Association of traits promoting resistance to the adzuki bean beetle, Callosobruchus chinensis (L.) (Coleoptera: Bruchidae) in pea genotypes grown at various soil fertility levels

Abstract

Adzuki bean beetle, Callosobruchus chinensis (L.) (Coleoptera: Bruchidae), is one of the most important pests of pea (Pisum sativum L.) crops in Ethiopia. The study focused on the association of resistance potential in the no-choice test of pea genotypes managed at different fertility levels and trait contributions. Based on the significance of fertility levels, genotypes were grouped into four, six, and five clusters, viz. Under neither rhizobium and phosphorus, rhizobium alone and rhizobium and phosphorus, respectively. Regardless of fertility levels, the inter-cluster distance (D²) values of the two potential clusters were highly significant (P < 0.01). At all fertility levels, the average performance of genotypes in each cluster for individual traits to infestation varied significantly. Genotype distribution patterns tended to group together into a small number of clusters. Eighty genotypes of the pea (Pisum sativum L. subsp. sativum and Pisum sativum L. subsp. abyssinicum A. Braun) were systematically managed under three fertility levels, and the first four principal components accounted for 94%, 92.3%, and 94.2% of the total variation. The primary trait that determines the resistance potential of pea genotypes is the trait susceptibility index (SI), which exhibits highly significant and adverse associations with critical traits such as the date of adult emergency and the percentage of seed coat, while exhibiting highly significant and favorable associations with the remaining traits at all fertility levels. The remaining characteristics showed highly significant positive or negative correlations within and particularly with the characteristics that determine resistance. Therefore, the cultivar "Adi" from "Pisum sativum L. subsp. sativum" had higher susceptibility compared to other genotypes, while the small-seeded pea genotypes "Pisum sativum L. subsp. abyssinicum A. Braun"; fpcoll-1/07, fpcoll-2/07, fpcoll-21/07, and fpcoll-43/07 were moderately resistant.

1. Introduction

Pea (Pisum sativum L.) is an annual herbaceous legume in the Fabaceae family that grows in the north, south, west, and central parts of the country, as well as pockets in the highlands and mid-highlands at altitudes ranging from 1800 to 3000 m.a.s.l. East Africa and West Asia have recently been identified as centers of origin and diversity of pea. In contrast, South Asia and the South and East Mediterranean sub-regions have been identified as secondary centers [1]. In Ethiopia, P. sativum is the dominant species, although the primitive forms, Pisum sativum var. abyssinicum A. Braun, locally named "dekoko," also exist, mainly in the northern part and Arsi Basin of the country [2].

Field pea is attacked by both field and storage insect pests, among which is the adzuki bean beetle, Callosobruchus chinensis (L.) (Coleoptera: Bruchidae) is one of the main biotic production constraints worldwide [3]. The adzuki bean beetle is distributed throughout the tropics and sub-tropics, and is a significant pest of stored legume seeds, including pea, Pisum sativum L. [[4], [5], [6]], chickpeas, Cicer arietinum L. [7,8], faba bean, Vicia faba L. [4], mungbean, Vigna radiata L. [9], common bean, Phaseolus vulgaris L. [3,10], cowpea, Vigna unguiculata L. [11], lentil, Lens culinaris Medik. [12], and soybean, Glycine max L. Merril. [13].

Infestation of legumes by the adzuki bean beetle occurs both in the field and during storage [[14], [15],16]. Although infestation and damage in the field are generally low, storage of infested seeds resulted in rapid insect multiplication and high grain damage in storage [16]. The adzuki bean beetle causes significant damage to stored beans, resulting in loss of nutrients and quantity and decreased germination rates [4]. The level of infestation may vary depending on the conditions of the crop and storage. An infestation of the adzuki bean beetle results in quantitative physical weight loss, qualitative declines caused by contaminants or biochemical changes, loss of seed viability as well as losses in nutritional value and economic value, making affected crops unfit for human consumption [17].

As with other storage insect pests, the management of adzuki bean beetles relied mainly on chemical insecticides. However, chemical insecticides can cause adverse effects on human and environmental health. Furthermore, the cost of chemical insecticides is unaffordable for most smallholder farmers in developing countries [14,18]. This highlights the need to develop a safe and affordable alternative option for controlling the adzuki bean beetle. The development of resistant varieties may be an option for the management of the adzuki bean beetle, mainly for smallholder farmers in developing countries [7,19,20].

Various efforts have been made to screen food legumes for Bruchid resistance (see, e.g. Refs. [4,5,14,[21], [22], [23], [24], [25]], and Tesfaye et al. [6] evaluated cultivated peas, and Esen et al. [5] cultivated and wild species of Pisum accessions for resistance to adzuki bean beetles.

According to Smith [26], during the evaluation of plants for resistance to insects, environmental factors such as soil nutrients have a significant effect on the expression of host plant resistance to insect pests. Altieri and Nicholls [27] observed that this could have a significant effect on the susceptibility of host plants to insect pests. Various studies also demonstrated that soil nutrients applied to the plant affect the expression of plant resistance to insect pests (see, e.g., Refs. [[27], [28], [29], [30], [31]].

As host seed cultivars significantly differ in their susceptibility levels to insect attack, Lambrides and Imrie [32] reported encouraging results for this issue. This frequently leads various researchers to screen host seed resistance to develop resistant varieties, which are crop plants or varieties that can naturally prevent, delay, or overcome pest infestations [19,22]. On the contrary, although the infestation of adzuki bean beetles has occasionally increased, the current situation regarding fertilizer use (inorganic) for the production of peas in Ethiopia is common (personal observation). However, despite this, there is no known reported evidence on the natural associations of traits that impact the potential for resistance in the field.

2. Materials and methods

2.1. Description of the study site

The experiment was carried out in Kulumsa (08^o00'02"N 39^o09'11"E, and altitude of 2210 m.a.s.l.) and Melkasa Agricultural Research Center (08^o24'N 39^o21'"E, and altitude of 1550 m.a.s.l.) during the main cropping season of 2017–2018. The minimum and maximum temperatures of Kulumsa Agricultural Research Center is 10 °C and 22.4 °C, respectively, and rain falls of 811 mm, and the relative humidity is 60.6%. The minimum and maximum temperature at the Melkasa Agricultural Research Center is 14 °C and 28.4 °C, respectively with 763 mm of rain fell on average, and 43% relative humidity [33].

