRS was more important than POWS.Similar observation was made by Nasiri and Fayazi (2008) in sheep production in Iran.Pre-weaning survival was more important than POWS because kids are more vulnerable than adults (Annor et al., 2000).Effects of diseases, poor nutrition and stress are more severe in kids than in adults.Ability of an animal to survive and produce up to a certain age reflects its ability to adapt to prevailing conditions.Adaptability increases as the animal grows.Burden of survival associated with the pre-weaning stage of growth is therefore higher than that for the post-weaning stage.Parturition interval, being the most important reproductive trait in this study, has also been recognized as one of the most important female reproductive traits in many livestock farming systems in the tropics (ESGPIP, 2010).Lower intervals result in lower culling of breeding females, lower cost of mating, fewer female kids for replacement and a higher length of productive life.
Relative Economic Value of mature weight was higher than birth weight and weaning weight, and post-weaning growth rate was also higher than pre-weaning growth rate.This observation was also made by MacNeil et al. (1994) and Annor et al. (2000) in cattle.Cost of production associated with product (growth or weight) production during the pre-weaning stage is lower than that of the post-weaning period and returns from product produced during the pre-weaning period is also lower than that of the post-weaning stage, because the postweaning period is longer than the pre-weaning.Cost of product production is determined by passage of time.Relative Economic Values of feed intake of both young does and young bucks at the pre-weaning stage were lower than those of the same class of animals at the postweaning stage.The above explanation given for body weight and growth rate also holds for feed intake.
Relative to all traits, the most important traits were mature body weight of males (bMWT) and females (dMWT).This was followed by blWWT, dlWWT, PRS and POWS, with the least important traits being BLFI and DLFI.Combining similar physiological traits, results of this work indicate that on average, body weight was the most important trait that made the greatest contribution to profit in relation to other traits.This was followed by survival, reproduction, growth rate and feed intake, with docility making the lowest contribution.This work contrasts earlier results reported by Annor et al. (2000) in beef cattle and Annor et al. (2007) in sheep in Ghana.
Results of those two studies showed that survival and reproductive traits were more important than body weight and growth traits.Difference between the last two reports and this is that whereas economic evaluation in this study was based on relative economic values, the other reports were based on absolute economic values.When this work is based on absolute economic values, survival and reproduction become more important than body weight and growth traits (Table 3).This study questions the statistic to use to standardize absolute economic values involving different physiological traits in order to calculate REVs.Genetic or phenotypic standard deviations have been used for computing REVs (Wilder and Van Vleck, 1988;Jagannatha et al., 1998;Wolfava et al., 2005;Eady and Garreau, 2008).However, this may not be appropriate when different physiological traits are considered because trait mean and the standard deviation tend to change together (Steel and Torrie, 1980;Gregory et al., 1995).Therefore, traits with large means such as body weight are likely to have higher REVs compared to traits with small means such as litter size.Furthermore, different traits have different units of measurements, and their comparison using the standard deviation, may be subjected to scale of measurement effects (Steel and Torrie, 1980).The coefficient of variation may be more appropriate for different physiological traits because it is used for relative comparison of variability of different means (Gregory et al., 1995).In some studies, the coefficient of variation of a trait is even used as its economic value (de Carvalho et al., 1999).

REVs when feed was set to zero
Effect of setting feed cost to zero increased EVs of PRS and all reproductive traits but decreased those of body weights.However, EVs of POWS, birth weight, growth rate and docility did not change.Relative Economic Values of body weight regressed towards zero and those of growth rate did not change because the tendency for animals to grow without feed was limited (Wolfava et al., 2005).The model ignored the capacity of animals to put on weight without feed, and therefore modified the results (Phocas et al., 1998).Annor (1996) reported similar results to this work.He observed that when feed cost was set to zero EVs of growth traits regressed towards zero.It is therefore important that feed intake should be included in models of economic evaluations (Wolfava et al., 2005).
Changes that occurred in EVs when feed cost was set to zero were minor, and did not affect ranking order of combined REVs of traits.Body weight traits were still more important than other traits.Similar results have been reported in dairy cattle.Vargas et al. (2002) and Kahi and Nitter (2004) found minor effects of feed price change on economic values of traits.

Using coefficient of variation to standardize EVs
The use of CVs to calculate REVs is not popular in animal breeding, although CVs are used as measures of genetic evolvability and diversity (variation) in the animal industry (Morris et al., 1978;McLennan and Lewer, 2005).The procedure points to the fact that growth rate, docility and litter size are the most important traits influencing the financial rewards of the farmer.In fact, litter size and docility would have been left out of the top traits if economic evaluation were based on genetic standard deviation.Reports from the literature justify the importance of these traits in genetic improvement programmes of the grasscutter industry.Grasscutter has poor growth rate (Adu, 2005).This has resulted in longer age at maturity or slaughter.The slaughter age is reported to be 43 to 53 weeks (Van der Merwe and Van Zyl, 2009), compared to a slaughter age of 8 to 16 weeks for the rabbit (Adu et al., 2005).Grasscutter in captivity has poor reproductive performance.This is reflected in their small litter size, long gestation periods and parturition intervals (Adu, 2005).The most important behavioural trait in grasscutter is docility.The National Research Council of the United States of America has recognized that there is a particular need to select and breed docile grasscutters because the animals are aggressive (NRC, 1991).

Traits to be included in the breeding objective
It was concluded that growth rate, litter size and docility are the traits suggested to be included in the breeding objective.Growth rate is an ideal trait because it has a positive genetic relationship with post-weaning survival and body weight, and negatively correlated with feed conversion ratio (Annor et al., 2012b).Improving growth rate will therefore improve survival, body weight and feed conversion ratio.In addition, growth rate has high genetic variance and heritability.Though litter size has low genetic variance and heritability (Annor et al., 2012a), its moderate relative economic value gives it an advantage to be included in the breeding objective.It also has a positive moderate genetic relationship with docility.Litter size has a negative genetic relationship with growth rate (Annor et al., 2012b).If it is excluded from the breeding objective, its mean value will deteriorate due to selection for improved growth rate (Goddard, 2009).Docility has moderate economic variance, high heritability and high relative economic value.

Conclusion
It was concluded that growth rate, docility, litter size and pre-weaning survival are the most important traits that will increase the financial rewards of the farmer.The results could be used as a guide to define breeding objectives for the grasscutter industry because emphasis on each trait in the objective is determined by its relative economic value.It is suggested that post-weaning growth rate, litter size at weaning and docility are the traits to be included in the breeding objective of grasscutter breeding programmes in Ghana.However, final traits to be included in the objective need to be determined in consultation with all actors (farmers, processors, retailers and consumers) in the industry.

Figure 1 .
Figure 1.Model flow diagram for 4 does and their offspring derived from average values of all traits.

Table 1 .
Assumptions used to construct grasscutter model.

Table 2 .
Biological traits influencing income and expenses.
a Acronyms of traits are defined in Table1.

Table 3 .
Relative economic values of traits when feed intake was included in objective.

Table 4 .
Relative economic values of traits when feed cost was set to zero.

Table 5 .
Relative economic values of traits when CV g is used.