Survival of the bean anthracnose fungus (Colletotrichum lindemuthianum) on crop debris in Canada

Abstract

Anthracnose caused by the fungus Colletotrichum lindemuthianum (Sacc. & Magnus) Lams.-Scrib. has been a major constraint to dry bean (Phaseolus vulgaris L.) production in Manitoba and Ontario, as well as elsewhere in the world. There is a lack of consensus on the length of time the anthracnose fungus can survive on infected crop debris as well as the effectiveness of crop rotation to control the disease under temperate climatic conditions. A multi-year study was established at field sites near Morden, MB and Exeter, ON to examine the survival of C. lindemuthianum on seed, pod and stem tissue samples of infected beans on the soil surface and buried in the soil over different 2-year periods. The survival of the anthracnose fungus was influenced by location, type of infected tissue and burial in the soil. On average, C. lindemuthianum could persist for 18 months at Morden and about 9 months at Exeter, which was long enough to serve as a source of primary infection to subsequent crops of dry beans. At both field sites, the anthracnose fungus overwintered longer on infected stems and pods than it did on infected bean seed. The burial of infected bean tissue reduced the viability of C. lindemuthianum in all tissue types compared with samples that remained on the soil surface. A 3-year crop rotation study at Morden in which either a wheat, fallow or bean treatment were grown in the second year, confirmed that the anthracnose fungus could survive long enough under zero-tillage conditions to infect a bean crop in the third year of the study and cause considerable downgrading of the seed. This study demonstrated the importance of tillage and crop rotation in the integrated control of bean anthracnose.

Résumé

L’Anthracnose causée par le champignon Colletotrichum lindemuthianum (Sacc. & Magnus) Lams.-Scrib. est un obstacle majeur à la culture du haricot sec (Phaseolus vulgaris L.) au Manitoba, en Ontario et ailleurs dans le monde. Il n’y a pas de consensus sur la durée de survie du champignon anthracnose sur les débris de culture infectés, ni sur l’efficacité de la rotation des cultures pour contrôler la maladie dans des conditions climatiques tempérées. Une étude pluriannuelle a été établie sur le terrain près de Morden au Manitoba, et d’Exeter en Ontario, pour examiner la survie de C. lindemuthianum sur des échantillons de graines, de gousses et de tissus de tiges de haricots infectés à la surface du sol et enfouis dans le sol pendant une période de deux ans. La survie du champignon anthracnose était influencée par l’emplacement, le type de tissu infecté et l’enfouissement dans le sol. En moyenne, C. lindemuthianum pourrait persister 18 mois à Morden et environ neuf mois à Exeter, une période suffisamment longue pour constituer une source d’infection primaire pour les cultures ultérieures de haricot sec. Sur les deux sites le champignon anthracnose a hiverné plus longtemps sur les tiges et les gousses infectées que sur les semences de haricot infectées. L’enfouissement de tissus de haricot infectés a réduit la viabilité de C. lindemuthianum dans tous les types de tissus par rapport aux échantillons restant à la surface du sol. Une étude triennale de la rotation des cultures à Morden, dans laquelle un traitement au blé, à la jachère ou au haricot ont été cultivés au cours de la deuxième année, a confirmé que le champignon anthracnose pouvait survivre suffisamment longtemps dans des conditions de sol non cultivé pour infecter les plants de haricot au cours de la troisième année, et provoquer un déclassement considérable de la graine. Cette étude a démontré l’importance delà culture du sol et de la rotation des cultures dans la lutte intégrée contre l’anthracnose du haricot.

Keywords:

• anthracnose
• Colletotrichum lindemuthianum
• Phaseolus vulgaris
• survival

Mots-clés:

• Anthracnose
• Colletotrichum lindemuthianum
• Phaseolus vulgaris
• survie

Introduction

Bean anthracnose, caused by Colletotrichum lindemuthianum (Sacc. & Magnus) Lams.-Scrib., is an important disease in most of the bean (Phaseolus vulgaris L.) producing regions of the world (Schwartz et al., 2005). Disease symptoms can appear on all above-ground bean tissues. Typically, symptoms on leaves occur as brown to black lesions that run along the veins. Brown, sunken, ovoid lesions form on the stems and pods. Pod infection usually results in discolouration, and sometimes shrivelling, of the seed. Severe infection by C. lindemuthianum can sometimes eliminate most of the crop yield (Melotto et al., 2000), while mild infection often leads to lesions being formed on the seed by the end of the growing season, which results in downgrading of the seed and reduced economic returns to the producer (Gillard et al., 2012).

