Compatibility Test and Agronomic Performance of Coffee Genotypes (Coffea canephora Pierre ex Froehner) in the State of Rondônia, Brazil

The objective of this work was to evaluate the degree of compatibility and agronomic performance in clonal genotypes of canephora coffee plants (Coffea canephora ex Froehner) for cultivation in the state of Rondônia, Brazil. The study was conducted with nine genotypes with three replications of Coffea canephora Arranged in the field: UFRO-60; UFRO-31; UFRO-61; UFRO-25; UFRO-03; UFRO-08; UFRO-21; UFRO-05 and UFRO-138 In adulthood and in two years of harvest (2013/2014 and 2014/2015). We evaluated: productivity, profitability, conversion index fruit cherry/grain benefited, mass of one hundred grains benefited, average diameter of the fruits, number of rosettes per branch, average distance between rosettes and number of fruits per rosette. Additionally, the degree of compatibility between the clones of clonal. The genotypes of Coffea canephora with the best agronomic performances evaluated for the edafoclimatic conditions of the State of Rondônia in this study were UFRO-08, UFRO-25, UFRO-03 and UFRO-138. According to the compatibility tests, the sequence of correct disposition in the field of Coffea canephora among the genotypes studied in this study are: UFRO-138; UFRO-31; UFRO-25; UFRO-08; UFRO-60; UFRO-21; UFRO-61; UFRO-03 and; UFRO-05. Close to 74% of the tests Were and self-pollination tests resulted in low fruiting, evidencing the characteristic of self-incompatibility.


Introduction
Worldwide, it is estimated that 500 million people are involved in the coffee production chain (Damatta et al., 2007;Ferrão et al., 2017). Brazil stands out as the world's largest producer of coffee and has one of the most efficient coffee growers in the world with a planted area of approximately 2.2 million hectares in 2018 (CONAB, 2018). The species Coffea arabica L. and Coffea canephora Pierre Ex Froehner are widely cultivated in the country due to the edaphoclimatic conditions favorable, mainly in areas of Southeastern and Northern Brazil (Ferrão et al., 2017). In areas of the Amazon Region, the species is mainly cultivated Coffea canephora, due to specific altitude and rainfall requirements (Partelli et al., 2014).
In the State of Rondônia, located in the southern part of Western Amazonia, coffee (Coffea canephora) is configured as the most planted perennial culture and contributes to the formation of family income in rural areas (CONAB, 2018).It is estimated, according to CONAB (2018), that the planted area is about 63.900 hectares and approximately 2 million of benefited sacks produced annually. Although the average productivity is still low, with about 31 sacks/ha, there is an annual increase in productivity influenced by the technological package that involves fertilization, liming, irrigation and pest control in addition to the expressive use of new materials production for renewal of clonal crops (Dubberstein et al., 2017;CONAB, 2018).
Being fundamental for the increase of coffee productivity in the state of Rondônia, the genetic improvement, among other benefits, allows for uniformity of maturation, resistance to periods of drought, decrease of the effect of coffee bienality and greater Yield of the benefited fruits (Ivoglo et al., 2008). However, as the Coffea canephora is an alodic species, that is, dependent on other plants so that there is flowering and fruiting, the genetic compatibility among coffee genotypes becomes essential so that the crop can be productive (Marcolan & Espindula, 2015). The self-incompatibility between genotypes is the rejection of their own pollen or of clone plants that carry allelic characteristics that prevent self-fertilization (Nettancourt, 1997). According to Castric & Vekemas (2004), the capacity of the Coffea canephora in avoiding self-fertilisation is a vital characteristic of the species in order to avoid the deleterious effects caused by endogamy depression.
For this reason, it is indicated that there is a sufficient quantity of clones or genotypes in the planting area, in order to increase the chances of compatibility and production. Therefore, it is imperative that genetic materials undergo compatibility tests that ensure sufficient pollination, flowering and fruiting capacity, since clones not sufficiently compatible compromise these phenological phases, causing miscarriage and defective grains and consequently decreasing the productivity and quality of the final beverage (Lopes, 2015;Ferrão et al., 2017). Thus, the present study aimed to evaluate the degree of compatibility and agronomic performance in clonal genotypes of canephora coffee plants (Coffea canephora) for cultivation in the state of Rondônia, Brazil.

