To the best of our knowledge, this is the first exhaustive phenotyping study where a cattle population has been constructed by a defined cross between trypanotolerant and susceptible breeds, and evaluated under natural tsetse- and trypanosomosis challenge. Phenotypes measured included number of infections, treatments, body weight, PCV and parasitaemia scores. The animals were allowed to graze in natural pastures and habitats infested with different species of tsetse flies transmitting different species and subspecies of trypanosomes, of varying virulence and intensities. This gave an opportunity to assess the actual behavior of the experimental animals in real life conditions in the presence of interaction of the environment, disease (trypanosomosis) and the vector (tsetse fly). To a large extent it followed the practices of an average, modern African livestock farmer and therefore, the results presented may be assumed to reflect the situation on the ground.
Gender and genetic type effects
Overall, the results of this study identify the female gender, and the genetic types with higher proportion of ND genome (i.e., F1 vs. BCB, BCB vs. KB, and F1 vs. KB), as the more trypanotolerant types. Within each specific comparison (e.g., female vs. male comparison in BCB, F1 and KB; or higher vs. lesser proportion of ND genome in F1 to BCB, BCB to KB, or F1 to KB comparison), there are some traits for which the supposedly less trypanotolerant type manifests the more tolerant phenotype. Similarly, within each specific gender or genetic type comparison, different traits show different levels of significance (non-significant, significant or highly significant). Examination of Tables 3 and 4 shows that degrees of significance and anomalous trait directions appear to distribute more or less randomly among the specific comparisons, and are not concentrated at specific traits. We believe that the most plausible explanation for these observations is that they represent sampling variation from a more trypanotolerant population value. That is, we propose that on a population basis, the trypanotolerant types (female vs. male, F1 vs. BCB or KB, BCB vs. KB) are more trypanotolerant across all of the traits, and variation in observed direction and degree among specific traits and comparisons are a matter of sampling variation only. With so many traits, types and comparisons, sampling variation in the less tolerant direction from a more tolerant population coupled with the converse from the less tolerant population, can give a contrast for which the more tolerant type have the less tolerant phenotype. Sampling in the opposite direction can yield a highly significant difference in the expected direction in one trait, non-significant difference in the expected direction for some other trait. Thus, we do not attribute variation in direction and degree of effect among the difference specific comparisons as representing true population differences in these attributes, and hence we will not discuss these differences in detail.
Combining all observations, then, we find that greater trypanotolerance whether conveyed by female gender or ND genome is associated with a lower number of positive trypanosome detections over the entire challenge period (STR); fewer infection cycles (NINF) and fewer treatments (NT); more days until first and second infection after exposure (DF1 and DF2), and from date of infection to treatment (DT1 and DT2); and higher average PCV across the entire infection period (MPC). Thus, this experiment fully achieved one of its primary goals in showing that trypanotolerance is expressed across the entire range of a challenge period, and in all aspects of trypanotolerance.
The finding of greater trypanotolerance of the female gender is in agreement with previous work reported [12, 13], the latter in Galana ranch involving Orma Boran cattle where in both cases, it was shown that females required fewer treatments than their male counterparts. Also, male vs. female weight-gain difference across the challenge period was less than the difference between male and female when reared for a comparable period in the absence of trypanosomiasis challenge.
Equally, this finding that female gender may be more trypanotolerant than the male is an important one as it confirms a similar hypothesis arising from a comparative study on susceptibility of male and female albino mice to Trypanosome brucei brucei, whose results suggested that the female mice were more trypanotolerant than their male counterparts, due to the fact that the polyclonal β-Lymphocyte factor is more active against trypanosomes in females than in males . It has also been reported that male N’Dama have higher plasma cholesterol levels than females and susceptible cattle have higher levels of plasma cholesterol and triglycerides than trypanotolerant cattle . Reasons for gender differences between levels of plasma cholesterol and triglycerides in both breeds are not known. However, it has been postulated that the low plasma lipid values of N’Dama would suggest that this animal has limited lipid nutrients to support the growth of trypanosomes, hence development of lower parasites during infection . Trypanosomes take up cholesterol, phospholipids and total lipids for growth. They also serve as sources of energy for trypanosome metabolism. These postulations may also explain the differential expression of trypanotolerance in the different genders but more exhaustive studies are needed to confirm this. Other possible explanation for the effect of gender on trypanotolerance points to the fact that trypanotolerant QTLs may be sex-linked or they may be in linkage disequilibrium with sex genes.
All in all, keeping female animals may be an attractive option in tsetse endemic areas, particularly when farmers are encouraged to put more emphasis on milk production and less use of draft power by bulls for field cultivation.