Seed-borne transmission of anthracnose is an important factor in the spread of the pathogen to new bean producing regions of the world, as well as between fields in a growing region and can result in the introduction of new races into a region (Tu, 1992; Conner et al., 2009). Survival of the anthracnose pathogen in the debris of infected dry bean crops has been reported (Ntahimpera et al., 1997), so that crop rotations of 2–3 years with non-host species is generally recommended as an important component in the integrated control of anthracnose (Dillard & Cobb, 1993; Schwartz et al., 2005).

Consensus does not exist regarding estimates of the length of time that C. lindemuthianum can survive on infected bean stubble or on crop debris buried in the soil (Tu, 1983; Dillard & Cobb, 1993; Ntahimpera et al., 1997). Differences in the estimated length of survival have been attributed to differences in environmental conditions and cultural practices as well as the methodology used to detect viable inoculum of C. lindemuthianum in field studies. Tu (1983) reported that in Ontario, the anthracnose fungus died out in infected bean debris that had been buried for less than 6 months, so bean debris was not considered to be a source of inoculum in the following spring. Field studies in New York State by Dillard & Cobb (1993) indicated that the pathogen survived on bean debris for up to 22 months and served as an important source of inoculum for subsequent bean crops. In Canada, infected crop debris is an important inoculum source for other diseases of pulse crops and is a crucial consideration in recommendations for disease control (Gossen, 2001; Gossen & Miller, 2004). This study was undertaken to examine the survival of the anthracnose fungus in different types of bean tissues on the soil surface and buried in the soil at field sites located in Manitoba and Ontario. A crop rotation study was also conducted to further evaluate the survival of the bean anthracnose fungus.

Materials and methods

Survival of the anthracnose fungus in crop debris

In October 2008, 2010 and 2016, 2-year studies on the survival of the anthracnose fungus were established at the Morden Research and Development Centre (49º11ʹN, 98º5ʹW) and at the Huron Research Station (43º32ʹN, 81º5ʹW) near Exeter, ON. Survival of C. lindemuthianum in the stems, pods and seeds was separately examined. The treatments in each experiment consisted of infected tissue samples that were either buried 15 cm below the soil surface or left on the soil surface. Each treatment was replicated four times and 32 samples of each type of tissue were prepared to allow the collection and assaying of the infected tissue on eight specific dates spaced at 3-month intervals over a 2-year period at each location. Plant material was separated into stem pieces, pod halves and seed. Stem pieces were trimmed to a 10 cm length. The pods were opened and separated. All plant material (pods, stems and seed) were visually inspected to confirm the presence of anthracnose symptoms. Freshly collected samples were plated on potato dextrose agar (PDA) to verify infection by C. lindemuthianum.

To facilitate the rapid retrieval of the tissue samples, they were placed in mesh pockets that were constructed of nylon mosquito netting (Huang, 1983). The netting was cut into pieces measuring 10 cm × 23 cm and was folded in half to make a 10 cm² envelope with a flap to cover the opening at one end. The sides were joined together with staples. For each 2-year cycle, the plant material used at each field site was collected from a field site at the Morden Research and Development Centre where inoculum of race 73 of C. lindemuthianum had previously been foliarly applied onto the canopy of susceptible bean cultivars. The plant material samples (20 infected stems in 10 cm lengths or 20 infected seeds or 10 infected pod halves) were placed inside the mesh envelopes, the opening was stapled closed and a plastic plot label was attached. The mesh envelopes were placed in 15-cm plastic pots containing soil. Half of the mesh envelopes were buried 15 cm deep in the plastic pots and the others were pinned to the soil surface with bamboo stakes, then the pots were buried up to the rim in a field at each of the two locations. Samples were placed in the field in October, after which four samples of each tissue type were collected at 3-month intervals throughout 2 years. Every 3 months the samples were retrieved, air dried and then stored at 3ºC until they were processed within a few days.

