The genetic diversity of coffee plants cultivated in Western Amazonia was comprehensively analyzed using molecular markers. This analysis, which included 140 genotypes and 39,329 SNP molecular markers, enabled the classification of coffee plants based on their origin, botanical variety, and genealogy, distinguishing those derived from open pollination from those produced through directed pollination.

Among these genotypes, some clones were selected by the coffee growers themselves, with some being intensively cultivated [35]. However, the pedigree of these clones remains uncertain, as it is not known to which botanical variety they belong or from which they derive the greatest genomic contribution. For instance, clone C25 is believed to be a Robusta coffee plant, as it is included in group I and is genetically closer to the genotypes in group II, which comprises the Robusta group. Among the producers’ clones, GJ2 and GJ20 were the most distinct and divergent from the other genotypes, forming a separate group (group IV) (Figure 1).

Our results confirm the hybrid nature of the Western Amazon coffee cultivation. In the largest group (Group V), hybrid coffee plants obtained through open pollination are grouped, with a predominance of genotypes selected in the field by local farmers.

This genetic diversity can be leveraged both for plant adaptation to different regions and in the context of climate changes and variations. The gradual increase in temperature and irregular rainfall patterns may negatively impact coffee cultivation, leading to yield losses, increased incidence of diseases, and compromising the viability of most growing areas [37]. The lower altitudes and higher temperatures typical of the tropical Amazon climate define the environment where these plants were selected. The consistent performance of these genotypes has also been reported in other environments [10]. While the BRS1216 and BRS2336 cultivars have shown good performance in crops grown in Amazonas, Acre, and Roraima, hybrid clones such as R22, GJ8, GJ25, AS2, and LB15 have been transferred from the Northern region to the Southeast, being cultivated on a large scale [6].

The multiclonal cultivar BRS Ouro Preto, composed of 15 genotypes, is characterized by traits of the Conilon botanical variety [18]. In the present study, thirteen clones from this cultivar were characterized, with six grouped with the Conilons (group III) and seven grouped in the large group V, where the hybrid coffee plants are found, indicating the hybrid nature of these plants within the cultivar.

Genetic diversity data also suggest that the following genotypes maintained in the collection, despite having distinct identifications, may be the same clone or closely related materials: RobGuarini2107 and RobGuarini2106; AP3 and AP2; Rob32 and Rob126; BAG26 and BRS3213.

In addition to enabling selection gains, the greater genetic variability in this coffee species is linked to both the expression of hybrid vigor and the complementary traits of the Conilon and Robusta botanical varieties [38,39]. In the dendrogram, individuals assigned to different groups exhibiting higher genetic divergence may be selected for hybridization. Individuals in groups IIa, IIb, IIc, IId, and IIe, which are characterized by a higher frequency of genotypes from the Robusta botanical variety, have greater potential for hybridization with individuals from Conilon grouped in IIIa, IIIb, IIIc, and IIIe. High-performing hybrid progenies can be obtained from the most genetically divergent plants identified in this study, maximizing the potential for hybrid vigor expression and leveraging the complementary traits of these botanical varieties.

The DAPC analysis identified two main genetic clusters corresponding to the traditionally recognized Robusta and Conilon groups, confirming the effectiveness of SNP markers in elucidating genetic structure. From a breeding perspective, these findings provide valuable insights into the use of genetic variability and the development of new cultivars. The registered cultivars, BRS Ouro Preto and Robustas Amazônicos, showed distinct genetic profiles. BRS Ouro Preto was primarily assigned to the Conilon cluster, consistent with its origin, while Robustas Amazônicos exhibited a higher level of genetic admixture, reflecting a broader genetic foundation shaped by targeted intervarietal crosses.

Additionally, the analysis successfully identified key clones within their respective cultivar groups, such as clone BRSOuroPreto199 in BRS Ouro Preto and clone BRS2357 within Robustas Amazônicos. These findings confirm the consistency between field observations and molecular profiles, underscoring the utility of molecular markers in characterizing breeding materials.

