Farmers' breeding practices and traits of economic importance for indigenous chicken in RWANDA.

Data on breeding practices and traits of economic importance for the indigenous chicken (IC) were collected through personal interviews using structured questionnaires and direct observations of chicken management practices. The study was conducted from November 2015 to January 2016 in Rwamagana, Rulindo, Ruhango, Kicukiro and Muhanga districts of Rwanda. Data were collected and analysed through computation of indices, which represented a weighted average of all rankings of a specific trait. Spearman's non-parametric rank correlation was calculated for ranking of traits of economic importance to indicate the directional effects. The results on chicken ecotypes and their attributes showed that prolificacy, mature weight, disease tolerance, egg number and heat tolerance were highly preferred. The dwarf ecotype was most abundantly reared (38.84%) and considered to be significantly smaller and to have poorer growth rate, but to have better prolificacy than other indigenous chicken ecotypes. Selection of breeding cock and hen was based on disease tolerance, body weight at sexual maturity, body size and growth rate. In addition, for hen, mothering ability and egg fertility (Fer) were considered. Indices for the traits perceived by farmers as of primary economic importance were egg yield (0.093), disease tolerance (0.091), high growth rate (0.089), prolificacy (0.088), high body weight (0.087) and egg fertility (0.083). The most important traits considered by the marketers were body weight (BW), disease tolerance (Dtol), plumage colour (Pcol), egg yolk colour (EYC), meat quality (MQ), growth rate (GR) and egg yield (EY) whereas for consumers, meat quality, egg yolk colour, egg yield, body weight and growth rate were considered. Among traits perceived as important by farmers, a positive and significant correlation was found between BW and GR and Fer. Correlation was moderate for BW and prolificacy, drought tolerance (Drtol), Dtol and EYC. BW was negatively correlated with temperament (Temp), heat tolerance, Pcol and egg shell colour (ESC). Regarding marketers and consumers' preference rank correlation, positive and significant correlation was between BW and GR and MQ. As such, appropriate ecotypes (indigenous chicken) which have these characteristics need to be identified and utilised more based on their performance and adaption to the environment conditions to ensure efficient IC production.

Characterization of indigenous chicken production systems in Rwanda.

Characterization of indigenous chicken (IC) production systems in Rwanda was conducted from November 2015 to January 2016 with the aim of understanding socio-economic characteristics, management of IC, production parameters, feed resources, and constraints faced by farmers rearing IC. A survey involving 206 households was carried out in 5 districts with the highest populations of IC using structured questionnaires. Data were analyzed using SAS software. The results showed that the majority of respondents (62.4%) were males and 37.6% were females. The majority of respondents (83.6%) had formal education. All family members were involved in the chicken husbandry practices. However, women (78%) were highly responsible for IC management activities. The family size averaged 5 persons (ranging 2 to 13) per household. Land was privately owned by farmers (84%) with a mean holding of 0.87 ha per household. The production system was mainly extensive with minimum provision of supplementary feeds. Chickens were being housed in separate houses at night and mating was uncontrolled. Breeding and replacement stocks were mainly sourced from the hatching of eggs using broody hens (60.68% of respondents). There were no vaccination programs, and ecto- and endo-parasites control was done when need arose. Egg production, income generation, meat production, and production of breeding stock were the main reasons of keeping IC. The first 3 main challenges facing IC production were diseases outbreaks, lack of investment capital, and predators. Others challenges, such as feed shortage, thieves, fluctuation of market price, lack of information on poultry rearing, and lack of chicken houses, also were mentioned. The IC production constraints mentioned need urgent mitigation measures to sustain utilization of IC against the changing climatic and economic conditions. Therefore, individual, public institution, and non-governmental organization efforts are required to develop sustainable breeding objectives directed to functions of IC and production challenges faced in the systems.

Economic evaluation of progeny-testing and genomic selection schemes for small-sized nucleus dairy cattle breeding programs in developing countries.

