|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 B IOLOGY OF Phakopsora pachyrhizi THE CAUSAL AGENT OF SOYBEAN RUST IN FLORIDA By HEATHER MARIE YOUNG A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2012
2 2012 Heather Marie Young
3 To all those who h ave encouraged and supported me
4 ACKNOWLEDGMENTS I would like to thank Dr. Jim Marois, my major professor, for his support, guidanc e, and help throughout my Doctoral degree program. I am thankful to Drs. David Wright, Ariena van Bruggen, and Phil Harmon, members of my advisory committee, for their useful counsel, critique and instruction. I am also very grateful to Drs. Dario Narve s, Daniele Liberti, support and help with greenhouse and field experiments and laboratory analyses. I am also thankful for Maynard Douglas, Chuck Dickson, and Brian Kidd for their assista nce with establishing and maintaining soybean plots. I would also like to extend my admiration to all the faculty and staff of North Florida Research and Education Center UF IFAS in Quincy, Florida and to the Plant Pathology Department at the University of Florida in Gainesville, Florida for their camaraderie, inspiration, and support, which were very important in improving my scientific skills and knowledge of academics. I want to thank Dr. David Wright and North Central Soybean Research Program for fin ancial and operational support without their assistance this research would not have been possible. For those not mentioned, who had contributed to my success; I offer my sincere gratitude. Finally, I thank my boyfriend, Brett Kelly, and my parents, De nnis and Linda Young, for their love, encouragement, and support.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 8 ABSTRACT ................................ ................................ ................................ ................... 10 CHAPTER 1 INTRODUCTION AND RESEARCH OBJECTIVES ................................ ................ 13 Soybean ................................ ................................ ................................ .................. 13 Physiology ................................ ................................ ................................ ........ 13 Uses and Cultivation History ................................ ................................ ............ 13 U.S. Cultivation ................................ ................................ ................................ 14 Soybean Rust ................................ ................................ ................................ ......... 15 Geographic Distribution ................................ ................................ .................... 15 Economic Impact and Yield Loss ................................ ................................ ..... 16 Pathogen Taxonomy and Morphology ................................ .............................. 18 Host Range ................................ ................................ ................................ ...... 20 Disease Symptoms and Signs ................................ ................................ .......... 20 Infection, Reproduction, and Dispersal ................................ ............................. 21 Soybean Rust in the U.S. ................................ ................................ ................. 23 Soybean Rust Epidemiology and Risk Models ................................ ................. 23 Management ................................ ................................ ................................ .... 27 Resistance to Soybean Rust in Soybean ................................ ......................... 32 Kudzu ................................ ................................ ................................ ...................... 34 Physiology ................................ ................................ ................................ ........ 34 Geographic Distribution ................................ ................................ .................... 35 History in U.S. ................................ ................................ ................................ .. 35 Reproduction ................................ ................................ ................................ .... 36 Management ................................ ................................ ................................ .... 37 Species Variation ................................ ................................ ............................. 38 Reactions to Soybean Rust ................................ ................................ .............. 38 Summary ................................ ................................ ................................ ................ 39 Research Objectives ................................ ................................ ............................... 40 2 EPIDEMIOLOGY OF SOYBEAN RUST IN SOYBEAN SENTINEL PLOTS IN FLORIDA ................................ ................................ ................................ ................ 41 Introduction ................................ ................................ ................................ ............. 41 Materials and Methods ................................ ................................ ............................ 42 Sentinel Plot Data ................................ ................................ ............................. 42
6 Data Collection ................................ ................................ ................................ 42 Data Analysis ................................ ................................ ................................ ... 46 Results ................................ ................................ ................................ .................... 47 Discussion ................................ ................................ ................................ .............. 49 3 EFFECT OF SOLAR RADIATION ON SEVERITY OF SOYBEAN RUST .............. 65 Introdu ction ................................ ................................ ................................ ............. 65 Material and Methods ................................ ................................ ............................. 67 Canopy Manipulation ................................ ................................ ........................ 67 Field In oculations and Evaluations ................................ ................................ ... 69 Detached Leaf Assays ................................ ................................ ...................... 71 Data Analysis ................................ ................................ ................................ ... 72 Results ................................ ................................ ................................ .................... 73 Discussion ................................ ................................ ................................ .............. 75 4 VARIATION IN THE NUC LEOTIDE BINDING SITE (NBS) REGIONS OF KUDZU ( PUERARIA MONTANA) : POSSIBLE S OURCE OF RESISTANCE TO PHAKOPSORA PACHYRHIZ I, THE CAUSAL AGENT OF SOBYEAN RUST ....... 90 Introduction ................................ ................................ ................................ ............. 90 Materials and Methods ................................ ................................ ............................ 93 Kudzu Populations for Initial Screening of NBS Region ................................ ... 93 P. pachyrhizi Inoculum ................................ ................................ ..................... 94 DNA extraction, polymerase chain reaction (PCR), and partial sequencing of NBS region ................................ ................................ ................................ 94 Sampling of Kudzu Populations across Florida ................................ ................ 95 Restriction Fragment Length Polymorphism (RFLP) Analysis .......................... 96 Detached Leaf Assay ................................ ................................ ....................... 98 Results ................................ ................................ ................................ .................... 99 Discussion ................................ ................................ ................................ ............ 102 5 EPIDEMIOLOGY OF SOYB EAN RUST ON SOYBEAN AND SUSCEPTIBLE AND RESISTANT KUDZU AND DECISION MODEL F OR FUNGICIDE APPLICATIO NS FOR SOYBEAN RUST IN FLORIDA ................................ ......... 116 Introduction ................................ ................................ ................................ ........... 116 Material and Methods ................................ ................................ ........................... 118 Estimation of Potential Inoculum ................................ ................................ .... 118 Decision Application Model ................................ ................................ ............ 123 Results ................................ ................................ ................................ .................. 125 Discussion ................................ ................................ ................................ ............ 128 LIST OF REFERENCES ................................ ................................ ............................. 146 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 166
7 LI ST OF TABLES Table page 2 1 Data summary of north Florida soybean sentinel plots (SSP) and the soybean rust epidemic data within those plots from 2005 through 2008 ............ 54 2 2 A description of variables ................................ ................................ ................... 55 2 3 Variables correlated with first disease detection date ................................ ......... 56 2 4 Variables correlated with maximum incidence ................................ .................... 57 2 5 Variables correlated with maximum severity ................................ ....................... 58 2 6 The num ber of soybean sentinel plots (SSP) in which a maturity group was first detected with soybean rust (SBR) ................................ ............................... 59 3 1 Energy values (W m 2 ) recorded in 0, 30, 40, and 60% shaded soybean plots .. 80 3 2 Energy values (W m 2 ) recorded in non shaded control soybean plots ............... 81 4 1 Primers used to amplify the nucleotide binding site ( NBS) domain in kudzu populations ................................ ................................ ................................ ....... 107 4 2 Distribution of kudzu populations in Florida. ................................ ..................... 108 4 3 Average percent similarity among clusters of kudzu populations in regard to the nucleotide binding site (NBS) region(s) ................................ ...................... 109 4 4 Analysis of molecular variance (AMOVA) of restriction fragment length polymorphism (RFLP) data ................................ ................................ ............... 110 5 1 Variables used to estimate the potential maximum amount of soybean rust inoculum escaping from different host canopies in Florida ............................... 134 5 2 Variables, at optimal conditions, used to calculate the number of new uredinia per day over time ................................ ................................ ................ 135 5 3 Monthly temperature for 2005 to 2008 across 5 different locations in no rth Florida ................................ ................................ ................................ .............. 1 36 5 4 Rates of disease increase ................................ ................................ ................ 137 5 5 Application decision model results and first observations of soybean rust (SBR) on kudzu and soybean each year ................................ .......................... 138 5 6 Rates of disease increase ................................ ................................ ................ 139
8 LIST OF FIGURES Figure page 2 1 Florida map of soybean sentinel plots and Florida Automated Weather Network stations ................................ ................................ ................................ 60 2 2 Weekly cumulative rain and disease incidence and severity over time .............. 61 2 3 Soybean growth stage and maturity group (MG) data over time ........................ 62 2 4 Precipitation threshold for average maxi mum disease severity of soybean rust in soybean sentinel plots (SSP) recorded from 2005 through 2008 ............. 64 3 1 Average plant height ................................ ................................ ........................... 82 3 2 Average leaf area indices ................................ ................................ ................... 83 3 3 Ambient temperature and relative humidity recorded in shade treatments ......... 84 3 4 Comparative ultraviolet visible solar spectrum recorded under shade treatments ................................ ................................ ................................ ........... 85 3 5 Severity of soybean rust (SBR) from field plots over time ................................ ... 86 3 6 Final soybean rust (SBR) disease severity among shade treatments ................ 87 3 7 Average soybean rust (SBR) severity on detached soybean leaves .................. 88 3 8 Average epicuticular wax (mg/cm 2 ) extracted from soybean leaves ................... 89 4 1 Distribution of kudzu populations in Florida ................................ ...................... 111 4 2 Neighbor joinin g phylogenetic tree ................................ ................................ ... 112 4 3 Restriction fragment length polymorphisms (RFLP) of kudzu populations susceptible, resistant, and immune to Phakopsor a pachyrhizi ......................... 113 4 4 Average number of uredinia and urediniospores produced per inoculation site on detached leaflets from different hosts ................................ .......................... 114 4 5 Unweighted pair group method using arithmetic mean (UPGMA) dendrogram 115 5 1 The number of uredinia developing on different hosts in Florida over time at optimal co ndition for disease development. ................................ ...................... 140 5 2 Soybean rust infection progress curves ................................ ............................ 141 5 3 Enumeration of visual urediniospores from passi ve wind vane spore traps ..... 142
9 5 4 Disease severity of soybean rust (SBR) within fungicide treatment plots ......... 143 5 5 Average soybean yields (kg/H) from fungicide application treatment plots ....... 144 5 6 Estimated number of sporulating uredinia over time on soybean and susceptible kudzu ................................ ................................ ............................. 145
10 Abstract of Dissertation Presented to the Gra duate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy B IOLOGY OF Phakopsora pachyrhizi THE CAUSAL AGENT OF SOYBEAN RUST IN FLORIDA By Heather Marie Young May 2012 Chair: James J. Marois Major: Plant Pathology Soybean rust (SBR) caused by the biotrophic fungus Phakopsora pachyrhizi is a potentially destructive disease to soybean ( Glycine max ) production in the United States. Winter temperatures and viable hosts limit the fungus to overwintering in the southeastern states, on the alternative host kudzu ( Pueraria montana ). The initial inoculum from kudzu and subsequent development of the disease within southern commercial soybean fields provide the inoculum for po tential epidemics in the major soybean production areas in the U.S. An improved understanding of SBR epidemics and movement of initial inoculum from kudzu in the southeastern states is important in developing fungicide application decision models for mana ging epidemics and for improving the accuracy of current risk assessment models for long distance dispersal of SBR spores. This study examined the effects of environmental factors (i.e. temperature, relative humidity, leaf wetness, rainfall, and solar rad iation) and soybean growth stage on SBR development and severity in soybean, examined genetic variation in kudzu populations with respect to possible resistant genes and quantified resistance to SBR in kudzu and soybean through detached leaf assays. A di sease and decision
11 model for fungicide application for SBR was developed and validated by the data obtained. Disease and soybean data, from 2005 through 2008 collected in north Florida, indicated that the majority of SBR infections occur at reproductive gr owth stage R3 (beginning pod) or later, even though earlier maturing varieties are at reproductive stages earlier in the year. Precipitation was found to be the principal environmental factor affecting disease severity, where average maximum severity was 2 0.8 and 9.8% if precipitation was more and less than 300 mm, respectively, over a 1 month period after initial infection. Early or late infection in soybean fields was attributed in part to weather conditions occurring 1 month prior to disease detection a nd potential build up of inoculum over time. Data from shading treatments in the field and detached leaf assays suggest that solar radiation affects not only disease development in soybean canopies by affecting temperature and relative humidity and reduci ng spore germination, but also by increasing epicuticular wax on soybean leaves creating a thicker barrier that P. pachyrhizi has to pe netrate through for infection. K udzu populations in Florida were screened for genetic resistance to SBR using primers des igned in conserved nucleotide binding site (NBS) domain of the known soybean resistant gene Rpp4 and restriction fragment length polymorphisms (RFLP) High genetic variation among kudzu populations and low variation within populations were found with resp ect to NBS region(s). No patterns of variation were observed in regard to geographic location of kudzu populations or infection history of SBR from 2005 through 2011. Results indicate that kudzu is a good source of resistance to SBR not only due to the g enetic variation among kudzu populations but also due to an overall
12 reduction in severity of disease development on both susceptible and resistant kudzu as compared to soybean. Detached leaf assays quantified different levels of resistance in susceptible, resistant, and immune kudzu with respect to susceptible and resistant soybean. From the results of the detached leaf assays the number of uredinia on susceptible kudzu was only 10% and on resistant kudzu was only 1% of the number of uredinia produced on s usceptible soybean. Similarly, the number of spores produced per inoculation area, 20 days after infection on susceptible and resistant kudzu was only 23 and 2%, respectively, of that produced on susceptible soybean. Sporulation was delayed on resistant k udzu as compared to susceptible soybean and kud zu. Sporulation, infection efficiency, and latent period were utilized from the detached leaf assay results in a spatial and temporal disease model estimating the number of infections occurring over time and a mount of spores escaping into the atmosphere from soybean and kudzu areas in Florida. Utilizing the knowledge on environmental variables and potential inoculum from soybean and kudzu a decision model for fungicide application was developed and validated o ver the growing seasons of 2009 through 2011. Results from this dissertation identified environmental factors that contribute to SBR epidemics in the United States and have been used to better estimate potential spore loads available for long distance tra nsport from southeastern states to the major soybean productions areas in the northern and mid western states to improve risk assessment for the major soybean production areas.
13 CHAPTER 1 INTRODUCTION AND RES EARCH OBJECTIVES Soybean Physiology Soybean [ Gly cine max (L.) Merrill] is an annual legume which typically grows 75 to 125 cm in height with 14 26 nodes ( 172 ). Leaves are simple compound trifoliates in an alternating pattern ( 172 ). Soybeans can either grow indeterminately, where the terminal bud continues to be active during most of the growing season, or determinately, where the terminal bud ceases growth at the onset of reproductive development ( 63 215 ). Soybean flowers have a typical papilionaceous flower with a high percentage of self pollination ( 215 ). A soybean plant can produce 60 80 pods, each containing 3 pea sized beans of 130 220 mg ( 172 184 ). Seed production is favored by warm mean temperatures of 20 C to 30 C. The seed content consists of approximately 38 56% protein and 8 27% oil, depending on variety ( 135 ) The root systems consist of a primary taproot and numerous lateral roots arranged in four longitudinal rows along the primary tap root; where most root dry weight (80 90%) occurs in the top 15 cm of the soil ( 172 188 ). Soybean above ground vegetative structures consist of an erect, branched stem, attaining a height of 75 125 cm and possess 14 26 nodes ( 172 ). Uses and Cultivation History In 2009 soybeans represented 53 seed production and within the U.S. soybeans provide 69% of the edible consumption of fats and oils ( 184 ). The various uses of soybean include not only human and animal feed, but also many industr ial uses from adhesives to biofuels. Linguistic, geographical, and historical evidence suggests domestication of soybean emerged around the eleventh century
14 B.C. in the eastern half of north China, most likely during the Shang dynasty (ca. 1700 1100 B.C.) ( 74 ). The movement of soybean within China and the peninsular Korea was most likely associated with the development, consolidation of territories, and degeneration of Chinese dynasties ( 74 ). Soybea n seed was used primarily for human and livestock feed and for medicinal purposes. By the 15 th and 16 th century, soybeans were introduced in Japan, Indonesia, the Philippines, Vietnam, Thailand, Malaysia, Burma, Nepal and north India. soybeans were cultivated in Europe and in 1765 plantings of soybeans were recorded in the Colony of Georgia to manufacture soy sauce and vermicelli ( 74 ). USDA published its first bulletin about the use of soybeans as a forage crop in 1899, and by 1904 soybeans were evaluated for their use as an oil and protein source. By 1909, commercial soybean production had developed in many states from the eastern seaboard into the central states, with an estimated 800 ha of soybeans being planted that year ( 134 ). The top five U.S. states with the largest production areas of soybean in 2010 were Iowa, Illinois, Minnesota, Nebraska, and Indiana ( 134 ). U.S. Cultivation In the U.S. soy beans are generally planted in early May to mid June and harvested in mid September through mid November, but it is possible for soybeans to be planted as early as mid April and as late as early August in some areas ( 13 147 217 ). Soybean cultivars are classified into 13 maturity groups (MG) based on the time of flowering and maturity due to geographical adaptation ( 228 ). Specifically, a soyb ean cultivar is assigned a numeric number based on the days from planting to maturity at a defined latitude (or day length) and a specific planting date under optimum environmental conditions ( 228 ). The number of different var ieties a farmer may plant in
15 one season varies based on the area being planted and other variables, and may range from 6 to 10 different varieties ( 201 ). In 1996 genetically modified soybeans were commercially released and by 2009 an estimated 91% of the soybean hectares planted in the U.S. were genetically modified soybean ( 134 ). Genetically modified soybean contains a gene derived from Agrobacterium sp. strain CP4, encoding a glyphosate tolerant enzyme, called CP4 enzyme 5 enolpyruvylshikimate 3 phosphate (EPSP) synthase, which results in glyphosate tolerant soybeans ( 53 ). Glyphosate is the main active ingredient in the herbicide RoundUp, which can be used on the gen etically Soybeans are grown in more than 30 states, making it the second largest crop in cash sales and the number one export crop in the U.S. ( 184 ). In 2009, the 31.4 milli on ha of soybeans planted in the U.S. was the largest soybean crop ever planted in the U.S and the 80.7 million metric tons harvested was also the largest in U.S. history, harvest of soybean with 57.0 million metric tons in 2009 ( 184 ). Soybeans accounted for 30% of the U.S. crop area planted in 2009, second only to corn. It accounted for 38% of on value of $32.1 billion in 2009 ( 134 ). Soybean Rust Geographic Distribution Soybean rust (SBR) caused by the fungal biotroph Phakopsora pachyrhizi Syd. & P. Syd. is present in all the major soybean producing areas of the wo rld. The first report of the disease was from Japan in 1902 ( 20 118 ). By 1934, the pathogen had spread to several other Asian countries as well as Australia ( 21 ); and by 1951 SBR was reported in India ( 171 ). In 1996, the first confirmed report of the pathogen on the
16 African continent was published from Kenya, Rwanda, and Uganda ( 118 ). Since 1996, the pathogen h as continued to spread across the African Continent with reports from Zambia and Zimbabwe in 1998, Nigeria in 1999, Mozambique in 2000, and South Africa in 2001 ( 3 118 152 ). The disease was found on soybeans on the Hawaii islands of Oahu, Kauai, and Hawaii in 1994, but dissemination to the continental U.S. from this location was never reported ( 86 ). In February of 2001 the first detection of P pachyrhizi in the New World was reported on soybeans in Paraguay. By 2002, the disease was widespread throughout Paraguay, limited areas of Brazil bordering Paraguay and northern Argentina ( 128 158 224 ). The pathogen spread north of the equator in Brazil and possibly into Venezuela, Guyana, and Suriname ( 79 224 ). In November 2004, the pathogen was found on soy beans in Baton Rouge, LA ( 167 ) and on soybeans and kudzu in Florida ( 65 ). It was hypothesized that Hurricane Ivan carried P. pachyrhizi spores from South America into the U.S. ( 79 ), thus making SBR present in every major soybea n producing region in the world Since its introduction to the continental U.S. the disease has spread from the Southeast to as far west as Nebraska, to the north in the Canadian province of Ontari o and to the south in Mexico ( 102 157 212 ). In the 2009 growing season, SBR was detected in 576 counties in 17 states ( 102 ). Economic Impac t and Yield Loss SBR is one of the most detrimental diseases to soybeans worldwide causing devastating yield losses and increased production costs ( 66 102 118 ). The dis ease can reduce seed weight of soybeans by 40 80%, shoot weight by 20%, the overall number of pods per plant, and the photosynthetic area ( 222 ). Experiments conducted in Brazil
17 in the 2005 2006 and 2006 2007 growing season att ributed SBR induced yield loss to premature leaf loss, reduction in canopy green leaf area due to lesions, reduction in dry matter accumulation, and reduction in harvest index due to reduced seed set and seed mass ( 96 ). In Chin a, yield losses from SBR commonly range between 10 and 30%, and have been observed to be over 50% in severe disease years ( 61 ). In southern Japan, estimated yield losses have ranged up to 80% for individual fields and total c rop losses have been reported as high as 30% in Taiwan ( 222 ). Yield losses of 68% in susceptible varieties and 22% in tolerant cultivars have been reported in field studies in Korea ( 137 ). Researc h plots in Australia that had not received fungicide applications had yield losses of 60 to 70% ( 139 ). In Africa, yield losses of 10 to 80% were observed in South Africa and from 60 to 80% in Zimbabwe ( 28 ). In recent years, yield losses of 100% have been reported in continuous mono culture soybean production in Africa ( 28 ). Paraguay and Brazil have observed yield losses up to 60 and 63%, respectively ( 224 ). Not only does yield loss due to SBR have an economic impact on soybean production, but also the cost of fungicides and their applications in response to SBR increase production cost. In the 2003 2004 soybean production season in Brazil, economic losses due to the disease, including costs of control, were estimated at more than $2 billion US$ ( 72 180 ). SBR poses a serious threat to the U.S. soybean production, as it does in other ma jor soybean production areas in the world. In 2004, USDA estimated that in the first year the pathogen is established in the country the value of net economic losses would
18 range from $640 million to $1.3 billion, and annual losses in ensuing years could a verage anywhere between $240 million and $2.0 billion ( 104 ). Further estimates stated that yield losses could exceed 10% in the U.S. with up to 50% in the Mississippi delta and southeastern states ( 151 220 222 ). Since its introduction in 2004, SBR has spread throughout the U.S. more gradually than was expected ( 45 169 ). The pathogen h as not been detected in the major soybean producing regions of the Midwest until late in the growing seasons and the relatively late season arrival in the Southeast has resulted in only minor yield losses in commercial soybean fields. While the impact of yield loss in U.S. commercial fields has been minimal thus far, experimental plot losses have ranged from 19 to 35% in Alabama, Georgia, and Florida ( 37 130 ). Even though SBR has not yet caused maj or yield losses in the U.S., it has still had an economic impact through fungicide application on soybeans. According to data from the National Agriculture Statistics Service (NASS), very little fungicide application occurred on soybeans prior to 2006, bu t in 2006 over 362,000 pounds of fungicides were applied to soybeans across 7 different states, with that number continuing to increase in the following years ( 134 ). Since its introduction to the continental U.S. in 2004, rese archers, producers, and industry representatives have been gaining valuable information from field trials about the biology and control of SBR in the U.S., despite its spread and severity continuing to be variable each year. Pathogen Taxonomy and Morpholog y P. pachyrhizi (synonym: Malupa sojae Uredo sojae Phakopsora erythrinae and Uredo erythrinae ( 140 ) ) is in the common rust fungi order Uredinales ( 73 118 213 ). Fungi in Uredinales are generally biotrophs, and can have up to 5 different spore stages
19 on two plant hosts which are taxonomically unrelated. The 5 spore stages include urediniospores, teliospores (both on principal host), basidios pores, spermagonia (pycnia), and aeciospores (latter on alternate host) ( 213 ). In nature, only urediniospores and teliospores have been observed for P. pachyrhizi ( 66 118 ) Basidiospores have been produced in laboratory conditions; however no alternate host has been identified on which the fungus completes its life cycle For this fungus, basidiospores are not considered a critical spore type ( 162 ). Hence, P. pachyrhizi is an autoecious, microcyclic rust fungus. The urediniospores are produced in anamorphic fruiting structures called uredinia, which can be amphigenous (growth on both sides of leaf), but are mostly hypophyllous (growth on aba xial surface of leaf) ( 20 66 ). Uredinia are tan to reddish brown in color, range from 50 200 m in diameter, and spatially sprinkled to clustered on yellowish lesions. A uredinium is surrounded by incurved paraphyses arising from a cellular basal peridium, which forms a volcano like structure, with urediniospores being borne on sporophores inside the structure. Paraphyses are cylindric to clavate, hyaline to yellowish brown in color, range in size from 25 50 x 6 14 m, and are slightly thickened at the apices. Urediniospores are released through an opening (ostiole) at the top of the volcano like structure. Urediniospores are sessile, obovoid to broadly ellipsoid, with a minute and dense echi to yellowish brown in color and range in size from 18 38 m long by 13 29 m wide with walls that are approximately 1 m thick ( 140 ). Under appropriate environmental condit ions the production of telia may be observed, often seen mixed on the abaxial leaf surface with uredinia. Telia are
20 chestnut brown to dark brown in color, crustose and range in size from 150 250 m. Teliospores are single celled, with walls approximate ly 1 m thick, oblate spheroid, and irregularly arranged in 2 to 7 spore layers within the telium. The spores range in size from 15 26 x 6 12 m and are yellowish brown to colorless ( 21 140 223 ). Host Range P. pachyrhizi has been reported to infect, under controlled conditions, 158 species in 54 genera of the subfamily Papilionoideae ( 57 65 105 140 160 170 179 ). Unlike other rusts, Phakopsora spp can directly penetrate host epidermal cells, which in P. pachyrhizi, mi ght be, in part, the reason it can infect a wide range of legume species ( 91 ). Due to the early misidentification of the two rust species ( P. pachyrhizi and P. meiobmiae ) and the multiple responses of host species to different rust pathotypes, a complete host range has not been clearly defined ( 118 ). Current reported hosts in the U.S. are Glycine max (soybeans) ( 167 ), Pueraria lobata (kudzu) ( 65 ), Desmodium tortuosum (Florida beggarweed) ( 170 ), Erythrina herbacea (Coral bean) ( 57 ), Phaseolus coccineus P. lunatus and P. vulgaris ( 105 ). In greenhouse evalua tions in 2008, 65 species were newly identified as hosts to P. pachyrhizi of which 62 occur in Alabama, Florida, Louisiana, Mississippi, and Texas ( 179 ). The large host range of P. pachyrhizi reflects the diversity and comple xity of its virulence patterns, as well as helps contribute as kudzu in the southern U.S. Disease Symptoms and Signs P. pachyrhizi may infect soybean at any growth stage ( 47 ), but symptoms are most often observed at late vegetative to early reproductive stages ( 45 91 111 ). The most common visual symptoms of SBR are tan to dark brown t o reddish brown angular
21 polygonal lesions (0.5 5 mm 2 ) ( 20 66 118 ). The color of the lesion is dependent upon its age and host/pathogen genotype interactions. Typical ly, symptoms are first observed on leaves in the lower canopy and mainly on leaves, but can be found on petioles, pods, and stems of soybean plants ( 176 ). Within lesions, multiple erumpent and globose uredinial may develop, on both abaxial and adaxial surfaces of soybean leaves ( 20 66 ). Telia are rare to observe, they are dark brown to black and are subepidermally distributed among uredinia. Production of multiple ured inia with a circular ostiole located at the top of each uredinia where spores emerge is a distinct sign of P. pachyrhizi Severe or early infection can result in premature yellowing and defoliation, causing yield loss ( 197 ). W hile P. pachyrhizi produces characteristic rust symptoms it may be confused with other diseases, such as brown spot ( Septoria glycines Hemmi), downy mildew ( Peronospora manshurica Naumov), frogeye leafspot ( Cercospora sojina Hara), bacterial blight ( Pseudo monas syringae van Hall), and bacterial pustules ( Xanthomonas axonopodis pv. Glycines Nakano), especially at disease onset and very early stages of development ( 66 198 ). Infection, Reproduction, and Dispersal Urediniospores are predominately wind dispersed and serve as the only known infective spore type ( 118 ). Infection process begins with germination of urediniospores across host tissue, which can occur within 2 hours after spore deposition, at 20 C when moisture is readily available ( 20 ). The germ tube can grow 5 to 400 m long, with appressoria frequently forming when germ tube is less than 100 m in length. Once the germ tube contacts an anticlinal wall or extends over epidermal cells, an appressorium
22 forms. Rarely will the appressorium develop over stomata ( 15 90 103 111 ). The appressorium will directly penetrate into epidermal cells, reaching the intercellular space of the mesophyll by way of the appressorial penetration peg, which works under turgor pressure to puncture the host epidermis ( 118 ). If an appressorium develops over a stomatum, it will penetrate one of the guard cells instead of entering through the stomatal opening. Direct penetration of host epidermis is a unique trait of Phakopsora rust species as compared to other rusts tha t require a natural openi ng or wound to enter the host. Between 24 and 48 hours after penetration, growth and development of specialized feeding hyphae, called haustoria, occur in the intercellular space of the mesophyll ( 90 91 ). Uredinia have been observed as early as 5 days after infection with urediniospores being produced after 9 days ( 90 108 ). Sporulation can continue for 3 to 6 weeks f or a single uredinium and up to 15 weeks in a single lesion while primary and secondary uredinia develop ( 90 118 ). On average, each uredinium can produce more than 12,000 urediniospres ( 21 45 90 108 ). Temperatures between 15 and 28 C ( 45 66 108 ) and a minimum of 6 hours of continuous leaf wetness ( 20 114 ) are required for infection to occur. As temperatures decrease below 18 C, the length of the moisture period needed for infection increases ( 109 ) and extended periods of leaf wetness in the field can increase the rate of spread of the pathogen in the upper canopy and overall disease severity ( 133 ). Temperatures near 20 C with 10 to 12 hours of free moisture promote maximum infection, which can result in high disease severity in les s than 20 days after infection.
