Figures
Abstract
Botrytis cinerea causes gray mold disease of strawberry (Fragaria ×ananassa) and is a globally important pathogen that causes fruit rot both in the field and after harvest. Commercial strawberry production involves the use of plastic mulches made from non-degradable polyethylene (PE), with weedmat made from woven PE and soil-biodegradable plastic mulch (BDM) as emerging mulch technologies that may enhance sustainable production. Little is known regarding how these plastic mulches impact splash dispersal of B. cinerea conidia. The objective of this study was to investigate splash dispersal dynamics of B. cinerea when exposed to various plastic mulch surfaces. Mulch surface physical characteristics and conidial splash dispersal patterns were evaluated for the three mulches. Micrographs revealed different surface characteristics that have the potential to influence splash dispersal: PE had a flat, smooth surface, whereas weedmat had large ridges and BDM had an embossed surface. Both PE mulch and BDM were impermeable to water whereas weedmat was semi-permeable. Results generated using an enclosed rain simulator system showed that as the horizontal distance from the inoculum source increased, the number of splash dispersed B. cinerea conidia captured per plate decreased for all mulch treatments. More than 50% and approximately 80% of the total number of dispersed conidia were found on plates 10 and 16 cm away from the inoculum source across all treatments, respectively. A significant correlation between the total and germinated conidia on plates across all mulch treatments was detected (P<0.01). Irrespective of distance from the inoculum source, embossed BDM facilitated higher total and germinated splashed conidia (P<0.001) compared to PE mulch and weedmat (P = 0.43 and P = 0.23, respectively), indicating BDM’s or embossed film’s potential for enhancing B. cinerea inoculum availability in strawberry production under plasticulture. However, differences in conidial concentrations observed among treatments were low and may not be pathologically relevant.
Citation: Wang X, Mattupalli C, Chastagner G, Tymon L, Wu Z, Jung S, et al. (2023) Physical characteristics of soil-biodegradable and nonbiodegradable plastic mulches impact conidial splash dispersal of Botrytis cinerea. PLoS ONE 18(5): e0285094. https://doi.org/10.1371/journal.pone.0285094
Editor: Abhay K. Pandey, Tocklai Tea Research Institute, INDIA
Received: November 9, 2022; Accepted: April 14, 2023; Published: May 8, 2023
Copyright: © 2023 Wang et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: This project was funded by the Washington State Department of Agriculture Specialty Crop Block Grant program (#K2863). Additional funding was provided by the Specialty Crops Research Initiative Award 2022-51181-38325 from the USDA National Institute of Food and Agriculture. Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect the view of the U.S. Department of Agriculture. Corresponding author, L.W.D., received the awards. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Fungi have evolved different mechanisms to aid dispersal of their propagules [1]. The main dispersal mechanisms of fungal plant pathogens are wind and rain splash [1–6]. Understanding dispersal mechanisms is important for limiting fungal propagule movement and reducing the spread of global diseases that threaten food security and plant survival [7]. Fungal plant pathogens belonging to phylum Ascomycota are capable of producing asexual spores known as conidia and these conidia are regarded as the major dispersal unit of many important plant pathogens [8]. Conidia can be dispersed over short distances (e.g., a few centimeters) by rain-generated splash droplets or over long distances (e.g., several kilometers) by wind [9]. Infected plant tissues are a source of conidial inoculum and conidia incorporated into water droplets can be transported to neighboring plants by rain or irrigation splash events [10]. Rain splash dispersal can become critical in the development of fungal disease epidemics when a large number of conidia are transported to neighboring plants [11,12]. Previous research has demonstrated there is variability in conidial dispersal numbers through rain splash [10,13,14] and this variability is attributed to multiple factors including the surfaces that splashed droplets are interacting with [9,15,16]. However, the overall role of surface characteristics on splash dispersal of plant pathogens is understudied and in need of further investigation to inform disease risk and associated management.
Gray mold is an important fungal disease caused by Botrytis cinerea that severely impacts yields in commercial strawberry (Fragaria ×ananassa) production systems [2,17,18]. The pathogen is found worldwide [3], produces abundant inoculum in the form of conidia [4], and is considered a primary pathogen responsible for significant yield losses of harvested strawberry across the world [17]. Yield losses due to this pathogen can exceed 80% in the absence of fungicides and under favorable environmental conditions [18]. The pathogen is currently considered hemibiotrophic [19] and can infect a variety of tissues including flowers and fruits (Fig 1). Ripe and green strawberry fruits are both susceptible to this pathogen in the field [20,21]. Botrytis is also regarded as a post-harvest pathogen because the disease can develop during transportation and storage [22,23]. Rain splash dispersal may cause severe secondary gray mold infections through the dispersal of conidia from infected flowers, buds, or fruits to susceptible tissues [11,24,25], especially in high humidity (>80%) environments or when rainfall occurs preharvest [26,27].
Infected strawberry fruit covered with gray mold (Botrytis cinerea) (A). Sporulating B. cinerea after inoculation of strawberry fruits with a SDHI (succinate dehydrogenase inhibitor; FRAC-7) fungicide-resistant isolate (B). Sporulation demonstrates the pathogenicity of the B. cinerea isolate used in the experiment. Photos by DeVetter (A) and Baral (B).
Commercial strawberry production is often done under plasticulture using non-degradable polyethylene (PE) mulch (Fig 2) [28,29]. PE mulch utilization promotes plant growth and increases yield primarily through weed suppression and optimization of soil temperatures [30–32]. Weedmat (a semi-permeable, woven, PE- or polypropylene-based geotextile) is an alternative to PE that is preferred by some farmers due to its durability [33] and potential to be re-used over multiple seasons. Soil-biodegradable plastic mulch (BDM) is perceived as a more sustainable alternative to PE and weedmat because it is designed to biodegrade in soils and consequently reduce plastic waste generation [34,35].
