<%BANNER%>

Monitoring Irrigation Water Quality in Horticulture

Permanent Link: http://ufdc.ufl.edu/UFE0045049/00001

Material Information

Title: Monitoring Irrigation Water Quality in Horticulture
Physical Description: 1 online resource (92 p.)
Language: english
Creator: Meador, Dustin P
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2013

Subjects

Subjects / Keywords: bacteria -- biological -- chemical -- physical -- quality -- sanitation -- treatment
Environmental Horticulture -- Dissertations, Academic -- UF
Genre: Horticultural Sciences thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The biological quality of water used for horticulture irrigation affects risk of crop losses from waterborne pathogens, algae, and biofilm. Efficacy of treatment technologies such as chlorine used to control biological contaminants in water is affected by other quality parameters including water chemistry,physical particles, and sanitizing agent concentration. The first objective of this research was to determine physical, chemical and biological irrigation quality in horticultural greenhouses and nurseries. The second objective was to determine whether a dehydrated culture media (Petrifilm) was a suitable onsite monitoring tool to quantify colony forming units (cfu) per ml of aerobic bacteria and fungus. The third objective was to determine free chlorine, total chlorine and oxidation reduction potential (ORP) when sodium hypochlorite was added to nutrient solutions that contained different sourcesof nitrogen. Results from the first objective showed that recirculated irrigation had lower physical and biological irrigation quality than source, well or municipal water. Recirculated water had low ultra violet (UV)transmission (68% to 72% compared with recommended 75%) and high concentrations of aerobic bacteria (above the 10,000 cfu·mL-1 recommended for controlling biofilm clogging of irrigation equipment). For objective 2, the density of aerobic bacteria (cfu·mL-1) from irrigation water samples estimated on Petrifilm-AC at 3 days was lower than estimates on Reasoner andGoldrich agar (R2A) culture medium bya factor of 2.92. Concentration (cfu·mL-1) of a single strainculture of Xanthomonas campestris pv. Begoniaceae cultured on Petrifilm-AC were also significantly lower (Phytophthora cactorum did not culture on Petrifilm-YM, whereas PDA and vegetable agar with antibiotics were able to culture Phytophthora. When chlorine from sodium hypochlorite was added at 2.6 mg·L-1 Cl to nutrient solutions at 100 mg·L-1 N, there was a rapid decrease of freechlorine to near zero within 2-min, because of complexation of hypochlorous acid with ammonium. Main findings were that improved monitoring and treatmentof recirculated water is needed, Petrifilm-AC is a useful onsite monitoring tool for aerobic bacteria in irrigation water, and chlorination efficacy islikely to be affected by interaction with water soluble fertilizer.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Dustin P Meador.
Thesis: Thesis (Ph.D.)--University of Florida, 2013.
Local: Adviser: Fisher, Paul R.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2013
System ID: UFE0045049:00001

Permanent Link: http://ufdc.ufl.edu/UFE0045049/00001

Material Information

Title: Monitoring Irrigation Water Quality in Horticulture
Physical Description: 1 online resource (92 p.)
Language: english
Creator: Meador, Dustin P
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2013

Subjects

Subjects / Keywords: bacteria -- biological -- chemical -- physical -- quality -- sanitation -- treatment
Environmental Horticulture -- Dissertations, Academic -- UF
Genre: Horticultural Sciences thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The biological quality of water used for horticulture irrigation affects risk of crop losses from waterborne pathogens, algae, and biofilm. Efficacy of treatment technologies such as chlorine used to control biological contaminants in water is affected by other quality parameters including water chemistry,physical particles, and sanitizing agent concentration. The first objective of this research was to determine physical, chemical and biological irrigation quality in horticultural greenhouses and nurseries. The second objective was to determine whether a dehydrated culture media (Petrifilm) was a suitable onsite monitoring tool to quantify colony forming units (cfu) per ml of aerobic bacteria and fungus. The third objective was to determine free chlorine, total chlorine and oxidation reduction potential (ORP) when sodium hypochlorite was added to nutrient solutions that contained different sourcesof nitrogen. Results from the first objective showed that recirculated irrigation had lower physical and biological irrigation quality than source, well or municipal water. Recirculated water had low ultra violet (UV)transmission (68% to 72% compared with recommended 75%) and high concentrations of aerobic bacteria (above the 10,000 cfu·mL-1 recommended for controlling biofilm clogging of irrigation equipment). For objective 2, the density of aerobic bacteria (cfu·mL-1) from irrigation water samples estimated on Petrifilm-AC at 3 days was lower than estimates on Reasoner andGoldrich agar (R2A) culture medium bya factor of 2.92. Concentration (cfu·mL-1) of a single strainculture of Xanthomonas campestris pv. Begoniaceae cultured on Petrifilm-AC were also significantly lower (Phytophthora cactorum did not culture on Petrifilm-YM, whereas PDA and vegetable agar with antibiotics were able to culture Phytophthora. When chlorine from sodium hypochlorite was added at 2.6 mg·L-1 Cl to nutrient solutions at 100 mg·L-1 N, there was a rapid decrease of freechlorine to near zero within 2-min, because of complexation of hypochlorous acid with ammonium. Main findings were that improved monitoring and treatmentof recirculated water is needed, Petrifilm-AC is a useful onsite monitoring tool for aerobic bacteria in irrigation water, and chlorination efficacy islikely to be affected by interaction with water soluble fertilizer.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Dustin P Meador.
Thesis: Thesis (Ph.D.)--University of Florida, 2013.
Local: Adviser: Fisher, Paul R.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2013
System ID: UFE0045049:00001


This item has the following downloads:


Full Text

PAGE 1

1 MONITORING IRRIGATION WATER QUALITY IN HORTICULTURE By DUSTIN PAUL MEADOR A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR O F PHILOSOPHY UNIVERSITY OF FLORIDA 2013

PAGE 2

2 2013 Dustin Paul Meador

PAGE 3

3 In memory of my brother Preston

PAGE 4

4 ACKNOWLEDGMENTS First I thank my beautiful wife Alexandra Meado r for our marriage of 13 years and for being my best friend and love of my life, for over 20 years! She has help ed to provide me with an am azing quality of life that I will cherish forever Without our family, I could not have achieved my goal of earning a doctoral degree and I thank God for surrounding me with loving, caring and compassionate people. My dissertation was a significant achievement for me and I believe it was possible because of the many lessons that I learned outside the classroom. The Perd Jenny have always been examples of a strong work ethic and positive outlook that is needed to be successful and happy in life I thank my grandparents and parents, brothers and sisters and wonderful in laws. Special thanks to my advisor, Paul Fisher, for his confidence, patience, and guidance I would also like to thank the University of T hanks to my committee members: Max Teplitski, Charlie Guy, Phil Harmon and Natalia Peres for their support, guidance and discussion throughout my research and writing. Special recognition is given to The Young Plant Research Center Partners for their continued dedication to the horticulture industry Finally thank you to my f ellow co workers, lab mates, fellow students, friends and teachers The opportunity to meet Tim Tebow and Dr Rebecca Darnell was truly inspiration al and thank you to Wagner Vendrame and staff at U. of Florida T ropical R esearch and E ducation Center for pro viding me with my first opportunity to become a professional scientist The Gator Nation is everywhere and has provided me with a strong foundation as a scientist educator and horticulture professional. onor to have this opportunity and I wish to thank th e animals of the earth and our pet family for sharing their habitat with us.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 9 ABSTRACT ................................ ................................ ................................ ................... 10 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 12 2 SURVEY OF PHYSICAL, CHEMICAL AND MICROB IAL WATER QUALITY IN GREENHOUSE AND NURSERY IRRIGATION WATER ................................ ........ 20 Materials and Methods ................................ ................................ ............................ 22 Results and Discussion ................................ ................................ ........................... 26 3 THE USE OF DEHYRATED MEDIA FOR ONSITE MONITORING OF AEROBIC BACTERIA IN IRRIGATION WATER ................................ .................... 41 Materials and Methods ................................ ................................ ............................ 43 Results and Discussion ................................ ................................ ........................... 44 4 COMPARISON BETWEEN STANDARD AND DEHYDRATED AGAR TO QUANTIFY FUNGI, PHYTOPHTHORA CACTORUM AND XANTHOMONAS CAMPESTRIS IN WAT ER SAMPLES ................................ ................................ .... 47 Materials and Methods ................................ ................................ ............................ 50 Experiment 1: Fungal Counts During Subirrigation Cycles ............................... 50 Experiment 2: Quantification of Colonies From Zoospores of P. cactorum ...... 52 Experiment 3: Quantification of Colonies From Single Isolated Xanthomonas campestris ................................ ................................ .............. 54 Statistical Analysis ................................ ................................ ............................ 56 Results ................................ ................................ ................................ .................... 57 Experiment 1: Fungal Coun ts During Subirrigation Cycles ............................... 57 Experiment 2: Phytophthora cactorum ................................ ............................. 57 Experiment 3: Xanthomonas campestris ................................ .......................... 58 Discussion ................................ ................................ ................................ .............. 58

PAGE 6

6 5 AMMONIUM AND UREA IN NUTRIENT SOLUTIONS DECREASE FREE CHLORINE CONCENTRATION FROM SODIUM HYPOCHLORITE ..................... 68 Materials and Methods ................................ ................................ ............................ 71 Results and Discussion ................................ ................................ ........................... 73 6 CONCLUSIONS ................................ ................................ ................................ ..... 84 APPENDIX LIST OF REFERENCES ................................ ................................ ............................... 88 BIOGRAPHICAL SKETCH ................................ ................................ ............................ 92

PAGE 7

7 LIST OF TABLES Table page 2 1 Descriptive characteristics for the production environments from each greenhouse or nursery (Location) included in the survey of irrigation quality. Each Location is listed by row and identified by the code n umber (1 24) presented in the first column. Each row represents a different Location, and specifies the production area, water sources, and irrigation technologies used. ................................ ................................ ................................ ................... 34 2 2 ANOVA irrigatio n quality results from the surveyed population of 24 greenhouses and nurseries (Locations). Irrigation quality variables were analyzed by Location, Sample Type (Source, Tank, Furthest Outlet, Subirrigation, or Catchment Basin), and their interaction (Locat ion Sample Type). Least squared means are presented for each Sample TypeValues followed by a superscript notation indicate they are outside recommended guidelines (see footnotes below). ................................ ................................ ....... 36 2 3 ANOVA table for aerobic bacteria and yeast and mold cfumL 1 for each greenhouse and nursery (Location) surveyed. At each Location, samples were collected at the Source and Furthest Outlet. For Location 7, there were two Source and Furthest Outlet irrig ation circuits which were analysed separately. Locations 20 and 24 did not have Furthest Outlet sampling points and were not included. The reported cfumL 1 of aerobic bacteria and yeast and mold are least square means for each Location, with statistical c omparison between Source and Furthest Outlet by Location. .......................... 38 2 4 ANOVA table for effect of sanitation treatments on the cfumL 1 of aerobic bacteria and yeasts and molds. Samples were collecte d immediately before (Pre Treatment) and immediately after (Post Treatment) each sanitation treatment. The reported density of aerobic bacteria and yeast and mold are least square mean cfumL 1 for each Location followed by significance between Pre Treatme nt and Post Treatment. ................................ .................... 40 3 1 Ingredients for the two culture substrates. ................................ .......................... 46 3 2 ANOVA effects Effect of substrate (Petrifilm or R2A) and measurement day (3 or 7) on bacteria cfumL 1 Substrate, day, and their interaction were level for the substrate day least square means. Data were log transformed for ANO VA analysis, and back transformed for presentation in this table. ......... 46 4 1 Ingredients for the four fungal culture substrates tested ................................ ..... 61 4 2 Ing redients for the three culture substrates used to culture Xanthomonas campestris pv. ................................ ................................ ................................ .... 62

PAGE 8

8 4 3 agar with ch loramphenicol (SDA) and measurement day (7 or 14) on cfu mL 1 of fungi. Substrate, day, and their interaction were significant at p<0.001. Data were LOG 10 transformed cfu mL 1 for both ANOVA and this table. the p=0.05 level for the effect of substrate and measurement day. ................................ ......................... 63 4 4 Estimated cfu mL 1 of least square mean counts on a hemocytometer, Potato dextrose agar (PDA), and the selective media agar (PARP H) to culture Phytophthora cactorum the p=0.05 level for the effect of substrate. Counts obtained from Petrifilm YM estimated zero cfu mL 1 of P. cactorum after the standard incubation time and w ere not included in the analysis. ................................ ................................ 63 4 5 Effects of measurement meth od (R2A, PDA, Petrifilm AC, and hemocytometer) and inoculum density (low, intermediate, and high) with Xanthomonas campestri s Main and interaction effects of measurement method and inoculum density were significant (p<0.0001). Letters represent measurement method and inoculum density. ................................ ..................... 63 5 1 Nutrient sources used in the study (reported by manufacturer). ......................... 78 5 2 ANOVA summary for Experiment (a) Multiple N sources, where free chlorine was applied to 11 nutrient solutions at 200 mg.L 1 N and a DI water control (N Source) at 2.6 mgL 1 Cl with 2 min and 60 min contact times (Time). ......... 79 5 3 ANOVA summary for Experiment (a) Mult iple N Concentrations where free chlorine was applied at 2.6 mgL 1 Cl to 4 reagent nutrient solutions (N Source) at 6 concentrations from 0 to 600 mgL 1 N ([N]) with 2 or 60 min contact time (Time). ................................ ................................ ............................ 79 5 4 ANOVA summary for Experiment (c) High Applied Chlorine where free chlorine was applied at two concentrations of 10 or 20 mgL 1 Cl ([Cl]) to 11 nutrient solutions at 200 mgL 1 N and a DI control (N Source) and free chlorine was measured a fter 2 min contact time. ................................ ................ 79

PAGE 9

9 LIST OF FIGURES Figure page 4 1 Fungal counts (cfumL 1 ) recovered from a tank and plated on 2 different substrates. F ungal counts on yeast and mold (Petrifilm YM) were determined on measurement day 7 and SDA on measurement day 14. Data presented were log transformed. ................................ ................................ ......................... 64 4 2 Comparison between counting method and substrate, hemocytometer (HEMO) were initial estimated colonies, aerobic count (Petrifilm AC) on measurement day 3 and R2A, PDA on measurement day 5. Data presented were Log10 transformed. ................................ ................................ .................... 65 4 3 Photographs (a) to (e) show the method developed to use Petrifilms for on farm monitoring of microbial cfumL 1 in irrigation system. Fig. 4 3a (above), run irrigation to sufficiently flush lines before sampling. Collect irrigation water as it fl ows from an emitter, using a sterile disposable vial. When sampling from tanks and catchment basins, draw the sample from the irrigation intake depth using a clean sampling container attached to a pole, and then transfer to the sterile vial. Take care not to contaminate the sample by touching the rim of the sterile vial with the irrigation emitter or hands. Materials (Petrifilm YM and Petrifilm AC, vials with 99 ml of sterile diluent, and disposable pipettes) are available to growers through a variety of di stributors. Estimated cost for the minimum order quantity at the time of this study was about $2.00 per sample on either Petrifilm AC or Petrifilm YM. ................................ ............. 66 5 1 Nitrogen source and its effect on free and total chlorine and ORP in Experiment (a) Multiple Fertilizers. Chart (A) shows the percent contribution of ammonium, nitrate, or urea to total nitrogen in each nutrient source. Charts (B) to (D) show the effect of 200 mg L 1 N from commercial fert ilizer blends and reagent grade nutrient solutions on (B) free chlorine, (C) total chlorine, and (D) ORP after injecting sodium hypochlorite at 2.6 mg L 1 Cl with 2 min and 60 min contact times. Letters represent mean comparisons using =0.05 level for the interaction of nutrient source and contact time. ................................ ................................ ................................ ....... 81 5 2 Effect of nitrogen concentration from four reagent grade N sources on free chlorine after injecting sodium hypochlor ite at 2.6 mgL 1 Cl in Experiment (b) Multiple N Concentrations, measured after (A) 2 min and (B) 60 min. Letters each measurement time. ................................ ................................ .................... 82 5 3 Effect of sodium hypochlorite at 10 mgL 1 Cl and 20 mgL 1 Cl measured after 2 min in 11 nutrient solutions at 200 mgL 1 N and a DI Control, in Experiment (c) High Applied Chlorine. Letters represent mean comparisons s HSD at the p=0.05 level ................................ ............................... 83

PAGE 10

10 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy MONITORIN G IRRIGATION WATER QUALITY IN HORTICULTURE By Dustin Paul Meador May 201 3 Chair: Paul R Fisher Major: Horticultural Sciences The biological quality of water used for horticulture irrigation affects risk of crop losses from waterborne pathogens, algae, and biofilm. Efficacy of treatment technologies such as chlorine used to control biological contaminants in water is affected by other quality parameters including water chemistry, physical particles, and sanitizing agent concentration. The first objectiv e of this research was to determine physical, chemical and biological irrigation quality in horticultural greenhouses and nurseries. The second objective was to determine whether a dehydrated culture media (Petrifilm) was a suitable onsite monitoring tool to quantify colony forming units (cfu) per ml of aerobic bacteria and fungus. The third objective was to determine free chlorine, total chlorine and oxidation reduction potential (ORP) when sodium hypochlorite was added to nutrient solutions that contained different sources of nitrogen. Results from the first objective showed that recirculated irrigation had lower physical and biological irrigation quality than source well or municipal water. Recirculated water had low ultra violet (UV) transmission (68% to 72% compared with recommended 75%) and high concentrations of aerobic bacteria (above the 10,000 cfumL 1 recommended for controlling biofilm clogging of irrigation equipment). For objective 2, the density of

PAGE 11

11 aerobic bacteria (cfumL 1 ) from irrigation wa ter samples estimated on Petrifilm AC at 3 days was lower than estimates on Reasoner and Goldrich agar (R2A) culture medium by a factor of 2.92. Concentration (cfumL 1 ) of a single strain culture of Xanthomonas campestris pv. Begoniaceae cultured on Petri film AC were also significantly lower (P<0.05) than counts on R2A culture medium or using a hemocytometer, but was not different from counts on Potato Dextrose Agar (PDA) culture medium. Zoospores of Phytophthora cactorum did not culture on Petrifilm YM, w hereas PDA and vegetable agar with antibiotics were able to culture Phytophthora When chlorine from sodium hypochlorite was added at 2.6 mgL 1 Cl to nutrient solutions at 100 mg L 1 N, there was a rapid decrease of free chlorine to near zero within 2 min because of complexation of hypochlorous acid with ammonium. Main findings were that improved monitoring and treatment of recirculated water is needed, Petrifilm AC is a useful onsite monitoring tool for aerobic bacteria in irrigation water, and chlorinat ion efficacy is likely to be affected by interaction with water soluble fertilizer.

