Destructive fungal disease survey of common bean (Phaseolus vulgaris L.) rust (Uromyces appendiculatus) in Southern Ethiopia

Abstract

Common bean (Phaseolus vulgaris L.) is important legume crop world-wide and in Ethiopia for its multipurpose uses. Common bean rust, is the most destructive fungal disease that severely reduces bean yield. For years, rust appeared in a farmer's field in Southern Ethiopia; however, the disease's significance remains unclear. The research aimed to ascertain the distribution and intensity of common bean rust, as well as elucidate the association of biophysical parameters. The field survey was conducted in southern Ethiopia in 2022. Ninety percent of the 78 commonbean fields were affected by common bean rust. Mareko, Meskan, Duguna Fango, Damot Woide, and Demba Gofa had 100% of the fields affected, and Boricha had 90%. Damot Woide and Lanfuro had the highest and lowest mean rust incidence rates, respectively, 59.2% and 22.5%. Duguna Fango had the highest rust severity (35.5%), while Lanfuro had the lowest (13.5%). In the research areas, the biophysical factors, either alone or in combination, have a significant impact on the intensity of common bean rust. The current investigation verified the distributionand the association biophysical factors with common bean rust. In addition, the survey of the disease and the identification of factors should continue over time and space.

Introduction

The common bean (Phaseolus vulgaris L.) is a multipurpose legume crop that grows throughout the world in tropical, subtropical, and temperate agro-ecologies. It is a major food legume that is consumed as cooked and snacked beans in a variety of dishes around the world¹. Ethiopia, Kenya, and Uganda produce the most in Africa¹. During the 2019/2020 main cropping season, pulses covered more than 12% of Ethiopia's total grain crop area, with common beans accounting for 2.2%². It is a rich source of protein (seeds contain 20–25% protein) and micronutrients for small-scale farmers³. In addition to this; it is also important in providing fodder for feeding livestock and it contributes to soil fertility improvement through atmospheric nitrogen fixation during the cropping season³.

Regardless of its importance, several biotic and abiotic factors hinder the national mean common bean productivity which is less than 2 t ha^–1 as compared to the world mean productivity of 2.5–3.5 t ha^–12. Diseases, insect pests, drought, and low-input farming methods have all been identified as biotic and abiotic factors limiting common bean production and productivity⁴. In East Africa, yield loss caused by biotic (i.e. diseases) is greater than yield loss caused by abiotic (i.e. drought and infertility constraints)⁵. Uromyces appendiculatus, which causes bean rust (Fig. 1), is one of the main diseases affecting common beans. Common bean rust is a major barrier to bean production in Ethiopia. The diseases is present throughout a large geographic area. It can severely damage fields planted with susceptible common bean varieties during favorable growing seasons⁶. While the effect of bean rust on achievable yield varies depending on the cultivar, the year, and the location, in early disease onset conditions, yield loss on susceptible cultivars may reach 85%⁷. Often, individual field collections of spores of the bean rust pathogen contain several different virulence strains⁸. Different strains of the bean rust pathogen once characterized for their virulence spectrum by inoculating them on a set of bean rust differential cultivars are called races. Numerous races of U. appendiculatus from many bean-producing areas of the world have been reported⁹. The urediniospores of U. appendiculatus can survive as a source of inoculum for the next season when the humidity and temperature are favorable¹⁰. Climates that are colder, with high humidity and temperatures between 17 and 20 °C may exacerbate the infection¹¹. In addition, plant debris left over from the previous season can also be a source of the disease¹². Plants with pathogen infection show symptoms such as pale, elevated areas on the underside of the leaf, which grow larger over the course of six days and eventually form brownish uredinia¹³. Depending on the environmental conditions, the pathogen's uredinia can spread quickly between 10 and 22 days after infection (the vegetative-flowering stages)¹⁴. According to Liebenberg and Pretorius¹⁵ wind and other agents are the main ways that common bean rust spreads. Helfer¹⁶ reported that, the most important global drivers for the geographical and host plant range expansion and prevalence of rusts, are global plant trade, host plant genetic homogenization and the regular occurrence of conducive environmental conditions, especially the availability of moisture.

Fig. 1.

Rust affected bean plants in southern Ethiopia. Photo by Yisahake Tsegaye.