Photo by DeVetter.
Ground covers including mulches can vary in surface characteristics and impact splash dispersal of plant pathogens. For example, splash dispersal patterns of Colletotrichum acutatum (cause of anthracnose) and Phytophthora cactorum (cause of leather rot on strawberry fruits and Phytophthora crown and root rots) were highly related to ground cover physical characteristics [6,36,37]. Random roughness was determined to be one of the key physical features of ground cover surface microtopography that influence splash dispersal [38] with decreasing roughness increasing the magnitude of splash distance [36,39]. For example, smooth PE mulch generated more conidial splash dispersal of C. acutatum compared to a straw mulch [36]. However, the impact of groundcover random roughness is not consistent as a buckwheat (Fagopyrum esculentum) husk mulch increased gray mold incidence in strawberry compared to other mulches including black plastic, wood chips, and straw [40]. Other ground cover characteristics such as permeability and surface adhesion of water may impact the dynamics of pathogen splash dispersal. However, there are no recent studies on splash dispersal from mulch surfaces and a systematic understanding of B. cinerea conidial dispersal via rain splash in strawberry productions systems under plasticulture is lacking.
The objective of this study was to investigate splash dispersal dynamics of B. cinerea conidia with three plastic mulch surfaces: PE mulch, weedmat, and BDM. Specific sub-objectives were to: 1) Evaluate mulch surface physical characteristics that could impact splash dispersal dynamics and 2) Determine whether different plastic mulches affect conidial splash dispersal patterns of B. cinerea.
Materials and methods
Mulch surface physical characteristics
New, unweathered PE, weedmat, and embossed BDM mulches included in this study (Table 1) were the same products that had been used in a previous mulch experiment [41]. Surface microtopographic features of mulch samples were assessed using a digital microscope (Keyence VHX-7000) in the Composite Materials and Engineering Center at Washington State University (WSU) in Pullman, Washington, USA. Mulch permeability was characterized to assess how material properties affect infiltration of liquid involved in splash dispersal and subsequent conidial dispersal outcomes. For each mulch treatment, a large plastic Petri dish (14 cm in diameter) was filled with greenhouse growing medium (Steuber Promix BX General Purpose, Premier Tech Horticulture, Quakertown, PA, USA) and covered with one of the mulch treatments. A 20 cm-long square piece of mulch was secured tightly around the dish by taping the mulch to the bottom of the Petri dish and adjusting the mulch surface tension to 0.74 N at the center using a tension meter (Chatillon 516 Series linear push/pull scale, Ametek, Berwyn, PA, USA), which is comparable to mulch tension measurements collected in the field (S1 Table). A silicone ring (6 cm diameter and 1 cm high) was placed on the mulch surface in the center of the Petri dish and sealed to the mulch using a white, 100% silicone sealant (Gorilla Glue Company; Berlin, Germany). The sealant was allowed to dry completely (~30 min) before 10 ml of water was gently poured into the area within the silicone ring. The volume of remaining water within the silicone ring was measured after 2, 4, 8, 16, and 32 minutes at 25°C using a 10 ml-graduated cylinder. Permeability measurements were repeated three times as technical replicates for each mulch treatment using new mulch for each repetition.
Characterization of splash dynamics
To better understand splash dynamics of water on PE mulch, weedmat, and BDM surfaces, a preliminary non-replicated experiment using a single droplet was conducted. A water nozzle, syringe pump (NE-1000, Pump Systems LLC., Dickinson, ND, USA), and needle attached to the nozzle were used to generate individual water droplets falling onto either PE mulch, weedmat, or BDM surfaces. A high-speed camera (NOVA S6, Photron USA, Inc., San Diego, CA, USA) with a backlit LED light source was used to record the splash dynamics initiated by a second drop falling onto the former drop placed on mulch surfaces. The number of resulting satellite droplets as well as qualitative data describing splash dynamics were recorded.
Conidial splash dispersal of B. cinerea
This experiment was conducted in a fully enclosed screen house (15 m long, 6 m wide, 5.5 m tall with 270 × 770 μm mesh size; Gable Series 7500, U.S. Global Resources, Seattle, WA, USA) at the WSU Mount Vernon Northwestern Washington Research and Extension Center (WSU NWREC) in Mount Vernon, WA to minimize potential contamination. The ground of the screenhouse was covered with woven, polyethylene landscape fabric. Within the screenhouse an Eurmax Premium instant canopy tent with enclosed sidewalls (3 m long, 2.9 m wide, 2.3 m tall; Eurmax Canopy, EI Monte, CA, USA) was assembled to further minimize contamination potential and wind effects (Fig 3).
Three pulsating sprinklers were mounted on tripods and uniformity tests were conducted to adjust the arrangement of sprinklers (A). A canopy tent was established to enclose the experiment and block the potential for wind dispersal impacts (B). Three sprinkler tripods were arranged in a triangle shape at specific distances (C) and the mulch treatment was placed at the centroid of the triangle shape. At the center of the mulch treatment, a conidial suspension (2 ±0.2 ×106 conidia/ml, symbolized as blue dot in the picture where three orange arrows point) was placed on the mulch treatment to serve as the inoculum source; five Botrytis spore trap media plates (BSTM) were placed adjacent to each other and arranged in a line. The mulch treatment and BSTM plates were placed on the water absorbent mat (C). Photo by Wang.
A rain simulator system was built within the canopy tent consisting of three metal pulsating sprinklers (Hunter MP Rotator® MP800, Hunter Industries, San Marcos, CA, USA) mounted on tripods (Melnor 65066-AMZ; Melnor Inc., Frederick County, VA, USA; Fig 3). There was one sprinkler head per tripod and individual sprinkler heads were arranged in an equilateral triangle with each side measuring 1.8 m long between sprinkler heads (Fig 3C). Each sprinkler head consisted of a high-efficiency nozzle paired with a pop-up body (Hunter Pro-Spray PRS30, Hunter Industries, San Marcos, CA, USA) for pressure regulation to 207 KPa. A schedule 80 PVC nipple (Grainger Industrial Supply, Lake Forest, IL, USA) was used to connect the sprinkler head and irrigation pipes. Water was discharged from nozzles at a height of 1.5 m.
Uniformity tests were conducted prior to conidial dispersal experiments to position the nozzles, maximize head-to-head coverage, and optimize uniformity of water discharged from the rain simulator system (Fig 3A). Each uniformity test was performed for five minutes and repeated three times as technical replicates for each configuration until maximum uniformity was achieved. Water was collected into five empty Petri plates (10 cm diameter) placed in the central area where water was discharged by the rain simulator system. A total of 40 plates were used for each uniformity test with plates arranged in eight cardinal and ordinal directions (with 5 plates in each direction) radiating from the center (North, South, West, East, Northwest, Northeast, Southwest, and Southeast). Accumulated water was collected from each plate and measured at the end of each uniformity test. All three nozzles were adjusted to a 210˚ arc and a low flow rate (68 liter/hour). The central position of the sprinkler system was arranged so it was 1 m equidistant from the inoculum source (Fig 3B and 3C). The configuration of the rain simulator system used for subsequent conidial dispersal experiments achieved an 85–90% uniformity within 1.8 m (S2 Table), which included the area where the dispersal experiment was conducted. Additionally, multiple drops of water emitted from the rain simulator system were collected on water-sensitive paper. From this, the average size (i.e., diameter) of the drops was measured. Then 10, 20, 30, 40, 50, and 60ul of water were dropped individually on a separate piece of water-sensitive paper using a pipette. The size of the drops from the rain simulator system was compared to the drops containing a known volume of water to estimate average droplet volume from the rain simulator system. This approach allowed us to determine the average volume of a drop from the rain simulator, which was 50ul. Precipitation rate was also measured during uniformity tests and the pressure was adjusted to achieve a rate of 21 mm/hour.
Following establishment of the rain simulator system, a conidial suspension was prepared using a pathogenic, SDHI (succinate dehydrogenase inhibitor; FRAC-7) fungicide-resistant isolate of B. cinerea to distinguish it from other airborne Botrytis spores during this experiment (Fig 1B). This isolate was characterized as having the P225F mutation which confers resistance to all SDHI active ingredients (Kozhar, personal communication). To prepare the conidial suspension, the isolate was cultured on full-strength potato dextrose agar amended with 100 μg/ml salicylhydroxamic acid (SHAM), 32 μg/ml isofetamid, and 35 mg/ml chloramphenicol. Then B. cinerea culture plates were incubated at 22°C in an incubator (Percival Intellus Ultra temperature controller, Percival Scientific, Inc., Perry, IA, USA) with 18 hours of white light (light intensity 320 lux) and 6 hours dark for 10 days. Sterile deionized (DI) water (20 mL) was added onto 10-day-old B. cinerea culture plates and a lab spatula was used to dislodge conidia into the water solution. The solution was then filtered through one layer of cheesecloth and transferred into 50 ml centrifuge tubes. The conidial suspension was vortexed for 5 minutes to distribute clumped conidia and then the conidial concentration was determined using a hemocytometer (Bright-line Model 1492, Hausser Scientific, Horsham, PA, USA). The final conidial concentration was adjusted to 2 ±0.2 ×106 conidia/ml. The conidial germination rate prior to running the splash dispersal experiment was nearly 100% (S3 Table).
Small Petri dishes (60 mm diameter) containing Botrytis spore trap media (BSTM) (modified from [42]) was used to collect the conidia splashed from the inoculum source. BSTM consisted of 2 g glucose, 20 g bactoTM agar, 0.1 g NaNO3, 0.1 g K2HPO4, 0.2 g MgSO4⋅7H2O and 0.1 g KCl in one liter of DI water. The solution was autoclaved and cooled to 56°C before adding 2.5 g/l tannic acid and adjusting the pH to 4.5 with 1 mol/l NaOH. Next, 0.2 g/l chloramphenicol, 0.02 g/l pentachloronitrobenzene (PCNB), 0.02 g/l dithane, and 0.1 ml/l 12% fenarimol were added and the media was poured into Petri dishes.
Before running the rain simulator, a large piece of Pig® Absorbent Mat Pad (30 cm × 50 cm) was placed on the ground to limit secondary splash from outside of the primary inoculum source and each replicate was placed on the mat at a marked position within the center of the rain simulator system (Fig 3C). The absorbent mat was immediately replaced after each replicate. For each replicate, five of the BSTM plates were placed immediately adjacent to the Petri dish containing the inoculum and one of the three mulch treatments and arranged linearly in a southward direction (Fig 3C). This direction was based on uniformity results and maintained throughout the experiment for consistency. Each BSTM plate was immediately adjacent to the other, so the BSTM plate distance between each center of an adjacent BSTM plate was standardized at 6 cm. The center of the first BSTM plate was placed 10 cm away from the center of inoculum source and mulch treatment. Succeeding plates were placed at 16, 22, 28, and 34 cm from the center of the first plate. These distances from the inoculum source were assigned as dispersal distances (denoted as i, i = 10, 16, 22, 28, 34).
The experiment was repeated five times as technical replicates per mulch treatment with the construction of the mulch treatment done using the procedures outlined above for mulch permeability characterization. Additionally, after the silicone sealant dried, the annulus of the mulch surface was covered by a water absorbent mat (Pig® Absorbent Mat Pad, model MAT 203; New Pig Corporation, Tipton, PA) to reduce production of secondary splash droplets (Fig 3C). New mulch and absorbent mats were used for each replicate. The mulch surface and silicone band were then disinfected with 70% ethanol immediately before exposure to the rain simulator system. The B. cinerea conidial suspension (10 ml per replicate) was pipetted into the well created by the silicone ring and formed a pool on the treatment mulch. Immediately after the placement of BSTM plates and the addition of conidial suspension, the rain simulator system was initiated for two minutes with three people turning the irrigation on and off at the same time. BSTM plates were immediately collected and incubated at 22 ˚C for 18 hours with lids open in the same incubator listed above to dry the media surface and provide time for conidia to germinate. Prior to counting conidial germination, lids were placed back on the BSTM plates, and maintained at 1°C for a maximum of 4 hours while conidial counts were being performed to minimize changes in spore germination during the counting process. Presence of a germ tube (length was greater than half the width of the conidia) growing out of a conidium was used to assign germination status. All conidia from BSTM plates were counted within 3 hours after removal from the incubator. Total dispersed conidia (TCi) and the number of dispersed conidia that germinated after 18h (GCi) were enumerated for each dispersal distance (i) from each BSTM plate using a compound microscope (Nikon Eclipse 50i, Nikon Inc., Melville, NY, USA) at 40X magnification.
Data analysis
All quantitative data were analyzed using R Studio software (Version 1.4.1106, Rstudio PBC, Boston, MA, USA). Mulch permeability was graphed using ggplot2 package [43], while surface characteristics and splash dynamics were visually assessed. Findings from the non-replicated splash dynamics study were treated as qualitative observations and not analyzed due to the lack of replication. Conidial splash dispersal TCi data were first fit to an exponential model =
(1)
with N0 as the prefactor of the exponential model, which represents the number of conidia at the impact location (d = 0). τ is the decay length and N represents the number of conidia at distance d. Data are presented in their original units. Then two additional variables were calculated and added into further data analysis as follows:
The percentage (%) of TCi of total dispersed conidia at all distances [abbreviated as TibyT] =
(2)
The percentage (%) of GCi of germinated conidia at all distances [abbreviated as GibyG] =
(3)
Technical replicate was regarded as a random factor, while mulch treatment and dispersal distance were treated as fixed factors. Four variables (TCi, GCi, TibyT, and GibyG) were checked to ensure they met distribution assumptions and fitted into a generalized linear mixed-effects model. Normal distributions of those four variables were checked before depositing data into analysis using the his() function on residuals. According to different distribution patterns, TCi and GCi data were subjected to a Poisson distribution and log transformed, while TibyT and GibyG data were subjected to a normal distribution. An interaction between mulch treatment and dispersal distance was considered for these four variables. Multivariate Analysis of variance (MANOVA) was conducted using estimated marginal means (least-square means) with a post hoc Tukey–Kramer test for detection of treatment effects at a significance of α = 0.05. Data were presented in their original units. Additionally, a Pearson product-moment correlation coefficient analysis was conducted between each of TCi, GCi, TibyT, GibyG and dispersal distance, between TCi and GCi, and TibyT and GibyG. All data are available in Excel file S5.
Results and discussion
Splash dynamics and dispersal of B. cinerea conidia differed due to mulch treatment and distance with treatment effects attributed to differences in the physical characteristics of the mulch surfaces. PE mulch, weedmat, and embossed BDM showed different surface topographic characteristics (Fig 4A–4C). PE mulch had a smooth, flat surface compared to the embossed BDM. Weedmat mulch had a rigid surface due to the woven fibers of PE. Furthermore, PE mulch and embossed BDM were impermeable to water and weedmat was semi-permeable based on laboratory observations (Fig 5). There was no water loss through infiltration for PE mulch and embossed BDM, whereas >94% of the water infiltrated through the weedmat after 32 mins. Mulch permeability in the screenhouse was not measured given there was additional water delivered via the rain simulator system and water displacement occurring from rain droplets impacting the inoculum pool. Although the woven construction of weedmat seemed to produce visibly similar surface characteristics as embossed BDM, the semi-permeability of weedmat led to infiltration of water containing the inoculum through the mulch and effectively reduced the volume of the conidial suspension.
Surface microtopographic characteristics of polyethylene (PE; A), embossed soil-biodegradable plastic (BDM; B), and weedmat mulches (C). Scale bar denotes 500 μm. The number of dispersed B. cinerea conidia over distance is below (D-F). Blue lines are a model-fit using the function N = N0exp(−d/τ). Here N0 is the prefactor of the exponential model, which represents the number of conidia at the impact location (d = 0). τ is the decay length and N represents the number of conidia at distance d.
Results from the rain simulator system showed the means of all TCi across the five distances for PE mulch, weedmat, and embossed BDM were 26.8, 31, and 47 conidia, respectively. All mulch treatments showed an inverse relationship in that as the horizontal distance from the inoculum source increased, the number of TCi and GCi decreased (Table 2). Inverse relationships were also observed between TibyT, GibyG and horizontal distance from the inoculum source (Table 2). Interestingly, the TCi and GCi differed due to mulch treatment (P<0.001 for TCi and GCi,) and dispersal distance from the inoculum source (P<0.001) (Table 3). The TibyT and GibyG differed due to dispersal distance (P<0.001) but not due to mulch treatment (P = 0.97–0.98). There was an interaction between mulch treatment and dispersal distance (P = 0.006) for TCi, but not for GCi (P = 0.20), TibyT (P = 0.64), and GibyG (P = 0.88). The TCi and GCi splashed from embossed BDM were the highest, followed by PE and weedmat mulches. PE and weedmat had similar TCi and GCi (P = 0.43 and P = 0.23 respectively). The dispersal distance at 10 cm and 34 cm had the highest and lowest TCi and GCi, respectively, across all mulch treatments. GC28 equaled to zero for PE mulch and GC34 equaled to zero for PE and weedmat mulches. T10byT and G10byG was > 60% and T10byT and T16byT were approximately 80% across all mulch sources.
Note i represents dispersal distance (cm) with i = 10, 16, 22, 28, 34.
Values following the means are standard error.
Differences in conidia dispersal distance were likely due to mulch surface topographic characteristics as the number of conidia over distance followed a simple exponential decay model (Fig 4D–4F). The measured N0 [number of conidia at impact location (d = 0)] was 50.50 for PE mulch, 187.58 for embossed BDM, and 320.53 for weedmat. The decay lengths (τ) were 8.33 cm for PE mulch, 5.26 cm for embossed BDM, and 3.70 cm for weedmat. This indicates that as mulch roughness increases, more conidia spread with shorter horizontal distance (N0 increases and τ decreases). Physically, this may be observed as the rough surface of mulches create splashes more upward compared to a smooth surface.
Random roughness of groundcover surface microtopography has been determined to be a key physical feature that influences splash dispersal [36,38]. Preliminary observations from the non-replicated splash dynamics experiment showed that the second single drop falling onto an existing single drop on the weedmat and embossed BDM surfaces produced similar numbers of splashed satellite droplets (approximately 20–22), whereas PE produced fewer satellite droplets (approximately 15). Both PE and embossed BDM showed a well-connected splash crown upon secondary drop impact (Fig 6D and 6F). Furthermore, the rim of the crown ejected what appeared to be visually smaller splashed satellite droplets when interacting on the PE mulch surface than embossed BDM. Weedmat exhibited an irregular pattern of splashing (Fig 6F), and this was attributed to the large-scale roughness of the interwoven weedmat fibers. We also observed PE and embossed BDM surfaces produced a similar spherical adhesion to a single drop of water whereas the weedmat surface produced a flat shape (Fig 6A–6C), which may be due variations in mulch permeability (Fig 5) [44]. Although this was not a replicated trail, observations suggest that an irregular splashing pattern may produce a variable but sometimes large number of splashed satellite droplets. Additionally, the embossed surface of BDM and interwoven fibers of weedmat lead to relatively upward splash patterns and a greater number of droplets than the PE mulch and this may increase the number of conidia dispersed in satellite droplets but quickly decay horizontally as described in the exponential model fit.
Splash dynamics of polyethylene (A and D), embossed soil-biodegradable plastic (B and E) and weedmat (C and F) mulches.
Regardless of mulch treatment, most B. cinerea conidia were dispersed within 10 cm from the inoculum source in this study. Conidia captured on media plates decreased greatly from 10 to 22 cm and were nearly absent at 34 cm from the inoculum source. A similar trend was observed for GCi. It is also to be noted that there was a highly significant linear correlation between TCi and GCi, as well as between TibyT and GibyG (Table 2). Results from this study align well with previous research on plant pathogen spread by splash dispersal in rain-simulated systems. Previous work has demonstrated that the majority of splashed droplets are dispersed over short distances and larger droplets containing dozens to hundreds of conidia fall within 30 cm from inoculum sources [10,36,45–49]. Perryman et al. [10] also found larger splashed droplets containing an average of 308 conidia/droplet of Pyllosticta citricarpa (cause of Citrus black spot) were dispersed within 30 cm from the inoculum source while smaller splash droplets can be dispersed further and as far as 70 cm. Findings from our study are also in agreement with the general understanding that rain splash dispersal of fungal plant pathogens occurs at smaller scales (e.g., centimeters) compared to wind dispersal, which can disperse propagules several meters or even kilometers from an inoculum source [50–53].
The conidial density of B. cinerea necessary for infection and lesion development on strawberry flowers and fruits as well as the splash dispersal distance of water droplets with conidial concentrations needed for infection are currently unknown. Previous inoculation studies on grape (Vitis vinifera) berry surfaces showed an individual airborne conidium could infect berry tissue and lead to lesion development, but overall infection was governed by host resistance that varied between fresh and cold-stored fruits [54]. For strawberry, a B. cinerea concentration of 3×103 conidia/m3 in air caused only 0.56% Botrytis fruit rot [55]. Studies demonstrating the density of B. cinerea needed in a liquid suspension to cause disease are limited. In lentil (Lens culinaris) seedlings, a Botrytis spp. suspension at 104 spores/ml concentration caused a 20% chance of infection whereas a 102 spores/ml concentration did not result in any symptom development [56]. The concentration of our suspension was 2 ±0.2 ×106 conidia/ml, which may be sufficient to cause disease. Hence, although our splash dispersal results showed a mulch treatment effect, whether the greater conidial numbers generated from embossed BDM cause more severe disease needs to be demonstrated using field-relevant conidial concentrations in suspension and is an area for future research. Decreasing random roughness (defined as the standard deviation of elevations over an area of interest from a base plane) [38] of ground covers, including mulches, has been previously reported to increase the magnitude of pathogen splash distance [6,24,57,58]. In this study, mulch random roughness was not measured, but visual assessment of mulch surface microtopography indicated PE to have lower mulch roughness than embossed BDM and weedmat due to their respective embossed and ridged surface features. PE can also be made with an embossed surface, so it is still an open question if similar responses will be observed between PE and BDM depending on whether or not they are embossed. Despite this variation in surface microtopography, all mulch treatments in the current study were associated with a decrease in the number of dispersed conidia with increasing horizontal dispersal distance. Furthermore, other factors such as the degree, uniformity, and scale of roughness may contribute to the magnitude of splash dispersal of pathogens.
It is to be noted that the assayed dispersal distances under our experimental conditions was designed on a small scale to reflect the range of typical plant spacings (25 to 30 cm) observed in commercial strawberry systems [23,59]. Our results suggest that wider within-row strawberry plant spacing could reduce the spread of B. cinerea conidia transported through splash dispersal. Wider within-row plant spacing has been shown to reduce disease incidence of Botrytis in field-grown strawberry in Florida with 40% less disease at a wider spacing of 45.6 cm compared to 22.9 cm [60]. Other splash dispersal studies similarly found plant density can affect splash dispersal of C. acutatum in strawberry [61]. Although wider spacing can be beneficial for reducing plant pathogens spread by rain splash and reducing the period of free moisture on plants, there is a trade-off as this effectively reduces plant density and yield potential by having fewer plants per unit area.
Rain characteristics, such as rain density, duration, and droplet size are also known to impact pathogen dispersal distance [12,24,36,39,62]. Pathogens could furthermore have different splash dispersal distances due to different density and/or sizes of pathogen spores incorporated within a droplet [6,10,24,39]. However, studies on splash dispersal epidemiology of fungal plant pathogens are few and there are many unexplored aspects that remain understudied. It is also unclear if the feedstock ingredients of mulches contribute to differences in surface characteristics and how they interact with water, which could impact splash dispersal. BDMs are comprised of 75–95% biodegradable feedstock ingredients with the remainder being additives such as plasticizers, fillers, colorants, pigments, UV-stabilizers, antioxidants, nucleating agents, and antibacterial additives [28]. The PE and weedmat mulches used in this study were both made of PE feedstock whereas the embossed BDM feedstock comprised of PLA (polylactic acid) and PBAT (polybutylene adipate-co-terephthalate). If ingredients differ in hydrophobicity or hydrophilicity, this could in turn influence the interaction between water droplets and resultant splash dispersal. In addition, different plastic mulches have varying surface characteristics due to their manufacturing (i.e., film vs. woven material vs. embossing), which could impact splash dispersal of pathogen spores and disease risk of fungal pathogens such as B. cinerea.
In conclusion, physical characteristics including permeability, surface physical features, and splashing dynamics of PE mulch, weedmat, and embossed BDM can impact splash dispersal dynamics of B. cinerea. Differences in the number of conidia dispersed among PE mulch, weedmat, and embossed BDM was observed. The embossed BDM facilitated greater dispersal of splashed conidia compared to PE mulch and weedmat. Based on these results, mulch physical features created by embossing may contribute to enhanced B. cinerea inoculum availability in strawberry production under plasticulture. However, the conidial concentrations observed among treatments were overall low and differences in splashed conidia may not be great enough to elicit differential responses in disease incidence at field scale. Future studies should focus on translating these differences into potential real-world risks within commercial strawberry fields for B. cinerea as well as other pathogens dispersed through rain splash.
Supporting information
S1 Table. Mulch tension measurements in a mulched ‘Albion’ strawberry field in Mount Vernon, WA.
https://doi.org/10.1371/journal.pone.0285094.s001
(DOCX)
S2 Table. Uniformity tests for standardizing rain simulator system before running splash dispersal experiments.
https://doi.org/10.1371/journal.pone.0285094.s002
(DOCX)
S3 Table. Assessment of conidial germination rates prior to running the splash dispersal Experiment.
https://doi.org/10.1371/journal.pone.0285094.s003
(DOCX)
S4 Table. Germination rate of all dispersed Botrytis cinerea conidia recovered from the rain splash dispersal experiment.
https://doi.org/10.1371/journal.pone.0285094.s004
(DOCX)
Acknowledgments
We are thankful to Emma Rogers, Brenda Madrid, Rachel Schick, Dakota McFadden, and Roshani Baral for their help with application of the rain simulator system and data collection. We also would like to thank David DeVetter for input on rain simulator design.
References
- 1. Magyar D, Vass MM, Li DW. Dispersal strategies of microfungi. In: Li DW, editor. Biology of microfungi. Fungal Biology. Springer, Cham; 2016. pp. 315–372.
- 2. Kozhar O, Larsen MM, Grunwald NJ, Peever, TL. Fungal evolution in anthropogenic environments: Botrytis cinerea populations infecting small fruit hosts in the Pacific Northwest rapidly adapt to human-induced selection pressures. Appl Environ Microbiol. 2020;86: e02908–19. pmid:32086310
- 3.
Carisse O. Epidemiology and aerobiology of Botrytis spp. In: Fillinger S, Elad Y, editors. Botrytis–the fungus, the pathogen and its management in agricultural systems. Springer, Cham; 2016. pp.127–148. https://doi.org/10.1007/978-3-319-23371-0_7
- 4.
Jarvis WR. Taxonomy. In: Coley-Smith JR, Verhoeff K, Jarvis WR, editors. The biology of Botrytis. Academic, London; 1980. pp.1–18.
- 5.
Jarvis WR. Epidemiology. In: Coley-Smith JR, Verhoeff K, Jarvis WR, editors. The biology of Botrytis. Academic, London; 1980. pp. 219–250.
- 6. Madden LV. Aggregation of Colletotrichum acutatum in response to simulated rain episodes. J Phytopathol. 1993;138: 145–156.
- 7. Fisher MC, Henk DA, Briggs CJ, Brownstein JS, Madoff LC, McCraw SL, et al. Emerging fungal threats to animal, plant and ecosystem health. Nature. 2012;484: 186–194. pmid:22498624
- 8.
Bärlocher F. Research on aquatic hyphomycetes: historical background and overview. In: Bärlocher F, editor. The ecology of aquatic hyphomycetes. Berlin: Springer; 1992. vol. 94, pp.1–15. https://doi.org/10.1007/978-3-642-76855-2_1
- 9. Fitt BDL, McCartney HA, Walklate PJ. Role of rain in the dispersal of pathogen inoculum. Annu Rev Phytopathol. 1989;27:241–270.
- 10.
Perryman SAM, West JS. Splash dispersal of Pyllosticta citricarpa conidia from infected citrus. EFSA Supporting Publication. 2014;11(2): EN–560. https://doi.org/10.2903/sp.efsa.2014.EN-560
- 11. Madden LV, Wilson LL, Ellis MA. Field spread of anthracnose fruit rot of strawberry in relation to ground cover and ambient weather conditions. Plant Dis. 1993;77: 861–866.
- 12. Gilet T, Bourouiba L. Rain-induced ejection of pathogens from leaves: revisiting the hypothesis of splash-on-film using high-speed visualization. Integr Comp Biol. 2014;54(6): 974–984. pmid:25326426
- 13. Gregory PH, Guthrie EJ, Bunce ME. Experiments on splash dispersal of fungus spores. J Gen Microbiol. 1959;20: 328–354. pmid:13654728
- 14. Brennan RM, Fitt BDL, Taylor GS, Colhoun J. Dispersal of Septoria nodorum pycnidiospores by simulated rain and wind. J Phytopathol. 1986;112: 291–297.
- 15. Ntahimpera N, Wilson LL, Ellis MA, Madden LV. Comparison of rain effects on splash dispersal of three Colletotrichum species infecting strawberry. Phytopathology.1999;89(7): 555–563. pmid:18944690
- 16. Ntahimpera N, Hacker JK, Wilson LL, Hall FR, Madden LW. Characterization of splash droplets from different surfaces with a phase doppler particle analyzer. Agric For Meteorol. 1999;97: 9–19.
- 17. Petrasch S, Knapp SJ, van Kan JAL, Blanco‐Ulate B. Grey mould of strawberry, a devastating disease caused by the ubiquitous necrotrophic fungal pathogen Botrytis cinerea. Mol Plant Pathol. 2019;20(6): 877–92. pmid:30945788
- 18.
Ries SM. Gray mold of strawberry. University of Illinois IPM Reports. 1995;RPD No. 704. Available from http://ipm.illinois.edu/diseases/series700/rpd704/.
- 19. Veloso J, van Kan JAL. Many shades of grey in Botrytis-host plant interactions. Trends Plant Sci. 2018;23(7): 613–622. pmid:29724660
- 20. Agudelo-Rommero P, Erban A, Rego C, Carbonell-Bejerano P, Nascimento T, Sousa L, et al. Transcriptome and metabolome reprogramming in Vitis vinifera cv. Trincadeira berries upon infection with Botrytis cinerea. J Exp Bot. 2015;66(7): 1769–1785. pmid:25675955
- 21. Rhouma A, Hajji-Hedfi L, Ben Othmen S, Kumari Shah K, Matrood AAA, Okon OG et al. Strawberry grey mould, a devastating disease caused by the airborne fungal pathogen Botrytis cinerea. Egypt J Phytopathol. 2022;50(2): 44–50.
- 22. Bristow PR, McNicol RJ, Williamson B. Infection of strawberry flowers by Botrytis cinerea and its relevance to grey mould development. Ann Appl Biol. 1986;109: 545–554.
- 23. Erhardt W, Walters T. Growing day-neutral strawberries in western Washington. Washington State University Extension Factsheet. 2014;FS132E. Available from https://research.wsulibs.wsu.edu/xmlui/bitstream/handle/2376/4912/FS132E.pdf?sequence=2.
- 24. Madden LV, Yang X, Wilson LL. Effects of rain intensity on splash dispersal of Colletotrichum acutatum. Phytopathology. 1996;86: 864–874.
- 25. Ntahimpera N, Madden LV, Wilson LL. Effect of rain distribution alteration on splash dispersal of Colletotrichum acutatum. Phytopathology. 1997;87(6): 649–655. pmid:18945084
- 26. Jarvis WR. The infection of strawberry and raspberry fruits by Botrytis cinerea Fr. Ann Appl Biol. 1962;50(3): 569–75.
- 27. Jarvis WR. The dispersal of spores of Botrytis cinerea Fr. in a raspberry plantation. Trans Br Mycol Soc. 1962;45: 549–559.
- 28. DeVetter LW, Goldberger JR, Miles C, Gomez J. Grower acceptance of new end-of-life management strategies for plastic mulch in strawberry systems. Acta Hortic. 2021;1309: 659–662.
- 29. Samtani JB, Rom CR, Friedrich H, Fennimore SA, Finn CE, Petran Aet al. The status and future of the strawberry industry in the United States. Hort Tech. 2019;29: 11–24.
- 30. DeVetter LW, Zhang H, Ghimire S, Watkinson S, Miles CA. Plastic biodegradable mulches reduce weeds and promote crop growth in day-neutral strawberry in western Washington. HortScience. 2017;52(12): 1700–1706.
- 31. Fernandez GE, Butler LM, Louws FJ. Strawberry growth and development in an annual plasticulture system. HortScience. 2001;36(7): 1219–1223.
- 32. Freeman S, Gnayem N. Use of plasticulture for strawberry plant production. Small Fruits Rev. 2005; 4(1): 21–32.
- 33. Fernandez-Salvador J, Chernoh E, Bobo-Shisler C. Evaluating the state of the Oregon strawberry industry in 2019. Acta Hortic. 2021;1309: 387–394.
- 34. Miles CA, DeVetter L, Ghimire S. Suitability of biodegradable plastic mulches for organic and sustainable agricultural production systems. HortScience. 2017;52: 10–15.
- 35.
European Norms (EN). Biodegradable mulch films for use in agriculture and horticulture requirements and test methods. EN-17033. 2018; Brussels, Belgium.
- 36. Yang X, Madden LV, Wilson LL, Ellis MA. Effects of surface topography and rain intensity on splash dispersal of Colletotrichum acutatum. Phytopathology. 1990;80: 1115–1120.
- 37. Yang X, Madden LV, Reichard DL, Wilson LL, Ellis MA. Splash dispersal of Colletotrichum acutatum and Phytophthora cactorum from strawberry fruit by single drop impactions. Phytopathology. 1992;82: 332–340.
- 38. Currence HD, Lovely WG. The analysis of soil surface roughness. Trans Am Soc Agric. Eng. 1970;13: 710–714.
- 39. Madden LV. Effects of rain on splash dispersal of fungal pathogens. Can J Plant Pathol. 1997;19: 225–230.
- 40. Prokkola S, Kivijarvi P, Parikka P. Effects of biological sprays, mulching materials, and irrigation methods on grey mould in organic strawberry production. Acta Hort. 2003;626: 169–176.
- 41. Wang X, Shrestha S, Tymon L, Zhang H, Miles C, DeVetter L. Soil-biodegradable mulch is an alternative to non-biodegradable plastic mulches in a strawberry-lettuce double-cropping system. Front. Sustain. Food Syst. 2003;6: 942645.
- 42. Edwards SG, Seddon B. Selective media for the specific isolation and enumeration of Botrytis cinerea conidia. Lett Appl Microbiol. 2001;32: 63–66. pmid:11169044
- 43.
Wickham H. ggplot2: Elegant graphics for data analysis. Springer-Verlag New York. 2016.
- 44. Sen U, Roy T, Chatterjee S, Ganguly R, Megaridis CM. Post-impact behavior of a droplet impacting on a permeable metal mesh with a sharp wettability step. Langmuir. 2019;35: 12711–12721. pmid:31499000
- 45. Penet L, Guyader S, Pétro D, Salles M, Bussière F. Direct splash dispersal prevails over indirect and subsequent spread during rains in Colletotrichum gloeosporioides infecting yams. PLoS ONE. 2014;9(12): e115757. pmid:25532124
- 46. MacDonald OC, McCartney HA. Calculation of splash droplet trajectories. Agric For Meteorol. 1987;39: 95–110.
- 47.
McCartney HA, Fitt BDL. Construction of dispersal models. In: Gilligan CA, editor. Mathematical modelling of crop disease. Advances in Plant Pathology. London: Academic; 1985. pp. 107–143.
- 48. Fitt BDL, McCartney HA, Creighton NF, Laceyp ME, Walklate J. Dispersal of Rhyncosporium secalis conidia from infected barley leaves or straw by simulated rain. Ann Appl Biol. 1988;112: 49–59.
- 49. Yang X, Madden LV, Reichard DL, Fox RD, Ellis MA. Motion analysis of drop impaction on a strawberry surface. Agric For Meteorol. 1991;56: 67–92.
- 50. Ooka JJ, Kommedahl T. Wind and rain dispersal of Fusarium moniliforme in corn fields. Phytopathology. 1977;67: 1023–1026.
- 51.
Fitt BDL, McCartney HA. Spore dispersal in splash droplets, in water, fungi and plants. In: Ayres PG, Boddy L, editors. Cambridge: Cambridge University Press; 1986. pp.87–104.
- 52. McCartney HA. Dispersal of spores and pollen from crops. Grana. 1994;33: 76–80.
- 53. Pietravalle S, van den Bosch F, Welham SJ, Parker SR, Lovell DJ. Modelling of rain splash trajectories and prediction of rain splash height. Agric For Meterorol. 2001;109: 171–185.
- 54. Coertze S, Holz G. Surface colonization, penetration, and lesion formation on grapes inoculated fresh or after cold storage with single airborne conidia of Botrytis cinerea. Plant Dis. 1999;83(10):917–924. pmid:30841073
- 55. Blanco C, de Santos BL, Romero F. Relationship between concentrations of Botrytis cinerea conidia in air, environmental conditions, and the incidence of grey mould in strawberry flowers and fruits. Eur J Plant Pathol. 2006;114: 415–425.
- 56. Davidson JA, Krysinska-Kaczmarek M. Effects of inoculum concentration, temperature, plant age and interrupted wetness on infection of lentil (Lens culinaris) by Botrytis spp. conidia. Australas Plant Pathol. 2007;36: 389–396.
- 57. Yang X, Wilson LL, Madden LV, Ellis MA. Rain splash dispersal of Colletotrichum acutatum from infected strawberry fruit. Phytopathology. 1990;80: 590–595.
- 58. Yang X, Madden LV. Effects of ground cover, rain intensity and strawberry plants on splash of simulated raindrops. Agric For Meteorol. 1993;65: 1–20.
- 59. Bolda MP, Tourte L, Klonsky KM, De Moura RL. Sample costs to produce organic strawberries. University of California Cooperative Extension. 2014;ST-CC-10. Available from https://ucanr.edu/sites/uccesc/files/210231.pdf.
- 60. Legard DE, Xiao CL, Mertely JC, Chandler CK. Effects of plant spacing and cultivar on incidence of Botrytis fruit rot in annual strawberry. Plant Dis. 2000;84(5): 531–538. pmid:30841344
- 61. Boudreau MA, Madden LV. Effect of strawberry density on dispersal of Colletotrichum acutatum by simulated rain. Phytopathology. 1995;85: 934–941.
- 62. Madden LV, Wilson LL, Yang X, Ellis MA. Splash dispersal of Colletotrichum acutatum and Phytophthora cactorum by short-duration simulated rains. Plant Pathol. 1992;41: 427–436.