A filter paper assay was used to detect the presence of viable inoculum of the anthracnose fungus. Infected bean tissue was removed from the mesh envelopes and placed in 15-cm Petri plates containing two Whatman #3 filter papers soaked with 8 mL of sterile distilled water. The plates were sealed and placed in the dark at a temperature of 21ºC. At the same time, disease-free seeds of the susceptible pinto bean cultivar ‘AC Pintoba’ were planted in 10-cm diameter greenhouse pots containing Sunshine Media #3 (Sun Gro Horticulture Canada Ltd, Seba Beach, AB) at a rate of three seeds per pot and allowed to grow for 10 days. One pot of bean seedlings was grown for each sample of bean tissue that was assayed from both field sites. All the plants were grown under controlled environmental conditions at a temperature setting of 21ºC with a 16-h day and 8-h night photoperiod. After 10 days, the infected bean tissue was removed from the moist filter papers in the Petri plates and the tissue was rubbed onto the leaves of three 10-day-old plants in a single pot, which were sprayed with a fine mist of water and then placed in a sealed plastic bag to maintain a high humidity conducive for infection and returned to the growth cabinet. For each sampling date and location, the pots with the bean plants inoculated with bean tissue from the various treatments were arranged in a completely randomized design with four replications. After 10 days, the plants in each pot were individually rated for anthracnose severity on the inoculated leaves using a scale of 0–9 (Balardin et al., 1997). The results were summarized as the incidence of plants with symptoms of anthracnose. The sampling of the crop debris from a specific location ended when anthracnose symptoms were no longer observed on the assay plants for two consecutive assay periods.

Weather conditions

All meteorological data were obtained from weather stations located at Morden, MB and London, ON at: http://climate.weather.gc.ca/climate_data/daily_data_e.html.

Year-round weather information was not available from the Huron Research Station, so data from London, ON (Exeter region) was used, since it was the closest location (48 km apart) with year-round meteorological information. The mean daily temperatures and total precipitation were summarized on a monthly basis.

Crop rotation study

In 2006, a crop rotation experiment was established on a field site at the Morden Research and Development Centre. The site was located in a field with sandy loam soil in which beans had not been grown for several years and was more than 200 m away from the closest bean crop. The experiment consisted of three 3-year cropping sequences (i.e. bean/wheat/bean (BWB), bean/fallow/bean (BFB) and bean/bean/bean (BBB)), which were planted using a zero-tillage system. In year 1 (2006), trifluralin was applied at a rate of 864 g a.i. ha⁻¹ and a fertilizer blend at a rate of 90-28-11-11 kg ha⁻¹ of N, P, K and S, respectively, was broadcast and incorporated prior to seeding. The experiment was seeded on 29 May 2006. In each year of the study, a seeder with a double-disk opener with an 18 cm row spacing was used to plant bean seed at a rate of 70 seeds m⁻². In 2006, anthracnose infected seed of the pinto bean cultivar ‘AC Pintoba’ was seeded over the entire experimental site. On 19 June, bentazon (BASF Canada Inc.) was applied at a rate of 1100 g a.i. ha⁻¹ with the surfactant Merge® (BASF Canada Inc.) at 1.0 L ha⁻¹. The herbicide sethoxydim plus Merge® was sprayed at a rate of 450 g a.i. ha⁻¹ on 7 July. Irrigation was used only in the first year to establish uniformly severe symptoms of anthracnose throughout the field. The beans were harvested on 10 August 2006 and all the infected straw remained on the field.

In the second year (2007), the site was split into four replications of the three rotation treatments with plot dimensions of 5.7 × 18 m. Fertilizer was broadcast in the spring, using a blend of 67 and 33 kg ha⁻¹ of N and P, respectively. The treatments included plots seeded to spring wheat (Triticum aestivum L.), disease-free seed of the susceptible navy bean cultivar ‘Cirrus’ and a summer fallow treatment in the second year. The treatments were arranged in a randomized complete block design with four replications. The experiment was seeded on 8 May 2007 using the same zero tillage system as in 2006. Wheat was seeded at a rate of 125 kg ha⁻¹. Fallow treatments were left undisturbed and sprayed with glyphosate (Monsanto Canada Inc., Ottawa, ON) at a rate of 3.7 L ha⁻¹. Bromoxynil and MCPA (Bayer Crop Science Canada, Calgary, AB) was applied at 1.0 L ha⁻¹ to the wheat plots on 14 June. Bentazon and Merge® were sprayed at rates of 1.8 and 0.5 L ha⁻¹, respectively, on 19 June. At the end of August, anthracnose severity was rated as the percentage of the leaf veins with symptoms on 10 plants randomly selected within each of the plots of only the BBB rotation, since no bean plants were grown in the other two rotation treatments that year. Wheat and bean crops were harvested on 10 September 2007 and 30 September, respectively, and the straw was returned to their plots.

