Genetic Gain Trend in Yield and Advanced Selection of Yayo Coffee (<i>Coffea arabica</i> L.) Land Race Collection

Merga, Dawit; Beksisa, Lemi; Alemayehu, Desalegn; Tefera, Fekadu; Adisu, Melaku; Benti, Tadesse; Ayano, Ashenafi; Giddisa, Gabisa; Kidane, Mebrate; Zakir, Mohammedsani

doi:https://doi.org/10.1155/2022/8322180

Advances in Agriculture

On this page

Abstract Introduction Materials Materials Results and Discussion Conclusion Data Availability Conflicts of Interest Acknowledgments References Copyright Related Articles

Research Article | Open Access

Volume 2022 | Article ID 8322180 | https://doi.org/10.1155/2022/8322180

Genetic Gain Trend in Yield and Advanced Selection of Yayo Coffee (Coffea arabica L.) Land Race Collection

Dawit Merga,¹Lemi Beksisa,¹Desalegn Alemayehu,¹Fekadu Tefera,¹Melaku Adisu,¹Tadesse Benti,¹Ashenafi Ayano,¹Gabisa Giddisa,²Mebrate Kidane,¹and Mohammedsani Zakir¹

Academic Editor: Amelia Salimonti

Received16 Apr 2022

Revised12 Sept 2022

Accepted26 Sept 2022

Published15 Oct 2022

Abstract

Coffea arabica L., the dominant cash crop in the world market, is native to rain forest of Ethiopia where it is believed to exist with high genetic diversity. Estimating genetic parameters are momentous in deciding breeding method to be followed for crop genetic improvement including Arabica coffee. The study was conducted with the intention to gauge genetic gain trend in coffee yield and to select advanced promising lines of Yayo coffee landrace for the next breeding step. The study was laid down at Metu research subcenter in 2013, using 124 coffee accessions that were established in simple lattice design under two sets each comprising 62 accessions including two checks. The over six year’s pooled analysis of variance indicated the handiness variability in yield performance among accessions. Moderate genotypic variance (15.46 to13.56%), heritability (56.16–81%), and expected genetic gain (15.52–20.8%) were observed. The genetic parameters and the superiority of check in yield over accessions elucidated that high yielder variety development by selection is difficult unless heterosis attaining breeding method followed, particularly for these Yayo coffee landrace origin. Common high genetic gain trend (49.19 and 100 kg·ha⁻¹) and response to selection (196.76 and 400 kg·ha⁻¹), selection differential 471.9 and 739.23 kg·ha⁻¹ were revealed in over four harvesting seasons mean value for both sets. Thus, selection is more effective in earlier season than late. High yielding accessions, Y27 and Y93, gave 3013.1 and 125.8 kg·ha⁻¹ yield gain over the high yielder check correspondingly. Despite the top 15 and 10 high yielders were selected from set-I and set-II, respectively, a total of 20 accessions with contrasting desirable traits were selected and established in crossing block for genetic improvement purposes via heterotic hybrid variety development program. These accessions were tolerant to major coffee disease and have desirable agronomic traits.

1. Introduction

Coffee is a herbaceous tree that belongs to family Rubiaceae and genus Coffea. It is a cash crop that can be propagated principally by seed. The dominant and noble in quality beverage from coffee species is Coffea arabica L. This coffee species is allotetrapliod and self-pollinated with some extent of outcross (10%) [1, 2]. It is native to south western Ethiopia where ample of diversity is authenticated by different scholars. Arabica coffee in its native ecology Ethiopia is growing under diverse environmental areas which range in altitude about 560–2600 m. a. s. l. that receives annual rainfall of 800–2000 mm [3–5]. This crop is mostly grown in tropical and subtropical regions [6, 7].

Arabica coffee is a predominant species which shares about 70% of the world coffee production [8–10]. Coffee contributes directly and indirectly for the livelihood of 125 million peoples in the world [5, 11]. Ethiopia is the only country that produces Arabica coffee only. Around 15 million (16% of population) of Ethiopian people lead their livelihood by the income generated from this commodity [11, 12]. Besides, Arabica coffee shares the largest percentage (29%) among the exported materials that earn foreign exchange income in the country.

Despite its high economical values and immense efforts made to develop 42 improved varieties by Jimma Agricultural Research Center from the last five and half decades [13, 14], coffee production in Ethiopia still remained low. Coffee production is hindered by different factors such as diseases, insect pest, and climate change [5, 7, 15]. The current climate change became opportunistic for newly emerging diseases like thread blight and insect pests like trips which are currently devastating coffee production in south western Ethiopia. Other contributing factors to low coffee productivity/production could be low technology adoption by farmers and the use of local varieties and traditional practices, very weak and nonuniform extension work by regionals agricultural offices, feeble on continuing the research legacy through extension work by the local experts, and exploiting genetic potential to the maximum for each coffee producing areas is still less than expected. Thus, to overcome low yield via the use of improved varieties and maximum exploitation of the genetic potential in each of the specific coffee producing regions of the country focus needs to be given to local landrace variety development program of coffee breeding strategy.

