Genetic origin of donkeys in Brazil

The genetic groups of native donkeys in Brazil are characterized by adaptation to the local environment. However, the donkey population in the country is declining, mainly because of agricultural mechanization and transportation that has led to the abandonment and the consequent indiscriminate slaughter of these animals. There are three local genetic groups of distinct geographic and temporal formation. However, analyses of their origin, phylogenetic relationship, and population structure are scarce. Within this context, molecular markers such as the mitochondrial control region (D-loop) are useful for these analyses. This study aimed to evaluate the variation and origin of maternal lineages of groups of naturalized donkeys in Brazil (Brazilian, Nordestino, and Pêga). We detected five mitochondrial haplotypes with 19 polymorphic sites, two of them exclusively found in the Nordestino donkey; this group is in fact more distant from the others. Phylogenetic analysis indicates maternal contributions of two clades (Nubian and Somali) to the formation of the genetic groups of donkeys, a fact that explains the high diversity, structure, and distances of the groups, reported here for the first time. This analysis contributes production and conservation of native donkey breeds. It also gives clues about the formation of the Iberian breeds from which Brazilian donkeys originated.


Introduction
Donkeys (Equus asinus) were first introduced in Brazil around 1534 during the period of Portuguese colonization (Mariante and Cavalcante 2006). Currently, the country harbors three local genetic groups of donkeys: Brazilian donkey, Nordestino donkey, and Pêga donkey ( Figure S1). Alves et al. (2021) found wide intra-and inter-breed genetic variability and a strong genetic structure, indicating a high differentiation between the local groups of the country.
The most threatened genetic group (ecotype) is the Nordestino donkey. It originated in the northeastern region of the Brazil and is characterized by adaptation to the adverse conditions of the local semi-arid climate (e.g., hooves adapted to dry soil) and a small size, feature that results in lower maintenance requirements. The Nordestino donkey is commonly used for transportation, traction and as a saddle animal associated with family farming. It has been suffering intensely from the indiscriminate slaughter to produce a gelatin-based traditional medicine, called eijao, and it is officially under the protection of nongovernmental organizations (McManus et al. 2010;Carneiro et al. 2018;Bittencourt et al. 2021;Brandão et al. 2021).
The Pêga donkey is the only genetic group identified as a breed with a genealogy record in an association in Brazil, the Brazilian Association of Pêga Donkey Breeders (ABCJPêga, in Portuguese acronym). Originating in the state of Minas Gerais, it is used for the production of gaited animals for competitions, in addition to the production of gaited mules (ABCJPêga 2022).
The Brazilian donkey has a more recent origin and is located in the state of São Paulo. It originated from crosses between animals brought from Italy (McManus et al. 2010;Carneiro et al. 2018), and it is known for his aptitude for saddle, transportation of goods, traction, and production of mules (McManus et al. 2010).
The donkey population in Brazil has decreased dramatically in recent years. The estimated population was 886,506 animals in 2020, with a decrease of almost 500,000 animals over 24 years (https:// www. fao. org/ faost at/ en/# data/ QCL/). Several factors may have contributed to this decline, particularly their reduced use due to agricultural mechanization and abandonment associated with indiscriminate slaughter (Carneiro et al. 2018).
It is known that the origin and domestication of donkeys occurred in Northwest Africa in two distinct clades: clade I, whose representative is the Nubian wild ass (Equus africanus africanus) and clade II, whose representative is the Somali wild ass (Equus africanus somaliensis) (Beja-Pereira et al. 2004). Both clades influenced the formation of European breeds and clade II also influenced the formation of Asian breeds (Xia et al. 2019). However, little is known about the origin of the local donkeys in Brazil. Alves et al. (2021) suggested an origin from different clades, given the differentiation and structure detected. Identification of the genetic origin contributes to the understanding of the process of historical formation of genetic groups, as well as to the definition of conservation strategies. Therefore, this study aimed to evaluate the phylogenetic relationship between the local groups of donkeys in Brazil that suffered a population decline and to identify the clades of origin based on the mitochondrial D-loop region.

Sample collection
Biological samples (hair follicles) were collected from 30 donkeys (Equus asinus) of three native Brazilian breeds: Brazilian donkey (10 samples); Nordestino donkey (10 samples), and Pêga donkey (10 samples) ( Figure S1, Table S1). For phylogenetic inferences, the species Equus caballus (NC_001640.1; Xu and Arnason 1994) we selected as outlier group. It was used ten samples per group since a strong population structure was previously reported (Alves et al. 2021) and the aim was to identify the clade origin.