2.2. Experimental materials

In this study, 80 genotypes were used (Annex 1). The primary production areas provided 43 collections of pea accessions (Pisum sativum L. subsp. abyssinicum A. Braun) (South Tigray and North Wollo). Twenty-six accessions were obtained from the International Center for Agricultural Research in Dry Areas (ICARDA). Two accessions and nine improved varieties/breeding lines were obtained from the Ethiopian Institute of Agricultural Research (EIAR).

2.3. Experimental field layout and management

The seeds of the pea genotype were grown at three different levels of inorganic soil fertility during the main cropping season 2017–18: neither rhizobium nor phosphorus, rhizobium alone and phosphorus plus rhizobium. Eighty seeds were evenly distributed among the four rows of genotype (4 m × 0.8 m) plots. The spacing of the rows and plants were 20 cm and 5 cm, respectively. According to the advice of Negash and Mulualem [34], 32 g of triple superphosphate (TSP) per plot (3.2 m²) of land was applied as phosphorus. Based on the recommendation of Menagesha Bio-Fertilizer Manufacturer PLC, which calls for 500 g/ha of inoculums, or roughly 0.16 g for 3.2 m² of inoculums, or 320 seeds per plot, an effective isolate of Rhizobium was inoculated. All agronomic procedures were carried out according to its suggestions. To eliminate any infestation prior to storage (eggs, larvae and adult bruchids), harvested seeds from each genotype were manually cleaned from foreign materials, adjusted to 9–10% moisture contents, and disinfected in a deep freeze at about −20 °C for a month before the study [22,35].

2.4. Mass-rearing of the insect

Mass-rearing of adzuki bean beetle was carried out at the Kulumsa and Melkassa Agricultural Research Center, Entomology Laboratory. The procedures are based on the recommendation of Keneni et al. [22] of the susceptible chickpea variety ‘Shasho’. The beetles were introduced to each 4 kg of seeds of the susceptible variety and kept at ambient temperature and relative humidity for seven days to allow oviposition. The parent insects were sieved out after seven days. Then the new progeny was used for re-culturing and kept again at optimum condition within the susceptible variety and removed after seven days. This re-culturing was continued and after enough new emerging insects was obtained, i.e. 1-2 day-old adult, unsexed insects were used for the different experiments.

2.5. Experimental design and infestation

In the experiment, no alternative conditions for adzuki bean beetle infestation were used, which was carried out in a laboratory setting with room temperature and relative humidity using a randomized complete block design with three replications. The harvested seeds was cleaned, stored in a cold room for a month, and cleared of any insect eggs. Per experimental unit, 200 seeds of each genotype were distributed (a 250 ml plastic jar of 250 ml; 6 cm × 7 cm). Each jar is regarded as a separate experimental unit. Jars with different pea genotypes were randomly assigned for each block. 1–2 day old, 14 unsexed adults adzuki bean beetles were randomly chosen and placed in each jar after being collected from the maintained culture. This insect had a 1:1 male-to-female ratio [36]; it is assumed that each jar contained seven males and seven females, for a total of 14 insects in each jar [22,35]. Hill [37] described serration antennae as a parameter for sex identification. After the introduction, we kept the adults in the jars for seven days before removing them. Every day, the plastic jars holding the seeds were checked to see if the first offspring had emerged. The first progeny is taken out of the jar after the emergence process is completed to assess the degree of attack and damage sustained by the first progeny. Using a thermo hygrometer, the room's temperature and relative humidity were measured every day until the end of the experiment to detect daily variations.

2.6. Data collected

The following information was collected from each pot with respect to its insect-based characteristics, including the total number of seeds and insect-based traits. Eggs laid in total (TNE): From the fourth day of infestation to the fourteenth day of infestation, the total number of eggs laid on the surface of the seeds of each genotype was counted each day, and records were taken for each treatment. Days to adult emergence: Beginning on the 20th day of the infestation and continuing until the adult emerged on the 32nd day, the number of days needed to reach adulthood was kept track of daily. Adults that emerged in numbers (NAE): Every day from 22 to 32 days, the total number of adults of each genotype was counted. The susceptibility index (SI) determined using Eq. (1) as described by Dobie [38]:

$$\mathrm{S}\mathrm{I}=\frac{\mathrm{L}\mathrm{o}\mathrm{g}\mathrm{Y}}{\mathrm{T}}\times 100$$ (1)

Where T is the estimated number of developmental periods (days) from the middle of the ovipositon period to the 50% emergence of the F1 progeny, SI is the susceptibility index and Y is the number of F1 emerging adults. According to Mensah [39], the values of the susceptibility indices were used to classify genotype susceptibility to bruchids into five categories.

• i.Genotypes with values ranging from 0.0 to 2.5 were considered resistant (R).
• ii.Genotypes with values ranging from 2.6 to 5.0 were considered moderately resistant (MR).
• iii.Genotypes with values ranging from 5.1 to 7.5 were considered moderately susceptible (MS).
• iv.Genotypes with values ranging from 7.6 to 10.0 were considered vulnerable (S).
• v.Genotypes with more than 10.0 were considered highly susceptible (HS).

The percentage of seed damage (PSD):

The percentage of damage of each genotype was calculated by separating healthy grains (without holes) from the sieved samples and using the formula described by Khattak et al. [40] Eq. (2).

$$\mathrm{P}\mathrm{e}\mathrm{r}\mathrm{c}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{a}\mathrm{g}\mathrm{e}\mathrm{o}\mathrm{f}\mathrm{s}\mathrm{e}\mathrm{e}\mathrm{d}\mathrm{d}\mathrm{a}\mathrm{m}\mathrm{a}\mathrm{g}\mathrm{e}=\frac{\mathrm{N}\mathrm{d}\mathrm{s}}{\mathrm{T}\mathrm{n}\mathrm{s}}\times 100$$ (2)

Where Nds is the number of damaged seeds and Tns is the total number of seeds.

Percentage of adult recovery (PAR): The ratio of the number of adults that emerged to the number of eggs laid on the surface of the seeds.

TSW (thousand seed weight in grams): Clean grain samples were taken from each genotype after adjusting the standard moisture content (10%), and 1000 grain seeds were counted from each sample grown under the same conditions.

The proportion of seed coat by weight in percent (PSC): The seed coat weight was calculated as a percentage of the total seed weight for the same genotypes grown under the same conditions.

