**5. Breeding and genetic diversity based on morphological and agronomic traits**

In plant breeding, it is crucial to identify the most critical phenotypic traits to boost plant production. Consequently, the evaluation of trait occurrences and differences in a population is a key to determining probably valuable crosses among accessions. Although most studies focus on genetic diversity with molecular markers, it is also useful for plant breeders to recognize the morphological diversity of traits of interest [57].

Around the world, the arabica breeding programmes has the primary purpose of developing new cultivars taking into account the economic benefits to be returned to coffee growers. The target characteristics in the desired arabica cultivar are productivity, mainly focused on bean size as well as cup quality and resistance to major diseases and pests. On the other hand, each breeding programme has its own particularities that establish the priorities of selection criteria usually defined based on multifactorial variations in specific circumstances of weather conditions, soil, biotic and abiotic stresses, cropping systems, socio-economic factors, market dynamics and consumer preferences. In arabica coffee, typically, four primary methods of breeding and selection are used: 1- Pure line selection; 2- Pedigree selection after hybridization (sometimes also backcrossing); 3 - Intraspecific F1 hybrids; 4 - Interspecific hybridization (arabica x robusta), backcrossing and pedigree selection. The comprehensive overview of selection criteria and outcomes from each breeding method is presented in detail by Van der Vossen [58].

Gathering a series of studies, Monge and Guevara [52] make the compilation of the critical phenotypic markers for evaluation of coffee and suggests a list of appropriated traits evaluation markers: morphological descriptors - viz.: architectural (ramification degree, number of internodes, and length of plagiotropic branches) and physical (dimensions and color of leaves, flowers and fruits, flush color, stem diameter, et*c.*); phenological descriptor (flowering dates, fructification cycle duration); ecological adaptation descriptors (altitude, dry or humid regions, resistance to pest and diseases); productive descriptors (productivity level and early or late flowering, and fruit set); technological descriptors (coffee quality, the weight of 100 beans, caracoli rate, etc).

Monge and Guevara [52] in a review also outlined a compilation result of two studies concerning the phenotypic evaluation of 300 wild *C. arabica* collected in eight Ethiopia area, those accessions were added into CATIE collections in 1985. It highlighted the high variability in fruit maturation length (ranging from 130 to 258 days), a caracoli rate (varying from 1 to 71%), size of leaves, internode length and bean size. Furthermore, there was a detected correlation concerning morphologic variables - viz.: the lower ramification of the tree, the bigger the leaves and the bean that produces.

Cilas et al. [59], in a study concerning genetic value prediction for *C. arabica* production through evaluation of morpho-agronomic traits, having the yield registered throughout the first four years of production. They concluded that better coffee yield may be increased by the addition of the medium level of heterozygosity, once the hybrid present immense superiority in comparison to the parental line. Furthermore, these authors also affirm that the prediction of yield may also be fully achieved by combining morphological traits, for instance, stem diameter, number of primary branches and tree height.

Bertrand et al. [60], addressing efforts towards sustainability, performed a study in three Central American countries comprising of 15 trials between 2000 to 2006 aiming to assess F1 hybrids of *C. arabica* in the agroforestry system (shade) compared to full-sun (unshade) crop system. The experiment involved thirteen lines and twenty-one F1 hybrids that were measured to average production throughout the first production cycle earlier than pruning and coppicing. The results point out that the green coffee per tree yield was higher among F1 hybrids in contrast to traditional cultivar in 58%, aggregating to 170 g in agroforestry, whereas in the full-sun system this increment was 34%, accumulating 190 g. In this respect, the economic outcomes of both systems look quite similar. This study also discussed the economic advantage in the agroforestry system renovation with hybrids, indicating that after six years of replacing the traditional cultivar by hybrids could earn up 5000 USD/ ha. They were also pointing to the facilitation of credit policies and the opportunity of reaching new market niches with differentiated prices.

The first original phenotypic structure within *C. arabica* was present by Montagnon and Bouharmont [61]. The authors observed eighteen morphological and agronomic characteristics in a field collection of 148 accessions used the analyzed by multivariate approach. Interestingly, the result allowed identifying a sharp structure split into two main groups, comprise respectively 53 and 76 accessions. The other six groups are composed of less than five entries. The principal component analysis explained 77% of the accumulated variation within the first two axes, which is reasonably good. Also, the authors believe that the arrangement of the two main structured groups combined with the historical evidence of those accessions infers that group 1 has not been engaged within the domestication pathway of *C. arabica*. The traits modified by the course of domestication partly explained the well-defined separation of those two main groups.

The genetic diversity study conduced in Tepi National Spices Agricultural Research Center on 93 *C. arabica* accessions based 22 quantitative characteristics was able to detect five clusters by using multivariate techniques of hierarchical cluster and principal component analysis [62]. According to Klief [62], the significant inter-cluster distances between clusters point out that there is a high probability for obtaining transgressive segregates and maximize heterosis by crossing germplasm accessions across distinct clusters.