23 P. pachyrhizi spore dispersal can be illustrated in 5 stages: 1) urediniospore production on the host, 2) urediniospore escape from the canopy, 3) turbulent transport and dilution of urediniospore clouds through the atmosphere, 4) survival in transport, and 5) urediniospore dep osition onto a host population ( 8 62 ). Soybean Rust in the U.S. In the U.S., the pathogen generally occurs on soybeans late in the growing season, when the majority of the soybeans are in the R3 or R4 reproductive growth stages or later ( 45 169 ). Based on observations recorded through the U.S. Department SBR surveillance and monitoring program, SBR overwinters in the Gulf Coast regions of the U.S. from Florida to Texas, and given suitable environmental conditions it spre ads from a south to north direction in the U.S. ( 45 77 80 145 169 ). Being a biotroph, P. pachyrhizi requires l iving host tissue, as well as temperatures greater than 4 C to survive ( 20 192 ), although urediniospores can survive short freeze/thaw cycles ( 83 ). Overwintering hosts may include volunteer soybeans or any of the numerous alternative hosts P. pachyrhizi can infect, but presently kudzu ( Pueraria montana [Lour.] Merr. v ar. lobata [Willd.] ) is the only known host that could provide large amounts of living tissue for overwi ntering in the southeastern U.S. in areas where it does not freeze ( 17 26 77 83 102 140 160 178 192 ). Soybean R ust Epidemiology and Risk Models SBR models can be separated into two major groups: simulation and empirical models. Simulation models are process driven, based on concepts derived from a pathosystem and help improve understanding of structure and behavior of such a biological system ( 193 ). Empirical models are statistically driven, based on statistical relations hip(s) of explanatory variables with experimental data. The reliability of model
24 predictions depends heavily on the quality of the dataset used for model development ( 93 ). SBR simulation models can be further divided into epi demiological and aerobiological models. Epidemiological models aim to mimic biological processes in the disease cycle to estimate disease development considering local inoculum availability. T hree such models are SOYRUST, specific to SBR ( 219 220 ), a general disease model for rusts ( 150 ), and CLIMEX ( 149 ) a climate computer model. SOYRUST is a computer model that simulates daily increase of disease severity by rate variables that estimate leaf area development and disease components, including infection, latent period, and uredinia senescence. Where infection rate is influenced by leaf wetness duration and temperature, latent period and ure dinia senescence are a function of only temperature. SOYRUST was further refined by linking it to the soybean growth model, SOYGRO, to simulate disease effects on soybean yield and assess potential crop losses in some locations of the U.S. by generating d isease progress curves and final yield in the presence or absence of SBR ( 219 220 ). A general disease model using infection efficiency, latent period, infectious period, the number of spores produce d per lesion, and proportion of spores landing on a potential infection site was adjusted to predict epidemic development for SBR and four other rust diseases ( 150 ). The general disease model was adjusted by the temperature ef fect on infection efficiency and latent period at the initial stages of an epidemic, in order to assess establishment potential by the estimation of daily increase of infection units and rate of disease increase. Results indicated that longer periods of t ime would
25 be required for SBR to increase from trace to visual detection levels in the field for spec ific locations within the U.S. The computer model, CLIMEX ( 182 ), uses climatic suitability for biological processes and has be en adjusted to determine stress free zones for year round survival of P. pachyrhizi A stress free index for a single location was calculated using the stress indices of cold, warm, and moisture stresses as defined in the literature. Results identified F lorida and other southeastern U.S. states boarding the Gulf of Mexico as potential regions where year round inoculum survival would be most likely to occur. It also identified northern states in the U.S. as regions where SBR epidemics would depend on prim ary inoculum originating from long distances ( 149 ). Aerobiological models aim to predict the transport and dissemination of airborne inoculum over long distances. SBR aerobiological models have been used to assess entry poten tial (HYSPLIT, 1 43 88 ), entry and establishment potential (HYSPLIT, 132 ), and entry, establishment, and epidemic potential (SRAPS, 79 ). Th e atmospheric transport model NOAA ARL HYSPLIT_4 (HYSPLIT Hybrid Single Particle Lagrangian Integrated Trajectory) ( 136 ) simulates ingle trajectories of air parcels, dispersal, concentration and deposition of particles, origi nating from a source geographical location and time of year. The parameters for strength of inoculum sources, spore production, survival and deposition that were mechanistically and empirically estimated based on current knowledge on the system and physic al principals of SBR were used in the HYSPLIT model ( 142 ). The climate prediction data from the Experimental Climate Prediction Center (ECPC) of the Scripps Institution of Oceanography was fed into the Pennsylvania State Unive rsity/National Center for Atmospheric Research mesoscale
26 model (MM5) that estimates meteorological variables to input into the NOAA ARL HYSPLIT_4 model to predict spore dissemination months ahead. This SBR aerobiological model was used to hindcast likely dispersal of P. pachyrhizi spores from Africa into South America in 2002 and to forecast dispersal from central to northern soybean production regions in South America in 2004. Furthermore, spore dispersal maps generated by the model using a source of P. pachyrhizi in Colombia showed likely spread of spores to the southern U.S. which was confirmed by the findings of the disease in November 2004, in several locations in the gulf coastal states ( 142 ). The HYSPLIT model was furth er adjusted to predict SBR risks in the U.S. during the 2005 growing season by combining biological and meteorological elements into a qualitative evaluation of the risk of disease development associated with one or a group of sources. This was done by in cluding a description of the factors pertinent to the disease status, a general outlook that assessed the risk of disease development, and a map showing the spore laden wind flowing away from an active source or source region using multiple, centrally loca ted trajectories ( 132 ). In such the model allowed the prediction of entry and establishment potential by estimating qualitative risks. SBR empirical models can be separated into 3 kinds according to the epidemic component est imated : suitability or critical periods for infection, disease progress, and maximum or final disease severity ( 40 ). Empirical models that estimate suitability or critical periods for infection were developed and applied in Br azil ( 29 154 ). Reis et al ( 154 ) used a non linear regression of interaction of leaf wetness duration and mean temperature during periods of wetness and infection ( 114 ) to develop daily values of probability for infection. Canteria et al. ( 29 ) also used a non linear regression function
27 of the number of hours of leaf wetness and mean temperature ( 109 ) to calculate relative intensity of infection. Risk maps of infection efficiency were generated daily and results were summarized for periods of a week or a month to map regions where higher climatic favorability was predicted ( 29 ) Management There have been multiple approaches for controlling SBR including cultural practices, fungicide application, biological control, and breeding for resistance. Planting early or using an early maturing variety are cultural p ractices that may allow the crop to mature before conditions become favorable for disease development ( 20 118 ). In general, most cropping systems and cultural practices such as narrow row or wide r ow planting, double crop soybean, and tillage system does not directly affect SBR incidence or severity; however, disease pressure may vary under some cropping systems and cultural practices due to extended time that the crop is exposed in the field and re gulated crop free periods. In Brazil, mandatory 90 day soybean free periods, in which no soybeans are to be grown and all volunteer soybean plants are to be destroyed, were put into regulation to reduce inoculum accumulation and overall disease pressure ( 183 ). Other cultural practices for reducing SBR incidence and severity include mid day and nighttime irrigation to avoid extended periods of leaf wetness ( 28 ).Furthermore, the destruction of altern ative hosts will decrease the amount of potential initial inoculum; however, the eradication of kudzu and other wild alternative hosts of SBR are not practical. Fungicide application has been most successful in controlling SBR epidemics. Historically, soy bean diseases have not been managed by fungicide applications, but there are certain conditions that may necessitate the use of foliar fungicides. Other
28 soybean diseases that have been managed with fungicide application include frogeye leaf spot ( Cercospo ra sojina ) ( 1 ), Cercospora leaf blight ( C. kikuchii ) ( 168 ), Sclerotinia stem rot ( Sclerotinia sclerotiorum ) ( 129 ), brown spot ( Septoria glycines ) ( 106 ), and Phomopsis seed infection ( 216 ). Factors contributing to the effectiveness of fungicide applications for SBR control include the severity of the disease/disease pressure, the duration for which the chemical rem ains active on the host tissue, and the length of the reproductive stages of the plant ( 120 ). It has been observed when 20 to 30% of soybean leaves in the mid canopy are infected fungicide applications are no longer effective in controlling the disease ( 45 ). Hence, early detection and prediction of SBR is critical for fungicide applications. While one to two sprays from R1 through R6 is recommended for control of SBR in the U.S. ( 45 ), the timing and effectiveness of any given fungicide treatment is often dependent on when rust is first detected and the intensity of its development ( 130 ). In Africa, it is recommended to apply fungicides as a preventative or within a couple of weeks after first observations of the disease are confirmed since yields tend to decrease as the time between first detection and first application increase ( 101 ). Fungicide applications made prior to first detection of SBR tend to result in greater yields than those made after first detection. In general, occurrence of rust at any time during the first half of seed fill could justify protective fungicide applications ( 14 130 ). A forecasting system based on monitoring source regions for SBR and weather variables is established through the Integrated Pest Management Pest Information Platform for Extension and Educat ion (ipm PIPE, sbr. ipmpipe.org). This system provides growers with information useful for determining timing and frequency of fungicide applications
29 and has and will continue to reduce the number of applications of fungicides applied to control SBR ( 80 130 ). Another important factor for controlling disease with fungicides is the spray coverage on the host plant. Methods used to apply fungicides depend on the area that needs to be treated and the equipment available t o make the application ( 120 ). Aerial applications of fungicides in Brazil provide adequate control of the disease over large soybean acreage ( 121 ). Alternatively, maximum disease control was achie ved in Zimbabwe by penetrating the fungicide spray into the soybean canopy, by use of air assist and high pressure lateral discharge equipment ( 100 ). In the U.S., effective application of fungicide have been attained using a c onventional sprayer and flat fan nozzle with high spray pressures to attain fine to medium droplet sizes ( 45 130 141 169 ). The 2 chemical c l asses of fungicides predominantly used to manage SBR are strobilurins and triazoles Strobilurins are within the Quinone Outside Inhibitors (QoI) fungicide group blocking energy transfer at the site of quinol oxidation (Qo site) in the cytochrome bc 1 co mplex inhibiting fungal mitochondrial re spiration resulting in the fungus death. T his is a single site of inhibition; hence risk of the development of resistance is high. Resistance to QoI fungicides in plant pathogens has been detected and 3 amino acid s ubstitutions in the cytochrome b gene were identified as mediating the resistance ( 87 173 ). The QoI fungicides have the ability to inhibit both spore germination and host penetration, but have litt le or no effect once the fungus has penetrated or colonized host tissue; hence they are most effective if applied prior to infection ( 45 ). Strobilurin fungicides have locally systemic translaminar movement,
30 where some active ingredients can leak through the leaf and bind to the cuticle on other side, this aids in compensating for incomplete spray coverage ( 206 ). Most strobilurins have a residual period of approximately 14 to 21 days. Triazoles are within the demethylation inhibitors (DMI) fungicide group. They target C14 demthylase in sterol biosynthesis in fungi, affecting membrane structure and function, resulting in abnormal growth and eventual death of the fungus ( 97 ). They have the ability to inhibit or stop development of infections that are already established at low levels, but have little or no effect on spore germination. Hence, DMIs are the fungicide of choice if SBR is established at low levels in a field. DMIs are locally systemic with a residual period of approximately 14 days. While DMIs also have a single site of inhibition, risk of development of resistance is medium due to resistance being mediated by the accumulation of several independent mutation s ( 95 ). Trials in India ( 162 ), Africa ( 100 ), and South America ( 121 122 ) have identified the triazole compoun ds of flusilazole, difenoconazole, triadimenol, tebuconazole, and tetraconazole, as well as strobilu rins and strobilurin mixes of az o x ystrobin, pycraclostrobin, pyrcalostrobin and boscalid, and trifloxystrobin and propiconazole as valuable fungicides for m anaging SBR Trials conducted in the U.S. in 2005, indicate that Folicur (tebuconazole), Headline (pyraclostrobin), Headline SBR (pyraclostrobin and tebuconazole), Laredo (myclobutanil), and Stratego (triflozystrobin and propiconazole) all provided a cceptable control of SBR ( 169 ). However, Folicur is no longer labeled for soybean and Headline SBR is no longer on the market. Current U.S. and international research on the effectiveness of fungicides against SBR continues to support triazole and strobilurin chemistries in reducing
31 severity and yield loss in treated versus control soybean plots ( 45 123 130 ), although in the 2008 2009 grow ing season in South America, some P. pachyrhizi populations showed increasing tolerance to certain fungicides ( 59 ). Limited options are available at present to manage SBR in organic soybean production, although recent research has reported a biological control agent that was effective in delaying disease development and reducing disease severity. Using some organic approved copper fungicides against SBR have provided greater yields compared to untreated soybean in research tria ls and silicon amendments that delay disease onset and reduce final area under the disease progress curve have also been identified in research trials ( 45 99 211 ). Furt hermore, application of saccharin to hydroponically grown soybeans can induce the systemic acquired resistance response of soybean to P. pachyrhizi ( 185 ). Simplicillium lanosoniveum was documented, through scanning electron mi croscopy, to wrap around urediniospores of P. pachyrhizi and colonize uredinia ( 210 ). Although S. lanosoniveum requires SBR to be present in some form on soybean leaves in order to colonize, it was effective in detached leaf a ssays and field trials in reducing disease severity of SBR ( 211 ). Additional research is still needed to evaluate the effects of sunlight on survival and colonization of leaf surface by S. lanosoniveum and the development of f ormulations with extended shelf life ( 211 ). Further research also is needed to investigate silicon application timings and combinations of silicon amendments with other fungicides or organic approved products, to manage SBR in organic soybean production ( 99 ). Breeding for resistance is an important management strategy against any plant pathogen and P. pachyrhizi is no exception. Planting resistant cultivars of soybean still
32 remains the most economi cal way to manage SBR although due to the pathogens ability to move long distances by windborne urediniospores and evolve quickly, race shifts in regional P. pachyrhizi populations may occur rapidly ( 67 155 ). Such an example was observed in South America when resistance genes, Rpp1 and Rpp3 in overcome within 2 years ( 175 ). Resistance t o SBR in soybean is discussed in more detail in the following paragraphs. While using cultivar mixtures can reduce disease severity due to the epidemiological effect of host diversity, as Caldwell & Myer ( 27 ) observed mixtures of soybean varieties relative to pure stands restricted the spread of a pathogenic form of cyst nematode. There is no documented research on soybean cultivar mi xtures to manage SBR Resistance to Soybean Rust in Soybean Three lesion types have been obser ved on soybean, i) tan lesions (TAN) which have a shorter latent period, many uredinia, and abundant sporulation, ii) reddish brown lesions (RB) which have a longer latent period, fewer uredinia and reduced sporulation, and iii) no visible uredinia; repre senting i) susceptible, ii) resistant, and iii) immune reactions ( 16 20 21 ). The formation of the visible reddish brown lesions is suggestive of a hypersensitive like r esponse. Five genes mediating resistance to SBR have been Resistance to P. pachyrhizi ( Rpp ): Rpp1 ( 112 ) Rpp2 ( 21 ) Rpp3 ( 22 69 ) Rpp4 ( 68 ) and Rpp5 ( 54 ). Rpp1 Rpp4 were reported in the 1980s ( 21 22 68 69 112 ), while Rpp5 was discovered in 2008 ( 54 ). Rpp1 is the only Rpp gene that provides immunity to certain isolates of P. pachyrhizi where no visible sympto ms occur. The other Rpp genes provide resistance responses characterized by limited fungal growth and sporulation through the formation of RB
33 lesions when challenged with incompatible fungal isolates ( 21 ). None of the known R pp genes provide resistance against all isolates of P. pachyrhizi ( 67 118 ). Rpp1 and Rpp4 have been mapped to two different loci on chromosome 18 ( 75 175 ), Rpp2 to chromosome 16 ( 175 ), Rpp3 to chromosome 6 ( 76 ), and Rpp5 to chromosome 3 ( 54 ). Other resistant alleles have been mapped to the vicinities of the previously mentioned loci, including Rpp? (Hyuuga) to Rpp3 locus ( 127 ), and Rpp1 b to the Rpp1 locus ( 37 ). With Rpp5 resistance was dominant in plant introduction (PI) 200487 and PI 200526, incomp letely dominant in PI 471904, and recessive in PI 200456 ( 54 ). Furthermore, a recessive resistance allele at or near the Rpp2 locus in PI 224270 was reported also ( 54 ). Mapping of Rpp genes offers breeders the opportunity to pyramid two or more Rpp genes to obtain broader and/or more durable SBR resistance that may delay and/or reduce SBR growth and sporulation ( 148 ). Of the five Rpp genes identified to date, extensive studies have been conducted on Rpp2 and Rpp4 In microarray analyses of Rpp2 it was observed that expression of basal defense pathway genes increased during the first 12 hours after inoculation, returned to mock levels before a second phase of differenti al gene expression occurred in both susceptible and resistant ( Rpp2 ) plants. The second phase of differential expression was stronger and detected 1 to 2 days earlier in the resistant ( Rpp2 ) plants than in the susceptible plants ( 204 ). Rpp2 through Rpp5 display similar resistant phenotypes (RB lesions) suggesting that Rpp3 through Rpp5 mediate responses similar to Rpp2 and are likely governed by disease resistance genes that mediate gene for gene recognition ( 117 ).
34 In 2009, the Rpp4 gene was identified as a member of the coiled coiled (CC) nucleotide binding site (NBS) Leucine rich repeat (LRR) family of disease resistance genes based on significant sequence similarity to other CC NBS LRR genes, in particular the lettuce Resistant Gene Candidate2 (RGC2) ( 117 ). The NBS LRR gene family accounts for the largest number of known disease resistance genes, and is one of the largest gene families in plant genomes. NBS LRR a re identified by their unique sequence motifs. In plants, LRR motifs have been found in a large number of proteins involved in both disease resistance and development and provides a potential binding surface for protein protein interactions ( 89 98 ). The probable function of NBS regions is to interact with ATP or GTP and act as a signaling molecule ( 196 ). In general, sources of resistance to SBR in soybean are relativel y rare possibly due to the cost of maintaining resistance genes when no pathogen is present overriding the benefits of resistance during pathogen attack ( 117 ). Also, the loss of resistance genes by selection may explain their rarity in soybean. The multitude of alternative leguminous hosts may be potential sources with an abundance of resistance mechanisms to SBR Kudzu Physiology Kudzu ( Pueraria spp. ) has been identified as the only known host that could provide large amounts of living tissue for overwintering of SBR in the southeastern U.S. ( 83 ). With economic loss caused by SBR in the U.S. depending largely on whether or not P. pachyrhizi can survive winters in the absence of soybean, kudzu is o f great importance in understanding and predicting SBR epidemics in the U.S. In addition to being important as an initial inoculum source of urediniospores, kudzu could also be a source of selection pressure on the pathogen and/or a source of resistance t o the
35 pathogen. Kudzu is a perennial semi woody, climbing leguminous vine, belonging to the tribe Phaseoleae Benth., subtribe Glycininae Benth. ( 203 ). There are approximately 17 species in the genus Pueraria recognized globall y ( 203 ) and different scientific names are often used in literature to refer to kudzu in the U.S., including Pueraria montana (Lour.) Merr., Pueraria lobata (Willd.) Ohwi. Pueraria thomsoni (Benth) Maesen, or Pueraria montana var. lobata (Willd.) Maesen and Almeida ( 209 ). Lobed leaflets and the size of wing and keel petals have been used to differentiate P. lobata from P. montana and P. thomsoni but all these morphological characteristics can be q uite variable; furthermore, possible hybrids exist between these related Pueraria species ( 203 ). Based on investigation by Ward ( 209 ), the widespread taxon, if treated at varietal rank, must bear the name Pueraria montana (Lour.) Merr. v ar. lobata (Willd.) Maesen & Almeida. Geographic Distribution Kudzu is considered native to Indo China, Korea, Japan, Malaysia, Oceania, and the Indian subcontinent ( 189 203 ). Successful introductions of kudzu have occurred in South America, Switzerland, Queensland, New South Wales, Australia, and United States. It is only considered a serious pest in the southeastern U.S. ( 23 ). It was first introduced into the U.S. in 1876 at the Centennial Exposition as an ornamental ( 189 ). Estimates of kudzu infestation in the southeastern U.S. vary from 2 million acres (~ 800,000 hectares) to 7 million acres (~ 3, 000,000) ( 34 48 ). Its closest relative in the continental U.S. is soybean ( Glycine max ) ( 23 ). History in U.S. In 1907, kudzu was promoted as a drought resistant, high n itrogen forage crop exhibition at Jamestown, Virginia. One farmer in Chipley, Florida was so enthralled with
36 the growth potential of kudzu that he grew 35 acres of it to sell as a fodder crop and sold rooted cuttings through the mail ( 23 ). By the 1920 30s, kudzu was propagated and promoted by the Soil Conservation Service as a means of holding soil on eroding gullies of the deforested southern landscape, predominantly in the southern states of Alabama, Georgia, and Mississippi ( 23 ). More than 1.2 million acres (485 thousand hectares) were planted under the Soil Erosion Service, which paid farmers, a onetime payment of $8.00 per acre (~$20 per hectare) planted in kudzu. Kudzu seedling and nurseries produced and distributed more than 73 million seedlings between 1936 and 1941 ( 23 ). Kudzu is now estimated to cover between 2 and 7 million acres (0.81 and 2.8 million hectares) of the U .S. primarily in the southeast ( 23 ). Present spread of kudzu is slow through local movement of infested soil. Eventually, farmers found drawbacks to kudzu including difficulty baling the kudzu hay and the tendency for it to grow so rapidly and extensively that it c overed all other vegetation. In the 1950s, kudzu was recognized as a weed and removed from the list of species acceptable for use under the Agricultural Conservation Program. Finally, in 1998 Congress placed kudzu on the federal list of noxious weeds. E conomic impact of kudzu is estimated at $48 acre 1 year 1 ($118 ha 1 year 1 ) for infested forest land which has lost productivity and control costs by power companies alone at $1.5 million per year ( 23 ). Reproduction Kudzu prod uces purple flowers with a sweet aroma in relative abundance. Its corolla is papilionaceous and 14 20 mm long. Kudzu can reproduce sexually through insect pollination and asexually through rhizome spread. It has been estimated that seed production by ku dzu in the U.S. varies from 0 to 1,800 seeds per m 2 soil surface,
37 with higher values occurring where vines are climbing on structures ( 195 ). However, kudzu seeds have extremely low viability demonstrated to be the result of ar thropod damage, mostly due to feeding by native Hemiptera ( 195 ). Kudzu seedlings develop woody root crown, with multiple runners and extensive tuberous roots, which contain carbohydrate reserves that enable the plant to surviv e repeated mow ing and herbicide application. Management While no cultural or chemical controls exist for kudzu, some biological control agents native to the U.S. have been identified, but have limited success in controlling the weed. The bacterium Pseudom onas syringae pv. p haseolicola has been reported to kill eight to ten week old kudzu seedlings, but produced few, if any, secondary infections under dry conditions in the field ( 230 ). The fungus Myrothecium verrucaria provided 95 to 100% control of kudzu in field tests within 14 days of inoculation by girdling runner stems with the fungus. Although, it can affect a number of important crops, ornamentals, and weeds, representing 6 different families, a patent for kudzu control has been applied for using this fungus ( 19 ). Also, an isolate of the fungus Colletrotrichum gloeosporioides was collected from kudzu in Houston County, Georgia and grown on Czapek Cox medium amended with kudzu extract to incre ase virulence on kudzu. Field inoculations showed a synergistic effect was achieved by a mixture of spores with 20% of the organochlorine compound dicamba. The fungus attacks both leaves and vines ( 51 ). Despite these studies an effective and efficient control measure for kudzu is yet to be discovered.