In developing countries minimal and erratic performance and pedigree recording impede implementation of large-sized breeding programs. Small-sized nucleus programs offer an alternative but rely on their economic performance for their viability. We investigated the economic performance of 2 alternative small-sized dairy nucleus programs [i.e., progeny testing (PT) and genomic selection (GS)] over a 20-yr investment period. The nucleus was made up of 453 male and 360 female animals distributed in 8 non-overlapping age classes. Each year 10 active sires and 100 elite dams were selected. Populations of commercial recorded cows (CRC) of sizes 12,592 and 25,184 were used to produce test daughters in PT or to create a reference population in GS, respectively. Economic performance was defined as gross margins, calculated as discounted revenues minus discounted costs following a single generation of selection. Revenues were calculated as cumulative discounted expressions (CDE, kg) × 0.32 (€/kg of milk) × 100,000 (size commercial population). Genetic superiorities, deterministically simulated using pseudo-BLUP index and CDE, were determined using gene flow. Costs were for one generation of selection. Results show that GS schemes had higher cumulated genetic gain in the commercial cow population and higher gross margins compared with PT schemes. Gross margins were between 3.2- and 5.2-fold higher for GS, depending on size of the CRC population. The increase in gross margin was mostly due to a decreased generation interval and lower running costs in GS schemes. In PT schemes many bulls are culled before selection. We therefore also compared 2 schemes in which semen was stored instead of keeping live bulls. As expected, semen storage resulted in an increase in gross margins in PT schemes, but gross margins remained lower than those of GS schemes. We conclude that implementation of small-sized GS breeding schemes can be economically viable for developing countries.

Defining a breeding objective for Nile tilapia that takes into account the diversity of smallholder production systems.

In general, livestock and fish farming systems in developing countries tend to be highly diverse in terms of agro-ecological conditions and market orientation. There are no studies that have investigated if and how this diversity translates to varying preferences for breeding objective traits. This is particularly important for breeding programmes that are organized on a national level (e.g. government-supported nucleus breeding programmes). The aim of this study was to investigate whether Nile tilapia farmers with diverse production systems and economic constraints have different preferences for breeding objective traits. The second objective was to derive a consensus breeding goal, using weighted goal programming that could be used for a national breeding programme for Nile tilapia. A survey was conducted among 100 smallholder Nile tilapia farmers in Kenya to obtain preference values for traits of economic importance, by using multiple pairwise comparisons. Individual and group preference values were estimated using analytical hierarchy process. Low-income farmers preferred harvest weight, while medium- and high-income farmers preferred growth rate and survival. Grouping farmers according to market objective (fingerling production or fattening) showed that fingerling producers preferred growth rate and survival, while fattening farmers preferred harvest weight, height and thickness. Weighted goal programming was used to obtain consensus preference values, and these were used to derive desired gains for a breeding goal of a national breeding programme that takes into account the diversity of smallholder production systems.

Genetic and economic evaluation of a basic breeding programme for Kenya Boran cattle.

The objective of this study is to describe the present organisational structure of Boran cattle breeding and develop a model breeding programme using a deterministic approach under the current structure of the Boran Cattle Breeders Society. A breeding unit of 13,000 cows supplying bulls to a commercial population of 39,000 cows was assumed. Selection criteria used were growth and reproduction traits while breeding objective traits targeted were growth, carcass, reproduction, survival, milk yield and feed intake traits. Higher selection intensity was possible for breeding sires resulting in higher genetic gains compared to dams even though selection accuracies were generally low. Annual genetic gains were positive except for dressing percentage, cow survival rate, and age at first calving. However, a reduction of age at first calving by 9.5 days obtained a return of Ksh 49.76. The overall monetary genetic gain obtained was Ksh 84.89 with a profit per cow per year of Ksh 377.98. The basic breeding programme could be expanded to accommodate a larger population of Boran cattle. The implications of the results are also discussed.

Economic implications of using Japanese Black sires carrying recessive genes associated with genetic defects.

The objective of this study was to calculate cumulative discounted expressions (CDE) for Japanese Black sires carrying a single defective allele in a herd by applying the gene-flow method to investigate the expression pattern of homozygous recessive genotype and to evaluate the monetary loss of using these sires. A single biallelic locus was considered with A representing the dominant allele and a representing the recessive allele. The gene-flow method was modified to consider the fitness of homozygous recessive genotype. Input parameters representing a typical situation in a Japanese Black cattle herd were used to calculate the CDE and the loss of using carrier sires. The effects of initial allele frequency and fitness on the CDE were determined for Aa and AA sires. The CDE of Aa sires were larger than those of AA sires under all initial allele frequencies and fitness. The difference in the CDE between using Aa and AA sires was largest when fitness was 0 (lethal recessive condition). The differences in the loss between Aa and AA sires were larger with increasing initial allele frequencies in lethal recessive condition. Applying the method used in this study to defects reported in Japanese Black cattle and with a population size of 628 000, the difference in the loss between using Aa and AA sires was US$48 575 800 and US$74 418 000 in the case of Band-3 and Claudin-16 deficiencies, respectively. The approach used in this study could be applied to other genetic defects in different breeds and species.