An study carried out in southwestern Saudi Arabia evaluated the genetic variation of accessions of *C. arabica* conserved *in-situ* in 19 localities, where stressful conditions prevail. Multivariate approach applied on 17 quantitative traits detected five groups. Interestingly, four accessions from the same place were grouped in four different clusters, supporting the importance of *in-situ* conservation strategy. All cluster showed significant inter-cluster distance, where two clusters present highest cluster distance. Therefore, Tounekti et al. [63] affirms that from these findings, it is suitable to explore this variability in breeding programmes to overcome environmental stresses.

The biochemical aspect of coffee liquor is highly essential. From this point of view, it was made a study addressing the genetic diversity based on caffeine content level concurrently with physical aspects of green bean characteristics and coffee

**69**

*Genetic Diversity of* Coffea arabica

*DOI: http://dx.doi.org/10.5772/intechopen.94744*

**6. Genetic diversity based on molecular markers**

the accurate discrimination of genotypes [71, 72].

The progress achieved in plant breeding programmes culminated in reduced genetic variability in the improved populations [36, 67–69]. This problem may be worse in species with a narrow genetic base, such as Arabica coffee (*C. arabica*). The narrow genetic base of this species is associated with its autogamy, the low number of plants that were initially distributed worldwide, and the recent evolution of the species [30, 36, 70]. Thus, genotype discrimination based on differences in phenotypic characteristics may be difficult because individuals who are genetically distinct may be phenotypically similar, which reduces the selective efficiency. To overcome this difficulty, molecular markers have been used as an important tool in

DNA markers allow the detection of variations in DNA sequences between individuals of the same species. Because they identify variations in DNA, they are stable and are unaffected by the environment or by pleiotropic or epistatic effects [73]. Thus, molecular markers have been used in breeding programmes as an efficient tool for the discrimination of genotypes and the analysis of genetic

cup quality. The examination of dissimilarities involved cluster analysis based on unweighted pair group arithmetic average (UPGMA), together with correlation among those variables analyzed. The outcome results consisted of two main groups were distinguished. The first cluster formed by 11 accessions distinguished by high caffeine content, undesirable physical characteristics of green bean and poor coffee cup quality. The other cluster split into two subgroups: the first with 26 accessions with caffeine content varying from low to average level and cup quality; the next subgroup with five accessions characterized by a medium level of caffeine content, desirable physical qualities of green coffee bean and high-grade cup quality. The authors also identify negative and significant associations linking caffeine content and all other variables related to cup quality. From that perspective, it is possible a simultaneous improvement of desirable cup quality plus low caffeine content [64]. A research, performed in IAC, evaluated the effectiveness of a minimum set of descriptors established for the conduct of test for distinctness, uniformity and stability in *C. arabica*. Twenty-nine cultivars were scattered in 11 groups when assessed by 35 morphological characteristics and three agronomic traits during three years. The results demonstrate that those descriptors were skilled in discriminating cultivar groups but a minor role in the identification of cultivars within each group. Therefore, the authors recommend the adoption of molecular markers and biochemical descriptors to identify cultivars to be protected more accurately [65]. Weldemichael et al. [66] conducted one well-designed study estimating genetic parameters in 49 accessions of *C. arabica*. It was used 26 carefully chosen appropriated quantitative traits aiming to estimate the phenotypic variation. The statistical analyses approach consisted of a series of adequate genetic parameters estimation. The findings exhibited the occurrence of variability for some morphological traits among coffee germplasm accessions. Interestingly, coffee berry disease recorded a pronounced genetic gain per population mean (88.8%); this point draws particular attention, once in arabica coffee disease resistance is a breeding objective of the chief priority to plant breeders. The detected low genetic advance as per cent mean and/or low genotypic coefficients of variation exhibited in most traits indicating these characteristics could not be developed through simple section rather heterosis breeding. Conversely, they advise that high morphological variation is not a guarantee of pronounced genetic variation; in this viewpoint, it is helpful to take into consideration the molecular and biochemical studies as a complementary approach.

#### *Genetic Diversity of* Coffea arabica *DOI: http://dx.doi.org/10.5772/intechopen.94744*

*Genetic Variation*

of primary branches and tree height.

of reaching new market niches with differentiated prices.

explained the well-defined separation of those two main groups.

accessions across distinct clusters.