38 Species Variation In 2005, inter simple sequence repeat (ISSR) analysis was used to investigate the genetic variation of P ueraria lobata and closely related species P. montana P. thomsoni P. edulis and P. phaseoloides ( 189 ). A total of 260 kudzu samples were collected both from China and 19 states of the U.S. The Shannon diversity indices of P. lobata for Chinese and U.S. samples were 0.208 and 0.221, respectively ( 189 ). Such similar diversity indices is contradictory to most cases of plant colonization, because when invasive species colonize a new area, genetic variation is often lower than in the source population due to frequent founder effects ( 161 199 ). Furthermore, P. lobata samples from the U.S. had high genetic diversity and low population differentiation ( 189 ). This suppo rts the theory that in the U.S. multiple introductions from multiple sources, with subsequent gene exchange among different regions or sources of kudzu occurred, resulting in high genetic diversity and low genetic differentiation. Reactions to Soybean Rust Previous studies have reported kudzu exhibiting the same reactions to P. pachyrhizi as soybean: TAN, RB, and no visible symptoms ( 17 82 ). Contrary to soybean reactions, RB lesions were more common (50% frequency) when evaluating 125 kudzu plants with 3 different P. pachyrhizi isolates ( 17 ). Furthermore, in 64% of instances where multiple plants from a site were tested, each reacted the same to the individual pathogen i solate, suggesting a tendency for plants at specific sites to be genetically identical with respect to rust reaction ( 17 ). Only 15% of individual plants produced a different reaction to one isolate than to the other two isolat es suggesting that susceptibility or resistance to P. pachyrhizi in individual kudzu sites is often broad, extending over multiple isolates of P. pachyrhizi
39 Further characterization of kudzu resistance to P. pachyrhizi was conducted in Florida in 2008 ( 82 ). An evaluation conducted on 139 kudzu sites in north Florida resulted in 25 sites free of SBR infection and 32 sites with reduced sporulation ( 82 ). Ten accessions of kudzu from north central Flo rida were examined in detached and attached leaf assays with a single isolate of P. pachyrhizi under controlled laboratory conditions. Of the 10 accessions examined, 6 were susceptible, 3 were immune, and 1 was resistant. Resistant interactions had early onset of a multicell hypersensitive response (HR), while immune interactions were the result of a cell wall deposition that blocked penetration in combination with early onset of HR ( 82 ). Furthermore, using quantitative real time polymerase chain reaction 15 days after inoculation, resistant kudzu had 10 fold less P. pachyrhizi DNA as compared to susceptible kudzu and P. pachyrhizi DNA was below detection level in immune kudzu ( 82 ). Susceptible ku dzu had approximately half the amount of P. pachyrhizi DNA present when compared with susceptible soybean cultivar ( 82 ). Summary With the introduction of SBR to the continental U.S., environmental factors previously identified to affect SBR in other areas of the world need to be examined in the soybean fields of the U.S. to determine major environmental factors that would drive the epidemic in the U.S. Furthermore, the unique relationship of SBR on the overwintering, alternativ e host, kudzu needs further investigation to examine possible resistance genes in the alternative host which would indicate selection pressure on the pathogen due to resistant populations. Potential initial inoculum produced on the alternative host is als o of great importance in understanding SBR epidemics in the U.S. and needs further quantification and modeling. Lastly, knowledge of the environmental
40 factors influencing SBR epidemics in north Florida and potential initial inoculum can be utilized to dev elop a disease and decision model for fungicide application for SBR Based on the previous listed needs the following objectives were created. Research Objectives The objectives of the present research are: 1) to determine environmental and other site spe cific variables that influence SBR epidemics in soybean in north Florida, 2) to determine molecular attributes in resistant kudzu accessions, 3) to estimate potential initial inoculum from kudzu populations in Florida to better characterize SBR epidemics i n Florida, and 4) to evaluate a spray forecasting model for SBR in Florida. To achieve objective 1, data from soybean sentinel plots and corresponding environmental data were analyzed to identify climatic and environmental factors important to SBR epidemi cs in soybean in north Florida (Chapter 2). Furthermore, field experiments were conducted where shading was manipulated in soybean fields and disease development and environmental factors were recorded and analyzed; and detached leaf assays analyzed disea se development without the influence of environmental factors on disease development (Chapter 3). The identification of candidate resistant genes to P. pachyrhizi in the alternative host kudzu (Chapter 3) was performed to achieve objective 2. To achieve objective 3, historical infection data of kudzu populations were utilized and restriction fragment length polymorphisms (RFLPs) and detached leaf assays were carried out to identify and quantify potential inoculum sources from kudzu populations in Florida (Chapter 4). Examining the use of different parameters for a spray forecasting model based on previous studies and potential initial inoculum (Chapter 5) was con ducted to achieve objective 4.
41 CHAPTER 2 EPIDEMIOLOGY OF SOYB EAN RUST IN SOYBEAN SENTINEL PL OTS IN FLORIDA Introduction Soybean rust (SBR), caused by the fungus Phakopsora pachyrhizi is one of the most destructive diseases to soybean production and is monitored in the eastern United States by a series of coordinated sentinel plots. Since SBR wa s discovered in the southeast United States in 2004 ( 167 ) its distribution and severity has been variable from year to year. This study examined data from sentinel plots in north Florida and corresponding weat her conditions to hel p understand this variability. Del Ponte et al. ( 40 ) highlighted the importance of precipitation influencing SBR epidemics, especially for regions where temperature is not a limiting factor for disease development, such as Florida. Examples of the importance of precipitation on SBR development include: severe epidemics in high rainfall areas of the lower Richmond Valley in Australia ( 187 ), yield l osses of up to 100% during rainy seasons and only 10 to 15% losses during a dry season in Thailand ( 163 ), and suppressed disease development in Brazil and Paraguay during moderate to severe droughts in mid to la te growing season ( 224 ). The objective of this study was to examine environmental and climatic factors that affected SBR development within the soybean sentinel plots of north Florida. It was hypothesized that disease onset and development would be highly correlated with moisture values calculated from the Florida Automated Weather Network. Results from this study may lead to a better understanding of the impact of weather on the epidemiology of this disease.
42 Materials and Methods Sentinel Plot Data Beginning in 2005, a national soybean sentinel plot program was established through the Integrated Pest Management Pest Information Platform for Extension and Education (ipm PIPE, sbr.ipmpipe.org) ( 80 ), to which the data for this study are reported. Data were utilized from forty six soybean sentinel plots (SSP) from 2005 through 2008 from 12 different counties in north Florida. Location of the SSPs was dependent upon a gronomic land availability in the counties and when possible were planted in the same location each year. SSPs that did not get planted within 4.8 km of 2007, Santa Rosa 200 6, Okaloosa 2007 & 2008; Walton 2006 & 2008; Holmes 2006; Jackson 2008; Jefferson 2006; Madison 2006; Suwannee 2008; Columbia 2008 ( Figure 2 1). Seven to 11 different SSPs became infected with SBR each year and were used for analysis. On average, SSPs wer e planted in a 225 m 2 area with 25 cm row spacing. All locations across the years were planted in March or April (Julian date 83 119) to detect SBR at the earliest date it developed in the county. Two different maturity group 3 (MG3) varieties, 2 differ ent MG5 varieties, and 3 different MG7 varieties were used in SSPs, from 2005 through 2008, (Table 2 1). All varieties used are susceptible to SBR The SSPs were managed using standard production practices for Florida ( 217 ) and were not irrigated. Data C ollection Each SSP was scouted bi weekly and then scouted weekly for SBR after one of the following conditions occurred: first bloom of soybean (growth stage R1) ( 113 ), presence of SBR in the surrounding counties, forecast or spore deposition data
43 suggested increased risk, or environmental conditions became conducive to SBR development. Scouting consisted of arbitrarily collecting 100 leaflets from each maturity grou p, so each site had a total of 300 leaflets collected per sampling date. Leaflets collected from each maturity group were placed in 3 separate plastic bags, stored at room temperature (~ 25 C) and evaluated within 1 to 7 days. Maturity group data were r ecorded individually each year, except in 2005 and SSP data from Santa Rosa County in 2006 where maturity group data were compiled together and reported. As such 2005 data and Santa Rosa County data from 2006 were omitted when analyzing maturity group dat a. Disease severity was determined using a dissecting microscope for symptomatic leaf area on individual leaflets. An arbitrary rating scale was used with the following values and corresponding severity ranges based on a visual key developed by Bayer Crop Science (Kansas City, Kansas), 0 = no disease, 1 = up to 2.5%, 2 = 2.5 5%, 3 = 5 10%, 4 = 10 15%, 5 = 15 25%, 6 = 25 35%, 7 = 35 67.5%, and 8 = 67.5 100% of the leaf area affected by SBR. The following mid point rating was used to create severity values, i.e., 0=0%, 1=1.25%, 2=3.75%, 3=7.5%, 4=12.5%, 5=20%, 6=30%, 7=51.25%, and 8=83.75%. Incidence and severity data for each individual MG was averaged for each sampling date, and this average was used to calculate weekly and yearly averages. Maximum incid ence and severity values for each SSP were averaged to determine yearly maximum incidence and severity, regardless of date (Table 2 1). Incidence and severity values were aligned chronologically and used to obtain the area under the disease progress curve (AUDPC) (Table 2 1, Figure 2 2). The previous
44 analyzed (i.e. when analyzing Jefferson SSP 2006 maximum incidence data, the Jefferson SS P 2005 maximum incidence data maximum severity to use in analysis (i.e. no data was recorded in 2004). In the circumstance where a SSP did not become infected, 0 was used for maximum incidence and severity. In such a case if a SSP did not become infected until 2008. In the event the SSP was lost due to poor germination or deer d amage these data were omitted. The duration of the epidemic was calculated from the week of first detection to the week of maximum severity (Table 2 2 ). Average incidence and severity were plotted over time ( Figure 2 2 ) and the AUDPC was calculated using Equati on 2 1 ( 205 ). (2 1) w here x i = proportion of tissue affected (disease severity) at the i th observation, t = time (days), and n of all of the individual trapezoids or areas from i to n 1. i and i +1 represent observations from 1 to n Julian days were used for date calculations, where January 1st is Julian day 1. Yearly disease incidence and severity averages, discussed previou sly, were used to calculate the AUDPC for each year. AUDPC was calculated through the second week in September each year based on natural senescence of plots (Fig ure 2 2 ). The apparent infection rate per week for each year was calculated from weekly seve rity averages from the week of detection to the week of maximum severity, these
45 were also averaged together to get average infection rate per year for the duration of the epidemic (Table 2 1). Apparent infection rate was calculated using Equation 2 2 ( 146 ). (2 2) W is the natural logarithm. Disease severity at time t 1 =x 1 (disease severit y at first disease detection) and disease severity at time t 2 =x 2 (disease severity w hen maximum severity reached). Six climatic time periods were used for analysis: 2, 3, and 4 weeks before disease detection (including first week of detection) and 4, 6, an d 8 weeks before maximum incidence and severity. These time periods were used to correlate preceding environmental conditions with first disease detection, maximum incidence, and maximum severity. The Florida Automatic Weather Network stations (FAWN) wer e utilized to obtain environmental data collected from five counties in the Florida Panhandle (Escambia, Jackson, Gadsden, Jefferson, and Suwannee) ( Figure 2 1).Weekly average temperatures were calculated from the daily average air temperatures from FAWN s tations, recorded at 60 cm above the soil. Similarly, the weekly average relative humidity (RH %), solar radiation (watts per square meter), and precipitation (mm) was calculated from FAWN stations at 2 m above soil level, and soil temperature taken at 10 cm below soil surface. These weekly averages were used to create the yearly averages presented in Figure 2 2 The dew point was calculated from the daily average air temperature and RH using E quation 2 3 ( 10 )
46 (2 3) Where where T is the temperatures in degrees Celsius RH is the measured relative humidity, T d is the calculated dew point temperature ( C) and "ln" refers to the natural logarithm The constants are: a = 17.27 and b = 237.7 ( C). The uncertainty in the measured dew point temperature is a function of the measured temperature and relative humidity and the uncertainties associated with those measurements. The uncertainty in the calculated dew point temperature is 0.4 C. This expression is based on the "Magnus" (or "Magnus Tetens") approx imation for the saturation vapor pressure of water in air as a function of temperature ( 10 ). It is considered valid for 0 C < T < 60 C; 1% < RH < 100%; 0 C < Td < 50 C. Dew point temperatures were then compared with the hourly average air temperature and hours where the difference between the two variables was 0 to 2 were counted as dew leaf wetness hours for daily calculations ( 55 ). Daily leaf wetness hours were averaged over the different time periods. The number of days with >0 mm, >1 mm, and >5 mm of precipitation and cumulative precipitation were recorded for each time period in question using FAWN data. All variables utilized in this study are listed in Table 2 2 Data Analysis Statistical analyses were performed using SAS versi on 9.1 (SAS Institute, Inc., correlation test. Correlations were examined between first detection date of SBR and weekly average air temperature, soil temperature, relative humidity, solar radiation,
47 precipitation, daily leaf wetness hours, number of days with precipitation >0mm, >1mm, date, in each year, was correlated with climatic variables averaged over 2, 3, and 4 weeks prior to detection the correlations between SBR maximum incidence and severi ty (percentage), first detection date, weekly average air temperature, soil temperature, relative humidity, solar radiation, precipitation, daily leaf wetness hours, number of days with precipitation >0mm, >1mm, and >5mm, the number of years a site was pre previous maximum incidence and severity. Maximum incidence and severity for each site and year was correlated with climatic variables averaged over 4, 6, and 8 weeks R esults Al l correlations were significant with a p value less than 0.05. Significant correlations between climatic variables and first detection date of SBR, maximum incidence and maximum severity are listed in Tables 2 3, 2 4, and 2 5, variables not listed were nonsignf icant. Heavy rainfall (> 80 mm) occurred during 2005 and 2008 with earlier disease detection and higher maximum incidences and severities. In contrast, dry to drought conditions (< 80 mm of total rain) occurred during 2006 and 2007, with later disease de tection, and lower maximum incidences and severities ( Figure 2 2 ). Disease duration (time period from first detection to maximum severity) increased from 2005 through 2008, while area under the disease progress curve (AUDPC), first disease detection, maxi mum incidence and severity varied between years (Table 2 1). Five sentinel plots in 2005 and one in 2007 and 2008 were lost due to poor germination or
48 deer damage before infection could occur, and in 2006 three sentinel plots never became infected. The ma jority of SSPs (33 plots) did not become infected with SBR until growth stage R4 or later; however, 2 SSPs were infected by growth stage R2 and 1 SSP was infected by growth stage R3 ( 113 ) ( Figure 2 3). In SSPs from 2006 through 2008, SBR infections were first discovered in MG3, MG5, MG7 a total of 4, 17, and 14 times, respectively (Table 2 6). First disease detection was significantly correlated with the number of years a site had been infected ( 0.35), previous 0.49), planting date ( 0.34), and maximum incidence and severity ( 0.66 and 0.44, respectively) (Table 2 3). Environmental variables to which first disease detection were negatively correlated with included the number of days with greater than 0 mm and 1 mm precipitation ( 0.39 and 0.49, respectively), cumulative solar radiation ( 0.54), and average leaf wetness ( 0.53) during the 4 week time period prior to first disease detection (Table 2 3). Temperature and relative humidi ty did not have a significant correlation with first disease detection. Maximum disease incidence significantly decreased 23%, 17%, and increased 59% in 2006, 2007, and 2008, respectively, in comparison with maximum incidence of the previous year (Table 2 2 ). Maximum incidence had a significant positive correlation with average weekly precipitation (0.39) for the 8 week period prior to maximum incidence date (Table 2 4). Maximum severities increased 22%, decreased 64%, and increased 275% in 2006, 2007, an d 2008, respectively, in comparison with the maximum severity of the previous year (Table 2 2 ). Climatic variables that significantly correlated with maximum severity included average weekly precipitation (0.38) and average leaf wetness (0.37) 4 weeks pri or to maximum severity date. Cumulative solar
49 radiation 4 weeks prior to maximum severity date was weakly correlated to maximum severity (p value 0.0417 and correlation coefficient 0.35) (Table 2 5). Plotting cumulative precipitation (mm) recorded 4 wee ks prior to maximum severity date and the maximum severity (%) from 2005 through 2008 revealed a threshold of 300 mm cumulative precipitation ( Figure 2 4). Discussion This study examined the epidemic development of SBR from 2005 through 2008. Epidemic deve lopment was correlated to multiple environmental variables with important trends apparent between the first detection of SBR and the plant descriptor variables of growth stage and maturity group. While SBR can infect soybean at any growth stage, it has be en indicated that earlier SBR infections within a growing season increases the chances of yield loss due to higher disease incidence and severity ( 222 ). Rarely did SBR infection occur on MG3 soybeans before the other varieties or before the MG3 soybeans reached growth stage R8 (full maturity) (Table 2 6, Figure 2 3). Most often, MG5 and MG7 soybean varieties were first detected with SBR, and at a range of growth stages including R2 (full bloom) through R8 (full maturity) (Table 2 6, Figure 2 3). The majority of SSPs were in growth stage R4 (full pod) or later when SBR was first detected ( Figure 2 3), even though MG3 varieties were at reproductive stages earlier in the year. Hence, soybeans did not become infec ted at earlier growth stages, most likely due to lack of inoculum and/or condu cive environmental conditions. From 2006 through 2008, fluctuating first disease detection was correlated with the previous number of years a site had been infected and the previ incidence, suggesting that accumulated inoculum could contribute to first disease detection. While first disease detection was negatively correlated with planting date,
50 this might be attributed to other factors over time, such as accumu lated inoculum, in that the planting dates were slightly later each year (Table 2 2 ). First disease detection data also were correlated strongly with maximum incidence and severity, which can be expected, because with earlier disease onset greater maximum incidence and severity can be ac hieved and at an earlier date. The environmental variables most strongly correlated to SBR first detection were precipitation of greater than 1 mm, cumulative solar radiation intensity, and average leaf wetness calculated o ver the 4 week period prior to SBR detection (Table 2 3). Although temperature and relative humidity were not directly correlated with first disease detection, the significant correlations of calculated leaf wetness provide an indirect correlation of thes e variables. The observed correlations with first disease detection imply days with greater than 1 mm of precipitation, increased leaf wetness hours, and decreased solar radiation promote earlier disease onset. This concurs with reports by Marchetti et a l. ( 109 ), Melching et al. ( 114 ) and Narvaez et al. ( 133 ) that increased moisture, increased SBR disease intensity. Howe ver, the lack of a correlation between first disease detection and the number of days with greater than 5 mm of precipitation may be explained by the observation that subsequent rainfall for a 30 minute period can remove up to 91% of urediniospores from so ybean leaves ( 46 ). Furthermore, the correlation between first disease detection and solar radiation coincides with Isard et al. ( 78 ) report of the proportion of P. p achyrhizi spores that germinate decrease with increasin g exposure to solar radiation. Environmental conditions, such as periods of high moisture and low solar radiation, can occur throughout the soybean growing season in Florida during
51 hurricanes and tropi cal storms. For example, on July 10 th of 2005 Hurricane Dennis made landfall on Santa Rosa Island in north Florida, and resulted in substantial amounts of precipitation (max = 118 mm) across the north Florida area. The environmental conditions created by Hurricane Dennis could account for the 2005 average disease detection date of August 19 (Julian day 231), being earlier than the following years of 2006 and 2007, which had dry to drought conditions. Notably, precipitation is a readily available variable throughout many regions across the U.S. and the correlation of first disease detection with precipitation can be used to determine areas that will become infected after possible spore deposition events. Del Ponte et al. ( 40 ) demonstrated that accumulated precipitation from the following month that SBR was initially detected could be used as a predictor for final disease severity. Similarly, in this study we found maximum incidence and severity correlated with av erage precipitation prior to maximum incidence and severity dates. In particular, maximum incidence correlated with average precipitation over 8 weeks prior to maximum incidence date (Table 2 4), whereas maximum severity was most significantly correlated to the average precipitation over 4 weeks prior to maximum severity date (Table 2 5). The positive correlation of maximum incidence with precipitation over an 8 week time period suggests that incidence is influenced more than maximum severity by precipita tion events occurring prior to disease onset. also a greater average maximum incidence in 2005 as compared to 2006 and 2007. In addition, the greatest maximum incidenc e and severity averages in 2008 may be attributed to the environmental conditions resulting from Tropical Storm Fay, which
52 traveled across north Florida August 17 24 and accounted for the peak of average precipitation during the third week in August (Jul ian day 230) ( Figure 2 2 ). The greatest maximum severity and precipitation occurred in 2008, indicating that rain, after initial infection, can contribute to the severity of a SBR epidemic in north Florida and that relatively dry to drought conditions, ob served in 2006 and 2007, can reduce the maximum severity reached during a growing season ( Figure 2 2 ). Accumulated precipitation 4 weeks prior to maximum severity displays a threshold of 300 mm cumulative precipitation ( Figure 2 4); where average maximum s everity was 20.8% and 9.8% if precipitation was more and less than 300 mm, respectively. Similarly, Del Ponte et al. ( 40 ) found thresholds in rainfall correlating to SBR severity in Brazil during the 1 month pe riod after disease detection, where severity was >70% and <30% if rainfall ranged between 250 459 mm and 10 125 mm, res pectively. Compounding factors contributed to fluctuating first disease detection, maximum incidences and severities from 2005 to 2008 in north Florida soybean sentinel plots. These factors include, but are not limited to, varying levels of initial inoculum, leaf wetness hours, and amount of precipitation before and after infection. Correlations between climatic variables and maximum incid ence and severity imply certain environmental conditions are favorable for SBR and may allow the pathosystem to reach a maximum impact earlier in the season. Furthermore, correlation between early disease detection and early maximum incidence and severity reached may lead to earlier available inoculum contributing to rust epidemics in the major soybean regions in the U.S. This study indicates that improved prediction of SBR epidemic severity and understanding of the variability of epidemics from year to ye ar in the United States can
53 be gained from continued research and monitoring of SBR in the southeastern United States as we have yet to see the maxim um potential of a SBR epidemic.
54 Table 2 1. Data summary of north Florida soybean sentinel plots (SSP) and t he soybean rust epidemic data within those plots from 2005 through 2008 2005 2006 2007 2008 Variety Planted MG3 DP 3861RR MG5 DP 5915RR MG7 DP 7870RR MG3 HS 3456RR MG5 DP 5915RR MG7 DP 7220RR MG3 HS 3861RR MG5 DP 5915RR MG7 DP 7220RR MG3 DP 3861RR MG5 AG 5905RR MG7 DP 7330RR Total number of (SSP) 12 10 12 12 Number SSP infected 7 7 11 11 AUDPC a (SEV b ) 146 39 20 204 AUDPC a (INC c ) 19 4 4 21 Plant date in Julian day 83 88 94 99 First disease detection in Julian day 231 258 243 216 Days after plant ing disease was detected in Julian day 148 170 148 115 Duration (days) 24 33 34 58 Avg. max. SEV b (%) 9 11 4 15 Avg. max. INC c (%) 82 64 52 83 Avg. apparent Infection rate 0.22 0.11 0.17 0.24 a AUDPC = area under disease progress curve. b SEV = disea se severity. c INC = disease incidence. Abbreviations for varieties planted include, MG = maturity group, RR = roundup ready, DP = Delta and Pine Land seed, HS = Hyland Seed, AG = Asgrow seed.
5 5 Table 2 2. A description of variables Variables were u sed to examine the relationship between soybean rust first disease detection, maximum disease incidence, severity, and environmental conditions in north Florida soybean sentinel plots from 2005 through 2008. Variable Description Onset First disease detect ion date MaxIncid Maximum disease incidence (%) reached at a particular site and year MaxSev Maximum disease severity (%) reached at a particular site and year Planting date Date planting occurred PrevYRMaxSev %) PrevYRMaxIncid NumyrInf Number of years site was previously infected InfectRate Infection rate calculated using initial and maximum disease severity AvgRain* Average precipitation (mm) RD>0mm* Number of days with precipitation >0mm RD>1mm* Number of days with precipitation >1mm RD>5mm* Number of days with precipitation >5mm SolRad* Cumulative solar radiation (W/m 2 ) AvgTemp* Average air temperature at 60 cm height AvgSoilTemp* Average soil tempera ture at 10cm depth RH* Average relative humidity LeafWetAvg* Average of leaf wetness hours *Each variable was determined for each of the six time periods of climate examined: 2, 3, and 4 weeks before first disease detection (including first week of det ection) and 4, 6, and 8 weeks before maximum incidence and severity was achieved.
56 Table 2 3. Variables correlated with first disease detection date. Variables correlated, t during the growing seasons of 2005 through 2008 in soybean sentinel plots in north Florida. Onset a Variable P value Correlation coefficient MaxIncid b <0.0001 0.66 MaxSev c 0.0080 0.44 Planting date 0.0434 0.34 PrevYRMaxIncid d 0.0048 0.49 Numy rInf e 0.0381 0.35 RD>0mm4wks f 0.0193 0.39 RD>1mm4wks g* 0.0026 0.49 SolRad 2 wks h 0.0025 0.49 SolRad 3 wks i 0.0007 0.54 SolRad 4 wks j* 0.0006 0.54 LeafWetAvg 4wks k* 0.0009 0.53 a Onset = First disease detection date. b MaxIncid = Maximum disea se severity (%) reached at a particular site and year. c MaxSev = Maximum disease severity (%) reached at a particular site and year. d e NumyrInf = Number of years site was previously infecte d. f RD>0mm4wks = Number of days with precipitation >0mm, over 4 weeks prior to first disease detection date. g RD>1mm4wks = Number of days with precipitation >1mm, over 4 weeks prior to first disease detection date. h SolRad 2 wks = Cumulative solar radia tion (W/m 2 ), over 2 weeks prior to first disease detection date. i SolRad 3 wks = Cumulative solar radiation (W/m 2 ), over 3 weeks prior to first disease detection date. j SolRad 4 wks = Cumulative solar radiation (W/m 2 ), over 4 weeks prior to first disease detection date. k LeafWetAvg 4wks = Average of leaf wetness hours, over 4 weeks prior to first disease detection date. Most significant climatic variables correlated.
57 Table 2 4. Variables correlated with maximum incidence. Variables correlated, using P growing seasons of 2005 through 2008 in soybean sentinel plots in north Florida. Maximum Incidence Variables P value Correlation coefficient MaxSev a <0.0001 0.70 AvgRai n 8wks b 0.0196 0.39 a MaxSev = Maximum disease severity (%) reached at a particular site and year. b AvgRain8wks = Average precipitation (mm), over 8 weeks prior to maximum incidence date.