The knowledge of prominent genetic parameters determines the success in genetic improvement of any crop whether via pure line selection or hybridization activities. Estimation of phenotypic and genetic variance is basic for heritability estimation from which genetic gain of any desirable traits is derived [16, 17]. Plant breeders anticipated genetic gain periodically to compare the efficiency of different breeding strategies in attaining the improvement of desirable traits [18, 19]. Hence, estimation of genetic gain in addition to other genetic parameter in any desirable traits indicates breeders’ success in selection of promising line and developing outstanding varieties [20]. For Coffea arabica L., the genetic gain of 213.89 kg·ha⁻¹ recorded in yield per cycle of selection and 7.41 cm, 5 in number, 3 in number, and 7.41% genetic gain were reported for primary branch length, number of secondary branch, number of bearing primary branch, and percent of bearing primary branch, respectively [21]. For selection at 5% superior genotypes of Arabusta (C. Arabica x C. Robusta hybrid), 699.3 kg·ha⁻¹ genetic gain was reported in coffee yield per cycle of selection [22]. Similarly, Atinafu et al. [23] reported 201.8 kg·ha⁻¹, and Lemi and Ashenafi [24] obtained 345.68 kg·ha⁻¹ genetic gains in clean yield per cycle for the top 5% high yielders’ selection from the population of Arabica coffee.

Most genetic gain that authors reported in clean coffee yield and yield-related traits is per cycle of selection or genetic gain of a year, which gives less information about genetic gain progress with respect to desirable trait/s over multiple harvesting seasons. The estimation of realized changes in genotypic values for coffee yield over multiple cycles is referred to as realized genetic trend [19] which is very important in the selection of advanced lines for the next breeding program. However, such kinds of information which can be used to predict the multiple seasons’ genetic gain in clean coffee yield concurrent advanced line selection in improvement program has been lacking. Using local landrace variety development program of coffee breeding strategy, huge number of accession from Yayo coffee landrace were evaluated for many years. However, from the results of this study, there has been a gap to identify the genetic gain trend in yield and its fate needs to be determined. Hence, the current field experiment is designed to determine the genetic gain trend in clean coffee yield and to identify advanced selection of high yielding Yayo coffee accessions for the next breeding program.

2. Materials and Methodology

2.1. Description of the Study Area

The current experiment was conducted at Metu agricultural research subcenter of Jimma agricultural research center. The subcenter is at a distance of 272 km from Jimma agricultural research center and located 8°19' 0″ to North and 35°35' 0″ to East. This area has an altitude of 1558 m a. s. l. and with rainfall annually rain fall of 1829 mm, and minimum and maximum annual temperature 12.7°C and 28.9°C, respectively.

3. Materials and Experimental Design

The experimental materials used for the present study consist a total of 124 accessions of Yayo coffee land race that were collected from different coffee growing ecologies Yayo district, south western Ethiopia. The study was laid down in two sets (set-I and set-II) using simple lattice design in August 2013. In each set, 62 coffee accessions with two checks were comprised to conduct the study (Table 1). The coffee seedling was planted using spacing of 2°m × 2°m between row and plant with a total of six coffee seedlings planted per plot. All agronomic management such as fertilizer, shade, weeding, sucker management, and mulching were applied uniformly in all the plots as per the recommendation [25].

3.1. Method Used and Data Recorded

The coffee yield data was recorded per plot using fresh red cherry in Gram per plot from the crop bearing trees over harvesting time in cropping seasons [26]. Data for the dried coffee fruit at the last period of harvesting was recorded as drying Gram per plot, and the dry yield data was multiplied by 2.6 to convert to red cherry in Gram before computing the mean yield of genotypes. The mean of red cherry data was computed by dividing the total amount of red cherry in Gram per plot for the total number of bearing coffee trees per plot. Then, the mean of red cherry per tree was converted to clean coffee yield in Qha⁻¹, by multiplying red cherry by 0.00417 (conversion factor). Finally, the yield data in Q ha⁻¹ was converted to kg·ha⁻¹ which is the SI unit for weight. Yield data of the studied coffee accession was undertaken over six consecutive harvest years.

