Evaluation of genetic divergence among clones of conilon coffee after scheduled cycle pruning

Transcription

1 Evaluation of geneti ivergene among lones of onilon offee after sheule yle pruning J.M. Dalomo 1, H.D. Vieira 2, A. Ferreira 3, W.L. Lima 1, R.G. Ferrão 4, A.F.A. Fonsea 5, M.A.G. Ferrão 5 an F.L. Partelli 6 1 Instituto Feeral o Espírito Santo, Campus Alegre, ES, Brasil 2 Centro e Ciênias e Tenologias Agropeuárias, Universiae Estaual o Norte Fluminense Dary Ribeiro, Campos os Goytaazes, RJ, Brasil 3 Departamento e Proução Vegetal, Centro e Ciênias Agrárias, Universiae Feeral o Espírito Santo, Alegre, ES, Brasil 4 Instituto Capixaba e Pesquisa, Assistênia Ténia e Extensão Rural, Vitória, ES, Brasil 5 Empresa Brasileira e Pesquisa Agropeuária / Instituto Capixaba e Pesquisa, Assistênia Ténia e Extensão Rural, Vitória, ES, Brasil 6 Departamento e Ciênias Agrárias e Biológias, Centro Universitário Norte o Espírito Santo, Universiae Feeral o Espírito Santo, São Mateus, ES, Brasil Corresponing author: A. Ferreira aesioferreira@gmail.om Genet. Mol. Res. 14 (4): (2015) Reeive Marh 20, 2015 Aepte July 10, 2015 Publishe November 30, 2015 DOI ABSTRACT. Coffea anephora genotypes from the breeing program of Instituto Capixaba e Pesquisa e Extensão Rural were evaluate, an geneti iversity was estimate with the aim of future improvement strategies. From an initial group of 55 genotypes, 18 from the region of Castelo, ES, were selete, an three lones of the ultivars Vitória an robusta tropial. Upon ompletion of the sheule yle pruning,

2 J.M. Dalomo et al morphoagronomi traits were measure in the 22 genotypes selete. The prinipal omponents metho was use to evaluate the ontributions relative to the traits. The geneti issimilarity matrix was obtaine through Mahalanobis generalize istane, an genotypes were groupe using the hierarhial metho base on the mean of the istanes. The most promising lones of Avaliação Castelo were AC02, AC03, AC12, AC13, AC22, AC24, AC26, AC27, AC28, AC29, AC30, AC35, AC36, AC37, AC39, AC40, AC43, an AC46. These methos etete high geneti variability, grouping, by similarity, the genotypes in five groups. The trait that ontribute the least to geneti ivergene was the number of leaves in plagiotropi branhes; however, this was not eliminate, beause isaring it altere the groups. There are superior genotypes with potential for use in the next stages of the breeing program, aime at both the omposition of lonal variety an hybriizations. Key wors: Geneti improvement; Coffea anephora; Multivariate statistis INTRODUCTION The offee agribusiness is important, both soially an eonomially, sine offee is ultivate in more than 60 ountries, most of whih are less evelope or eveloping, an beause its erivatives are primarily onsume in rih an evelope ountries (Ceon et al., 2008). In Brazil, offee proution is onsiere the first nonolonial merantile ativity an is a key omponent of the national eonomy, yieling billion ollars for the trae balane in 2012 (MDIC/SECEX, 2013). The state of Espírito Santo is the seon largest Brazilian prouer of offee an the largest national prouer of onilon offee, with 311,200 ultivate hetares (CONAB, 2014). In this state, the offee business is the most important agriultural ativity, beause it reates jobs, wealth, an helps to maintain populations in rural areas (Rorigues et al., 2012). The proution of Conilon offee in the 2013 harvest was 8211 thousan saks (CONAB, 2014). Muh of the suess of the Brazilian offee ulture is ue to geneti improvement (Ivoglo, 2008; Melo an Sousa, 2011), an this, in the ase of perennial speies suh as offee, is epenent on knowlege of the available germplasm, the biologial variation among speies in the genus an among populations within speies, an of the variation among iniviuals (Costa et al., 2005), whih makes the stuy of geneti ivergene a neessity. In the searh for superior ultivars, the use of geneti variability in rosses of genetially ivergent groups is an important strategy for ahieving gains