Molecular data collection and alignment
The DNA samples of the Brazilian donkey were kindly provided by the Brazilian Agricultural Research Corporation (Embrapa) through a Material Transfer Agreement (23,066.043532/2019-91). Total DNA was extracted from the hair follicle samples using the DNA NucleoSpin® Tissue kit (Macherey-Nagel).
The primers described by Lopez et al. (2005) were used for amplification and sequencing of fragments of a mitochondrial locus, control region (displacement loop, D-loop) (F: CTG GTC TTG TAA ACC and R: ACA GTT ATG TGT GAG CAT GG). The sequences were amplified between positions 15,351 and 15,949 bp. Donkey sequences available in GenBank were added, as well as sequences of Equus caballus as outlier group, Equus africanus africanus of clade I, and Equus africanus somaliensis of clade II ( Table 1, Table S2).
The amplification conditions were one cycle of initial denaturation at 94 °C for 5 min, followed by 35 cycles at 94 °C for 1 min, 54 °C for 1 min and 72 °C for 1 min, and a final extension at 72 °C for 5 min. The PCR products were purified using precipitation with 20% polyethylene glycol (Sambrook et al. 1989) and sequenced in an ABI PRISM 3500 automatic sequencer (Applied Biosystems).
For Bayesian inference, two independent runs of 10 million iterations, with four Markov Monte Carlo chains (MCMC), sampling a tree every 10,000 iterations, were performed. Stationarity and convergence of the runs (effective sample size-ESS > 200) were evaluated with Tracer v.1.6 (http:// beast. bio. ed. ac. uk/ Tracer). The potential scale reduction factor was also used to check convergence of the chain and burn-in (Gelman and Rubim 1992). For maximum likelihood estimation, 1000 bootstrap replicates we used.

Results
Specifically in the three genetic groups of donkeys, 19 polymorphic sites were identified in the mitochondrial DNA control region (524 bp) of the 30 individuals. The sites were parsimony informative and one of them carried singletons, generating five haplotypes (H1-H5) with haplotype (Hd = 0.740) and nucleotide (π = 0.01477) diversities. The same parameters were also presented for each group separately ( Table 2).
Two of the five haplotypes identified were exclusive to the Nordestino donkey and one to the Pêga donkey. The predominant haplotypes were H1 and H4. The H1 haplotype was found in seven individuals, five belonging to the Pêga donkey and two to the Nordestino donkey. The H4 haplotype was present in 13 individuals, three of the Pêga donkey as well as all the individuals of the Brazilian donkey ( Figure S2, Table 2).
The topologies recovered by Bayesian inference and maximum likelihood estimation based on the dataset were similar to other clusters of the species. Bayesian inference ( Fig. 1) rooted in Equus caballus revealed two main branches. Most samples (80%) of the Nordestino donkey were allocated along with the Nubian wild ass (Equus africanus africanus), clade I, while all samples (100%) of the Brazilian donkey and most samples (80%) of the Pêga donkey were assigned to the cluster of the Somali wild ass (Equus africanus somaliensis), clade II.

Discussion
With this study was identified, for the first time, that the three donkey genetic groups from Brazil were influenced by both domestication clades of the species. Both ass clades are known to have influenced the formation of European breeds (Xia et al. 2019). The local breeds in Brazil originated during Portuguese/Iberian colonization, thus having representatives of both clades (like the European breeds). Although there are no similar studies available for Portuguese and Spanish donkey breeds, the contribution of Iberian breeds may have been influenced by the Muslim invasion in the Middle Ages (which differs from the rest of Europe). This fact is also reflected in the genetic  It is possible to believe that the formation of the local breeds from Brazil occurred at different times and has different origins of Portuguese/Iberian donkey breeds since they belong to different clades. Sample size (n), total polymorphic sites (S), parsimony informative (S PI ), and singleton site (S S ), number of haplotypes (NH), private haplotype (PH), shared haplotype (SH), haplotype diversity (hd), nucleotide diversity (π) with their standard deviations (s.d.), and average number of nucleotide differences (k) within and across the three populations The formation of the Brazilian donkey occurred more recently, and its history is clear since it is known to be the result of mating between breeds of Italian origin. This can be easily observed in Figure S3, in which the Brazilian donkey is close to Italian breeds (clade II). Figure S3 shows a Bayesian phylogenetic tree constructed based not only on the sequences of wild donkeys (as in Fig. 1) but also of domestic donkeys from other countries.
Differently from the results herein obtained, Xia et al. (2019) reported the Nordestino donkey belonging to clade II, while we observed a greater participation of individuals in clade I. This finding may be due to a sampling effect. The unique genetic group from Brazil studied by Xia et al. (2019) was the Nordsetino donkey. Moreover, Xia et al. (2019) suggested the great differences between Peruvian donkey breeds and Nordestino donkey are because the donkey from Brazil came from the USA. We do not believe this to be the most adequate explanation. The contribution of the two clades to the formation of the Iberian donkey breeds is a more suitable hypothesis.
AMOVA results show significant variation between the donkey breeds (Table S2), as also reported by Alves et al. (2021) using another mitochondrial DNA region (D-loop).
The local donkey breeds in Brazil have different clades of origin, indicating different maternal contributions to their formation and, possibly, to the formation of breeds from the Iberian Peninsula. A better understanding of the history of formation of these genetic groups, their high genetic variability and stratification will contribute to the conservation and production of these animals. Moreover, it provides genetic information that can be used as the first criterion to identify ecotypes (Nordestino and Brazilian donkeys) as breeds.