PAGE 12

12 CHAPTER 1 INTRODUCTION Horticultural operations are using recirculating subirrigation systems and runoff water collected in catchment basins to supplement potable water s ources for irrigation (Obreza et al., 2010), in an effort to reduce runoff and conserve scarce water resources. Irrigation water that is collected for use in catchment basins (capture and reuse) or recirculating (ebb and flood) irrigation systems is charac terized by increased levels of physical, chemical, and biological contaminants that can decrease water quality, compared to potable water (Gilbert et al., 1980; Runia et al., 1994). Decreased irrigation quality can result in micro irrigation emitter and fi lter clogging and promote biofilm formation on critical surfaces within the irrigation distribution systems (Bucks et al., 1979; Ravina et al., 1997). Elevated biological loads in recirculated water may result in biofilm formation. Biofilms are multicellul ar communities of microorganisms that may include bacteria, fungi and algae as well extracellular polymers produced and excreted by the organisms within the biofilm. Any wet surface or an interface between air and a highly eutrophic (nutrient rich liquid) can host biofilms (Cullimore, 1993). Biofilms are common inside water storage tanks and irrigation lines (Ravina et al., 1997), where they can reduce water flow, corrode metals, reduce efficacy of sanitation agents, and serve as sources of pathogens. Biofi lms are an important part of microbial lifecycle and Sutton et al. (2006) has shown that oospores of Pythium spp. can be encountered in biofilms, and serve as a source of pathogen inoculum in hydroponic production. Monitoring and controlling biofilms, ther efore, is important to maintain the quality of the irrigation water.

PAGE 13

13 Sanitation treatment technologies are incorporated into horticulture irrigation systems to improve water quality and provide control of pathogens, improve plant health and prevent the dev elopment of secondary inoculum or protective overwintering structures (Agrios, 1997). These technologies include filtration, chlorination, copper ionization, ozonation, UV, use of activated peroxygens, chlorine dioxide, heat, or other technologies to provi de disinfestation of irrigation water (van Os, 2009). Water quality variables such as suspended particles, dissolved organic and inorganic molecules, and microbes create a demand on sanitizing agent active ingredients, thereby reducing the efficacy of wate r treatment technologies for control of microbes, pathogens and algae (Copes et al., 2004; Ravina et al., 1997; Sutton et al., 2006). Monitoring of biological, physical, and chemical variables is necessary to ensure water quality is adequate for sanitatio n treatment to be effective. Physical contaminants from plant debris, peat moss, sand, and insoluble salts can restrict water flow, create a sanitizing agent demand, reduce UV transmission, and shield target pathogens from contact with a sanitizer (Copes e t al., 2004; Gauthier et al., 1999; Pettygrove and Asano, 1985; Pirovanni et al., 2004; Rogers et al., 2003). Chemical aspects of water quality, including pH, total dissolved solids, alkalinity, and nitrogen, can decrease the efficacy of sanitizing treatme nt to provide pathogen and pest control. High solution pH and reduced forms of nitrogen in organic and inorganic compounds, such as ammonium in water soluble fertilizers, can reduce the availability of strong oxidizers such as hypochlorous acid (APHA, 1995 ; Feben and Taras, 1955; Heibling and VanBriessen, 2008; Tanwar et al., 2008).

PAGE 14

14 Regulations and standard monitoring protocols are in place to promote health and operational safety in food and beverage processing and recreational swimming facilities, and the se approaches can be applied to monitoring irrigation water. Monitoring biological variables such as counts of aerobic bacteria and fungi are used to indicate contamination or ineffective treatment (American Public Health Association (APHA), 1995; Maier et al., 2009). The reduction of bacterial counts from water samples collected before versus after water treatment is used as a measure of effective control by a sanitizing agent, according to the U. S. Environmental Protection Agency (USEPA), 1999. An accept able level of control is a two to three log reduction of bacterial cfumL 1 from pre treatment to post treatment and indicates treatment reliability for disinfection (USEPA, 1999). The population density of aerobic bacteria can be expressed as colony formi ng units per milliliter of the original sample, (cfumL 1 ). Total cfumL 1 of aerobic bacteria tested at points along the irrigation system and immediately before and after treatment (for example, chlorine injection) can serve as a non specific indicator o f disinfection efficacy (APPHA, 1995; Maier et al., 2009) and monitored to determine the risk of clogged irrigation lines from biofilm (Bucks et al., 1979). B iofilm formation in micro irrigation systems can be quantified in terms of aerobic bacteria cfumL 1 with a threshold of less than 10,000 cfumL 1 recommended for acceptable risk of biofilm clogging of drip and mist emitters (Rogers et al., 2003). Measured cfumL 1 for aerobic bacteria using culture based techniques can depend on the substrate, and me thods used for sample collection, handling and shipment (Watkins and Jian, 1997). The recommended levels of cfumL 1 for aerobic

PAGE 15

15 bacteria are traditionally based on culture counts using approved liquid gel agar preparations described in standard method 921 5A (APHA, 1988). Another challenge to monitoring microorganisms in irrigation is that they often become environmentally stressed during shipment and storage of samples, and be rendered viable but non culturable. Under such conditions, quantitative assessme nt of population numbers may underestimate actual microbial load. Quantification of aerobic bacteria may be more easily accomplished as an onsite test for growers than quantification of plant pathogens in irrigation water, because pathogen numbers are ofte n low compared with beneficial or benign micro organisms, and quantification of aerobic bacteria does not require a specialized assay. An onsite method to monitor microbial density in irrigation water would assist management of water quality and treatment in the ho rticulture industry. Petrifilms TM (3M Microbiology, St. Paul, MN) are a plating technology that may be useful fo r irrigation water monitoring. The ready made, dehydrated culture medium includes nutrients and a tetrazolium salt that fluoresces livi ng bacteria colonies on a flat card with a laminate cover slip. Petrifilms are suited for food producers and processors because they contain the standard nutrient components as described in method 940.36 (AOAC, 1999). The nutrients in Petrifilms are water soluble and become available to microbes when the dehydrated gelling agent is solubilized and an indicator dye is incorporated to facilitate accurate colony counts. The U. S. Department of Agriculture and international regulatory agencies have validated se veral standardized monitoring techniques that utilize Petrifilms to evaluate sanitation processes such as chlorination in post harvest handling (Materon, 2003), and to determine sources of contaminants and test the ability of water treatment systems to con trol human pathogens (Riordan et al.,

PAGE 16

16 2001). Culture based monitoring of microbes is difficult with traditional liquid gel agar because of cross contamination with airborne microbiological spores under aseptic conditions. There is extensive plant patholog y research on effects of free chlorine on waterborne pathogens. Control of Pythium and Phytophthora zoospores has been reported with 2 mgL 1 of free chlorine (Cayanan et al., 2009; Hong et al., 2003; Hong and Richardson, 2004; Lang et al., 2008). Chlorine oxidizes and chlorinates living tissue and organic compounds, resulting in damage to membranes, regulatory enzymatic functions and nucleic acids of microorganisms (Stewart and Olson, 1996). Increasing concentration of chlorine, lowering solution pH and in creasing contact time are factors that improve pathogen control (Lang et al., 2008). Chlorine from calcium hypochlorite, sodium hypochlorite, or chlorine gas sources has a complex chemistry in irrigation water, and interaction with fertilizers and acid in jection are likely to affect sanitation efficacy of chlorination. Once added to water, sodium hypochlorite is converted to the hypochlorite ion (OCl ) and hypochlorous acid (HOCl), which along with dissolved chlorine gas are collectively termed free chlori ne. The balance between hypochlorous acid and hypochlorite is determined by water pH, whereby hypochlorous acid predominates at solution pH below 7.5 and hypochlorite ions at pH above 7.5 (Morris, 1966). Water pH also influences the sanitizing strength of a chlorine solution, because HOCl is estimated to be approximately 100 more effective as a biocide than OCl (White, 1992). The effect of pH on waterborne pathogens was illustrated in research by Lang et al. (2008), whereby control of Pythium

PAGE 17

17 aphanidermat um and Pythium dissotocum zoospores was achieved at either 2.0 mgL 1 Cl at pH 8.1, or 0.5 mgL 1 Cl below pH 6.3. In the presence of organic and inorganic nitrogen including ammonia, a range of olecules, such as chloramines in Eq. [1 to 3], adapted from EPA (1999). The concentration of both free NH 3 + HOCl NH 2 Cl (monochloramine) + H 2 O [1] NH 2 Cl + HOCl NHCl 2 (dichloramine) + H 2 O [2] NHCl 2 + HOCl NCl 3 (nitrogen trichloride) + H 2 O [3] In municipal water treatment facilities, a dose of one part of inorganic N to every three parts of free chlorine re sults in 99% conversion of free chlorine to chloramines after 0.2 seconds at pH 7, or 147 seconds at pH 4 (White, 1992). In horticulture irrigation, water soluble fertilizer are often supplied in irrigation between 100 and 200 mgL 1 N and then treated wit h free chlorine concentrations between 1 and 2 mgL 1 Cl. With such a high N to Cl ratio, the majority of free chlorine would therefore be expected to rapidly convert to chlorinated nitrogen forms. Chlorine also reacts with urea, although the chain of reac tions is more complex and slow acting than chlorination of ammonia (Blatchley and Cheng, 2010). Combined forms of chlorine and nitrogen molecules or chloramines are considered weaker sanitizers than free chlorine from hypochlorous acid (White, 1992) becau se of the much longer contact time that chloramines require to disinfect with

PAGE 18

18 compared to an equal concentration of free chlorine. To control 99% of Escherichia coli with hypochlorous acid at 1 mgL 1 Cl required for a contact time of 60 seconds, whereas w ith combined chlorine forms, such as NH 2 Cl and NHCl 2 at 1 mgL 1 Cl required more than 6000 seconds (Akin et al., 1982). Based on the reported control of human pathogens in water supplies, combined chlorine treatments may require longer contact times and i ncreased total chlorine concentrations compared with free chlorine to provide a similar level of control for horticulture applications, although data are not available for plant pathogens. However, the greater stability of chloramines in the presence of or ganic compounds compared with hypochlorous acid may increase penetration of chloramines into biofilm, resulting is greater inactivation of biofilm bacteria (LeChevallier et al., 1988). The American Water Works Association (AWWA, 1991) provided guidelines f or residual concentration of between 0.5 to 1 mgL 1 chloramine for disinfection of ground water (White, 1992), such as from wells or irrigation catchment areas. The first objective was to quantify physical, chemical and microbial load and overall sanitizi ng demand at key sampling points within irrigation systems, using a field survey approach at multiple greenhouse and nursery locations. These values were compared against published irrigation guidelines for effective sanitation treatment for control of pat hogens, algae and biofilm, and for adequate nutrient management of horticultural crops. The second objective of this research was to compare cfumL 1 of aerobic bacteria cultured onto Petrifilm AC and standard media, sampled from a recirculated irrigation system in Florida greenhouse. The next research objective was to compare cfumL 1 of fungi cultured onto Petrifilm YM and standard media, sampled from

PAGE 19

19 recirculated irrigation. Next, an evaluation with known cfumL 1 of single isolates of bacterium Xanthomo nas campestris was cultured and compared between Petrifilm AC and standard media used for bacteria culture and known isolates of the oomycete Phytophthora cactorum (zoospores) was cultured and compared between Petrifilm YM and standard media for fungi cult ure. The final research objective was to determine the free chlorine, total chlorine and ORP responses over time when nutrient solutions were blended with sodium hypochlorite. Three experiments were run. In Experiment (a) Multiple fertilizers 200 mgL 1 N from 7 commercial blended fertilizers containing macro and micronutrients and reagent grade nutrient solutions were combined with 2.6 mgL 1 Cl. In Experiment (b) Multiple N Concentrations, five concentrations of N from 0 to 600 mgL 1 from 4 reagent gra de nutrient solutions were combined with 2.6 mgL 1 Cl. Experiment (c) High Applied Chlorine quantified whether 10 or 20 mgL 1 were sufficient to provide a target 2 mgL 1 Cl after a 2 min contact time in the presence of 200 mgL 1 N from commercial ferti lizers and reagent grade N salts.

PAGE 20

20 CHAPTER 2 SURVEY OF PHYSICAL, CHEMICAL AND MICROBIAL WATER Q UALITY IN GREENHOUSE AND NURSERY IRRIGATION WATER Horticultural operations are increasingly using recirculating subirrigation systems and runoff water collected in catchment basins to supplement decreasingly available potable water sources for irrigation (Obreza et al., 2010). Irrigation water that is collected for use in catchment basins (capture and reuse) or recirculating (ebb and flood) irrigation systems are characterized by elevated levels of physical, chemical, and biological contaminants that result in lower water quality, compared to potable water (Gilbert et al., 1980; Runia et al., 1994). When irrigating with low quality water, the combined physical, ch emical, and microbial variables can result in microirrigation emitter and filter clogging and promote biofilm formation on critical surfaces within the water distribution systems (Bucks et al., 1979; Ravina et al., 1997). A report by Hong and Moorman (2005 ) compiled research that identified 16 species of Phytophthora (including the plant pathogen P hytophthora ramorum ) 26 species of Pythium, 10 viruses, waterborne plant pathogens and other microorganisms in irrigation water samples collected from nursery an d greenhouse operations. Sanitation technologies are incorporated into horticulture irrigation systems to provide control of pathogens, improve plant health and prevent the development of secondary inoculum or protective overwintering structures (Agrios, 1 997). These technologies include filtration, chlorination, copper ionization, ozonation, UV, use of activated peroxygens, chlorine dioxide, heat, or other technologies to provide disinfestation of irrigation water (van Os, 2009). Water quality variables su ch as suspended particles, dissolved organic and inorganic molecules, and microbes create a demand on sanitizing agent active ingredients, thereby reducing the efficacy of water

PAGE 21

21 treatment technologies for control of microbes, pathogens and algae (Copes et al., 2004; Ravina et al., 1997; Sutton et al., 2006). Monitoring of biological, physical, and chemical variables is necessary to ensure water quality is adequate for its intended use. For example, a standard evaluation for health and operational safety at food and beverage processing and recreational swimming facilities monitors biological variables such as counts of aerobic bacteria and fungi to indicate contamination or ineffective treatment (American Public Health Association (APHA), 1995; Maier et al., 2009). The reduction of bacterial counts from water samples collected before versus after water treatment is used as a measure of effective control by a sanitizing agent, according to the U.S. Environmental Protection Agency (USEPA), 1999. Physical contami nants from plant debris, peat moss, sand, and insoluble salts can restrict water flow, create a sanitizing agent demand, reduce UV transmission, and shield target pathogens from contact with a sanitizer (Copes et al., 2004; Gauthier et al., 1999; Pettygrov e and Asano, 1985; Pirovanni et al., 2004; Rogers et al., 2003). Chemical aspects of water quality, including pH, total dissolved solids, alkalinity, and nitrogen, can decrease the efficacy of sanitizing treatment to provide pathogen and pest control. High solution pH and reduced forms of nitrogen in organic and inorganic compounds, such as ammonium in water soluble fertilizers, can reduce the availability of strong oxidizers such as hypochlorous acid (APHA, 1995; Feben and Taras, 1955; Heibling and VanBrie ssen, 2008; Tanwar et al., 2008). Surveys have been conducted to assess chemical water quality (Argo et al., 1997) and pathogen presence in ornamental greenhouses and nurseries (Hong and Moorman, 2005). Water quality information is lacking on whether level s of water