resulting from seletion. The importane of geneti iversity for improvement lies in the fat that it provies parameters for the ientifiation of superior genotypes, sine the hoie of genitors to form segregating populations is one of the main eisions that the breeer nees to make (Bertan et al., 2006). Aoring to Fonsea et al. (2006) an Moreira et al. (2009), multivariate tehniques an be use for this purpose, thus enabling multiple ombinations of information to our within the experimental unit. Thus, several multivariate tehniques an be use to preit geneti iversity, among whih the issimilarity measures involve Mahalanobis generalize istane, hierarhial grouping methos, suh as UPGMA, an graphi ispersion tehniques involving prinipal omponent analysis (Ceon et al., 2008). The hoie of metho to be use is base on the auray esire by the researher, as well as on the ease of analysis an ata olletion (Fonsea et al.,

3 Evaluation of geneti ivergene in onilon offee lones ; Bezerra Neto et al., 2010). Geneti variability among the genotypes of Coffea anephora has been reporte (Ferreira et al., 2005; Fonsea et al., 2006; Ceon et al., 2008; Ferrão et al., 2008; Ivoglo et al., 2008; Rorigues et al., 2012). However, the majority of biometri information use to estimate the geneti ivergene of offee plants has been measure in plants onute outsie the regime of pruning an isbranhing. Pruning an isbranhing are alreay wiesprea praties among onilon offee growers, an are use routinely to rejuvenate an maintain lan proutivity an to improve profitability for the prouer (Pereira et al., 2007). The option of sheule yle pruning (PPC) ompletely renews the aerial part of the rop every four or five years. Consiering this bakgroun an the lak of stuies on geneti iversity base on soli multivariate analysis involving traits measure after the management of sheule yle pruning, the objetive of this stuy was to evaluate promising genotypes of C. anephora an to estimate their geneti iversity, aiming to ompose varieties through future hybriizations, an reovery of superior genotypes from segregating populations. MATERIAL AND METHODS This offee test was onute in the experimental farm Bananal o Norte, belonging to Instituto Capixaba e Pesquisa e Extensão Rural (Inaper), in the istrit of Paotuba, muniipality of Cahoeiro e Itapemirim, south of the State of Espírito Santo, Brazil. The experimental rops were plante in June 2005, following a ranomize bloks esign, four repliations, with 55 treatments (genotypes of C. anephora var. Conilon). The genotypes belong to the Inaper breeing program, an 51 were lones erive from the phenotypi seletion of matrix plants from agriultural properties in the area of Castelo, State of Espírito Santo (Avaliação Castelo, AC), three were lones belonging to the ultivar Inaper 8142 (onilon Vitória), an one genotype was the open-pollinate ultivar Emaper 8151 (Robusta Tropial). The plots were ompose of a row of five plants, spae 3.0 x 1.2 m, an the seon an fourth plants were onsiere useful. Cultural praties were performe as reommene by Inaper for ommerial rops, with the aition of supplemental irrigation. After ompletion of the fourth harvest (September 2010), the experiment was subjete to PPC, maintaining two of the five orthotropi stems of the plants, an, of these, the plagiotropi branhes that ha proue grains in more than 50% of its rosettes were eliminate. Base on the riteria of vigor an tolerane to rust fungus, observe in the first 5-year evaluation perio ( ), an on proutivity, yiel stability, uniformity of maturation, an grain size, regaring the first four rops ( ), the 18 most promising lones were selete from the group AC, whih were assesse uring the agriultural year 2010/2011, together with the three lones belonging to the ultivar onilon Vitória (12V, early; 02V, meium; an 13V, late) an the genotype of the ultivar robusta tropial (RT), totaling 22 genotypes stuie. The morphoagronomi traits evaluate were: 1) number of orthotropi branhes proue per plant (NRO), obtaine by the monthly an umulative ount of new branhes proue (length 10 m), maintaining five new branhes in eah plant, to generate a new anopy, an removing the others; 2) ry matter of eliminate orthotropi branhes (MSO), obtaine by monthly an umulative weighing after rying in an oven at 65 o C for 72 h; 3) length of the new orthotropi branhes (CRO), obtaine by the istane between the insertions of the five new branhes with

