Abstract
A wide range of water-treatment technologies is used to control waterborne microbial problems in greenhouse and nursery irrigation. An online modified Delphi survey was carried out to identify the perceived key attributes that growers should consider when selecting among water-treatment technologies and to characterize a list of 14 technologies based on those same attributes. The expert panel consisted of ornamental plant growers (n = 43), water-treatment industry suppliers (n = 28), and research and extension faculty (n = 34). The survey was delivered to the expert panel in two rounds. Response rate was 59% and 60% for the first and second rounds, respectively. Growers identified control of plant disease, algae, and biofilm as primary reasons for adopting technologies, whereas mandatory regulation was not a major reason for adoption. All 23 attributes (related to cost, system size, control of microorganisms, chemistry, ease of use, and regulation) were perceived to be important when selecting between water-treatment technologies. Injectable sanitizing chemicals such as chlorination were considered to have low capital cost, unlike technologies that required installation of more complex equipment, such as heat treatment, hydrogen peroxide, ozone, reverse osmosis, or ultraviolet radiation. Filtration (excluding membrane filtration) was the only technology not perceived to be effective to control microorganisms. Filtration and copper were not considered effective to control human food-safety pathogens. Ozone was rated the highest as a technology that removes or oxidizes agrochemicals. Chemical water treatments, as opposed to physical water treatments, were perceived to be sensitive to water quality parameters and to have residual effect through the irrigation. Chlorine gas was perceived to be the only technology for which regulatory permission would be an obstacle. All technologies were perceived to be effective in water with low electrical conductivity (EC) or in solutions containing water-soluble fertilizers. This survey documents perceived attributes of water-treatment technologies, which are most useful where experimental data are not yet available. Research and outreach needs were highlighted by cases where perceived attributes differed from available experimental data or where there was a lack of consensus between experts.
Commercial greenhouse and nurseries use high water volumes, reportedly averaging over 22,000 gal/acre per day (Camberato and Lopez, 2010), and compete with other industries and drinking water for high quality sources. Irrigation water in greenhouse and nursery operations can be a source or dispersal mechanism for pathogens, algae, and biofilm, especially when recirculated and surface-water sources are used. A broad spectrum of plant pathogens, including oomycetes, fungi, bacteria, viruses, and nematodes, have been detected in water sources or distribution systems (Hong and Moorman, 2005). Algae and biofilm can also be problematic in greenhouse production by clogging emitters (Bucks et al., 1979; Dehghanisanij et al., 2005; Gilbert et al., 1981; Juanico et al., 1995), harboring plant pathogens (El-Hamalawi, 2007; Goldberg and Stanghellini, 1990; Hyder et al., 2009), and causing worker-safety issues such as slippery walkways.
Review articles on technical aspects of water treatment (Ehret et al., 2001; Fisher, 2013; Newman, 2004; Raviv and Lieth, 2007; Runia, 1995; Stewart-Wade, 2011; Van Os, 2010; Zhou and Smith, 2002) have shown there is incomplete information about the technologies and their interactions with chemical and physical characteristics of water. Considerable research has been undertaken by plant pathologists on the efficacy of water-treatment technologies for controlling waterborne pathogens (Raudales, 2013). However, investigation has focused largely on chlorination and control of oomycetes with less information available for other combinations of technologies and target organisms. There are limited data estimating phytotoxicity thresholds for chemical water disinfectants. However, the high levels of active ingredient required to control resistant pathogens suggest that in many cases the effective dose may cause phytotoxicity. Knowledge gaps remain about the compatibility of water-treatment technologies with the many biological, chemical, physical, social, and economic factors that characterize greenhouse and nursery operations. All these possible factors make it difficult for growers to select a technology that best fits their operation.
Horticulture growers are willing to adopt risk-management strategies to prevent disease spread in their operations (Breukers et al., 2012). However, risk and uncertainty about disease-management strategies affect decision making and represent a barrier to adoption (McRoberts et al., 2011). Horticulture growers identified uncertainties about efficacy and costs as some of the barriers for adoption of disease-management strategies (Breukers et al., 2012). Growers of ornamental crops indicated that a major obstacle to implement sustainable practices was the perception that the practices were not compatible with current operation systems (Dennis et al., 2010). Uncertainties about adopting a practice or technology lessen with time as growers acquire of information by experience or knowledge (Breukers et al., 2012; Feder and Umali, 1993; Marra et al., 2003; McRoberts et al., 2011; Rogers, 2003). Rogers (2003) described the “hardware” aspects of innovations as the technology itself and “software” aspects of innovations as the information or knowledge about the performance of the technology. Some researchers have proposed that models on adoption of innovations in agriculture should consider interrelationships associated with the innovation, such as changing chemical applications and fertilization practices when adopting a new plant variety (Feder, 1982; Smale and Heisey, 1993). Adoption of innovations in agriculture therefore requires consideration of additional factors other than the technology itself. Such factors can include operation size, adaptability to current infrastructure, perceived technical benefits, ease of use, costs, financial benefits, external incentives, and public policy among others (Adesina and Zinnah, 1993; Feder and Umali, 1993; Loo et al., 2012; Mangiafico et al., 2008; Padel, 2002; Rogers, 2003; Sunding and Zilberman, 2001).
Plant-disease management is a human decision-making process that would benefit from interdisciplinary research that combines social and natural sciences (Gent et al., 2011; McRoberts et al., 2011). Interdisciplinary research by plant pathologists, horticulturists, chemists, engineers, environmentalists, economists, and social scientists would be required to make a thorough analysis of all the factors that characterize water-treatment technologies for use in greenhouses and nurseries. However, given budget constraints on experimental research to resolve existing knowledge gaps, it is appropriate to use nonexperimental, but systematic, research designs.
The Delphi method involves soliciting a group of experts on an individual basis using a series of questionnaires to reach consensus about a problem for which quantitative data are not available or experimental-research designs are not feasible (Brown and Helmer, 1964; Dalkey and Helmer, 1963; Murray, 1979). In other words, the Delphi method is a structured communication process used to predict outcomes based on consensus of a group of experts (Mitroff and Turoff, 2002). The Delphi survey method is based on the epistemology of inexact sciences which convey quasi-laws based on past knowledge, experience, and critical analysis of the experts for prediction of a problem (Helmer and Rescher, 1959). Quasi-laws cannot be derived from simplified observations and are instead inferred by the experience and views of the experts of complex situations (Aligica and Herritt, 2009; Helmer and Rescher, 1959). A major distinction between the Delphi survey and other types of survey is that the expert panel has the opportunity to anonymously reconsider their answers based on their reaction to the group responses (Helmer and Rescher, 1959; Mitroff and Turoff, 2002). Selection of the expert panel involves nonprobability sampling as random selection of the experts is not possible given the demand of specific knowledge and experience (Hasson et al., 2000). Alternatives to deal with researcher bias on selection of the expert panel are self-rating of expertise (Brown and Helmer, 1964), peer nomination of experts (Murray, 1979), and selection based on indicators such as publications, research programs, or public presentations (Helmer and Rescher, 1959). Dalkey and Helmer (1963) designed the method as an anonymous survey to stimulate independent thinking by each of the participants. The Delphi method is carried out under anonymous conditions to prevent influence from other members of the panel, uneasiness to express opinion, and “bandwagon effect” in which all agree because it is the opinion of the majority (Helmer and Rescher, 1959). There is no standard definition of consensus, and some external factors (such as budget or time frame of project) might influence how consensus is defined. Consensus among the experts is defined a priori by the researchers with the goal of matching consensus to the objective of the research question (Hasson et al., 2000; Mitroff and Turoff, 2002; Sackman, 1974). The original Delphi method started with open-ended questions that lead to a list of parameters to be evaluated (Brown and Helmer, 1964). However, diverse modifications to the Delphi method have been developed to ease the process and to target more specifically attitudinal data (Mitroff and Turoff, 2002).
Specific objectives of this research were 1) to identify the perceived key attributes of water-treatment technologies that should be considered by growers when selecting a technology for their operation and 2) to characterize 14 available water-treatment technologies in terms of these attributes. A modified Delphi survey method was used to evaluate these objectives. Results of this survey were intended as a first step to develop a framework of selection criteria for water-treatment technologies in irrigation for control of waterborne pathogens, algae, and biofilm.
Materials and methods
The expert panel consisted of 43 growers, 28 industry suppliers, and 34 research and extension faculty (N = 105, divided into the three “expert types”). The “grower” group members consisted of ornamental plant growers known to have one or more water-treatment technologies in their operation and who were identified during cooperative extension water-treatment workshops hosted by the Water Education Alliance for Horticulture, which is a university/industry education consortium. The “industry” group consisted of suppliers and consultants from companies with products or services focused on the design, installation, and maintenance of water-treatment technologies. The “researcher” group represented scientists who had published articles, had known experience on water treatment, or both. The names of the individuals who participated and their association remained anonymous throughout the process.
A list of 23 “attributes” (shown in Table 1) that were potentially important in technology selection was developed based on informal grower and industry feedback by surveying during cooperative extension workshops. Water-treatment technologies included calcium hypochlorite, chlorine dioxide, chlorine gas, copper ionization, copper sulfate, filtration (excluding membrane filtration), heat treatment and pasteurization, hydrogen peroxide, hypochlorous acid, ozone, reverse osmosis/membrane filtration, silver ionization, sodium hypochlorite, and ultraviolet radiation.
Perceptions of an expert panel solicited via a modified Delphi survey that each attribute listed is important when selecting between water-treatment technologies. The expert panel was composed of growers of ornamental crops, researchers, and water-treatment industry suppliers with direct experience in one or more water-treatment technologies for control of pathogens in irrigation. The mean rating is based on a Likert scale from 1 (strongly disagree) to 5 (strongly agree).
A draft survey was developed with online Qualtrics® software (Qualtrics Laboratories, Provo, UT). The survey was approved by the University of Florida Institutional Review Board 02 under the protocol no. 2010-U-1178. In a pilot study, this version of the survey was sent for clarification and verification to a subset of the expert panel of eight people from the three expert types. The survey was amended with minor changes based on the feedback received.
Round 1.
The revised survey was sent to expert panel members, who were contacted multiple times based on Dillman’s-tailored design method (Dillman et al., 2009). The first contact was via e-mail 2 weeks before sending the survey informing them about their selection as part of the expert panel and the upcoming survey. The survey was first sent via e-mail on 7 Jan. 2011. Over a period of 4 months, four reminders were sent to the expert panel including one postcard and three e-mails. Round 1 of the survey was closed on 4 May 2011. A booklet was sent to the expert panel in appreciation for completing round 1 of the survey, with notification that round 2 was forthcoming.
Round 1 of the survey consisted of several sections. The demographics section requested information on their job title, organization, years of experience with water-treatment technologies, and an open-ended question on that experience. Growers only were also surveyed as to their reasons for adopting water-treatment technologies, the area of crop production, and which technologies were currently in use at their operation. For objective 1, which aimed to identify important factors in technology selection, all expert types were asked to evaluate the importance of different attributes in the selection between water-treatment options. In this section, the expert panel rated their level of agreement (on a Likert scale from 1 = strongly disagree, 3 = neither agree or disagree, to 5 = strongly agree) that each of the 23 attributes listed was an important selection criterion. A third section asked the expert panel to self-rate their level of expertise for each of 14 technologies using a Likert scale from 1 = no knowledge, 2 = beginner, 3 = informed without direct experience, 4 = informed with direct working experience, to 5 = expert. A fourth section related to objective 2, which was to identify the attributes of each technology. The Qualtrics® software dynamically presented questions only on technologies for which a participant had a self-rated expertise level of at least “3 informed without direct experience.” The technology was rated by the survey participant based on a level of agreement (with a Likert scale from 1 = strongly disagree to 5 = strongly agree) that each of the attributes listed accurately described a specific technology. In round 1, open questions were also provided to allow comments about each of the attributes and technologies.
Round 2.
The survey was sent to the subgroup of the expert panel who answered the questionnaire in round 1. The round 2 questionnaire included only sections 3 and 4 described for round 1, and in this round the average expert panel response for each attribute from round 1 (“the round 1 response”) was also included. For any question where the participant agreed with the round 1 response, the participants were asked to leave this value unchanged. If a participant disagreed with any round 1 response, the participants were asked to indicate their own response on the Likert scale. Round 1 responses left unchanged were therefore assumed to indicate agreement by the participant.
Average rating and the percentage of agreement among experts were the two criteria used to determine the expert panel’s perception of the attributes that characterize water-treatment technologies. An average rating >3.0 indicated that the expert panel perceived the attribute to characterize the technology and an average rating ≤3.0 indicated that the attribute did not characterize the technology. Average rating of 3.0 has been used as point of acceptance or rejection by other researchers (Duffield, 1993). When fewer than 30% of respondents changed the rating between rounds, the average round 1 response was considered “consensus” of the expert panel, whereas changes by 30% or more of the respondents indicated “disagreement.” This threshold was based on the consensus level of 70% that has been previously used by other researchers (Loughlin and Moore, 1979; McKenna, 1994; Slaughter et al., 1999; Van Steenkiste et al., 2002).
Analysis of variance (ANOVA) was used (PROC GLM) in SAS (version 9.2; SAS Institute, Cary, NC) to analyze main and interaction effects of “expert type” and “attribute,” and the main effect of “years of experience” in objective 1 (importance of “attributes” in technology selection). For objective 2, to determine “attribute” values for each technology, ANOVA was run by “attribute” using PROC GLIMMIX (SAS version 9.2), with main and interaction effects of “expert type” and “technology,” and the main effect of “years of experience.” Tukey’s honestly significant difference test was used to compare means in all analyses at the P ≤ 0.05 level. Delphi surveys are typically analyzed with descriptive statistics (Hasson et al., 2000). However, inferential statistics in survey analysis provides more insights into the relationships between responses and rounds in surveys, including Delphi surveys (Hasson et al., 2000; Shah and Kalaian, 2009). Therefore, inferential statistics was used to analyze the data of this survey. The F test was used because it has been shown to be robust to interval data assumptions from Likert scales (Breukers et al., 2012; Carifio and Perla, 2007), including Delphi surveys with sample sizes larger than 30 experts (Kalaian and Kasim, 2012; Shah and Kalaian, 2009).
Results and discussion
Demographics
The response rate of the survey was 59% and 60% for rounds 1 and 2, respectively. The sample population that responded to round 1 included 27 grower, 15 industry, and 21 researcher experts, compared with 15 grower, 9 industry, and 14 researcher experts in round 2. The majority (54%) of the expert panel self-rated as having more than 10 years of experience, with 26% or 20% self-rated as having 5 to 10 years, or less than 5 years of experience, respectively. Silver ionization was the only technology where only two of the three expert types were represented (industry and researcher, n = 7 individuals), with no grower in the survey group with experience. Growers identified “to prevent plant diseases” (74%), “to control algae” (65%), “to control biofilm or clogging of equipment” (65%), and “as a response to plant loss” (35%) as “important reasons to adopt a water-treatment system in your operation.” No growers indicated “it is a mandatory regulation in my area” as a reason to adopt a water-treatment technology. The operation size of the growers who participated in the survey ranged from 2.5 to 125.0 acres of covered greenhouse, 2.5 to 50.0 acres of open field production area, and 2.5 to 150.0 acres of total production area. The areas of greenhouse and open area production in the grower group totaled 430.0 and 246.0 acres, respectively. The mean value for operation areas for greenhouse and open area were 20.5 and 17.6 acres, respectively. Growers indicated that the technologies currently used either alone or in combination with other technologies were copper (n = 9), chlorine (n = 4), activated peroxygen (n = 4), chlorine dioxide (n = 4), ozone (n = 2), reverse osmosis (n = 2), and rapid sand filtration (1).
Objective 1: Attributes for selection of water-treatment technologies
Attributes.
In choosing the attributes that should be considered when selecting water-treatment technologies, the perceived importance differed between attributes (P < 0.001) and there was an interaction between expert type and attribute (P < 0.001). There was no main effect of expert type (P = 0.244), although years of experience did affect response (P = 0.001) (Table 1).
All 23 attributes had an average rating above 3.0, a value meaning “neither strongly agree or disagree” that an attribute was important in the selection of water-treatment technologies. An average rating of 3.0 was previously set as the threshold for consideration in round 2, and all 23 attributes were therefore conserved (Table 1). The highest rating for the attribute “if residues are present, they are not toxic to plants” was not significantly different from 11 other attributes that included aspects related to efficacy against a range of target organisms, practical aspects of worker safety and operation, and regulatory considerations. Operating and capital costs were lower-ranked attributes. Two potential reasons to install water-treatment systems, “controls human food-safety pathogens” or “removes or oxidizes pesticides and herbicides,” were ranked lower than controlling plant pathogens, algae, and biofilm.
Expert types and experience.
The perceived importance of attributes associated with control of algae, biofilm, and human food-safety pathogens differed between expert types. However, all expert types strongly agreed that “controls plant pathogens” was an important attribute. Researcher group perception on the importance that a technology “controls algae” [mean ± sd (3.75 ± 0.48)] was lower than the opinion of grower (4.93 ± 0.72) and industry groups (4.50 ± 0.27). Industry’s perception of the importance that a technology “controls biofilm” (4.93 ± 0.27) was higher than growers (4.21 ± 0.88) and researchers group (3.86 ± 0.66). Growers’ perception on the importance that a technology “controls human food-safety pathogens” (3.21 ± 0.93) was lower than the perception of industry (3.92 ± 0.92) or researchers (3.94 ± 0.57). Lower rating of biofilm control and control of human food-safety pathogens by growers compared with other expert types presumably meant that growers perceived these problems to be of lower importance in their own production systems. The grower population was primarily producing ornamental rather than edible crops, and may therefore have been less sensitive to human food-safety pathogens than other expert types. Researchers had a lower opinion of the importance that a technology “has a residual effect through the entire irrigation system” (3.75 ± 0.68) compared with the growers’ opinion (4.38 ± 0.56). Researchers’ perceived importance that a technology is “effective at high water pH” (3.71 ± 0.61) was higher than the industry’s opinion (3.00 ± 1.04).
The overall perceived importance of attributes was greater for experts who self-rated as having either more than 10 years of experience or less than 5 years of experience (4.19 for both groups), compared with the group who self-rated as having the intermediate 5 to 10 years of experience (4.00). Despite a statistical significant effect of years of experience, there may not be a practical significance in this difference.
Objective 2: Comparison of attributes between technologies
In the next survey section, the rating scale indicated how strongly participants perceived that a particular attribute characterized each individual technology. In the ANOVA that evaluated effects of “technology,” “expert type,” and “years of experience” on attribute ratings, there were differences between technologies (P < 0.05) for all 23 attributes. For six attributes, ratings differed across technologies between expert types. Across technologies, industry considered that water treatment had a higher initial capital cost and was not as suitable for small operations or small volumes, compared with other expert types. Researchers had a higher perception that water-treatment technologies were able to control plant pathogens, biofilm, human food-safety pathogens, and removal of pesticides and herbicides compared with growers and industry. The highest rating for “Controls biofilm” and “Controls human food-safety pathogens was by researchers, followed by growers and industry. Silver ionization was removed from the analysis because of small sample size (n = 7).
Cost.
The water-treatment technologies calcium hypochlorite, chlorine dioxide, chlorine gas, copper sulfate, hydrogen peroxide, hypochlorous acid, and sodium hypochlorite were considered to have a perceived “low initial capital cost” based on an average rating above 3.0 (Table 2). All of these technologies require a fairly simple, low cost injection system to deliver the dosage (Parke and Fisher, 2012). Copper ionization, filtration, heat treatment, ozone, reverse osmosis, and ultraviolet radiation were not considered to have low initial capital cost, with average ratings below 3.0. Copper ionization was the only technology where at least 30% of experts disagreed in ratings of initial capital cost between rounds 1 and 2, indicating a lack of consensus on this attribute.
Comparison between technologies of attributes associated with cost, system size, and control of microorganisms. Ratings were provided by an expert panel, via a modified Delphi survey, who was asked to rate their level of agreement that each attribute characterized the listed technology. In a second round, the expert panel was asked to express their agreement with the average rating of round 1. Mean values are based on a Likert scale from 1 (strongly disagree) to 5 (strongly agree).
Calcium hypochlorite, chlorine dioxide, chlorine gas, copper ionization, copper sulfate, filtration, hypochlorous acid, and sodium hypochlorite had average ratings above 3.0 for “has low operating cost per volume of irrigation water” (Table 2), although more than 30% of experts disagreed on ratings of ozone and hydrogen peroxide between rounds 1 and 2. A preliminary cost analysis indicates that operation costs for the different technologies can range widely between ≤$0.25 and ≥$1.00 per 1000 gal of water treated (Fisher, 2013). Considering the large volumes of water used for irrigation, minor changes in operation costs could result in large annual costs. A cost comparison of ultraviolet radiation and chlorine dioxide for control of Phytophthora cinnamomi in nurseries in Australia indicated that the capital and operation costs of chlorine dioxide were higher than ultraviolet radiation when the daily water use ranged between 5000 and 100,000 L per day (Mebalds et al., 1996). However, the capital cost of chlorine dioxide did not change between 10,000 and 100,000 L of water per day, suggesting that chlorine dioxide may be more cost-effective for treating large volumes of water. No comprehensive studies that compare the cost of other water-treatment technologies for use in irrigation are available.
System size.
All treatments other than chlorine gas were considered to be “suitable for small operation sizes (under 2 acres) or low water volumes such as mist systems,” with an average rating above 3.0 (Table 2). However, there was no consensus (more than 30% of experts disagreed between rounds 1 and 2) regarding the suitability of chlorine dioxide, chlorine gas, hypochlorous acid, ozone, reverse osmosis, and ultraviolet radiation for small operations. Heat treatment, copper sulfate, and hydrogen peroxide were not considered to be suitable for large operations (Table 2). More than 30% of the expert panel disagreed on the ratings for the suitability of chlorine dioxide, hydrogen peroxide, and ozone for large operations.
Among the technologies being used by the grower group at the time of the survey, the only technologies used in small operations <2 acres was chlorine dioxide in stabilized packets with single stock tanks (a 1.3-acre operation) and copper ionization was used in a 2.5-acre operation. Copper ionization was the water-treatment technology used in the widest range of operation sizes (2.5 to 100 acres), followed by chlorine dioxide generators (9 to 100 acres), and sodium hypochlorite (25 to 100 acres). Although only 13% of growers indicated they were using nonmembrane filtration, it is likely that screen or other filters were in use in all operations. Calcium hypochlorite and chlorine gas were used in individual operations of 23 and 10 acre, respectively. Hydrogen peroxide, ozone, and reverse osmosis were used in operation sizes ranging between 10 and 25 acres, 25 and 50 acres, and 6 and 16 acres, respectively.
Control of microorganisms.
Filtration (excluding membrane filtration) was rated ≤3.0 for all the attributes related to control of microorganisms including plant pathogens, algae, biofilm, and human-safety pathogens (Table 2). All other technologies had mean ratings above 3.0 for controlling algae and plant pathogens. Both forms of copper and filtration had a mean rating ≤3.0 for the control of human food-safety pathogens. There was more than 30% disagreement among the expert panel on two or more attributes related to control of algae, plant pathogens, biofilm, or human food-safety organisms for the technologies chlorine gas, copper ionization, filtration, hydrogen peroxide, reverse osmosis, and ultraviolet radiation.
Membrane filtration with small pore size (<10 μm) or slow media filtration are the only filtration options for removal of microorganisms from water (Ehret et al., 2001; Stewart-Wade, 2011; Van Os, 2010). Multiple stages of membrane filtration were more efficient in removing plant parasitic nematodes than a single filtration stage in experimental research (Moens and Hendricks, 1992; Ohtani et al., 2000). Disagreement among the experts for ratings related to filtration (excluding membrane filtration) may have arisen because slow media filtration was not specifically included as a technology in the survey. In contrast to screen filters, slow media filters significantly decreased Fusarium oxysporum, pelargonium flower break virus, Phytophthora cinnamomi, Phytophthora cryptogea, Radopholus similis, and Xanthomonas campestris (Berkelmann et al., 1995; Calvo-Bado et al., 2003; Minuto et al., 2008; Wohanka, 1995; Wohanka et al, 1999; Van Os et al., 1999).
Copper ionization was rated 3.6 for the attribute “controls plant pathogens.” However, 46% of experts disagreed between rounds with this rating, indicating a lack of consensus. A high level of disagreement may have arisen because copper ionization requires extended contact times (≥1 h) at a typical dose of 1 to 2 mg·L−1 for effective control of plant pathogens. For example, Agrobacterium tumefaciens (Wohanka et al., 2009a), Clavibacter michiganensis (Wohanka and Fehres, 2007a), Erwinia carotovora (Wohanka et al., 2007), Ralstonia solanacearum (Wohanka and Fehres, 2007b), and Xanthomonas hortorum (Wohanka and Fehres 2006a) treated with 2 mg·L−1 required 24, 2, 1, 1, and 4 h, respectively, to achieve 90% mortality of pathogen propagules. In contrast, other combinations of lower doses or contact times were ineffective in controlling these pathogens, and certain pathogens (Fusarium oxysporum, Pythium aphanidermatum, and Trichoderma asperellum) were resistant to copper ionization treatment at 4 mg·L−1 with contact times ≥1 h (Wohanka and Fehres, 2006b, 2006c; Wohanka et al., 2009b). Extended contact times are unlikely to occur in commercial greenhouse and nursery operations, unless treatment is applied in storage tanks with long residual time.
All technologies other than filtration, reverse osmosis, and ultraviolet radiation were considered to be effective at controlling biofilm. Because biofilm resides inside irrigation distribution lines, the point-treatment nature of the three aforementioned technologies, and therefore the lack of residual effect, reduces their potential efficacy in controlling biofilm. Biofilms are a matrix of polymers with a complex community of microorganisms attached to a surface (Costerton et al., 1995). The metabolic behavior and structure of biofilms is highly dependent on the microbial composition of the biofilm (Stoodley et al., 2002). In general, microbes attached to surfaces are more resistant to control than planktonic organisms (Bois et al., 1997; LeChevallier et al., 1988). Treatment with 0.5 mg·L−1 chlorine reduced the density of coliform pathogens in the biofilm and resulted in release of biofilm particles, but it did not completely control biofilm (Morin et al., 1996).
All technologies except copper ionization and filtration were considered to be effective at controlling human food-safety pathogens (Table 2). Inactivation of human pathogens has been widely studied in drinking water (Maier et al., 2009). Irrigation water has been identified as a potential source of human food-safety pathogens (Alsanius et al., 2011; Gortáres-Moroyoqui et al., 2011; Ijabadeniyi et al., 2011; Lewis Ivey and Miller, 2013; Steele and Odumeru, 2004). Irrigation water treated with chlorine gas or chlorine dioxide (both at residual concentrations of 1 mg·L−1) in combination with slow sand filtration were not effective in reducing coliform bacteria to the U.S. Environmental Protection Agency (USEPA) required levels (Lewis Ivey and Miller, 2013). Control and monitoring of food-safety pathogens in irrigation will be necessary given increasing regulation to reduce the risk of waterborne contamination of fresh produce by human pathogens (U.S. Food and Drug Administration, 2013). More research is required to determine synergistic efficacy of control of plant pathogens and human pathogens.
Effect on nutrient or inorganic molecules.
Calcium hypochlorite, copper ionization, copper sulfate, filtration, hypochlorous acid, and ultraviolet radiation were rated ≤3.0 for the attribute “removes or oxidizes pesticides or herbicides” (Table 3). Ozone, which was given the highest rating (3.6) for this attribute, is commonly used to remove pesticides from natural water resources and drinking water supplies (Chiron et al., 2000). Ozone oxidizes directly or via hydroxyl free radicals, which is the strongest oxidizer form (Hoigné and Bader, 1976). Hydroxyl free radicals can result in complete mineralization of molecules (Chiron et al., 2000). Other technologies may be used alone or in combination with other technologies to reduce pesticide levels in irrigation water, including reverse osmosis, carbon filtration, chlorination (or other oxidizers), and ultraviolet radiation (Chian et al., 1975; Chiron et al., 2000; Kamel et al., 2009; Lafi and Al-Qodah, 2006; Legrini et al., 1993).
Comparison between technologies of attributes associated with sensitivity to water quality parameters, effects on nutrients or inorganic molecules, and residual effect. Ratings were provided by an expert panel, provided via a modified Delphi survey, who were asked to rate their level of agreement that each attribute characterized the listed technology. In a second round, the expert panel was asked to express their agreement with the average rating of round 1. Mean values are based on a Likert scale from 1 (strongly disagree) to 5 (strongly agree).
Calcium hypochlorite, copper sulfate, hydrogen peroxide, hypochlorous acids, ozone, reverse osmosis, and sodium hypochlorite were the only technologies with ratings ≤3 for the attribute “fertilizers including micronutrients are not affected.” Ozone can oxidize manganese, iron, and any other micronutrient chelates (Ohashi-Kaneko et al., 2009; Vanachter et al., 1988). Reverse osmosis removes most nutrients from irrigation water (Runia, 1995). Calcium hypochlorite, hydrogen peroxide, hypochlorous acid, and sodium hypochlorite were also rated ≤3.0, presumably because it is assumed that they are a strong enough oxidizers to affect organic chelates, although data are not available.
Sensitivity to water quality parameters.
Several technologies including calcium hypochlorite, chlorine dioxide, chlorine gas, copper ionization, copper sulfate, hydrogen peroxide, hypochlorous acid, and sodium hypochlorite had ratings below 3.0 for the attribute “is effective at high water pH” (Table 3). More than 30% of experts disagreed for ratings of calcium hypochlorite, chlorine gas, copper ionization, and sodium hypochlorite for this attribute. Hypochlorous acid, which is the active ingredient of calcium hypochlorite, chlorine gas, hypochlorous acid, and sodium hypochlorite, has decreased sanitizing activity as pH increases (Lang et al., 2008; White, 1992). Chlorine dioxide is effective at a wider pH range than other chlorine forms (Faust and Aly, 1983). Copper solubility also decreases with increasing pH (Lindsay, 1979), which could reduce efficacy of copper ionization and copper sulfate. All the technologies had ratings above 3.0 for “is effective with low EC water,” although there was no consensus for copper ionization. Higher EC facilitates electrolysis in solutions, and therefore ionization of copper or silver (Fischer et al., 2008). Copper ionization units currently used in horticulture differ in their ability to adjust to the changing EC that commonly occurs in greenhouses using water-soluble fertilizers, and the authors have frequently measured doses of copper lower than the target concentrations in commercial greenhouse installations.
Filtration (excluding membrane filtration), heat treatment and reverse osmosis were the only technologies rated above 3.0 for the attribute “effective in water with suspended organic and inorganic particles” (Table 3). More than 30% of the experts disagreed about the average rating of heat treatment and ultraviolet radiation. Chemical sanitizers are affected by the presence of organic particles in the solution because the particles exert chemical sanitizer demand, therefore reducing the residual active ingredient. for sanitizing water (Fisher et al., 2013). It is well known that ultraviolet radiation efficacy is affected by the presence of organic and inorganic particles that block the transmission of radiation to the target organism (Yiaosumi et al., 2005). One of the principle roles of filtration in water treatment is to reduce suspended solids and sanitizing agent demand (Maier et al., 2009), and survey results reinforce the importance of filtration as an essential component of any water treatment system design. In addition, the rating and the disagreement for ultraviolet radiation indicates the need to educate growers on water clarity and pretreatment when using ultraviolet radiation for sanitation.
All the technologies were rated ≥3.2 for the attribute “effective in solutions containing water-soluble fertilizers” (Table 3). However, in a study where chlorine interacted with ammonium, a reduction in oxidation reduction potential and free chlorine were observed (Meador and Fisher, 2013), which would be expected to occur because of conversion of hypochlorous acid (free chlorine) to chloramine (combined chlorine) (Deborde and von Gunten, 2008). It is unknown how chloramines affect the efficacy to control plant pathogens. Research on human health found chloramine was more effective than hypochlorous acid in reducing the incidence of Legionella and biofilm buildup (Lin et al., 2011); however, longer contact times are required (Degrémont, 1979).
Residual effect.
Filtration (excluding membrane filtration), heat treatment, ozone, reverse osmosis, and ultraviolet radiation were not considered to have the attribute “has a residual effect through the entire irrigation system” (Table 3). As discussed early in relation to biofilm control, most of these technologies are characterized as point treatments, in contrast with chemicals that are transported through the irrigation system (Stewart-Wade, 2011). Ozone is an unstable oxidant with fast reaction in water (Hoigné and Bader, 1994), which by definition reduces its residual activity.
Copper ionization, filtration (excluding membrane filtration), heat treatment, ozone, reverse osmosis, and ultraviolet radiation were perceived to have the attribute “if residues are present, they are not toxic to plants” (Table 3). However, more than 30% of the expert panel disagreed on the ratings for chlorine dioxide, copper ionization, heat treatment, ozone, reverse osmosis, sodium hypochlorite, and ultraviolet radiation for one of the attributes related to residual effect (Table 3). Chlorine applied at 2.4 mg·L−1 (Cayanan et al., 2009), chlorine dioxide at 2.5 mg·L−1 (Rens, 2011), ozone at 0.9 mg·L−1 (Graham et al., 2009), and hydrogen peroxide at 8 mg·L−1 (Vänninen and Koskula, 1998) have resulted in phytotoxicity of diverse crops.
Ease of use.
Copper sulfate, hypochlorous acid, ozone, and ultraviolet radiation were not considered to have the attribute “easy to monitor active ingredient and efficient operation” (Table 4). More than 30% of the expert panel disagreed with the ratings on ease of monitoring for copper ionization, heat treatment, hydrogen peroxide, and ozone. Inline and handheld devices and kits are available for onsite monitoring active ingredients of sanitizing agents (Fisher, 2013). Chlorine gas and sodium hypochlorite were the only technologies perceived as not being “safe for workers to use” (Table 4). Chlorine-based products have been categorized in Toxicity Category I, which is the highest degree of acute toxicity established by the USEPA, 1999). Chlorine dioxide, chlorine gas, ozone, and reverse osmosis were not considered to have the attribute “simple to use and does not require specialist training for workers” (Table 4). Chlorine gas, ozone, reverse osmosis, and ultraviolet radiation were not considered to have the attribute “requires minimal maintenance for calibration, cleaning, or replacing parts” (Table 4). All technologies were rated ≥3.0 for the attribute “has a long shelf life of materials (>3 months)” (Table 4).
Comparison between technologies of attributes associated with ease of use and regulation. Ratings were provided by an expert panel, via a modified Delphi survey, who were asked to rate their level of agreement that each attribute characterized the listed technology. In a second round, the expert panel was asked to express their agreement with the average rating of round 1.
Regulation.
Chlorine gas was the only technology not considered to have the attribute “regulatory permission is unlikely to be an obstacle” (Table 4). More than 30% of the expert panel disagreed with the ratings for the attribute “regulatory permission is unlikely to be an obstacle” for the technologies chlorine dioxide and chlorine gas (Table 4). USEPA under the National Primary Drinking Water Regulations created a list of disinfectants and disinfection byproduct contaminants, specifying the maximum contaminant level (MCL) for various sanitizers and their byproducts for public water sources (Tarver, 2008; USEPA, 2013). The MCL for disinfectants are 0.8 mg·L−1 for chlorine dioxide and 4 mg·L−1 for chlorine as annual average (Tarver, 2008; USEPA, 2013). The MCL for disinfectant byproducts are 0.01 mg·L−1 for bromate, 1.0 mg·L−1 for chlorite, 0.06 mg·L−1 for halo acetic acids (HAA5), and 0.10 mg·L−1 for total trihalomethanes (Tarver, 2008; USEPA, 2013). Copper, which is listed under the inorganic chemicals contaminant, has an action level of 1.3 mg·L−1 (USEPA, 2013). The action level “indicates the amount necessary to reduce the level of contaminant in drinking water” (Tarver, 2008).
Conclusions
The results of this survey confirmed that decision making for selection of water-treatment technologies is multifaceted. Hence, multidisciplinary approaches are required to accurately model the selection process of water-treatment technologies in irrigation. Disagreement among experts, and lack of published quantitative research data, highlight areas of needed research and outreach. For example, the expert panel disagreed on 11 of 24 attributes that characterized copper ionization and in 8 of 14 technologies for the attribute “controls biofilm.” The lack of consensus on attributes of copper ionization may result from limited research on this technology in the United States, given that most efficacy testing of copper ionization for control of plant pathogens has been carried out in Europe (Hopkins et al., 2009; Toppe and Thinggaard, 1998, 2000; Wohanka and Fehres, 2006a, 2006b, 2007a, 2007b; Wohanka et al., 2009a, 2009b). Technologies vary widely in the number of published efficacy studies (Raudales, 2013), with chlorine being the most studied technology. Aspects other than control of plant pathogens, such as control of biofilm or algae, or economic analysis of technology options, have received little attention despite the importance of these factors as identified by the survey panel.
More quantitative research on this topic will ensure that growers successfully meet future regulatory requirements to capture and recirculate water, and comply with food-safety regulations. Although the results of this survey do not substitute for quantitative experimental data, the information surveyed current perceived attributes of each technology. This analysis provided the necessary information for the development of a framework that will facilitate the decision-making process for growers by providing selection criteria of water-treatment technologies in irrigation for control of waterborne microbial problems. Growers and extension agents can identify the selection criteria that are priorities for the target operation and match these priorities with the strengths and weaknesses of each technology.
Units
Literature cited
Adesina, A.A. & Zinnah, M.M. 1993 Technology characteristics, farmer’s perceptions and adoption decisions: A Tobit model application in Sierra Leone Agr. Econ. 9 297 311
Aligica, P.D. & Herritt, R. 2009 Epistemology, social technology, and expert judgment: Olaf Helmer’s contribution to futures research Futures 41 253 259
Alsanius, B.W., Alam, M., Larsson, C., Rosberg, A.K., Ahrné, S., Molin, G. & Jensén, P. 2011 Decontamination of irrigation water under field conditions: Preliminary results Acta Hort. 922 61 66
Berkelmann, B., Wohanka, W. & Krezal, G. 1995 Transmission of pelargonium flower break virus (PFBV) by recirculating nutrient solutions with and without slow sand filtration Acta Hort. 382 256 262
Bois, F.Y., Fahmny, T., Block, J.C. & Gatel, D. 1997 Dynamic modeling of bacteria in a pilot drinking water distribution system Water Res. 31 3146 3156
Breukers, A., van Asseldonk, M., Bremmer, J. & Beekman, V. 2012 Understanding growers’ decisions to manage invasive pathogens at the farm level Phytopathology 102 609 619
Brown, B. & Helmer, O. 1964 Improving the reliability of estimates obtained from a consensus of experts. RAND Rpt. P3925, RAND Corp., Santa Monica, CA
Bucks, D.A., Nakayama, F.S. & Gilbert, R.G. 1979 Trickle irrigation water quality and preventive maintenance Agr. Water Mgt. 2 149 162
Calvo-Bado, L.A., Pettitt, T.R., Parsons, N., Petch, G.M., Morgan, J.A.W. & Whipps, J.M. 2003 Spatial and temporal analysis of the microbial community in slow sand filters used for treating horticultural irrigation water Appl. Environ. Microbiol. 69 2116 2125
Camberato, D.M. & Lopez, R.G. 2010 Commercial greenhouse and nursery production: Controlling algae in irrigation ponds. Purdue Ext. Publ. HO-247-W
Carifio, J. & Perla, R.J. 2007 Ten common misunderstandings, misconceptions, persistent myths and urban legends about Likert scales and Likert response formats and their antidotes J. Soc. Sci. 3 106 116
Cayanan, D.F., Dixon, M., Zheng, Y. & Llewellyn, J. 2009 Response of container-grown nursery plants to chlorine used to disinfest irrigation water HortScience 44 164 167
Chian, E.S., Bruce, W.N. & Fang, H.H. 1975 Removal of pesticides by reverse osmosis Environ. Sci. Technol. 9 52 59
Chiron, S., Fernandez-Alba, A., Rodriguez, A. & Garcia-Calvo, E. 2000 Pesticide chemical oxidation: State-of-the-art Water Res. 34 366 377
Costerton, J.W., Lewandowski, Z., Caldwell, D.E., Korber, D.R. & Lappin-Scott, H.M. 1995 Microbial biofilms Annu. Rev. Microbiol. 49 711 745
Dalkey, N. & Helmer, O. 1963 An experimental application of the Delphi method to the use of experts Mgt. Sci. 9 458 467
Deborde, M. & von Gunten, U. 2008 Reactions of chlorine with inorganic and organic compounds during water treatment-kinetics and mechanisms: A critical review Water Res. 42 13 51
Degrémont, S.A. 1979 Water treatment handbook. Halsted Press, New York, NY
Dehghanisanij, H., Yamamoto, T., Ould Ahmad, B., Fujiyama, H. & Miyamoto, K. 2005 The effect of chlorine on emitter clogging induced by algae and protozoa and the performance of drip irrigation Trans. Amer. Soc. Agr. Eng. 48 519 527
Dennis, J.H., Lopez, R.G., Behe, B.K., Hall, C.R., Yue, C. & Campbell, B.L. 2010 Sustainable production practices adopted by greenhouse and nursery plant growers HortScience 45 1232 1237
Dillman, D.A., Smyth, J.D. & Christian, L.M. 2009 Internet, mail and mixed-mode surveys: The tailored design method. 3rd ed. Wiley, Hoboken, NJ
Duffield, C. 1993 The Delphi technique: A comparison of results obtained using two expert panels Intl. J. Nurs. Stud. 30 227 237
Ehret, D.L., Alsanius, B., Wohanka, W., Menzies, J.G. & Utkhede, R. 2001 Disinfestation of recirculating nutrient solutions in greenhouse horticulture Agronomie 21 323 339
El-Hamalawi, Z.A. 2007 Attraction, acquisition, retention and spatiotemporal distribution of soilborne plant pathogenic fungi by shore flies Ann. Appl. Biol. 152 169 177
Faust, S.D. & Aly, O.M. 1983 Chemistry of water treatments. Butterworth Publ., Woburn, MA
Feder, G. 1982 Adoption of interrelated agricultural innovations: Complementarity and the impacts of risk, scale, and credit Amer. J. Agr. Econ. 64 94 101
Feder, G. & Umali, D.L. 1993 The adoption of agricultural innovations: A review Technol. Forecast. Soc. Change 43 215 239
Fischer, R., Fisher, P. & Frances, A. 2008 Water treatment with copper ionization GMPro 2008 Dec 18 21
Fisher P. 2013 Water treatment: A growers guide for nursery and greenhouse irrigation. Water Education Alliance for Horticulture, Gainesville, FL
Fisher, P., Mohammad-Pour, G., Haskell, D.W., Huang, J. & Meador, D. 2013 Water sanitizing agents such as chlorine and chlorine dioxide interact with peat substrate and suspended solids Acta Hort. 1013 279 284
Gent, D.H., De Wolf, E.D. & Pethybridge, S.J. 2011 Perceptions of risk, risk aversion, and barriers to adoption of decision support systems and integrated pest management: An introduction Phytopathology 101 640 643
Gilbert, R.G., Nakayama, F.S., Bucks, D.A., French, O.F. & Adamson, K.C. 1981 Trickle irrigation: Emitter clogging and other flow problems Agr. Water Mgt. 3 159 178
Goldberg, N.P. & Stanghellini, M.E. 1990 Ingestion-egestion and aerial transmission of Pythium aphanidermatum by shore flies (Ephydrinae: Scatella stagnalis) Phytopathology 80 777 779
Gortáres-Moroyoqui, P., Castro-Espinoza, L., Naranjo, J.E., Karpiscak, M.M., Freitas, R.J. & Gerba, C.P. 2011 Microbiological water quality in a large irrigation system: El Valle del Yaqui, Sonora México J. Environ. Sci. Health Part A 46 1708 1712
Graham, T., Zhang, P., Zheng, Y. & Dixon, M.A. 2009 Phytotoxicity of aqueous ozone on five container-grown nursery species HortScience 44 774 780
Hasson, F., Keeney, S. & McKenna, H. 2000 Research guidelines for the Delphi survey technique J. Adv. Nurs. 32 1008 1015
Helmer, O. & Rescher, N. 1959 On the epistemology of inexact sciences Mgt. Sci. 6 25 52
Hoigné, J. & Bader, H. 1976 The role of hydroxyl radical reactions in ozonation processes in aqueous solutions Water Res. 10 377 386
Hoigné, J. & Bader, H. 1994 Characterization of water quality criteria for ozonation processes. Part II: Lifetime of added ozone Ozone Sci. Eng. 16 121 134
Hong, C.X. & Moorman, G.W. 2005 Plant pathogens in irrigation water: Challenges and opportunities Crit. Rev. Plant Sci. 24 189 208
Hopkins, D.L., Thompson, C.M. & Lovic, B. 2009 Management of transplant house spread of Acidovorax avenae subsp. citrulli on cucurbits with bactericidal chemicals in irrigation water. Plant Health Prog. Bul. 0129
Hyder, N., Coffey, M.D. & Stanghellini, M.E. 2009 Viability of oomycete propagules following ingestion and excretion by fungus gnats, shore flies, and snails Plant Dis. 93 720 726
Ijabadeniyi, O.A., Debusho, L.K., Vanderlinde, M. & Buys, E.M. 2011 Irrigation water as a potential pre-harvest source of bacterial contamination of vegetables J. Food Saf. 31 452 461
Juanico, M., Azov, Y., Teltsch, B. & Shelf, G. 1995 Effect of effluent addition to a freshwater reservoir on the filter clogging capacity of irrigation water Water Res. 29 1695 1702
Kalaian, S.E. & Kasim, R.M. 2012 Terminating sequential Delphi survey data collection Pract. Assess., Res. Eval. 17 5 1 10
Kamel, A., Byrne, C., Vigo, C., Ferrari, J., Stafford, C., Verdi, G. & Hedrick, J. 2009 Oxidation of selected organophosphate pesticides during chlorination of simulated drinking water Water Res. 43 522 534
Lafi, W.K. & Al-Qodah, Z. 2006 Combined advanced oxidation and biological treatment processes for the removal of pesticides from aqueous solutions J. Hazard. Mater. 137 489 497
Lang, J.M., Rebits, B., Newman, S.E. & Tisserat, N. 2008 Monitoring mortality of Pythium zoospores in chlorinated water using oxidation reduction potential. Plant Health Prog. Bul. 0922
LeChevallier, M.W., Cawthon, C.D. & Lee, R.G. 1988 Inactivation of biofilm bacteria Appl. Environ. Microbiol. 54 2492 2499
Legrini, O., Oliveros, E. & Braun, A.M. 1993 Photochemical processes for water treatment Chem. Rev. 93 671 698
Lewis Ivey, M.L. & Miller, S.A. 2013 Assessing the efficacy of pre-harvest, chlorine-based sanitizers against human pathogen indicator microorganisms and Phytophthora capsici in non-recycled surface irrigation water Water Res. 47 4639 4651
Lin, Y.E., Stout, J.E. & Yu, V.K. 2011 Controlling Legionella in hospital drinking water: An evidence-based review of disinfection methods Infect. Control Hosp. Epidemiol. 32 166 173
Lindsay, W.L. 1979 Chemical equilibria in soils. 1st ed. Blackburn Press, Caldwell, NJ
Loo, S.L., Fane, A.G., Krantz, W.B. & Lim, T.T. 2012 Emergency water supply: A review of potential technologies and selection criteria Water Res. 46 3125 3151
Loughlin, K.G. & Moore, L.F. 1979 Using Delphi to achieve congruent objectives and activities in a pediatrics department Acad. Med. 54 101 106
Maier, R.M., Pepper, I.L. & Gerba, P. 2009 Environmental microbiology. 2nd ed. Elsevier Academic Press, Burlington, MA
Mangiafico, S.S., Newman, J.P., Mochizuki, M.J. & Zurawski, D. 2008 Adoption of sustainable practices to protect and conserve water resources in container nurseries with greenhouse facilities Acta Hort. 797 367 372
Marra, M., Pannell, D.J. & Ghadim, A.A. 2003 The economics of risk, uncertainty and learning in the adoption of new agricultural technologies: Where are we on the learning curve? Agr. Syst. 75 215 234
McKenna, H.P. 1994 The essential elements of a practitioners’ nursing model: A survey of clinical psychiatric nurse managers J. Adv. Nurs. 19 870 877
McRoberts, N., Hall, C., Madden, L.V. & Hughes, G. 2011 Perceptions of disease risk: From social construction of subjective judgments to rational decision making Phytopathology 101 654 665
Meador, D.P. & Fisher, P.R. 2013 Ammonium in nutrient solutions decreases free chlorine concentration from sodium hypochlorite HortScience 48 1304 1308
Mebalds, M., van der Linden, A., Bankier, M. & Beardsell, D. 1996 Using ultraviolet radiation and chlorine dioxide to control fungal plant pathogens in water Nurs. Pap. 5 1 2
Minuto, A., Gaggero, L., Gullino, M.L. & Garibaldi, A. 2008 Influence of pH, nutrient solution disinfestation and antagonists application in a closed soilless system on severity of fusarium wilt of gerbera Phytoparasitica 36 294 303
Mitroff, I.I. & Turoff, M. 2002 11.B.Philosophical and methodological foundations of Delphi. In: The Delphi method: Techniques and applications. 30 Sept. 2012. <http://is.njit.edu/pubs/delphibook/>
Moens, M. & Hendricks, G. 1992 Drainwater filtration for the control of nematodes in hydroponic-type systems Crop Protection 11 69 73
Morin, P., Camper, A., Jones, W., Gatel, D. & Goldman, J.C. 1996 Colonization and disinfection of biofilms hosting coliform-colonized carbon fines Appl. Environ. Microbiol. 62 4428 4432
Murray, T.J. 1979 Delphi methodologies: A review and critique Urban Syst. 4 153 158
Newman, S. 2004 Disinfecting irrigation water for disease management. Proc. 20th Annu. Conf. Pest Mgt. Ornamentals. Soc. Amer. Florists. p. 1–10
Ohashi-Kaneko, K., Yoshii, M., Isobe, T., Park, J.S., Kurata, K. & Fujiwara, K. 2009 Nutrient solution prepared with ozonated water does not damage early growth of hydroponically grown tomatoes Ozone Sci. Eng. 31 21 27
Ohtani, T., Kaneko, A., Fukuda, N., Hagiwara, S. & Sase, S. 2000 Development of a membrane disinfection system for closed hydroponics in a greenhouse J. Agr. Eng. Res. 77 227 232
Padel, S. 2002 Conversion to organic farming: A typical example of diffusion of an innovation? Sociol. Ruralis 41 40 61
Parke, J. & Fisher, P. 2012 Treating irrigation water to eliminate water molds, p. 5–49 In: J.W. Pscheidt and C.M. Ocamb (eds.). Pacific Northwest plant disease management handbook. 9 Apr. 2014. <http://pnwhandbooks.org/plantdisease/pesticide-articles/treating-irrigation-water-eliminate-water-molds>
Raudales, R. 2013 Characterization of water treatment technologies in irrigation. Univ. Florida, Gainesville, PhD Diss. UFE0046234
Raviv, M. and J.H. Lieth (eds.). 2007. Soilless culture: Theory and practice. Elsevier, Burlington, MA
Rens, L.R. 2011 Chlorine dioxide as sanitizing agent in recirculating irrigation for greenhouse hydroponic bell peppers. Univ. Florida, Gainesville, MS Thesis UFE0043811
Rogers, E.M. 2003 Diffusion of innovations. 5th ed. Free Press, New York, NY
Runia, W.T. 1995 A review of possibilities for disinfection of recirculation water from soilless cultures Acta Hort. 382 221 229
Sackman, H. 1974 Delphi assessment: Expert opinion, forecasting and group process. RAND Corp. Rpt. 1283, RAND Corp., Santa Monica, CA
Shah, H.A. & Kalaian, S.A. 2009 Which is the best parametric statistical method for analyzing Delphi data? J. Mod. Appl. Stat. Methods 8 226 232
Slaughter, A., Katz, R.V. & Grasso, J.E. 1999 Professional attitudes toward denture adhesives: A Delphi technique survey of academic prosthodontists J. Prosthet. Dent. 82 80 89
Smale, M. & Heisey, P.W. 1993 Simultaneous estimation of seed-fertilizer adoption decisions: An application to hybrid maize in Malawi Technol. Forecast. Soc. Change 43 353 368
Steele, M. & Odumeru, J. 2004 Irrigation water as source of foodborne pathogens on fruit and vegetables J. Food Protection 67 2839 2849
Stewart-Wade, S.M. 2011 Plant pathogens in recycled irrigation water in commercial plant nurseries and greenhouses: Their detection and management Irr. Sci. 29 267 297
Stoodley, P., Sauer, K., Davies, D.G. & Costerton, J.W. 2002 Biofilms as complex differentiated communities Annu. Rev. Microbiol. 56 187 209
Sunding, D. & Zilberman, D. 2001 Chapter 4: The agricultural innovation process: Research and technology adoption in a changing agricultural sector, p. 207–261. In: B.L. Gardner and G.C. Rausser (eds.). Handbook of agricultural economics. Vol. 1. Elsevier, Amsterdam, The Netherlands
Tarver, T. 2008 “Just add water”: Regulating and protecting the most common ingredient J. Food Sci. 73 1 13
Toppe, B. & Thinggaard, K. 1998 Prevention of Phytophthora root rot in gerbera by increasing copper ion concentration in the nutrient solution Eur. J. Plant Pathol. 104 359 366
Toppe, B. & Thinggaard, K. 2000 Influence of copper ion concentration and electrical conductivity of the nutrient solution on Phytophthora cinnamomi in ivy grown in ebb-and-flow systems J. Phytopathol. 148 579 585
U.S. Environmental Protection Agency 1999 R.E.D. facts: Chlorine gas. EPA-738-F-99-001. 19 Feb. 2014. <http://www.epa.gov/oppsrrd1/reregistration/REDs/factsheets/4022fact.pdf>
U.S. Environmental Protection Agency 2013 Drinking water contaminants. 19 Feb. 2014. <http://water.epa.gov/drink/contaminants/index.cfm>
U.S. Food and Drug Administration 2013 Food Safety Modernization Act. 9 Apr. 2014. <http://www.fda.gov/Food/GuidanceRegulation/FSMA/default.htm>
Van Os, E.A. 2010 Disease management in soilless culture systems Acta Hort. 883 385 394
Van Os, E.A., Amsing, J.J., van Kuik, A.J. & Willers, H. 1999 Slow sand filtration: a potential method for the elimination of pathogens and nematodes in recirculating nutrient solutions from glasshouse-grown crops Acta Hort. 481 519 526
Van Steenkiste, B.C., Jacobs, J.E., Verheijen, N.M., Levelink, J.H. & Bottema, B.J.A.M. 2002 A Delphi technique as a method for selecting the content of an electronic patient record for asthma Intl. J. Med. Inform. 65 7 16
Vanachter, A., Thys, L., Van Wambeke, E. & Van Assche, C. 1988 Possible use of ozone for disinfestation of plant nutrient solutions Acta Hort. 221 295 302
Vänninen, I. & Koskula, H. 1998 Effect of hydrogen peroxide on algal growth, cucumber seedlings and the reproduction of shoreflies (Scatella stagnalis) in rockwool Crop Protection 17 547 553
White, G.C. 1992 Handbook of chlorination and alternative disinfectants, 3rd ed. Wiley/Van Nostrand, New York, NY
Wohanka, W. 1995 Disinfection of recirculating nutrient solutions by slow sand filtration Acta Hort. 382 246 255
Wohanka, W. & Fehres, H. 2006a Efficacy of water treatment with the Aquahort® - system against Xanthomonas hortorum pv. pelargonii. Geisenheim Res. Ctr., Rheingau, Germany
Wohanka, W. & Fehres, H. 2006b Efficacy of water treatment with the Aquahort® - system against Fusarium oxysporum f.sp. cyclaminis. Geisenheim Res. Ctr., Rheingau, Germany
Wohanka, W. & Fehres, H. 2006c Efficacy of water treatment with the Aquahort® - system against Pythium aphanidermatum. Geisenheim Res. Ctr., Rheingau, Germany
Wohanka, W. & Fehres, H. 2007a Efficacy of water treatment with the Aquahort® - system against Clavibacter michiganensis ssp. michiganensis. Geisenheim Res. Ctr., Rheingau, Germany
Wohanka, W. & Fehres, H. 2007b Efficacy of water treatment with the Aquahort® - system against Ralstonia solanacearum Race 3. Geisenheim Res. Ctr., Rheingau, Germany
Wohanka, W., Becker, D. & Fehres, H. 2009a Efficacy of water treatment with the Aquahort® - system against Agrobacterium tumefaciens. Geisenheim Res. Ctr., Rheingau, Germany
Wohanka, W., Becker, D. & Fehres, H. 2009b Efficacy of water treatment with the Aquahort® - system against Trichoderma asperellum. Geisenheim Res. Ctr., Rheingau, Germany
Wohanka, W., Fehres, H. & Schoenbach, W. 2007 Efficacy of water treatment with the Aquahort® - system against Erwinia carotovora spp. carotovora. Geisenheim Res. Ctr., Rheingau, Germany
Wohanka, W., Luedtke, H., Ahlers, H. & Luebke, M. 1999 Optimization of slow sand filtration as a means for disinfecting nutrient solutions Acta Hort. 481 539 544
Yiaosumi, W., Evans, L. & Rogers, L. 2005 Farm water quality and treatment. Ag Facts AC 2 (April), 9th ed. Dept. Primary Industries, Australia Government, Victoria, Australia
Zhou, H. & Smith, D.W. 2002 Advanced technologies in water and wastewater treatment J. Environ. Eng. Sci. 1 247 264