58 Table 2 5. Variables correlated with maximum severity. Variables correlated, using growing seasons of 2005 through 2008 in soybean sentinel plots in north Florida. Maximum Severity Variables P value Correlation coefficient MaxIncid a < 0.0001 0.70 AvgRain 4 wks b 0.0256 0.38 SolRad 4 wks c 0.0417 0.35 LeafWetAvg 4 wks d 0.0266 0.37 a MaxIncid = Maximum disease incidence (%) reached at a particular site and year b AvgRain 4 wks = Average precipitation (mm), over 4 weeks prior to maximu m severity date. c SolRad 4 wks = Cumulative solar radiation (W/m 2 ), over 4 weeks prior to maximum severity date. d LeafWetAvg 4 wks = Average of leaf wetness hours, over 4 weeks prior to maximum severity date.
59 Table 2 6. T he number of soybean sentinel pl ots (SSP) in which a maturity group was first detected with soybean rust ( SBR) Data from 2006, 2007, and 2008 SSP Maturity group first detected with SBR MG3 MG5 MG7 Total Number of diseased SSP/year 2006 1 4 a 4 a 6 d 2007 2 b 4 b 7 b 11 2008 1 9 c 3 c 11 Total 4 17 14 a At 3 SSP both MG5 and MG7 were detected with SBR at the same date. b Only 6 of the 7 SSPs in 2006 had maturity group data reported and used in this study. c At 1 SSP all 3 maturity groups were detected with SBR at the same date. d At 2 SSPs both MG5 and MG7 were detected with SBR at the same date.
60 Figure 2 1. Florida map of soybean sentinel plots and Florida Automated Weather Network stations. A Florida map indicating, with the circle, the study area of north Florida. B Enlarged a rea of north Florida with approximate locations of soybean sentinel plots from 2005 2008 used to monitor soybean rust and Florida Automated Weather Network (FAWN) stations that were used to obtain environmental data utilized in this study.
61 Figure 2 2. W eekly cumulative rain and disease incidence and severity over time. The disease incidence (square symbol, primary y axis) and severity (triangle symbol, primary y axis) of soybean rust, and weekly cumulative rain (bars, secondary y axis) were averaged acro ss soybean sentinel plot locations in north Florida and plotted over time (in Julian days, x axis) during the seasons of 2005 through 2008.
62 Figure 2 3. Soybean growth stage and maturity group (MG) data over time. Soybean growth stage and maturity group (MG) data are plotted over time (in Julian (SBR) during the seasons of 2005 through 2008. Soybean MG was not recorded for 2005 SSPs or Santa Rosa SSP in 2006, only soybean growt h stage is reported in this figure for these data. In 2005, 3 of the 7 SSPs were first detected with SBR at growth stage R5 on 15 August (Julian day 227). out of 7 SSPs were first detected with SBR on MG5 and MG7 on 27 July (Julian date 208) at growth stages R6 and R5, respectively. The grayed other SSP in 2006 were first detected with SBR on MG5 and MG7 on 28 September (Julian date 271) all at growth stage R8; and 1 other SSP was first detected with SBR on 28 September (Julian ate 271) at growth stage R8. represents these 3 SSP dat a. In 2007, 2 out of 11 SSPs were first detected with SBR on MG5 on 8 August (Julian date 220), both at growth stage R5. other SSP in 2007 was first detected with SBR on MG3, MG5, an d MG7 on 7 September (Julian date 250) at growth stages R8, R8, and R6, respectively. In 2008, 1 out of 11 SSPs were first detected with SBR on MG5 and MG7 on 8 July (Julian date 190) at growth stages R4 and R2, respectively. Another SSP in 2008 was first detected with SBR on MG5 and MG7 on 27 August
63 (Julian date 240) at growth stages R5 and R6, respectively. The grayed squares and grayed triangles aligned vertically on the 2008 graph represent these SSP data. Two other SSP in 2008 were first detected with SBR on MG5 on 20 July (Julian date 202) at growth stage R4. The bolded square on the 2008 graph represe nts those 2 SSP data. Figure 2 3. Continued.
64 Figure 2 4. Precipitation threshold for average maximum disease severity of soybean rust in soybean sentinel plots (SSP) recorded from 2005 through 2008. Symbols represent each SSP average maximum disease s everity, as a percentage, reached each year and cumulative precipitation (mm) over the 4 weeks prior to each site reaching maximum disease severity each year (different symbols representing each year); and the line is the threshold at 300 mm of precipitati on.
65 CHAPTER 3 EFFECT OF SOLAR RADI ATION ON SEVERITY OF SOYBEAN RUST Introduction Soybean rust (SBR), caused by Phakopsora pachyrhizi Syd. and P. Syd., has the potential to be a damaging fungal disease of soybean (Glycine max (L.) Merr). SBR often is first detected in low light environments and regularly progresses from the lower to upper soybean canopy. Favorable meteorological and environmental conditions such as temperature, relative humidity and rainy days with prolonged cloudiness are critical component s that affect the progression of disease development ( 38 ) Identification and quantification of the key factors that drive a SBR epidemic are critical components for assessing and managing SBR risks. Such studies provide a rati onal basis to improve disease management and subsequently prevent yield loss. There is very little quantitative information on the effect of environmental and meteorological conditions in the field when compared to the body of literature from growth chambe r and greenhouse studies. Since long distance dispersal of SBR urediniospores has been confirmed ( 94 ), there has been a surge in new epidemiological studies for SBR to accurately predict and manage the disease across a large ge ographical area. As a consequence, several disease models have been developed that integrate meteorological data such as wind speed, wind direction, and solar radiation as predictors to assess the risk of the disease at various geographical scales ( 40 79 142 ) While airborne, the survival of fungal spores is affected by temperature, solar radiation, ultraviolet radiation, and relative humidity ( 79 ). Pathogenicity of fungal spores, in the course of spores long distance aerial spread, is dependent on spore survival
66 during variation in temperature, relative humidity, and solar radiation over the time that they are suspended in the atmosphere and a fter landing on a susceptible host. Solar radiation has been demonstrated to have the potential to increase canopy temperature, reduce relative humidity (RH), and reduce SBR urediniospore survival ( 79 ), but limited field inform ation is available on how solar radiation triggers physiological changes in the structure of the soybean plant canopy, which might also influence SBR development. This information could be another important factor for accurate predictions using aerobiologi cal models for SBR forecasting ( 79 142 ). Components of solar radiation (UVB radiation; 280 to 320 nm) can affect crop growth directly through photosynthesis and photomorphogenic systems, up regulati on of pathways producing defense compounds (e.g., flavonoids and related phenolic compounds or waxes), decreased vegetative growth, and decreased developmental times. In contrast, the UVB (280 320 nm) component of solar radiation also has a range of effect s on invading pathogens that include damage to DNA and proteins thus affecting metabolic and physiologic responses of the organism ( 7 ). Ghajar et al. ( 58 ) and Rotem and Aust ( 159 ) reported that UV radiations in sunlight, and not temperature, are the major factor affecting germination of fungal propagules. The extent of damage caused by biotic factors (pest, pathogens and weeds) on growth and development of crop plants can be modified by abiotic factors such as solar radiation, possibly making the plants more or less susceptible to pest pressure ( 84 ). Soybean leaves from canopies exposed to solar UVB showed significantly higher levels of soluble phenolics and lower levels of lignin than leaves that developed in canopies covered by polyester films ( 227 ). Canopy shading can reduce solar radiation
67 and impact the microclimate affecting plant physiology including cuticular wax. Experimental and observational data reports show that high radiant energy, low humidity, and higher temperatures (23 to 30 C) are favorable conditions for maximum wax production ( 9 33 177 ). Once the canopy closure is 100%, creating a completely shaded area between rows, the entire exposed upper canopy begins to act as a more uniform evaporative surface diminishing the differences in shading of the soil surface and variation in heat from soil between rows due to reflection of radiation ( 181 ). Recently, the effects of sunlight (both total solar and ultraviolet radiation) on urediniospore survival have been used to predict P. pachyrhizi ure diniospore germination ( 78 ). The recommendations, based on the findings of Isard et al. ( 78 ), have SBR epidemiology. While field experiments on influ ence of total solar radiation on P. pachyrhizi urediniospore germination have been conducted by Isard et al. ( 78 ), these factors have not been studied under laboratory conditions in the absence of other environmental parameters that may otherwise confound the correlation between them and urediniospore survival and viability. The objectives of this study were: to investigate the effect of solar radiation on progress of SBR in the field and to evaluate the effect of solar radiatio n on epicuticular wax content of soybean and if it is correlated to disease susceptibility. Material and Methods Canopy M anipulation To investigate the effect of solar radiation on the progress of SBR four soybean plantings were established at the North F lorida Research and Education Center (NFREC) in Quincy, Florida. Early plantings occurred on 11 April 2008 and 24 April 2009 and late plantings on 4 August 2008 and 23 July 2009; henceforth designated as
68 early or late plantings, respectively. All plantings were in the same field. Asgrow cv. 5905 Roundup Ready, maturity group 5 soybeans were planted on 76 cm rows, over a total area of 15,600 square meters, on average. Three 3 m x 3 m square structures of black knitted shade cloth (Rimol Greenhouse Systems, I nc. Hooksett, NH) were constructed at least 4.6 m apart from each other and had a minimum of 4.6 m soybean boundary around the perimeter. Control plots without shade cloth with the same area and parameters were also established. The shade cloth attenuated 30, 40, or 60% sunlight as stated by the manufacturer and determined by Optronic OL754 spectroradiometer ( Optronic Inc., Orlando, FL ) measurements of energy values (W m 2 ) under each shade cloth in field plots ( Table 3 1 ). All plantings had 3 replicates (t otal of 12 plots). The shade cloth structures were constructed over the soybeans at vegetative growth stage V3 V4 ( 113 ) in a complete randomized block design. Subsequently, leaves were marked on 3 individual plants in each plot starting at growth stage V3, to be utilized for leaf assays Plant height was measured at reproductive growth stage R3 in the early and late planting of 2008 in all treatments. At reproductive growth stage R3 there is little to no further increase in hei ght in determinant soybeans and hence this stage was chosen to measure the overall effect the treatments had on plant height. Data log tags (MicroDAQ.com, Ltd. Contoocook, NH) were used to record temperature and relative humidity every 20 minutes in the m iddle of each plot at approximately 36 cm above the soil for the duration of the experiment. The weeks during reproductive growth stage R3 R4 (pod growth) of each planting were selected as typical examples of temperature and relative humidity observed thro ughout the growing season and critical time periods for
69 disease development. Leaf area indices (LAI) were obtained for early and late plantings of soybeans in 2008 on 9 July and 3 October at reproductive growth stage R3and for the late planting in 2009 o n 11 October at reproductive growth stage R6 by measuring photosynthetically active radiation (PAR; 400 to 700 nm) approximately 10 to 20 cm above the top of the canopy and 25 30 cm below the top of the canopy using a LI COR 191 quantum sensor and LI 1400 Data Logger (LI COR Biosciences; Lincoln, NE). All measurements were taken between 11:00 AM and 2:00 PM (EDT) (UTC 16:00 19:00). The following equation was used to calculate LAI from the PAR values: LAI = ( 1/k ) ln( Q b /Q a ) ( 214 ), where Q a is the above canopy PAR, and Q b is the below canopy PAR. The extinction coefficient, k of 0.4 was used based on Flenet et al. ( 52 ) study on row spacing effects on extinction coefficients of soybean. Average light transmissions at the upper soybean canopies in plots covered by the 30, 40, and 60% shade material were measured using an Optronic OL754 spectroradiometer ( Optronic Inc., Orlando, FL ) measuring energy values (Wm 2 ) between 270 and 775 nm ( Table 3 1 ). Simi lar measurements were also made in the lower, middle, and upper canopies of control plots ( Table 3 2 ). All light transmission measurements were taken on 28 September, 2009. Field Inoculations and E valuations Urediniospores used for inoculation were collect ed the previous week from soybeans (Asgrow cv. 5905 Roundup Ready, maturity group 5) infected with a population of urediniospores collected from naturally infected soybean at the North Florida Research and Education Center in Quincy, Florida the previous y ear. Diseased plants were maintained in a greenhouse and urediniospores were collected using a cyclone spore collector (G R manufacturing Co., Manhattan, KS) from the underside of infected leaves and stored in a 10 ml glass vial at 20 C until spore susp ension was
70 made. The early plantings of 2008 and 2009 were inoculated at growth stage R1 with a suspension of approximately 75,000 urediniospores per ml of water and 10 l of Tween 20 between 6 and 8 p.m. Eastern Standard Time. Inoculation occurred in the evening to ensure all plots were exposed to similar environmental conditions for inoculation and to enhance probability of infection. Late plantings were naturally infected with SBR with no addition of inoculum. Germination of urediniospore was assessed by plating 100 l of urediniospore suspension on 1% water agar and recording the number of germinated urediniospores out of 100. A spore was considered germinated if the germ tube was suspension was applied to the lower canopy in each plot using a hand held spray bottle. Weekly evaluations were made, starting 10 to 14 days after inoculation for early plantings or for late plantings at growth stage R1 (flowering) because natural SBR inf ection usually occurs at the beginning of soybean reproductive stages. Ten randomly pre assigned areas in each plot were evaluated at the lower, middle, and upper canopies for severity of SBR based on visual assessment of symptomatic leaf area using the sc ale, 0= no disease, 1 = up to 2.5%, 2 = 2.5 5%, 3 = 5 10%, 4 = 10 15%, 5 = 15 25%, 6 = 25 35%, 7 = 35 67.5%, and 8 = 67.5 100% of the leaf area affected by SBR. Average severity rating was determined by converting each rating to the midpoint of its range, i.e., 0=0%, 1=1.25%, 2=3.75%, 3=7.5%, 4=12.5%, 5=20%, 6=30%, 7=51.25%, and 8=83.75% ( 225 ). The lower, middle, and upper canopy included plant area from nodes 1 to 3, 3 to 6, and 6 to 10, respectively. In the early plantings, the lower, middle, and upper canopy were designated between 10 and 20 cm, 20 to 40 cm, and 40 to 70 cm from the ground, respectively. In the late plantings, the lower, middle, and upper canopy were
71 designated between 10 to 30 cm, 30 to 60 cm, and 60 to 10 0 cm from the ground, respectively. Detached Leaf A ssays At growth stage R1 (flowering)( 113 ), two pre marked leaflets each from the lower, middle, and upper canopies, as previously described, were collected for leaf assays to a ssess the amount of epicuticular wax and susceptibility of leaves to SBR. An additional leaflet was collected from the lower canopy of each plot to verify SBR was not present in the field at that time. The detached leaflets were rinsed with sterile distill ed water and allowed to dry. One set of leaflets was immediately used to assess susceptibility and the other set was placed in resalable plastic bags and placed in a freezer at 20 C for later analysis of epicuticular wax. Leaflets collected from both ear ly plantings and the late planting in 2008 were used in susceptibility assays and leaflets from the early planting in 2009 and both late plantings were used in epicuticular wax assays. For the susceptibility assay the detached leaflets were inoculated with urediniospores of P. pachyrhizi which were collected from diseased soybeans in a greenhouse the previous day and kept in frozen storage (approximately 20 C) until use. A suspension of 15,000 urediniospores per mL of water with 10 l of Tween 20 was used to inoculate the upper and lower surface of each leaflet using approximately 3 mL of suspension per leaflet by using a KC 566CG air sprayer (Lowes; North Wilkesboro, NC). The suspension was also sprayed on a plate of 1 % water agar and germination was ass essed as previously described. After inoculation, the petiole of each leaflet was cut diagonally with a scalpel to produce a clean, fresh surface. Each leaflet was then placed in an individual Petri plate (150 x 15 mm) half filled with 1% water agar
72 with t he petiole inserted into the a gar. Petri plates were kept at room temperature (approximately 24C) and exposed to approximately 30 % of natural sunlight for 8 hours per day. Leaflets were evaluated 10 to 14 days after inoculation using the rating scale prev iously given. Non inoculated control leaflets from the lower canopy were also maintained in Petri plates half filled with 1% water agar and evaluated along with inoculated leaflets. The other set of leaflets were removed from the freezer and allowed to tha w 2 to 3 months later. Epicuticular wax was then measured using the following modified procedures from Beattie and Marcell ( 12 ). Individual leaflets were submerged in 10 mL of chloroform in pre weighed aluminum foil boats for 3 0 s. The waxes were dried by placing the foil boats on a hotplate (approximately 150 C) until chloroform had evaporated. After an hour of incubation at room temperature the foil boats were weighed. The area of each leaflet was obtained using the LI COR 30 00 portable area meter with LI COR 3000 transparent conveyer belt accessory (LI COR Biosciences; Lincoln, NE). The amount of wax extracted from a leaflet was expressed in mg of wax per cm 2 of leaf area. Data A nalysis Statistical analyses were performed usi ng SAS 9.1 (Statistical Analysis Software version 9.1, SAS Institute Inc., Cary, NC). Temperature and relative humidity were averaged across repetitions and 7 days soybean was in reproductive growth stage R3 R4 (pod development) each planting, resulting wi th data in 20 minute intervals over a 24 hour period typical of temperature and relative humidity observed throughout the growing season. These data were analyzed by the general linear model procedure and
73 nificant level of 0.05. Disease severity recorded at growth stage R5 was averaged across each planting date at each canopy level and treatment. Disease severity recorded in field experiments at growth stage R5, severity of detached leaves, and epicuticular wax from detached leaves were analyzed by two factor analysis of variance (ANOVA) with interaction. Factors used as main difference test at a significant level of 0. 05 was used for multiple comparisons among treatments and canopy height by means of the general linear model procedure of SAS when significant differences were found in the analysis of variance table. Significant s correlation test. Results At growth stage R3, in the early planting (11 April) of 2008, plant height under 40 and 60% shade were significantly greater than those under 0 and 30% shade, whereas in the late planting (4 August) of 2008 there was no signific ant difference in plant height among treatments ( Fig ure 3 1 ). Furthermore, plant height from the early planting of 2008 was significantly greater than plant height from the late planting of 2008. Leaf area indices from the early planting of 2008, with the exception of LAI from 60% shaded soybean, were significantly greater than LAI from the late plantings in 2008 and 2009 ( Fig ure 3 2 ). There were no significant differences in LAI across treatments from the 2008 plantings, however LAI from the 2009 planting non shaded and 30% shaded treatments had significantly greater LAI than that from 40 and 60% shaded treatments. The two way ANOVA of the disease severity at reproductive stage R5 showed a significant effect of the shade treatments ( F = 10.78, P = 0.0001) a nd canopy heights ( F
74 = 67.97, P < 0.0001), but not of their interaction ( F = 8.06, P = 0.339). Plant date, as expected, also had a significant effect ( F = 8.06, P = 0.002 ). Temperature was significantly affected by shade treatments ( F = 8.80, P < 0.0001), although only temperature in the non shaded treatment was significantly greater than the shaded treatments ( Fig ure 3 3 ). Temperatures ranged from 20 to 30 C throughout typical days of the growing season, where morning and evening temperatures were within 10 to 28.5 C. Relative humidity was not significantly affected by shade treatments ( F = 2.00, P = 0.11) and ranged from 76 to 100% throughout typical days of the growing season and had an average of 93% across all treatments. UVB, UVA, and PAR values, wh en converted to a percentage of t he upper canopy energy values, were affected approximately in proportion to the percent shade ( Figure 3 4 and Table 3 1 ). In both inoculated and naturally infected plots, disease was first detected in the lower canopy and p rogressed upwards ( Figure 3 5 ). Final disease severity, recorded at growth stage R5, was significantly greater in the lower canopy of all treatments and in the middle canopy of 40 and 60% shade treatments ( Fig ure 3 6 ). In the middle canopy, the final disea se severity in the non shaded and 30% shade treatments were significantly lower from the 40 and 60% shade treatments. In the upper canopy, the final disease severity in the 40 and 60% shade treatments were significantly greater than the non shaded treatmen ts. Similar to severity results from the field, the greatest disease severity was observed on leaves collected from the lower canopy of all treatments ( Figure 3 7 ). Disease severity on leaves collected from the lower canopy was significantly greater
75 then s everities observed on leaves from the upper canopy within the same treatment. Amount of epicuticular wax from leaves taken from the upper canopy in non shaded treatments was significantly greater than all other shaded treatments at all canopy heights, with the exception of wax from the upper canopy in 30% shaded treatments ( Figure 3 8 ). Amount of epicuticular wax was not significantly different across all treatments at the lower and middle canopies and from 40 and 60% shaded treatments at the upper canopy. Only the epicuticular wax and disease severity in the upper canopy, regardless of treatment, had significant negative correlation at P = 0.009, R = 0.84. Discussion The present study examined the effects of solar radiation on the progress of SBR under fiel d and detached leaf conditions and susceptibility of detached leaves developed under different solar radiation intensities. The greater difference of disease severities over time across treatments in the early planting of 2008 as compared to the late plant ings ( Figure 3 5 ) may be attributed to increased plant height ( Figure 3 1 ) and LAI ( Figure 3 2 ) across treatments in the early planting. The increase in plant height, LAI and greater differences among treatments in the early planting are most likely due to longer day length allowing solar radiation to have a greater effect on plant physiology and disease development. Furthermore, the early planting of 2008 may have had a more uniform inoculation in the lower canopy, allowing the effect of the shade treatmen ts to be enhanced over time and across the treatments. Temperatures recorded throughout typical days of the growing season are within the range for SBR development. Morning and evening temperatures are within the range for P. pachyrhizi germination and inf ection of 10 to 28.5 C ( 109 ), and never
76 reached above 30 C which can retard disease development ( 30 ), suggesting temperature was not a limiting factor in this experiment. Temperature may have caus ed minor differences in disease development between non shaded and shaded treatments, but not to differences between shaded treatments where temperature was not significantly different. The high relative humidity recorded throughout typical days of the gro wing season was within the optimum range for SBR development and was not significantly affected by shade treatments, suggesting relative humidity was not a limiting factor for disease development. Similarly, Batchelor et al. ( 11 ) suggested moisture was a nonlimiting factor when adjusting SBR neural network models by the inclusion of a variable for cumulative days when relative humidity exceeded 90% did not improve the accuracy of the model. Intensity of solar radiation is more l ikely influencing disease development and final disease severity in the field, especially with the 60% shade treatment where the amount of UVB, UVA, and PAR was reduced to 36, 34, and 32%, respectively, of the initial intensity in the non shaded treatment ( Table 3 1 ). These reduced intensities are comparable to those in the non shaded middle canopy ( Table 3 2 ), which might be contributing to the lack of significant difference between the final disease severity of the upper canopy of the 60% shade treatment and the middle canopy of the non shaded treatment ( Figure 3 6 ). Our data suggests the greater final disease severities in the lower canopy and shaded treatments are most likely due to difference in solar radiation and its direct and indirect effects on dis ease development. Similarly, Dias et al. ( 41 ) concluded that sunlight likely affects post germination or post infection processes in SBR contributing to final severity levels
77 In addition to the direct effect of solar radiatio n on disease development in the field, its influence on soybean physiology may have been a contributing factor in the difference in disease severities observed in the detached leaf assays. The amount of epicuticular wax on soybean leaves could affect the p especially if the pathogen penetrates directly, as P. pachyrhizi does. In rice cultivars, infection rates by Rhizoctonia solani were negatively correlated with the amount of cuticular wax deposits on the outer sheath surface ( 107 ). It was found that the wax deposits interfered with the formation of appressorial and infection cushions of Rhizoctonia solani and when wax deposits were removed with chloroform, infection of previously resistant cultivars occurred ( 107 ). Previously reported, UVB radiation can affect leaf ultrastructure and anatomy, photosynthetic pigments, UVB absorbing compounds, and growth and development ( 84 ). In p articular, UVB radiation could affect development of the epicuticular wax layer, which acts as the first line of defense against biotic and abiotic factors. While there have been reports of environmental stresses affecting the epicuticular wax layer ( 84 ), only a few studies ( 60 186 ) have reported UVB effects on epicuticular waxes of crop plants. Steinmuller and Tevini ( 186 ) rep orted that enhanced UVB radiation produced 23 and 28% increase in wax content on leaf area in barley and bean, respectively. Another study with six pea genotypes differing in amount of leaf surface wax, showed an increase in wax content when the genotypes were exposed to UVB radiation ( 60 ). Additionally, UVB radiation can also alter the chemical composition of leaf surface waxes ( 186 ). Production of UV absorbing compounds such as flavonoids and othe r phenolics that function as phytoalexins or antifungal compounds have also
78 been attributed to UVB radiation ( 85 125 ). Marta et al. ( 110 ) reported that grapevine leaves maintained at 100% sun light had high polyphenolic content and significantly lower downy mildew severity compared to leaves under shading nets. While the upper canopy epicuticular wax amounts were significantly correlated to the disease severity in the det ached leaf assays more than just the amount of epicuticular wax could be contributing to disease reduction Other components of the wax, such as chemical composition, need to be analyzed to further investigate the effect of solar radiation on the epicuticu lar wax layer and its interaction with P. pachyrhizi urediniospores. Solar radiation could be affecting many aspects of soybean physiology with the amount of epicuticular wax being just one aspect. Similar results where SBR was consistently more prevalent and more severe under shaded conditions ( 36 ) and soybean plants kept in the dark up to 16 h following inoculation had greater amount of lesions even though penetration rates remained constant after 8 h of darkness ( 15 ) imply sunlight affects SBR development after infection. S olar radiation may elicit defense mechanisms affecting post infection processes. One possible defense mechanism induced by light is the production of the phytoalexin glyceollin. Greater resistance to Phytophthora megasperma f. sp. glycinea was reported in soybean seedlings incubated under light for more than 12 h and had a 30 times greater concentration of glyceollin as compared to seedlings incubated in the dark ( 208 ). More research needs to be conducted to further identify physiological and biochemical factors affected by solar radiation which might contribute to disease development. Results from field trials and detached leaf assays indicate that sol ar radiation influences progress of SBR caused by P. pachyrhizi by affecting disease development
79 across canopy heights and epicuticular wax content. Significant differences in disease severity in upper canopy heights in detached leaf assays correlated well with epicuticular wax values, with wax values decreasing and disease severity increasing as canopy height decreased. Solar radiation has a direct influence on SBR severity as observed across canopy height s and shade treatments. This study suggests that i n order to make useful epidemiological predictions in forecasting models by factoring in the influence of solar radiation on disease progress and severity of SBR effect of its UV component at given environmental conditions (temperature, relative humidity etc.) cannot be undermined.