4 J.M. Dalomo et al the ol an their apial meristems (m); 4) Diameter of the new orthotropi branhes (DRO) with stanarize measurement in the entral region of the seon internoe of eah of the five branhes (mm); 5) number of noes of the orthotropi branhes (NNO), etermine by ounting eah of the five new branhes; 6) average length of the orthotropi internoes (CEO), obtaine by the ratio between the length of the new orthotropi branhes (CRO) an their respetive number of noes (NNO) (m); 7) number of plagiotropi branhes proue (NRP), obtaine by ounting in eah of the five new orthotropi branhes; 8) average length of the plagiotropi branhes (CRP), measure in two selete branhes per plant, one on eah sie of the planting row, an obtaine through the istane between the insertion of these branhes in the orthotropi branh an its apial meristem (m); 9) number of plagiotropi noes (NNP), obtaine by iret ounting in the selete branhes; 10) number of leaves proue in the plagiotropi branhes (NFP), obtaine by monthly an umulative ount in the branhes mentione in item eight; 11) length of the plagiotropi internoes (CEP), obtaine by the ratio between the length of the branhes (CRP) an the number of noes (NNP) of the selete plagiotropi branhes (m); 12) largest iameter of the base of the anopy (DBC), measure in the transverse iretion in relation to the planting rows, having as limits the projetion of the branhes of greatest length (m); 13) Perentage of mature rosettes on the plagiotropi branhes (RM), obtaine by the ratio between the number of rosettes that proue flowers an the number of plagiotropi noes (NNP), ounte iretly in the selete branhes; 14) number of flowers proue by rosettes (FL/RM), obtaine by the iret an umulative ount, at every flowering, in the selete branhes; 15) number of remaining fruits per rosette (NF/RM), obtaine by ounting 30 ays after the last flowering, in the selete branhes; 16) perentage of evelope flowers (VING), obtaine by the ratio between the number of flowers proue by rosettes (FL/RM) an the number of remaining fruits per rosette (NF/RM); an 17) number of remaining fruits per branh (FR/RA), obtaine by multiplying the number of remaining fruits per rosette (NF/RM) an the number of rosettes that proue flowers per branh. Growth traits were evaluate monthly until September For the purpose of analyses, the umulative values were use for the traits NRO, MSO, CRO, DRO, NNO, NRP, CRP, NNP, NFP, an DBC an, the average values were use for the traits CEO an CEP. Proutive traits RM an FL/RM were evaluate after eah major flowering (August 24, September 07, September 17, Otober 05, an Otober 29, 2011), an the umulative values were use in the analyses. The traits NF/RM, VING, an FR/RA were evaluate 30 ays after the last flowering (November 29, 2011). Iniviual variane analysis was performe in GENES (Cruz, 2013) for the 17 traits uner evaluation, an, later, the Sott-Knott test (1%) was employe to luster the means to onfirm variation among the genotypes. In the next stage, the Mahalanobis generalize istane was etermine in orer to obtain the geneti issimilarity matrix an, epening on the istane between the iniviuals, grouping was performe aoring to the unweighte pair group metho with arithmeti mean (UPGMA). The relative importane of eah trait was verifie to isriminate geneti iversity through the prinipal omponent metho, base on whih variables were retaine. RESULTS AND DISCUSSION Base on vigor an tolerane to rust fungus observe in the first 5-year evaluation perio ( ), an on proutivity, yiel stability, uniformity of maturation, an grain size, regaring the first four rops ( ) (INCAPER, unpublishe ata), among those lones belonging to the AC group, the following were selete: AC02, AC03, AC12, AC13, AC22, AC24, AC26, AC27, AC28, AC29, AC30, AC35, AC36, AC37, AC39, AC40, AC43, an AC46, whih were onsiere