Breeding goals for the Kenya dual purpose goat. II. Estimation of economic values for production and functional traits.

Economic values for production traits (milk yield, MY, kg; 12-month sale weight, LW, kg; consumable meat percentage, CMP) and functional traits (doe live weight, DoWT, kg; number of kids weaned, NKW; kidding frequency, KF; kidding rate, KR, %; doe weaning rate, DoWR, %; doe survival rate, DoSR, %; post-weaning survival rate, PoSR, %; pre-weaning survival rate, PrSR, % and; residual feed intake of yearlings, RFIgamma, kg and does RFId, kg) were estimated for the Kenya Dual Purpose goat (KDPG) for systems under two bases of evaluation. The production systems included smallholder low-potential (SLP), smallholder medium-potential (SMP) and smallholder high-potential (SHP), while the bases of evaluation considered were fixed flock-size and fixed feed resource. Under both bases of evaluation, economic values were highest in SMP apart from the economic values for feed intake-related traits (RFIy and RFId). In SMP, the economic values under fixed flock-size scenario were KSh 71.61 (LW), 20.90 (MY), 45.20 (CMP), 13.68 (NKW), 3.61 (KF), 6.52 (KR), 12.39 (DoWR), 22.96 (DoSR), 22.87 (PoSR), 13.18 (PrSR), -2.76 (RFIy) and -3.00 (RFId). The corresponding economic values under fixed feed resources scenario were KSh 73.28, 29.39, 45.20, 16.91, 4.76, 9.45, 13.84, 25.67, 25.15, 16.19, -2.76 and -3.00. Generally in all production systems, economic values for most traits were higher under fixed feed resource than under fixed flock-size scenario. In all systems, the economic values for most of the traits were sensitive to changes in prices of feed, milk and meat. The positive economic values for most traits under fixed flock-size scenario and fixed feed resource indicates that a unit increase in genetic merit for the traits would have a positive effect on the profitability of the systems.

Breeding goals for the Kenya dual purpose goat. I. Model development and application to smallholder production systems.

A deterministic model was developed and applied to evaluate biological and economic variables that characterize smallholder production systems utilizing the Kenya Dual Purpose goat (KDPG) in Kenya. The systems were defined as: smallholder low-potential (SLP), smallholder medium-potential (SMP) and smallholder high-potential (SHP). The model was able to predict revenues and costs to the system. Revenues were from sale of milk, surplus yearlings and cull-forage animals, while costs included those incurred for feeds, husbandry, marketing and fixed asset (fixed costs). Of the total outputs, revenue from meat and milk accounted for about 55% and 45%, respectively, in SMP and 39% and 61% in SHP. Total costs comprised mainly variable costs (98%), with husbandry costs being the highest in both SMP and SLP. The total profit per doe per year was KSh 315.48 in SMP, KSh -1352.75 in SLP and KSh -80.22 in SHP. Results suggest that the utilization of the KDPG goat in Kenya is more profitable in the smallholder medium-potential production system. The implication for the application of the model to smallholder production systems in Kenya is discussed.

Effect of direct and indirect selection criteria for efficiency of gain on profitability of Japanese Black cattle selection strategies.

The objective of this study was to determine the effect of use of residual feed intake (RFI) and the blood concentration of IGF-I (IGF) as selection criteria for efficiency of gain on profitability of Japanese Black cattle selection strategies with restricted test capacity. A breeding objective that integrated the cow-calf and feedlot segments was considered. Selection strategies were defined that differed in whether information on IGF and RFI during performance testing (RFIpt) was used to make selection decisions and in the number of animals measured for IGF. In all strategies, sires were selected from the proportion chosen during the first selection stage (performance testing), modeling a 2-stage selection process. The effect on genetic gain and profitability of variations in test capacity, of the genetic correlations of IGF with marbling score (MS) and RFIpt, and nonzero economic values for and, hence, inclusion of RFI of the cow and feedlot animals in the breeding objective was examined. Additional genetic gain and profitability were generated when information on IGF concentration and RFIpt in the performance-tested young bulls was included in the selection criteria. Profit per cow was optimal when measurement of IGF and RFIpt were incorporated together in the selection index. Increasing test capacity resulted in an increase in genetic gain in all strategies, and profit per cow was optimal in all strategies when 900 places were available for performance testing. Profit per cow was more sensitive to changes in the genetic correlation between IGF and MS than between IGF and RFIpt, especially when more animals were measured for IGF, or else the favorable relationship between IGF and MS had no significant effect on profit per cow. Additional genetic gain and profitability were generated in each strategy when RFI of the cow and feedlot animals were included in the breeding objective with nonzero economic values. These results may be used to provide guidance to Japanese Black cattle breeders and, in the absence of more breed-specific information, may also have application in other Japanese beef breeds.