Despite, the existing fact, bean growers in the study areas are not well informed to manage the disease. Comprehensive information with regard to the pathosystem of common bean rust has not been well documented in southern parts of the country. Different cropping systems and production situations can influence disease occurrence, development and damage to crops and surveying is an important tool to studying these diversified conditions¹⁷. On the other hand, management of diseases requires an understanding of how such interrelated factors lead to epidemics¹⁸. It will be easier to identify the most crucial factors and concentrate efforts on creating an integrated and sustainable management package if it is assumed that disease intensity is correlated with cropping systems, crop combinations, planting times, field and disease management techniques, and other conditions during the production season¹⁹. However, there is a lack of information in southern Ethiopia regarding the frequency and severity of the disease as well as how cropping systems, field management techniques, environmental factors, and seasonal variations affect common bean rust. Furthermore, little is known about the pathosystem in South Ethiopia in particular and Ethiopia in general. It is unclear what factors affect the disease's progression and severity.

Therefore, the study was designed to meet the following objectives: to elucidate the association of common bean rust with biophysical factors and determine the prevalence and severity of the disease in southern Ethiopia.

Materials and methods

Description of study areas

Field surveys were conducted in major common bean production areas of Southern Ethiopia; namely Silte, Gurage, Wolaita, and Gofa Zones, and Boricha district of Sidama region (Fig. 2). The survey locations were purposively selected based on their production potentials. Some of the districts produce the crop twice annually depending on the rainfall distribution (using short season-rain that extends from March to May, and the main rainy season begins from late June, and end of August). The survey was carried out during the main rainy season from August to October in all areas of surveyed districts. The districts vary in their weather conditions (Table 1).

Fig. 2.

Map of areas surveyed for common bean rust disease during the 2022 main cropping season.

Table 1.

Characteristic features of common bean rust fields surveyed in southern Ethiopia.

Zone	District	Altitude (m.a.s.l.)	Temperature (°C)	Rainfall (mm)
Gurage	Mareko	1876	24	1203
	Meskan	2074	25.5	1010
Silte	Sankura	1876	19	1259
	Lanfuro	1987	27	825
Wolaita	D.Woide	2121	19.2	1962
	D.Fango	2026	21	1800
Gofa	D. Gofa	1500	22.5	1150
	Oyida	1650	20.5	1350
Sidama	Boricha	1866	19.8	1002

Source: National Meteorology Agency, Hawassa branch, Ethiopia.

D.Woide = Damot Woide, D.Fango = Duguna Fango, D.Gofa = Demba Gofa.

Sampling procedure and sampling units

From mid-September to mid-October 2023, a field survey covering common beans was conducted, covering the crop's development from late flowering to maturity. Each district's two kebeles were chosen based on their potential for production. Three to five farmers' common bean fields were chosen from each kebele (Kebele refers to farmers association) and examined for the presence of the common bean rust disease. A total of 78 fields were assessed during the survey. The common bean fields were randomly sampled at intervals of 5–7 km along main and feeder roads to kebeles using vehicle and motorbike odometers. Depending on the size of the field 2–5 quadrats were dropped at 1 m × 1 m (1 m²) and the plants were inspected for rust, and disease scores were recorded. Disease incidence and severity were assessed for every quadrat by moving across each field from one end to the other in an ‘X’ pattern. From each quadrat, the number of plants assessed and the number of plants with rust pustules and associated symptoms were counted.

Disease assessment

Surveys were conducted on common bean fields to determine the frequency, severity, and incidence of common bean rust. The percentage of bean fields afflicted with common bean rust disease relative to the total number of fields evaluated in each district is known as the prevalence. Typically, it is expressed as follows in percentage form:

$$\text{Common bean rust prevalence}(\%)=\left(\text{N}/\text{TN}\right)\ast 100$$

Where N = Number of fields with the disease per district TN = Total number of fields surveyed per district.

Disease incidence was recorded as the percentage of plants showing rust symptoms and signs (rust pustule) in each quadrat, and the averages of the number of quadrats in the respective field were calculated for each field.

$$\text{Disease Incidence}(\%)=\left(\text{Nq}/\text{Tn}\right)\ast 100$$

Where, Nq = Number of rust infected plants per quadrat Tn = Total number of plants assessed per quadrat.