5 Evaluation of geneti ivergene in onilon offee lones the most promising for use in researh on the geneti improvement of onilon offee. The iniviual analysis of variane for 22 genotypes (18 from the AC group, three from the ultivar Vitória, an one from the ultivar Robusta Tropial) showe signifiant ifferenes at 1% probability as etermine by the F test, in 16 of the 17 traits measure, exept for the perentage of mature rosettes, RM (Table 1), emonstrating that there is geneti variability among the genotypes stuie. These results are appropriate, an sine they are assoiate with highly proutive genotypes (INCAPER, unpublishe ata), they provie favorable iniations of suess in the seletion of superior lones, with the possibility of use in breeing programs (Rorigues et al., 2012). Table 1. Analysis of variane, oeffiients of variation (CV), genotypi etermination oeffiient (H 2 ), an means of 17 morphoagronomi traits (1) evaluate in 22 genotypes of Coffea anephora belonging to the onilon offee geneti improvement program of Instituto Capixaba e Pesquisa e Extensão Rural - Inaper, ultivate at Estação Experimental e Bananal o Norte - EEBN, muniipality of Cahoeiro e Itapemirim, State of Espírito Santo, Brazil. F.V. GL NRO MSO CRO CRP DBC DRO NNO NRP NNP NFP CEO CEP RM FL/RM FR/RM VING FR/RA Mean squares Bloks Genotype ** ** 552.6** 181.9** 718.9** 11.7** 13.3** 39.4** 4.1** 18.3** 2.1** 1.2** 51.9 ns 67.7** 59.0** 471.1** ** Resiue CV(%) H 2 (%) (1) NRO: number of orthotropi branhes proue per plant; MSO: ry matter of eliminate orthotropi branhes (g); CRO: length of new orthotropi branhes (m); CRP: length of plagiotropi branhes (m); DBC: largest iameter of base of anopy (m); DRO: iameter of new orthotropi branhes (mm); NNO: number of noes of orthotropi branhes; NRP: number of plagiotropi branhes proue; NNP: number of plagiotropi noes; NFP: number of leaves proue on plagiotropi branhes; CEO: length of orthotropi internoes (m); CEP: length of plagiotropi internoes (m); RM: perentage of mature rosettes per plagiotropi branh (%); FL/RM: number of flowers proue by mature rosettes; FR/RM: number of fruits evelope per mature rosette; VING: evelopment of fruits (%); FR/RA: number of remaining fruits per branh. **Signifiant at 1% probability by F test. ns Not signifiant by F test. The experimental oeffiients of variation (CV e ) were within the range % (Table 1), whih is onsiere aeptable for experiments in perennial ultures suh as offee (Ferrão et al., 2008). Aoring to those authors, higher values of CV e an be observe in experiments involving genetially ifferent materials, beause of their ifferent responses to stresses suh as high temperatures an rought, iniene of pests, iseases, wins, an pruning. The traits that were less influene by the environment were CRO, DBC, DRO, NNO, NNP, CEP, an RM, presenting CV e of <10%. The traits that were more influene by the environment were NRO, MSO, FR/RM, an NF/RA, whih were between 20 an 30%. The genotypi etermination oeffiients (H 2 ) estimate from the means of the treatments, are steay from to 93.21%, exept for the trait RM, whih presente H 2 of 23.21% (Table 1). The high values estimate for the vast majority of traits iniate that the geneti variability is preominant in relation to the environmental (Ferrão et al., 2008), that there was aequate ontrol of experimental error (Ivoglo et al., 2008), an that there are favorable onitions for seletion an improvement of the evaluate traits. When means were lustere base on the Sott-Knott test at 1% probability (Table 2), five issimilar groups of genotypes were observe for NRO an CRO, three groups for MSO, CRP, DRO, CEP, FL/RM, FR/RM, an FR/RA, two groups for DBC, NNO, NNP, NRP, NFP, CEO, an VING, an, for the trait RM, all the genotypes were similar.