In the third and final year, disease-free seed of the susceptible navy bean cultivar ‘Cirrus’ was planted across the entire site in Morden on 23 May 2008. Prior to zero till seeding, a fertilizer blend was broadcast at a rate of 90-34-18 kg ha⁻¹ of N, P and K, respectively. Glyphosate herbicide was used for weed control at a rate of 3.7 L ha⁻¹ prior to emergence on 31 May. Bentazon with Merge® was applied at a rate of 1056 g a.i. on 25 June. Weeds were further controlled with a tank mix of bentazon, fomesafen and the surfactant Agral 90 (Syngenta Canada) at a rate of 864, 144 g a.i. ha⁻¹ and 0.25% (v/v), respectively, on 3 July. The borders of each plot from the previous year had been marked, so the location of different rotation treatments in each replication could be readily determined in 2008. Data on anthracnose incidence and severity on the canopy and pods were collected from all the plots in the study. The initial anthracnose incidence in each plot was assessed on 9 July 2008 and was based on the number of plants with anthracnose symptoms out of a random sample of 20 plants in each plot. In year 3, the severity of anthracnose on the leaves was visually evaluated at 10 randomly distributed sites within each plot. The rating sites in each plot were marked with a stake, so they could be located again and rated over time. Foliar anthracnose severity ratings were made on 25 July, 1 August, 8 August and 28 August in 2008. The percentage of pod tissue with disease symptoms also was recorded at the same sites in each plot on 28 August.

All the plots were harvested with a small-plot combine on 16 September 2008. Then, the bean seed was dried to 15% moisture content and weighed to determine seed yield (kg ha⁻¹). The percentage of anthracnose-discoloured seed was visually assessed by counting the number of seed with anthracnose lesions in a sub-sample of 200 seeds from each plot.

Statistical analysis

Data on the survival of anthracnose in tissue samples from both locations were tabulated for the replications over years and blocks and the results were presented in tabular form for both provinces and type of plant tissue samples throughout the sampling periods. Thus, the longest surviving samples are included in the tables, and were not averaged over years.

For the rotation study, the median percentage across the 20 plants evaluated in each plot was calculated for canopy incidence and pod infection, and seed discolouration; these were used for analysis. Median values were utilized rather than averages to better deal with extreme values (outliers) in the data set. Those attributes and plot yield were subjected to ANOVA, and the residuals were evaluated to discern whether the ANOVA assumptions were met or not. Treatment differences for canopy (4th rating), pod infection, seed discolouration and yield were discerned by the LSD (5%) procedure. All statistical computations were made using the statistical programming language, GenStat® (Payne, 2017).

Results

Weather

The weather conditions at Morden and the Exeter region varied considerably among the years of this study (Table 1). Morden was consistently much colder and drier than the Exeter region. At Morden, from November 2007 through to August 2009, the mean monthly temperatures were similar to the long-term average, but in February, May and December of 2008, as well as in January through to May of 2009, the average temperatures were much cooler than the long-term average. In contrast, the mean temperatures at Morden were much higher than normal throughout the winter months of 2016 and January and February of 2017. Temperatures were warmer than normal in the Exeter region in November and December 2007 as well as in January and May 2008. The temperatures in the Exeter region were similar to or higher than the long-term mean throughout the test period in 2015–2017.

Table 1. Mean monthly average temperatures (°C) and precipitation (mm) at Morden, MB and London, ON (Exeter region) during the time periods of the anthracnose survival study.

		Morden								London
	Season	2007	2008	2009	2010	2015	2016	2017	LTA	2007	2008	2009	2010	2015	2016	2017	LTA
Temp.	Jan	-	−15	−18	−13	-	−12	−11	−16	-	−3	−9	−6	-	−4	−2	−6
	Feb	-	−15	−14	−13	-	−8	−8	−12	-	−6	−4	−5	-	−3	0	−6
	Mar	-	−6	−8	−1	-	0	−5	−5	-	−2	1	4	-	4	0	−1
	Apr	-	3	3	9	-	3	6	5	-	9	8	10	-	6	10	6
	May	-	9	9	11	-	15	12	13	-	11	13	15	-	14	12	13
	June	-	16	15	17	-	18	18	18	-	19	18	19	-	18	19	18
	July	-	18	17	21	-	22	21	20	-	21	18	22	-	22	21	20
	Aug	-	19	18	20	-	20	19	19	-	19	20	22	-	22	19	20
	Sept	-	13	17	13	-	15	15	13	-	17	16	16	-	18	17	15
	Oct	-	6	3	2	-	7	7	6	-	9	8	10	-	11	13	9
	Nov	−4	−4	2	-	0	4	-	−4	7	2	6	-	7	6	-	3
	Dec	−14	−18	−15	-	−6	−12	-	−13	3	−4	−3	-	3	−3	-	−3
Precip.	Jan	-	20	13	14	-	6	18	21	-	94	44	37	-	50	80	75
	Feb	-	11	22	14	-	8	13	19	-	131	87	33	-	77	71	60
	Mar	-	9	26	7	-	17	14	25	-	89	89	35	-	142	86	78
	Apr	-	19	22	43	-	56	8	39	-	58	137	76	-	49	113	82
	May	-	32	54	133	-	67	22	65	-	89	104	114	-	31	133	83
	June	-	155	73	53	-	128	65	85	-	71	92	133	-	63	68	84
	July	-	33	57	103	-	109	38	74	-	75	76	110	-	91	50	82
	Aug	-	49	37	77	-	67	37	70	-	60	95	40	-	170	42	82
	Sep	-	91	107	72	-	90	71	53	-	62	41	108	-	70	32	96
	Oct	-	61	33	66	-	28	5	43	-	75	76	88	-	50	85	81
	Nov	12	29	0	-	16	33	-	26	58	130	27	-	58	55	-	93
	Dec	40	23	39	-	23	41	-	21	68	133	65	-	82	82	-	88