80 Table 3 1 Energy values (W m 2 ) recorded in 0, 30, 40, and 60% shaded soybean plots Data recorded on 28 September 2009 at the North Florida Research and Educat ion Center in Quincy, Florida. Shade Treatment (%) UVB ab z UVA ac z PAR ad z W m 2 Percent of non shaded treatment W m 2 Percent of non shaded treatment W m 2 Percent of non shaded treatment 0 3.37 100 47.53 100 349.2 100 30 2.32 69 31.57 66 226 .0 65 40 1.84 55 24.66 52 173.7 50 60 1.23 36 16.22 34 111.9 32 a First columns for each UV band report raw data (W m 2 ) recorded using an Optronic OL754 spectroradiometer (Optronic Inc., Orlando, FL) measuring energy values between 270 and 775 nm. Second col umns of each UV band represent the UV flux relative to the upper canopy in the full sun (non shaded) soybeans. b UVB ultraviolet radiation at 300 to 320 nm c UVA ultraviolet radiation at 320 to 400 nm d PAR photosyntetically active radiation at 40 0 to 700 nm
81 T able 3 2. Energy values (W m 2 ) recorded in non shaded control soybean plots Data recorded on 28 September 2009 at the North Florida Research and Education Center in Quincy, Florida. Canopy Level UVB ab z UVA ac z PAR ad z W m 2 Percent of upper canopy W m 2 Percent of upper canopy W m 2 Percent of upper canopy Upper 3 .00 100 43.8 0 100 328.3 100 Middle 1.19 40 10.5 0 23 78.7 23 Lower 0.34 11 3.26 7 14.9 4 a First colum ns for each UV band report raw data (W m 2 ) recorded using an Optronic OL754 spectroradiometer ( Optronic Inc., Orlando, FL) measuring energy values between 270 to 775 nm. Second columns of each UV band represent the UV flux relative to the upper canopy. b U VB ultraviolet radiation at 300 to 320 nm c UVA ultraviolet radiation at 320 to 400 nm d PAR photosyntetically active radiation at 400 to 700 nm
82 Figure 3 1. Average plant height. Average plant height (cm) was recorded at growth stage R3 for early ( 11 April) and late (4 August) planted soybeans in 2008, under 0, 30, 40, and 60% shade treatment installed at growth stage V3 V4. Different letters on the bars indicate significant differences between early and late planting dates and among treatments base d on averages analyzed by least significant differences at P
83 Figure 3 2. Average leaf area indices Average leaf area indices were measured at reproductive growth stage R3 for early (11 April 2008) and late (4 August 2008) planted soybeans on 9 July 2008 and 3 October 2008, respectively; and at reproductive growth stage R6 for late (23 July 2009) planted soybeans on 11 October 2009 under 0, 30, 40, or 60% shade installed at growth stage V3 V4. Leaf area indices were measured using LI COR 191 quantum sensor and LI 1400 Data Logger (LI COR Biosciences; Lincoln NE). Different letters on the bars indicate significant differences between planting dates and among shade treatments based on averages analyzed by least significant differences
84 Figure 3 3. Ambient temperature and relative humidity recorded in shade treatments. Ambient temperature (A) and relative humidity (B) were recorded approximately 36 cm above the soil (mid canopy) within soybean rows e very 20 minutes in soybeans grown under 0, 30, 40, and 60% shade installed at vegetative growth stage (V3 V4). B A
85 Figure 3 4. Comparative ult raviolet visible solar spectrum recorded under shade treatments. Comparative ultraviolet visible solar spectrum (270 to775 nm) of overhead sun at 2:00 PM EDT was measured under 0, 30, 40, and 60% shade plot treatments on 28 September 2009. The spectrum was obtained using an Optronic OL754 spectroradiometer (Optronic Inc., Orlando, FL)
86 Fig ure 3 5. Severity of soybean rust (SBR) from field plots over time SBR was rated weekly under different shade treatments (0, 30, 40, and 60% shade), canopy heights (lower, middle, and upper), and planting dates. X axis displays date and growth stage of soybean at each evaluation. S hade cloth structures were installed at vegetative growth stage V3 V4. Severity of SBR was based on visual assessment of symptomatic leaf area using the scale, 0= no disease, 1 = up to 2.5%, 2 = 2.5 5%, 3 = 5 10%, 4 = 10 15%, 5 = 15 25%, 6 = 25 35%, 7 = 35 67.5%, and 8 = 67.5 100% of the leaf area affected by SBR and averages were calculated by converting each rating to the midpoint of its range.
87 Figure 3 6. Final soybean rust (SBR) disease severity among shade treatments. Average SBR severity at reprodu ctive growth stage R5 (seed development) at lower, middle, and upper soybean canopies under 0, 30, 40, and 60% shade treatments in the field were calculated to determine final SBR severity Shade cloth structures were installed at vegetative growth stage V 3 V4. Disease severity was assessed at ten randomly pre assigned areas in each plot by visual assessment of symptomatic leaf area using the scale, 0= no disease, 1 = up to 2.5%, 2 = 2.5 5%, 3 = 5 10%, 4 = 10 15%, 5 = 15 25%, 6 = 25 35%, 7 = 35 67.5%, and 8 = 67.5 100% of the leaf area affected by SBR and averages were calculated by converting each rating to the midpoint of its range. Different letters on the bars indicate significant differences among soybean canopy heights and shade treatments based on ave rages analyzed by least significant differences at P heights.
88 Figure 3 7 Average soybean rust (SBR) severity on detached soybean leaves The a verage SBR severity on detached soybean leaves taken from lower, middle, an d upper canopies under 0, 30, 40, and 60% shade treatments which were installed at vegetative growth stage (V3 V4). Leaves were detached at reproductive growth stage R1, and inoculated with Phakopsora pachyrhizi urediniospore suspension containing approxi mately 15,000 spores/ml. Disease severity was determined based on visual assessment of symptomatic leaf area using the scale, 0= no disease, 1 = up to 2.5%, 2 = 2.5 5%, 3 = 5 10%, 4 = 10 15%, 5 = 15 25%, 6 = 25 35%, 7 = 35 67.5%, and 8 = 67.5 100% of the l eaf area with soybean rust. Averages severity was calculated by converting each rating to the midpoint of its range. Different letters on the bars indicate significant differences among soybean canopy heights and shade treatments based on averages analyzed by least significant differences at P
89 Figure 3 8. Average epicuticular wax (mg/cm 2 ) extracted from soybean leaves Average epicuticular wax (mg/cm 2 ) was extracted from soybean leaves at reproductive growth stage, R1, from lower, middle, and upper canopies under 0, 30, 40, and 60% shade treatments. Different letters on the bars indicate significant differences among soybean canopy heights and shade treatments based on averages analyzed by least significant differences at P treatments a nd canopy heights.
90 CHAPTER 4 VARIATION IN THE NUC L E OTIDE BINDING SITE ( NBS) REGIONS OF KUDZ U ( PUERARIA MONTANA ): POSSIBLE SOURCE O F RESISTANCE TO PHAK OPSORA PACHYRHIZI, THE CAUS AL AGENT OF SOBYEAN RUST Introduction Soybean rust (SBR), caused by the biotrop hic fungus Phakopsora pachyrhizi is potentially destructive to soybean ( Glycine max ) production worldwide. In the U.S., the fungus over winters on exotic, invasive kudzu species, predominantly on the species Pueraria montana (Lour.) Merr. var. lobata (Wi lld.) Maesen & Almeida. ( 83 209 ) in the southeastern states Urediniospores originating from infection on kudzu provide the initial inoculum during e very soybean growing season and has the potentia l to disperse to northern soybean producing areas ( 79 80 83 ) Kudzu was first promoted in the U.S. in 1907 as a drought resistant, high nitrogen forage crop. By the 192 0 30s, kudzu was propagated and promoted by the Soil Conservation Service as a means of holding soil on eroding gullies of the deforested southern landscape. Kudzu seedlings and nurseries produced and distributed more than 73 million seedlings between 193 6 and 1941 ( 23 ). Kudzu is now estimated to cover between 0.81 and 2.8 million hectares of the U.S. primarily in the southeast. Eventually, farmers found drawbacks to kudzu including difficulty baling kudzu hay and the tenden cy for it to grow so rapidly and extensively that it covered all other vegetation. In 1998 Congress placed kudzu on the federal list of noxious weeds. Kudzu exhibits similar susceptible, resistant, and immune response reactions to P. pachyrhizi as soybean with susceptible responses characterized by tan lesions with profuse sporulation, resistant responses characterized by reddish brown lesions with delayed, reduced sporulation, and immune responses characterized by no lesion
91 development ( 82 ). Of 10 kudzu accessions obtained from north central Florida and tested by Jordan et al. ( 82 ), 6 were identified as susceptible, 3 immune, and 1 resistant to P. pachyrhizi The resistant kudzu had 10 fold less P. pachyrhizi DNA present 15 days after inoculation compared with susceptible kudzu quantified by quantitative real time polymerase chain reaction ( 82 ). In the U.S, a resistant reaction, characterized by reddish brown lesions, was common on kudzu plants collected from areas of Mississippi, Louisiana, Kentucky, and North Carolina when challenged with three isolates of P. pachyrhizi one each from the U.S states of Alabama (AL) and Louisiana (LA), and one from Brazil ( 16 ). It was further reported that plants at specific sites were genetically identical with respect to their reaction to urediniospores from a wide range of P. pachyrhizi isolates when tested with 8 additional isolates from several areas o f the world. Of the kudzu populations tested from the southeastern U.S., 46% produced a resistant response to the three P. pachyrhizi isolates from AL, LA and Brazil ( 16 ). In kudzu populations from north Florida, 32 out of 13 9 tested had reduced sporulation and red brown lesions characteristic of resistant reactions to P. pachyrhizi ( 82 ). Data from the Integrated Pest Management Pest Information Platform for Extension and Education (ipm PIPE) co llected from 2005 through 2010 provide evidence of further resistance to P. pachyrhizi in k udzu populations throughout Florida ( 80 ) Current SBR prediction models for the U.S. are based upon initial inoculum from kudzu and assu me all kudzu populations to be susceptible ( 79 80 ). A more precise understanding of resistance in kudzu may provide better estimates of initial inoculum for forecast ing models, shed light on probab le evolutionary mechanisms of host parasite
92 interactions of SBR, and lead to identifying a potential source of resistance that could be utilized against SBR in soybean. To date, five major sources of SBR resistance termed resistant to Phakopsora pachyrhi zi ( Rpp ), ( Rpp1 Rpp 2 Rpp 3 Rpp 4 and Rpp5 ) h ave been identified in soybean ( 21 22 54 68 69 112 ) A resistance candidate gene, Rpp4C belonging to the coiled coiled nucleotide binding site leucine rich repeat (CC NBS LRR) gene family has been associated with resistance conferred by Rpp4 ( 117 ) Resistant g enes with regions encoding products that contain nucleotide binding site (NBS) and a series of leucine rich repeats (LRRs) have been shown to mediate a hypersensitive reaction (HR) to defend against pathogen attack ( 35 42 43 126 226 ). The NBS region is responsible for the binding and hydrolysis of ATP and GTP and acts as a signaling molecule ( 190 ), whereas the LRR encodes for a potential binding surface for protein protein interactions ( 89 ). It has been confirmed that the majority of 149 NBS LRR encoding genes in Arabidopsis ( Arabidopsis thaliana ) are clust ered physically in the genome, although a cluster can contain highly dissimilar NBS LRR encoding genes ( 115 116 156 ). Similarly, NBS LRR encoding gene clusters have bee n identified in many other species including rice ( Oryza sativa ; 535 NBS LRR encoding genes; ( 229 )) and common bean ( Medicago truncatula ; 333 NBS LRR encoding genes; ( 5 )). Yet, a BLASTX comparison of the soybean genome scaffold with Rpp4 candidate sequences did not identify any additional NBS LRR encoding genes in soybean ( 4 117 ). Likewise, an evaluation of the entire U.S. germplasm collecti on (16,000 accessions) against a mixture of five P. pachyrhizi isolates identified less than 5% (805 accessions) with partial tolerance or resistance reactions ( 119 ). The rarity of resistance to SBR in
93 soybean suggests other s ources should be investigated for resistance to P. pachyrhizi such as wild legumes, including kudzu. Previously identified SBR resistance in kudzu ( 82 ) may be associated with the NBS LRR gene family which exhibits a resistanc e reaction similar to Rpp4 mediated resistance in soybean when challenged with a compatible P. pachyrhizi isolate. There is a need for improved understanding of the degree of resistance in kudzu to P. pachyrhizi as well as the underlying variables to whic h such resistance can be attributed. The objectives of this study were to identify and characterize probable resistance conferring NBS regions and examine the genetic diversity of NBS regions of kudzu populations in Florida by using restriction fragment l ength polymorphisms (RFLP). Materials and Methods Kudzu Populations for Initial Screening of NBS Region Individual kudzu plants representing 5 kudzu populations, FLAL15, FLAL13, FLMA01, FLAL11, located in north central Florida and FLMD12 located in north Florida were collected in the spring of 2007 and in the summer of 2010, respectively. Kudzu plants were collected by digging up a healthy, asymptomatic root and vine from each location, planted in 5 gal plastic pots containing Metro Mix 300 potting soil ( Sun Gro Horticultural Distributors, Inc., Bellevue, WA) and maintained in a rust free greenhouse without shade until spring of 2011, at the University of Florida campus located in Gainesville, Florida. In the spring of 2011 the potted kudzu plants were rel ocated to rust free greenhouse at the North Florida Research and Education Center campus located in Quincy, Florida. Kudzu population FLAL15 was previously reported as resistant, FLAL13 and FLMA01 as susceptible, and FLAL11 as immune to the P.
94 pachyrhizi isolate (FL05) collected from infected soybean in the summer of 2005 from the Plant Science Research and Education Center near Citra, FL ( 82 ). Kudzu population FLMD12 had not previously been characterized with respect to its r eaction to P. pachyrhizi P. pachyrhizi Inoculum The isolate of P. pachyrhizi used in this study was collected from infected kudzu (population FLGA09) in the summer of 20 10 from the North Florida Research and Education Center in Quincy F lorida Th is isola te was maintained on susceptible maturity group 5 soybean (Pioneer 95Y20 RR ) in a greenhouse on the North Florida Research and Education Center campus in Quincy, Florida. Newly produced urediniospores were collected by vacuum cyclone separation ( G R manufac turing Co., Manhattan, KS USA) from the abaxial side of infected soybean leaves that were washed the previous day to remove old urediniospores. Urediniospores were suspended in a solution of 0.2% (vol/vol) Tween 20 in sterile distilled water and gently m ixed for 20 to 30 s using a vortexer. Final urediniospore concentration was adjusted to 10 5 spores per ml using a hemacytometer. DNA extraction, polymerase chain reaction (PCR), and partial sequencing of NBS region Immature leaflets approximately 0.6 to 2.6 cm in length and 0.25 to 2.0 cm in width were collected from individuals from each kudzu population and pulverized in mortars with liquid nitrogen. Total genomic DNA was extracted using the DNeasy Plant Mini kit ( QIAGEN, Valencia CA, USA). Primers de signed in the conserved region of the nucleotide binding site (NBS) domain of the resistant gene, Rpp4 of soybean ( 117 ) were used to amplify NBS domain(s) in kudzu populations (Table 1). The PCR mixture
95 contained 12.5 l of Ex tract N Amp TM (Sigma Aldrich, Inc., St. Louis, MO, USA), 1.25 l each of the forward and reverse primers 5 l of deionized (DI) water, and 5 l of template DNA (approximately 250 ng) The PCR program used for amplification was as follows: initial dena turation at 94 C for 1 min; 34 cycles of 94 C for 30 sec, annealing at 55 C for 30 sec, elongation at 72 C for 1.4 min, and final extension at 72 C for 7 min. PCR products were purified using the MiniElute PCR Purification kit ( QIAGEN, Valencia CA ), and cloned into Escherichia coli JM109 High Efficiency Competent cells ( Promega, Fitchburg WI, USA ) using pGEM T Easy vector system II Kit ( Promega, Fitchburg WI, USA ) A total of 19 clone s were sequenced (I CBR, University of Florida, Gainesville, FL ) Sequences were a lign ed using Clustal W ( 194 ) and phylogenetic trees were generated using the software Mega5 ( 191 ). Sampling of Kudzu Populations across Florida Kudzu populations i n Florida were identified and monitored for the presence of P. pachyrhizi beginning in 2005, with additional populations added in subsequent years. The latitude and longitude coordinates and the approximate area of each kudzu population were recorded. Kudzu populations were separated geographically by at least 500 m or by physical barriers, such as a paved roadway. Multiple samples were taken from 4 kudzu populations (FLAL15, FLJef12, FLMA01, and FLMD12), at least 3 m apart and screened separately to examine variation within a kudzu population. For RFLP analysis, kudzu populations were sampled from 31 counties, across the 6 regions of Florida (Figure 4 1). Kudzu samples were collected from as far south as Miami Dade county (15.50805, 80.30267 decimal degrees (DD) ) a s far northeast as Duval county
96 (30.29280, 81.61590 DD) and as far northwest as Escambia county (30.99080, 87.57790 DD) Using data collected from 2005 through 2011, 179 kudzu populations out of 266 were infected 3 of the years scouted or less and wer e identified as possibly resistant to P. pachyrhizi Of the 179 possibly resistant populations identified 1 05 populations were sampled with an additional 8 susceptible kudzu populations that were infected 4 years or more. Immature leaflets of kudzu, as d escribed previously, were collected from the 113 populations during 2011. Restriction Fragment Length Polymorphism (RFLP) Analysis DNA extraction and PCR amplification of the NBS region from all sampled kudzu populations was performed as described previous ly. Using aligned sequences and NEBcutter V 2.0 ( http://tools.neb.com/NEBcutter2/ ) the enzyme Taq I was identified and utilized to further differentiate between kudzu populations. A RFLP analysis of the amplified NBS region of the kudzu populations was performed. The restriction enzyme, Taq I (New England Biolabs, Inc.,Ipswich, MA; 0.3 l containing 30 units) was incubated with 10 l of purified PCR product for approximately 2 h at 65 C following the gel (2% agarose in 1X TBE buffer containing 0.002% ethidium bromide) fo r approximately 1.5 h at 50 V. Gels were photographed on a UV transilluminator. RFLP band profiles were analyzed using the software package GelCompar II (version 6.5; Applied Maths, Kortrijk, Belgium). Molecular weights were assigned to each fragment us
97 unweighted pair group method with arithmetic means (UPGMA) phylogenetic tree was generated using D ICE similarity coefficients to determine genetic similarities among the kudzu populations sampled. To infer population genetic structure and differentiation an analysis of molecular variance (AMOVA; ( 50 )) was done by using the program ARLEQUIN (version 3.1; ( 49 )). Populations were assigned to 6 groups on the basis of the region they are located (Northwest, Northeast, Central, South Central, and South) (Figure 4 1 and Table 4 2) to assess the effect of geography on genetic variation. Another grouping based on number of years of previous SBR infections (obtained from historical records) grouped the sampled populations into 7 groups: 47 populations never observed to be infected, 37 populations infected 1 year, 21 populations infected 2 years, 3 populations infected 3 years, 2 populations infected 4 years, 1 population infected 5 years, and 2 populations infected 6 years. In the AMOVA, banding patterns are treated as molecular haplotypes of NBS regions. A distance matrix based on pairwise squared Euclidean distances between the haplotypes was generated, based on which, the total observed variance was partitioned into covariance components at different levels of genetic structure: among populations, among groups of populations, and among populations within groups, to generate the F statistics ( F ST F CT or F SC ). The measure of F ST is the proportion of the total observed variance that resides among all sampled populations, whereas F CT represents the propor tion of the total variance that resides among the groups of populations. The statistic F SC estimates the proportion of total variance in a group that resides among populations in that group ( 49 50 56 ). The F values can range from 0 to 1. The P values
98 obtained are the proportion of 1000 permutations that gave an F value larger than the observed value and represent the probability of the F statistic being measured. Detac hed Leaf Assay Detached leaf assays were conducted using kudzu plants, representative of populations FLAL15, FLMA01, FLAL11, and FLMD12, to verify resistance in FLMD12 as identified by RFLP. Susceptible soybean variety (Pioneer 95Y20 RR ) and resistant soyb ean, PI 459025B the original source of the Rpp4 gene, were also used in the detached leaf assay. Two or four leaflets were excised from each kudzu plant and soybean variety. Fully expanded leaflets were collected from the third to sixth node of kudzu and fully expanded leaflets from the upper most nodes of soybean at vegetative growth stage V8 (last vegetative stage before flowering) to reproductive growth stage R3 (pod development) and used in the assays. The detached leaflets were rinsed with sterile distilled water and allowed to air dry in the laboratory The petiole of each leaflet was cut diagonally with a scalpel to produce a clean, fresh surface. Each leaflet was placed in an individual Petri plate (150 x 15 mm) half filled with 1% water agar w ith the petiole inserted into the agar. Within each Petri plate a sterile, moist paper towel (8 x 8 cm) was placed to increase moisture. Each leaflet was inoculated on its adaxial side in 9 to 10 defined areas with 5 l of spore suspension (10 5 spores/ml ) (prepared as described previously). After inoculation Petri plates were placed in the dark in a moist chamber for approximately 48 h. After removing Petri plates from the moist chamber they were parafilmed and maintained at room temperature (approximat ely 24 C) and exposed to approximately 30% intensity of natural sunlight for 8 hours per day for the entire duration of the assay.
99 At 10, 15, and 20 days post inoculation (DPI), leaf discs, (approximately 8 mm in diameter) were removed from the detached le aflets at the inoculated sites using a cork borer, and inspected for number of sporulating uredinia. The leaf discs were then placed individually in 1.5 ml microcentrifuge tubes and 100 l of 0.2% (vol/vol) Tween 20 in sterile distilled water was added to each tube, followed by mixing for 30 s using a vortexer. Spore concentration was determined for each leaf disc using a hemacytometer. The number of spores produced at each inoculated site was determined using the spore concentration and collection volum e. At each evaluation date (10, 15, and 20 DPI), 3 inoculated sites were harvested from each leaflet, giving a total of 6 or 12 sites for each plant. The assay was carried out three times (once with 2 leaflets of each plant and twice with 4 leaflets of e ach plant) for a total of 30 inoculation sites for each plant at each time point of evaluation. Statistics were applied to the data a significance level of 0.05 in SAS 9.1 (Statistical Analysis Software version 9.1, SAS Institute Inc., Cary, NC). Results Primers targeting the conserved NBS domain of the soybean Rpp4C resistance gene were used to amplify similar regions from kudzu populations Amplicons from kudzu po pulations were the same in size as that from soybean (1.2 kb using Rpp4 _NB_F/R, Table 4 1). Sequencing of 19 clones of the NBS domain from kudzu populations FLMA01, FLAL13, FLAL11, and FLAL15 revealed motifs identifying the regions as NBS domains. Clones from the different kudzu populations had high sequence similarity (81 to 100%) with each other and a neighbor joining phylogenetic
100 tree delineated multiple clusters (Figure 4 2). However, clones representing a given kudzu population did not always cluste r together implying the primers were amplifying multiple NBS regions from kudzu. A GenBank search, performed with the BLASTX algorithm showed homology of the kudzu sequences with Rpp4 candidate genes from soybean and Phaseolus vulgaris respectively shari ng from 80 to 86% and 73 to 77% protein identity. Individuals that were at least 3 m apart within a kudzu population, did not display variation in the RFLP banding pattern, indicating that individuals within populations examined in this study are clonal Compared to the known susceptible kudzu populations, FLMA01 and FLAL13, the known resistant kudzu population, FLAL15, had a distinct banding pattern when digested with the restriction enzyme, Taq I (Figure 4 3). Out of the 113 kudzu populations screened, 1 population (FLMD12) had the same banding pattern as the known resistant kudzu population FLAL15. Results from the detached leaf assays further indicated kudzu population FLMD12 is resistant to P. pachyrhizi The number of uredinia and urediniospores produced per inoculated area on FLAL15 and FLMD12 in the detached leaf assay were not significantly different (Figure 4 4). Whereas, the number of uredinia and urediniospores produced per inoculat ed area were significantly lower on susceptible and resistant kudzu compared to susceptible and resistant soybean. Furthermore, uredinia did not develop on resistant kudzu populations (FLAL15 and FLMD12) until 15 DPI and majority of sporulation did not oc cur until 20 DPI, and no uredinia developed on the immune kudzu (FLAL11) (data not shown).
101 A UPGMA dendrogram that was generated based on RFLP of the NBS domain identified grouped the sampled kudzu populations into 28 distinct clusters (Figure 4 5 and Tabl e 4 3). Although kudzu populations did not always cluster based on infection history, some clusters with high cophenetic correlation and average percent similarity may indicate those populations have a genetic pre disposition of resistance or susceptibili ty to P. pachyrhizi Specifically, the identified resistant kudzu populations FLAL15 and FLMD12 clustered together with population FLMD13, which has never been found to be infected, with a cophenetic correlation coefficient of 95 and average similarity of 81.4% among those populations (cluster 18 in Table 4 4 and Figure 4 5). Cluster 22 and 24 groups kudzu populations that have never been observed to be infected with cophenetic correlation coefficients of 82 and 64, respectively; and average similarity of respectively, 66.5 and 85.8%, among those populations (Table 4 4; Figure 4 5). Whereas, cluster 11 and 19 groups kudzu populations that have been observed to be infected every year that they were scouted along with populations that have never been obser ved to be infected or infected 1 to 3 years, with cophenetic correlation coefficients of 92 and 88, respectively, and average similarity of, respectively, 92 and 88%, among those populations. Similarly, only 6 clusters group kudzu populations that are in the same geographic region in Florida and of those 6 clusters only 1 (cluster 18) has a high cophenetic correlation coefficient and average similarity among the populations. Population structure determined by AMOVA showed almost identical results when grou ped by geographic region and by the number of years a population was observed to be infected with SBR (Table 4 4). All the variation was attributed to variation among
102 populations and among populations within groups. The proportion of total variance resid ing among populations within groups ( F SC ) was 1.000 ( P = 0.000) and 1.000 ( P = 0.000) w hen populations were grouped by geographic location or number of years of previous infection, respectively. Discussion Based on the results, primers designed in conserve d regions of the NBS domain from soybean can be used to amplify similar regions from kudzu populations in Florida. While sequencing of the NBS region from 4 kudzu populations had 80 to 86% sequence identity to Rpp4 candidate genes in soybean multiple poly morphisms, evidenced by multiple clusters on the phylogenetic tree, indicate that the conserved primers were amplifying multiple NBS regions within a single kudzu population. The RFLP results further provide evidence of great variation in the NBS region i n kudzu which may be due to variability within a single NBS region and/or variability among NBS regions in kudzu populations. This is expected, as other studies have identified multiple NBS genes within an organism using primers designed in conserved regio ns of the NBS ( 124 ). The sequence analysis and phylogenetic tree imply greater genetic diversity within kudzu compared to soybean, with respect to the NBS regions. Bonde et al. ( 18 ) reported that k udzu plants previously shown to produce resistant reaction to 3 P. pachyrhizi isolates also produced resistant reactions to an additional 8 isolates from around the world. This broad resistance displayed by kudzu against multiple isolates of P. pachyrhizi may be due to multiple resistance genes, indicated by multiple NBS regions identified in this study. Furthermore, the high variation of the NBS region(s) in both susceptible and resistant kudzu populations may explain the reduction in uredinia and uredin iospore production on all kudzu populations examined compared to soybean.
103 Geographic location and infection history of the sampled kudzu populations does not appear to influence the genetic structure of these populations. The low F CT values from an AMOVA indicate that populations within the same region of Florida and those with similar infection history from 2005 through 2011 are genetically distant from each other as they are from populations from different regions of Florida or from populations that diff er in the number of years they have been observe to be infected with SBR. Most of the sampled kudzu populations did not cluster according to their infection history, although clusters 18, 22, and 24 were clusters grouping only those kudzu populations that were never observed to be infected with SBR These clusters may represent NBS region(s) that contribute to resistance to P. pachyrhizi with reduction in uredinia and reduction and delay in sporulation. Other clusters that grouped kudzu populations that h ad a range in infection history from never being observed to be infected to being infected every year may be representing NBS region(s) that likely do not contribute to resistance. Furthermore, the lack of any pattern based on infection history in these c lusters could be due to lack of inoculum or conducive environmental conditions for SBR infection and disease development. Further investigation and exhaustive sampling of susceptible kudzu populations may better resolve polymorphisms in the NBS region con tributing to resistance to P. pachyrhizi Consistent with the findings from this research, Pappert et al. ( 144 ) found southeastern U.S. kudzu populations to have high genetic diversity with no regional patterns of variation and relatively little overlap of genotypes between populations when examining 14 allozyme loci. Furthermore, kudzu populations from the southeastern U.S. have been reported to have similar diversity index to the native Chinese kudzu
104 populations from examinat ion of inter simple sequence repeat (ISSR) variations ( 189 ). Such high levels of genetic diversity among populations and low diversity within kudzu populations can be explained by limitation of gene flow among populations and high genotypic diversity. With growth in kudzu populations being primarily clonal by rhizomes, low levels of seed set and seed viability and restriction of gene flow via seed dispersal could lead to genetic differentiation among populations ( 1 138 ). When kudzu populations were initially becoming established, pollination and seedling establishment may have contributed to genotypic variation within a population. But once a population became establish ed with a dense structure of growth and rhizomes, only a very small proportion of seedlings may become established. This was reported by Abramovitz ( 1 ), where out of 245 naturally occurring seedlings observed at the periphery of populations only one became established. The high level of genetic variability among kudzu populations in the southeastern U.S. can be primarily attributed to establishment by introductions from different sources and gene exchange and recombination dur ing initial establishment. Variability in the NBS region(s) in kudzu may partially explain why kudzu, both susceptible and resistant populations, display greater resistance to P. pahcyrhizi with reduction in uredinia and reduction and delay in sporulation as compared to soybean. Meyers et al. ( 117 ) found that amino acid differences among the Rpp4 candidate genes may play a key role in resistance. Resistance to rice blast and flax rust, conferred by LRR NBS gene(s), has been re versed by changing as few as 1 and 6 amino acids, respectively ( 25 44 ). This shows that even small variations in the NBS region could significantly alter susceptibility or resistance in a kudzu pop ulation which perhaps is the
105 basis of a lack of correlation between apparent high similarity among RFLP of the NBS region(s) and resistance or susceptibility to SBR observed in this study. While the NBS marker used in this study did have high variation amo ng kudzu populations a more robust marker may better differentiate between susceptible and resistant kudzu populations or possibly a more extensive digest of the NBS region(s) using a combination of enzymes may identify more polymorphisms contributing to r esistance. Furthermore, there may be other resistance genes, non LRR NBS genes, contributing to the observed reduction in uredinia and urediniospore production in kudzu populations and multiple defense mechanisms. Different defense mechanisms in kudzu po pulations were reported by Jordan et al. ( 82 ). Where P. pachyrhizi infection of susceptible kudzu only differed from susceptible soybean 120 h after infection by limiting colonization of the mesophyll, infection of resistant k udzu displayed a hypersensitive response (HR) between 24 and 48 h after inoculation, and infection of immune kudzu displayed an early onset of an HR with a cell wall deposition around the penetration hyphae ( 82 ). These defense mechanisms may be mediated by different genes hence other markers than the NBS region(s) used in this study should be used to further investigate potential sources of resistance to P. pachyrhizi in kudzu. Based on the results, using the number of previous years a kudzu population has been observed to be infected may be a poor parameter to estimate resistance in kudzu populations, as this may only be a result of low disease pressure or unfavorable conditions in the area for SBR development. However, the re duction in disease development on susceptible, resistant, and immune kudzu observed in this study and patterns in infection history could be used to better estimate initial inoculum from kudzu
106 populations which begin SBR epidemics each year in the U.S. Fu rthermore, as SBR severity has been variable since its introduction into the continental U.S., continued observation and research on kudzu populations will provide valuable information for improved forecasting of SBR movement and development, and estimatin g the potential risk SBR poses to U.S. soybean production. This research indicates that kudzu is a good source of resistance to SBR not only due to the genetic variation among kudzu populations but also due to an overall reduction in severity of disease d evelopment on both susceptible and resistant kudzu as compared to soybean.
107 Table 4 1. Primers used to amplify the nucleotide binding site ( NBS ) domain in kudzu populations The primers were designed in the conserved NBS domain of the resistant gene, Rpp 4 of soybean ( 117 ). Primer Name Sequence Rpp4_ NB_F CCTTATAATCCTAGATGATC Rpp4 _NB_R CGTAATTTCTTTAACTCTCC
108 Table 4 2. Distribution of kudzu populations in Florida. Region Kudzu Populations Possible Resistant Populations a Sampled Kudzu P opluations b Northwest 160 88 63 (67%) Northeast 52 41 30 (63%) Central 24 20 11 (55%) South Central 11 11 7 (64%) South 19 19 2 (11%) a Kudzu populations were identified as possibly resistant to Phakopsora pachyrhizi if f rom observational data from 2005 through 2011 a population was infected 3 years or less b Number and percent (in parentheses) of kudzu populations sampled out of the kudzu populations identified as possible resistant population
109 Table 4 3. Average perc ent similarity among clusters of kudzu populations in regard to the nucleotide binding site (NBS) region(s). Average percent similarity with standard deviation calculated based on restriction fragment length polymorphisms (RFLP) of NBS region(s), among ku dzu populations within a cluster in an unweighted pair group method with arithmetic means (UPGMA) phylogenetic tree (Figure 4 4), generated using DICE similarity coefficients. Cluster a Average percent s imilarity with Std. Dev. b Cophenetic correlation coeff icient c 1 58.0 +/ 0.0 66 2 64.3 +/ 5.8 67 3 74.6 +/ 5.6 73 4 79.1 +/ 2.2 81 5 73.2 +/ 1.9 84 6 72 7 69.8 +/ 2.8 62 8 76.9 +/ 1.8 89 9 85.0 +/ 0.0 75 10 84.1 +/ 2.5 78 11 84.9 +/ 1.5 92 12 79.2 +/ 3.4 76 13 77.0 +/ 3.3 78 14 78.0 +/ 1.7 74 15 75.2 +/ 0.0 62 16 68.2 +/ 0.76 97 17 77.1 +/ 1.6 85 18 81.4 +/ 1.2 95 19 84.9 +/ 0.0 88 20 76.3 +/ 0.0 79 21 81 22 66.5 +/ 0.0 82 23 82 24 85.8 +/ 0.0 64 25 61.7 +/ 6.1 60 26 68.4 +/ 1.7 87 27 58.4 +/ 5.3 86 28 70 a Kudzu population clusters generated by UPGMA tree using RFLP of the NBS region ( s ) b Average percent similarity and standard deviation for clusters with 2 or more populations. c Cophenetic correlation coefficient computed by Gel Compar II (version 6. 5; Applied Maths, Kortrijk, Belgium)
110 Table 4 4. Analysis of molecular variance (AMOVA) of restriction fragment length polymorphism (RFLP) data ARLEQUIN 3.1 was used to determine the genetic structure among populations of kudzu ( Pueraria montana ) sampled in 2011 by AMOVA of RFLP data The populations were grouped by geographic region: Northwest (NW), Northeast (NE), Central (C), South Central (SC), and South (S); and by the number of years a population was observed to be infected with Phakopsora pachyrhi zi from 2005 through 2011 (0, 1, 2, 3, 4, 5, or 6). Computations for AMOVA are based on a matrix of pairwise squared Euclidean distances between populations. Percent variation is the distribution of variation at a given level of hierarchy (among groups/a mong populations within groups/among populations). The F statistics are a measure of genetic structure at different levels of hierarchy with values ranging from 0 to 1. The P value is the proportion of 1000 permutations that gave an F statistic value the observed value. Grouping of populations Source of variation Variation (%) F statistic a P By geographic region Five groups: NW, NE, C, SC, and C Among groups 1 60 F CT = 0 .0 16 0. 060 Among populations within groups 9 8 40 F SC = 1 000 0.000 Among populations 100 .00 F ST = 1 .000 0.000 By number of years infected Seven groups: 0, 1, 2, 3, 4, 5, and 6 Among groups 0. 00 F CT = 0.00 0 0.6 90 Among populations within groups 100. 00 F SC = 1 0 00 0.000 Among populations 10 0.00 F ST = 1 000 0.000 a F CT = proportion of total variance residing in groups; F SC = proportion of total variance in a group residing among populations in that group; F ST = proportion of total variance residing among populations
111 Figure 4 1. Distribution of kudzu populations in Florida. The percentage (and number) of possibly resistant kudzu populations that were sampled from each region is listed. Populations were identified as possibly resistant based on being infected with s oybean rust 3 years or less from 2005 through 2011.
112 Figure 4 2.Neighbor joining phylogenetic tree The tree and bootstrap values are generated by the maximum composite likelihood method using the nucleotide binding site (NBS) sequences of 19 clones fro m 4 kudzu populations (FLAL11,FL AL13,FL MA01, and FLAL15) and rooted with the sequence of soybean PI 459025B, the original source of Rpp4
113 Figure 4 3. Restriction fragment length polymorphisms (RFLP) of kudzu populations susceptible, resistant, and im mune to Phakopsora pachyrhizi Lane 1: 100 bp ladder, lane 2: FLAL13 known susceptible, lane 3: FLMA01 known susceptible, lane 4: FLAL15 known resistant, lane 5: FLMD12 not previously identified as susceptible, resistant, or immune, and lane 6: FLAL11 kno wn immune.
114 Figure 4 4. Average number of uredinia and urediniospores produced per inoculation site on detached leaflets from different hosts. Hosts used included: susceptible soybean (Pioneer 95Y20RR), resistant soybean (PI 459025B), susceptible kudzu p opulation (FLMA01), resistant kudzu population (FLAL15), and kudzu population (FLMD12) identified as possibly resistant. Fully expanded leaflets were detached from the third to sixth node of kudzu and from the upper most nodes of soybean at vegetative gro wth stage V8 to reproductive growth stage R3. Leaflets were inoculated at 9 to 10 defined areas on the adaxial side with approximately 500 spores at each area. Three inoculated areas were evaluated 10, 15, and 20 days post inoculation (DPI) for number of uredinia and urediniospores produced per inoculated area. Different letters on the bars indicate significant differences among DPI and hosts based on averages analyzed by least significant differences at P 0.05, across DPI and hosts.
115 Figure 4 5 Unweighted pair group method using arithmetic mean (UPGMA) dendrogram A, UPGMA dendrogram generated with Gel Compar II (version 6.5; Applied Maths, Kortrijk, Belgium), based on RFLPs of NBS region ( s ) in the sampled kudzu populations in Florida. Boxes indicate areas enlarged in B. B, clusters with high cophenetic correlation coefficients and average percent similarities. a ID color and number identification for clusters, b Kudzu population name, c County k udzu population is located. d Region of Florida kudzu population is located: Northwest (NW), Northeast (NE), Central (C), South Central (SC), and South (S) e The number of years the kudzu population has been observed to be infected with Phakopsora pachyrh izi from 2005 through 2011 A B
116 CHAPTER 5 EPIDEMIOLOGY OF SOYB EAN RUST ON SOYBEAN AND SUSCEPTIBLE AND RESISTANT KUDZU AND DECISION MODEL FOR F UNGICIDE APPLICATION S FOR SOYBEAN RUST IN FLORIDA Introduction S oybean rust (SBR), caused by the obligate fungal par asite Phakopsora pachyrhizi is a potentially devastating disease of soybean in the U.S. Each year initial inoculum is produced on the alternative host kudzu ( Pueraria lobata ) in the southeastern U.S. Kudzu has been estimated to cover 800,000 to 3,000,00 0 hectares in the southeastern U.S. with most severe infestations occurring in the piedmont regions of Mississippi, Alabama, and Georgia ( 48 34 23 ). Based on surveys of kudzu in Florida from 2007 through 2011, it is estimated Florida has approximately 201,000 hectares of kudzu, predominately in the north central to northeast and northwest regions (K. personal communication ). Epidemiological models for SBR have been developed that estimate disease development considering biological processes in the disease cycle and local inoculum availability ( 39 ). For example, Pivonia and Yang ( 150 ) calculated the numbe r of infections and rate of disease increase amongst five rust pathogens, including P. pachyrhizi by adjusting a general disease model ( 64 ) and climatology data for specific locations within the U.S. The models for the U.S. a re based upon the initial inoculum originating in kudzu and assume all kudzu populations to be susceptible to SBR ( 79 ). While multiple risk assessment and risk prediction models have been developed for SBR, uncertainties in in oculum availability weaken their accuracy ( 150 ). Management of SBR epidemics in soybean in the southeast is necessary not only to avoid yield loss, but to also reduce inoculum that can potentially infect other soybean
117 producing areas of the country. Fungicides can be applied, but are costly. T o be efficient and effective at managing disease, growers need to know when application s will be most beneficial The effectiveness of different fungicides and timing of applications for SBR have often depended on when SBR was first detected, the intensity of its development, recent weather conditions, and proximity to sources of SBR ( 131 174 ). To assess proximity to sources of SB R the Integrated Pest Management Pest Information Platform for Extension and Education (Ipm PIPE, www.sbrusa.net ) can be used to view reports of SBR in the U.S. on soybean and alternative hosts. SBR development has been predicted using simulation and empirical models based on the effects of environmental variables on epidemiological parameters such as infection rate, latent period, and senescence of disease lesions. Data from Marchetti et al ( 109 ) on temperature and dew duration for maximal infection on soybean have been utilized in models to assess the potential impact of SBR on U.S. agriculture prior to its arrival in 2004 ( 220 222 ). Similarly, daily weather variables, such as rainfall events, have been used to predict epidemic potential in field epidemics ( 40 ). To better understand and manage SBR epidemic s in north Florida a simple decision model w as developed based on presence of P. pachyrhizi and leaf wetness duration or rainfall events, and tested against traditional timing of fungicide applications. The results were validated and further interpreted based on the estimated SBR potential inoculum calculated using the differences in SBR epidemiology on soybean and susceptible and resistant kudzu and the area of these hosts in Florida (Chapter 4). The objectives of this study were to use the differences in epidemiology of SBR on soybean and suscep tible and resistant kudzu and the area of these hosts in Florida to 1)
118 estimate the potential maximum inoculum load, 2) the time of year the maximum inoculum load can occur in Florida and 3) use the inoculum estimates to validate simple decision applicatio n models. Material and Methods Estimation of Potential Inoculum Spore production per area per day on susceptible and resistant soybean and kudzu was obtained from detached leaf assays with a P. pachyrhizi isolate collected from naturally infected kudzu gro wing on the University of Florida, North Florida Research and Education Center in Quincy, Florida in the fall of 2010 and maintained on susceptible soybean in the greenhouse (Chapter 4). The difference between the amount of spores produced per inoculation area (50 mm 2 ) 10 and 15 days post inoculation for each host was divided by 5 days to calculate the sporulation per 50 mm 2 per day for each host and is considered to be representative of the exponential phase. The total area of soybean planted in Florida in 2010 ( 134 ) and the estimated area of personal communication ) and the corresponding leaf area indices were used to estimate the total available host tissue in Florida. Based on previous research on the characterization of resistance to SBR in kudzu ( 17 82 ) and screening of genetic resistance in kudzu populations (Chapter 4) the proportion of immune and resistant kudzu to the total area was estimated at 18 and 32%, respectively. Furthermore, the total available host tissue was estimated at half of the actual area due to the observation of leaf senescence after approximately 50% of the total area exhibits SBR symptoms and it is assumed to no longer contribute to the inoculum escaping the canopy. The spore escape rate from Andrade et al. ( 6 ) based on a simple
119 turbulence closure method and a parameterization of the canopy porosity in soybean was used to calculate th e amount of spores escaping the canopy from the total amount produced. The spore production per 50 mm 2 ( SP a ), 50% of the total area of host in Florida ( A 50 ), host leaf area index ( LAI ), and spore escape rate ( E ) parameters (Table 5 1) were used to estimat e the potential maximum amount of inoculum escaping from hosts in Florida using Equation 5 1( 6 ). (5 1) Where M i is the maximum amount of potential inoculum escaping host canopies; SP a is the spor e production per 50 mm 2 per day, A 50 is 50% of the total area (ha) of host; LAI is leaf area index; and E is the spore escape rate from the host canopy. From detached leaf assay data (Chapter 4) the amount of spores applied to each inoculation area (500 sp ores) and the average germination rate (79.5%) were used to calculate the total number of viable spores applied to each inoculation area (398 spores per 50 mm 2 ). The average number of uredinia produced 20 days post inoculation on each host was divided by the number of viable spores applied to calculate the infection efficiency ( IE ). The number of uredinia produced per day was calculated using the difference between the uredinia produced 10 and 15 days post inoculation for each host and dividing by 5 days. The spore production per 50 mm 2 per day, previously discussed was divided by the number of uredinia produced per day per host to calculate the spore production per uredinium per day ( SP u ). Previously Twizeyimana et al. ( 200 ) proved the utility of detached leaf assay for screening soybean for rust resistance. Furthermore spore production per uredinia on susceptible soybean used in this study concurrs with results from Twizeyimana et al. ( 200 ). Fr om detached leaf assay data
120 and other research (Chapter 4, 92 ) the latent period ( LP ), the number of days from inoculation until uredinium eruption, was calculated for each host. The proportion of spores deposited within the ca nopy of the spore origin ( D ) used was set at the upper limit of 0.1 ( 64 ) and the infectious period ( IP ) was obtained from the literature ( 150 224 ). These parameters (Ta ble 5 2) were assumed optimal and used to calculate the number of new uredinia per day over time using Equation 5 2 ( 150 ). (5 2) With the initial conditions, Y(t) = 0 at ( LP + IP ) < t < 0 and Y(t) = IE x I at t = 0 w here Y(t) is the number of new uredinia at day t ; IE is infection efficiency; D is the proportion of spores deposited within the canopy of the spore origin; SP u is the spore production per uredinium per day; LP is latent period; IP is th e infectious period in days as previously described; and I is the initial influx number of spores at the beginning of the year. In the model, 100 viable spores were assumed successfully infecting host plants ( Y ( t ) = IE x I = 100 at t = 0). The measure = Y ( t )/ Y ( t 1 ) represents the rate of disease increase per day. While Equation 5 2 had no spatial component, the maximum amount of new uredinia was set by the number of uredinia required to produce the maximum amount of spores on 50% of the total area of each host as calculated by Equation 5 1 and the number of spores produced per uredinium. The presence of SBR in Gadsden County in north Florida on kudzu and soybean was recorded from 2005 thru 2011 and the first observation of the pathogen on each host w as used as the initial infection date ( t = 0). To increase the accuracy of Equation 5 2, infection efficiency ( IE ) and latent period ( LP ) parameters were adjusted based on the effect of temperature. Monthly average
121 temperatures were calculated from daily observations recorded from Florida Automatic Weather Network (FAWN) stations in five counties in north Florida (Escambia, Jackson, Gadsden, Jefferson, and Suwannee, accessed from http://fawn.ifas.ufl.edu/ ) from 20 05 through 2008 (Table 5 3). The effect of temperature on the infection efficiency of SBR was assumed to follow a classical response curve for biological systems with three cardinal temperatures: optimum ( T opt = 22.5 C), maximum ( T max = 27.5 C), and min imum ( T min = 10 C) ( 109 ) and was calculated using Equation 5 3 ( 24 153 ). (5 3) Where = Y max /[( T opt T min )( T max T opt ) = 0.042 and = ( T max T opt )/( T opt T min ) = 0.40. Y is the infection index and Y max = 1 when wetness conditions are optimal. Chances for infections in a field are usually higher during the night and early morning than day time because relative humidity is higher a nd leaf wetness duration is longer in the night than in daytime ( 150 ). Therefore, the average infection index for each month was determined using mean nighttime temperatures. Mean night time temperature was estimated by avera ging the mean and the minimum monthly temperatures (Table 5 3). Multiplying infection efficiency at optimal conditions ( Y max ) and infection index ( Y ) gave the infection efficiency value for each temperature. To calculate the average latent period for a gi ven temperature latent period was defined as a function of physiological days ( PD ) of SBR. PD i can be a linear function of temperature with Equation 5 4 ( 70 221 ): PD i = ( T i T min )/( T opt T min ), if T min T i < T opt PD i = ( T max T i )/( T max T opt ), if T max T i > T opt (5 4)
122 Optimal temperature ( T opt = 22 C) is where latent period was found to be the shortest and maximum temperature ( T max = 42 C) and minimum temperature ( T min = 11 C) were from the l iterature ( 92 ). T i is the mean temperature calculated from 2005 through 2008 data. The average latent period during a given month was calculated with Equation 5 5 ( 150 ): LP (at T i ) = LP (at T opt )/ PD i (5 5) After disease initiation, development continues according to the average temperature conditions of the month examined. Data from the 2008 and 2009 National Wind Vane Spore Trap Program for SBR was used to validate the estimate of spores e scaping host canopies in Florida. In a wind vane spore trap a white petrolatum coated slide was placed at approximately a 45 degree angle and exchanged every 2 to 7 days from June to October. Wind vane spore traps were aligned approximately 165 cm above the ground and placed approximately 95 cm apart. Three wind vanes with diameters of approximately 15 cm were used in 2008 and one in 2009. Microscopic evaluation was performed on each slide and DNA was extracted from whatever residue was present on the s lide and assayed by quantitative polymerase chain reaction (qPCR) for P. pachyrhizi ( 71 ). In 2008 slides that contained greater than 50 visual spores on a slide were only recorded as having 50 visual spores. To better adjust these values time periods in 2008 where 2 to 3 slides were collected the number of visual spores was totaled over the time period. In 2009, all visual spores were recorded. Monthly temperatures for Gadsden County from 2009 to 2011 were calculated, as prev iously described, from the FAWN station located on the North Florida Research and
123 Education Center in Quincy, Florida ( http://fawn.ifas.ufl.edu/ ). The temperatures were used in Equations 5 3 and 5 4 to calculate the infection efficiency and latent period and used in Equation 5 2 to estimate the rate of disease increase and number of sporulating uredina. This was used to better understand disease severity and yield results from the decision application model in 20 09 to 2011. Decision Application Model Decision application model for SBR was develop and tested in 2009 to 2011, with one soybean planting in 2009 and 2010 and two soybean plantings in 2011 at the North Florida Research and Education Center in Quincy, FL. Pioneer 97M50 RR soybean was planted 5 May 2009, 17 May 2010, 25 June 2011, and 18 July 2011 in a 700 m 2 area with 51 cm row spacing and 6 to 7 seeds per 30 cm. Each treatment plot was approximately 4 x 6 m with approximately 2 to 4 m between each plot. Approximately 2.5 m border of soybean was planted around the entire area. All plantings had 4 replicates arranged in a complete randomized block design. Disease severity was evaluated at four corners (approximately one row and 2 m in from the border ro ws) and one central point within each plot. Severities were averaged to calculate overall disease severity in each plot. Approximately 9 m 2 area in the center of each plot was harvested at the end of each season and seed moisture was taken and seed weigh t and yield were adjusted to 13% moisture to calculate final yield. Applications of the triazole fungicide Folicur 3.6 (tebuconazole) were determined based upon the presence of SBR, on soybean or kudzu, in Gadsden County and environmental variables determ ined by data from Marchetti et al. ( 109 ) on maximal infection or by rainfall events determined by data from Del Ponte et al. ( 40 ) on maximum severity. Marchetti et al. ( 109 ) reported maximal infection occurred on soybean at 20 to
124 25 C with 10 to 12 hours of dew. Correlation analysis in Del Ponte et al. ( 40 ) indicated that rainfall variables were more strongly correlated with final disease s everity than temperature variables over a 1 month period from the time of first disease detection. Specifically, severity was >30% in southern Brazil locations with rainfall events greater than 12.5 cm, distributed over 5 to 10 rainy days during the 1 mon th period. In accordance with the previous mentioned environmental variables the following parameters were set up for application treatments if SBR was present in the Gadsden County and: if 3 consecutive days with 10 hours or more of leaf wetness and nigh t time temperatures (from 2200 hours to 0500 hours) between 20 to 25 C occurred fungicide was applied to the temperature/dew duration treatments; if over 5 days within the past 30 days, cumulative rain was equal or greater than 13 cm fungicide was applied to the rainfall event treatments. These treatments were compared to a control (no fungicide application) and a traditional fungicide application schedule with one application at soybean growth stage R1 (flowering) and one application at growth stage R3 ( pod development). Presence of SBR in the county, on soybean or kudzu was obtained from the Ipm PIPE website ( www.sbrusa.net ). Fungicide was applied using a platform sprayer with a 7 nozzle 11002 Turbo teejet at 35 p si and 25.7 L per ha. Weather data were obtained from a FAWN station located on the North Florida Research and Education Center campus in Quincy, FL, which records observations every 15 min at 60 cm above the soil. Hourly averages of temperature and relat ive humidity were calculated and used to calculate the dew point using Equation 5 6 ( 10 ). (5 6)
125 Where where T is the temperatures in degrees Celsius RH is the measured relative humidity, T d is the calculated dew point temperature ( C) and "ln" refers to the natural logarithm The constants are: a = 17.27 and b = 237.7 ( C). The uncertainty in the measured dew point temperature is a function of the measured temperature and re lative humidity and the uncertainties associated with those measurements. The uncertainty in the calculated dew point temperature is 0.4 C. This expression is based on the "Magnus" (or "Magnus Tetens") approximation for the saturation vapor pressure of water in air as a function of temperature ( 10 ). It is considered valid for 0 C < T < 60 C; 1% < RH < 100%; 0 C < Td < 50 C. Dew point temperatures were then compared with the hourly average air temperature and hours where the difference between the two variables was 0 to 2 were counted as dew lea f wetness hours for daily calculations ( 55 ). Weather data, from the previously mentioned FAWN station, was also used to calculate the number of hours below 0 C from 30 November to 17 April prior to each planting and total rai nfall from March through September. Results Based on infection of 50% of the total area of susceptible soybean, susceptible kudzu, and resistant kudzu in Florida the calculated maximum potential amount of spores escaping from the host canopy is 8.1 x 10 12 5.3 x 10 12 and 8.5 x 10 11 respectively. The number of uredinia required to produce the maximum amount of inoculum from susceptible soybean, susceptible kudzu, and resistant kudzu is 1.5 x 10 12 1.2 x 10 12 and 5.4 x 10 11 respectively. The average dat e that infection has been first observed on kudzu and soybean in Florida from 2005 to 2010 was used as the
126 initial infection date, which on average was in mid March for kudzu and mid June for soybean. Using the average initial infection date and optimum p arameters (Table 5 2) the maximum amount of inoculum produced on soybean occurred 93 days after initial infection (DAI) in mid September and on susceptible kudzu 196 DAI at the end of September (Figure 5 1 ). The point in time that the maximum amount of in oculum would be produced on resistant kudzu was not predicted (Figure 5 1) due to the epidemic staying in the early lag phase (Figure 5 2). Based on the calculated rates of disease increase ( during different months of a growing season in north Florida (Table 5 4), temperature conditions from April through October are suitable for the development of SBR on susceptible soybean and susceptible kudzu, but not on resist ant kudzu (Figure 5 2). SBR epidemics on resistant kudzu never reached the logistic phase and hence never temperature conditions examined are more suitable for SBR development on soybean than k udzu. May through October have the highest rates of disease increase for soybean and susceptible kudzu, with May and September temperature conditions near optimal promoting the shortest latent period ( LP = 6 to 7 on soybean and LP = 9 to 10 on susceptibl e kudzu). The first increase in spores collected by the wind vane spore traps occurred in August, followed by the highest amount of spores collected in mid September in both 2008 and 2009 (Figure 5 3). This agrees with the high calculated rate of disease increase, and the shorten latent period calculated by the disease model (Table 5 4, Figure 5 2). Disease pressure varied over the 3 years the decision model was tested in the field, as indicated by the first observation of SBR on kudzu and soybean (Tabl e 5 5).
127 Furthermore, visual disease symptoms in plots were only observed in 2009 (Figure 5 4), although soybean yield was only significantly lower in control plots regardless of treatment in 2009 (Figure 5 5). In 2009 the rainfall event and leaf wetness based decision applications were effective for controlling SBR (Figure 5 4), although the yields from these treatments were not significantly greater than a traditional application at growth stages R1 (flowering) and R3 (pod development) (Figure 5 5). In 2010 a traditional application and leaf wetness based application resulted in the highest yield, although they were not significantly greater than the control yield. Furthermore, in 2010 the yields from the control and rain based application were not sign ificantly different, which was expected because of the lack of significant rain events no applications were applied in the rain based treatments. In 2011 yields across treatments and controls were not significantly different, due to the lack of disease pr essure. The calculated rates of disease increase, of SBR on soybean and susceptible kudzu during the growing seasons in 2009, 2010 and 2011 in Quincy, FL (Table 5 6) indicate that sufficient temperature conditions occurred May through September each yea r; although greater and more consistent disease increase occurred in 2009. The lower values in 2010 and 2011 season indicate later disease onset and slower disease average temp erature conditions (Table 5 using temperature conditions in 2009 to 2011 was not greater than 1.00, because epidemics on resistant kudzu never reach the logistic phase (data not shown). Based on first detecti on of SBR on kudzu and soybean in 2009 through 2011 and monthly temperature conditions high disease severity was estimated in 2009, but not in 2010
128 and 2011 (Figure 5 6). Similarly, high disease severity estimated in 2008 can be contributed to early disea se onset and previously reported rain events ( 225 ). Infections and potential initial inoculum survival through winter temperatures most likely affected disease onset each year. The winter of 2008 to 2009 in Quincy, Florida ha d approximately 180 hr below 0 C, whereas winters of 2009 to 2010 and 2010 to 2011 had 265 and 306 hr below 0 C, respectively. In both 2010 and 2011 growing seasons there were drought conditions. Total rainfall from March through September in 2009 was 1107 mm in Quincy, FL, whereas in 2010 and 2011 total rainfall was only 719 and 739 mm, respectively. Discussion Under optimal conditions, the available area of hosts in Florida, and the average disease onset of SBR, soybean and susceptible kudzu have the potential to contribute a significant number of spores for long distance transport, while resistant kudzu has very low to no potential to contribute spores for long distance transport. The lack of severe SBR epidemics on resistant kudzu is due to the redu ced infection efficiency and spore production. This further reduces the total area of kudzu available to contribute spores for long distance transport and initial inoculum for infection of soybean. Temperature conditions in May, June, and September were near optimal for SBR development (Table 5 4, Figure 5 2) on soybean and susceptible kudzu. Similarly, Pivonia and Yang ( 150 ), reported the optimal month for SBR reproduction in Baton Rouge, LA based on temperature conditions w as in May and September. Comparing SBR incipience times in Florida, the results indicate conditions for disease development are suitable for infection of soybean 2 months prior to the actual
129 appearance on soybean using the average of mid June. Although co mparing SBR development on susceptible kudzu in Florida the results indicate less suitable conditions until May. This may explain why SBR epidemics on kudzu have been observed to develop slowly during the spring and early summer despite early detection ( 32 ). Furthermore, a lower disease increase rate implies a longer time for disease to develop to a detectable level ( 150 ). Although kudzu canopies may have a more conducive environment for SBR establishment and development than portrayed by the temperature conditions used in this study. The estimates of the potential maximum inoculum load occurring in September in Florida correspond to the maximum amount of spores collected in September from wind vane spore traps in Quincy, FL in 2008 and 2009 (Figure 5 1 and 5 3). Further more, the greater number of sporulating uredinia on susceptible kudzu in 2009 (Figure 5 6), due to earlier disease onset and more conducive temperature conditions, may account for the higher number of spores collected in 2009 than in 2008. These data indi cate the parameters used in this study can accurately predict time periods of potential inoculum based on disease onset and temperature conditions. This study evaluated the effect of host and temperature to address suitability for disease development and e stimation of potential inoculum for long distance transport. The effect of temperature conditions on infection efficiency and latent period taken from the literature were from studies conducted at constant temperatures, while effects on infection efficien cy and latent period may differ under natural ly fluctuating temperature conditions. Comparing the temperature e ffect on the latent period of lettuce downy mildew using average temperatures from naturally and theoretically fluctuating
130 temperatures to previ ous work conducted at constant temperature conditions revealed that under fluctuating temperatures there was a smaller temperature effect on the length of the latent period than reported under constant temperature conditions ( 16 5 166 ). Hence, effect of temperature on latent period and infection efficiency may be improved by the inclusion of daily temperature oscillations. Moreover, calculated dew durations are affected by the actual temperature and relative humidity at a particular time of day, the estimates being generally least accurate early in the morning ( 164 218 ). Future studies which include more factors into a disease simulation model might further increase the accuracy of time period of SBR onset and development. In 2009 all soybean plants that received fungicide treatments had significantly greater yields than untreated soybean plants. The rain and leaf wetness based applications we re significantly greater at controlling SBR in 2009, although the yields from these treatments were not significantly greater than the traditional application at growth stages R1 (beginning bloom) and R3 (pod development) (Figure 5 4 and 5 5). While disea se severity reached approximately 75% in traditional application plots the yield was not significantly different from those that did not develop disease symptoms, most likely because disease did not develop until reproductive stage R7 (beginning maturity) after seed fill. Fungicide spray programs for SBR tested in 2006 and 2007 in Alabama, found fungicide applications at soybean reproductive growth stages R2 (full bloom) and R5 (beginning seed) resulted in significantly higher yields ( P 0.1) than untreated soybean when SBR was first detected at growth stages R4 R5 (pod fill and beginning seed) ( 174 ). Hence, soybean that has fungicide applied at early soybean reproductive growth stages (R1 and R3 or R2 and R 5) can provide significantly greater
131 yield than untreated soybean when SBR is present in the surrounding area and conducive conditions for disease development occur. Depending on the soybean growth stage when symptoms of SBR are first detected and environm ental conditions additional fungicide applications may or may not result in greater yields. In this study, additional fungicide applications (more than the two applications at soybean growth stages R1 and R3) did not result in significantly greater yield due to the late appearance of SBR symptoms at soybean growth stage R7 (Table 5 5 and Figure 5 5). However, three applications of tebuconazole at R1, R3, and R5 did result in significantly greater yields than two applications at R1 and R3 when SBR was firs t detected at R4 in fungicide trials conducted in Quincy, FL in 2006 ( 131 ). Therefore, as first disease detection occurs earlier in the growing season (i.e. on earlier soybean growth stages) additional fungicide applications m ay provide greater protection from yield loss due to SBR. In 2010 and 2011 fungicide applications did not significantly increase yield compared to non treated soybean due to the lack of substantial SBR inoculum and conducive environmental conditions for di sease development. While SBR was detected in other soybean fields in Quincy, FL in 2010 and 2011 when soybeans were at growth stages R6 and R1 R2, respectively, SBR symptoms were never observed in the experimental plots. This can be explained by the low number of estimated sporulating uredinia based on first detection and temperature conditions (Figure 5 6). Hence, a model should use both soybean growth stage when SBR is first detected as well as environmental conditions to guide fungicide applications t o reduce impact on yield loss.
132 Furthermore, increased hours of winter temperatures below 0 C prior to the 2010 and 2011 seasons, compared to 2009, reduced initial inoculum and contributed to later disease onset. Similarly, drought conditions in 2010 and 2011 reduced inoculum dispersal and disease development. Therefore, while temperature conditions were conducive for SBR development in 2010 and 2011, the lack of initial inoculum due to winter temperatures delayed disease onset and lack of substantial rai n events retarded disease development. First detection of SBR within the county of planted soybean can be used as a risk indicator, but would be best utilized in combination with environmental data to assess potential dispersal of inoculum and rate of dis ease increase to aid in fungicide application decisions. Based on temperature conditions in north Florida from 2005 through 2011, there is a narrow window of suitable conditions for SBR establishment and development in May and June and again in September, which is similar to that found for Baton Rouge, LA ( 150 ). This may hinder early SBR establishment in the south and consequently its northward spread, but may still potentially impact yields of late planted soybean in the south This study provides fundamental data on inoculum potential from infected kudzu in Florida which can improve the accuracy of SBR forecasting models, but further incorporation of other variables affecting inoculum dispersal are needed to analyze potential risk in northern soybean production areas. The potential impact on yield due to low levels of SBR severity and the lack of resistant varieties commercially available affirms the destructive potential of SBR. The destructive potential of SBR and milder wi nters and significant changes in rainfall events due to climate change emphasizes the
133 need for further research and forecasting of SBR to better assess and predict SBR risk to U.S. soybean production.
134 Table 5 1. Variables used to estimate the potential m aximum amount of soybean rust inoculum escaping from different host canopies in Florida Variables Susceptible Soybean Susceptible Kudzu Resistant Kudzu Spore production per 50 mm 2 per day ( SP a ) 224 26 7 Half of total area (ha) of host in FL ( A 50 ) 4500 50500 32500 Leaf area index ( LAI ) 3 6 6 Spore escape rate ( E ) 0.12 0.12 0.12
135 Table 5 2. Variables at optimal conditions, used to calculate the number of new uredinia per day over time Variables Susceptible Soybean Susceptible Kudzu Resistant Kudzu Infection efficiency ( IE ) 0.19 0.09 0.009 Spore production per uredinium ( SP u ) 45 36 13 Proportion of spores deposited ( D ) 0.1 0.1 0.1 Latent period ( LP ) 6 9 12 Infection period ( IP ) 21 21 21
136 Table 5 3. Monthly temperature for 2005 to 2008 across 5 different locations in north Florida March April May June July August September October Mean 15.3 18.4 22.9 26.0 26.6 26.8 24.9 21.1 Minimum 8.0 10.7 15.7 20.0 21.6 22.0 19.5 15.2 Maximum 22.8 26.2 30.4 33.3 33.8 33.9 31.9 28.5 Mean night a 11.6 14.5 19.3 23.0 24.1 24.4 22.2 18.2 a Mean night temperature is estimated as the average of the mean and minimum temperatures.
137 Table 5 4. Rates of disease increase a calculated for different hosts and months as more suitable conditions for rust reproduction Host Optimal conditions Marc h April May June July August September October Susceptible soybean 1.3 0 1.06 1.16 1.29 1.25 1.21 1.21 1.25 1.24 Susceptible kudzu 1.13 1.02 1.07 1.12 1.12 1.11 1.11 1.12 1.11 Resistant kudzu b 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 a Rate of disea se increase per day [ = Y ( t )/ Y ( t 1) ] was calculated using Y ( t ) = IE x D x SP u [ Y ( t LP ) Y ( t LP IP )] + Y ( t 1) using the initial conditions, Y(t) = 0 at ( LP + IP ) < t < 0 and Y(t) = IE x I at t = 0 w here Y(t) is the number of new sporulating uredinia at day t ; IE is infection eff iciency; D is the proportion of spores deposited within the canopy of the spore origin; SP u is the spore production per uredinium per day; LP is latent period; IP is the infectious period in days; and I is the initial influx number of spores at the beginni ng of the year. In the model, 100 viable spores were assumed successfully infecting host plants ( Y ( t ) = IE x I = 100 at t = 0). b Soybean rust epidemics on resistant kudzu never reach the logistic phase and hence never attain a rate of disease increase ( ) greater than 1.00
138 Table 5 5. Application decision model results and first observations of soybean rust (SBR) on kudzu and soybean each year Plant date SBR detection on kudzu a SBR detection on soybean b Treatment c Applications d Yield (Kg/H) 5 May 2009 24 April 2009 7 July 2009 Control 0 1181 Tradational 2 2193 Rain event 4 2632 Leaf wetness 6 2569 17 May 2009 15 July 2010 10 Sept. 2010 Control 0 3350 Tradational 2 4011 Rain event 0 3912 Leaf wetness 4 3231 24 June 2011 12 July 2011 25 Aug. 2011 Control 0 3149 Tradational 2 3123 Rain event 1 3215 Leaf wetness 3 3465 18 July 2011 12 July 2011 25 Aug. 2011 Control 0 2810 Tradational 2 2648 Rain event 1 2627 Leaf wetness 2 2696 a First report of SBR esta blishment on kudzu after winter conditions b First report of SBR establishment on soybean after winter conditions c Treatments included: Control no application of fungicide; Traditional fungicide application at soybean growth stage R1 (flowering) and R 3 (pod development); Rain event fungicide applied if SBR present in county and if over 5 days within the past 30 days cumulative rain was equal or greater than 13 cm; Leaf wetness fungicide applied if SBR present in county and if 3 consecutive days wit h 10 hours or more of leaf wetness and night time temperatures (from 2200 to 0500 hours) between 20 to 25 C occurred d The number of application of the triazole fungicide Folicur 3.6 (tebuconazole)
139 Table 5 6. Rates of disease increase Rates of disease a calculated for soybean and susceptible kudzu and months of different years as affected by temperatures, during the exponential phase of Year Host M arch April May June July August September October 2009 Susceptible soybean 1.00 1.26 1.27 1.22 1.22 1.22 1.24 Susceptible kudzu 1.00 1.11 1.11 1.10 1.11 1.10 1.12 2010 Susceptible soybean 1.00 1.27 1.27 1.23 1.21 1.24 1.24 Susceptible kudzu 1 .00 1.11 1.11 1.10 1.11 1.12 1.12 2011 Susceptible soybean 1.08 1.20 1.27 1.23 1.22 1.23 1.24 Susceptible kudzu 1.02 1.08 1.11 1.10 1.11 1.11 1.12 a Rate of disease increase per day [ = Y ( t )/ Y ( t 1) ] was calculated using Y ( t ) = IE x D x SP u [ Y ( t LP ) Y ( t LP IP )] + Y ( t 1) using the initial conditions, Y(t) = 0 at ( LP + IP ) < t < 0 and Y(t) = IE x I at t = 0 w here Y(t) is the number of new sporulating uredinia at day t ; IE is infection eff iciency; D is the proportion of spores deposited within the canopy of the spore origin; SP u is the spore production per uredinium per day; LP is latent period; IP is the infectious period in days; and I is the initial influx number of spores at the beginni ng of the year. In the model, 100 viable spores were assumed successfully infecting host plants ( Y ( t ) = IE x I = 100 at t = 0).
140 Figure 5 1. The number of uredinia developing on different hosts in Florida over time at optimal condition for disease deve lopment.
141 Figure 5 2. Soybean rust infection progress curves. S oybean rust infection progress curves were calculated for hypothetical epidemics, staying in the exponential phase, on 3 different hosts in Florida, under temperature conditions of several mo nths during the soybean growing season.
142 Figure 5 3. Enumeration of visual urediniospores from passive wind vane spore traps Wind vane spore traps utiliz ed white petrolatum coated slides to capture spores, which were then analyzed by microscopic evaluat ion. Wind vane spore traps were location at North Florida Research and Education Center, Quincy, FL. from June to October.
143 Figure 5 4. Disease severity of soybean rust (SBR) within fungicide treatment plots Severity of SBR recorded in plots in 2009 loc ated at the North Florida Research Education Center in Quincy, FL. Treatments included Control no application of fungicide; Traditional fungicide application at soybean growth stage R1 (flowering) and R3 (pod development); Rain event fungicide appli ed if SBR present in county and if over 5 days within the past 30 days cumulative rain was equal or greater than 13 cm; Leaf wetness fungicide applied if SBR present in county and if 3 consecutive days with 10 hours or more of leaf wetness and night time temperatures (from 2200 to 0500 hours) between 20 to 25 C occurred
144 Figure 5 5. Average soybean yields (kg/H) from fungicide application treatment plots Plots were planted at the North Florida Research and Education Center, in Quincy, FL in 2009, 20 10, and 2011. Treatments included: Control no application of fungicide; Traditional fungicide application at soybean growth stage R1 (flowering) and R3 (pod development); Rain event fungicide applied if soybean rust ( SBR ) present in county and if ov er 5 days within the past 30 days cumulative rain was equal or greater than 13 cm; Leaf wetness fungicide applied if SBR present in county and if 3 consecutive days with 10 hours or more of leaf wetness and night time temperatures (from 2200 to 0500 hour s) between 20 to 25 C occurred
145 Figure 5 6. Estimated number of sporulating uredinia over time on soybean and susceptible kudzu Estimation based on temperature conditions effect on infection efficiency and latent period; usi ng first detection data and temperature data from Gadsden County in north Florida in 2008 through 2011.
146 LIST OF REFERENCES 1. Abramovitz, J. N. 1983. Pueraria lobata Willd. (Ohwi) kudzu: limitations to sexual reproduction. MS thesis, University of Maryland, College Park, MD, USA. 2. Akem, C. N. 1995. The effect of timing of fungicide applications on the control of frogeye leaf spot and grain yield on soybeans European J. of Plant Path. 101:183 187. 3. Akinsanmi O. A. Ladipo J. L. and Oyekan P. O 2001. First re port of soybean r ust ( Phakopsora pachyrhizi ) in N igeria Plant Dis 85: 97. 4. Altschul, S. F., Madden, T. L., Schaffer, A. A., Zhang, J., Zhang, Z., Miller, W., and Lipman, D. J. 1997. Gapped BLAST and PSI BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25:3389 3402. 5. Ameli ne Torregrosa, C., Wang, B. Roe, B., Young, N. D., and Cannon, S. B. 2007. Identification and characterization of nucleotide binding site leucine rich repeat genes in the model plant Medicago truncatula Plant Phys. 146:5 21. 6. Andrade, D., Pan, Z., Dannevik, W., and Zidek, J. 2009. Modeling soybean rust spore escape from infected canopies: model description and preliminary results. Journal of Applied Meteorology and Climatology. 48:789 803. 7. Asthana, A., and Tuv eson, R W. 1992. Effects of UV and phototoxins on selected fungal pathogens of citrus. Int. J. Plant Sci. 153:442 452. 8. Aylor, D. E. 1986. A framework for examining inter regional aerial transport of fungal spores Ag. and For. Meteorol. 38:263 288. 9. Baker E. A. 1974. The influence of environment on leaf wax development in Brassica oleracea var. gemmifera New Phytol. 73:955 966 10. B arenbrug, A. W. T. 1974. Psychrometry and Psychrometric Charts, 3rd Edition, Cape and Transvaal Printers Ltd.,Cape Town, S.A. 11. Ba tchelor, W. D., Yang, X. B., and Tschanz, A. T. 1997. Development of a neural network for soybean rust epidemics. Tans. ASAE 40:247 252 12. Beattie, G. A., and Marcell, L. M. 2002. Effect of alterations in cuticular wax biosynthesis on the physicochemical pr operties and topography of maize leaf surfaces. Plant, Cell Environ. 25:1 16
147 13. Beuerlein, J. and Dorrance, A. E. 2005. Chapter 5: Soybean production Ohio Agronomy Guide, 14th Edition. 14. Board, J. E., and Harville, B. G. 1998. Late planted soybean yield res ponse to reproductive source/sink stress. Crop Sci. 38:763 771. 15. Bonde, M. R., Melching, J. S., and Bromfield, K. R. 1976. Histology of the suscept pathogen relationship between Glycine max and Phakopsora pachyrhizi the cause of soybean rust. Phytopatholo gy 66:1290 1294. 16. Bonde, M. R., Nester, S. E., Austin, C. N., Stone, C. L., Frederick, R. D., Hartman, G. L., and Miles, M. R. 2006. Evaluation of virulence of Phakopsora pachyrhizi and P. meibomiae isolates. Plant Dis. 90:708 716. 17. Bonde, M. R., Nester, S. E., Berner, D. K., and Frederick, R. D. 2008. Comparative susceptibility of legume species to infection by Phakopsora pachyrhizi Plant Dis. 92:30 36. 18. Bonde, M. R., Nester, S. E., Moore, W. F., and Allen, T. W. 2009. Comparative susceptibility of kudzu acce ssions from the southeastern United States to infection by Phakopsora pachyrhizi Plant Dis. 93:593 598. 19. Boyette, C. D., Abbas, H. K., and Walker, H. L. 2000. Bioherbicidal control of kudzu with Myrothecium verrucaria Phytopathology 90:S9. 20. Bromfield, K. R. 1984. Soybean rust. Monograph. The American Phytopathological Society: St. Paul, MN. pg. 1 58. 21. Bromfield, K. R., and Hartwig, E. E. 1980. Resistance to soybean rust and mode of inheritance. Crop Sci. 20:254 255. 22. Bromfi eld, K. R., and J. S. Melching. 1982 (Abstract) Sources of specific resistance to soybean rust. Phytopathology 72:706. 23. Britton O K Orr D ., and Sun J H 2002 Biological control of Kudzu, Pueraria montana var lobata In: Biological Control of Invasive Plants in the Eastern United St ates (eds R Van Driesche, B Blossey, M Hoddle, S Lyon & R Reardon), 323 325, USDA, Forest Service FHTET 2002 04, Morgantown, WV, USA. 24. Butler, D. R., and Jadhav, D. R. 1991. Requirements of leaf wetness and temperature for infection of groundnut by rust. Plant Pathol. 40:395 400. 25. Bryan, G. T., Wu, K. S., Farrall, L., Jia, Y., Hershey, H. P., McAdams, S. A., Faulk, K. N., Donaldson, G. K., Tarchini, R., and Valent, B. 2000. A single amino acid difference distinguishes resistant and susceptible alleles of t he rice blast resistance gene Pi ta Plant Cell 12:2033 2046.
148 26. CAB International. 2008. Crop Protection Compendium. CABI. Online publication. 27. Caldwell, S. P. and Myers, O. J. 1980. Seasonal field mapping of soybean cyst nematode populations as influenced by cultivars and blends. In World Soybean Research Conference II, 1979, ed. F. T. Corbin. Boulder, CO: Westview (Abstr.). 28. Caldwell, P. and Laing, M. 2002. Soybean rust A disease on the move. Southern African Society for Plant Pathology Monthly Feature. March 2002. Online publication. 29. Canteri, M. G., Caramori, P., Tsukahara, R., Silva, O. C., Faria, R., and Godoy, C. V. 2005. A system to map risk of infection by Phakopsora pachyrhizi for Parana State, Brazil. Phytopathology 95:S16. 30. Casey, P. S. 1979. Th e epidemiology of soybean rust Phakopsora pachyrhizi Syd. Ph.D. thesis, University of Sydney, 203 pp. 31. Chakraborty, N., J. Curley, R. D. Frederick, D. L. Hyten, R. L. Nelson, G. L. Hartman, and Diers, B. W. 2009. Mapping and confi rmation of a new allele at R pp1 from soybean PI 594538A conferring RB lesion type resistance to soybean rust. Crop Sci. 49:783 790. 32. Christiano, R. S. C. and Scherm, H. 2007. Quantitative aspects of the spread of Asian soybean rust in the southeastern United States, 2005 to 2006. Phy topathology 97:1428 1433. 33. Clark, J. A., and Levitt, J. 1956.The basis of drought resistance in the soybean plant. Physiol. Plant. 9:598 606 34. Corley, R. N. Woldeghebriel, W., and Murphy, M. R. 1997. Evaluation of the nutritive value of kudzu ( Pueraria loba ta ) as a feed for rumints. Animal Feed Science and Technology 68:183 188. 35. Dangl, J. L. and Jones, J. D. G. 2001. Plant pathogens and integrated defense responses to infection. Nature 411:826 833. 36. Dias, A. P. S., Li, X., Harmon, P. F., Harmon, C. L., and Yang, X. B. 2011. Effects of shade intensity and duration on Asian soybean rust caused by Phakopsora pachyrhizi Plant Dis. 95:485 489. 37. Delaney, D. P., Lawrence, K. S Sikora, E. J., Lawence, G. W., Delaney, M. A., and Pegues, M. D. 2007. Efficacy of fo liar fungicides for Asian soybean rust disease management and yield enhancement in Alabama. Proc. 2007 National Soybean Rust Symposium.