6 J.M. Dalomo et al Table 2. Means of 17 morphoagronomi traits (1) evaluate in 22 genotypes of Coffea anephora belonging to the onilon offee geneti improvement program of Instituto Capixaba e Pesquisa e Extensão Rural - Inaper, ultivate at Estação Experimental e Bananal o Norte - EEBN, muniipality of Cahoeiro e Itapemirim, State of Espírito Santo, Brazil. Clones NRO MSO CRO CRP DBC DRO NNO NRP NNP NFP CEO CEP RM FL/RM FR/RM VING FR/RA AC b 117.4b 51.50b b 20.21a 24.30a 32.95a 9.69b 19.37b 4.83b 5.34b 76.87a 25.24a 21.64a 85.6a 151.0a AC b 121.3b 64.19a a 20.57a 25.0a 27.30b 11.37a 22.75a 4.85b 5.65b 77.49a 20.62b 16.03b 77.7a 142.6a AC b a b 21.30a 22.5b 30.12b 9.81b 19.62b 4.68b 5.68b 77.94a 27.54a 15.18b 53.6b 110.1b AC a 288.8a b a 28.60b 9.25b 18.31b 3.57b a 18.86b a 92.6b AC b 133.7a 62.06a a 20.60a 24.62a 33.02a 10.75a 21.37a 5.44a 5.82a 74.72a 24.28a b 102.8b AC AC AC AC AC AC AC AC AC AC AC AC AC a a b 19.24b 21.82b 28.42b 9.81b 19.69b 4.99b 5.54b 78.82a 27.57a 19.51a 69.6a 146.1a 163.5b 116.5b 56.31a b 21.58a 26.22a 34.95a 11.62a 23.19a 4.44b a 19.88b 16.20b 81.7a 139.2a 153.1b 123.1b 61.25a a 20.82a 20.72b 24.87b 10.37b 18.37b 6.66a 5.87a 73.25a 21.21b b a 60.31a a 20.27a 23.22b 28.87b 10.06b 20.12b 5.54a 6.00a 78.54a 25.47a 19.35a 75.3a 150.5a 190.5b 128.2a 65.75a a 23.58a 23.20b 36.85a 12.12a 24.19a 5.53a 5.43b 75.01a 23.16a 15.20b 65.5b 132.8a b 226.7a b b a 30.45b 11.87a 23.56a 4.01b a 18.67b b 91.1b 180.9b b b 19.17b 22.55b 26.12b 9.25b 18.19b 4.80b 5.56b 78.11a 27.01a b 98.3b b 55.69a b b 28.00b 10.69a 21.19a 5.74a 5.19b 70.87a 20.85b b 101.1b 165.3b a b 18.94b 25.07a 28.55b 11.19a 22.19a 4.29b 5.30b 73.50a 21.21b b 103.8b b 273.0a 114.9b 62.44a a 21.34a 25.22a 34.57a 12.87a 25.75a 4.56b a a 100.3b 175.8b a b 19.10b 23.62a 28.02b 11.19a 22.12a 4.53b a 19.94b b 95.1b b 53.62a b 18.44b 20.27b 26.40b 10.00b 20.00b 5.73a 5.36b 75.22a 26.78a 21.49a 79.7a 148.4a 241.9a 133.2a 64.87a a 21.54a 23.85a 32.22a 10.06b 20.19b 5.58a 6.45a 73.65a b 72.7 (2) 12 V b a b 19.65b 24.17a 33.63a 11.81a 23.56a 4.52b a 23.57a 15.11b 63.2b 132.0a 02 V 13 V (3) b a b 20.38a 25.02a 30.92b 11.00a 21.94a 4.53b 5.29b 80.53a 19.38b 15.60b 79.5a 132.9a (4) b b 28.95b 9.12b 18.19b 3.75b a b 52.0 (5) RT b 115.4b 58.75a b 19.24b 22.42b 30.67b 11.00a 22.06a 5.14a 5.30b 70.89a b 56.8 (1)NRO: number of orthotropi branhes proue per plant; MSO: ry matter of eliminate orthotropi branhes (g); CRO: length of new orthotropi branhes (m); CRP: length of plagiotropi branhes (m); DBC: largest iameter of base of anopy (m); DRO: iameter of new orthotropi branhes (mm); NNO: number of noes of orthotropi branhes; NRP: number of plagiotropi branhes proue; NNP: number of plagiotropi noes; NFP: number of leaves proue on plagiotropi branhes; CEO: length of orthotropi internoes (m); CEP: length of plagiotropi internoes (m); RM: perentage of mature rosettes per plagiotropi branh (%); FL/RM: number of flowers proue by mature rosettes; FR/RM: number of fruits evelope per mature rosette; VING: evelopment of fruits (%); FR/RA: number of remaining fruits per branh. (2), (3), (4) Clones of ultivar Inaper onilon Vitória. (5) Open-pollinate ultivar - Emaper robusta tropial. The others are elite lones resulting from phenotypi seletion of matrix plants from agriultural properties of the region of Castelo, State of Espírito Santo, Brazil. Means followe by the same letters in the olumn belong to the same similarity group aoring to the Sott-Knott test at 5% probability.