Seed weight loss in grams (SWL): Seeds were separated into damaged and undamaged categories, and weight loss was adjusted to account for 10% moisture content. The seeds were counted and weighed, both damaged and undamaged. Seed weight loss was calculated using Eq. (3) following the formula [41]:

$$\frac{\left(\mathrm{W}\mathrm{u}*\mathrm{N}\mathrm{d}\right)-\left(\mathrm{W}\mathrm{d}*\mathrm{N}\mathrm{u}\right)*100}{\mathrm{W}\mathrm{u}\left(\mathrm{Nd}+\mathrm{N}\mathrm{u}\right)}$$ (3)

Where, Wu is the weight of undamaged seeds, Nu is the number of undamaged seeds, Wd is the weight of damaged seeds, and Nd is the number of damaged seeds.

2.7. Data analysis

2.7.1. Analysis of variance (ANOVA)

Analysis of variance (ANOVA) was performed using SAS version 9.3 statistical software (SAS Institute, 2010) to quantify genotypic differences and genetic variation in response to adzuki bean beetle infestation in peas based on quantitatively collected data from both locations for all parameters. The F-max method, which is based on the ratio of the larger mean square of error (MSE) from the separate analysis of variance to the smaller mean square of error, was used to test the homogeneity of error variance using methods of combined analysis of variance, as shown in Eq. (4) [42].

$$\mathrm{F}-\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}=\frac{\mathrm{L}\mathrm{a}\mathrm{r}\mathrm{g}\mathrm{e}\mathrm{r}\mathrm{M}\mathrm{S}\mathrm{E}}{\mathrm{S}\mathrm{m}\mathrm{a}\mathrm{l}\mathrm{l}\mathrm{e}\mathrm{r}\mathrm{M}\mathrm{S}\mathrm{E}}$$ (4)

According to Gomez and Gomez [43], the error variance means is regarded as homogeneous if the larger error means that the square was not three times larger than the smaller error means square. Tukey's test was used to determine the mean separation at the 5% and 1% probability levels.

The following model was used to perform separate analyses of variance to calculate the total variation among genotypes:

Pijkf = + (b/l) ik + gj + lk + f + (gl) j + (glf) fjk + efijk

The presence of here Pijkf is short for "phenotypic observations of genotype j in block i." (at location k and fertility level f) I = 1 … B, j = 1 … G, k = 1 … L, and f = 1 … f) where G, L, B, and f are, respectively, the number of genotypes, location, block, and fertility. Grand mean, block i's influence within location k, genotype j's influence within location k, fertility f's influence within location k, genotype j's influence within fertility f's influence within genotype j's influence within genotype j's influence within genotype j's influence within genotype j's influence within genotype j's influence within genotype j's influence within genotype j's influence within genotype j'. The F-test was used to determine whether there were significant differences between genotypes, locations, fertility rates, and their interaction.

2.7.2. Estimation of genotypic correlations

In Sing and Chaudhary [44] formulas, the correlation coefficients between all possible trait combinations at the genotypic (r_g) level were estimated. Using Proccandisc's SAS software version 9.3 [45], correlation analyses were performed to determine traits that were correlated to produce genotypic correlation as described in Eq. (5):

$$\mathrm{G}\mathrm{e}\mathrm{n}\mathrm{o}\mathrm{t}\mathrm{y}\mathrm{p}\mathrm{i}\mathrm{c}\mathrm{c}\mathrm{o}\mathrm{r}\mathrm{r}\mathrm{e}\mathrm{l}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{s}{\mathrm{r}}_{\mathrm{x}\mathrm{y}}\left(\mathrm{G}\right)=\frac{{\mathrm{c}\mathrm{o}\mathrm{v}}_{g_{xy}}}{\sqrt{{\sigma }^2}{g}_x{\sigma }^2{g}_y}$$ (5)

2.7.3. Cluster and distance analyses

When grouping sets of genotypes into homogeneous classes, procDiscrim in SAS software version 9.3 (SAS Institute, 2010) estimated cluster analysis and genetic differences between clusters. To create the dendrograms, MINITAB statistical software [46] was used. The genotypes were clustered using the average linkage method into various homogeneous clusters. Based on the Euclidean distance, which is a measure of dissimilarity, dendrograms were created (the distance). Mahalanobis’ D²statistics [47] were also used to estimate the genetic separation between clusters.

Dij² = (xi - xj)' cov-1(xi - xj), where Dij² is the distance between cases I and j, xi and xj are vectors of variable values for cases I and j, and cov-1 is the variance-covariance matrix pooled within groups.

2.7.4. Principal component analysis

In order to resolve the total variation of a set of traits into linear, independent composite traits that successively maximize variability in the data, it is a crucial breeding tool that breeders frequently use [48]. It was also analyzed to identify traits that contribute a significant portion of the total variation between genotypes [49]. The general formula for calculating the first component score in the principal component analysis is as follows (Eq. (6)):

C₁ = b_i1(X₁) + b_i2+ … …. b_ip (X_P) (6)

Where C1 is the principal component 1 score for the subject (the first component extracted), b1p is the principal component's regression coefficient (or weight) for the observed variable p. 1 X_p = the subject's result for the variable that was observed.

3. Results

3.1. Cluster analysis for quantitative traits

The genotypes were divided into six, five and four clusters based on measured traits when grown with rhizobium alone, rhizobium and phosphorus together, or neither of the two in their respective orders (Fig. 1). The least number of clusters is found in genotypes grown with no application (neither rhizobium nor phosphorus). On the contrary, genotypes grown with other applications (rhizobium inoculation, rhizobium, and phosphorus) include more clusters.

Fig. 1.

Dendrograms of 80 pea genotypes grown at various fertility levels; no rhizobium nor phosphorous (F1), only rhizobium (F2), and mixtures of rhizobium and phosphorous (F3). Overall, the graph above demonstrated that the level of genetic expression of the term "genetic diversity" with a different color would be the best level of soil fertility.

A total of 28 (P. sativum L. subsp. abyssinicum A. Braun) landrace collections in total, which have relatively small seeds and thicker seed coats, were grouped into the first cluster (C1), and 15 (P. sativum L. subsp. abyssinicum A. Braun) landrace collections, which have a relatively better seed size and less thick seed coats than the first cluster, were grouped under cluster (C2) per zero application. Furthermore, a total of 20 (five improved varieties, 14 newly introduced ICARDA materials, and one crossed pipeline) have medium to large seeds and a decreased seed coat weight compared to the genotypes in clusters (C1) and (C2) that were grouped under cluster (C) (C3). Finally, at this level of fertilization, 17 (four improved varieties, twelve newly introduced ICARDA materials, and one pipeline) with larger seed sizes and thin seed coats were grouped in a cluster (C4).