LTA = Long-term Average (1968–1998).

At Morden, precipitation was usually lower than normal throughout the test periods in 2007–2009, but there were heavy accumulations of rain in June, September and October of 2008 and September of 2009. Precipitation at Morden was close to normal or below average from November 2015 to March 2016, then above average rainfalls occurred from May to July of 2016, which was followed by periods of low precipitation in March to August in 2017. Precipitation occurred on a regular basis in the Exeter region throughout the time periods of the anthracnose survival study. Frequent precipitation occurred as either rain or snow in each month throughout the winter months in the Exeter region in 2008, and February to June 2009, and again throughout most of May to September in 2010. Weather conditions in the Exeter region were generally drier than normal during the test period of the anthracnose survival study from 2015 to 2017, but precipitation was heavier than usual in February, March and September of 2016 and in February to June of 2017.

Anthracnose survival study

From 2007 to 2010 and 2016 to 2017, the stems, pods and seed samples from anthracnose infected bean were removed from the burial sites at Morden and Exeter. The filter paper assay essentially used susceptible bean plants as a selective medium to detect the presence of viable inoculum of the anthracnose fungus. The results from the filter paper bioassay for each of the three sampling cycle periods were fairly consistent, so the final results from the test period cycles at each field location were combined (Table 2). At the Morden site, there were large differences in the survival of C. lindemuthianum between the tissue samples that were buried versus those that remained on the soil surface. By the third sampling date (i.e. summer of year 1), buried samples of infected pods (20%), seed (27%) and stems (22%) produced symptoms on a low percentage of the bean plants in comparison to the high incidence of infection (63–84%) on the bean plants inoculated with tissue samples that remained on the soil surface.

Table 2. Average incidence of infection (%) on plants of the bean cultivar AC Pintoba following exposure to inoculum of Colletotrichum lindemuthianum on pod, seed and stem samples over 2-year periods beginning in 2007, 2009, and 2015 at Morden, MB and Exeter, ON.

	Placement		Surface									Buried
	Plant part		Pod			Seed			Stem			Pod			Seed			Stem
Province	Year	Season	n	p	(%)	n	p	(%)	n	p	(%)	n	p	(%)	n	p	(%)	n	p	(%)
Manitoba	Yr1	winter	33	32	97	33	10	30	36	33	92	33	31	94	36	10	28	36	16	44
		spring	33	28	85	33	8	24	36	20	56	33	17	52	36	5	14	30	9	30
		summer	36	30	83	36	24	67	36	23	64	15	3	20	15	4	27	27	6	22
		autumn	36	22	61	30	12	40	36	23	64	ns	ns	ns	ns	ns	ns	33	4	12
	Yr2	winter	24	12	50	24	2	8	24	12	50	ns	ns	ns	ns	ns	ns	24	0	0
		spring	18	14	78	24	1	4	24	12	50	ns	ns	ns	ns	ns	ns	18	1	6
		summer	15	0	0	21	0	0	18	1	6	ns	ns	ns	ns	ns	ns	18	0	0
		autumn	12	0	0	12	0	0	12	0	0	3	0	0	ns	ns	ns	12	0	0
Ontario	Yr1	winter	36	33	92	30	5	17	33	30	91	36	21	58	33	7	21	36	23	64
		spring	30	27	90	36	11	31	36	34	94	36	15	42	36	1	3	33	19	58
		summer	36	12	33	24	1	4	36	18	50	30	3	10	24	4	17	36	5	14
		autumn	33	0	0	21	0	0	36	0	0	12	0	0	3	0	0	36	0	0
	Yr2	winter	21	0	0	12	0	0	24	0	0	9	0	0	ns	ns	ns	21	0	0
		spring	ns	ns	ns	ns	ns	ns	ns	ns	ns	ns	ns	ns	ns	ns	ns	ns	ns	ns
		summer	ns	ns	ns	ns	ns	ns	ns	ns	ns	ns	ns	ns	ns	ns	ns	ns	ns	ns
		autumn	ns	ns	ns	ns	ns	ns	ns	ns	ns	ns	ns	ns	ns	ns	ns	ns	ns	ns