Disease severity is defined as the proportion of plant area (leaves, stems, pods) affected with diseases from the total part of the tissue. It measures the amount of damage to the plant tissue and is assessed based on symptomatic lesions (pustules). Six to ten plants in every 2–5 quadrats were assessed for common bean rust. Rust disease severity was rated using the CIAT 1 to 9 scale following Van Schoonhoven and Pastor-Corrales²⁰, where 1–3 = resistant (no visible pustules to few pustules covering 2% of foliar area), 4–6 = intermediate (small pustules covering 5% foliar area to large pustules often surrounded by chlorotic halos covering 10% foliar area) and 7–9 = susceptible (large to very large pustules covering 25% foliar area).

Disease severity values were converted into Percentage Severity Index (PSI) for the analysis as suggested by Wheeler ²¹

$$\text{PSI}=\left(\text{Sn}/\text{NM}\right)\ast 100$$

Where Sn = Sum of numerical rating N = Number of plants rated M = Maximum score on the scale.

Biophysical factors assessment

Biophysical factors are biotic and abiotic variables that affect how the common bean rust epidemic spreads. To ascertain their correlation with disease incidence and severity, these variables were gathered and examined. In each field, Global Positioning System (GPS) instrument was used to express the geographical coordinates (latitude and longitude) as well as altitude. Field data recording sheet was used to report the crop growth stage (flowering, podding, or maturity), sowing date, sowing method (row or broadcasting), a variety grown (local or improved), fertilization (fertilizer applied or not fertilized), cropping system (sole or intercropped) and weed management (poorly managed, intermediate or well managed). Data collection was conducted by on-spot field observation and interview of growers during the assessment.

Data analysis

To explain the distribution and relative importance of common bean rust, data from field assessments were summarized using descriptive statistics in SPSS version 20. As explained by Yuen²² and Fininsa and Yuen¹⁸, disease incidence and severity were categorized into discrete groups of binomial qualitative data. Class boundaries and cut points were established by calculating the average of the severity and incidence of all evaluated fields. Therefore, in the survey, the binary variables for rust incidence (> 43 and ≤ 43%), and severity (≤ 25 and > 25%) of common bean rust were selected. To depict the bivariate distribution of the field in accordance with the categorization, a categorization table of disease incidence, severity, and independent variables was constructed (Table 2). Incidence and severity means were separated using Fisher’s protected Least Significant Difference (LSD) test at P < 0.05. The association of common bean rust incidence and severity with independent variables were analyzed using logistic regression model Yuen²²) with the SAS procedure of GENMOD SAS Institute Inc²³. The logistic regression was used to evaluate the importance of multiple independent variables that affect the response variable. The importance of the independent variables was evaluated twice in terms of their effect on incidence and severity. First, the association of all the independent variables with the incidence and severity was tested in a single-variable model. Second, the association of an independent variable with the incidence and severity was tested when entered first and last with all the other variables in the model. Lastly, those independent variables with significant association with the incidence and severity were added to a reduced multiple-variable model. The parameter estimates and their standard errors were analyzed using the GENMOD procedure both for the single and multiple models. The odds ratio was obtained by exponentiation of the parameter estimates for comparing the effect based on a reference point, which is interpreted here as the relative risks Yuen²².

Table 2.

Categorization of variables used in logistic regression analysis for common bean rust epidemics in 9 districts (n = 78) during the 2022 main cropping season.