7 Evaluation of geneti ivergene in onilon offee lones Conerning the features NRO an MSO, the genotype AC13 was highlighte, whih proue new branhes uring the evaluation perio, wasting grams of ry matter with the elimination. At the other extreme there was AC36, whih proue branhes, losing grams of ry matter. The proution of orthotropi branhes is a esirable trait beause of the nee for multipliation by the rooting of uttings an renewal of the anopy; however, it is unesirable from the perspetive of labor require for its removal an waste of energy to proue ry matter to be remove. Regaring the traits that efine the size of the plants (CRO, DRO, CEO, an NNO), AC22, AC46, AC28, an AC29 were highlighte as tall plants, an AC13 an 13V were of lower stature. In relation to traits that efine the arhiteture of plants (CRP, DBC, NRP, NNP, CEO, an NFP), the lones AC03 an AC12 present intermeiate CRO an one of the greatest DBC. Usually, the plants of lower stature an ompat arhiteture are selete, whih are appropriate to inrease ensity an generate tolerane to iseases, to the main pests, an to rought (Ferreira et al., 2005). In Table 2, the most prominent genotype for FL/RM was AC24, whih proue flowers per rosette; however, 30 ays after flowers were proue, AC02 was the most notable with fruits remaining per rosette, for reahing the highest evelopment rate (85.60%), an resulting in the highest number of remaining fruits per branh (151.0). On the other han, the negative aspet was represente by the genotypes RT an 13V: the first, in aition to prouing few flowers per rosette (14.66), also showe the worst rate of evelopment (49.82%), an the lowest number of fruits per rosette (6.94); the seon proue the lowest number of flowers per rosette (14.65) an remaining fruits per branh (52.0). Aitionally, the Sott-Knott test (1%) establishe two groups of genotypes for the trait NNP an only one group for RM (Table 2), suggesting that branhes with a lower NNP an ifferentiate a greater perentage of rosettes (RM). Arhitetural traits of the plant are highly hereitary an some present strong geneti orrelation with yiel. In partiular, the proportion of frutifiation of the plagiotropi noes ounte at 15 m from the uppermost region of the plant was foun to be a goo iniator of yiel over two frutifiation yles (Cilas et al., 2006). Ivoglo et al. (2008) reommene that some variables shoul be ispose in stuies on geneti ivergene in robusta offee beause the original grouping of progenies is not altere an they have high geneti orrelation with other evaluate traits. Prinipal omponents analysis estimate that the relative importane of eah trait for geneti iversity was, in inreasing orer: NFP, RM, VING, FR/RM, CRO, NRO, DBC, FL/RM, DBC, NNO, NNO, NNP, NRP, MSO, FR/RA, CEO, an CRO. Although NFP is a minor ontributor to geneti iversity, its elimination or the elimination of any other trait was not possible, sine, the number of groups was altere after isaring. With the enrogram obtaine through the hierarhial metho