Determined from bagged plant (pod, seed, stem) samples over two years following the dry bean crop in 2007, 2009 and 2015 at Morden, MB, and Exeter, ON, where n = total number of samples recovered, p = number of positive samples and ns = sample material not found.

By the fourth sampling date (i.e. autumn of year 1), the buried pod and seed tissue samples at Morden had completely rotted and could not be retrieved to assay for infectivity. In contrast, the buried stems were still able to act as a source of infection on a low percentage of the inoculated plants until the third last sampling date (spring of year 2). The breakdown of bean tissue samples on the soil surface was relatively slow, which enabled the tissue to be assayed for infectivity for the duration of each sampling period. At the fourth sampling date (autumn of year 1) all three types of bean tissue on the soil surface were able to produce sufficient inoculum to induce high incidences of infection on the bean seedlings. However, by the fifth sampling date (winter of year 2), the incidence of seedling infection arising from the infected pods and stems was much higher than that observed from the infected bean seeds. At the subsequent sampling date (spring of year 2), the infected seed produced a very low incidence of seedling infection while the pod and stem tissue caused a relatively high incidence of infection. At the seventh sampling (summer of year 2), only the stem tissues produced a trace amount of infection and this occurred at a low incidence. None of the infected bean tissue samples from the soil surface at Morden were able to incite anthracnose symptoms on the bean seedlings at the final sampling date (autumn of year 2).

Differences in the infectivity of the buried and surface samples of the infected bean tissue also were large at the field site near Exeter, ON (Table 2). The bean tissue samples at Exeter remained intact for the time periods in which they were sampled, but the ability of the tissue samples on the soil surface to cause infection persisted for a shorter time period than was observed at Morden. In contrast to the results observed at the Morden site, there were no differences in length of time the three types of tissue could incite symptoms on bean seedlings. However, after the initial sampling date (winter of year 1), the numbers of intact tissue samples available for the assay from the infected seed declined much more rapidly than those from the pods and stems. Buried seed tissue was not available for the filter paper assay by the fifth sampling date (winter of year 2). Despite the differences in the breakdown of the infected bean tissue, none of the buried tissue samples from Exeter caused infection on the fourth sampling date (autumn of year 1). At each sampling date at Exeter, the incidence of seedling infection was much higher for all types of tissue except the seed from the soil surface in comparison with the same types of tissue that had been buried. As observed in Morden, the infected seed samples from the soil surface produced a much lower incidence of infection than occurred for the pod and stem samples. In contrast to the results from Morden, none of the infected bean tissue samples caused any anthracnose symptoms after the third sampling date (summer of year 1).

Crop rotation study

In August 2007 (year 2), bean anthracnose symptoms were observed in the canopy of the BBB rotation and disease severity was assessed (mean severity 22.3 ± 2.8%), which was the only treatment that year that included dry beans. No volunteer bean seedlings were observed in any of the plots in 2007 or 2008. In 2008, anthracnose symptoms developed early in the growing season within the continuous bean rotation and became quite severe by the end of the growing season (Table 3). Anthracnose symptoms in the bean-fallow-bean and the bean-wheat-bean rotations were initially weak and were consistently significantly less severe than those in the continuous bean rotation on each of the four dates that anthracnose symptoms in the plant canopy were rated. There were no differences in anthracnose severity in the crop canopy between the BFB and BWB treatments on any of the assessment dates. All three rotations also differed significantly for anthracnose severity on the pods at the end of August 2008. Differences in pod infection resulted in a much higher incidence of seed discolouration in 200-seed samples from the BBB rotation in comparison to the low incidences in the seed samples from the other two rotations. The yield in the BBB rotation was 43–45% lower than the other two rotation treatments. The seed yield and incidence of seed discolouration of the BFB and BWB rotations were similar.