Second, when the genotypes were grown with inoculation with rhizobium, a total of 19 (P. sativum L. subsp. abyssinicum A. Braun) landrace collections with small seeds, lower seed weight, and thicker seed coats were grouped in the first cluster (C1), and 24 (P. sativum L. subsp. abyssinicum A. Braun) landrace collections with relatively better seed sizes and thinner seed coats than genotypes grouped under cluster (C1) were (C2). Approximately 11 (one improved variety, nine newly introduced ICARDA materials, and one pipeline) with medium to large seeds and relatively thinner seed coats than (C1) and (C2) were categorized as part of the cluster (C3). More than 21 per large seeded and inverse seed coat thickness were grouped in cluster (C4), including six improved varieties and 15 ICARDA-introduced materials. Likewise, a single improved variety with a large seed and a thin seed coat was grouped with others in a cluster (C5). Last but not least, at this level of fertility, a total of four genotypes (one improved variety, two introduced ICARDA materials and one pipeline) with very large seed sizes and lower seed coat weight were grouped in the cluster (C6).

Although the genotypes are grown under rhizobium and phosphorus, a total of 32 (P. sativum L. subsp. abyssinicum A. Braun) landrace collections with small seed sizes and thicker seed coats are grouped in clusters (C1). Over 11 (P. sativum L. subsp. abyssinicum A. Braun) landrace collections that have nearly better seed size and less seed coat weight than genotypes in cluster C1 were grouped in the cluster (C2). These have medium to large seed sizes of six-improved, ten-introduced from ICARDA, and one-pipeline per a total of 17- pea genotypes were grouped in clusters (C3). Once again, 13 genotypes with better seed size than those in previous clusters were grouped in cluster (C4) (three improved, nine introduced materials from ICARDA, and one pipeline). Finally, seven (introduced materials from ICARDA) seeds with very large seed sizes and lighter seed coat weights were grouped in the cluster (C5).

3.2. Distance analysis

Based on their genetic distances (D²), the response traits to adzuki bean beetle infestation of eighty (80) pea genotypes grown under three soil fertility levels were grouped into different clusters (Table 1). Regardless of all soil fertility levels, inter-cluster distance (D²) values were found to be highly significant (P < 0.01) depending on the potential pairs of clusters. The average inter-cluster D² was found for genotypes grown with neither rhizobium nor phosphorus, only rhizobium, and a mixture of rhizobium and phosphorus conditions in that order, ranging from 27.23 (between clusters C1 and C2) to 175.78 (between clusters C2 and C4), and 23.79 (between clusters C1 and C2) to 237.16 (between clusters C1 and C6) and 27.09 (between clusters C1 and C2) to 215.69 (between clusters C1 and C4). From these findings, it can be concluded that fertility levels perform similarly to the maximum genetic divergence (D² = 237.16), and six number clusters were observed when genotypes were grown with rhizobium inoculation. Secondly, when the genotypes were grown with the application of rhizobium and phosphorus, the maximum genetic divergence (D² = 215.69) and five number clusters were also observed. On the other hand, when the genotypes were grown without the use of rhizobium or phosphorus, less genetic divergence (D² = 175.78) with four number clusters was found.

Table 1.

A couple of squared distances (D2) between 6 clusters of 80 different pea genotypes that were grown at various soil fertility levels during an adzuki bean beetle infestation.

Clusters
Neither rhizobium nor phosphorus
Clusters	C1	C2	C3	C4	C5	C6
C1	0				–	–
C2	27.23**	0			–	–
C3	124.73**	139.68**	0		–	–
C4	157.06**	175.78**	43.66**	0	–	–
With rhizobium
C1	0
C2	23.79**	0
C3	145.75**	133.81**	0
C4	187.02**	173.39**	43.32**	0
C5	95.14**	76.62**	100.21**	127.56**	0
C6	237.16**	223.04**	93.03**	50.32**	172.91**	0
With rhizobium and phosphorus
C1	0					–
C2	27.09**	0				–
C3	172.62**	152.87**	0			–
C4	215.69**	194.95**	44.36**	0		–
C5	147.30**	130.29**	35.25**	77.58**	0	–

** = significant at (p < 0.01), Chi-square (χ²) = 21.666.

3.3. Average performances of genotypes in different clusters

The average response of the genotypes in each cluster to adzuki bean beetle infestation for specific traits at all fertility levels revealed significant variation (Table 2). Due to this, the clusters (C1 and C2) were distinguished by the lowest mean number of eggs laid (83.6 and 61.7), adults emerged (47.2 and 32.8), percent of seed weight loss (13.1 and 9.1), percent of seed damage (18.2 and 12.8), percent of adult recovery (56.7 and 54.5), and mean the number of holes per seed (0.24 and 0.16) in which both clusters are only included P. sativum L. subsp.abyssinicum A. Braun genotypes. However, all improved and introduced genotypes and pipelines were grouped into clusters (C3 and C4) that had the highest average number of eggs laid (111.5 and 139.9), adults emerged (86.7 and 113.9), percentage of seed weight loss (24.2 and 31.8), percentage of seed damage (37.8 and 48.9), percentage of adult recovery (75.8 and 81.1) and the average number of holes per seed (0.43 and 0.51).

Table 2.

Differences among clusters of 80 pea genotypes grown under different soil fertility levels and mean performance of ten response traits to infestation by adzuki bean beetles.

Clusters
Character	C1	C2	C3	C4	C5	C6	Grand mean
Neither rhizobium nor phosphorus
NE	83.6	61.7	111.5	139.9	–	–	98.4
DAE	28.7	28.3	26.7	26.2	–	–	27.6
NA	47.2	32.8	86.7	113.9	–	–	68.6
PSWL	13.1	9.1	24.2	31.8	–	–	19.1
PSD	18.2	12.8	37.8	48.9	–	–	28.6
PSC	17.9	18.1	10.2	10.4	–	–	14.4
TSW	68.2	66.1	179.0	190.9	–	–	121.6
SI	5.8	5.4	7.3	8.0	–	–	6.6
AR	56.7	54.5	75.8	81.1	–	–	66.2
MNHPS	0.24	0.16	0.43	0.57	–	–	0.34
With rhizobium
NE	73.6	95.6	134.0	139.6	165.4	191.8	120.1
DAE	28.7	29.2	23.8	26.1	25.4	23.8	27.3
NA	47.9	55.7	110.0	117.2	143.9	170.9	91.9
PSWL	13.3	15.5	30.6	32.7	40.1	47.6	25.6
PSD	21.1	23.9	51.0	55.5	66.0	76.4	41.6
PSC	18.0	17.7	24.8	10.6	10.1	9.5	14.5
TSW	68.1	66.9	70.1	168.8	186.7	217.2	120.2
SI	5.8	5.9	8.0	8.1	8.7	9.5	7.2
AR	66.1	63.0	82.6	84.1	87.5	89.7	74.7
MNHPS	0.24	0.28	0.55	0.6	0.72	0.86	0.46
With rhizobium and phosphorus
NE	73.4	87.0	135	161.2	185.9	–	117.7
DAE	27.2	26.7	28.1	29.8	30.6	–	28.3
NA	39.2	57.7	101	123.5	151.6	–	83.4
PSWL	10.9	16.1	28.4	34.4	42.2	–	23.2
PSD	19.9	25.8	50.7	60.2	71.3	–	40.3
PSC	18.4	17.2	10.7	10.2	10.3	–	14.5
TSW	67.1	68.9	177.3	177.8	196.9	–	121.6
SI	5.9	6.7	7.0	7.1	7.3	–	6.6
AR	55.1	67.0	75.1	76.9	81	–	67.3
MNHPS	0.2	0.3	0.5	0.6	0.8	–	0.4