Variables	Variable class	Prevalence	Sev		INC
			> 25	≤ 25	> 43	≤ 43
Districts	Mareko	100	6	3	7	2
	Meskan	100	6	3	8	1
	Sankura	75	3	5	1	7
	Lanfuro	66.67	0	6	1	5
	Duguna Fango	100	9	1	9	1
	Damot Woide	100	3	3	6	0
	Demba Gofa	100	1	9	6	4
	Oyida	70	1	9	3	7
	Boricha	90	4	6	7	3
Altitude (m.a.s.l)a	1500–2000	88.57	23	39	37	25
	> 2000	20.51	10	6	12	4
Variety	Local	62.82	26	23	36	13
	Improved	37.12	8	21	12	17
Fertilizer application	Not	28.20	16	6	19	3
	Applied	71.79	18	38	29	27
Weed managementb	Poor	43.58	10	24	27	7
	Intermediate	38.46	8	22	16	14
Sowing method	Broad casting	29.48	21	12	27	6
	Row	57.69	13	32	22	23
Cropping systemc	Sole	47.43	29	8	32	5
	Intercrop	52.56	5	36	16	25
Common bean growth staged	Pod formation	29.48	10	13	15	8
	Pod filling	34.62	13	14	16	11
	Physiological maturity	35.89	11	17	18	10
Plant population (m−2)e	Greater than 10	58.97	25	21	33	13
	Less than 10	42.02	9	23	15	17
Sowing date	June end–end of July	55.12	23	20	28	15
	1st August–end of August	44.87	11	24	20	15
Survey site	Farmers field	91.02	31	40	44	27
	FTC	3.84	2	1	2	1
	Trial	5.12	1	3	2	2
Farm size (ha)	Less than 0.25	67.94	26	27	33	20
	Greater than 0.25	32.05	8	17	15	10

Sev. = Severity; INC = Incidence.

^aAltitude ranged from 1500 to 2000 and > 2000 masl are considered as mid- and highland areas.

^bWeed management practices were recorded as poor (presence of high weed infestation), intermediate (few weeds present), and good (fields free of any weed infestation).

^cCropping system refers to planting only common beans as sole cropping and sowing common beans simultaneously with other crops (sorghum and maize) as intercropping.

^dGrowth stage referred to as pod formation when half of the plants in the quadrat are forming pod, pod filling is when plants in the quadrat start to pod development, and maturity is when the crop reaches its physiological maturity.

^ePlant population was determined in 1 m⁻² quadrat as highly dense (> 10 bean plants m⁻²) and sparsely populated (≤ 10 bean plants m⁻²).

FTC = Farmers Training Center.

Result and discussion

General characteristics of assessed common bean fields

The current study's survey of farmer fields found that about 91% of them were situated between 1500 and 2000 m above sea level (m.a.s.l.) (Table 1). Locals and improved bean variety were widely cultivated by farmers in the survey areas: Red wolaita, Hawassa-dume, and Nasir were improved varieties grown in the survey areas. The majority of fields (55.1%) were planted between the end of June and the end of July, while the remainder (44.9%) were planted between the beginning of August and the end of August. In the survey area, row planting was the most common sowing technique (57.7%). In the districts under study, NPS fertilizer accounted for 71.8% of the total fertilizer applied to common beans. More than half of the fields (52.6%) used intercropping. During the field surveys, the common bean growth stage was 35.9% at the physiological maturity stage, 34.6% during pod filling, and 29.5% during pod formation. Aside from having observed the disease in action, farmers and development agents didn't seem to know much about the plant disease and how to manage it. Furthermore, the farmers employ no management strategies, including the use of pesticides, other than weeding, which was also done inadvertently to control rust disease.

Common bean rust prevalence, incidence, and severity

In Southern Ethiopia's bean-growing regions and the Sidama Region's Boricha district, the common bean rust was widespread. Nevertheless, there are differences in prevalence between districts. The districts with the highest disease prevalence—Mareko, Meskan, Duguna Fango, Damot Woide, and Demba Gofa—were all 100%. Boricha came in second with a 90% prevalence of bean rust. In comparison, the prevalence of the disease was lower in Sankura (75%), Oyida (70%), and Lanfuro (66.7%) (Table 2). Survey districts also differed in common bean rust incidence and severity. The mean disease incidence was the highest at Damot woide (59.2%) followed by Duguna Fango and Mareko, which had mean incidence of 54.6% and 52.2%, respectively, (Table 2). On the other hand, the lowest rust incidence was observed at Lanfuro (22.5%) and Sankura (27.5%). The districts of Meskan, Demba Gofa, Boricha and Oyida had intermediate bean rust incidence ranging from 29 to 49%. Regarding the rust severity, the highest common bean rust severity was scored in Duguna Fango (35.5%), followed by Damot woide (32.5%), and Meskan (31.5%), whereas; the lowest rust severity was scored in Oyida (13.7%) and Lanfuro (13.5%) (Table 2). In addition, the severity of 28.8% in Mareko, 25.4% in Boricha, 22.4% in Demba Gofa and 20% in Sankura were also recorded.