Genotypes classified as public domain, located near the center of the DAPC plot, exhibited significant Robusta ancestry. This pattern suggests a history of well-documented introgression between Conilon and Robusta in the Amazon region [35]. Furthermore, the AGB Embrapa collection displayed substantial genetic variability, including highly differentiated genotypes. This diversity provides a strong foundation for identifying contrasting parental lines and optimizing crossing strategies.

Although C. canephora is a species that carries resistance genes, studies have shown that there are varying levels of resistance among different genotypes, with some coffee plants being resistant, moderately resistant, or even susceptible [34].

Coffee rust, caused by H. vastatrix, can be found in nearly all coffee-growing regions and can lead to significant production losses [33,40,41]. At least nine dominant genes present in different coffee species have been characterized, ranging from S_H1 to S_H9. The resistance genes S_H1, S_H2, S_H4, and S_H5 are derived from C. arabica, S_H3 from C. liberica, while the genes S_H6, S_H7, S_H8, and S_H9 have been identified in C. canephora. These genes are also present in the Timor Hybrid (HdT), a natural hybrid between C. arabica and C. canephora. Recently, Barka et al. [28] and Almeida et al. [24] identified and characterized two candidate SH resistance genes to H. vastatrix in the Timor Hybrid.

The presence of other resistance alleles and loci that confer coffee plants’ resistance to rust has also been studied in the coffee collection from Rondônia [25,26,42]. The more resistance genes a parent plant possesses, the greater the likelihood of obtaining resistant progenies. Field resistance is also linked to the emergence of new races of H. vastatrix, which can evolve more rapidly than the response of a breeding program. The high variability of the pathogen poses a risk to the durable resistance of coffee plants in the field, while this variability also serves as an important source of resistance [24].

Although predicting field resistance responses using molecular markers has limitations due to the complexity of factors influencing resistance, the segregation results of resistance markers are valuable for hybridization and the development of new breeding populations.

Limitations in this prediction can be observed in the group of genotypes that did not exhibit resistance markers for rust. While the clone BRS2357 is characterized by its susceptibility to rust, the genotypes of the Apoatã botanical variety (AP5, AP8, AP1, AP3) and the cultivar BRS1216 are known in the field for their high resistance to this pathogen. However, in the context of producing resistant progenies, hybridizing parent plants with a higher frequency of resistance markers increases the likelihood of pyramiding these genes and obtaining resistant plants from previously characterized parents.

Among the more than 50 races of H. vastatrix identified, race II stands out as the most common and widely distributed in the country [33]. Molecular data for QTL/GL2 (Locus A) indicated 47 homozygous resistant coffee plants (AA), 42 heterozygous resistant plants (Aa), and 22 homozygous susceptible plants (aa). For QTL/GL5 (Locus B), the marker used displayed a dominant pattern, allowing the identification of 51 resistant coffee plants (B_).

The resistance of coffee plants to races I, II, and pathotype 001 of H. vastatrix is conferred by two independent dominant loci [43]. Therefore, the presence of a dominant allele at either locus indicates a higher frequency of resistance alleles [24]. The best sources of resistance correspond to coffee plants with the genotype AAB_, as they possess both genes, leading to greater durability of resistance and homozygosity at locus A. For locus B, it was not possible to distinguish between homozygous and heterozygous plants, as the available molecular marker exhibits dominant behavior. The coffee plants with the genotype AAB_ include BAG 34, BAG 33, BAG 30, N1, AS5, Rob128, Rob36, GB4, WP6, Con530, Con69, LB60, LB88, LB110, and LB160.

The marker CARF005 (Locus C) enabled the identification of 82 resistant coffee plants. A significant number of coffee plants exhibited this dominant resistance marker, which may be attributed to the fact that this gene was cloned from Timor Hybrids that possess resistance genes from the species C. canephora [28].