Table 3. Anthracnose incidence and severity ratings (%), seed discolouration and yield in three crop rotations at Morden in the third year of the rotations (2008).

Rotation^Z	Incidence out of 20^Y	Severity in the crop canopy (%)				Pod^X Infection (%)	Seed^X discolouration (%)	Yield^X (kg ha^−1)
		1st rating	2nd rating	3rd rating	4th^X rating
BBB	19	6	6	8	12a	14a	14a	1230b
BFB	1	1	1	1	4b	9b	7b	2250a
BWB	1	0	0	0	2b	3c	3b	2130a
LSD (0.05)^X					4	5	6	102

^ZAbbreviations for rotation crops: B = Bean, F = Fallow and W = Wheat.

^YBased on initial observations that were made on 20 plants per plot on 7 August 2008.

^XMeans followed by different letters within a column are significantly different as determined by the LSD (5%) procedure.

Discussion

This study demonstrated that infected crop debris on the soil surface and at times buried in the soil could survive under a wide range of environmental conditions and act as a primary source of inoculum of C. lindemuthianum for infection on dry beans in subsequent growing seasons. In Manitoba, infected bean tissue on soil surface and stem tissue buried in the soil could still infect a bean crop in the second spring after the harvest of a bean crop with anthracnose. The period of infectivity was shorter for bean tissue samples from Exeter than similar samples from Morden. Burial of tissue samples sped up the decline in infectivity at both locations. Studies on other crop pathogens (Bailey, 1996; Ntahimpera et al., 1997) have shown that burial of crop debris acts in at least two ways to reduce the spread of diseases. First, burial of the crop physically separates the inoculum of the plant pathogens from their host plants (Gossen & Miller, 2004). Second, burial of infected plant material speeds up the decomposition of the infected crop debris (Katan, 2000; Gossen, 2001). Burial of the crop debris usually exposes it to damper conditions than the soil surface, which results in a faster rate of decomposition (Gossen, 2001). Many plant pathogens like C. lindemuthianum cannot effectively compete with soil saprophytes, so they quickly die out once the crop residue decomposes (Ntahimpera et al., 1997).

In Manitoba, stem and pod debris were able to produce anthracnose lesions on the seedlings after they had been left on the soil surface for up to 21 months. At both field sites, the percentages of infected plants were much higher for plants inoculated with tissue samples from pods and stems than from the seed samples. The incidence of seedling infection from seed samples on the soil surface was very low after the first 6 months at Exeter and the first 9 months at Morden. The anthracnose fungus in buried pod and seed tissue samples at Morden died out after 9 months of burial, but the stem tissue was still able to infect a small percentage of plants for up to 21 months. This difference in the viability of C. lindemuthianum in the different type of tissues has not previously been reported. It appears that differences in the infectivity of the different types of tissue and reductions in viability of buried samples are directly related to the decomposition of the infected bean tissues. Infected bean seed would be a richer source of starch (Nassar et al., 2010) for saprophytic fungi and bacteria than pods and stems, which would be more resistant to decay. The obvious differences in the decline of infectivity of the tissue samples that was observed between the two test locations also likely relates to differences in the speed of the breakdown of the infected bean tissue. The rate of decomposition of infected bean tissue would be greater under the warmer and wetter conditions that occurred at Exeter than under the colder and drier conditions at Morden.

This was the first study to examine the survival of C. lindemuthianum in western Canada. The test sites had an obvious effect on the length of time that C. lindemuthianum remained viable on the various types of bean tissue. The anthracnose fungus remained viable for at least 9 months longer at Morden than it did at the Huron Research Station. This difference in the duration of infectivity of C. lindemuthianum in the tissue samples was likely due to the differences in weather conditions at these locations, particularly mean daily temperature. Winter conditions can be quite severe on the Canadian prairies (Gossen, 2001). Throughout the years of the study, mean temperatures were much higher at Exeter than they were at Morden. The warmer temperatures likely contributed to the decline of viable inocula of the anthracnose fungus in bean tissues in Ontario within 9 months as opposed to more than 18 months at Morden.