PAGE 22

22 contaminants such as suspended solids, aerobic bacterial density, and chemical oxygen demand meet recommended guidelines. If physical, chemical and microbial load are outside of recommended guidelines, there is increased risk that treatment tech nologies will not provide effective control of target pathogens, algae, and biofilm in recirculated irrigation systems. The objective of this study was to quantify physical, chemical and microbial load and overall sanitizing demand at key sampling points w ithin irrigation systems, using a field survey approach at multiple greenhouse and nursery locations. These values were compared against published irrigation guidelines for effective sanitation treatment for control of pathogens, algae and biofilm, and for adequate nutrient management of horticultural crops. Materials and Methods 11 states in the U.S., during a period of active crop production from Apr. to July 2009. Surveyed Loc ations had an average production area of 985,546 ft 2 (Table 2 1 ), ranging from 72,000 ft 2 to 6, 533,989 ft 2 The reported average daily water use among the Locations was 160,724 gal with a range of 6001 to 1,000,105 gal. Overall, an average of 33% or 43,6 05 gal (range from 0% to 100%, or 0 to 1,000,105 gal) was from recycled water, defined as Subirrigation (recirculated) or Catchment Basin (a combination of rainwater, runoff, and other sources). The sampled population of greenhouse and nursery Locations pr oduced a variety of seedling plug and cutting transplants and finished containerized ornamental crops. Growers who managed recycled or treated irrigation, and who had participated in previous water quality extension workshops by

PAGE 23

23 the authors in several U.S. states, were selected to participate if they were willing to comply with a standardized sample collection and handling protocol. Water samples were collected at points within each Location that represented up collected from an inline main (6 10 inch) pipe or pressurized expansion tank at the first access point in each nursery (most upstream access poin t within each of the Locations). Tank water was collected from enclosed containments of treated (chemical sanitizer and/or filtration) irrigation solution, prior to going through distribution plumbing. Furthest Outlet was the irrigation solution at the poi nt of application to the crop located furthest away from the Source, after passing through the distribution plumbing, and was collected from overhead, drip or hand held nozzles. Subirrigation samples were collected from recycled ebb and flood irrigation du ring a flood cycle and prior to filtration or sanitizer treatment. Catchment Basin samples were collected from an outdoor, open surface area that received runoff and recycled irrigation water and from the depth of the pump intake valve, prior to filtration or treatment. To determine the effect of water treatment on microbial density in irrigation water during crop production, counts were made on additional samples were collected from Locations that treated water with sanitizer injection or membrane filtrati on (Table 2 1). Eight Locations provided water treatment to recycled irrigation, nine Locations treated well or municipal Source water, and seven Locations did not use sanitizers or membrane filtration. Samples collected at the closest sampling points bef ore and after sanitizer or membrane filtration were identified as Pre Treatment and Post Treatment.

PAGE 24

24 Pre Treatment samples from Source water had not received any sanitizer or membrane filtration after water entered the greenhouse, whereas Pre Treatment from Subirrigation and Catchment Basins were collected after an irrigation event, prior to additional filtration or sanitizer treatment. Recycled water was reported as the percentage of daily water used at each Location that was collected and used again for ir rigation. The experimental design was a split plot with Location as the main plot and Sample Type as the subplot. Effects were analyzed using analysis of variance (ANOVA) using GLM (SAS version 9.1; SAS institute, Cary, NC) including Location (24 levels), Sample Type (five levels) and Sample Type Location interaction, where the sampling time during the day within each Location was a random variable used as the error term. Because of differences in irrigation system design, not all of the 24 Locations had all five Sample Types. Additional analysis of Pre Treatment and Post Treatment was conducted using MIXED (SAS version 9.1). Techniques for collection, handling and laboratory processing followed standard methods (APHA, 1995). Three replicates of 350 mL sa mples were collected in Whirl Pak TM (Nasco, Modesto, CA) sterile collection bags from the five Sample Types. A total of 15 samples were collected at each of the Locations, by a trained grower or supervised employee during 1 d ay of production. Each Location was sampled once during the six we eks of collection and analysis. Four Locations per week shipped their samples overnight to the University of Florida in insulated coolers with ice packs and digital data loggers (HOBO, Onset Computers, Bourne, MA) to ensu re environmental stability and sample integrity during collection and shipment. All samples were received

PAGE 25

25 for processing and analysis within 20 h from the time they were collected and were maintained between 22 to 25 C. Physical water quality was determi 1 ), as described in Current Method 2540 D (APHA, 1995). An Emerson model 4144 (St. Louis, MO) vacuum pump applied a constant suction of 200 kPa to filter each 200 mL sample through a 4.25 cm pre weighed Whatman GF/C 0.45 (Whatman GE, Maidstone, United Kingdom) to recover all suspended particles was analyzed using the loss on ignition procedure where weighed filters from previous TSS analysis were ignited in a muffle furnace at 550 C to volatilize the organics and then reweighed according to standard methods (APHA, 1995). The UV transmission was determined by spectrophotometry at 254 nm, as described by Chang et al. (1998) using a SpectraMax Plus, model 384 spectrophotometer and SoftMax Pro, analytical software package (Molecular Devices, Sunnyvale, CA). Chemical water quality was evaluated using several testing methods outlined by the USEPA (2004). The chemi 1 ), was determined by colorimetric analysis using a digital ORION 4000 meter and COD AQUAfast IV test ampoules containing potassium dichromate and mercuric sulfate (Thermo Fisher, Barrington, IL). Digestion with a COD block inc ubator (Hach, Loveland, CO) was followed by analysis using the closed reflux colorimetric method 5220 B (APHA, 1995). Chemical measurements of pH, alkalinity, dissolved oxygen and electrical conductivity (EC) were conducted using methods described by the A ssociation of Agricultural

PAGE 26

26 Chemists (AOAC), 1999. Chlorine demand was evaluated at 2 min using standard method 4500 Cl G (APHA, 1995). Microbial analysis was performed using a dry rehydratable culture substrate, 3M Petrifilm TM direct count plates (3M Micro biology, St. Paul, MN) formulated for aerobic bacteria cultures or yeast and molds. Each culture contained a dehydrated substrate mounted on a flexible card where microbial colonies are detected and counted with the aid of a color indicator to enumerate li ving microbial densities ( cfu mL 1 ). Irrigation samples were transferred into culture immediately after samples were removed from insulated shipping coolers and allowed to reach to room temperature. Microbiological methods followed official methods: 990.12 Petrifilm Aerobic Count Plates and 997.02, Yeast and Mold Count Plate (AOAC, 1999). Serial dilutions were carried out at one to 10 8 dilutions for aerobic bacteria and one to 10 4 for yeasts and molds, and sterilized water was used as control. Each serial dilution was cultured on aerobic bacteria as well as yeast and mold substrate with three replicates each. Cultures were incubated at 74 F in the dark for 3 to 7 d, until counts were stable. Magnification (10) was used to obtain colony counts for each cul ture. Counts were made from the selected dilution which resulted in 10 to 250 cfu mL 1 on the aerobic count plates and from 10 to 150 cfu/mL on yeast and mold count plates. Results and Discussion Physical water quality variables had significant main effect s for Sample Type and Location with TSS and UV transmission variables, and main and interaction effects for organic carbon. Clogging risk was evaluated using the Bucks et al. (1979) clogging index for microirrigation systems, whereby levels of physical, ch emical, and biological

PAGE 27

27 contaminants were each rated a value from 0 (no risk) to 10 (severe risk); a combined total of 10 or less represented low clogging risk; and a combined total of 20 or more would represent a severe risk an d indicate need for treatment Total Suspended Solids level was insufficient to indicate a physical clogging risk according to the Bucks et al. (1979) scale (Table 2 2). A TSS below 5 mgL 1 is recommended when reuse water will be treated with a chemical oxidizer (USEPA, 1999), and t his recommended level was exceeded for Furthest outlet, Subirrigation, and Catchment Basin samples. Water used to irrigate food crops in Florida has a TSS limit of 5 mgL 1 with a limit of 20 mgL 1 for non food crops (USEPA, 2004) that would apply to the ornamental crops grown in survey Locations. The average physical water quality for Sample Types other than the Source did not meet published guidelines for irrigation use in terms of organic carbon and UV transmission (Table 2 2 ). The guideline for effec tive disinfestation with UV treatment is a UV transmission of at least 75% at the 254 nm wavelength (Newman, 2004). Lower water quality is indicated if total organic carbon (TOC, mgL 1 ) is above 2 mgL 1 because UV transmission and efficacy of sanitizing agents both decrease (Cha ng et al., 1998; USEPA, 1999). Samples collected from the Source had the highest physical quality among the Sample Types in terms of higher UV transmission and lower levels of TSS and organic carbon compared with Subirrigation and Catchment Basin samples. Although TSS in Source water samples was below 10 mgL 1 in all Locations, six Locations had organic carbon levels above the recommended 2 mgL 1 (between 2.1 and 7.1 mgL 1 ). Only one Location had low physical quality from the Sou rce indicated by a low UV transmission that was below the recommended 75%. At all of the other

PAGE 28

28 Locations the higher UV transmission (above 75%) indicated suitability of sanitation with UV treatment. Physical water quality variables did not differ ( P >0.05) between Subirrigation and Catchment Basins or between Tank and Furthest Outlet. After the Source the decreased UV transmission may have resulted from a combination of TSS and dissolved micronutrient chelates and fertilizer dyes. Organic carbon comprised b etween 51% and 69% of TSS, depending on the Sample Type, with the remaining 31% to 49% consisting of non ignitable mineral particulates. Because the organic carbon component of TSS can be oxidized, samples with a high proportion of organic carbon would be likely to increase sanitizing agent demand. There was a moderate risk of clogged drip emitters from chemical contaminants according to the Bucks et al. (1979) scale. However, alkalinity, total dissolved solids, and EC of Source water were within the water quality recommendations for use on floriculture crops. Chemical water quality variables had significant main effects for Location for all chemical water quality measurements. Sample Type affected chemical water quality variables other than alkalinity (Tabl e 2 2). Management of pH is important both for nutrient management and its effects on activity of sanitizers such as chlorine (decreasing the availability of the strong sanitizing form of hypochlorous acid with increased pH) and copper (decreased solubili ty with increasing pH). Average pH was outside the recommended guidelines for Source and Catchment Basin. Appropriate treatment to decrease alkalinity and pH would include acid injection and possibly reverse osmosis. Total dissolved solids were measured wi th an EC meter, in contrast dissolved solids and alkalinity can result in precipitation of inorganic components and

PAGE 29

29 combined compounds such as calcium carbonate. Total dis solved solids and EC for Source water were within recommended values for production ornamental crops (Whipker et al. 2003). Higher total dissolved solid and EC values in Tank, Furthest Outlet and Subirrigation may not indicate a problem, because these Samp le Types included water soluble fertilizer that was injected immediately after the Source. According to criteria from Bucks et al. (1979), the Source and Subirrigation had a 6/10 chemical risk of emitter clogging, and a rating of 5 for other Sample Types. Source samples had the lowest aerobic bacteria counts of all Sample Types, and Subirrigation samples had the highest yeast and mold counts. There were significant main effects for Sample Type and Location for aerobic bacteria as well as yeast and mold, wit h a significant interaction for aerobic bacteria. Microbial water quality variables for Source water were within the recommended guidelines, whereas Tanks, Furthest Outlet, Subirrigation and Catchment Basin had aerobic bacteria above the recommended 10,000 cfu mL 1 (Table 2 2 ). There was a severe microbial clogging risk from Tank, Furthest Outlet, Subirrigation and Catchment Basin, with a biological contamination index of 10/10 (Bucks et al., 1979). Biofilms are microbial multicellular structures, often con sisting of bacteria, fungi and algae and extracellular polymers that are produced and excreted by them used to establish on any wet surface including pumps and pipe materials (Kerr et al., 1999; Rogers et al., 2003), plants, fruits and vegetables (Carmicha el et al., 1999). High levels of microbial growth indicate limited water treatment efficacy and the potential to harbor plant pathogens in biofilm. The Source had low COD and chlorine demand compared with other Sample Types, which did not meet COD guideli nes (Table 2 2 ). Sample Type and Location

PAGE 30

30 significantly affected COD, and chlorine demand was only affected by Sample Type. Sanitizing agent demand can be quantified directly by measuring the difference between applied and residual active ingredient concen tration when a sanitizer is injected into a water sample, or indirectly by the COD. The COD represents the amount of oxygen required to completely oxidize contaminants in water, and is therefore correlated with the required sanitizing agent concentration ( Tanwar et al., 2008). Biodegradable organic matter in reused irrigation water and industrial effluent is monitored for compliance with COD to ensure the sanitizing agent demand is not too high (USEPA, 2004). The COD was highest from Catchment Basin samples although Catchment Basins had a low chlorine demand compared with most other Sample Types. The low chlorine demand probably occurred in Catchment Basin samples compared with Subirrigation, Tanks, and Furthest Outlet samples, because the latter three Samp le Types would have contained nitrogen from water soluble fertilizer. Published guidelines for chlorine demand suggest a minimum free chlorine residual of 1 mgL 1 after a 30 mi n. contact time (USEPA, 1999). However, this free chlorine level is unlikely to occur when reduced forms of nitrogen such as ammonium fertilizer are present, resulting in complexed chlorine. Chlorine demand of 8.2 mgL 1 for Subirrigation water, for example, can be interpreted as a requirement for 10.2 mgL 1 of applied chlorine in o rder to provide 2 mgL 1 of residual free chlorine. Changes in microbial water quality from Source to Furthest Outlet (Table 2 3) indicated a general trend of declining water quality as water was distributed through irrigation lines. For both aerobic bacte ria and yeast and mold counts, there were

PAGE 31

31 yes/no variable corresponding to whether water was captured and reused between the Source and Furthest Outlet). In one case (Loca tion #21) there was a significant reduction in aerobic bacteria counts from Source to Furthest Outlet (Table 2 3), where the Source was from a shallow well followed by treatment with peroxyacetic acid (Table 2 1). Significant increases in aerobic bacteria between Source and Furthest Outlet occurred in 13 Locations, despite chemical treatment in 9 of these cases (Table 2 3). In general, there were higher microbial counts in recycled irrigation. However, there were exceptions such as the high counts in Locati on 21, 7 and 8 despite no recycling. The change in total microbial counts at sampling points before and after sanitizer injection was used as a bio indicator to monitor the capability of the water treatment system to control microorganisms at each Location Interpretation of the efficacy of the treatment systems is limited by only one day of sampling, and limited replication of technologies between Locations. There was a significant interaction between Location and Treatment for aerobic bacteria as well as yeasts and molds (Table 2 4). Eight Locations exceeded the recommended 10,000 cfu mL 1 for aerobic bacteria pre treatment, and in four of those cases the Post Treatment level was below 10,000 cfu mL 1 Irrespective of Pre Treatment counts of aerobic bacter ia were above 10,000 cfu mL 1 at the Post Treatment sampling point, including one Location with calcium hypochlorite, four with copper ionization, one with peroxyacetic acid, and one with reverse osmosis. The high Post Treatment bacteria counts indicated s everal cases where treatment systems were not providing biological irrigation quality within recommended guidelines.

PAGE 32

32 Monitoring of the physical, chemical, and microbial water quality can improve the efficacy of treatment with reliable control of pathogens without causing corrosion of pumps and equipment (Chang et al., 1998) and to improve environmental safety factors (USEPA, 1999). The highest water quality occurred in Source water, with the exception of pH (Table 2 2). Appropriate treatment for alkalinit y and pH in Source water would include acid injection and possibly reverse osmosis. Sample Types after than the Source tended to have lower UV transmission, and high organic carbon, aerobic bacteria and COD, and not all systems were managing microbial load The greatest potential risk of clogged emitters in irrigation systems resulted from aerobic bacteria levels above 10,000 cfu mL 1 Of the eight Locations with Pre Treatment bacterial counts were above 10,000 cfu mL 1 (Table 2 4), treatment lowered bacter ia counts below 10,000 cfu mL 1 in only half the locations. Improvements to irrigation systems are required to reduce bacteria counts, including increased filtration to remove suspended particles and overall sanitizing demand, followed by more effective ch emical treatment. These results highlight that irrigation monitoring of biological, chemical, and physical water quality can be used to improve irrigation system design when implementing a water treatment technology. The density of aerobic bacteria can be counted to indicate the relative efficacy of a treatment technology to control microbes in greenhouse and nursery irrigation, particularly when providing chemical treatments or micro emitter irrigation using recycled water. More detailed analysis of Pre a nd Post Treatment for each Location should be conducted to determine if the treatment systems were effective or failed in control of aerobic bacteria, yeast and mold. This survey was

PAGE 33

33 not intended for comparison between treatment technologies because it was conducted at only one point in time. Installation of sampling ports before and after each treatment point should be a standard design feature to improve the ability to monitor irrigation quality. Regular monitoring of sanitizing agent levels and biologica l water quality may provide assessment of the water quality changes that are likely to occur over time and within a horticultural irrigation system. Before this survey, none of the Locations had measured physical water quality and Pre and Post Treatment w ater quality variables such as aerobic bacteria counts, which are important monitoring variables to ensure correct operation of a treatment system. The results of this study provide a reference for irrigation quality and may be used as a guideline to devel op and implement monitoring procedures to improve irrigation management.