The current survey suggested the variation in bean rust intensity as affected by different associated factors (Table 3). For instance, the highest rust incidence (52.6%), and the highest severity (32.2%) were recorded in fields located at an altitude of > 2000 masl. In contrast, the low mean rust incidence of 40.6% and severity of 23.2% were scored at an altitude range of 1500–2000 masl. High rust intensity at an altitude of > 2000 masl might be due to the prevalence of lower temperature and ample rainfall. Conversely, Singh et al.²⁴ reported that areas with low altitudes had a high bean rust incidence and severity which may be attributed to high rainfall and relative humidity that favors infection and development of bean rust disease. The rust disease intensity also slightly varied in surveyed locations in accordance with the varieties planted. Hence, relatively high mean disease incidence (47.2%) and severity (27.0%) were observed on fields planted with local variety followed by incidence (36.0%) and severity (21.6%) of fields planted with improved variety. This finding was consistent with the results of Menza et al.²⁵, who reported that locally used varieties were had higher susceptibility to a broader range of diseases than improved varieties based on germplasm aspects. The improved varieties were found to be resistant to pathogens that cause disease, even though the rates of disease infection varied. They were found to be within the same range of infection. This is because, in contrast to locally cultivated variety, whose genetic composition is not recently improved, the improved varieties genetic composition, which combines a variety of positive traits, may be responsible for their higher diseases resistance. The sowing date in the assessment areas starts from June and extends to August in the previous seasons. However, in 2022 main growing season due to the late start of rain the planting starts in the middle of June to the end of August, and most of the assessed fields were planted at the end of June to end of August. Consequently, the surveyed fields were categorized based on planting dates as those from the end of June to the end of July, and from the beginning of August to the end of August. The rust disease intensity score did not differ significantly with the sowing date in the presently assessed fields. In addition crop growth stage did not have much influence on rust intensity (Table 3). In contrast to this, rust incidence and severity significantly varied with the common beans density (population). The highest rust incidence of 48.4% and highest severity of 28.6% were recorded on fields with plant populations of more than ten plants per meter square (m⁻²); while the lowest rust incidence of 35.3% and severity of 19.9% were scored on fields with a plant population having less than ten plants per meter square (m⁻²). This might be attributed to less favourable microenvironment in lower plant density due to the presence of enough air circulation and low leaf wetness. The present discovery was supported by the findings of Biddle and Cattlin²⁶, who found that the faba bean’s dense population preserved a suitable humid microclimate that promoted the inoculum’s production and subsequent infection.

Table 3.

Mean incidence and severity of common bean rust (Uromyces appendiculatus) for different independent variables.

Variable	Variable class	Incidence (%)	Severity (%)
		Mean ± SE	Mean ± SE
District	Mareko	52.22 ± 3.24	28.88 ± 1.23
	Meskan	49.44 ± 1.94	31.51 ± 1.94
	Lanfuro	22.50 ± 7.50	13.50 ± 4.50
	Sankura	27.50 ± 6.34	20.00 ± 4.72
	Duguna Fango	54.61 ± 3.52	35.50 ± 2.41
	Damot Woide	59.17 ± 4.36	32.50 ± 3.59
	Demba Gofa	46.00 ± 2.45	22.44 ± 1.17
	Oyida	29.00 ± 6.66	13.70 ± 3.17
	Boricha	43.50 ± 5.17	25.44 ± 3.53
Altitude (m.a.s.l)	1500–2000	40.56 ± 2.25	23.15 ± 1.40
	> 2000	52.57 ± 3.59	32.19 ± 2.19
Survey site	Farmer field	43.33 ± 2.09	25.13 ± 1.31
	FTC	36.70 ± 18.60	22.20 ± 11.10
	Trial	42.50 ± 3.23	24.80 ± 4.58
Farm size	< 0.25	43.49 ± 2.41	26.27 ± 1.57
	> 0.25	42.04 ± 3.67	22.32 ± 2.05
Variety	Local	47.17 ± 2.23	27.04 ± 1.37
	Improved	36.03 ± 3.55	21.57 ± 2.39
Sowing date	June end–July end	43.17 ± 3.13	24.98 ± 1.84
	1st August–August end	42.86 ± 2.33	25.03 ± 1.71
Plant population(m-2)	> 10	48.39 ± 1.53	28.56 ± 1.05
	< 10	35.31 ± 4.02	19.89 ± 2.44
Growth stage	Pod formation	43.91 ± 3.51	26.21 ± 2.42
	Pod filling	45.00 ± 3.27	25.11 ± 1.99
	Maturity	40.23 ± 3.67	23.86 ± 2.26
Sowing method	Broad cast	50.18 ± 2.13	29.75 ± 1.35
	Row	37.78 ± 2.88	21.52 ± 1.79
Cropping system	Sole	52.73 ± 1.72	31.87 ± 1.13
	Intercropped	34.27 ± 2.88	18.81 ± 1.67
Fertilizer application	Applied	38.66 ± 2.40	22.36 ± 1.51
	Not	54.14 ± 2.38	31.72 ± 1.61
Weed management	Poor	49.74 ± 1.84	30.51 ± 1.27
	Intermediate	41.33 ± 3.01	23.55 ± 1.91
	Good	30.36 ± 6.86	14.76 ± 3.35