NE stands for "total egg number," DAE for "days to adult emergence," and NA for "number of adults emerged." PSD stands for percentage of seed damage, PSC for percentage of seed coat weight, TSW for thousand seed weight, SI for susceptibility index, AR for adult recovery, and MNHPS for mean number of holes per seed.

In inoculation with rhizobium, the genotypes were divided into six clusters, with clusters (C1 and C2) containing the genotypes with the lowest average mean performance of the traits that were taken into account: the percentage of seed weight loss (13.3 and 15.5), the percentage of seed damage (21.1 and 23.9), the percentage of adult recovery (66.1 and 63) and the average number of holes per seed (0.24 and 0.28). While a cluster (C6) contains the genotype with the highest average mean egg laid performance (191.8), adult emerged (170.9), percentage of seed weight loss (47.6), percentage of seed damage (76.4), percentage of adult recovery (89.7), and average number of holes per seed (0.86). Additional clusters (C3, C4, and C5) contain genotypes with medium to large seeds that showed intermediate mean values for both traits related to seeds and insects.

For rhizobium plus phosphorus application, Cluster (C1 and C2) had the least average mean performance of the egg laid (73.4 and 87), the adult emerged (39.2 and 57.7), the percentage of seed weight loss (10.9 and 16.1), the percentage of seed damage (19.9 and 25.8), the percentage of adult recovery (55.1 and 67), and the mean number of holes per seed (0.2 and 0.3), whereas in a cluster (C5) includes the genotype by the superior (0.8). For the remaining clusters, both the insect and seed traits have shown intermediate mean performance.

3.4. Genetic diversity based on the geographic pattern of genotypes

The genotype distribution patterns concentrated on a small number of clusters (Table 3). Materials collected from north Wollo and south Tigray was mainly clustered in C1 and C2 at all fertility levels based on their origin. Except for a few materials grouped into clusters (C3 and C4), most improved varieties introduced materials and pipelines are concentrated in clusters (C5 and C6). Number wise, 43 collected genotypes (North Wolloand South Tigray) were concentrated in clusters (C1) and (C2), while 37 genotypes (improved varieties, pipelines, and introduced materials) were grouped primarily under (C3), (C4), and (C5) except for one improved variety, two introduced materials, and one pipeline, which were grouped in (C6) when the genotypes were grown with neither rhizobium nor phosphorus.

Table 3.

Pattern of clustering based on the average performance of ten traits in 80 genotypes of pea grown at different soil fertility levels and subjected to adzuki bean beetle attack.

Collection Place	Genotypes	Different clusters under each fertility levels
		Neither rhizobium nor phosphorus				With rhizobium						With rhizobium and phosphorus
		C1	C2	C3	C4	C1	C2	C3	C4	C5	C6	C1	C2	C3	C4	C5	C6
North Wollo	17	12	5	–	–	8	9	–	–	–	–	13	4	–	–	–	–
South Tigray	26	16	10	–	–	11	15	–	–	–	–	19	7	–	–	–	–
Improved Varieties	9	–	–	5	4	–	–	1	6	1	1	–	–	6	3	–	–
Introduced Materials	26	–	–	14	12	–	–	9	15	–	2	–	–	10	9	7	–
Advanced lines	2	–	–	1	1	–	–	1		–	1	–	–	1	1	–	–
Total	80

3.5. Principal component analysis

The 80 field pea genotypes managed under three fertility levels (with neither rhizobium nor phosphorus, with rhizobium, and once again with rhizobium and phosphorus) with their respective orders showed that the first four principal components explained 94%, 92.3%, and 94.2% of total variation (Table 4).

Table 4.

Results of a principal component analysis of ten traits of 80 genotypes of peas grown at various soil fertility levels in relation to adzuki bean beetle infestation.

Neither rhizobium nor phosphorus					With rhizobium				With rhizobium and phosphorus
Principal components					Principal components				Principal components
Traits	PC1	PC2	PC3	PC4	PC1	PC2	PC3	PC4	PC1	PC2	PC3	PC4
NE	−0.001	−0.025	0.994	0.095	−0.004	0.269	0.962	−0.022	−0.01	−0.048	0.996	0.015
DAE	0.283	0.497	0.033	−0.358	0.320	0.128	−0.023	0.244	0.332	0.125	0.046	−0.423
NA	0.350	−0.185	−0.048	0.392	0.341	−0.221	0.065	0.407	0.372	−0.141	−0.020	−0.285
AR	−0.241	0.552	−0.049	0.638	0.353	0.099	−0.015	0.257	0.126	0.731	0.015	−0.045
MNHPS	0.379	−0.034	−0.021	0.263	0.349	−0.319	0.083	−0.289	0.390	−0.061	−0.006	0.039
TSW	0.379	−0.034	−0.021	0.263	0.367	0.179	−.043	0.049	0.390	−0.061	−0.006	0.039
PSC	0.369	−0.089	−0.037	0.244	0.378	0.089	−0.024	0.043	0.377	0.069	−0.012	0.063
PDS	0.309	0.307	0.048	−0.093	0.378	0.089	−0.024	0.043	0.307	0.153	0.014	0.823
PSWL	0.304	0.448	0.023	−0.221	−0.237	0.634	−0.171	0.446	0.368	0.106	0.039	−0.207
SI	0.355	−0.324	0.011	−0.210	0.236	0.546	−0.169	−0.646	0.237	−0.616	−0.036	0.093
Eigen Values	6.61	1.27	1.01	0.65	6.71	1.02	0.94	0.60	6.28	1.63	1.00	0.49
Proportion (%)	66.3	11.7	10.2	6.6	66.98	11.0	8.9	6.1	62.8	16.3	10.2	4.9
Cumulative (%)	66.1	77.8	86.9	94	66.9	78.6	88.1	92.3	62.8	79.2	89.2	94.2

NE stands for "total egg number," DAE for "days to adult emergence," and NA for "number of adults emerged." PSD stands for percentage of seed damage, PSC for percentage of seed coat weight, TSW for thousand seed weight, SI for susceptibility index, AR for adult recovery, and MNHPS for mean number of holes per seed.