PAGE 34

34 Table 2 1. Descriptive characteristics for the production environments from each greenhouse or nursery (Location) included in the survey of irrigation quality. Each Location is l isted by row and identified by the code number (1 24) presented in the first column. Each row represents a different Location, and specifies the production area, water sources, and irrigation technologies used. Code State Production area z (ft 2 ) y Main water source x Well depth (ft) y Daily water used v (gal) y Recycled % u Filtration Filter pore size (m) y Treatment system Acid 1 Florida 174,235 well 141 20,002 0 screen > 300 calcium hypochlorite sulfuric 2 Florida 999,999 well 440 100,011 30 screen > 1,000 cal cium hypochlorite none w 3 Colorado 78,749 municipal none w 10,001 0 centrifuge none w calcium hypochlorite phosphoric 4 Florida 500,005 well 180 135,014 0 screen > 300 chlorine dioxide none w 5 Florida 653,003 well 331 64,267 0 centrifuge, screen 120 sodiu m hypochlorite sulfuric 6 Michigan 600,001 municipal none w 1,000,105 75 screen, paper 28 sodium hypochlorite, ozone sulfuric 7 Florida 72,000 well 131 35,004 90 screen > 300 copper ionization inline none w 8 Florida 900,002 well 200 80,008 0 sand, screen 100 copper ionization inline sulfuric/phosph oric 9 New Jersey 799,995 well 131 6,001 0 screen 100 copper ionization inline none w 10 North Carolina 6,533,989 pond none w 1,000,105 100 centrifuge, sand, paper 28 copper ionization inline none w 11 Californi a 3,999,994 pond none w 490,052 100 centrifuge, screen 100 copper ionization inline sulfuric/phosph oric 12 New Hampshire 200,004 well 801 20,002 75 centrifuge, screen 100 copper ionization tank none w 13 New Hampshire 121,998 well 125 45,005 50 paper 28 fi ltration only sulfuric 14 New Hampshire 93,000 well 121 30,003 50 centrifuge, screen 100 filtration only sulfuric 15 California 2,599,999 municipal none w 60,006 0 screen none w filtration only none w 16 Michigan 200,004 well 89 90,010 40 screen 100 filtra tion only nitric 17 Florida 1,012,474 well 26 15,002 0 screen > 300 filtration only sulfuric 18 Michigan 729,997 well 112 80,008 10 paper 18 filtration only sulfuric 19 Minnesota none w well 200 none w none w centrifuge > 300 filtration only none w

PAGE 35

35 z Production area represents the size of the production area that irrigation was sampled from. y 1 ft 2 = 0.0929 m 2 1 ft = 0.3048 m; 1 gal = 3.7845 L; 1 m = 1 micron x Main water source identifies where the irrigation originated; well was irrigation pumped from below ground, municipal was treated city tap water and pond was irrigation collected in a Catchment Basin. w Indicated a value was not reported. v Estimat ed total daily irrigation volume used, as reported for each Location. u Recycled is the percentage of irrigation runoff that was captured and recirculated in a subirrigation loop or held in a Catchment Basin. Table 2 1. Continued Code State Production area z (ft 2 ) y Main water source x Well depth (ft) y Daily water used v (gal) y Recycled % u Filtration Filter pore size (m) y Treatment system Acid 20 Michigan 479,995 well 121 55,006 0 centrifuge 100 hypochlor ous acid sulfuric 21 Florida 232,124 well 26 6,001 0 centrifuge > 300 peroxyacetic acid sulfuric 22 New Jersey 769,996 well 49 250,026 15 screen, paper 28 peroxyacetic acid sulfuric 23 Michigan 215,999 well 161 35,004 0 reverse osmosis < 1 peroxyacetic acid sulfuric 24 New Mexico 699,998 well 98 70,007 40 reverse osmosis < 1 reverse osmosis none w

PAGE 36

36 Table 2 2 ANOVA irrigation quality results from the surveyed population of 24 greenhouses and nurseries (Locations). Irrigation quality variables were analyzed by Location, Sample Type (Source, Tank, Furthest Outlet, Subirrigation, or Catchment Basin), and their interaction (Location Sample Type ). Least squared means are presented for each Sample TypeValues followed by a superscript notation indicate they are outside recommended guidelines (see footnotes below). ANOVA effect y Least square means by Sample Type m Recommended ranges Water quality variable Sample Type Location Sample Type Location Source Tank Furthest Outlet Subirrigation Catchment Basin Physical Total suspended solids (mgL 1 ) z *** *** NS 3.1 a 4.4 a 5.5 a 9.6 b 9.7 b < 30 t Organic carbon (mgL 1 ) z *** *** ** 2. 0 a 2.9 s a 3.8 s a 4.9 s b 6.2 s b s UV x transmission (%) *** *** NS 86 a 75 bc 76 b 68 r c 72 r bc > 75% r Chemical Alkalinity v (mgL 1 ) z NS *** NS 128 c 59 ab 130 bc 55 a 120 c < 150, < 300 q Total dissolved solids (mgL 1 ) z *** NS 425 a 528 q ab 502 q ab 676 q b 354 a < 480, < 960 q pH *** *** NS 7.6 p cd 6.8 ab 7.3 bc 6.3 a 8.1 d < 7.5 p Electrical conductivity (Scm 1 ) z ** *** NS 660 a 840 q ab 787 q ab 1110 q b 553 a < 750, < 150 q Microbial Aerobic bacteria ( cfumL 1 ) z *** ** ** 1,108 a 65,229 o ab 184,728 o b 117,431 o b 435,336 o b < 10,000 o Yeast and mold ( cfumL 1 ) z ** NS 6 a 871 a 94 a 3,181 b 980 a ND w

PAGE 37

37 ANOVA effect y Least square means by Sample Type m Recommended ranges Water quality variable Sample Type Location Sa mple Type Location Source Tank Furthest Outlet Subirrigation Catchment Basin Overall sanitizing demand Chemical oxygen demand (mgL 1 ) z *** ** NS 12.0 a 37.7 t bc 30.5 t b 48.2 t cd 61.3 t d <30 t Chlorine demand at 2 min (mgL 1 ) z ** NS NS 1.4 a 8.7 b 6.4 ab 8.2 ab 3.2 ab ND w Clogging risk Physical 0 0 0 0 0 < 10 n Chemical 6 5 5 6 5 < 10 n Microbial 2 10 n 10 n 10 n 10 n < 10 n Total 8 15 n 15 n 16 n 15 n < 10 n z 1 mgL 1 = 1 ppm; 1 scm 1 = 0.001 mmho/ cm; 1 cfumL 1 mL = 29.5735 cfu/ fl oz. y Statistical significance indicated by NS, *, **, *** are nonsignificant or significant at P x Ultraviolet light (UV) transmission. w ND is used to indicate no data. v Al kalinity value reported indicated the equivalent concentration of calcium carbonate (CaCO 3 ) that would need to be neutralized by a strong acid. u Colony forming units (cfu). t Guidelines from U.S. Environmental Protection Agency (2004). s Recommendations pro vided by Chang et al. (1998). r Recommendations from Newman (2004). q Guidelines ranges for plug production and floricultural crops from Whipker et al. (2003). p Recommendations from Copes et al. (2004). o Recommendations from Rogers et al. (2003). n Guideline s from Bucks et al. (1979). m P Table 2 2 Continued

PAGE 38

38 Table 2 3. ANOVA table for aerobic bacteria and yeast and mold cfumL 1 for each greenhouse and nursery (Location) surveyed. At each L ocation, samples were collected at the Source and Furthest Outlet. For Location 7, there were two Source and Furthest Outlet irrigation circuits which were analysed separately. Locations 20 and 24 did not have Furthest Outlet sampling points and were not i ncluded. The reported cfumL 1 of aerobic bacteria and yeast and mold are least square means for each Location, with statistical comparison between Source and Furthest Outlet by Location. Location code Treatment w Recycled (Catchment Basin or Tank) v Aero bic bacteria ( cfu z /mL) y Yeast and mold ( cfu z /mL) y Source Furthest Outlet Significance x Source Furthest Outlet Significance x 1 yes no 1,203 1 NS 0 0 NS 2 yes yes 1,066 576,825 *** 12 560 3 yes no 1 1 NS 0 0 NS 4 yes no 1,522 12,504 NS 0 42 N S 5 yes yes 9,376 2,463,801 *** 141 31 NS 6 yes yes 0 1,100 NS 0 153 NS 7 yes no 92 217,884 *** 4 459 ** 7 yes yes 92 395,213 *** 4 0 NS 8 yes no 11 103,948 *** 0 8 NS 9 yes no 236 570 NS 1 0 NS 10 yes yes 14,460 65,142 *** 123 730 11 yes yes 1 172,732 *** 0 69 NS 12 yes no 340 54,602 *** 0 2 NS 13 no yes 33 206,016 *** 0 2,910 ***

PAGE 39

39 Table 2 3 Continued Location code Treatment w Recycled (Catchment Basin or Tank) v Aerobic bacteria ( cfu z /mL) y Yeast and mold ( cfu z /mL) y Source Furthest Outlet Significance x Source Furthest Outlet Significance x 14 no yes 3 34,303 *** 0 8,508 *** 15 no no 8,714 671 NS 0 60 NS 16 no yes 232 199,469 *** 7 107 NS 17 no no 1,135 253 NS 0 0 NS 18 no no 20,722 28,500 *** 7 0 NS 19 no no 9 1 N S 0 0 NS 21 yes no 248,054 294 *** 688 1,447 NS 22 yes yes 21 35,604 *** 0 13,435 *** 23 yes no 38 18 NS 0 196 NS Location *** *** Sample Type *** Treatment NS NS Recycled *** *** Location Sample Type *** *** Treatment Rec ycled *** NS z Colony forming units (cfu). y 1 cfu/mL = 29.5735 cfu/ fl oz. x Statistical significance indicated by NS, *, **, *** are either not significant or significant at P respectively. w Indicates whether chemical sanitation was used (yes) or not used (no) between the Source and Furthest Outlet. v Indicates whether irrigation water was collected and held in a Catchment Basin or Tank between the Source and Furthest Outlet (yes) or was not held in a Catchment Basic or Tank (no).

PAGE 40

40 Table 2 4. ANOVA table for effect of sanitation treatments on the cfumL 1 of aerobic bacteria and yeasts and molds. Samples were collected immediately before (Pre Treatment) an d immediately after (Post Treatment) each sanitation treatment. The reported density of aerobic bacteria and yeast and mold are least square mean cfumL 1 for each Location followed by significance between Pre Treatment and Post Treatment. Location Sanitat ion treatment Aerobic bacteria (cfu z /mL) y Yeasts and molds (cfu z /mL) y Pre Treatment Post Treatment Significance x Pre Treatment Post Treatment Significance x 1 calcium hypochlorite 224,818 1 *** 33 0 2 calcium hypochlorite 325,151 19,791 *** 33 17 NS 3 calcium hypochlorite 1 1 NS 0 0 NS 4 chlorine dioxide 1,522 286 NS 0 2 NS 5 sodium hypochlorite 215,705 40 *** 10 3 NS 6 sodium hypochlorite 0 0 NS 0 0 NS 6 ozone circulation tank 940 478 NS 149 49 NS 7 copper ionization inline 92 46 NS 4 2 NS 8 copper ionization inline 104 25,606 *** 0 34 9 copper ionization inline 236 665 NS 1 0 NS 10 copper ionization inline 6,516 65,142 NS 315 730 NS 11 copper ionization inline 393,443 1,462,866 *** 484 140 NS 12 copper ionization tank 340 39,689 *** 0 1 NS 21 peroxyacetic acid 261,683 300 *** 1,405 1 *** 21 peroxyacetic acid 86,201 9 *** 75 10 22 peroxyacetic acid 33,015 122,724 NS 18,002 15,841 NS 23 peroxyacetic acid 75 58 NS 0 0 NS 24 reverse osmosis 149,027 65,183 *** 372 28 *** Treatment *** *** Location *** *** Treatment Location *** *** z Colony forming units (cfu). y 1 cfu/mL = 29.5735 cfu/ fl oz. x Statistical significance indicated by NS, *, **, *** are not significant or significant at P respectively.

PAGE 41

41 CHAPTER 3 THE USE OF DEHYRATED MEDIA FOR ONSITE MONITORING OF AEROBIC BACTERIA IN IRRIGATION WATER An onsite method to monitor microbial density in irrigation water would assist management of water quality and treatmen t in the horticultur e industry. Population density of aerobic bacteria (expressed as colony forming units per milliliter of the original sample, cfu mL 1 ) in irrigation water can be monitored to determine the risk of clogged irrigation lines from bio film (Bucks et al., 1979). Biofilms are multicellular communities of microorganisms, and may include bacteria, fungi and algae as well extracellular polymers produced and excreted by the organisms within the biofilm. Any wet surface or an interface between air and a highly eutrophi c (nutrient rich liquid) can h ost biofilms (Cullimore, 1993). Biofilms are common inside water storage tanks and irrigation lines (Ravina et al., 1997), where they can reduce water flow, corrode metals, reduce efficacy of sanitation agents, and serve as so urces of pathogens. Biofilms are an important part of microbial lifecycle and they can serve as both a sink and source of microbes, depending on the environmental conditions (Cullimore, 1993). Monitoring and controlling biofilms, therefore, is important to maintain the q uality of the irrigation water. At present, growers can send irrigation water samples to a laboratory, where standard methods are available for bacterial analysis under aseptic conditions (Darbie et al., 200 6). However, an onsite test for ae robic bacteria cfu mL 1 could provide rapid comparison with a recommended threshold of less than 10,000 cfu mL 1 for acceptable risk of biofilm clogging of drip and mist emitters (Rogers et al., 2003). Total cfu mL 1 of aerobic bacteria tested at points al ong the irrigation system and immediately before and after treatment (for example, chlorine injection) can serve as a

PAGE 42

42 non specific indicator of disinfection efficacy (APP HA, 1995; Maier et al., 2009). A 2 to 3 log reduction of bacterial cfu mL 1 from pre t reatment to post treatment indicates acceptable treatment reliability f or disinfection (USEPA, 1999). Quantification of aerobic bacteria may be more easily accomplished as an onsite test for growers than quantification of plant pathogens in irrigation wate r, because pathogen numbers are often low compared with beneficial or benign micro organisms, and quantification of aerobic bacteria does not require a specialized assay. Petrifilm TM (3M, Saint Paul, MN) is a plating technology with a ready made, dehydrate d culture medium with nutrients and tetrazolium salt that fluoresces living bacteria colonies on a flat c ard with a laminate cover slip. Petrifilm is widely used in food and beverage processing to determine cfu mL 1 of aerobic bacteria, yeasts & molds, and sometimes pathogens such as pathogenic E. coli or Salmonella spp. depe nding on the type of Petrifilm. The Petrifilm technology has also been used to evaluate sanitation processes such as chlorination in post harvest handling (Materon, 2003), to determine sources of contaminants and test the ability of water treatment systems to control human pathogens (Riordan et al., 2001). Many standardized industrial processes are monitored with Petrifilm and are recognized as a proven technology for food safety by the USDA, USEPA and inte rnational regulatory agencies. The research objective was to validate the use of Petrifilm to accurately quantify total aerobic bacteria in horticulture irrigation water compared with standard c ulture substrate. Subirrigation return wat er from a Florida nursery was plated in a dilution series with Petrifilm and R2A agar substrates.

PAGE 43

43 Materials and Methods Samples were collected from a subirrigation recovery tank in a large scale commercial green house in Apopka, FL. On Nov. 4, 2011 a total of 18 irrigation samples were randomly collected after a coin flip determined whether or not to collect (heads) or not collec t (tails) at 15 min intervals. All samples were collected over a 6 h monitoring period from one subirrigation return tank, du ring m ultiple flooding events. Water quality measurements (pH, electrical conductivity, dissolved oxygen and temperature) were stable duri ng collection (data not shown). Samples (300 mL) were collected into sterile 500 mL Whirl pak bags (Nasco, Ocala, FL) and m aintained in temperature controlled vessels at 22 to 25 C, similar to t he initial sample temperature. Samples were immediately transferred to the laboratory at UF and placed into culture within 6 h of collection. The R2A medium was prepared by mixing 15. 2 g of the dehydrated mix (Table 3 1, Remel Labs, Lenexa, KS) dissolved in 1 L of purified boiled water and adjusted to pH 7.2 at 25 C, autoclaved and cooled to 45 50 C in a water bath prior to pouring into sterile 50 mm diameter Petri dis hes (Fisherbran d #09 720 500). Petrifilm (Table 3 1) Total aerobic bacteria for nursery water samples were quantified using Petrifilm (method 990.12, Petrifilm Aerobic Count Plate method (AOAC, 1999) ) or R2A as described by method 988.18, Aer obic Plate Count (AOAC, 1999). The sample was distributed over the substrate using the Spread Plat e Method 9215 C (APPHA, 1995). Samples were mixed with a vortex shaker and diluted in a series of direct (10 0 ) to 1 0 8 with sterile saline/ peptone diluent, and two repl icates per dilution were used. Transfers to culture followed methods described for aseptic transfer and culture (APPHA, 1995).