Among the common bean varieties used by growers, local (farmers’) variety (Woye, Zalle, Fara Acha) were more susceptible than improved varieties. Higher bean rust pressure was also associated with broadcasting than row planting. The intensity (incidence and severity) of the rust was also influenced by cropping system. Hence, the higher rust disease incidence (52.7%) and a severity of 31.9% were observed on sole cropped fields while the lowest rust disease incidence of 34.3% and a severity of 18.8% were scored on intercropped fields. This finding was in contrast with the finding of Wafula et al.²⁷, who reported no difference in rust disease intensity between common beans as a sole crop or intercrop. However, many previously reported findings showed the significance of intercropping on disease development. A Study by Hailu²⁸ confirmed the importance of intercropping in the reduction of anthracnose disease incidence and severity compared with sole cropping. The finding was also consistent with the finding of Boudreau et al.^29,30, who reported reduction in disease intensity by intercropping in 79% of studies involving fungal pathogens. Fertilizer application is also one of the factors that play a vital role on intensity of the bean rust. The highest rust incidence and severity were recorded in fields without fertilization than fields on which fertilizer was applied. Although inconsistent results are common in various pathosystems, balanced and adequate fertility for any crop reduces plant stress, improves physiological resistance and lower the risks of diseases Kruginsky et al.³¹. Similarly, Cunfer et al.³² indicated adequate supply of nutrition (NPK) to the crop reduced the wheat leaf rust and increased the yield of wheat.

The assessed common bean fields have different level of weed infestation. The highest rust disease intensity was observed on weed-infested fields followed by moderately weed-infested fields. In contrary, lower disease incidence of 30.4%, and the severity of 14.8% were observed on well (good) weed-managed fields. The present result was in agreement with the work of Yimer et al.³³, weeds reduced crop vigour as a result of intensive competition of weeds for available resources, leading host plants to both foliar and soil-borne disease predisposition. In addition, Palumbo³⁴ reported that weed species found in and around crop fields served as alternate hosts to many diseases and insect pests that can later infect and re-infest nearby crops.

Common bean production practices significantly influenced the intensity of bean rust in the surveyed areas (Table 4). Of the independent factors’ altitude, district, variety, sowing method, common bean plant density (m⁻²), cropping system, fertilizer application, and weed management showed significant variations in common bean rust intensity. Disease incidence and severity were very highly significantly (P < 0.001) affected by district, cropping system, and fertilizer application. Crop variety, plant density and weed management were highly significantly (P < 0.01) associated with the disease and altitude had significant (P < 0.05) on disease pressure. However, disease incidence and severity were not significantly affected by the sowing date and common bean growth stage. Most of the areas surveyed in the present study had favorable environmental conditions, particularly temperature (17–24 °C). In this particular survey, optimum rainfall was recorded in Duguna Fango, Damot Woide, Mareko, Meskan and Boricha than other surveyed areas, which could be responsible for the high rust epidemics in the areas. These findings were consistent with the work of Ogecha et al.³⁵, who stated that the environment is the main factor that influences the distribution of pulse biotic stressors. The conducive environmental conditions for bean rust development are cool to moderate temperatures (17–22 °C) concurrent with high humidity (> 90%) periods for at least 7–8 h interspersed with dryer, windy periods that favored spore dispersal³⁶.