Tu (1983) reported that C. lindemuthianum was not able to persist on buried, infected bean stubble and incite anthracnose in the following spring at field sites near Harrow, Ontario. However, in the current rotation study, the anthracnose fungus was able to survive at sufficient levels for over a year and a half and produce anthracnose symptoms on dry beans. The Bean-Fallow-Bean and the Bean-Wheat-Bean rotation treatments had anthracnose symptoms in the crop canopy early in the growing season, which indicated that the anthracnose fungus had survived for more than one growing season in the absence of a susceptible host. In year 3 of the rotation study, disease incidence was consistent across each plot, and across each replicate, indicating that the inoculum came from infected residue from the previous crop, and not from another source. The microclimate created by the Bean-Wheat-Bean rotation apparently did not influence the decomposition of bean residues on the soil surface and reduce anthracnose severity in the bean crop in the third year of the rotation. A similar study at the Huron Research Station in Exeter that was conducted at the same time also showed that the anthracnose fungus could persist and cause disease symptoms more than a year and a half after infected bean crops had been harvested (C.L. Gillard, unpublished data). Anthracnose symptoms were not expected at Exeter, since the assay of infected bean tissues had not detected any viable inoculum. These rotation results indicate that the sensitivity of the tissue assay for anthracnose requires improvement.

Tu (1982) ploughed under the bean debris in the autumn, while at both Morden and Exeter, the rotations were carried out under zero tillage, which would have left the infected bean debris on the soil surface where it would have been subject to slower decomposition. Similar studies in New York State (Dillard & Cobb, 1993; Ntahimpera et al., 1997) also determined that the anthracnose fungus could overwinter on infected bean stubble and cause severe symptoms of anthracnose in subsequent crops; they attributed this difference in survival from those of Tu (1982) to the lack of burial of the crop debris (autumn disking in the Ontario study as opposed to no autumn tillage in the study in New York state), a milder climate and possibly race composition (beta race in NY and delta race in ON). Yet in a separate study, Dillard & Cobb (1993) showed with a whole bean plant assay that viable inocula of C. lindemuthianum was detected in buried bean debris for up to 238 days after burial.

The studies by Dillard & Cobb (1993) and Ntahimpera et al. (1997) both used a live plant assay, similar to the modified plant assay used in the current study, to evaluate the viability of C. lindemuthianum rather than agar plating, which was used by Tu (1982). It appeared that the live plant assay provided a much more sensitive method for detecting infectious inoculum of C. lindemuthianum and the results would have not been affected by the presence of saprophytic fungi and bacteria (Dillard & Cobb, 1993) as was likely the case in Tu’s study (1982). Dillard & Cobb (1993) demonstrated that the anthracnose fungus could only be detected for 145 days after burial with agar plating, but was detected for 238 days after burial with a live plant assay. Those studies support the current disease recommendation that beans should be grown in 2–3 year rotations with non-host crops in order to prevent new outbreaks of anthracnose (Ntahimpera et al., 1997; Schwartz et al., 2005). However, this is contrary to the conclusions of Tu (1982, 1988) who stated that C. lindemuthianum could not overwinter in Ontario, so crop rotation would not be recommended for bean anthracnose control.

Burial of infected tissues greatly reduced the longevity of C. lindemuthianum in all three types of bean tissue. Ntahimpera et al. (1997) demonstrated that deep tillage greatly decreased anthracnose symptoms in subsequent bean crops. However, in recent years, minimum tillage operations have been almost universally adopted for soil conservation in western Canada (Larney et al., 2015) and elsewhere especially for low-residue producing crops such as dry bean. Therefore, the use of deep tillage to control bean anthracnose is no longer a common practice.

The survival of C. lindemuthianum in the Bean-Fallow-Bean and the Bean-Wheat-Bean rotation treatments in the current study did not decrease yield to the same extent as the continuous bean rotation. However, seed infection and discolouration in the Bean-Fallow-Bean and the Bean-Wheat-Bean rotations were sufficient to downgrade the value of the harvested seed and the infected seed would have served as a source of inoculum for subsequent bean crops. The results of the rotation and the overwintering studies underline the importance of recommendations that beans should be grown in crop rotations with non-host crops for at least two years and preferably for three years to prevent yield and seed quality losses caused by anthracnose in beans (Dillard & Cobb, 1993; Schwartz et al., 2005).

Acknowledgements

We thank Dennis B. Stoesz of the Morden Research and Development Centre and Donald Depuydt and Steve Willis of the Ridgetown Campus of the University of Guelph for their technical assistance in this study.

Additional information

Funding

This research was financially supported by the Agricultural Adaptation Council, the Manitoba Pulse and Soybean Growers and the Ontario Bean Growers are provincial organizations in Canada that fund research on pulse crops. Protection, BASF Canada and the Pulse Science Cluster (Growing Forward 2) of Agriculture and Agri-Food Canada; Syngenta Foundation for Sustainable Agriculture.