The other gene previously cloned and monitored in this study corresponds to Locus D. With a dominant profile, this marker allowed for the identification of 33 coffee plants containing the resistance marker. The marker RLK2 was developed based on the presence and absence of the gene HdT_LRR_RLK2 in various differential coffee clones. These clones contain different combinations of rust resistance genes. Almeida et al. [24] noted that the gene HdT_LRR_RLK2 does not correspond to any of the previously characterized resistance genes (SH 1-9).

Coffee plants with the other four pyramided resistance genes were also identified, constituting promising sources of resistance. Notably, the coffee plants BAG30, Rob36, Con530, Con69, WP6, N1, P42, LB10, and BAG34 contain Locus A in dominant homozygosity, which means it will not segregate in the field. These coffee plants are potential parents for breeding programs, as they possess four pyramided rust resistance alleles.

Molecular markers were also employed to monitor loci of resistance to another disease known as CBD, caused by the fungus Colletotrichum kahawae. Two SSR markers were used, both identified and mapped by Gichuru et al. [31] and validated by Alkimin et al. [23]. These markers flank the gene Ck-1, which confers resistance to the disease. Through the combined analysis of the markers, 24 homozygous resistant coffee plants were identified. These individuals exhibit resistance markers for both molecular markers (Sat207 and Sat235). The observed recombinations—meaning the presence of only one marker—can also lead to the loss of resistance genes [44]. This is due to the estimated distance between the two markers, which is 17.2 cM [31]. The greater the distance between markers, the higher the chance of recombination occurring [29].

Although CBD is a disease restricted to the African continent, preventive breeding has proven important in coffee cultivation. In the 1970s, anticipating the arrival of rust in Brazil, resistant cultivars of C. arabica were evaluated in other countries, facilitating better management of this disease at the time. The availability of these markers increases the likelihood of obtaining resistant progenies even in the absence of the pathogen [23,27,41].

In the characterization of marker segregation, it was possible to identify coffee plants containing the highest number of pyramided alleles for both diseases. The coffee plants N1, N12, and N16 carry the four resistance alleles for rust evaluated in this study and also possess the allele for CBD (A_B_C_D_EE). These three coffee plants are open-pollinated hybrids selected by the same producer (Table 2). The dendrogram also shows that the coffee plants N1 and N16 are genetically similar, grouped in subgroup Vd.

Divergent crosses have shown promising results in obtaining plants with greater vigor and productivity, as well as exhibiting complementary characteristics of the botanical varieties Conilon and Robusta. The results of this study allow for the selection of matrices with greater potential for hybrid progeny production, based on the selection of matrices from different botanical varieties that carry resistance genes and are grouped in different clusters. For example, one can perform a cross between the coffee plant identified as Con154, which possesses pyramided genes (AaB_C_D_) and is allocated in group IIIb of the dendrogram, with the coffee plant identified as Rob36 (AAB_C_D_ee). This plant also has pyramided rust resistance genes and is allocated in linkage group IId.

Coffee cultivation has been part of the Western Amazon landscape since its colonization in the mid-1970s [36]. However, significant production highlights were achieved years after the cultivation began [6]. Currently, this region is the second-largest national producer of C. canephora [45,46]. Although coffee is part of a robust and structured production chain, it faces challenges such as global climate change [32,37]. Climate change may pose risks to resistance against both biotic and abiotic stresses. For example, studies project a reduction in the incubation period of coffee leaf rust, which could result in more severe epidemics [1].

The results of this study demonstrate the hybrid nature of the plants cultivated in Rondônia, a coffee culture carried out in the tropics characterized by low altitudes, high temperatures, and acidic soils, which favor the selection of plants with greater resilience to climate change.

The use of molecular markers in breeding programs allows for advancements in obtaining new breeding populations, assisting breeders in their decision-making processes.