3.2. Statistical Analysis

All yield data collected were subjected to R-software and SAS version of 9.4 [27] for statistical data analysis. Analysis of variance was performed for yield and availability of variability indicated which is a rudimentary for estimating genetic gain, genetic gain trend, and heritability. The significant difference among accessions for yield was tested at 5% () probability level. Statistical random model was used for data analysis, and both genotypes and seasons were random factors. The following statistical model was followed, Yij = μ + Gi + Xk (j) + βj + εijk for single year data (Table 2) and Yijkr = μ + Gi + Cr + βj (Yr) + Xk (j) + (GC) ir + εijkr for over years data (Table 3). Where Yijk = response of Y trait from the i^th Genotype under j^th replication and Cr = effects of r^th level of years, µ = overall mean effects, gi = effects of i^th level of Genotypes, βj = effects of j^th level of replication within year, Xk (j) = effects of K^th level of incomplete blocks within replications, and εijk = the residual or random error component. PROC mixed procedure was used for statistical analysis. Homogeneity of variance was tested using Bartlett test [28] before pooled analysis of over years yield data. Mean separation test between Coffee genotypes was carried out using least significance difference (LSD). Advanced selection for the next breeding program was done at 24 % (the comprising the top 15 high yielders) and 16% (comprising the top 10 high yielders) for set-I and Set-II, respectively; the top fifteen and ten high yielders coffee accessions were selected according to their yield performance over six harvesting years.

3.2.1. Response to Selection

The genetic gain value of traits reponses to selection per cycle of selection; calculated as R = ihσp, where i is the selection intensity, h is the standard deviation of heritability, and σp is the phenotypic standard deviation.

3.2.2. Selection Differential

SD = iσp, where SD is the selection differential, σp is the phenotypic standard deviation, and i is the selection intensity.

3.2.3. Estimation of the Rate of Genetic Gain

The expected gain per unit of time is referred to as the rate of genetic gain which can be computed by a method developed by Falconer [29]which is as follows: (ΔGA = R/L); where ΔGA is the rate of genetic gain, R is the response to selection, and L is the time required for selection. In general, the mean predicted yield gain across years (Y) can be estimated as [30, 31]: where i = standardized selection differential, h² = estimated broad sense heritability on a genotype mean basis, and sp = square root of the estimated phenotypic variance across years. h² = σ²_g/(σ²_g + σ²_gy/Y + σ²_e/YR) for over years; where σ²_g is the genotype, σ²_gy is the GY interaction and σ²_e is the experimental error components of variances (all estimated from ANOVA for the combined data), and R and Y are, respectively, the number of replicates and harvesting years (cycles) assumed/hypothesized for selection.

3.2.4. Components of Variance

Error (σ²e), genotypic (σ²g), and phenotypic (σ²p) variance was computed following the formula suggested by Hallauer et al. [32], and Singh and Chaudhary [33]. σ²e = Mse/r; Mse–mean square of error, σ²g = (Msg-Mse)/r; Msg-mean square of genotypes, r-replication, and σ²P = σ²e + σ²g.

3.2.5. Broad Sense Heritability

Broad sense heritability is calculated as follows: H² = σ²g/σ²p for single year, where σ²g is the genotypic variance and σ²p is the phenotypic variance, and H² = σ²_g/(σ²_g + σ²_gy/Y + σ²_e/YR) for over years; where σ²_g is the genotype variance, σ²_gy GY is the interaction variance and σ²_e is the experimental error of variance (all estimated from combined data), and R and Y are the number of replicates and harvesting years (cycles) assumed/hypothesized for selection, respectively.

3.2.6. Phenotypic and Genotypic Coefficient of Variation

Phenotypic and genotypic coefficient of variation is computed as follows: GCV = (√σ²g)/x and PCV = (√σ²p)/x where GCV is the genotypic coefficient of variance, PCV is the phenotypic coefficient of variance and x-general mean.

4. Results and Discussion

4.1. Analysis of Variance

There was a highly significant difference () among accessions of set-I in 2015 and 2016 harvesting seasons; while, there was also a significantly difference () observed among accessions of this set in 2018 and 2019 harvesting seasons (Table 4). However, there was statistically nonvariable performance in clean coffee yield among the accession of set-I in 2017 and 2020 harvesting seasons. From set-II, a highly significant difference yield was revealed among accessions in 2018 and 2020 yield harvesting years; this expected in Arabica coffee [34, 35]. But, accessions of set-II showed nonsignificant difference in yield performance in 2015, 2016, 2018, and 2019 harvesting years. Accessions of both sets (set-I and II) showed highly significant difference in mean of clean bean yield over six years. This pooled mean points out the handiness of variability among the examined accessions. The present results confirmed the finding of Dawit et al. [21] and Kitila et al. [36] who reported significant difference in yield among coffee genotypes for clean coffee yield.