UPGMA (Figure 1) an onsiering the ut by the Mojena metho (1977) at 45% of the maximum fusion level, it was verifie that the, 22 genotypes were separate into five issimilar groups: Group I: AC02, AC24, AC26, AC36, AC37, AC40, AC43, 02V, 12V, an 13V; Group II: AC12, AC22, AC27, an AC35; Group III: AC03 an AC28; Group IV: AC29, AC39, AC46, an RT; an Group V: AC13 an AC30. In the geneti issimilarity matrix obtaine by Mahalanobis generalize istane, the most issimilar genotypes were AC03 an AC13, with a istane of , an these were ivie into ifferent groups through the UPGMA metho, a relevant result ue to rosses between lones from ifferent groups showing higher heterosis (Fonsea et al., 2006). The most similar genotypes were AC37 an AC40, with a istane of The most issimilar genotype was AC03, whih presente the highest sum of istanes among the pairs of istanes of whih this genotype was part, an

8 J.M. Dalomo et al AC40 was the most similar, an presente the lowest sum of istanes. Figure 1. Analysis of hierarhial grouping of 22 genotypes of onilon offee through the UPGMA metho base on the matrix obtaine through the Mahalanobis generalize istane. 02V, 12V, an 13V = lones of ultivar Inaper onilon Vitória; RT = open-pollinate ultivar - Emaper robusta tropial; the others (AC) are elite lones resulting from phenotypi seletion of matrix plants from agriultural properties of the region of Castelo, State of Espírito Santo, Brazil. Consiering the ut by the Mojena metho at 45% of the maximum fusion level. The genotypes 02V, 12V, an 13V, belonging to the ultivar Inaper 8142 (onilon Vitória), were part of the same group (I). The genotypes of AC that remaine in the same group (AC02, AC24, AC26, AC36, AC37, AC40, AC43) are biometrially similar to the lones stuie from the variety Vitória, an AC02, AC24, an AC43 obtaine the best ratings of the group in relation to proutive traits (RM, FL/RM, FR/RM, VING, an FR/RA). The genotypes of group II (AC12, AC22, AC27, an AC35) were haraterize by a meium size, an meium buing an evelopment of flowers; the genotypes of group III (AC03 an AC28), ha a large size an a high level of buing an evelopment of flowers; those of group IV (AC29, AC39, AC46, an RT), ha large size an a low level of buing an flower evelopment, an it was noteworthy that plants of AC (lonal plants) presente the same large size as the open-pollinate ultivar robusta tropial, whih is multiplie with sees. The genotypes of group V (AC13 an AC30) ha a small size an the proution of several orthotropi branhes per plant was a prominent trait. When an assoiation is establishe between the proutivities of the whole experiment in the first four rops (INCAPER, unpublishe ata) with the traits of groups previously esribe, lones AC02 an AC26 are highlighte in group I (fifth an sixth highest yiel), lone AC27 in group II (the highest yiel of the whole experiment), lone AC03 in group III (seon most proutive), lone AC29 in group IV (thir most proutive), an lone AC13 in group V (12th most proutive).