The first principal component (PC1), which accounted for 66.1% of the total variation when genotypes were grown without rhizobium or phosphorus, had high positive and negative values for DAE (0.284), NA (0.35), MNHPS (0.39), PSD (0.309), PSWL (0.304), SI (0.356), and AR (−0.241), but the least negative values for NE (−0.0012). While the second principal component (PC2), which accounted for 11.7% of the variation, had high positive and negative values for AR (0.552), SI (0.325), DAE (0.497), PSD (0.307), PSWL (0.449) and NA (−0.185), the least negative values were for NE (−0.025), MNHPS (−0.035), TSW (−0.034) and PSC (−0.089). Again, the third principal component (PC3) accounts for 10.2% of the total variation and contains the least negative values for the following traits: NA (−0.0481), AR (−0.0496), MNHPS (−0.021), TSW (−0.021), and PSC (−0.037). High and low positive values are found for most traits, including NE (0.994) and DAE (0.0337), PSD (0.0484), PSWL (0.0238), and SI (0.0117). Similarly to the third principal component, the fourth principal component (PC4) accounts for 6.6% of total variation and has high positive and negative values for some traits, including NA (0.392), AR (0.638), MNHPS (0.263), TSW (0.263), PSC (0.244) and DAE (−0.358), PSWL (−0.221), and SI (−0.21). However, the traits NE (0.096) and PSD (−0.093) were explained by the least favorable and favorable values.

For genotypes grown under rhizobium conditions, the first principal component accounts for 66.9% of the total variation, with high positive and negative values for DAE (0.3205), NA(0.341), AR(0.353), MNHPS(0.349), PSD(0.378), TSW(0.367), PSC(0.378), SI(0.236), and PSWL(-0.237), but the least negative value for NE (−0.119). The second principal component (PC2) contributed 11% of the total variation and had the highest positive and negative values for NE (0.269), PSWL (0.634), SI (0.546), NA (−0.221) and MNHPS (−0.319), but the lowest positive values for DAE (0.128), AR (0.099), TSW (0.179), PSC (0.089) and PSD (0.089). The third principal component (PC3), which represented 8.9% of the variation, also had high and least positive values for some traits, including NE (0.962), NA (0.065) and MNHPS (0.083). However, it also had the least negative values for traits such as DAE (−0.023), AR (−0.015), TSW (−0.043), PSC (−0.024), PSWL (−0.171), and SI (−0.169). Furthermore, the fourth principal component (PC4), which made up 6.1% of the total, had low positive and negative weights for the traits TSW (0.049), PSC (0.043), PSD (0.045) and NE(-0.022), but high positive and negative weights for the traits DAE (0.244), NA (0.407), AR (0.257), PSWL (0.446), MNHPS (−0.289) and SI (−0.646). On the other hand, each principal component (PC1, PC2, PC3 and PC4) shares 62.8%, 16.3%, 10.2% and 4.9% of a total variation with its corresponding orders when the genotypes are grown under rhizobium and phosphorus conditions. Principal component one (PC1), which accounted for all variation, had high positive weights for the traits DAE (0.332), NA (0.372), MNHPS (0.39), TSW (0.391), PSC (0.377), PDS (0.307), PSWL (0.368), and SI (0.237), but low positive and negative weights for AR (0.126) and NE (−0.01). Similarly to PC1, PC2 shows that the total variation has a high positive and negative weight for the AR (0.731) and SI (−0.616), traits but a low positive and negative weight for DAE (0.126), PSC (0.069), PSD (0.153), PSWL (0.106), NE (−0.048), NA (0.142), MNHPS (−0.062) and TSW (−0.061). While the third principal component (PC3) contains the total variation, the least negative values are found for the following traits. NA (−0.02), MNHPS (−0.006), TSW (−0.007), PSC (−0.012) and SI (−0.036), while high and low positive values are found for the majority of traits: NE (0.997), DAE (0.046), AR (0.015), PSD (0.014), PSWL (0.039), and SI (−0.036). However, the principal component (PC4) accounted for all variations and had the highest positive and negative values for the traits PSD (0.823) and DAE (−0.423), NA (−0.286), and PSWL (−0.207), while the traits NE (0.015), MNHPS (0.039), TSW (0.041), PSC (0.063), SI (0.093) and AR (−0.045) had the lowest positive and negative values.

3.6. Interrelationships between traits

The relationships between the traits showed some instances of strong associations in both positive and negative directions. Except for the percentage of seed coat weight, which showed a significant negative association, all traits were significant and positively correlated with the overall number of adults that emerged (Table 5). Days to adult emergence have a significant and inverse relationship with almost all traits, except the percentage of seed coat weight, which has a significant and inverse relationship. Except for the date of adult emergency and the percentage of seed coat weight, which had a highly significant negative association, all traits showed a highly significant positive correlation with thousand seed weight (seed size). However, the date of adult emergency showed a significant positive association with a percentage of seed coat. On the contrary, all other traits showed a highly significant negative correlation.

Table 5.

Combined correlation coefficients for ten traits of 80 genotypes of peas tested under all possible fertility-level conditions for adzuki bean beetle infestation.