PAGE 44

44 Plates were incubated in the dark at 25 to 27 C in a growth ch amber at 70 % relative humidity. Colonies were counted on both media at 3 and 7 days, to satisfy the recommended incubation period for Petrifi lm and R2A agar, respectively. The paired samples for Petrifilm and R2A at day 3 and 7 were compared with ANOVA in PROC GLM us ing log linear regression with PROC REG in SAS. Results and Discussion There were significant effects of substrate type, (p<0.001), day (p<0.001), and their interaction (p<0.01) on bacteria cf u mL 1 (with least s quare means shown in Table 3 2). As time increased f rom day 3 to day 7, bacteria cfu mL 1 increased. The R2A substrate tended to have higher bacterial colony counts than Petrifilm on a given measurement day. A comparison based on standa rd recommendations for incubation periods (3 days for Petrifilm, and 7 days for R2A (AOAC, 1999) showed that averaged counts on Petrifilm were lower f or most sample times (Table 3 2). For all methods and measurements, the bacterial cfu mL 1 greatly exceede d the recommended threshold of 10,000 (Rogers et al., 2003). Both methods would therefore indicate a h igh risk of biofilm formation. Conversion between bacteria cfu mL 1 estimates from the two substrates would depend on the measurement day, based on the si gnificant interaction between substrate and measurement day. To convert bacteria CFU/mL from Petrifilm on a given number of days of incubation to the bacteria CFU/mL estimated with R2A, the Petrifilm bacteria count CFU/mL (x 10 4 ) could be multiplied by 65.9/ 50.1 = 1.3 for day 3, or 146.2/ 83.1 = 1.79 on day 7 using least square means in Table 3 2. Based on standard

PAGE 45

45 recommendations for incubation periods (3 days for Petrifilm, and 7 days for R2A (AOAC, 1999), the conversion factor between R2A and Petrifilm would be 146.2/ 50.1 = 2.92. This study confirmed that Petrifilm can be used to quantify aerobic bacteria in irrigation water to estima te its microbiological quality. However, given the effect of substrate and measurement day on bacteria cfu mL 1 a comparison of counts from Petrifilm to recommended guidelines on R2A would need a calibration. Based on suggested incubation time the counts on Petrifilm after 3 days of incubation can be used to estimate the counts on R2A after 7 days of incuba tion time. Further validation with a known bacterium or multiple irrigation sources would be helpful to evaluate whether differences in counts between the substrates is consistent for all microorganisms of concern. Differences in cfu mL 1 between substrate s may have resulted from different growth response of aerobic bacteria species to the nutrient formulat ions and physical environment. A limitation of the Petrifilm is that selective substrates are not currently available for specific plant pathogens. There fore, the technique is a non specific measurement of aerobic bacteria that may be suitable for quantifying microbial load, estimating a potential for biofilm formation and relative efficacy of water treatment systems along points in an irrigation line, rat her than to identify presence or absence of a specific plant pathogen.

PAGE 46

46 Table 3 1 Ingredients for the two culture substrates. Ingredients R2A agar Petrifilm Agar 15% (agarose) 20% (guar) Gl ucose (anhydrous) 0.05% Soluble starch 0.05% P eptone, yeast extract 0.05% 0.02% Peptone, casein hydrolysate 0.5% 0.50% to 1.5% Peptone (other) 0.5% from beef 0.5% from soybean Magnesium sulfate 0.05% 0.05% Potassium chloride 0.05% Potassium (sulfate, phosphate) 0.3% 0.1% Sodium alginate 0.5 4% Sodium pyruvate 0.3% <0.2% Table 3 2 ANOVA effects Effect of substrate (Petrifilm or R2A) and measurement day (3 or 7) on bacteria c fu mL 1 Substrate, day, and their interactio n were significant at p<0.001. level for the subst rate x day least square means. Data were log transformed for ANOVA anal ysis, and back transformed for presentation in this table. Day 3 (cfu mL 1 ) Day 7 ( cfu mL 1 ) Least square mean R2A 6 5.9 B 1 46.2 D 9 8.1 Pet rifilm 5 0.1 A 8 3.1 C 6 4.5 Least square mean 5 7.4 1 10.2 Std. error 2 .9 2 .8

PAGE 47

47 CHAPTER 4 COMPARISON BETWEEN STANDARD AND DEHYDRATED AGAR TO QUANTIFY FUNGI PHYTOPHTHORA CACTORUM AND XANTHOMONAS CAMPESTRIS IN WATER SAMPL ES Introduction Recirculating irrigation systems provide ideal conditions for microbial growth, development of biofilms, and transport of waterborne pathogens. A wide range of pathogen species have been isolated and identified from greenhouse and nursery irrigation waters, including water molds ( Pythium and Phytophthora spp.), fungi, bacteria and viruses (Hong and Moorman, 2005). Biofilms, which are complexes of algae, bacteria, and other microbes, are a major cause of clogged micro irrigation equipment (B ucks et al., 1979). Sutton et al. (2006) showed that oospores of Pythium spp can be encountered in biofilms, and serve as a source of inoculum in hydroponic production. Waterborne p athogens can cause devastating damping off, root and stem rots, and blight s on crops of annuals, grasses, vegetable, and fruit crops (Hendrix and Campbell, 1973; Pernezny and Roberts, 2008; Simone and Momol, 1990). To control biofilm and pathogens, a range of water disinfestation technologies are used to inhibit the growth of bi ofilms and waterborne pathogens, including filtration, chlorination, copper ionization, heat, and ultraviolet radiation (Van Os, 2009). Risk of biofilm formation in micro irrigation systems can be quantified in terms of aerobic bacteria cfu mL 1 with a t hreshold of less than 10,000 cfu mL 1 recommended for acceptable control of biofilm clogging of drip and mist emitters (Rogers et al., 2003). The recommended cfu mL 1 threshold was based on approved liquid gel agar preparations, such as R2A and PDA, which are described in standard method 9215A (APHA, 1988). However, the measured microbial density is likely to depend on both the

PAGE 48

48 substrate and methods used to collect, handle, and ship samples (Watkins and Jian, 1997). Microorganisms can become environmentally stressed during shipment and storage of samples and be rendered viable but non culturable. Under such conditions, quantitative assessment of population numbers may underestimate actual microbial load. On farm culturing of microbes may therefore be desirab le. However, traditional liquid gel agar is prone to contamination with airborne microbiological spores under the non aseptic conditions common in the field. Petrifilms TM (3M Microbiology, St. Paul, MN) are a plating technology that includes a dehydrated culture medium on a plastic card with a cover slip, designed for quantification of microorganisms by industries such as crop and food production and processing. These ready made culture systems contain standard nutrients as described in method 940.36 (AOAC 1999), a cold water soluble gelling agent, and an indicator dye. Petrifilms are widely used to quantify ( cfu mL 1 ) aerobic bacteria, yeast and molds, and specific food safety organisms such as E. coli or Salmonella depending on the components in the deh ydrated culture media used for a particular Petrifilm test. The U.S. Department of Agriculture and international regulatory agencies have validated several standardized monitoring techniques that utilize Petrifilms for culturing aerobic bacteria, yeast and molds, and specific food safety organisms as a means to quantify contamination and efficacy of pathogen treatment in food production systems, irrigation and water supplies, and recreational waters ( Materon, 2003; Riordan et al., 2001). A protocol (Fig. 4 1) was developed to quantity aerobic bacterial and fungal cfu mL 1 in irrigation water using Petrifilm AC (#6400, #6406) and Petrifilm YM (#6407, # 6417). A erobic bacterial and fungal cfu mL 1 were quantified in a survey of 24

PAGE 49

49 greenhouse and nursery irrig ation systems throughout the U.S. (Meador et al., 2012). Petrifilm AC counts showed trends of increased microbial load from source water (averaging 1.1 x 10 3 cfu mL 1 ) to the outlets furthest in the irrigation system from the source (1.84 x 10 5 cfu mL 1 or catchment basins (4.35 x 10 5 cfu mL 1 ). The high cfu mL 1 indicate biofilm risk, increased demand for sanitizing chemicals such as chlorine, and therefore increased chlorine application rate required to control target pathogens. Fungal cfu mL 1 quantified with the Petrifilm YM showed similar trends of increased cfu mL 1 in recirculated water compared with well or municipal water. However, in contrast to aerobic bacteria, target guidelines are not available for cfu mL 1 for fungi or oomycetes in water sampl es largely because of their complex life stages, species diversity, and variability in viable colony numbers even under controlled laboratory conditions. Additional research is needed to compare Petrifilm estimates of cfu mL 1 with standard culture plating techniques. Preliminary research suggested that aerobic bacteria cfu mL 1 in a recirculated irrigation sample cultured onto Petrifilm AC was lower than cfu mL 1 of aerobic bacteria cultured onto R2A (Meador et al., 2013). Research is not available compari ng the Petrifilm YM against standard techniques to culture and quantify fungi. In addition, single species cultures of aerobic bacteria and oomycetes that are known to occur in irrigation water have not been validated for Petrifilms. The research objectiv e was to compare cfu mL 1 from Petrifilms and traditional agar preparations using either a recirculated irrigation solution or single known microbial species. Three experiments were run: (1) Quantitation of fungal cfu mL 1 on Petrifilm YM and a fungal sele ctive culture medium (SDA) with a recirculated irrigation solution, (2)

PAGE 50

50 Quantitation of cfu mL 1 of the oomycete Phytophthora cactorum (zoospores) on Petrifilm YM, the oomycete selective culture medium PARP H, and a non selective medium (PDA), and (3) Quan titation of cfu mL 1 of the bacterium Xanthomonas campestris on Petrifilm AC, PDA and R2A. Materials and Methods Experiment 1: Fungal Counts During Subirrigation C ycles Standard culture plates followed preparations as described in method 988.18 (AOAC, 1999 ). The SDA medium was prepared by suspending 15.2 g of the dehydrated mix (Table 4 1, Remel Labs, Lenexa, KS) in 1 L of DI. The suspension was buffered to pH 7.2 and was then autoclave sterilized and cooled and maintained between 45 50 C in controlled wat er bath. To select for fungi and inhibit the growth of bacteria, antibiotics were added to the media cooled to 50 C. Chloramphenicol was added at 50 mg/L (dissolved in 10 ml of 95% ethanol and filter sterilized) to the cooled molten media. Gentamycin was dissolved in 2 ml of acetone and then added to molten agar to provide 50 mg L 1 Molten agar with antibiotics were poured at 45 C into sterile 50 mm diameter Petri dishes (Fisherbrand #09 720 500) under a laminar flow hood. Plates were cooled to room temp erature (25 C) and solidified prior to culture plating. Petrifilms are made with 2 component layers: a bottom layer is an adhesive coated film with spray dried, solubilized dry medium. The second layer is a cover sheet consisting of a thin, transparent p olypropylene film (laminate sheet) that is also coated with dehydrated medium and an adhesive. A phosphatase indicator is then added to identify active colonies. The Petrifilms are 7.68.9 cm and the cover sheet and bottom layer are heat sealed together at one edge and then sterilized in ethylene oxide.

PAGE 51

51 Petrifilm YM preparation (Table 4 Paul, MN.). The Petrifilm YM plates contained a guar gum (#HP 11, Celanese Corp.), xantham gum (Kelco Company, Chicago, Il l.) and the standard nutrient components (supplied as half of the weight of the guar and xantham gum). The laminate sheet may also contain compounds that can inhibit certain groups of bacteria and fungi or water molds. Irrigation samples were collected fr om a subirrigation recovery tank in a large scale commercial greenhouse in Apopka, FL on Nov. 4, 2011. A total of 18 irrigation samples were collected randomly after a coin flip determined whether or not to collect (heads) or not collect (tails) at 15 min intervals. All samples were taken over 6 h from one subirrigation return tank, during multiple flooding events. Water quality measurements (pH, electrical conductivity, dissolved oxygen and temperature) were stable during collection (data not shown). Sampl es (300 ml) were collected into sterile 500 ml Whirl pak bags (Nasco, Ocala, FL) and maintained in temperature controlled vessels at 22 to 25 C, similar to the initial sample temperature. Samples were transferred immediately to the laboratory at UF and p laced into culture within 6 h of collection. Culture plating onto SDA and Petrifilm YM was performed under laboratory conditions and sample transfer onto culture medium followed standard methods for aseptic transfer of fungal species, method 9610C (APPHA, 1999). In summary, 1 ml aliquots from the dilution series were transferred to each medium type and plates placed in a dark incubator. Fungal cultures were incubated in the dark at 25 to 27 C in a growth chamber at 70% relative humidity. Colonies were cou nted after the

PAGE 52

52 recommended incubation period for Petrifilm YM and SDA, respectively. Petrifilm YM plates were seeded and incubated for 7 days, according to Standard Method (SM) 997.02 Yeast and Mold Counts in Foods, Dry Rehydratable Film Method (Petrifilm YM method) (AOAC, 1999). Samples plated onto SDA were incubated for 14 days, as described by SM 988.18 Fungal plate count method (AOAC, 1999). E xperiment 2: Quantification of Colonies F rom Z oospores of P cactorum Preparation of PARP H followed methods de scribed by Singleton et al. (1992). The PDA medium was prepared by mixing 15.2 g of the dehydrated mix (Table 4 1, Remel Labs, Lenexa, KS) suspended in 1 L of purified boiled water and adjusted to pH 7.2 at 25 C, autoclaved and cooled to 45 50 C in a wat er bath. Autoclaved medium was slowly cooled and held at 50 C until the molten agar was poured at 45 C into sterile 50 mm diameter Petri dishes (Fisherbrand #09 720 500). Petrifilm YM preparation (Table 4 t Paul, MN). Three separate isolates of Phytophthora cactorum were provided by Natalia Peres of the Univ. of Florida, Plant Pathology Department, Gulf Coast Research and Education Center (UF GREC, Wimauma, FL). Individual cultures were grown from pure cul tures on PDA. Mycelium growth and zoosporangium production of Phytophthora cactorum was maintained and multiplied on PDA as a pre enrichment medium using methods described by Singleton et al. (1992). Nine petri dishes (3 replicates per isolate) provided ab undant zoosporangium development and viable zoospore release was observed under a dissecting microscope after one hour at 25 C. Zoospore concentrations were prepared in sterile solutions and ranged from 0 to 10 8 zoospore

PAGE 53

53 cfu mL 1 Counts were estimated by microscopic examination with a Bright Line Counting Chamber Hemocytometer (Hausser Scientific Horsham, PA). Replicated treatments were mixed with a vortex shaker and diluted in a series of 4 dilutions, with sterile saline an d peptone diluent, and two replicates per dilution. Transfers to culture followed methods described for aseptic transfer and culture (APPHA, 1995). The sample was distributed over the substrate using the Spread Plat e Method 9215 C (APPHA, 1995). One ml sam ple aliquots from the dilution series were transferred to each medium type and transferred onto culture medium followed standard methods for aseptic transfer of fungal species, method 9610C (APPHA, 1999). Concentrations of confirmed living spores released ranged from 1 to 710 8 cfu mL 1 Phytophthora cactorum recovered on PARP H was determined after incubation at 18 C for 21 days. Counts were made after 2, 7, 10, and 14 days and determined again after 21 days, to satisfy the recommended incubation period f or the various medium types. After standard incubation times, the cfu mL 1 on each of the three medium preparations were compared. Plate counts on the Petrifilm YM followed the standard method 9610 C, F, G (APPHA, 2005), Petrifilm Interpretation Guide (3M, Saint Paul, MN), and Standard Culture Techniques for the Determination of Yeast and Molds (AOAC 995.21). Additional testing for Experiment 2 was undertaken, whereby combined inoculum using zoospores, mycelia, sporangia, cysts and other life stages were b lended to prepare stock inoculum and then prepared according to the same culture technique as described in Experiment 2. Individual Petrifilm YM components were then tested to

PAGE 54

54 determine if the top film or bottom card of the Petrifilm YM plates inhibited gr owth or detected different inoculum preparations of zoospores only and blended test suspensions with the various growth stages. Culture purity, contamination and morphological features were observed using various techniques for Gram Staining and mycelium c ell differentiation using Lacto Phenol Cotton Blue and magnification ranges from 10 to 100 (Leica objective #506228). Samples were fixed to frosted slides and observed under oil immersion or under a glass cover slip. E xperiment 3: Quantification of Col onies From Single I solated Xanthomonas campestris Substrate preparation used 15.2 g of R2A (Table 4 2, Remel Labs, Lenexa, Kansas) suspended in 1 L of purified boiled water, adjusted to pH 7.2 at 25 C, and autoclaved for 30 min at 121 C and cooled to 45 50 C in a water bath. Molten agar was then poured into sterile 50 mm diameter Petri dishes (Fisherbrand #09 720 500). The PDA was prepared as described in method 988.18, Aerobic Plate Count (AOAC, 1999). Autoclaved media was slowly cooled and held at 50 C until the molten agar was poured at 45 C into sterile 50 mm diameter Petri dishes (Fisherbrand #09 720 500). Plates were cooled to room temperature (25 C) to solidify the agar prior to culture. Petrifilm AC preparation (Table 4 2) followed manufacturer Paul, MN.). A single isolated strain of Xanthomonas campestris was obtained on October 25, 2011 from Jerry Minsavage at the Univ. Florida, Department of Plant Pathology (Gainesville, FL). Bacterial cells were revitalized from tu rbid suspension stored at 70 C. On November 1, 2011the suspension of X. campestris was brought to room temperature and maintained in the dark at 25 C. A 0.001 ml aliquot was aseptically