9 Evaluation of geneti ivergene in onilon offee lones The isolation of genotypes AC13 an AC30 in the last group emonstrate that they have a high geneti istane ompare to the other genotypes an that they an be use in programs of irete rosses aime at obtaining new or hybri ultivars of robusta offee, taking avantage of their small size when it is in the best interest of the breeer. The hoie of the breeer for a speifi group of genotypes epens on the harateristis an objetives of the researh; however, among those that possess the trait of interest, the most similar ones must be hosen when the objetive is releasing a new variety, an the most issimilar ones, when the objetive is hybriization. Base on yiel stability, uniformity of maturation, grain size, vigor, tolerane to rust fungus, an proutivity, the most promising genotypes within the AC group were AC02, AC03, AC12, AC13, AC22, AC24, AC26, AC27, AC28, AC29, AC30, AC35, AC36, AC37, AC39, AC40, AC43, an AC46. The existene of ivergent genotypes was observe, an five issimilar groups were forme. The trait that ontribute the least to geneti ivergene was the NFP; however, this was not eliminate, sine isaring it altere the groups. The evaluate genotypes present the potential for ontinuity in geneti improvement programs intereste in lonal variety an hybriization. Conflits of interest The authors have no onflit of interest to elare. ACKNOWLEDGMENTS Researh supporte by Instituto Feeral e Euação, Ciênia e Tenologia o Estao o Espírito Santo (Ifes), Universiae Estaual o Norte Fluminense Dary Ribeiro (UENF), Universiae Feeral o Espírito Santo (UFES), Instituto Capixaba e Pesquisa, Assistênia Ténia e Extensão Rural (Inaper), an Coorenação e Aperfeiçoamento e Pessoal e Nível Superior (CAPES). REFERENCES Bertan I, Carvalho FIF e, Oliveira AC e, Vieira EA, et al. (2006). Comparação e métoos e agrupamento na representação a istânia morfológia entre genótipos e trigo. Rev. Bras. Agroien. 12: Bezerra Neto FV, Leal NR, Gonçalves LSA, Rêgo Filho LM, et al (2010). Desritores quantitativos na estimativa a ivergênia genétia entre genótipos e mamoneira utilizano análises multivariaas. Rev. Cien. Agronom. 41: Ceon PR, Silva FFE, Ferreira A, Ferrão RG, et al. (2008). Análise e meias repetias na avaliação e lones e afé Conilon. Pesq. Agrope. Bras. 43: Cilas C, Bar-Hen A, Montagnon C an Goin C (2006). Definition of arhitetural ieotypes for goo yiel apaity in Coffea anephora. Ann. Bot. 97: Companhia Naional e Abasteimento (CONAB) (2014). Aompanhamento a Safra Brasileira - Café, Safra 2013/2014, quarta estimativa, abril/2014. Available at [ Aesse April 17, Costa RB a, Resene MDV e, Contini AZ, Rego FLH, et al. (2005). Avaliação genétia e inivíuos e erva-mate (Ilex paraguariensis St.Hil.) na região e Caarapó, MS, pelo proeimento REML/BLUP. Cien. Florestal 15: Cruz CD (2013) GENES - a software pakage for analysis in experimental statistis an quantitative genetis. Ata Si. 35: Ferrão RG, Cruz CD, Ferreira A, Ceon PR, et al. (2008). Parâmetros genétios em afé Conilon. Pesq. Agrope. Bras. 43: Ferreira A, Ceon PR, Cruz CD, Ferrão RG, et al. (2005). Seleção simultânea e Coffea anephora por meio a ombinação e análise e fatores e ínies e seleção. Pesq. Agrope. Bras. 40: Fonsea AFA a, Seiyama T, Cruz CD, Sakaiyama NS, et al. (2006). Divergênia genétia em afé onilon. Pesq. Agrope. Bras. 41: Ivoglo MG, Fazuoli LC, Oliveira ACB e, Gallo PB, et al. (2008). Divergênia genétia entre progênies e afés robustas.

10 J.M. Dalomo et al Bragantia 67: Melo B e an Sousa LB e (2011). Biologia a reproução e Coffea arabia L. e Coffea anephora Pierre. Rev. Vere 6: 1-7. Ministério o Desenvolvimento, Inústria e Comério Exterior (MDIC) - Seretaria e Comério Exterior (SECEX) (2013). Balança Comerial Brasileira. Available at [ Aesse January, 15, Mojena R (1977). Hierarhial grouping methos an stopping rules: an evaluation. Comput. J. 20: Moreira RMP, Ferreira JM, Takahashi LSA, Vasonelos MEC, et al. (2009). Potenial agronômio e ivergênia genétia entre genótipos e feijão-e-vagem e resimento eterminao. Semina Ciênias Agrárias 30: Rorigues WN, Tomaz MA, Ferrão RG, Ferrão MAG, et al. (2012). Estimativa e parâmetros genétios e grupos e lones e afé onilon. Coffee Si. 7: