Conflict of interest

selection of only a few individuals at the beginning of the process) and reproductive isolation between wild and domestic forms [52, 67]. However, a growing body of archaeological, genetic, and ethnohistorical evidence suggests that long-term gene flow between wild and domestic stocks was much more common than previously expected, and selective breeding of females was largely absent during the early phases of animal domestication [52, 67]. Therefore, complete separation between wild and domestic populations was relatively late and regionspecific [52]. These findings challenge assumptions about severe genetic bottlenecks during domestication and interpretations of genetic variability in terms of multiple instances of domestication and raise new questions regarding ways in which behavioral and phenotypic domestication traits were developed and maintained [52, 72]. The identity of the wild progenitor (or progenitors) of most domestic mammals remains also unclear because (i) the potential wild progenitors are often able to interbreed and produce fertile offspring with the domesticated congeners and (ii) many domestic animals can produce viable offspring with a host of wild, closely related sister taxa [32]. Therefore, the intuitive notion that each modern domestic animal (when discussed as a global population) is descended solely from a single wild species is almost certainly incorrect, and the genetic ancestry of domestics is likely to be relatively

Domesticated species are the result of a long and endless process that started millennia ago (Table 1). During about 98% of their domestication history, farm animals have been managed in a sustainable way by farmers, which lead to animals well adapted to local conditions [49, 50]. Yet, the situation changed dramatically 200 years ago as animals began to be selected for the same phenotypic characteristics to produce hundreds of well-defined breeds (Table 1), and reproduction among breeds was seriously reduced, leading to the fragmentation of the initial gene pool [49, 50, 70]. A few decades ago, the selection pressures were increased further, particularly with the use of artificial insemination, leading to a few industrial breeds with very high performances [49, 50, 70]. In the United States, the average milk production/cow of dairy cows increased by 1287 kg between 1993 and 2002, and 708 kg of this increase, or 55%, was due to genetics [78]. Interestingly, until the mid-1980s, most of the increase in milk yield was the result of improved management, in particular better application of nutritional standards and improved quality of rough age [78]. Since then, genetics became the major factor as a result of effective use of artificial insemination, intense selection based on progeny testing of bulls, and worldwide distribution of semen from bulls with high genetic merit for production [78]. This results in that, despite their total number of individuals, numerous industrial breeds have low effective population sizes [49, 50, 70]. This might explain that apart from a highly favorable increase in production, present-day selection for high production efficiency in livestock species in many cases was accompanied by undesirable side effects for several physiological, immunological, and reproduction traits [78, 79]. A new breeding goal aimed at improving fitness and tolerance of metabolic stress is necessary to prevent the decrease in the quality of life of farmed species and instead, perhaps, enhance it [70, 78–80]. More generally, an alternative to breeding for specific traits is to target "robustness" and "resilience," with the former focusing on current variation among environments and the latter on future variation [81]. Management strategies should be used to address short-term challenges from changing environments, and genetic selection should be used to address long-term problems [81]. Another solution might

complex [32, 40].

12 Animal Domestication

The author declares no conflict of interest.