PAGE 55

55 transferred and streaked onto nutrient agar (SNA) and incubated in the dark at 25 C for 3 days. Three stock culture isolations were made (A, B, C) from individual colonies of X. campestris aseptically transferred using an ethanol, flame sterilized wire loop and suspended in 3 ml of sterile, buffered (pH 7.2) DI and vorte xed for 30 seconds. The three stock culture isolations were estimated at 5, 10 and 15 million cfu mL 1 of X. campestris There were three replicates made for each stock culture (A, B, C) and provided 9 test suspensions to estimate cfu mL 1 from a dilution series. prepared inoculum of X. campestris A dilution series was made to obtain final concentrations of 10 8 10 7 10 6 10 5 10 4 10 3 10 2 and 10 1 1 and 0 (control) cfu mL 1 of X.campestris Plated inoculum from the direct, control, and each dilution series of X. campestris were transferred in accordance with standard method 9610C for aseptic spread plate transfer (APPHA, 1995). One ml aliquots of the dilution series were pl ated onto Petrifilm AC dehydrated culture media and 0.1 ml was plated on R2A and PDA in was used to distribute inoculum over the Petrifilm AC plate that has a culture are a of approximately 20 cm 2 A borosilicate glass rod was heated and bent to provide the same spread area of 20 cm 2 distributed over the R2A and PDA media. Counts from the direct and dilution series were determined with a high power field (HPF) microscopy an d an improved Neubauer, Bright Line Hemocytometer (Hausser Scientific, Horsham, PA). A calibration was made between the % absorbance at 600 nm (Hong et al., 2008). Total aerobic bacteria cfu mL 1 for samples were quantified using Petrifilm AC (method 990.1 2, Petrifilm AC Plate method (AOAC, 1999))

PAGE 56

56 or R2A and PDA as described by method 988.18, Aerobic Plate Count (AOAC, 1999). Each sample was distributed over the test agar using the Spread Plate Method 9215 C (APPHA, 1995). Bacterial cultures were incubated in the dark at 25 to 28 C with 80% relative humidity. Petrifilm AC plates were incubated according to methods described by 3M. Aerobic bacteria cfu mL 1 were quantified after 3 days (Petrifilm AC) or 5 days (R2A and PDA). Cultures were counted at 10 mag nified light source with standard colony counter. The Petrifilm AC Interpretation Guide was used as a guide to determine valid plate counts within the preferred counting range on a Petrifilm AC plate (25 to 250 colonies per plate). The same criterion was u sed to enumerate the colonies on the R2A and PDA and was comparable to the countable range as described by method 9215A (APPHA, 1995) range between 30 to 300 cfu mL 1 A second trial was conducted that repeated the above procedure on January 5, 2012. Stat istical Analysis The cfu mL 1 estimates were obtained by multiplying the respective dilution factor to colony numbers from plates that provided a countable range between 30 to 150 cfu mL 1 Experiment 1 was analyzed with a two factor analysis of variance ( ANOVA) to evaluate the effect of time and agar type. The estimated cfu mL 1 for each agar type was compared at day 7 and 14 using log transformed data and the GLM procedure (SAS version 9.1; SAS institute, Cary, NC). Mean separation was conducted using Tuk honest significant differences (HSD) and a linear regression model was evaluated using the REG procedure (SAS version 9.1; SAS institute). Experiments 2 and 3 were analyzed with one factor ANOVA to evaluate the effect of agar type at their

PAGE 57

57 recommended incubation times (Petrifilm YM and PDA at day 7 and PARP H at day 14. Log transformed data was used with the GLM procedure (SAS version 9.1; SAS Results Experiment 1: Fungal Counts Dur ing Subirrigation C ycles Fungi recovered from irrigation water samples grew on Petrifilm YM cultures. However, estimated fungal cfu mL 1 was lower than on SDA (Fig. 4 1). There were significant effects of substrate type, (p<0.001), measurement day (p<0.001 ), and their interaction (p<0.001) on fungal cfu mL 1 (Table 4 3). As incubation time increased from day 7 to day 14, the fungal cfu mL 1 increased on both media. The SDA substrate had higher fungal cfu mL 1 than Petrifilm YM on each measurement day. Stan dardized incubation times were used, as described in approved methods (AOAC, 1999; APHA, 1995) and varied for the two substrates, with 7 or 14 days for Petrifilm YM and SDA, respectively. The fungal LOG 10 cfu mL 1 from Petrifilm YM after 7 days in culture was 71% of the count estimated using SDA at day 14 (Table 4 3). When data were back transformed to absolute counts, fungal cfu mL 1 on Petrifilm YM was only 5.5% of the cfu mL 1 estimated on SDA. Experiment 2: Phytophthora cactorum There was no growth of P. cactorum zoospores on any Petrifilm YM culture plates, indicating the culture medium is not a suitable substrate for this organism. On PARP H medium, growth was observed two days after inoculation with zoospores and plate counts were stable for the enti re 3 weeks of the study. There were significant differences in cfu mL 1 between PDA, PARP H and hemocytometer counts (Table 4 4).

PAGE 58

58 PDA provided a higher cfu mL 1 than PARP H and based on LOG 10 transformed results, the percentage of cfu mL 1 recovered on PDA was 88% and PARP H was 78% of the hemocytometer counts. The additional study for Experiment 2 using entire mycelial mats (blended and homogenized) resulted in a faint or muted signal on Petrifilm YM after 7 days, further indicating that this medium was no t suitable for culturing P. cactorum (data not shown). Microscopic evaluation revealed degenerated mycelium, zoospores and sporangia. Experiment 3: Xanthomonas campestris The cfu mL 1 for X. campestris estimated using the Petrifilm AC was similar to the r esults from PDA, but counts from both media were lower than the R2A and hemocytometer counts (Table 4 4). There was no difference between the cfu mL 1 estimated by PDA or Petrifilm AC at any of the three inoculum levels. R2A had a lower estimated cfu mL 1 than the hemocytometer at the low and high inoculum levels, but was equal at the intermediate inoculum level. At the intermediate dose level using back transformed data, PDA was 16% and Petrifilm AC was 14% of the counts estimated by the hemocytometer. Ove rall, results show that interpreting biofilm risk from a cfu mL 1 count for an irrigation water sample against the recommended level of 10,000 cfu mL 1 would be dependent on the substrate. However, specific substrates are sometimes not explicitly mentioned in biofilm monitoring reports (for example, Rogers et al., 2003). Discussion The Petrifilm YM medium was able to culture fungi from recirculated irrigation water and microscopic examination of morphology of the fungi indicated presence of several life sta ges of filamentous fungi, yeasts and molds. However, the inability of

PAGE 59

59 Petrifilm YM to quantify P. cactorum limits the usefulness of this medium for irrigation monitoring, because of the importance of oomycetes as waterborne pathogens. Fungal counts using P etrifilm YM should therefore be interpreted with care because some fungi or oomycetes, such as P. cactorum may not be represented. A potential reason for the inability of Petrifilm YM to detect viable inoculum of P. cactorum is due to the sensitivity of s ome fungi and oomycetes to various antibiotics. The laminate sheet of Petrifilm YM contained methone sulfonate which may inhibit growth and culturability of oomycetes (Table 4 1). In a separate trial, the material and methods were repeated, except the Pet rifilm YM laminate sheet was removed. Growth of P. cactorum was noted after 7 days of incubation on Petrifilm YM without the laminate sheet. The removal of the laminate sheet is not recommended for testing samples because of the possibility of cross contam ination. The PARP H medium, which is selective for oomycetes, estimated a value similar to the number of the viable zoospores counted on the hemocytometer and is an established method for isolating and enumerating spores from environmental samples, such as testing irrigation water. Based on this study, Petrifilm YM would have limited use as an irrigation monitoring tool, and a standard culture method such as SDA (as a total fungal count) or PARP H (for culturing oomycetes) would be recommended. Aerobic bact eria could be quantified using Petrifilm AC, as evidenced by both the culture of X. campestris from this study, and previous research with monitoring of irrigation water (Meador et al., 2013). However, bacterial cfu mL 1 was lower on Petrifilm AC and PDA t han estimated density from R2A and hemocytometer counts. If the risk of equipment clogging and sanitation treatment efficacy is monitored according

PAGE 60

60 to total counts of aerobic bacteria cfu mL 1 an appropriate calibration factor is recommended. Petrifilms provide a monitoring technology that is easy to use as an onsite test, but is limited by the lack of either (a) selective medium designed to detect specific plant pathogens, or (b) a non selective medium that will culture oomycetes in addition to other fun gi. The Petrifilm AC is suitable as a relative measurement to quantify trends in total microbial load from different sampling points along an irrigation line, or as a non selective test of biocidal efficacy of treatment systems by sampling immediately befo re and after treatment. The Petrifilm YM would also provide a general indication of fungal growth, with the limitations discussed earlier related to oomycetes. Estimated bacterial cfu mL 1 depended on the substrate and quantification method. The Petrifilm AC resulted in bacterial counts that were the same as PDA, but were lower than R2A or a hemocytometer count. Therefore, if a level greater than 10,000 cfu mL 1 was quantified with Petrifilm AC, it suggests that the biofilm risk threshold would also be exce eded using other media or quantification methods.

PAGE 61

61 Table 4 1 Ingredients for the four fungal culture substrates tested Ingredients PARP H PDA SDA Petrifilm YM Agar 15% (corn meal agar) 15% (agar 15% (agar) 20% (guar ) V8 Juice 0.05% Glucose/dextr os e 5.0% 4.0% 2.0% Potato extract 5.0% Peptone 0.10% (beef extract) 0.20% (cassein hydrolysate and yeast extract) Magnesium sulfate 0.05% 0.05% Potassium chloride 0.05% Potassium (sulfate, phosphate) 0.30% 0.10% Sodium alginate 0.5 4. 0% Sodium pyruvate 0.30% 0.20% Antibiotics 0.3% a 0.050% b 0.050% c a Pimaricin, Ampicillin, Rifampicin, Pentachloronitrobenzine, Hymexazol b Chloramphenicol, Gentamicin c Methone sulfonate

PAGE 62

62 Table 4 2 Ing redients for the three culture substrates used t o culture Xanthomonas campestris pv. Ingredients R2A PDA Petrifilm AC Agar 15% (agar) 15% (agar) 20% (guar) Glucose/ dextrose (anhydrous) 0.05% 5.0% Soluble starch extract 0.05% 5.0% Peptone, yeast extract 0.05% 0.02% Peptone, casein hydrolysate 0 .50% 0.50 to 1.5% Peptone (other) 0.50% (beef extract) 0.50% (soybean) Magnesium sulfate 0.05% 0.05% Potassium chloride 0.05% Potassium (sulfate, phosphate) 0.30% 0.10% Sodium alginate 0.50 4.0% Sodium pyruvate 0.30% <0.20%

PAGE 63

63 Table 4 3 E agar with chloramphenicol (SDA) and measurement day (7 or 14) on cfu mL 1 of fungi. Substrate, day, and their interaction were significant at p<0.001. Data were LOG 10 transformed cfu mL 1 for both ANOVA and this table. Letters substrate and measurement day Day 7 (LOG 10 cfumL 1 ) Day 14 (LOG 10 cfumL 1 ) Least square mean (LOG 10 cfumL 1 ) Petrifilm YM 3.093 A 3.415 C 3.254 SDA 4.017 B 4.349 D 4.183 Least square mean 3.554 3.882 Table 4 4 Estimated cfu mL 1 of least square mean counts on a hemocytometer, Potato dextrose agar (PDA), and the selective media agar (PARP H) to culture Phytophthora cactorum comparison at the p=0.05 level for the effect of substrate. Counts obtained from Petrifilm YM estimated zero cfu mL 1 of P. cactorum after the standard incubation time and were not included in the analysis. Least squar e mean (LOG 10 cfumL 1 ) Hemocytometer 4.78 A PDA 4.22 B PARP H 3.75 C Table 4 5 Effects of measurement method (R2A, PDA, Petrifilm AC, and hemocytometer) and inoculum density (low, intermediate, and high) with Xanthomonas campestris Main and int eraction effects of measurement method and inoculum density were significant (p<0.0001). Letters represent measurement method and inoculum density. Measurement method and estimated bac terial count (LOG 10 cfumL 1 ) of i noculum Inoculum Level Petrifilm AC R2A PDA Hemocytometer Low 7.2171 F 7.8144 D 7.237 EF 7.9809 C Intermediate 7.2877 EF 8.0546 BC 7.3606 E 8.1447 B High 7.3545 E 8.1102 B 7.334 EF 8.2811 A

PAGE 64

64 Figure 4 1 Fungal c ounts ( cfumL 1 ) recovered from a tank and plated on 2 different substrates. Fungal counts on yeast and mold (Petrifilm YM) were determined on measurement day 7 and SDA on measurement day 14. Data presented were log transformed.

PAGE 65

65 Figure 4 2 Compari son between counting method and substrate, hemocytometer (HEMO) were initial estimated colonies, aerobic count (Petrifilm AC) on measurement day 3 and R2A, PDA on measurement day 5. Data presented were Log10 transformed.

PAGE 66

66 Figure 4 3 Photographs (a) to (e) show the method developed to use Petrifilms for on farm monitoring of microbial cfumL 1 in irrigation system. Fig. 4 3 a (above), r un irrigation to sufficient ly flush lines before sampling. Collect irrigation water as it flows from an emitter, using a sterile disposable vial. When sampling from tanks and catchment basins, draw the sample from the irrigation intake depth using a clean sampling container attached to a pole, and then transfer to the sterile vial. Take care not to contaminate the sample by touching the rim of the sterile vial with the irrigation emitter or hands. M aterials (Petrifilm YM and Petrifilm AC, vials with 99 ml of sterile diluent, and disposable pipettes) are available to growers through a variety of distributors. E stimated cost for the minimum order quantity at the time of this study was about $2.00 per sample on either Petrifilm AC or Petrifilm YM. Figure 4 3 .b Dilute 1 ml of the irrigation water sample by dispensing the sample with a sterile disposable dropper into 99 ml of diluent in a sterile vial. Close the 100 ml diluted sample and mix by inverting the vial until uniform (about 3 5 times). Figure 4 3.c. With a new sterile dropper, take 1 ml from the diluted sample and place onto the Petrifilm card.

PAGE 67

67 Figure 4 3.d. Spreader plate is used to spread the 1 ml sample evenly to cover 20 squares. Figure 4 3. e After 3 to 5 days (Petrifilm AC) or 7 days (Petrifilm YM) of incubation at near room temperature in the dark (office drawer), plates can be counted. Images sh ow Petrifilm AC (red dots, left) and Petrifilm YM (blue dots, right). Multiply the count by 100 to estimate cfumL 1 when using the 1 ml sample:99 ml sterile diluent ratio. It is typically possible to count up to 120 data points per Petrifilm. Multiple dil utions (for example, non diluted, 1/100, 1/10,000) may be needed for more accurate estimation of cfumL 1 in very clean o r highly contaminated samples. If only a single dilution will be prepared for Petrifilm AC, we recommend the 1/100 dilution because cou nts less than 10 per plate indicate fairly clean samples in terms of biofilm risk, whereas plates that are too numerous to count at this dilution indicate high biofilm risk. Photos courtesy of Dustin Paul Meador

PAGE 68

68 CHAPTER 5 AMMONIUM AND UREA IN NUTRIENT SOLUTIONS DECREASE FREE CHLORINE CONCEN TRATION FROM SODIUM HYPOCHLORITE Chlorine from calcium hypochlorite, sodium hypochlorite, or chlorine gas sources has a complex chemistry in irrigation water. Once added to water, chlorine is converted to the hypochlorite ion (OCl ) and hypochlorous acid (HOCl), which along with dissolved chlorine gas are collectively termed free chlorine. The balance between hypochlorous acid and hypochlorite is pH dependent, whereby hypochlorous acid predominates at solution pH below 7.5 or hypochlorite ions above pH 7.5 (Morris, 1966). Water pH also influences the sanitizing strength of a chlorine solution, because HOCl is estimated at up to 80 more effective as a biocide than OCl (White, 1992, p.269). The effect of pH on control of waterborne pathogens in a chlorine s olution was illustrated by Lang et al. (2008), whereby control of Pythium aphanidermatum and P. dissotocum zoospores was achieved at either 2.0 mgL 1 Cl at pH 8.1, or 0.5 mgL 1 Cl at a lower pH of 6.3. There is extensive research on effects of free chlo rine on waterborne plant pathogens. Control of Pythium and Phytophthora zoospores has been reported with 2 mgL 1 of free chlorine (Cayanan et al., 2009; Hong et al., 2003; Hong and Richardson, 2004; Lang et al., 2008). Chlorine oxidizes and chlorinates li ving tissue and organic compounds, resulting in damage to membranes, enzymes and nucleic acids of microorganisms (Stewart and Olson, 1996). Increasing concentration of chlorine, lowering solution pH, increasing ORP, and increasing contact time are factors that improve pathogen control (Lang et al., 2008). Oxidative strength of a chlorine solution can be measured in millivolts (mV) using an ORP meter. A positive correlation was found between ORP and control of coliform bacteria by chlorination of waste wate r (Yu et al., 2008). Coliforms and pathogenic

PAGE 69

69 bacteria were rapidly controlled in post harvest wash water if ORP was maintained between 650 to 700 mV (Suslow, 2004). Lang et al. (2008) found that Pythium aphanidermatum and P. dissotocum zoospores were kill ed within 0.25 to 0.5 minutes when ORP was above 780 mV in chlorinated water. In the presence of organic and inorganic nitrogen including ammonia, a range of chloramines in Eq. [1 to 3], adapted from EPA (1999). The concentration of both free NH 3 + HOCl NH 2 Cl (monochloramine) + H 2 O [1] NH 2 Cl + HOCl NHCl 2 (dichloramine) + H 2 O [2] NHCl 2 + HOCl NCl 3 (nitrogen trichloride) + H 2 O [3] In municipal water treatment facilities, a dose of one part of inorganic N to every three parts of free chlorine reportedly results in 99% conversion of free chlorine to chloramines after 0.2 sec. at pH 7, or 147 sec. at pH 4 (White, 1992, p. 255). In horticulture irrigation, water soluble fertilizer containing ammonium, nitrate, and/or urea N is often supplied in ir rigation between 100 and 200 mgL 1 total N and then treated with free chlorine concentrations between 1 and 2 mgL 1 Cl. With such a high N to Cl ratio, the majority of free chlorine would therefore be expected to rapidly convert to chlorinated nitrogen f orms. Chlorine also reacts with urea, although the chain of reactions is more complex and slow acting than chlorination of ammonia (Blatchley and Cheng, 2010).