4.2. Genotypic, Phenotypic Variance, and Heritability in Coffee Yield

The genotypic and phenotypic coefficient of variance including broad sense heritability of clean coffee yield is elucidated in Table 4. The whole performance or phenotype of crops wasconditioned by environment and the inherent part by genetic factors of the genotype; especially perennial crops such as coffee are highly affected by cumulative environmental factors as it persists for 15 years and above years. The current results indicated high genotypic coefficient variance (GCV > 20%) for yield in 2015 (GCV = 24.78%), 2016 (GCV = 27.09%), and 2018 (GCV = 24.07%), and moderate GCV (10–20%) was recorded in 2019 (GCV = 16.82%) and 2020 (GCV = 15.87%) harvesting seasons for set-I. In contrast, the lowest GCV (<10%) was revealed in 2017 (GCV = 9.17%) year in set-I. In line with this, Akpertey et al. [37] reported a high to moderate range of GCV (59.26 to17.58%) value in four harvesting seasons.

Besides in set I, high phenotypic coefficient of variance (PCV > 20%) had been observed in all harvesting seasons which ranged from 47.59% to 20.43% in 2020 and 2017 harvesting years, respectively. In set I, moderate heritability (H² 30–60%) 60.7, 50.9, 41.69, and 37.17% was recorded in 2015, 2016, 2018, and 2019 harvesting seasons; however, the lowest H² (<30%) was observed in 2017 (20.14%) and 2020 (11.12%) years which might be indicating genotypic variability as highly influenced by nonheritable factor. Such oscillation happens for H² across years as described by Mistro et al. [38] for coffee.

From yield trait in Set-II, the highest GCV value observed in 2020 and 2018 years was 29.11% and 26.05, respectively (Table 4). Whereas, moderate GCV (10–20%) was manifested in 2016 and 2017 harvesting years. The PCV of set-II is high (PCV > 20%) across all seasons; but the highest was recorded in 2015 (PCV = 40.37%) and in 2019 (PCV = 36.90%) harvesting seasons. Concurring, moderate GCV and high PCV were obtained per year in yield for coffee as shown from past findings [21, 39]. This illustrates the highest percentage of variation attributed from nonhereditary parts for phenotypic expression in these seasons. The highest H² (>60%) was recorded in 2020 (H² = 67.93%); moderate heritability was revealed in 2018 and 2016 seasons which were 54.13% and 31.95%, respectively. Negative H² (−42.39%) was revealed in 2019 season from set-II due to negative genotypic variance that resulted from high environmental variance effect. From the combined yield data over six years mean analysis, the estimated GCV, PCV, and H² values of 15.07%, 20.02%, and 61.47% for set-I and 14.79%, 19.37%, and 58.88% for set-II, respectively, were recorded which clearly indicated the existence of moderate variability and heritability in both sets. These imply that it is very difficult to release high yielder via direct selection especially considering accessions of Yayo coffee landrace comprised in this study.

4.3. Annual Expected Genetic Gain and Response to Selection in Coffee Yield

The genetic gain percentage of mean (GAM (%)) together with heritability and GCV determines the supremacy of useable by breeder (additive) or dominance gene in desirable traits; they are vital in deciding breeding method to be followed. Moderate GAMs (10–20%), 12.26%, and 11.64% were recorded in 2018 and 2019 harvesting seasons, respectively; but in the other harvesting seasons, lower GAM (<10%) was observed, which ranged from 2.78 to 8.96% in set-I (Table 5). In set-II, high GAM (>20%), 21,39%, and 32.94% had been revealed in 2018 and 2020 years, respectively. Similarly, high per cycle GAM was estimated as reported by Atinafu et al. [23] and Kitila et al. [40] in coffee yield; moreover, it was reported that GAM fluctuation across seasons is an expected phenomenon Arabica coffee in yield [36]. In contrast, low GAM was recorded in other seasons. The annual GAM (%), GCV, and H² indicated that there is no harvesting season in which these three components’ high value recorded simultaneously for both sets (Tables 4 and 5). The mean yield of over six harvesting seasons of both sets clearly manifested moderate GAM (16.45% and 14.755 for set-I and set-II, respectively), H² (61.4% and 58.88%), and GCV (15.70% and 14.79% in set-I and set-II correspondingly) (Tables 4 and 5). Thus, this calls heterosis achieving breeding method to develop heterotic genotype in yield over already released commercial local variety. The annual GAM in 2017 (2.78%) (Set-I) and (4.11%) (set-II) were very low due to low phenotypic variance and heritability value which were resulted from low annual genotypic variance (Tables 4 and 5).

From annual response to selection (R) and selection differential (SD), high values were recorded in 2018 (196.76 and 471.90) and 2019 (223.40 and 601.08, respectively) harvesting seasons in set-I (Table 5). Also, high rate of genetic gain 49.19 and 44.68 had been shown, respectively, in the same harvesting seasons. In set-II, high annual R and SD were observed in 2018 (400.17 and 739.23, respectively) and 2020 (664.19 and 977.79), respectively. The highest annual SD 777.77 and 977.79 were recorded in set-I and set-II, respectively. This pointed out high performance of the top 15 and 10 high yielder relative to the whole population in 2020 harvesting years. Under set-II, during early harvesting season rate of genetic gain and response to selection showed negative; this elucidated that the high yielders selected at 16% (top 10) show less performance at earlier season when compared with the original population. In both set-I and set-II, the R, GAM%, SD, and rate of genetic gain were inconsistent across harvesting seasons which might be accounted to be the genotypic expression being highly conditioned by environmental factors.

From the results of R and SD, it is possible to decide the fate of the experiment in 2018 or 2019 for set-I and in 2018 harvesting season for set-II; the top 15 and 10 had expressed clearly their yield potential from set- I and set-II correspondingly (Tables 5 and 6). The most top 15 high yielders found between 1 and 15 rank in 4YRS and 5YRS except four accession (Y14, Y45, Y61, and Y4 which ranked 21^th, 20^th, 18^th, and 16^th, respectively) and one accession Y14 (which ranked 24^th) correspondingly (Table 6). Under set-II in 4YRS, except Y91, Y73, and Y96 (ranked 31^th, 21th, and 15^th) and in 5YRS except Y91, Y89, Y81, and Y105 (ranked 20^th, 19^th, 17^th, and 12^th, respectively) could perform top 10 like 6YRS. Thus, the average yield performance of over five years (5YRS) for set-I and over four years for Set-II was ideal harvesting seasons to decide the next breeding program. Such earlier decision is prominent for perennial crops such as coffee to save budget and time to proceed to the next research program.

4.4. Genetic Gain Trend and Response to Selection

The annual genetic gain and response to selection were directly proportional to selection differential (Figure 1). Also, the rate of genetic gain had been positively correlated with response to selection. Genetic gain and response to selection showed discrepancy across years in both sets (set-I and set-II) mainly due to the selection at 24% (top 15) and 16% done using over six years yield performance. However, the discrepancy in set-I was negligible; except in 2^nd harvesting seasons, both R and GA were showed increasing trend to the end (5th season). Also, the bienniality nature of the selected high yielders coffee genotypes might have affected the repeatability of genetic gain trend and response to selection. In agreement to this finding, inconsistent genetic gain had been reported by Mistro et al. [38] across eight harvesting seasons in Arabica coffee. The exponential increment of response to selection and rate of genetic gain recorded from 3^rd to 5^th and from 3^rd to 4^th coffee yield harvesting seasons in set-I and set-II, respectively could be considered as decisive parts for determining the fate of this experiment. The trends of genetic gain and response to selection start declined after 5^th season in set-I; in set-II, abating after 4^th and start inclining after 5^th which may be accounted to biennial problem and top 10% incongruity in yield performance across seasons. Thus, it was economical and momentous if promising line selection was planned at 4^th harvesting season for both sets with special care for those having high biennial nature. In agreement, Mistro et al. [38] reported that selection for yield is more effective at the earlier (before five years) than the late crop bearing stage.

4.5. Combined Analysis of Variance Over Years

Highly significant variation in yield performance was observed among coffee accessions and years in both sets (Table 7). Coffee genotypes in set-I sowed consistent performance across years; but inconsistent yield performance were observed in set-II coffee genotypes (Table 7). Similarly, from the pooled mean of eight years clean coffee highly significant differences among coffee genotypes and significant GY variation has been reported [38]. The GCV and PCV of pooled analysis authenticated the existence of moderate variability (10–20%) among genotypes. High H² (81.91%) and moderate (56.16%) were recorded in set-I and set-II, respectively. Likewise, moderate GAM (19.05%), H² (38.22%), and GCV (14.56%) were estimated in Arabica coffee from the mean of over six years yield [24]. In contrast to the current result, Akpertey et al. [37] reported low H² using pooled mean of over five years’ coffee yield. High H² in set-I was not due to high genotypic variability but low contribution of genotypes by year interaction (−26.40%). Also, 20.82% and 15.52% of GAM were recorded in set-I and set-II, respectively. Moderate genetic gain was reported from the combined over five years’ mean of coffee yield at 10% of selection level [37] and at 5% selection level [38, 41]. Additionally, high contribution of nonheritable part detected in set-I and set-II (106% and 69.98%, respectively). The pooled analysis of H², GCV, and GAM from both sets confirmed that it is difficult to develop best performing line/s in yield via direct selection unless heterotic development via hybridization program is followed. The top 15% high yielders from set-I recorded better response to selection (R), rate of genetic gain (∆GA), and selection differential (SD) than the top 10% from set-II.

4.6. Advanced Selection for the Next Breeding Program

The top 15 and 10 high yielder coffee genotypes were selected depending upon their yield performance over six years (Tables 8 and 9). The high yielders’ yield potential ranged from 1614.9 to 2312 kg·ha⁻¹ and 1493.8 to 1795 kg·ha⁻¹ from set-I and set-II, respectively. Coffee genotype Y61 showed high yield next to Y27 from set-I (Table 8); despite statistically nonsignificant Y27 and Y61 recorded 313.1 kg·ha⁻¹ and 102.6 kg·ha⁻1, respectively, over high yielder check 74110. Also, Y93 gave 125.3 kg·ha⁻¹ yield advantage over high yielder check 74112 (Table 9). These three genotypes were resistance to major coffee diseases like coffee berry disease (CBD) and coffee leaf rust (CLR) under field condition; Y27 and Y61 showed 0.03 and 5.08 in CBD and 10.42 and 7.67 in CLR correspondingly (Table 10). Also, Y93 was highly resistant to CBD and CLR under field. Additionally, they had shown high survival rate, uniform performance, and vigor in overall growth performance. Most top 15 and 10 high yielders showed resistance to CBD and CLR under field. The CBD reaction of these selected genotypes ranged from 0.03 to 12.22% and 0.00 to 14.17 in CLR which indicates that they are resistant to these major diseases.

Also, these coffee genotypes had been selected depending upon other desirable agronomic traits such as vigorous, growth habit (open, medium, and compact), stem habit (stiff and flexible), many number of primary branch and many primary branch with secondary branch, high leaf to crop ratio, different bean size, and very less to no necking (Table 10). These traits are breeders’ desirable traits which are very important in genetic improvement. Out of the top 15 and 10 genotypes, 20 genotypes which have contrast traits and divergent were selected and established in crossing block to utilize for further coffee genetic improvement purpose; the rests were included in Germplasm conservation.

5. Summary and Conclusion

Significantly variable performance in clean coffee yield was observed among 124 coffee accessions that were tested under two sets. In most harvesting seasons, accessions from set-I and set-II performed significantly different. Oscillation of genotypic coefficient of variation (GCV), phenotypic coefficient of variation (PCV), expected genetic gain (GAM), and heritability (H²) was observed across harvesting seasons. Also, genetic gain trend and response to selection showed discrepancy from year to years which resulted from environmental effect and might be from the bienniality nature of Arabica coffee. The combined analysis revealed moderate GCV (15.46 and 13.56%), GAM (20.8 and 15.52%), and H² (81.9 and 56.16%) which implies the difficulty of variety development via direct selection unless heteirotic achieving breeding method was followed. High H² (81.91%) in set-I was not due to high GCV (>20%) but resulted from low contribution of genotypic by environmental interaction in coffee phenotype expression. High genetic gain trend was observed commonly in set-I (49.19 kg·ha⁻¹) and set-II (100.04 kg·ha⁻¹) from the mean of four harvesting seasons; this pointed out that the selection is more effective and economical in earlier season.

The top 15 high yielders were selected from set-I and 10 from set-II based on pooled mean of yield over six years. Despite nonsignificant difference, the high yielder accessions Y27 and Y93 gave 313.1 kg·ha⁻¹ and 125.8 kg·ha⁻¹ yield advantage over nationally released coffee varieties from set-I and set-II, respectively. In addition to high yielding, the top 15 and 10 high yielder genotypes were resistance to major coffee diseases: coffee berry disease (CBD) and coffee leaf rust (CLR). Out of these, 20 genotypes which are divergent in morphological characteristics were selected and established in crossing block to utilize for genetic improvement via hybridization, and the rest were included in Germplasm conservation.

Data Availability

The data of this finding study are available with the corresponding author upon reasonable request.

Conflicts of Interest

The authors declare that there are no conflicts of interest.

Acknowledgments

This experiment was financially supported by Ethiopian Institute of Agricultural Research.