The first original phenotypic structure within *C. arabica* was present by Montagnon and Bouharmont [61]. The authors observed eighteen morphological and agronomic characteristics in a field collection of 148 accessions used the analyzed by multivariate approach. Interestingly, the result allowed identifying a sharp structure split into two main groups, comprise respectively 53 and 76 accessions. The other six groups are composed of less than five entries. The principal component analysis explained 77% of the accumulated variation within the first two axes, which is reasonably good. Also, the authors believe that the arrangement of the two main structured groups combined with the historical evidence of those accessions infers that group 1 has not been engaged within the domestication pathway of *C. arabica*. The traits modified by the course of domestication partly

The genetic diversity study conduced in Tepi National Spices Agricultural Research Center on 93 *C. arabica* accessions based 22 quantitative characteristics was able to detect five clusters by using multivariate techniques of hierarchical cluster and principal component analysis [62]. According to Klief [62], the significant inter-cluster distances between clusters point out that there is a high probability for obtaining transgressive segregates and maximize heterosis by crossing germplasm

An study carried out in southwestern Saudi Arabia evaluated the genetic variation of accessions of *C. arabica* conserved *in-situ* in 19 localities, where stressful conditions prevail. Multivariate approach applied on 17 quantitative traits detected five groups. Interestingly, four accessions from the same place were grouped in four different clusters, supporting the importance of *in-situ* conservation strategy. All cluster showed significant inter-cluster distance, where two clusters present highest cluster distance. Therefore, Tounekti et al. [63] affirms that from these findings, it is suitable to explore this variability in breeding programmes to overcome environ-

The biochemical aspect of coffee liquor is highly essential. From this point of view, it was made a study addressing the genetic diversity based on caffeine content level concurrently with physical aspects of green bean characteristics and coffee

Cilas et al. [59], in a study concerning genetic value prediction for *C. arabica* production through evaluation of morpho-agronomic traits, having the yield registered throughout the first four years of production. They concluded that better coffee yield may be increased by the addition of the medium level of heterozygosity, once the hybrid present immense superiority in comparison to the parental line. Furthermore, these authors also affirm that the prediction of yield may also be fully achieved by combining morphological traits, for instance, stem diameter, number

Bertrand et al. [60], addressing efforts towards sustainability, performed a study in three Central American countries comprising of 15 trials between 2000 to 2006 aiming to assess F1 hybrids of *C. arabica* in the agroforestry system (shade) compared to full-sun (unshade) crop system. The experiment involved thirteen lines and twenty-one F1 hybrids that were measured to average production throughout the first production cycle earlier than pruning and coppicing. The results point out that the green coffee per tree yield was higher among F1 hybrids in contrast to traditional cultivar in 58%, aggregating to 170 g in agroforestry, whereas in the full-sun system this increment was 34%, accumulating 190 g. In this respect, the economic outcomes of both systems look quite similar. This study also discussed the economic advantage in the agroforestry system renovation with hybrids, indicating that after six years of replacing the traditional cultivar by hybrids could earn up 5000 USD/ ha. They were also pointing to the facilitation of credit policies and the opportunity

**68**

mental stresses.

cup quality. The examination of dissimilarities involved cluster analysis based on unweighted pair group arithmetic average (UPGMA), together with correlation among those variables analyzed. The outcome results consisted of two main groups were distinguished. The first cluster formed by 11 accessions distinguished by high caffeine content, undesirable physical characteristics of green bean and poor coffee cup quality. The other cluster split into two subgroups: the first with 26 accessions with caffeine content varying from low to average level and cup quality; the next subgroup with five accessions characterized by a medium level of caffeine content, desirable physical qualities of green coffee bean and high-grade cup quality. The authors also identify negative and significant associations linking caffeine content and all other variables related to cup quality. From that perspective, it is possible a simultaneous improvement of desirable cup quality plus low caffeine content [64].

A research, performed in IAC, evaluated the effectiveness of a minimum set of descriptors established for the conduct of test for distinctness, uniformity and stability in *C. arabica*. Twenty-nine cultivars were scattered in 11 groups when assessed by 35 morphological characteristics and three agronomic traits during three years. The results demonstrate that those descriptors were skilled in discriminating cultivar groups but a minor role in the identification of cultivars within each group. Therefore, the authors recommend the adoption of molecular markers and biochemical descriptors to identify cultivars to be protected more accurately [65].

Weldemichael et al. [66] conducted one well-designed study estimating genetic parameters in 49 accessions of *C. arabica*. It was used 26 carefully chosen appropriated quantitative traits aiming to estimate the phenotypic variation. The statistical analyses approach consisted of a series of adequate genetic parameters estimation. The findings exhibited the occurrence of variability for some morphological traits among coffee germplasm accessions. Interestingly, coffee berry disease recorded a pronounced genetic gain per population mean (88.8%); this point draws particular attention, once in arabica coffee disease resistance is a breeding objective of the chief priority to plant breeders. The detected low genetic advance as per cent mean and/or low genotypic coefficients of variation exhibited in most traits indicating these characteristics could not be developed through simple section rather heterosis breeding. Conversely, they advise that high morphological variation is not a guarantee of pronounced genetic variation; in this viewpoint, it is helpful to take into consideration the molecular and biochemical studies as a complementary approach.