PAGE 70

70 Chloramines are considered weaker sanitizers than free chlorine from hypochlorous acid (White, 1992, p.270) because of the longer contact time that chloramines require to control human pathogens compared with an equal concentration of free chlorine. Control of 99% of Escherichia coli with hypochlorous acid at 1 mgL 1 Cl required a 1 min contact tim e, whereas with combined chlorine forms including NH 2 Cl and NHCl 2 at 1 mgL 1 Cl required more than 100 min (Akin et al., 1982). However, the greater stability of chloramines in the presence of organic compounds compared with hypochlorous acid may increase penetration of chloramines into biofilm, resulting is greater inactivation of biofilm bacteria (LeChevallier et al., 1988). The American Water Works Association (AWWA, 1991) provided guidelines for residual concentration of between 0.5 to 1 mgL 1 chloram ine for disinfection of ground water (White, 1992), such as from wells or irrigation catchment areas. Based on experience with control of human pathogens in water supply, irrigation of edible crops with complexed chlorine is likely to require longer contac t times and/or higher total chlorine application concentrations compared with free chlorine for control of human pathogens. However, data are not available for plant pathogens. There are limited research data on the residual level of free or total chlorine in the presence of nutrient solutions, despite the common practice of dual injection of chlorine and water soluble fertilizer and the potential impact on pathogen control. The research objective was to determine the free and total chlorine and ORP respons es over time when nutrient solutions were blended with sodium hypochlorite.

PAGE 71

71 Materials and Methods Three experiments were run: Experiment (a) Multiple fertilizers 200 mgL 1 N from 7 commercial blended fertilizers containing macro and micronutrients and r eagent grade nutrient solutions were mixed with 2.6 mgL 1 Cl from sodium hypochlorite. In Experiment (b) Multiple N Concentrations, five concentrations of N from 0 to 600 mgL 1 from four reagent grade nutrient solutions were mixed with 2.6 mgL 1 Cl Exp eriment (c) High Applied Chlorine quantified whether 10 or 20 mgL 1 applied Cl was sufficient to provide a residual of 2 mgL 1 Cl after a 2 min contact time in the presence of 200 mgL 1 N from commercial fertilizers and reagent grade N salts. The 11 nut rient solutions for Experiment (a) Multiple Fertilizers are shown in Table 5 1. Nutrient solutions ranged in the contribution of ammonium, nitrate, and urea to total nitrogen, as well as presence of other macro and micronutrients (Fig. 1). Both commercia l blended fertilizers and reagent grade salts were tested, along with a deionized water control. Fertilizers were sourced from two different manufacturers, Greencare Fertilizers (Kankakee, IL.) and Everiss NA, Inc. (Dublin, OH). All commercial fertilizer b lends tested contained some ammonium or urea, including high nitrate fertilizer such as 13N 0.9P 10.8K (Fig. 1). In Experiment (b) Multiple N Concentrations, the 4 reagent grade salts from Table 5 1 were applied at five N concentrations (0, 75, 150, 300 a nd 600 mgL 1 N) in combination with 2.6 mgL 1 Cl. In Experiment (c) High Applied Chlorine, chlorine was applied at 10 or 20 mg L 1 Cl from sodium hypochlorite to the fertilizer treatments in Table 5 1 at 200 mgL 1 N. Experiment (c) was analyzed separa tely from Experiment (a), despite a similar experimental design, because the procedure was run at a different

PAGE 72

72 time, dilution of the samples was necessary when analyzing higher levels of free and total chlorine levels in Experiment (c), in order to provide in range concentrations for measurement, and measurements were made at 2 min only. No sample dilution was required for Experiment (a). Nutrient solutions were prepared in a closed system in 4 L plastic containers in all experiments. Nutrient solutions were adjusted to a pH to 6.0 using KOH or H 2 SO 4 with constant mixing and were maintained at 25C. Chlorine solution was prepared by adding Clorox Regular Bleach (6.15% sodium hypochlorite) (Clorox Company, Oakland, CA) to deionized water and then injecting suf ficient solution into the closed nutrient solution container through a septum to provide a final concentration of 2.6 mgL 1 Cl for Experiment (a) and (b), and 2.6, 10 or 20 mgL 1 Cl in Experiment (c). Free and total chlorine, pH, temperature, and ORP we re measured in the nutrient solution before addition of chlorine, and 2 and 60 min after addition of chlorine. Free and total chlorine concentrations were measured as described by methods 2350 B and 4500 Cl. F. 1d, respectively (APHA, 1995). Colorimetric d etermination of mgL 1 Cl was performed using a Thermo Scientific Orion AQUAfast IV AQ4000 colorimeter, AC4P71 Chlorine Free and Chlorine Total reagent packs (Thermo Fisher, Barrington, IL). Solution pH was measured using an Orion model 61 65 (Thermo Fish er), solid state probe and ORP measured using an Orion 91 79 Triode electrode and both pH and ORP were displayed simultaneously on a Four Star meter (Thermo Fisher, Barrington, IL). In each experiment, the design was a randomized complete block design. On each day (block), one replicate of each of the randomly ordered nutrient and chlorine treatment combinations was run. There were three replicates for Experiment (a) and (b),

PAGE 73

73 and two replicates for Experiment (c). For all experiments, data were analyzed us ing ANOVA with SAS PRO GLM (SAS version 9.1, SAS institute, Cary, NC). Means were Results and Discussion effect of N sourc e on the concentration of free chlorine, ORP and total chlorine (Table 5 There was a two w Control solutions of DI water did not differ in measured variables between 2 min and 60 min (Fig. 5 1 ). At 2 min, free chlorine (2.6 mgL 1 Cl) in the control was higher than free chlorine concentrati on in all nutrient solutions containing fertilizer, which was below the minimum reportable limit (0.15 mgL 1 Cl) of the colorimeter used to measure free chlorine concentrations. Total chlorine was less affected by fertilizer presence than free chlorine, h owever 13N 0.9P 10.1K had lower total chlorine than ammonium sulfate and ammonium nitrate. The 13N 0.9P 10.1K fertilizer contained urea, and urea alone had lower total chlorine than the control and some other solutions at 60 min. However, 15N 2.2P 12.5K al so contained urea and did not result in a drop in total chlorine. In addition, urea did not have lower free chlorine than the control, and urea measurements of free chlorine were higher than total chlorine at 60 min. Therefore, estimation of total and/or f ree chlorine in solutions containing urea was not accurate because free chlorine forms (hypochlorous acid, hypochlorite, and dissolved chlorine) are components of total chlorine, and free chlorine should therefore be equal to or lower than total chlorine. Urea

PAGE 74

74 had a higher ORP than any other solution at 60 min, emphasizing the complexity of reactions of chlorine and urea (Blatchley and Cheng, 2010). The ORP was lower in most commercial fertilizers compared with the control at both 2 and 60 mins, regardless of the proportion of each nitrogen form. Potassium nitrate and urea had free chlorine concentration equal to the control at 2 min and 60 min, whereas with ammonium nitrate and ammonium sulfate the free chlorine concentrations were estimated at zero because at 0.15 mgL 1 Cl the value was below the detectable limit of the meter. The ORP was higher in potassium nitrate than ammonium nitrate and ammonium sulfate, and was equal to the ORP of the control. In Experiment (b) Multiple N Concentrations, there were s ignificant interactions between reagent grade N sources, concentration, and time (Table 5 3). There was no difference in the free chlorine concentration for the control between 2 and 60 min (Fig. 5 2 ). Among the reagents, the higher level of free chlorin e was measured in nutrient solutions with potassium nitrate than urea. The lowest free chlorine levels occurred in ammonium nitrate or ammonium sulfate at 75 to 600 mgL 1 N. Free chlorine level did not decline significantly compared with the control for p otassium nitrate or urea at 75 to 600 mgL 1 N. In Experiment (c) High Applied Chlorine, there were significant main and interaction effects of N source and ap 5 4). Free chlorine concentrations for the control at 10 or 20 mgL 1 applied Cl were significantly different (9.9 and 19.9 mgL 1 Cl, respectively, Fig. 5 3). At both applied chlorine concentrations, ammonium sulfate and ammonium nitrate, and all commercial fertilizer blends resulted in the same free chlorine level, ranging from 0.20 to 1.26 mgL

PAGE 75

75 1 Cl. For urea and potassium nitrate, the free chlorine concentration was not significantly different from the cont rol at both applied Cl concentrations. The decrease in free chlorine levels observed with reagent grade salts and commercial fertilizer blends in these experiments has practical implications for combining fertilization and water sanitation in horticulture Where ammonium was present in the solution, the majority of chlorine is likely to be complexed, even at low concentrations (for example, 37.5 mgL 1 NH 4 N from ammonium nitrate applied at 75 mgL 1 N, Fig 5 2). Most commercial fertilizers contain some a mmonium, including ammonium sulfate, ammonium phosphate, ammonium nitrate, or as a component in calcium nitrate fertilizer. Nitrate did not decrease free chlorine concentration, and urea reacted slowly, indicating these nitrogen forms would be more compati ble with free chlorine than ammonium. Another management strategy would be to use holding tanks to inject chorine into unfertilized water in order to provide adequate contact time prior to injecting fertilizer. A further option to reduce chlorine demand of fertilizer would be to apply granular or controlled release fertilizer directly to the growing substrate rather than water soluble fertilizer. In the presence of ammonium, high applied concentration of chlorine would be required to achieve residual free chlorine for control of plant pathogens. Even at 20 mgL 1 Cl applied free chlorine in Experiment (c), in the presence of ammonium nitrogen the level of free chlorine rapidly decreased to 1.26 or less mgL 1 Cl. Given that as high as 2 mgL 1 of free chlor ine may be required for control of Pythium and Phytophthora zoospores (Cayanan et al., 2009; Hong et al., 2003; Hong and Richardson, 2004; Lang et al., 2008), it may be difficult to achieve that level of free chlorine without more than 10

PAGE 76

76 times the applied chlorine concentration. We have observed that some growers are indeed applying several times higher chlorine concentrations than 2 mgL 1 Cl when trying to control dosage based on a free chlorine target while also injecting water soluble fertilizer that c ontains ammonium. The implications in terms of phytotoxicity risk and pathogen control of that practice are unknown, and chlorination cost would increase. An alternative sanitizing technology other than chlorine could be used that is more compatible with water soluble fertilizer, for example chlorine dioxide. Copes et al. (2004) found the efficacy of chlorine dioxide in nutrient solutions for control of Fusarium oxysporum was more affected by pH and micronutrients than nitrogen, and chlorine dioxide at 2 m gL 1 was able to provide 50% decrease of F. oxysporum in nutrient solutions that contained 100 mgL 1 N and 3 mgL 1 of micronutrients (Cu, Fe, Mn and Zn). Our results emphasize the need for monitoring of chlorine levels and ORP in horticulture applicatio ns, where nitrogen and other contaminants create a chlorine demand that is likely to vary over time, thereby affecting residual free chlorine concentration. Research conducted by Lang et al. (2008) demonstrated that free chlorine concentration and ORP were important for disinfestation of Pythium aphanidermatum Measurements of pH, free chlorine concentration, and ORP are standard operating procedures to monitor to ensure treatment efficacy in water and wastewater treatment facilities (EPA, 1999), and these procedures should be adapted to horticulture operations.

PAGE 77

77 The low level of free chlorine found in these experiments when combined with fertilizers that contained ammonium or urea indicate that hypochlorous acid is likely to be rapidly converted to complexed chlorine forms when water soluble fertilizers are used in combination with chlorine. Several alternative strategies may maximize efficacy of water treatment systems in horticulture to avoid this problem, including use of solid fertilizers, injecting chlor ine in a holding tank before fertilizer injection, or using nitrate based water soluble fertilizer. Monitoring is essential with chlorine and any other water treatment technology, with important parameters including solution pH, active ingredient levels of the sanitizing agent, ORP (if oxidizers are used), the density of colony forming units of indicator organisms such as aerobic bacteria plate counts and the presence or absence of particular pathogenic organisms of concern Research is needed to determine the disinfestation strength of complexed chlorine for horticulture applications, where plant pathogens are the primary target. Additional research is needed to determine the potential interactions of chlorine with micronutrients and chelates, and to furthe r investigate the phytotoxicity effects of free, complexed and total chlorine, as well as the potential for toxic byproducts produced from the reactions of chlorine and fertilizer.

PAGE 78

78 T able 5 1. Nutrient sources used in the study (reported by manufactu rer). Fertilizer (N P K) Source z CaCO 3 equiv. y Acid Base (A/B) % NH 4 % NO 3 % Urea NH 4 NO 3 Urea P 2 O 5 K 2 O P K Ca Mg S B Cu Fe Mn Mo Zn N13 0.9P 10.1K E 167 B 0.3 11.9 0.8 0.0 0.9 0.1 2.0 13.0 0.9 10.8 6.0 3.00 0.00 0.02 0.02 0.06 0.03 0.01 0.03 N15 2.2P 12 .5K E 65 B 1.1 11.8 2.1 0.1 0.8 0.1 5.0 15.0 2.2 12.5 5.0 2.00 0.00 0.02 0.02 0.07 0.04 0.01 0.04 N14 1.8P 11.6K G 100 B 2.0 12.0 0.0 0.1 0.9 0.0 4.0 14.0 1.8 11.6 5.0 2.00 0.00 0.02 0.02 0.07 0.03 0.01 0.03 N16 2.1P 13.7K G 7 B 4.1 12.4 0.0 0.2 0.8 0.0 4.8 16.5 2.1 13.7 3.30 1.60 0.00 0.02 0.02 0.08 0.04 0.01 0.04 N15 2.2P 16.6K E 24 A 4.3 10.7 0.0 0.3 0.7 0.0 5.0 25.0 2.2 20.8 0.0 2.50 1.00 0.01 0.02 0.07 0.04 0.07 0.04 N20 0.9P 16.6K E 145 A 7.2 12.8 0.0 0.4 0.6 0.0 2.0 20.0 0.9 16.6 0.0 1.10 0.74 0. 02 0.02 0.10 0.05 0.01 0.05 N19 5.6P 19.4K G 340 A 9.7 9.7 0.0 0.5 0.5 0.0 5.6 19.4 2.5 16.1 0.0 0.00 0.00 0.02 0.02 0.10 0.05 0.01 0.05 Urea Reagent 0.0 0.0 100 0.0 0.0 1.0 Potassium Nitrate Reagent 0.0 100 0.0 0.0 1.0 0.0 Ammonium Sulfate Reagent 21.0 0.0 0.0 1.0 0.0 0.0 Ammonium Nitrate Reagent 50.0 50.0 0.0 0.5 0.5 0.0 z G = Greencare Fertilizers, E = Everris Co. y reported kg/metric ton

PAGE 79

79 Table 5 2. ANOVA summary for Experiment (a) Mult iple N sources, where free chlorine was applied to 11 nutrient solutions at 200 mg.L 1 N and a DI water control (N Source) at 2.6 mgL 1 Cl with 2 min and 60 min contact times (Time). Free Cl (mgL 1 ) ORP (mV) Total Cl (mgL 1 ) N Source *** *** *** Ti me ns ns N Source Time ns ** ns Rep ns ns *** Table 5 3. ANOVA summary for Experiment (a) Multiple N Concentrations where free chlorine was applied at 2.6 mgL 1 Cl to 4 reagent nutrient solutions (N Source) at 6 concentrations from 0 to 600 mgL 1 N ([N]) with 2 or 60 min contact time (Time). Free Cl (mgL 1 ) ORP (mV) Total Cl (mgL 1 ) N Source *** *** *** [N] *** *** *** Time ns *** *** N Source Time *** *** N Source [N] *** *** *** Time [N] ns ns *** N Source Time[N] ns *** *** Replicate ns ns ns Table 5 4. ANOVA summary for Experiment (c) High Applied Chlorine where free chlorine was applied at two concentrations of 10 or 20 mgL 1 Cl ([Cl]) to 11 nutrient solutions at 200 mgL 1 N and a DI control (N Source) and free chlori ne was measured after 2 min contact time. Free Cl (mgL 1 ) N Source *** [Cl] *** N Source [Cl] *** Replicate ns

PAGE 80

80

PAGE 81

81 Figure 5 1. Nitrogen source and its effect on free and total chlorine and ORP in Experiment (a) Multiple Fertilizers. Chart (A) sh ows the percent contribution of ammonium, nitrate, or urea to total nitrogen in each nutrient source. Charts (B) to (D) show the effect of 200 mg L 1 N from commercial fertilizer blends and reagent grade nutrient solutions on (B) free chlorine, (C) total c hlorine, and (D) ORP after injecting sodium hypochlorite at 2.6 mg L 1 Cl with 2 min HSD at the p=0.05 level for the interaction of nutrient source and contact time.