References

A. Carvalho, H. P. Medina Filho, L. C. Fazuoli, O. Guerreiro Filho, and M. N. A. Lima, “Genetic aspects of coffee,” Brazilian Journal of Genetics, vol. 14, pp. 135–183, 1991.
View at: Google Scholar
A. Carvalho and C. A. Krug, “Agentes de polinização da flor do cafeeiro (Coffea arabica L.),” Bragantia, vol. 9, no. 1–4, pp. 11–24, 1949.
View at: Publisher Site | Google Scholar
K. Taye and M. J. Burkhardt, “Plant composition and growth of wild Coffea arabica: implications for management and conservation of natural forest resources,” International Journal of Biodiversity and Conservation, vol. 3, pp. 131–141, 2011.
View at: Google Scholar
B. Bayetta, “Arabica coffee breeding for yield and resistance to coffee berry disease (C. kahawae Sp. nov),” University of London, London, UK, 2001, Dissertation.
View at: Google Scholar
J. Moat, J. Williams, S. Baena et al., “Resilience potential of the Ethiopian coffee sector under climate change,” Nature Plants, vol. 3, no. 7, Article ID 17081, 2017.
View at: Publisher Site | Google Scholar
J. Berthaud and A. Charrier, Genetic Resources of Coffea, In: R. J. Clarke, R. Macrae (ed.). Coffee: 4. Agronomy. Londres: Elsevier Applied Science, pp. 1–42, 1988.
M. B. P. D. Camargo, “The impact of climatic variability and climate change on arabic coffee crop in Brazil,” Bragantia, vol. 69, no. 1, pp. 239–247, 2010.
View at: Publisher Site | Google Scholar
L. Philippe, B. Benoít, and E. Hervé, “Breeding coffee (Coffea arabica L.) for sustainable production,” Breeding Plantation Tree Crops: Tropical Species, Springer, Berlin, Germany, pp. 525–543, 2009.
View at: Google Scholar
B. Bellachew and J. C. Sacko, “Coffee genetic resources under severe threat from genetic erosion in the centres of origin and diversity: an urgent need for conservation measures,” in Proceedings of the 22nd International Conference on Coffee Science (ASIC), pp. 1487–1496, Campinas, Brazil, September 2008.
View at: Google Scholar
F. M. Damatta and J. D. C. Ramalho, “Impacts of drought and temperature stress on coffee physiology and production: a review,” Brazilian Journal of Plant Physiology, vol. 18, no. 1, pp. 55–81, 2006.
View at: Publisher Site | Google Scholar
ICO, “Historical data on the global coffee trade,” 2016, https://www.ico.org/new-historical.asp.
View at: Google Scholar
A. P. Davis, T. W. Gole, S. Baena, and J. Moat, “The impact of climate change on indigenous arabica coffee (Coffea arabica): predicting future trends and identifying priorities,” PLoS One, vol. 7, no. 11, Article ID e47981, 2012.
View at: Publisher Site | Google Scholar
T. Demelash, “Achievements and prospects of coffee research in Ethiopia: a review,” International Journal of Research Studies in Agricultural Sciences, vol. 5, pp. 41–51, 2019.
View at: Google Scholar
M. Dawit and W. Zenebe, “Ethiopian coffee (Coffea arabica L.) germplasm genetic diversity: implication in current research achievement and breeding program: review,” Journal of Agricultural Research Pesticides and Biofertilizers, vol. 3, pp. 1–9, 2021.
View at: Google Scholar
M. Esser, Coffee, Climate Change and Adaption Strategies for German Coffee Producers, BoD–Books on Demand, Norderstedt, Germany, 2015.
N. Chand, S. R. Vishwakarma, O. P. Verma, and M. Kumar, “Worth of genetic parameters to sort out new elite barley lines over heterogeneous environments,” Barley Genetics Newsletter, vol. 38, pp. 10–13, 2008.
View at: Google Scholar
W. U. Haq, M. F. Malik, M. Rashid, M. Munir, and Z. Akram, “Evaluation and estimation of heritability and genetic advancement for yield related attributes in wheat lines,” Pakistan Journal of Botany, vol. 40, pp. 1699–1702, 2008.
View at: Google Scholar
A. R. Hallauer, M. J. Carena, and J. D. Miranda Filho, “Quantitative genetics in maize breeding,” Handbook of Plant Breeding, Springer Science & Business Media, Berlin, Germany, 2010.
View at: Google Scholar
J. E. Rutkoski, “Estimation of realized rates of genetic gain and indicators for breeding program assessment,” Crop Science, vol. 59, no. 3, pp. 981–993, 2019.
View at: Publisher Site | Google Scholar
S. Ganapathy, A. Nirmalakumari, and A. R. Muthiah, “Genetic variability and interrelationship analyses for economic traits in finger millet germplasm,” World Journal of Agriculture Science, vol. 7, pp. 185–188, 2011.
View at: Google Scholar
M. Dawit, M. Hussein, and A. Ashenafi, “Estimation of genetic variability, heritability and genetic advance of some wollega coffee (Coffea arabica L.) landrace in western Ethiopia using quantitative traits,” Journal of Plant Sciences, vol. 9, pp. 182–191, 2021.