With respect to effect of ND genome, the F1 was clearly superior to the KB in all aspects of trypanotolerance, with minor exceptions. Since all trypanotolerance loci in the F1 are present in heterozygous state, this means that to a large extent, gene action at the trypanotolerance loci must have a strong additive component, in addition to any dominance or recessive effects. This conclusion differs from that based on QTL mapping in an F2 population , which reported primarily recessive gene action at their mapped trypanotolerance loci. Those conclusions  were based on rather complex QTL mapping analyses, while the F1 to BCB and BCB to KB results in the present study, although based on a smaller sample, are strong and direct. Therefore, we believe that they should be taken as representing the actual reality. This means that QTL mapping for trypanotolerance based on the BCB should be effective. If trypanotolerance loci were indeed recessive, they would not come to expression in the F1 or BCB.
The apparent superiority of the F1 in trypanotolerance relative to the BCB was greater than the apparent superiority of the BCB relative to the KB. The F1 was superior in 22 of 33 traits, of which 16 were at the significant and highly significant levels; while the BCB was superior to the KB in only 17 of 33 traits of which 7 were at the significant and highly significant levels. Yet on an individual locus basis, the expected additive superiority of the BCB to the KB is the same as the expected superiority of the F1 to the BCB. In the F1, all trypanotolerance loci are present in all individuals in heterozygous state; in the BC1 sample, half of trypanotolerance loci are present in all individuals in heterozygous state; in the KB, none of the trypanotolerance loci are present. Thus, across all loci, the additive population value for trypanotolerance of the BCB is as far removed from the population value of the KB, as the population value of the F1 is removed from the population value of the BCB. Similar unexpected results were obtained for mean body weight. The F1 was distinctly lighter than the KB (−43.94 kg), indicating an appreciable additive component for gene effect of body weight loci from the KB. But here too, the difference between F1 and BCB (−28.2 kg) was almost twice as great as the difference between BCB and KB (−15.74 kg). The explanation for these anomalies may lie in the specific environment of Batches 4, 5 and 6. Batches 4 and 5, on which the BCB vs. KB comparison is based were characterized by low tsetse challenge, and poor nutrition due to drought. This would have reduced differences due to trypanotolerance and growth rate. In contrast, Batch 6 on which the F1 to BCB comparison is based was characterized by high tsetse challenge and good nutrition. This would have allowed fuller expression of differences in trypanotolerance and growth rate.
In summary, the BCB individuals demonstrated increased ability to manage trypanosomosis as compared to purebred KB and were intermediate to a sample of F1 individuals. This shows a strong additive component to trypanotolerance loci, validating the use of the BCB as a mapping population (the results of the QTL mapping component of the study will be reported separately), and as a platform for development of a synthetic breed combining trypanotolerance of the ND with body size of the Boran. In addition to the comparison of genetic types, the results demonstrate clearly superior trypanotolerance of the female gender as compared to the male within all three genetic types (KB, BCB and F1), again supporting our previous study  even more as in the present case many more traits were analyzed. Thus, female BCB animals may be superior cow-mothers under field challenge. The study was unique in following a population under challenge over the course of time, showed that trypanotolerance was also reflected in the time course of infection, treatment, recovery and re-infection.
Trypanotolerance and the three components of animal disease resistance
The defense strategy of a host animal with respect to disease resistance has three components: avoidance, which aims to reduce the risk of exposure or infection; resistance, mediated primarily by the immune system, which aims to reduce pathogen burden following infection; and tolerance, which aims to reduce the deleterious effects of pathogen burden on the host . In a striking study  based on construction of Boran/N’Dama chimeric twins, it was shown that N’Dama trypanotolerance included both resistance, expressed as ability to control parasitaemia; and tolerance, expressed as ability to control anemia. Moreover, control of parasitemia and control of anemia were shown to be separate traits. All three defense strategies are exemplified in the F1 to KB comparisons of the present study and also in our previous study . The F1 animals had better ability than KB to control parasitemia, as shown by a lower percentage of positive parasitemia scores (%PAR) and lower total parasitemia score (TPS), exemplifying resistance. They had better ability than the KB to control the effects of pathogen burden, as shown by higher mean PCV across the entire test period (MPC), higher percentage of parasitemia detections that did not require treatment (NT1), longer interval from first parasitemic detection to treatment (T1), and fewer trypanocide treatments across the challenge period (NT), exemplifying tolerance. Finally, as uniquely shown in the present study, the F1 had better ability to avoid infection, as shown by delayed onset of first infection (DF1), and greater elapsed time between end of first infection and onset of second infection (DF2), exemplifying avoidance.