PAGE 82

82 Figure 5 2. Effect of nitrogen concentration from four reagent grade N sources on free chlorine after injecting sodium hypochlorite at 2.6 mgL 1 Cl in Experiment (b) Multiple N Concentrations, measured after (A) 2 min and (B) 60 min. Letters represent mean compa each measurement time.

PAGE 83

83 Figure 5 3. Effect of sodium hypochlorite at 10 mgL 1 Cl and 20 mgL 1 Cl measured after 2 min in 11 nutrient solutions at 200 mgL 1 N and a DI Control, in Experiment (c) H igh Applied Chlorine. Letters represent mean comparisons

PAGE 84

84 CHAPTER 6 CONCLUSIONS Survey of Water Quality The survey of irrigation water in commercial nurseries and greenhouses (Chapter 2) emphasized the importance of monitoring physical, chemical, and microbial water quality, particularly when recirculating irrigation water. The highest water quality occurred in Source water, with the exception of pH which could be treated with acid injection. Sampling points after th e Source tended to have decreased water quality in terms of lower UV transmission, and higher organic carbon, aerobic bacteria and COD. The greatest potential risk of clogged emitters in irrigation systems resulted from aerobic bacteria levels above 10,000 cfumL 1 In eight Locations where Pre Treatment bacterial density was above 10,000 cfumL 1 treatment systems only lowered bacteria density below this threshold in half the locations. Improvements to irrigation systems at the survey locations would be r equired to reduce bacteria counts, including increased filtration to remove suspended particles and overall sanitizing demand, followed by more effective chemical treatment. Regular monitoring of sanitizing agent levels and biological water quality would i ndicate water quality changes over time and within a horticultural irrigation system. The density of aerobic bacteria sampled before and after water treatment provided a useful measure of the relative efficacy of a treatment technology to control microbes in greenhouse and nursery irrigation. More detailed analysis of Pre and Post Treatment for each Location should be conducted to determine if the treatment systems effectively controlled aerobic bacteria and fungi over time. Installation of sampling ports before and after each treatment points should be a standard design feature to improve the ability to

PAGE 85

85 monitor irrigation quality. Before this survey, none of the Locations had measured physical water quality and Pre and Post Treatment water quality variabl es such as aerobic bacteria counts, which are important monitoring variables to ensure correct operation of a treatment system. Petrifilm Technology Experiments reported in Chapters 3 and 4 confirmed that Petrifilm can be used to quantify aerobic bacteria in irrigation water. Substrate and measurement day affected the estimated bacteria cfumL 1 in irrigation water (Chapter 3). Counts using Petrifilm AC medium would need multiplication by a factor (2.92 in this one study) to allow comparison with R2A, and t o compare against an established threshold of 10,000 cfumL 1 for acceptable biofilm risk in micro irrigation systems. In Chapter 4, the estimated cfumL 1 of Xanthomonas campestris determined on Petrifilm AC and PDA was lower than cfumL 1 counted on R2A and determined by initial hemocytometer counts. Differences in cfumL 1 between substrates may have resulted from different growth response of aerobic bacteria species to the nutrient formulations and physical environment. We conclude that monitoring irrig ation water quality with Petrifilm AC can provide an on farm test to estimate the microbial load, potential for biofilm formation and relative efficacy of water treatment systems along points in an irrigation line. No growth of the oomycete P cactorum wa s detected on Petrifilm YM. Irrigation testing using Petrifilm YM should therefore be interpreted with care because, not all fungi, or fungal like organisms such as P cactorum, may not be represented in colony counts on Petrifilm YM. The Petrifilm YM may not have been suitable for detection of P. cactorum because of either the phosphatase enzyme indicator or antibiotics in the Petrifilm YM. In contrast, PARP H, the selective medium for oomycetes, resulted in a

PAGE 86

86 colony count that was comparable to the viable zoospore cfumL 1 estimated with the hemocytometer. A limitation of the Petrifilm is that selective substrates are not currently available for specific plant pathogens. Therefore, the technique is a non specific measurement of aerobic bacteria that may be suitable for quantifying microbial load, estimating a potential for biofilm formation and relative efficacy of water treatment systems along points in an irrigation line, rather than to identify presence or absence of a specific plant pathogen. Chlorine T reatment A low level of free chlorine was found when sodium hypochlorite was mixed with fertilizers containing ammonium or urea, indicating that hypochlorous acid was converted to complexed chlorine forms such as chloramines. The practical application is that hypochlorous acid would not be the dominant sanitizing form when chlorination and water soluble fertilizer are used together in irrigation. Several alternative strategies may maximize efficacy of water treatment systems in horticulture to avoid this p roblem, including use of soil applied fertilizers, injecting chlorine in a holding tank before fertilizer injection, or using nitrate based water soluble fertilizer. Monitoring is essential with chlorine and any other water treatment technology, with impo rtant parameters including solution pH, active ingredient levels of the sanitizing agent, ORP (if oxidizers are used), the density of colony forming units of indicator organisms such as aerobic bacteria plate counts and the presence or absence of particula r pathogenic organisms of concern Research is needed to determine the disinfestation strength of complexed chlorine for horticulture applications, where plant pathogens are the primary target. Additional research is needed to determine the

PAGE 87

87 potential inter actions of chlorine with micronutrients and chelates, and to further investigate the phytotoxicity effects of free, complexed and total chlorine, as well as the potential for harmful byproducts produced from the reactions of chlorine and other irrigation w ater variables.

PAGE 88

88 LIST OF REFERENCES Agrios, G 1997 Plant pathology 4th ed ., p 188 192 Academic Press, San Diego, Calif Akin, E W ., J C Hoff, and E C Lippy 1982 Waterborne outbreak control which disinfectant? Environ Health Perspective 46:7 12 American Public Health Association 1995 Standard meth ods for the evaluation of water and wastewater 18th ed Amer Public Health Assn ., Denver, CO American Water Works Association 1991 Guidance manual for compliance with the filtration and disinf ection requirements for public works systems using surface water sources U S Environmental Protection Agency, Office of Groundwater and Drinking Water, Washington, D C Ar go, W R ., J A Biernbaum, and D D Warncke 1997 Geographical characterization of greenhouse irrigation water HortTechnology 7:49 55 Ass ociation of Agricultural Chemists (AOAC) 1999 Official methods of analysis of AOAC 16 th ed AOAC International, Boston, Mass B ergendahl, J A and L Stevens 2005 Oxidation reduction potent ial as a measure of disinfection effectiveness for chlorination of wastewater Environmental Progress 24: 214 222 Bu cks, D A ., F S Nakayama, and R G Gilbert 1979 Trickle irrigation water quality and preventative maintenance Agric Water Mgt 2:149 162 Carm ichael, I ., I S Harper, M J Coventry, P W Taylor, J Wan, and M W Hickey 1999 Bacterial colonization and biofilm development on minimally processed vegetables J of Applied Microbiology (supplement) 85:45S 51S C ayanan, D F ., Y Zheng, P Zhang T Graham, M Dixon, C Chong, and J Llewellyn 2008 Sensitivity of Five Container grown Nursery Species to Chlorine in Overhead Irrigation Water HortScience, 43: 1882 1887 C ayanan, D F ., P Zhang, W Liu, M Dixon, and Y Zheng 2009 Effic acy of chlorine in controlling five common plant pathogens HortScience, 44: 157 163 C ayanan, D F ., M Dixon, Y Zheng, and J Llewellyn 2009 Response of container grown nursery plants to chlorine used to disinfest irrigation water HortScience, 44: 164 167 Ch ang, E E ., P C Chiang, and T F Lin 1998 Development of surrogate organic contamination variables for source water quality standards in Taiwan Chemosphere 37(4):593 606

PAGE 89

89 C opes, W E ., G A Chastaganer, and R L Hummel 2004 Activity of chl orine dioxide in a solution of ions and pH against Thielaviopsis basicola and Fusarium oxysporum Plant Disease 88:188 194 Feben, D., and M. J. Taras. 1955. Chlorine demand constants. J. of American Water Works Assoc. 43:922. Gauthie r, V., C. Rosin, L. M athieu, J.M. Portal, J. C. Block, P. Chaix, and D. Catel. 1999. Organic matter as loose deposits in a drinking water distribution system. Water Research 33:1014 1026. Gilbert, R.G., F.S. Nakayama, D.A. Bucks, O.F. French, and K. C. Adamson. 1980. Trickle ir rigation emitter clogging and other flow problems. Agric. Water Mgt. 3:159 178. Heibling, D.E., and J. M. VanBriessen. 2008. Continuous monitoring of free chlorine concentrations in response to controlled microbial intrusions in a laboratory scale distribu tion system. Water Research 42:3162 3172. Hendrix Jr ., F F ., Campbell, W A ., 1973 Pythiums as plant pathogens Annu Rev Phytopathol 11, 77 98 Hong, C ., Richardson, P A ., Kong, P ., 2002 Comparison of membrane filters as a tool for isolating pythia ceous species from irrigation water Phytopathology 92, 610 616 Hong, C X ., Moorman, G W ., 2005 Plant pathogens in irrigation water: challenges and opportunities Crit Rev Plant Sci 24, 189 208 Kerr, C.J., K.S. Osborn, G.D. Robinson, and P. S. Handl ey. 1999. The relationship between pipe material and biofilm formation in a laboratory model system. J. of Applied Microbiology (supplement) 85:29S 38S. Maier, R. M., I. L. Pepper, and C. P. Gerba. 2009. Environmental microbiology. 2nd ed., p. 485 499. Acad emic Press, Boston, Mass. Materon, L A ., 2003 Survival of Escherichia coli O157:H7 applied to cantaloupes and the effectiveness of chlorinated water and lactic acid as disinfectants World J Microbiol Biotechnol 19, 867 773 Meador, D P ., Fisher, P R ., Teplitski, M ., 2013 The use of petrifilm to quantify aerobic bacteria in irrigation Proc Fla State Hort Society Delray Bch FL June 5 7 Meador, D P ., Fisher, P F ., Guy, C L ., Harmon, P ., Peres, N ., Teplitski, M ., 2012 A survey of the physical, chemical and biological water quality in greenhouse irrigation HortTechnology, (In Press)

PAGE 90

90 Newman, S. E. 2004. Disinfecting irrigation water for disease management. In: 20th Annual Conference on Pest Management on Ornamentals, p. 1 10. San Jose, 20 22 Fe b. 2004. San Jose Society of American Florists. Obreza, T., M. Clark, B. Boman, T. Borisova, M. Cohen, M. Dukes, T. Frazer, E. Hanlon, K. Havens, C. Martinez, K. Migliaccio, S. Shukla, and A. Wright. 2010. A guide to environmental protection agencies nume ric nutrient water quality criteria for Florida. Univ. of Florida. Inst. of Food and Agric. Sci. Res. Bul. SL 316. Pettygrove, G. S., and T. Asano. 1985. Irrigation with reclaimed municipal wastewater a guidance manual. Lewis Publishers, Chelsea, Mich. Pi rovanni, M., A. Piagentini, D. Guemes, and S. Arkwright. 2004. Reduction of chlorine concentration and microbial load during washing disinfection of shredded lettuce. Intl. J. of Food Science and Tech. 39:341 347. Ravina, I. E. Paz, Z. Sofer, A. Marcu, A. Schischa, G. Sagi, Z. Yechialy, and Y. Lev. 1997. Control of clogging in drip irrigation with stored treated municipal sewage effluent. Agric. Water Mgt. 33:127 137. Riordan, D C ., Sapers, G M ., Hankinson, T R ., Magee, M ., Mattrazzo, A M ., Annous, B A ., 2001 A study of U S orchards to identify potential sources of Escherichia coli O157, H7 J Food Prot 64, 1320 1327 Roberts, P D ., Urs, R R ., French Monar, R D ., Hoffine, M S ., Seijo, T E ., McGovern, R J ., 2005 Survival and recovery of Phytophthora capsici and oomycetes in tailwater and soil from vegetable fields in Florida Ann Appl Biol 146, 351 359 Rogers, D.H., F. R. Lamm, and M. Alam. 2003. Subsurface drip irrigation systems water quality assessment guidelines. Irrigation manage ment Series, Kansas State University Agricultural Experiment Station and Cooperative Extension Service. Res. Bul. MF 2575. Runia, W. T. 1994. Disinfection of recirculation water from closed cultivation systems with ozone. Acta Hort. 361:388 396. Simone, G W ., Momol, M T ., 1990 Vegetable Disease Recognition and Management Florida Greenhouse Vegetable Production Handbook, Vol 3 University of Florida Cooperative Extension Technical Bulletin, EDIS Doc HS797 Singleton, L L ., Mihail, J D ., Rush, C M ., 199 2 Methods for Research on Soilborne Phytopathogenic Fungi APS Press, St Paul, Minn ., pp 265 266

PAGE 91

91 Sutton, J. C., C. R. Sopher, T. N. Owen Going, W. Liu, B. Grodzinski, J. C. Hall, and R. L. Benchimol. 2006. Etiology and epidemiology of Pythium root rot in hydroponic crops: current knowledge and perspectives. Summa Phytopathologica 32:307 321. Tanwar, P., N. Tapas, U. Pallavi and P. Manekar. 2008. Correlating on line monitoring variables, pH, dissolved oxygen and oxidative reduction potential with nutrie nt removal in an intermittent cyclic process bioreactor system. Bioresource Technology 99:7630 7635. U. S. Environmental Protection Agency (USEPA). 1999. Guidance manual, alternative disinfectants and oxidants, disinfectant use in water. Municipal Support Division. Office of Wastewater Management. Office of Water, Washington, D. C. Cir. U. S. Environmental Protection Agency (USEPA). 2004. Guidelines for water reuse. Municipal Support Division. Office of Wastewater Management. Office of Water, Washington, D. C. Report USEPA/625/R 04/108:167. Van Os, E A ., 2009 Comparison of some chemical and non chemical treatments to disinfect a recirculating nutrient solution Acta Hort 843, 229 234 Watkins, J ., Jian, X ., 1997 Cultural methods of detection for micro organisms: recent advances and successes Freshwater Bio Assoc Conference, Queen Annes Gate, London Dec 12 14 Whipker, B E ., J M Dole, T J Gavins, and J L Gibson 2003 Water quality, p 9 18 In: D Hamrick (eds ) Ball redbook crop production 1 7th ed Ball Publishing, Batavia, Ill

PAGE 92

92 BIOGRAPHICAL SKETCH Dustin Paul Meador was born in Indianapolis, I.N. to Paul A. Meador and Sandra L. Riddle. At a very young age his mother moved around the world and had the opportunity to live in Canada, Jamaica and several states across the U.S. At the age of five Dustin and his mother moved to Miami, F.L. and began a new family when his mother was remarried. Dustin has two sisters: Christina and Dana; and three brothers: Danny, Johnny, Marshall and one brother, Preston ( deceased ) During his time as a student at the U. of Florida he d eveloped s olutions for emerging problems in the horticulture industry and to improve environmental stewardship. As a n applied scientist he can provide analytical, interdiscip linary, research techniques to inform management decisions that improve greenhouse and nursery production and sustainability. He provides e ducation to all ages and skill levels, including industry professionals, recreational gardeners and garden clubs, elementary schools, undergraduate and graduate students Horticulture is part of his professional and private life and he incorporates science to educate and motivate people to care for their environment. Some intellectual contributions and technical support and prod uction systems have resulted in applied technologies that will improve the sustainability of the horticulture industry, such as a woody liner p roduction system that improves root form and quality of trees and tree products His hope is to expand the knowle dge base of horticulture and help industry provide quality crops with sustainable production techniques for irrigation treatment and nutrient management