Location and Experimental Design
The work was carried out in 'Ouro Verde', in the rural area of Novo Horizonte do Oeste, in the State of Rondônia, Brazilian Amazon, Brazil. The climate is predominantly the hot and humid tropical type Aw, according to Köppen, with a well-defined dry period, occurrence of water deficit from June to September, average annual temperature of 25 ºC, average annual precipitation of 2,400 mm (Ageitec, 2019). A randomized block design was used, containing nine treatments, nine genotypes of Conilon coffee (Coffea canephora): UFRO-60; UFRO-31; UFRO-61; UFRO-25; UFRO-03; UFRO-08; UFRO-21; UFRO-05 and UFRO-138, arranged in lines, within the field of cultivation, as recommended by Ferrão et al. (2017a) and three repetitions. Except for the evaluation of the 2014/2015 crop, which did not count on the UFRO-61 clone, thus limiting itself to eight treatments. Adult plants were used, producing in the second and third commercial harvest, which were conducted with all the technical recommendations recommended by Ferrão et al. (2017). It was adopted a spacing of 3m x 1,3m, between lines and between plants, respectively, totaling approximately 2.560 plants ha -1 , conducted with 3 to 4 rods.

Experimental Conditions and Variables Analyzed
The harvest of coffee genotypes in the 2013/2014 and 2014/2015 was carried out in April and May when the plants had 80% of the cherry grain stage fruits, ideal harvest point as determined (Nunes et al., 2005). The experimental variables, except for productivity, were estimated from twelve plagiotropic branches taken from the median portion of the canopy of the useful area of the plot, being one branch for each cardinal point in each plant that composed the experimental plot. The evaluations that involved the obtaining of the grain mass were corrected to 13% of moisture, using the drying at full sun on the yard covered by canvas, and performed the periodic turning, according to the orientations of Pinheiro et al. (2012).
Plants were evaluated in open pollination: grain yield (Bags ha -1 ) determined according to average production per plant, multiplying by the number of plants per hectare; beneficiation bield, consisting of the conversion of fruits into 'coco' type in benefited grains; cherry/grain benefit ratio; mass of 100 grains; average diameter, measured using a caliper, determined from the arithmetic mean obtained from 50 fruits. The morphological characteristics were also evaluated: number of rosettes per plagiotropic branch; average distance between rosette, determined by the distance between the first and last rosette, and divided by the number of rosette and the average number of fruits per rosette for each genotype.

Genetic Compatibility Test
The compatibility test was installed in the main florade of the genotypes of Coffea canephora, a stage resulting from temporary water deficit, Following the procedures and steps proposed by Ferrão et al. (2017b) and Teixeira et al. (2011). The plants that would serve as pollen receptors were determined and constituted in an experimental plot. Each parcel received nine crosses, consisting of manual pollination with pollen from the other varieties and a self-pollination test. In each branch plagiotrópico selected himself four rosettes from the apex of the branch, jas.ccsenet. and in the emission o pollen don were prote of the burn One day beginning cut off and four days, normality fruiting pe initial date interfering compatibil Figure 1 Source: Au