Table 4.

Analysis of variance of independent factors on the incidence and severity.

Source of variation	df	Mean square
		Incidence (%)	Severity (%)
Altitude	1	1832.7*	1038.76**
District	8	1318.91***	536.84***
Variety	1	2257.1**	545.74*
Sowing date	1	1.83 ns	0.04 ns
Sowing method	1	2930.69**	1290.96**
Plant density (m−2)	1	3229.19**	1418.65***
Crop growth stage	2	169.45 ns	34.44 ns
Cropping system	1	6630.54***	3318.74***
Fertilizer application	1	3785.03***	1383.32***
Weed management	2	1932.40**	1282.00***

df: Degree of freedom; MS: mean square, values with *, ** and *** implies significant at P = 0.05, P < 0.01 and P < 0.001 respectively; ns = not significant.

Association of common bean rust incidence and severity with biophysical factors

In southern Ethiopia, the impact of independent variables on epidemics of common bean rust varies (Table 5). When added as a single variable to a logistic regression model, the independent variables such as district (P < 0.000), cropping system (P < 0.000), plant population (P < 0.02), sowing method (P < 0.000), variety (P < 0.01), weed management (P < 0.005), and fertilizer application (P < 0.007) significantly correlated with the incidence of common bean rust in the surveyed districts (Table 6). Most of the predictor variables lost their significant association with common bean rust epidemics, when entered last into the model. The independent variable district was the only predictor that maintained its significant association with bean rust incidence, when entered last to the model with other factors. District was strongly associated (χ² = 26.32 and 30.62, df = 8) with common bean rust incidence. The probability of the bean rust incidence exceeding 43% was 2.3 and 1.7 times higher in Meskan and Duguna Fango, respectively, compared to Boricha. The rust incidence occurrence on intercropped common bean fields was 2.9 times lower than the sole cropped common bean fields. The current result was consistent with the finding of Boudreau et al.^29,30, who reported reduction in disease intensity by intercropping in 79% of studies involving fungal pathogens.

Table 5.

Logistic regression model for common bean common bean rust incidence (%) and severity (%) and likelihood ratio test on independent variables in Southern Ethiopia during the 2022 main cropping season.

Independent variable	df	Common bean rust incidence LRTa				Common bean rust severity LRTa
		Type 1 analysis (VEF)		Type 3 analysis (VEL)		Type 1 analysis (VEF)		Type 3 analysis (VEL)
		DR	P > χ2	DR	P > χ2	DR	P > χ2	DR	P > χ2
District	8	26.32	< 0.000	30.62	< 0.000	27.79	< 0.000	22.42	0.004
Altitude	1	0.10	0.74	4.51	0.03	4.67	0.03	2.50	0.11
Sowing date	1	0.08	0.76	0.51	0.47	4.70	0.03	0.19	0.65
Cropping system	1	18.95	< 0.000	2.43	0.01	41.01	< 0.000	11.19	< 0.00
Plant (m−2)	1	5.21	0.02	41.01	0.087	6.30	0.01	0.19	0.66
Sowing method	1	11.19	< 0.000	2.43	0.12	11.19	< 0.000	2.31	0.12
Variety	1	5.90	0.01	0.96	0.32	5.73	0.01	0.79	0.37
Growth stage	2	0.54	0.76	5.65	0.05	0.90	0.63	0.52	0.77
Weed management	2	10.29	0.005	4.58	0.10	21.73	< 0.000	3.73	0.15
Fertilizer application	1	7.10	0.007	1.00	0.31	13.39	< 0.000	0.06	0.79

^aLRT = likelihood ratio test; VEF = variable entered first; VEL = variable entered last; DR = deviance reduction; P = probability of an χ² value exceeding the deviance reduction; χ² = chi-square; and df = degrees of freedom.

Table 6.

Analysis of deviance, natural logarithms of odds ratio and standard error of common bean rust incidence and likelihood ratio test on independent variables in reduced regression model.