Genetic relationships between growth and carcass traits and profitability in Japanese Brown cattle.

The objectives of this study were 1) to examine the genetic relationship between growth and carcass traits and carcass price (CaP) and profitability in Japanese Brown cattle, 2) to estimate economic values of carcass and growth traits as regression coefficients of price and profit traits on growth and carcass traits using a multiple regression model, and 3) to compare direct and indirect predictions of the genetic merit of profit obtained from multitrait analysis and selection index, respectively. Growth and carcass traits considered in this study were ADG during the feedlot period, CWT, LM area (LMA), rib thickness (RT), subcutaneous fat thickness (SFT), and marbling score (MS). Carcass price was evaluated as a price trait independent of its influence on profit. Profit traits were defined as 1) net income per year (PROF1), 2) net income per year/energy requirement (PROF2), and 3) net income per year minus feed costs (PROF3). Correlations between direct and indirect predictions were estimated based on EBV of sires and dams with progeny records. The heritability estimate for CaP was 0.41. The heritability estimates for profit traits were high and were 0.62, 0.66, and 0.60 for PROF1, PROF2, and PROF3, respectively. The genetic correlations between CaP and ADG, CWT, LMA, RT, SFT, and MS were 0.19, 0.14, 0.30, 0.38, -0.11, and 0.98, respectively. Among the profit traits, PROF1 had the greatest genetic correlations with growth and carcass traits. The correlations with ADG, CWT, LMA, RT, SFT, and MS were 0.30, 0.21, 0.24, 0.39, -0.01, and 0.69, respectively. These estimates indicate that use of profit traits as a selection criterion may promote desirable correlated responses in growth and carcass traits. The economic values for growth and carcass traits estimated relative to CaP and each profit trait differed because of the apparent differences in the description of these traits. The correlations between EBV for the same profit traits from direct and indirect predictions were high and ranged from 0.818 to 0.846 based on EBV of sires and from 0.786 to 0.798 based on EBV of dams. The strong correlations between direct and indirect predictions for profit indicate that there is no advantage to selecting directly for profit compared with an index with all of the component traits.

Effect of carcass price fluctuations on genetic and economic evaluation of carcass traits in Japanese Black cattle.

The objectives of this study were 1) to investigate the effect of changes in carcass market prices due to bovine spongiform encephalopathy (BSE) occurrences on estimates of genetic parameters and economic weights for carcass traits; and 2) to compare direct and indirect approaches for prediction of genetic merit of Japanese Black cattle for profitability of their progeny. The direct approach utilized estimated breeding values of carcass prices, whereas in the indirect approach, selection indices were constructed as products of economic weights and breeding values of component traits. Data were composed of 80,191 carcass records divided into 5 periods based on changes in carcass prices as a result of occurrences of BSE in Japan and the United States. The periods ranged from a period before occurrence of BSE in Japan to a period of beef import restrictions and a rise in prices. Carcass traits analyzed included HCW, LM area, rib thickness, subcutaneous fat thickness, and marbling score (MS). Price traits included carcass unit price and carcass sale price. Estimates of heritability for price traits were moderate (0.32 to 0.46) and slightly sensitive to changes in carcass market prices. Genetic correlations of HCW and LM area with price traits increased and that between MS and carcass sale price decreased with period, whereas estimates of genetic correlation between MS and carcass unit price were high in all periods (0.96 to 0.98). Economic weights for carcass traits varied with periods because carcass prices were highly sensitive to economic importance of traits. Nevertheless, correlations between within-period breeding values for price traits estimated using direct and indirect approaches were high (0.92 to 0.99). This result indicates that selection realized by direct and indirect approaches will provide very similar results. A comparison among within-approach breeding values estimated in different periods showed that the largest differences in breeding values of sires for price traits were between the periods after occurrences of BSE in Japan and in the United States. Economic effects of BSE occurrences influenced the importance of carcass traits and economic merits of price traits through a change of carcass prices from period to period, irrespective of the approach taken in determining the genetic merit of breeding animals for profitability of their progeny.