Traits NE DAE NA PSWL PSD PSC TSW SI AR MNHPS
Neither rhizobium nor phosphorus
NE	0	−0.49**	0.95**	0.94**	0.94**	−0.75**	0.81**	0.87**	0.10NS	0.94**
DAE			−0.55**	−0.54**	−0.56*	0.51**	−0.56**	−0.51*	−0.14NS	−0.53**
NA				0.81**	0.99**	−0.82**	0.88*	0.92**	0.88**	0.99**
PSWL					0.98**	−0.83**	0.86**	0.93**	0.89**	0.91**
PSD						−0.84**	0.89**	0.90**	0.83**	0.96**
PSC							−0.91**	−0.78**	−0.75*	−0.82**
TSW								0.84**	0.58*	0.88**
SI									0.85**	0.92**
AR										0.87**
MNHPS
With rhizobium
NE	0	−0.75**	0.98**	0.92**	0.97**	−0.82**	0.92**	0.94**	0.7NS	0.98**
DAE			−0.76**	−0.75**	−0.76**	0.57**	−0.70**	−0.90**	−0.52*	−0.76**
NA				0.93**	0.93**	−0.83**	0.95**	0.96**	0.94**	0.97**
PSWL					0.95**	−0.81**	0.90**	0.95**	0.92**	0.96**
PSD						−0.85**	0.94**	0.93**	0.93**	0.99**
PSC							−0.90**	−0.72**	−0.37*	−0.83*
TSW								0.90**	0.40*	0.94**
SI									0.92**	0.96**
AR										0.93**
MNHPS
With rhizobium and phosphor
NE	0	−0.70**	0.81**	0.92**	0.98**	−0.86**	0.94**	0.75**	0.18NS	0.88**
DAE			−0.24NS	−0.70**	−0.70**	0.63**	−0.68**	−0.64*	−0.15NS	−0.69**
NA				0.89**	0.97**	−0.87**	0.91**	0.77**	0.87**	0.91**
PSWL					0.96**	−0.74**	0.78**	0.71**	0.66**	0.85**
PSD						−0.82**	0.96**	0.76**	0.86**	0.91**
PSC							−0.91**	−0.69**	−0.73**	−0.57**
TSW								0.72**	0.81**	0.94**
SI									0.76**	0.76**
AR										0.87**
MNHPS

NE stands for "total egg number," DAE for "days to adult emergence," and NA for "number of adults emerged." PSD stands for percentage of seed damage, PSC for percentage of seed coat weight, TSW for thousand seed weight, SI for susceptibility index, AR for adult recovery, and MNHPS for mean number of holes per seed.

4. Discussion

The response of pea genotypes to adzuki bean beetle infestation varied according to the traits measured at all fertility levels, resulting in distinct clusters. As a result, distinct clusters of pea genotype diversity were generated at all fertility levels. The level of infestation of adzuki bean beetles would be significantly reduced by these clusters, primarily based on the seed characteristics of each field pea genotype, particularly the size and thickness of the seed coat. The same result was also reported in chickpeas [31]. However, soil fertility levels also varied in response among tested genotypes, which may be directly proportional to lower fertility levels that affected the genetic expression of the genotypes' resistance to adzuki bean beetles and made it clear that higher soil fertility may contribute to the better genetic expression of all genotypes in terms of genetic diversity. Argaye [35] reported that the increase in the frequency of clusters may be related to the combined application of rhizobium and phosphorus to crops such as chickpeas. However, Jarso [2] found that there were high cluster differences when chickpea genotypes were grown under saturated and minimal water stress conditions. Furthermore, it was noted that phosphorus-grown chickpea genotypes had more clusters than their counterparts grown without phosphorus in terms of number [22]. The genetic distance matrix among pea genotypes from two geographical locations (North Wollo and South Tigray), one introduction source ICARDA, and the existing improved varieties and pipelines was managed at different fertility levels used to construct the cluster in which the possible pairs of clusters, inter-cluster distance (D²) values, were found to be highly significant (P < 0.01) regardless of all soil fertility levels. This indicates that the genotypes are divergent even when the magnitudes of the traits considered vary. Genotypes that coexist within a cluster are believed to be more closely related to each other than other clusters, even though the number and types of genotypes vary from cluster to cluster. That may be the case because, as another study demonstrates, a favorable test environment increases both the level of gene expression and genetic diversity for multiple traits while negatively affecting the expression of the genotypes' genetic potential for the traits under consideration. In a study involving the same insect and chickpeas, comparable results were reported [35]. The genotype grown under (rhizobium inoculation) with a genetic distance (D²) = 237.16 between clusters C1 and C6 receives immediate priority for selecting the parental line for hybridization based on this test indicator. The parental line grown in rhizobium and phosphorus genotypes also yields a genetic distance (D² = 215.69) between genotypes C1 and C4 as a binary alternative. The fundamental premise to be noted above is that these highly divergent clusters would have caused significant genetic recombination and divergence in the cross-breeding of advantageous offspring. That is why authors like Chahal and Gosal [49] concentrated on genetic differentiation for multiple traits without considering parental selection without genotype-specific effects for specific traits of interest like disease resistance and desired agronomic and quality performance, focusing on the failure.

Based on the average mean performance of the measured traits in clusters at each fertility level, the genotypes were more genetically divergent and expressed. Separately, among the genotypes considered, small seeds with a high proportion of seed coat weight genotypes performed worse on average and clustered in (C1) at each fertility level, with susceptibility index values that were moderately resistant to adzuki bean beetles. On the contrary, genotypes with large seeds have a lower percentage of seed coat weight, which may account for their higher average mean performance record for the traits listed above that indicated susceptibility to adzuki bean beetles compared to C1 and C2. For the remaining clusters (C3 and C4), genotypes with known characteristics showed better average mean results for the traits under consideration and indicated that they were only moderately susceptible to adzuki bean beetles. The infestation status of all large seeded field pea genotypes by adzuki bean beetles in this study showed a high average mean performance of the traits considered except for the proportion of seed coat weight at all fertility levels, which may have resulted in a more susceptible crop and vice versa for all small seeded field pea genotypes. Furthermore, the genotypes of small to medium seed chickpeas with thick seed coats had moderate resistance to adzuki bean beetles [35]. According to Sarwar [50], larger seeds resulted in more insect damage, and genotypes with soft, smooth seed coats, white seeds, and larger seeds were more susceptible to insect pests [22].

Depending on the geographic origin of the genotypes, each cluster may contain genotypes of the same origin or different origins. Most of the field pea genotypes in this study were concentrated in a small number of clusters, which may suggest that there is only limited genetic diversity for traits that affect how the plant responds to an infestation of adzuki bean beetles. The collections from two different locations, southern Tigray and northern Wollo, were almost all clustered together, which may have been caused by the genetic flow of the crop between the neighboring areas. Related findings on the responses of various chickpea genotypes to adzuki bean beetle infestation were also reported by Argaye [35] and Keneni et al. [22]. Breeders should also take care when selecting traits that result in divergent genotypes to create modern varieties. Unless these varieties have a propensity to be divided into a few clusters due to a shared genetic ancestry [51]. Since sample traits frequently exhibit varying degrees of inter-correlation, not all principal components are required to provide a comprehensive summary of the data. Given that multiple traits were involved in explaining the variation among the genotypes, this research also shows a significant share for the first four principal components that represented a sizable amount of diversity between the genotypes investigated. A high and positive eigenvector for a particular trait typically indicated a positive correlation between that trait and the given PC, while a high and negative eigenvector indicated a negative correlation between the trait and the given PC [2]. While the variables with eigenvector of large absolute magnitude (close to unity) reflect a strong influence, those of small magnitude (near zero) reflect little influence for a particular variable provided that the first principal component has a great role on the variation [49]. Similarly, the traits individually contributed small effects to the variation in a given PC and hence the differentiation of the accessions into different clusters was somewhat dictated by the cumulative effects of several traits; however, traits with relatively greater positive weights of eigenvectors in a given PC those, breeding efforts may need to simultaneously focus on genetic manipulation of these traits to reduce infestation and seed damage levels by adzuki bean beetle [22].