View at: Google Scholar
J. C. H. Jane, N. Kahiu, W. M. James, and O. O. Chrispine, “Genetic variability, heritability and correlation of quantitative traits for arabusta coffee (C. arabica L. X tetraploid C. canephora pierre),” Journal of Plant Breeding and Crop Science, vol. 12, no. 1, pp. 50–57, 2020.
View at: Publisher Site | Google Scholar
G. Atinafu, H. Mohammed, and T. Kufa, “Genetic variability of sidama coffee (Coffea arabica L.) landrace for agro-morphological traits at awada, Southern Ethiopia,” Academic Research Journal of Agricultural Science and Research, vol. 5, pp. 263–275, 2017.
View at: Google Scholar
B. Lemi and A. Ashenafi, “Genetic variability, heritability and genetic advance for yield and yield components of limmu coffee (Coffea arabica L.) accessions in south western Ethiopia,” Middle-East Journal of Scientific Research, vol. 24, pp. 1913–1919, 2016.
View at: Google Scholar
IAR, “Recommended production technologies for coffee and associated crops,” Institute of Agricultural Research, Addis Ababa, Ethiopia, 1996.
View at: Google Scholar
IPGRI, “Descriptors for coffea sp. & psilanthus sp,” International Plant Genetic Resource Institute, Rome, Italy, 1996.
View at: Google Scholar
SAS, “Statistical analysis software,” SAS Institute Inc, Cary, NC, USA, 2013.
View at: Google Scholar
A. R. Dabholkar, Elements of Biometrical Genetics, Concept Publishing Company, New Delhi, India, 1999.
D. S. Falconer, Introduction to Quantitative Genetics, Iowa State University, Ames, IA, USA, 1960.
D. S. Falconer, Introduction to Quantitative Genetics, Pearson Education India, Noida, India, 1989.
M. Cooper, I. H. DeLacy, and K. E. Basford, “Relationships among analytical methods used to study genotypic adaptation in multi-environment trials,” Plant Adaptation and Crop Improvement, CABI, Wallingford, UK, pp. 193–224, 1996.
View at: Google Scholar
A. R. Haullauer, W. A. Russell, and K. R. Lamkey, “Corn breeding,” Corn and Corn Improvement, US Department of Agriculture, Washington, D.C., USA, pp. 463–564, 1988.
View at: Google Scholar
R. K. Singh and B. D. Chaudhary, Biometrical Methods in Quantitative Genetic Analysis, Kalyani Publishers, New Delhi, India, 1985.
Y. Masreshaw, G. Wosene, and T. Abush, “Genetic diversity of coffee (Coffea arabica L.) collections for morpho-agronomic traits in southwestern Ethiopia,” East African Journal of Sciences, vol. 15, pp. 141–154, 2021.
View at: Google Scholar
T. Abdulfeta, M. Hussien, and A. Ashenafin, “Genetic diversity analysis of tepi surroundings coffee (Coffea arabica L.) germplasm accessions using quantitative traits in Ethiopia,” International Journal of Agricultural Bioscience, vol. 7, pp. 76–80, 2018.
View at: Google Scholar
O. Kitila, S. Alamerew, T. Kufa, and W. Garedew, “Genetic diversity analysis of limmu coffee (Coffea arabica L.) collection using quantitative traits in Ethiopia,” International Journal of Agricultural Research, vol. 6, pp. 470–481, 2011.
View at: Publisher Site | Google Scholar
A. Akpertey, E. Anim-Kwapong, and A. Ofori, “Genetic variation among robusta coffee genotypes for growth and yield traits in Ghana,” Journal of Agricultural Science, vol. 10, no. 4, p. 138, 2018.
View at: Publisher Site | Google Scholar
J. C. Mistro, L. C. Fazuoli, P. Gonçalves, and O. Guerreiro Filho, “Estimates of genetic parameters and expected genetic gains with selection in robust coffee,” Cropp Breeding and Applied Biotechnology, vol. 4, no. 1, pp. 86–91, 2004.
View at: Publisher Site | Google Scholar
W. Getachew, A. Sentayehu, and K. Taye, “Genetic variability, heritability and genetic advance for quantitative traits in coffee (Coffea arabica L.) accessions in Ethiopia,” African Journal of Agricultural Research, vol. 12, no. 21, pp. 1824–1831, 2017.
View at: Publisher Site | Google Scholar
O. Kitila, S. Alamerew, T. Kufa, and W. Garedew, “Variability of quantitative traits in limmu coffee (Coffea arabica L.) in Ethiopia,” International Journal of Agricultural Research, vol. 6, pp. 482–493, 2011.
View at: Publisher Site | Google Scholar
Y. Masreshaw, G. Wosene, and T. Abush, “Estimate of genetic variability components via quantitative traits in coffee (Coffea arabica L.) germplasm in Ethiopia,” Academic Research Journal of Agricultural Science and Research, vol. 8, pp. 492–504, 2020.
View at: Google Scholar

Copyright

Copyright © 2022 Dawit Merga et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

PDF Download Citation

Download other formats

Order printed copies

Views

336

Downloads

479

Citations