Evaluation of the efficiency of alternative two-tier nucleus breeding systems designed to improve meat sheep in Kenya.

A deterministic approach was used to genetically and economically evaluate the efficiency of five two-tier nucleus breeding systems for meat sheep in Kenya. The nucleus breeding systems differed in terms of whether the system was closed or open, in the type of animals that were involved in the movement of genetic superiority and in the number of selection pathways in each system. These systems were compared under four alternative breeding objectives based on monetary genetic gain and profit per ewe. The first objective simulated a situation where the flock size cannot be increased due to non-feed related constraints (FLOCK). The second specifically assumed that the flock size is restricted due to limited amount of feed resources (FEED). The third and fourth objectives assumed that sheep performed only tangible roles (TR) and both tangible and intangible roles (IR) in the production system respectively. Monetary genetic gains were highest for all objectives in an open nucleus system with a certain proportion of commercial-born ewes being introduced in the nucleus while at the same time utilizing young rams from the nucleus to breed sires and dams for the nucleus and commercial sector (ONyre). Utilizing young rams in a closed nucleus system for the dissemination of superior genes resulted in higher annual monetary genetic gain than utilization of old rams. Profit per ewe was significantly higher for FLOCK and IR in ONyre. In a closed system that allowed for downward movement of dams from the nucleus to the commercial sector to breed sires and dams, profit per ewe was highest for FEED and TR. The success of a nucleus breeding system should also focus on the profitability and logistics of establishing it. The implication of these results on the choice of two-tier nucleus breeding systems for the improvement of meat sheep is discussed.

Economic values for production and functional traits and assessment of their influence on genetic improvement in the Boran cattle in Kenya.

Economic values for production (sale weight in steers, SWs, kg and heifers, SWh, kg; dressing percentage, DP; consumable meat percentage, CMP and milk yield, MY, kg) and functional (cow weight, CoWT, kg; cow survival rate, CoSR, %; postweaning survival rate, PSR, %; feed intake of cows, FIc, heifers, FIh and steers, FIs, kg DM) traits were estimated for six production systems utilizing the Boran breed in Kenya. The influence of the estimated economic values on genetic improvement was also assessed using different selection indices. The six production systems were described according to their sale age, levels of input and final goal, namely, short-fed medium input beef (SMB); long-fed medium input beef (LMB); short-fed high input beef (SHB); long-fed high input beef (LHB); long-fed low input dual purpose (LLD); and long-fed medium input dual purpose (LMD). Two evaluation bases were considered: fixed herd size and fixed pasture input. In the fixed herd-size situation, the economic values for production (except MY in beef systems) and functional traits (except FI in all systems) were positive meaning a unit increase in genetic merit of these traits had greater influence on revenues than costs. As expected, the economic value for MY was negative in the pure beef systems (SMB, LMB, SHB and LHB) and positive in the dual-purpose systems (LLD and LMD). When compared with the economic values estimated in the fixed herd-size situation, in the fixed pasture input situation they were lower for feed intake in the three classes of livestock and other traits related to it in all systems. These traits were CoSR, CoWR, PSR, CoWT, SWh and MY. The economic values for CoWT in LLD and LMD were negative (KSh -11.14 and -15.33 respectively). The economic values did not influence much the direction of the genetic response in each trait in the different production systems. However, the magnitude of the actual gain was dependent on the index applied. The magnitude of the economic values for production and functional traits estimated in this study suggest that genetic improvement of these traits will have a positive effect on profitability of Boran cows kept in dual-purpose systems and when herd size is restricted. In beef systems, genetic improvement of MY will have a negative effect on profitability, especially when restrictions on herd size and feed exists.

Economic values for resistance to gastrointestinal helminths in meat sheep in Kenya.