Results
The . The n the ed by these s, the costlier is the production. Clonal crops, such as the ones in this study, present higher grain maturation uniformity, increasing the yield in relation to the seminiferous crops (Partelli et al., 2006).
The mass of 100 grains was significantly more in several clones, except in the genotype UFRO-61 (15.13 g). As can be seen in Table 1, the lower weight of the grains seems to be directly related to the Yield of this genotype, which was also the smallest. In general, the mean mass of 100 grains observed were higher than those observed by Kameyama et al. (2016) for several coffee genotypes in the state of São Paulo, Brazil, between 7.1 and 13.2 G. The genotypes studied by Ferrão et al. (2013a) obtained weights similar to the Obtained by the genotypes studied in this study. According to Medina Filho & Bordignon (2003) the mass of 100 grains is an excellent indicator of pollination and compatibility, because the emergence of 'Moca' grains indicates that only one of the grain compartments was fertilized, decreasing the weight of the grains and the productivity.   Regarding the fruit diameter, there were statistical differences only for the clones in the 2013/2014 harvest in which the clones UFRO-31, UFRO-08 and UFR0-0 were better (Table 2). According to Rocha et al. (2013), the largest grain size has a direct reflex in the final beneficiation of coffee with a positive correlation between this factor and the yield. Both the weight of the grains and the size are genetic characteristics controlled by several genes and may undergo variation among the genotypes and suffer with environmental factors such as water availability and fertilization management (Ferrão et al., 2017b). According to Silva et al. (2008) (Table 2). According to Rocha et al. (2013) the number of rosettes by branches favors the increase in the productivity of Coffea canephora and the selection of genotypes with superior characteristics. Verifying cultivars of Coffea canephora in the state of Rondônia, Rocha et al. (2013) observed a positive correlation between plant height and the number of rosettes per stem, allowing the selection of genotypes with lower heights, which facilitate treatment cultural, but with higher numbers of rosettes. The average number of rosettes obtained in this study are higher than those obtained by Ricci et al. (2013) in different shading conditions. According to the authors, the increase in luminosity influences the highest number of rosettes in the stem.
For the distance between the rosettes in the branches, the UFRO-05 clone showed the longest distance in the two harvests (2013/2014 and 2014/2015) ( Table 2). The longest distances between the rosettes are unfavorable in a selection program of superior genotypes, considering that it is related to the decrease in the number of rosettes that a plagiotropic branch may contain, decreasing the overall productivity of the coffee (Marcolan & Espindula, 2015). According to Charrier and Eskes (2004), the distance between rosettes is correlated to smaller grains, but a larger number of grains per rosettes. Ricci et al. (2013) verified distances lower than this work. In this respect,  Ferrão et al. (2017) both the spacing when the luminosity but limit the number of grains in the rosette, however, intrinsic characteristics of the plants can directly influence this item.   The compatibility test between clonal genotypes differed statistically for receivers (lines) e for donors (columns). There is the presence of at least two compatibility groups, being reserved the lowest level of compatibility for self-pollinations, evidencing the self-incompatibility existing in the coffee plant. Analyzing the interactions between the receptors, it was statistically observed that the UFRO-03 genotype presented a low level of compatibility with the UFRO-21 genotype, which is similar to that observed in the self-pollination test. However, it cannot be affirmed that these genotypes share one or more alleles, characterizing them as "half siblings", as evidenced by these results, since the reciprocal among the clones presents high compatibility. Similarly, it occurs with the UFRO-08 clone, which for the UFRO-21 and UFRO-05 clone, presented the aforementioned occurrence (Table 3).
These facts are justified by the loss of pollen viability, which suffers oscillations and the germination of the germinative power due to artificial manipulation that can result in different forms in each genotype (Mendes, 1949). Nevertheless, it is important to emphasize that the donor and receptor branches in this study were covered with sacks at a time that coincides with the warmest period of the year in the state of Rondônia, which may be an additional factor for the reduction of the viability of the pollen, and consequently contributed to the low number of fruits.
As for the clone of Coffea canephora UFRO-08, three compatibility groups were presented: the first of lower level, restricted to self-pollination, a second group to which the clones are inserted UFRO-21, UFRO-05 and UFRO-138. With median compatibility and a third group with the other genotypes, composing the high compatibility group. For the clone UFRO-21, four compatibility groups were identified: one of the lowest compatibilities linked to self-pollination, one second of median compatibility, where is the UFRO-138 clone, a third of high compatibility with the UFRO-05 clone and a fourth group with excellent level of compatibility, where all other genotypes are inserted (Table 3; Figure 2).    UFRO-25 That presented four groups, in which they were inserted in the first ones with a degree of compatibility equal to or above 86%, the second between 68 and 85%, the third between 10 and 67% and the fourth ones with 9% compatibility or less. In four occurrences, when the genotypes were pollens donors, there were statistically significant interactions between genotypes that manifested themselves as similar to the self-pollination test. The interactions between: Ufro-21: Ufro-03, UFRO-05: UFRO-31, UFRO-05: Ufro-08 and UFRO-05: UFRO-138 (Table 3).
Working with compatibility test between genotypes of Coffea canephora in the state of Rondônia, Brazil, Lopes (2015) verified compatibility between the tested plants and suggests the existence of more than one compatibility group, confirmed by this work. Some field observations indicated that in the Coffea canephora. There are three allelic forms of the "S" gene (S1, S2 and S3) controlling the gametophytic type self-incompatibility characteristic (Conagin & Mendes, 1961) (Figure 2). According to Omolaja and Falzun (2004), they have five allelic forms of the "S" gene can control compatibility and sustain the results of this work. According to the authors, the number of genotypic expressions would reach the ten case, confirming the existence of five alleles (S1S2, S1S3, S1S4, S1S5, S2S3, S2S4, S2S5, S3S4, S3S5, S4S5).
Through descriptive analysis, Monaco and Carvalho (1972) report that cross-matching between cultivars of Coffea canephora should present fruiting percentage equal to or greater than 50%, which was obtained in 74% of the tests performed in this study (Table 3). The literature reports the admission of 5 to 10% of self-fertilisation Schifino-Wittmann and Dall'agnol (2002), however, Monaco and Carvalho (1972) and Ferrão et al. (2017b) affirm that this percentage should be 0%, but this fact was not observed in this study. Further studies are needed to technically assess the compatibility between genotype of Coffea canephora, in order to identify the interaction of the materials for planting indication to farmers. Knowledge of compatibility or self-incompatibility between genotypes allows for increased productivity and final uniformity in Coffee cultivation (Ferrão et al., 2017;Schifino-Wittmann & Dall'agnol, 2002).