Most of the measured traits showed a strong positive and negative association, with a positive significant relationship between the number of adults and the number of eggs, the number of adults and the percent weight loss, the number of adults and the number of holes, the number of eggs and the percent weight loss, and the number of eggs and holes showing that an increase in the number of eggs and adults led to greater weight loss and a greater number of holes in grains. Similar findings were also reported by Aslam et al. [52] and Argaye [35]. For some traits, the percentage of adult recovery from battles showed a significant positive association across the board, but in contrast, the date of adult emergency and the percentage of seed coat weight showed a significant negative association. This denotes an increase in the total number of adults that emerged, the average number of holes in each seed, the percentage of damaged seeds, and the index of susceptibility as a direct result of the percentage of adult recovery. Along with these traits, a trait to improve the genotype for insect resistance would be seed traits such as seed coats, as well as insect traits such as an adult, emergency time, and adult traits. However, there were positive correlations, but no statistically significant differences between the total number of eggs and the recovery of the adult. In support of this, Argaye [35] also reported relationships of a similar nature. However, bruchid beetles thrive in genotypes with few eggs per seed, especially females, who use a variety of tactile, chemical and physical signals to help them avoid intense larval competition [53]. There is an inverse correlation between the quantity of adzuki bean beetle eggs and the recovery rate. There is an inverse relationship between the number of adzuki bean beetle eggs and their rate of recovery. However, the negatively correlated results showed fewer eggs, adults, holes, thousand seed weight, and percentage of damaged seeds, weight loss, and susceptibility index that stemmed. Furthermore, the results showed that the longer the days until adult emergence, the more resistant the genotypes would be to this attack by beetles. The findings of this study, therefore, agree with some scholars' reports [22,35].

The susceptibility index assumes that the more F1 progeny produced and the shorter the development period, the more susceptible the seeds would be [54]. In the current study, favorable temperature and relative humidity for these beetles led to nearly identical results, particularly at Melkasa. Even though the magnitude of the correlation coefficient was slightly altered, these results suggest that the findings of Redden and McGuire [55] and Keneni et al. [22] are more or less similar to this one. The thickness of the seed coat that allows insect larvae to penetrate the internal seeds may be the cause of the positive correlation between the date of adult emergency and the percentage of seed coat weight. The positive association resulted from the inverse relationships between seed size and seed coat thickness. This study report is in line of Argaye [35]. In parallel to this work, small seed legumes have less food for bruchid larvae to feed on, and seed size may limit to their ability to grow [56]. Again, Mei et al. [57] reported that seed size may also affect the oviposition preference of beetles and a strong correlation has been observed between bruchid resistance and small- or medium-sized seeds. The positive significant association between the percentage of seed damage and seed size indicates that large seeds are a factor that explains the susceptibility to bruchids [21]. One of the factors of resistance to bruchid species and predation is universally the preference for smaller seed selection in legumes [58]. Therefore, the hardiness and thickness of the seed coat of the investigated pea genotype, which act as deterrent traits for traits related to insects such as oviposition, adult development, and weight loss in those genotypes, as well as inverse associations, also compel the breeder to use independent selection. Other researchers have also reported similar results [35,59]. Overall, the negative correlation suggests that, as Shaheen et al. [59] reported, independent selection for each character will be effective for resistance to the beetle.

However, stability of trait performances is caused by linkage disequilibrium, pleiotropic gene actions, and epistatic effects of various genes, that is, the same sets of genes control each trait or there is no independent genetic control between the traits even though fertility level varies. As a result, the direction of the correlation coefficients between traits was completely similar under the three soil fertility levels for resistance to the adzuki bean beetle; however, with small magnitude changes (strong and weak association in both directions). Due to the stable performance of traits on a single fertility level being an indirect improvement of those traits under the other fertility levels, this can also rule out the possibility of genotype selection for growth with any fertility level. Due to the similarity of the trait in a different left environment, it is, therefore, better to select those traits at a single fertility level and, further, save time and resources. Both Argaye [35] and Keneni et al. [22] reported on research on chickpeas that was more or less similar. Both groups found a pattern of association of traits in chickpea genotypes grown with and without phosphorus.

5. Conclusions

In general, most of the measured characteristics showed strong positive and negative associations with a positive significant relationship. The direction of the correlation coefficients between the traits was completely similar at the three levels of soil fertility for the resistance of adzuki bean beetle, but with small changes in magnitude (strong and weak association in both directions). Therefore, due to the similarity of the trait in a different left environment, it is better to select these traits at a single fertility level and also save time and resources. Since the stable performance of traits at a single fertility level represents an indirect improvement of those traits at the other fertility levels, this may also preclude the possibility of genotype selection for growth at each fertility level. In particular, when there is negative correlation, it is advocated that independent selection for each trait is effective in resistance to beetle, and breeders should also be cautious in selecting traits that result in different genotypes to create modern cultivars. Another significant variability in the susceptibility index (preference for moderately resistant genotypes) was observed between field pea genotypes managed with different levels of fertility in the different environments, and this can be further considered as an opportunity for exploitation in the breeding program (Annex 2).

Declarations

Author contribution statement

Deressa Tefaye; Esayas Mendesil: Conceived and designed the experiments; Performed the experiments; Analyzed and interpreted the data; Wrote the paper.

Gemechu Keneni: Conceived and designed the experiments; Performed the experiments; Analyzed and interpreted the data; Contributed reagents, materials, analysis tools or data; Wrote the paper.

Funding statement

This research did not receive any specific grant from funding agencies in the public, commercial or not-for-profit sectors.

Data availability statement

Data included in article/supplementary material/referenced in article.

Declaration of interest’s statement

The authors declare no conflict of interest.

Additional information

Supplementary content related to this article has been published online at [URL].

Appendix A. Supplementary data

The following is the Supplementary data to this article.