Added variable	Residual deviancea	df	CBR incidence LRTb		Variable class	Est.logc	SEd	Odds ratio
			DR	P > χ2
Intercept	105.60		–	–		− 0.45	0.74	0.3
District	62.29	8	24.35	0.001	Mareko	− 0.27	1.23	0.5
					Meskan	− 0.07	1.35	2.3
					Lanfuro	0.31	1.43	0.05
					Sankura	0.65	1.55	0.01
					D. Fango	− 0.25	1.37	1.7
					D. Woide	0.31	1.43	0.1
					D. Gofa	− 0.09	1.07	0.4
					Oyida	0.27	1.11	0.1
					Boricha	*
Cropping system	86.65	1	18.95	< 0.000	Sole	0.46	0.87	2.9
					Intercropped	*

^aLRT = Likelihood ratio test; DR = deviance reduction; P = probability of χ² value exceeding the deviance reduction; CBR = common bean rust; SE = standard error.

All of the independent variables revealed a significant association with common bean rust severity except crop growth stage (P < 0.63), when evaluated as a single variable. The significance of association differs by the predictor. District, cropping system, sowing method, weed management and fertilizer application had the strongest association (P < 0.000) followed by plant population and variety (P < 0.01), while altitude and sowing date had relatively milder but still significant association (P < 0.03) with bean rust severity. However, among the independent variables district (P < 0.004), and cropping system (P < 0.000) sustained their association with common bean rust severity when entered last to the model with other independent variables. The assessed district (χ² = 27.79 and 22.42, df = 8) and cropping system (χ² = 41.01 and 11.19, df = 1) showed a strong influence on common bean rust severity. Concerning the severity, the high severity (> 25%) was 1.9 and 10.8 times at Meskan and Duguna Fango at p < 0.003. The probability of common bean rust severity exceeding 25% was 117.6 times higher in solely cropped common bean fields compared to intercropped common bean rust fields (p < 0.000) (Table 7). This result agreed with the work of Eshetu et al.³⁷, who reported that effect of different production and management practices as climate change resilient cultural management components in chocolate spot and rust management strategies.

Table 7.

Analysis of deviance, natural logarithms of odds ratio and standard error of common bean rust severity and likelihood ratio test on independent variables in reduced regression model.

Added variable	Residual deviance	df	CBR incidence LRTa		Variable class	Est.log	SE	Odds ratio
			DR	P > X2
Intercept	107.30		–	–		0.21	0.97	0.1
District	43.62	8	22.67	0.003	Mareko	− 0.45	1.65	0.7
					Meskan	− 0.16	1.42	1.9
					Lanfuro	0.20	2.04	0.1
					Sankura	0.32	1.74	0.1
					D. Fango	0.37	1.55	10.8
					D. Woide	0.24	1.79	0.1
					D. Gofa	0.62	1.84	0.01
					Oyida	0.50	1.91	0.03
					Boricha	*
Cropping system	66.29	1	41.01	< 0.000	Sole	0.67	1.24	117.6
					Intercropped	*

^aLRT = Likelihood ratio test; DR = Deviance reduction; P = Probability of χ² value exceeding the deviance reduction; CBR = common bean rust; SE = standard error.

Conclusion

The results of this study demonstrated that common bean rust disease was a serious and widespread disease in southern Ethiopian districts that grew common beans. The information from the survey data explained how various biophysical factors, either separately or in combination, are linked to epidemics of common bean rust. Districts and intercropping strongly correlated with the intensity of common bean rust in this study.

It is recommended that awareness about the rust disease, the conditions that favor and/or disfavor rust disease epidemics in growers’ fields should be provided to concerned bodies from the Agricultural office up to growers. In addition to further pathogen diversity studies that include more areas, disease assessment and the identification of factors linked to the disease should continue over time and space.

Author contributions

Y.T. composed concept, carried out the analysis and wrote the initial draft of the manuscript. A.C. provided insights in to the methodology, analysis and investigation and contributed to reviewing and editing the manuscript. Y.R. supervised the work and contributed to reviewing and editing the manuscript. All authors have approved the submitted version of the manuscript.

Data availability

The datasets generated or analyzed in the current study are available from the corresponding author upon reasonable request.

Competing interests

The authors declare no competing interests.