Economic values for resistance to gastrointestinal (GI) helminths in meat sheep were estimated based on previously published genetic and phenotypic parameters in a selection index objectively optimizing gains on yearling weight (YW). A two-trait index of YW and faecal egg count (FEC) were considered when accounting only for tangible roles of sheep production using individual records for index calculation (scheme 1) or including individual, sire, dam and male paternal half-sib information (scheme 2). Schemes 3 and 4 were similar to schemes 1 and 2, respectively, but accounted for both tangible and intangible roles of sheep production. Economic values within each scheme were estimated for five breeding objectives as follows: (i) index response in YW equals response from single trait selection, (ii) index response in YW is maximum, (iii) the level of FEC was held constant, (iv) A predetermined response of -0.22 (x1000) eggs per gram was assumed and (v) Monetary value of response in FEC is minimum. All estimated economic values were negative, and were Kenya Shillings (KSh) -34.90 (US dollar -0.50), 0.00, -92.20 (US dollar -1.32), -153.31 (US dollar -2.19) and -47.90 (US dollar -0.68), respectively, for options 1-5 in scheme 1. In all schemes the breeding options ranked 4, 3, 5 and 1 in descending order; however, options 1 and 5 exchanged positions in schemes 2 and 4. Economic values were responsive to changes in heritability of FEC and genetic correlation between the two traits. The magnitude of the economic values reflects the importance of including resistance to GI helminths in the breeding objectives for meat sheep in different production systems.

Genetic and economic evaluation of Japanese Black (Wagyu) cattle breeding schemes.

Deterministic simulation was used to evaluate 10 breeding schemes for genetic gain and profitability and in the context of maximizing returns from investment in Japanese Black cattle breeding. A breeding objective that integrated the cow-calf and feedlot segments was considered. Ten breeding schemes that differed in the records available for use as selection criteria were defined. The schemes ranged from one that used carcass traits currently available to Japanese Black cattle breeders (Scheme 1) to one that also included linear measurements and male and female reproduction traits (Scheme 10). The latter scheme represented the highest level of performance recording. In all breeding schemes, sires were chosen from the proportion selected during the first selection stage (performance testing), modeling a two-stage selection process. The effect on genetic gain and profitability of varying test capacity and number of progeny per sire and of ultrasound scanning of live animals was examined for all breeding schemes. Breeding schemes that selected young bulls during performance testing based on additional individual traits and information on carcass traits from their relatives generated additional genetic gain and profitability. Increasing test capacity resulted in an increase in genetic gain in all schemes. Profitability was optimal in Scheme 2 (a scheme similar to Scheme 1, but selection of young bulls also was based on information on carcass traits from their relatives) to 10 when 900 to 1,000 places were available for performance testing. Similarly, as the number of progeny used in the selection of sires increased, genetic gain first increased sharply and then gradually in all schemes. Profit was optimal across all breeding schemes when sires were selected based on information from 150 to 200 progeny. Additional genetic gain and profitability were generated in each breeding scheme with ultrasound scanning of live animals for carcass traits. Ultrasound scanning of live animals was more important than the addition of any other traits in the selection criteria. These results may be used to provide guidance to Japanese Black cattle breeders.

Influence of production circumstances and economic evaluation criteria on economic comparison of breeds and breed crosses.

The ranking of genotypes (i.e., breeds and breed crosses) for economic performance depends on the production circumstances of the herd and the criteria for economic evaluation. In this study, the effects of evaluation criteria and production circumstance are quantified using data from the literature on six genotypes. The economic evaluation criteria measured at herd level included lifetime profit and lifetime profit expressed per day of calving interval, per day of productive lifetime, per day of total herd life, per unit of milk, and per unit of feed energy. Four production circumstances were studied; these circumstances included constraints on the output of milk and on herd use of concentrates, roughages, and both concentrates and roughages. Profit was determined based on the sale of milk, calves, cull cows, and heifers and the expenses incurred for concentrates and roughages, animal deaths, milking time, interest, and other inputs. Results indicate that production circumstances and evaluation criteria largely influence the ranking of genotypes and, therefore, the outcome of economic comparisons. The genotype that ranks the best under a certain criteria and circumstance ranks differently when these situations are altered. In economic comparisons of genotypes aimed at offering recommendations for implementation in smallholder dairy production systems in the tropics, the choice of the evaluation criteria should be determined by the limitation or constraint that characterizes the area where the genotypes are to produce or are currently producing. For example, for situations in which feed availability is limiting, genotypes should be ranked on total profit per unit of feed energy.