Population dynamics of Epicoccum nigrum, a biocontrol agent against brown rot in stone fruit

Conidial culture and production

Epicoccum nigrum Link strain 282 (ATCC number 96794) was isolated originally from healthy peach twigs at an experimental orchard in Madrid, Spain (Melgarejo et al. 1985). The fungus was maintained on potato dextrose agar (PDA; Difco, Detroit, MI, USA) slants at 4°C, grown on PDA plates and incubated at 20–25°C in the dark for 10 days in a solid-state fermentation system for conidial production (Larena et al. 2004). Briefly, the fungus was grown on a mixture of peat (Gebr. BRILL substrate GmbH & Co. KG, Georgsdorf, Germany): vermiculite (Termita, Asfaltex, S.A., Barcelona, Spain): lentil meal (1 : 1 : 0·5, w/w/w). Five hundred grams of the above-described substrate (40% w/w moisture content) were placed in a plastic bag (1300 cm³) designed for solid-state fermentation (VALMIC^R, Filter 160 mm, 200 × 560 mm² large × hole, Sacherei de Pont-Audemer S.A., Pont-Audemer, France), sealed and sterilized by autoclaving at 1·0 kg cm⁻² and 120°C for 1 h on three consecutive days. Bags were then inoculated with a conidial suspension of E. nigrum to obtain 10⁵ conidia g⁻¹ dry substrate, sealed again and then incubated in the dark at 20–25°C for 10 days. Conidia were obtained from the solid fermentation bags by adding sterile distilled water (SDW) to the conidial-substrate mixture at a rate of 1 : 4, w/v. The conidial suspensions were then shaken in a rotary shaker (Model AO400; Bunsen, S.A., Madrid, Spain), at 200 rev min⁻¹ for 10 min and filtered through several layers of glass wool. Most conidia passed through the glass wool and were concentrated by centrifugation (14 040 g for 10 min). The final yield of fresh conidia was 10⁸ conidia g⁻¹ dry weight of substrate with viability greater than 80% (Larena et al. 2004).

Conidial formulations

Five types of conidial formulations were prepared: FOR1, FOR5, FOR6, FOR7 and FOR8 (Table 1). For production of FOR1, FOR5, FOR6 and FOR7, the fresh conidia were resuspended after the 14 040 g centrifugation (see above section) in different additive solutions (2·5% methyl cellulose for FOR1, 2·5% methyl cellulose + 25% glycerol for FOR5 and FOR6, 1% KCl + 1·25% sodium alginate for FOR7) in 250 ml centrifuge tubes, shaken on a vortex mixer (Reax Top, Heildolph, Rose Scientific Ltd. Alberta, Canada) for 10 s and kept for 10 min at room temperature (20–25°C). Each conidial suspension was filtered through 2-μm filter paper in a Büchner funnel and then mixed with one of the following desiccants: silica powder (in the case of FOR1, FOR6 and FOR7) or talc (in the case of FOR5) at a ratio necessary to get granulation (Larena et al. 2007). Then, the granules were dried in a fluid bed dryer (FBD model 350s, Burkard Manufacturing Co Ltd., Hertfordshire, UK) at the highest air flow rate at 40°C (Larena et al. 2003, 2007). The moisture content of each final conidial formulation was measured using a humidity analyser (BOECKEL, GmbH & Co, Hamburg, Germany). Germination of the dried conidia was assessed using a previously described bioassay (Larena et al. 2007). Conidia were dried to reduce their moisture content to between 5% and 10% and their germinability was as high as 80% after drying (Larena et al. 2003, 2007).

In the case of FOR8, fresh conidia were produced in the fermentation bags, as described above, but the additive 1% KCl was added to the substrate at the same time that the bags were inoculated with E. nigrum conidia. After the production process with KCl, fresh conidia were re-suspended in 1·3% Nu-film-17 (96% di-menthene, Miller Chemical and Fertilizer Co., Hanover, Pennsylvania, USA), maintained for 10 min at room temperature, and talc was then added to conidial paste until it granulated. The granules were then dried as described above.

Treatments and experimental design

Twelve field surveys were carried out in commercial peach orchards in Spain and Italy over four growing seasons between 2002 and 2005 (Table 2). The Spanish orchards were located in Alfarrás [Universal Transverse Mercator (UTM) Coordinate: 298197, 4634045], Albesa (UTM: 305302, 4624899), Sudanell (UTM: 297230, 4603615; Lleida, Cataluña, Spain) and Vinebre (UTM: 297930, 4562145; Tarragona, Cataluña, Spain). The Italian orchards were located in Mozeccane (UTM: 642644, 5018852; Veneto, Italy), Domegliara (UTM: 642522, 5042149; Veneto, Italy) and Salerano sul Lambro (UTM: 530099, 5015715; Lombardia, Italy). The areas of the orchards ranged from 3 to 10 ha and the planting distances between rows and trees ranged from 2 × 4 m to 4 × 5 m. The surveys involved different cultivars of peach [Prunus persica (L.) Batsch (‘Summer Lady’, ‘June Lady’, ‘Red September’, ‘Star Red Gold’ and ‘Red d′Albesa’)] and nectarines [P. persica var. nectarina (Aiton) Maxim (‘Caldesi 20–20’, ‘Venus’ and ‘Autumn Free’; Table 2]. The experimental design of each experiment was a randomized complete block design in which four replications (four blocks) were used, and three trees were used as the sample unit for each treatment and replication. To avoid interplot interference, three barrier trees were used to separate the randomised blocks and treatments. Daily measurements of mean environmental temperature (T), rainfall (mm) and percent relative humidity (RH) were collected by automated weather-monitoring equipment located 0·5–5 km from each orchard. Average daily T, rainfall and RH were calculated for each orchard and for each treatment period (day of first E. nigrum treatment to 1 day before the second E. nigrum treatment; day of second E. nigrum treatment to 1 day before the third E. nigrum treatment; day of third E. nigrum treatment to 1 day before the fourth E. nigrum treatment and day of fourth E. nigrum treatment to 1 day before the fifth E. nigrum treatment).

Two treatment doses of E. nigrum conidia were applied: 10⁶ conidia ml⁻¹ in 2002 and 2003, and 10⁷ conidia ml⁻¹ in 2004 and 2005. The application schedule of the E. nigrum treatments was: during the full flowering-to-shuck split period (when the growth of the fruit splits the shuck; BBCH = 65–71; BBCH general scale from Biologische Bundesantalt, Bundessortenamt and CHemische Industrie, Germany; Meier et al. 1994); at pit hardening (BBCH = 76) and during the month before harvest (BBCH = 81–87). Details of the treatments and their application dates are shown in Table 3.

In all Spanish surveys and in the DOM02 Italian survey, the application equipment for the various treatments was a backpack sprayer at 10 bars working pressure with a 1-mm hollow cone nozzle. In the SAL02 Italian survey, a cart at 20 bars working pressure with a spray gun nozzle was used to apply the treatments. In the MOZ03 Italian survey, a backpack sprayer at 14 bars working pressure with a spray gun nozzle was used. The details of the formulations and the field surveys are shown in Tables 1 and 2, respectively.

No specific fungicide against Monilinia spp. was applied during the surveys, except in 2004 and 2005, when trees in the same orchard were treated with either an experimental conidial formulation or a fungicide. When applied, the specific fungicides were: Caddy Pepite 10 (cyproconazole 10% WG; Bayer Cropscience, S.L. Valencia, Spain) in ALF04 and SUD04; and Folicur 25RW (tebuconazole 25% WG; Bayer Hispania Industrial, S.A., Barcelona, Spain) in ALF05 and SUD05. In order to protect untreated trees from the fungicidal sprays, three barrier trees were used to separate them from the treated trees. All treatments were applied in nonwindy conditions and are described in Table 4.

Population dynamics of Epicoccum nigrum

To investigate the population dynamics of E. nigrum following the application of each conidial formulation to peach surfaces, 10 flowers or five fruits per sample unit at each treatment were sampled in each orchard by collecting specimens randomly from either the centre of the tree or the quarters. The annual sampling frequencies were nine in 2002, 10 in 2003, 11 in 2004 and 12 in 2005. At the laboratory, each sample (10 flowers or five fruits per treatment and replicate) was suspended in containers containing SDW and shaken for 30 min at 150 rev min⁻¹ in a rotary shaker. Each solution was decanted to centrifuge tubes that were then centrifuged for 10 min at 14 040 g, and the recovered pellet was resuspended in 5 ml SDW (concentrate). The population size of E. nigrum in each concentrate was estimated from (i) the number of conidia and (ii) the number of colony-forming units (CFUs) of E. nigrum per flower or fruit. The numbers of E. nigrum conidia were counted in a haemocytometer under a light microscope (×100). The number of CFUs of E. nigrum per Petri dish containing PDA (Difco, Detroit, MI, USA) supplemented with 0·5 g l⁻¹ streptomycin (PDAs) was counted by the naked eye. Aliquots (100 μl) of undiluted and diluted concentrates were spread onto PDAs. Three replicate dishes were used for each replicate and dilution. The Petri dishes were maintained at 20–25°C for 5–7 days in the dark before counting the number of CFUs (Lacey et al. 1980). Flowers and fruits from SUD05 and ALF05 in 2005, which did not receive any treatment, were used as the controls to study the indigenous E. nigrum population. The identity of E. nigrum conidia in the CFUs was confirmed periodically by morphological characterization.

Biocontrol efficacy of Epicoccum nigrum conidial formulations

To evaluate the effect of the various E. nigrum conidial formulations on the number of Monilinia spp. conidia on peach surfaces, four field surveys (ALF04, SUD04, ALF05 and SUD05) were conducted. Trees that were treated with either fungicides (at flowering, pit hardening and 21 days before harvest) or not treated with either a fungicide or the conidial formulation in each orchard were used as the controls in these surveys (Table 3).

To compare sizes of the Monilinia spp. population on peach surfaces at each treatment, 10 flowers or five fruits (without overt disease symptoms) per sample unit were collected on 11 occasions in 2004, and on 12 occasions in 2005 from each orchard. Each sample (10 flowers or five fruits per treatment and replicate) was suspended in containers containing SDW and shaken for 30 min at 150 rpm in a rotary shaker. Each solution was decanted to centrifuge tubes that were then centrifuged for 10 min at 14 040 g. The recovered pellet was re-suspended in 5 ml SDW (concentrate). The number of Monilinia spp. conidia in each concentrate was counted in a haemocytometer under a light microscope (×100) and expressed as conidial number per flower or fruit.

Data analysis

Data on the conidial number and CFUs of E. nigrum for each treatment at the different sampling dates were analysed by analysis of variance (anova;Snedecor and Cochran 1980). The number of conidia and CFUs of E. nigrum per flower or fruit were log₁₀ (x + 1) transformed before the analysis in order to improve the homogeneity of the variances. When the F-test showed a significant difference at P = 0·05, treatment means were compared using the Student–Newman–Keul’s test (Snedecor and Cochran 1980).

Progress curves for conidial number of E. nigrum and M. laxa, and for CFUs of E. nigrum for each treatment at the different sampling dates were plotted for each orchard and year. The area under each progress curve [area under the curve for conidial numbers of E. nigrum (AUNCEPIPC) or Monilinia spp. (AUNCMONPC), and the area under the curve for the number of CFUs of E. nigrum (AUCFUEPIPC) for each orchard and year] were calculated by trapezoidal integration as described by Campbell and Madden (1990) using the following formulae with an Excel spreadsheet:

where ‘t’ is ‘days from full flowering to harvest’, ‘NCEPI’ or ‘NCMON’ are ‘the number of conidia at each sampling date for E. nigrum or Monilinia spp., respectively’, ‘CFUEPI’ is ‘the number of CFUs at each sampling date for E. nigrum’ and ‘n’ is ‘the number of samplings’.

Data of the AUNCEPIPC and the AUCFUEPIPC from ALB03 were analysed by anova (Snedecor and Cochran 1980) and when the F-test showed a significant difference at P = 0·05, treatment means were compared by Student–Newman–Keul’s test (Snedecor and Cochran 1980). Data of the AUNCEPIPC and the AUCFUEPIPC from 11 surveys (ALF02, VIN02, SAL02, DOM02, ALF03, SUD03, MOZ03, ALF04, SUD04, ALF05 and SUD05), or the AUNCMONPC from four surveys (ALF04, SUD04, ALF05 and SUD05) were analysed independently by contrast with the F-test at significance levels of 0·1, 0·05 and 0·01 (Snedecor and Cochran 1980).

The equations AUNCEPIPC or AUCFUEPIPC = f (T, HR, npit, npre, D) were used to investigate the relationship between the effects of T, RH, the number of E. nigrum applications at pit hardening (npit), the number of E. nigrum applications in the last month before harvest (npre) and the number of E. nigrum conidia (D) in AUNCEPIPC and AUCFUEPIPC, using data pooled from the 12 surveys, where f (T, HR, npit, npre, D) is a linear function of the terms T, HR, npit, npre, THR, Tnpit, Tnpre, HRnpit, HRnpre, npitnpre, T², HR², npit² and npre². Regression analysis was used to estimate these parameters. Data were fitted to the equation by the Model Regression Selection option in Statgraphics Plus for windows v. 4·1 (StatPoint, Inc., Herndon, VA, USA). The selection was performed on the basis of the significance of the estimated parameters: the coefficient of determination, the adjusted coefficient of determination, the Mallow’s Cp coefficient (evaluates the fit of regression model by the squared distance between its predictions and the true values), the Durbin–Watson statistic (a test for serial correlation in the residuals of a least squares regression analysis) and the normal distribution of residuals (Jacome and Schuh 1992).

Population dynamics of Epicoccum nigrum

Figure 1 shows (i) the number of conidia and CFUs of E. nigrum following the application of fresh conidia to peaches (VIN02 and DOM02) and nectarines (ALF02 and SAL02; Fig. 1a–d) and (ii) T, rainfall and RH (Fig. 1e–h) during each crop season at each orchard. Figure 2 shows (i) the number of conidia following the application of the FOR1 E. nigrum conidial formulation to peaches (ALF03 and MOZ03) and nectarines (SUD03; Fig. 2a–c) and (ii) T, rainfall and RH (Fig. 2d–f) during each crop season at each orchard. Figure 3 shows (i) the number of conidia and CFUs following the application of the FOR7 E. nigrum conidial formulation to peaches (ALF04) and nectarines (SUD04; Fig. 3a,b and (ii) T, rainfall and RH (Fig. 3c,d) during each crop season at each orchard. Figure 4 shows (i) the number of conidia and CFUs following six applications of the FOR7 E. nigrum formulation to peaches (ALF05) and five applications to nectarines (SUD05; Fig. 4a,b) and (ii) T, rainfall and RH (Fig. 4c,d) at pit hardening (BBCH = 76) and at preharvest (BBCH = 81–87). We found that the results of applying the FOR8 formulation were similar to those obtained with the FOR7 formulation and for this reason, as a matter of clarity, only results of FOR7 are presented in Fig. 3. In all instances, the number of E. nigrum conidia on peach surfaces (flowers or fruits) at each sampling date in every orchard was 100- to 1000-fold higher than the number of CFUs on the PDAs. Significantly larger E. nigrum populations (P = 0·05) were present on the fruit surfaces 15 days before harvest (20–30% for conidia and 80–100% for CFUs) than on the flower surfaces in the early spring (Figs. 1–4). The number of conidia remained fairly constant over the period from petal fall to the first preharvest treatment with the various conidial formulations (Figs. 1–4). The numbers of E. nigrum conidia were always significantly higher (P = 0·05) immediately following application of the various conidial formulations when compared with those found at the other sampling times (Figs. 1–4). The only exception was when it rained following an application.

The effect of applying a fresh conidial formulation on E. nigrum presence on peach surfaces (ALB03) compared with that following application of the FOR1, FOR5 and FOR6 formulations are presented in Table 5. The results of applying the FOR1, FOR5 and FOR6 formulations were similar to those obtained after applying the other formulations in the different orchards (Figs. 1–4). No significant differences in the AUNCEPIPCs between each E. nigrum conidial formulation were found in the ALB03 survey (Table 5). However, when the AUCFUEPIPCs of each of the four formulations were compared, significant differences were found: the AUCFUEPIPCs for the FOR1, FOR5 and FOR6 formulations were all significantly greater (P = 0·05) than that of the fresh conidial formulation (Table 5).

The effect of applying a fresh conidial formulation on E. nigrum presence on fruit surfaces (ALF02, VIN02, SAL02 and DOM02) was compared with that obtained after applying the FOR1, FOR7 and FOR8 formulations (ALF03, SUD03, MOZ03, ALF04, SUD04, ALF05 and SUD05) and the results are presented in Table 6. For all biological treatments on peach and nectarines trees, the AUNCEPIPCs and the AUCFUEPIPCs were significantly bigger than those for untreated peach trees (P = 0·01; Fig. 5 and Table 6). Treating trees three times at pit hardening was more effective in increasing the size of the E. nigrum population on peach and nectarine surfaces than one treatment at pit hardening: the AUNCEPIPCs and AUCFUEPIPCs on peach trees treated with an E. nigrum formulation three times at pit hardening were significantly bigger (P = 0·01) than those for trees treated only once. When the conidial formulations were applied at flowering, the population size of E. nigrum on peach surfaces did not increase (Table 6). Between formulations, the AUNCEPIPCs and AUCFUEPIPCs for those peach trees treated with the FOR7 conidial formulation were significantly bigger (P = 0·01) than those for trees treated with either the FOR1 or FOR8 conidial formulation (Table 6). The AUNCEPIPC for those peach trees treated with the fresh conidial formulation was significantly bigger (P = 0·05) than for those of trees treated with any of the other conidial formulations (Table 6). In contrast, the AUCFUEPIPC for those peach trees treated with the fresh conidial formulation was significantly smaller (P = 0·01) than those for trees treated with any of the other conidial formulations (Table 6). There were also significant differences between peach and nectarine trees on the effects of the biological treatments on E. nigrum (Table 6). The AUCFUEPIPCs for nectarines were significantly larger than those of peaches. The same level of statistical significance (P = 0·01) was also found for the AUCFUEPIPCs, for nectarines treated with 10⁷ conidia ml⁻¹ were compared with those for peaches treated with 10⁶ conidia ml⁻¹.

When AUNCEPIPC and AUCFUEPIPC were related to T, HR, npit, npre and D by applying linear regression analysis to data pooled from the 12 surveys, multiple equations resulted. From these equations, the following one was selected to best describe AUCFUEPIPC as a function of HR, npit and npre because it has the highest adjusted determination coefficient and the lowest Mallow’s Cp coefficient.

The SEs of the estimates were significant (P ≤ 0·05) and are shown in parentheses below the corresponding parameters. A statistically significant relationship was detected between the variables at 99% probability.

Biocontrol efficacy of Epicoccum nigrum formulations

Significant differences (P = 0·01) were found when the AUNCMONPCs for all biological treatments were compared with that for the untreated control group, namely trees treated with either fungicides (at flowering, pit hardening and 21 days before harvest) or not treated with either a fungicide or the conidial formulation (Table 6). When the AUNCMONPCs for fungicide applications were compared with those for biological treatments, no significant differences were observed (Table 6). The smallest AUNCMONPCs on peach surfaces was found for those trees treated with the FOR7 conidial formulation, especially when this formulation was applied three times at pit hardening and more than three times in the last month before harvest (Table 6).