149 38. Del Ponte, E. M., and Esker, P. D. 2008. Meteorological factors and Asian soybean rust epidemics A systems approach an d implications for risk assessment. Sci. Agric. (Piracicaba, Braz.) 65:88 97. 39. Del Ponte, E. M., Godoy, C. V., Canteri, M. G., Reis, E. M., and Yang, X. B. 2006. Models and applications for risk assessment and prediction of Asian soybean rust epidemics. Fi topatol. Bras. 31:533 544. 40. D el Ponte, E. M., Godoy, C. V., Li, X., and Yang, X. B. 2006. Predicting severity of Asian soybean rust epidemics with empirical rainfall models. Phytopathology. 96:797 803. 41. Dias, A. P. S., Li, X., Harmon, P. F., Harmon, C. L., and Yang, X. B. 2011. Effects of shade intensity and duration on Asian soybean rust caused by Phakopsora pachyrhizi Plant Dis. 95:485 489. 42. Dixon, M. S., Hatzixanthis, K. Jones, D. A. Harrison, K., and Jones, J. D. G. 1998. The tomato Cf 5 disease resist ance gene and six homologs show pronounced allelic variation in leucine rich copy number. Plant Cell 10:1915 1926. 43. Dixon, M. Jones, D. A., Keddie, J. S. Thomas, C. M., Harrison, K. and Jones, J. D. 1996. The tomato Cf 5 disease resistance locus comprises two functional genes encoding leucine rich repeat proteins. Cell 84:451 459. 44. Dodds, P. N., Lawrence, G. J., and Ellis, J. G. 2001. Six amino acid changes confined to the leucine rich repeat strand/ turn motif determine the difference between the P an P2 rust resistance specificities in flax. Plant Cell 13:163 178. 45. Dorrance, A. E., Draper, M. A., and Hershman, D. E. eds. 2008. Using foliar fungicides to manage soybean rust. Land Grand Universities Cooperation NCERA 208 and OMAF, Bulletin SR 2008. 46. Dufa ult, N. S., Isard, S. A., Marois, J. J., and Wright, D. L. 2010. Removal of wet deposited Phakopsora pachyrhizi urediniospores from soybean leaves by subsequent rainfall. Plant Dis. 94:1336 1340. 47. Dufresne, L. and Bean, G. 1987. Effects of temperature and light intensity on telia development by Puerto Rico and Taiwan isolates of Phakopsora pachyrhizi the soybean rust fungus. Plant Dis. 71:629 631. 48. Everest, J. W., Miller, J. H. Ball, D. M., and Patterson, M. G. 1991. Kudzu in Alabama. Alabama Cooperative Extension Service Circular ANR 65, Auburn University, Auburn, Alabama, USA.
150 49. Excoffier L Laval G Schneider S. 2005 Arlequin version 3.0: an integrated software package for population genetics data analysis. Evol Bioinform 1:47 50. 50. Excoffier, L., Smouse, P. E., and Quattro, J. M. 1992. Analysis of molecular variance inferred from metric distances among DNA haplotypes: application to human mitochondrial DNA restriction data. Genetics 131:479 491. 51. Farris, J. and Reilly, C. C. 2000. The biological co ntrol of kudzu ( Pueraria lobata ). Proceeding of the American Association for the Advancement of Science, Washington, D.C. 116: AB2. 52. Flenet, F. Kiniry, J. R., Board, J. E., Westgate, M. E., and Reicosky, D. C. 1996. Row spacing effects on light extinction coef ficients of corn, sorghum, soybean, and sunflower. Agron. J. 88:185 190 53. Funke, T. Han, H., Healy Fried, M. L., Fischer, M., and Schnbrunn, E. 2006. Molecular basis for the herbicide resistance of Roundup Ready crops. PNAS. 103. 13010 13015. 54. Garcia, A., Calvo, E. S. de Souza Kiihl, R.A. Harada, A. Hiromoto, D.M. and Vieira L.G.E. 2008. Molecular mapping of soybean rust ( Phakopsora pachyrhizi ) resistance genes: Discovery of a novel locus and alleles. Theor. Appl. Genet. 117:545 553. 55. Gates, D. M. 1968. Tr anspiration and Leaf Temperature. Annl Rev Plant Physiol 19 : 211 238. 56. George, S., Sharma, J., and Yadon, V. L. 2009. Genetic diversity of the endangered and narrow endemic Piperia yadonii (Orchidaceae) assessed with ISSR polymorphisms. Am. J. Bot. 96:2022 2030. 57. Gevens, A. J., Nequi, N., Vitoreli, A., Marois, J. J., Wright, D. L., Harmon, C. L., and Harmon, P. F. 2008. First Report of Soybean Rust Caused by Phakopsora pachyrhizi on Erythrina herbacea (Coral Bean) Plant Dis. 92:1472. 58. Ghajar, F., Holford, O., C other, E., and Beattie, A. 2006. Effects of ultraviolet radiation, simulated or as natural sunlight, on conidium germination and appressorium formation by fungi with potential as mycoherbistats. Biocontrol Science and Technology 16:451 469 59. Godoy, C. 2009 Changes in performance of soybean rust fungicides over years and new management strategies adopted in Brazil. 2009 Natl. Soybean Rust Symp., New Orleans, LA. 9 11 Dec. 2009. 60. Gonzalez, R., Paul, N. D., Percy, L., Ambrose, M. McLaughlin, C. K., Barnes, J. D. Aresea, M., and Wellburn, A. R. 1996. Responses to ultraviolet B radiation
151 (280 315 nm) of pea ( Pisum sativum ) lines differing in leaf surface wax. Physiol. Plant. 98: 852 860. 61. Green, A. 1984. Soybean Rust. Pest Not Known to Occur in the United States o r of Limited Distribution No. 56. USDA APHIS PPQ. 62. Gregory, P. H. 1973. Microbiology of the Atmosphere, ed. N. Polunin. Leonard Hill Books. Aylesbury. 377. 63. Guard, A. T. 1931. Development of floral organs of the soybean. Bot. Gaz. 91:97 102. 64. Gumpert, F. M ., Geiger, H. H. and Stahle, U. 1987. A mathematical model of the epidemics in homogeneous and heterogeneous host stands. Zeitschrift Pflanzendrankheiten Pflanzenschutz 94:206 215. 65. Harmon, P. F., Momol, M. T., Marois, J. J., Dankers, H., and Harmon, C. L. 2005. Asian soybean rust caused by Phakopsora pachyrhizi on soybean and kudzu in Florida. Plant Health Progress. doi:10.1094/PHP 2005 0613 01 RS. 66. Hartman, G. L., Sinclair, J. B., and Rupe, J. C. 1999. Compendium of Soybean Disease 4th Edition. APS Press. 67. Hartman, G. L, Miles, M. R., and Frederick, R. D. 2005. Breeding for resistance to soybean rust. Plant Dis. 89:664 666. 68. Hartwig, E. E. 1986. Identifi cation of a fourth major gene conferring resistance to soybean rust. Crop Sci. 26:1135 1136. 69. Hartwig, E. E., and K. R. Bromfield. 1983. Relationships among three genes conferring resistance to rust in soybeans. Crop Sci. 23:237 239. 70. Hau, B., Eisensmith, S. P., and Dranz, J. 1985. Construction of temporal model: II. Simulation of aerial epidemics. Pages 31 66 in: Advances in Plant Pathology, Vol 3. C. A. Gilligan, ed. Academic Press, Orlando, FL. 71. Haudenshield, J. S., Chaudhary, P., and Hartman, G. L. 2010. Recovery of Phakopsora pachyrhizi urediniospores from passive spore trap slides and extraction of their DNA for quantitative PCR. Phytopathology 100:S185. 72. Henning, A. A. and Godoy, C. e no mundo. In: Zambolim, L. (Ed.), Ferrugem Asia tica da Soja. UFV. Viosa, Brazil, pp. 1 14. 73. Holliday, P. 1998. A Dictionary of Pl ant Pathology, 2 ed. Cambridge University Press. Cambridge.
152 74. Hymowitz, T 1990. Soybeans: The Success Story. Advances in New Crops. Pages 159 163. 75. Hyten, D. L., Hartman, G. L. Nelson, R. L. Frederick R. D., Concibido, V. C. Narvel, J. M. and Cregan P. B. 2007. Map location of the Rpp1 locus that confers resistance to soybean rust in soybean. Crop Sci. 47:837 840. 76. Hyten, D. L., Smith, J. R. Frederick, R. D. Tucker, M. L. Song, Q. and Cregan P. B. 2009. Bulk segregate analysis using the Golden Gate ass ay to locate the Rpp3 locus that confers resistance to Phakopsora pachyrhizi (soybean rust) in soybean. Crop Sci. 49:265 271. 77. Isakeit, T., Miller, M. E., Saldana, R., Barnes, L. W., McKemy, J. M., Palm, M. E., Zeller, K. A., DeVries Paterson, R., and Levy L. 2006. First Report of rust 23 caused by Phakopsora pachyrhizi on soybean and kudzu in Texas Plant Dis. 90:971. 78. I sard, S. A., Dufault, N. S., Miles, M. R., Hartman, G. L., Russo, J. M., de Wolf, E. D., & Morel, W. (2006). The effect of solar irradia nce on the mortality of Phakopsora pachyrhizi urediniospores. Plant Disease. 90: 941 945. 79. Isard, S. A., Gage, S. H., Comtois, P., and Russo, J. M. 2005. Principles of the atmospheric pathway for invasive species applied to soybean rust. BioScience 55:851 8 61. 80. Isard, S. A., Russo, J. M., and De Wolf, E. D. 2006. The establishment of a national Pest Information Platform for Extension and Education. Published Online at Plant Heath Progress. doi: 10.1094/PHP 2006 0915 01 RV. 81. Ishii, H., Fraaije, B. A., Sugiyam a, T., Noguchi, K., Nishimura, K., Takeda, T., Amano, T., and Hollomon, D. W. 2001. Occurrence and Molecular Characterization of Strobilurin Resistance in Cucumber Powdery Mildew and Downy Mildew. Phytopathology 91:1166 1171. 82. Jordan, S. A., Mailhot, D. J., Gevens, A. J., Marois, J. J., Wright, D. L., Harmon, C. L., and Harmon, P. F. 2010. Characterization of kudzu ( Pueraria spp.) resistance to Phakopsora pachyrhizi the causal agent of soybean rust. Phytopathology 100:941 948. 83. Jurick, W. M., Narvez, D. F., Bren nan, M. M., Harmon, C. L., Marois, J. J., Wright, D. L., and Harmon, P. F. 2008. Winter survival of the soybean rust pathogen, Phakopsora pachyrhizi in Florida. Plant Dis. 92:1551 1558. 84. Kakani, V. G., Reddy, K. R., Zhao, D., and Sailaja, K. 2003. Field c rop responses to ultraviolet B radiation: a review. J. Agric. Forest. Meteorol. 120 :191 218.
153 85. Keller, M., Rogiers, S. Y., and Schultz, H. R. 2003. Nitrogen and ultraviolet 94 86. Killgo re, E., and Heu, R. 1994. First report of soybean rust in Hawaii Plant Dis 78:1216. 87. Kim, Y. S., Dixon, P., Vincelli, P., and Farman, M. L. 2003. Field resistance to strobilurin (QoI) fungicides in Pyricular grisea caused by mutations in the mitochondria l cytochrome b gene. Phytopathology 93:891 900. 88. Kim, K. S., Wang, T. C., and Yang, X. B. 2005. Simulation of apparent infection rate to predict severity of soybean rust using a fuzzy logic system. Phytopathology 95:1122 1131. 89. Kobe, B. and Deisenhofer, J. 1995. A structural basis of the interactions between Leucine rich repeats and protein ligands. Nature 374:183 186. 90. Koch, E., Ebrahim Nesbat, F., and Hoppe, H. H. 1983. Light and electron microscopic studies on the development of soybean rust ( Phakopsora pachyrhizi Syd.) in susceptible soybean leaves. Phytopathol. Z. 106:302 320. 91. Koch, E. and Hoppe, H. H. 1988. Development of infection structures by the direct penetrating soybean rust fungus ( Phakopsora pachyrhizi Syd.) on artificial membranes. J. Phytopa thol. 122:232 244. 92. Kochman, J. K. 1979. The effect of temperature on development of soybean rust. Aust. J. Agric. Res. 30:273 277. 93. Krause, R. A. and Massie, L. B. 1979. Predictive systems: modern approaches to disease control. Annual Review of Phytopatho logy 13:31 47. 94. Krupa, S., Bowersox, V., Claybrooke, R., Barnes, C. W., Szabo, L., Harlin, K., and Kurle, J. 2006 Introduction of Asian soybean rust urediniospores into the Midwestern United States A case study. Plant Dis. 90: 1254 1259. 95. Kuck, K. H. a nd Mehl, A. 2004. Prothiconazole: Sensitivity profile and anti resistance strategy. Pflanzenschutz Nachrichten Bayer 57:225 236. 96. Kumudini, S., Godoy, C. V., Board, J. E., Omielan, J., Tollenaar, M. 2008. Mechanisms Involved in Soybean Rust Induced Yield R eduction. Crop Sci. 48:2334 2342. 97. Kwok, I. M. Y. and Loeffler, R. T. 1993. The biochemical mode of action of some newer azole fungicides. Pestic. Sci., 39:1 11.
154 98. Leister, R. T. and Katagiri, F. 2000. A resistance gene product of the nucleotide binding site Leucine rich repeats class can form a complex with bacterial avirulence proteins in vivo. Plant J. 22:345 354. 99. Lemes, E. M., Mackowiak, C. L., Blount, A., Marois, J. J., Wright, D. L., Coelho, L., and Datnoff, L. E. 2011. Effects of silicon application on s oybean rust development under greenhouse and field conditions. Plant Dis. 95:317 324. 100. Levy, C. 2003. A field evaluation of eight fungicides in the control of soybean rust in Zimbabwe during 2002 03. Commercial Oilseeds Producer Association, Harare, Zimbab we. 101. Levy, C. 2005. Epidemiology and chemical control of soybean rust in southern Africa. Plant Dis. 89: 699 647. 102. Li, X., Esker, P. D., Pan, Z., Dias, A. P., Xue, L., and Yang, X. B. 2010. The Uniqueness of the Soybean Rust Pathosystem: An Improved Unders tanding of the Risk in Different Regions of the World. Plant Dis 94:796 806. 103. Lin, S. Y. 1966. Studies on the physiologic races of soybean rust fungus, Phakopsora pachyrhizi Syd. J. Agric. Res. (Taiwan) 15:24 28. 104. Livingston, M., Johansson, R., Daberkow, S ., Roberts, M., Ash, M., and Breneman, V. 2004. Economic and policy implications of wind borne entry of asian soybean rust into the United States. U. ERS. Editor. Online publication. 105. Lynch, T. N., Marois, J. J., Wright, D. W., Harmon, P. F., Harmon, C. L. Miles, M. R., and Hartman, G. L. 2006. First report of soybean rust caused by Phakopsora pachyrhizi on Phaseolus species in the United States. Plant Dis. 90:970. 106. Mantecon, J. 2001. Timing of fungicide applications for the control of soybean brown spot F & N Tests 60:FC126. 107. Marchall, D. S., and Rush, M. C. 1980. Infection cushion formation on rich sheaths by Rhizoctonia solani Phytopathology 70:947 950 108. Marchetti, M. A., Urecker, F. A., and Bromfield, K. R. 1975. Uredial development of Phakopsora pachyr hizi in soybeans Phytopathology 65:822 823. 109. Marchetti, M. A., Melching, J. S., and Bromfield, K. R. 1976. The effect of temperature and dew period on germination and infection by urediniospores of Phakopsora pachyrhizi Phytopathology 66:461 463. 110. Marta, A. D., Di Stefano, V., Cerovic, Z. G., Agati, G., and Orlandini, S. 2008. Solar radiation affects grapevine susceptibility to Plasmopara viticola Sci. Agric. 65:65 70.
15 5 111. McLean, R. J. 1979. Histological studies of resistance to soybean rust, Phakopsora pach yrhizi Syd. Aust. J. Agr. Res. 30:77 84. 112. McLean, R. J., and D. E. Byth. 1980. Inheritance of resistance to rust ( Phakopsora pachyrhizi ) in soybeans. Aust. J. Agric. Res. 31:951 956. 113. M cWilliams, D. A., Berglund, D. R., and Endres, G. J. 2004. Soybean Growt h and Management Quick Guide. Online Publication: NDSU Agriculture and University Extension. A 1174. 114. Melching, J. S., Dowler, W. M., Koogle, D. L., and Royer, M. H. 1989. Effects of duration, frequency, and temperature of leaf wetness periods on soybean r ust. Plant Dis. 73:117 122. 115. Meyers, B. C., Dickerman, A. W., Michelmore, R. W., Sivaramakrishnan, S., Sobral, B. W., and Young, N. D. 1999. Plant disease resistance genes encode members of an ancient and diverse protein family within the nucleotide bindin g superfamily. Plant J. 20:317 332. 116. Meyers B C Kozik A Griego A Kuang H and Michelmore R W 2003 Genome wide analysis of NBS LRR encoding genes in Arabidopsis. Plant Cell 15:809 834 117. Meyer, J. D. F., Silva, D. C. G., Yang, C., Pedley, K. F., Z hang, C., van de Mortel, M., Hill, J. H. Shoemaker, R. C., Abdelnoor, R. V., Whitham, S. A., and Graham, M. A. 2009. Identification and analyses of candidate genes for Rpp4 mediated resistance to Asian soybean r ust in soybean. Plant Physiol. 150:295 307. 118. Miles M. R., Frederick, R. D. and Hartman, G. L. 2003. Soybean Rust: Is the U.S. Soybean Crop At Risk? APSnet Features. Online. doi: 10.1094/APSnetFeature 2003 0603 119. Miles M Frederick R and Hartman G 2006 Evaluation of soybean germplasm for resistance t o Phakopsora pachyrhizi. Plant Health Prog 10: doi/ 10.1094/PHP 2006 0104 01 RS 120. Miles, M. R., Hartman, G. L., Levy, C., and Morel, W. 2003. Current status of soybean rust control by fungicides Pesticide Outlook:197 200. 121. Miles, M. R., Morel, W., and H artman, G. 2003. Summary of the USDA fungicide efficacy trials to control soybean rust in Paraguay 2002 2003. p. 1 8.
156 122. Miles, M. R., Morel, W., Steinlage, T., and Hartman, G. L. 2004. Summary of the USDA fungicide efficacy trials to control soybean rust in Paraguay 2003 2004. p. 1 11. 123. Miles, M. R., Levy, C., Morel, W., Meuller, T., Steinlage, T., van Rij, N., Frederick, R. D., and Hartman, G. L. 2007. International fungicide efficacy trials for the management of soybean rust. Plant Dis. 91:1450 1458. 124. M iller, R. N. G. Bertioli, D. J., Baurens, F. C., Santos, C. M. R., Alves, P. C., Martins, N. F., Togawa, R. C., Souza Junior, M. T., and Pappas Junior, G. J. 2008. Analysis of non TIR NBS LRR resistance gene analogs in Musa acuminate Colla: isolation, RFLP marker development, and physical mapping. BMC Plant Biol. 8:15 doi:10.1186/1471 2229 8 15. 125. Modolo, L. V., Cunha, F. Q., Braga, M. R., and Salgado, I. 2002. Nitric oxide synthase mediated phytoalexin accumulation in soybean cotyledons in response to the Diap orthe phaseolorum f. sp. meridionalis elicitor. Plant Physiol. 130:1288 1297. 126. Moffett, P., Farnham, G., Peart, J., and Baulaombe, D. C. 2002. Interaction between domains of a plant NBS LRR protein in disease resistance related cell death. EMBO Journal 21: 4511 4519. 127. Monteros, M. J., A. M. Missaoui, D. V. Phillips, D. R. Walker, and Boerma, H. R 2007. Mapping and confi brown lesion resistance gene for Asian soybean rust. Crop Sci. 47:829 836. 128. Morel W. and Yorinori J. 2002. Ministe rio de Agricultura y Ganaderia, Centro Regional de Investigacion Agri cola, Capitan Miranda, Paraguay Bol. Divulg. No. 44 129. Mueller, D. S., Dorrance, A. E., Derksen, R. C., Ozkan, E., Kurle, J. E., Grau, C. R., Gaska, J. M., Hartman, G. L., Bradley, C. A., and Pedersen, W. L. 2002. Efficacy of fungicides on Sclerotinia sclerotiorum and their potential for control of Sclerotinia stem rot on soybean Plant Dis. 86:26 31. 130. Wright, D. L. 2008. Evaluation of fungicides and fungicide timing for the control of soybean rust in northwest Florida, 2006. Plant Disease Management Reports (online). Report 2:FC086. doi :10.1094/PDMR02. American Phytopathological Society, St. Paul, MN. 131. Mueller T. A., Miles, M. R., Morel, W., Marois, J. J., Wright, D. L., Kemerait, R. C., Levy, C., and Hartman, G. L. 2009. Effect of fungicide and timing of application on soybean rust severity and yield. Plant Dis. 93:243 248.
157 132. NAPDFC. 2006. North American Plant Disease Forecast Center. Soybean Rust Forecast Homepage. Online publication. 133. Narvez, D. F., Jurick, W. M., Marois, J. J., and Wright, D. L. 2010. Effects of surface wetness periods on development of soybean rust under field conditions. Plant Dis. 94:258 264. 134. National Statistics for Soybeans National Agricultural Statistics Service (NASS). Agricultural Statistics Board USDA. 2010. Online publication 135. Nguyen, H. T. 2011. Seed Composition. Molecular Genetics & Soybean Genomics Laboratory, University of Missouri. O nline publication 136. NOAA. 2006. National Oceanic and Atmospheric Administration. Atmospheric Research Laboratory. NOAA ARL HYSPLIT Model. Online publication. 137. NPAG 2002. NPAG Data: Phakopsora pachyrhizi Australasian soybean rust. Exotic Fungus N ot in the Continental United States. p. 1 14. 138. Nybom, H. and Bartish, IV. 2000. Effects of life history traits and sampling strategies on genetic diversity estimates obtained with RAPD markers in plants. Perspectives in Plant Ecology, Evolution and Systema tics 3:93 114. 139. Ogle, H., Byth, D., and Mclean, R. 1979. Effect of rust ( Phakopsora pachyrhizi ) on soybean yield and quality in Southeastern Queensland. Australian J. of Ag. Res. 30:883 893. 140. Ono, Y., Buritic, P., and Hennen, J. F. 1992. Delimitation of Phakopsora Physopella and Cerotelium and their species on Leguminosae Mycol. Res. 96:825 850. 141. Ozkan, H. E., Zhu, H., Derksen, R. C., Guler, H., and Krause, C. R. 2006. Evaluation of spraying equipment for effective application of fungicides to control Asian soybean rust. Aspects of App. Bio. 77:423 431. 142. Pan, Z., Yang, X. B., Pivonia, S., Xue, L., Pasken, R., and Roads, J. 2006. Long term prediction of soybean rust entry into the continental United States. Plant Dis.90: 840 846. 143. Pan, Z., X ue, L., Del Po nte, E. M., Andrade, D., Li, X., Pasken, R. and Yand, X. B. 2005. Soybean rust dispersal prediction and analysis in the U.S. for 2005 growing season. Proceedings, National Soybean Rust Symposium, Nashville, TN, USA. Online publication.
158 144. Pappert, R. A., Ham rick, J. L., and Donovan, L. A. 2000. Genetic variation in Pueraria lobata (Fabaceae), an introduced, clonal, invasive plant of the southeastern United States. Am. J. Bot. 87:1240 1245. 145. Park, S., Chen, Z. Y., Chanda, A. K., Schneider, R. W., & Hollier, C. A. 2008. Viability of Phakopsora pachyrhizi urediniospores under simulated southern Louisiana winter temperature conditions. Plant Dis 92 : 1456 1462. 146. P arry, D. W. (1990). Plant pathology in agriculture. New York City, NY: Cambridge University Press. 147. Pedersen, P. 2007. Soybean planting date Soybean Extension and Research Program, Department of Agronomy, Iowa State University Online publication 148. Pedersen, W. L. and Leath S 1988. Pyramiding major genes for resistance to maintain residual eff ects. A nnu. Rev. Phytopathol. 26:369 378. 149. Pivonia, S. and Yang, X. B. 2004. Assessment of potential year round establishment of soybean rust throughout the world. Plant Dis. 88:523 529. 150. Pivonia, S. and Yang, X. B. 2006. Relating epidemic progress from a general disease model to seasonal appearance time of rusts in the United States: implications for soybean rust. Phytopathology 96:400 407. 151. Pivonia, S., Yang, X. B., and Pan, Z. 2005. Assessment of epidemic potential of soybean rust in the United States Plant Di s. 89:678 682. 152. Pretorius Z. A. Kloppers F. J. and Frederick R. D. 2001. First Report of Soybean Rust in South Africa Plant Dis 85: 1288. 153. Reed, K. L., Hamerly, E. R., Dinger, R. F., and Jarvis, P. G. 1976. An analytical model for field measurements of photosynthesis. J. Appl. Ecol. 13:925 942. 154. Reis, E. M., Sartori, A. F., and Camara, R. K. 2004. Modelo climatic para a previs o da ferrugem da soja. Summa Phytopathologica 30:290 292. 155. Ribeiro, A. S., Moreira, J. U. V., Pierozzi, P. H. B., Rachid, B. F., de Toledo, J. F. F Arias, C. A. A., Soares, F. M., and Godoy, C. V. 2007. Genetic control of Asian rust in soybean. Euphytica 157:15 25. 156. Richly E Kurth J Leister D 2002 Mode of amplification and reorganization of resistance genes during recent Arabidopsis thaliana evolution. Mol Biol Evol 19:76 84
159 157. Rodrguez A. C., Rios J. A., and Hernndez J. R 2006. First Report of Asian Soybean Rust Caused by Phakopsora pachyrhizi from Mexico Plant Dis 90: 1260. 158. Rossi R. L 2003. First Report of Phakopsora pachyrhizi the Causal Organism of Soybean Rust in the Province of Misiones, Argentina. Plant Dis 87: 102. 159. Rotem, J., Wooding, B., and Aulor, D. E. 1985. The role of solar radiation, especially ultraviolet, in the mortality of fungal spores. Phytopathology 75: 510 514 160. Rytter, J. L., Dowler, W. M., and Bromfield, K. R. 1984. Additional alternative hosts of Phakopsora pachyrhizi causal agent of soybean rust. Plant Dis. 68:818 81 9. 161. Sakai A K Allendorf F W Holt J S et al. 2001 The population biology of invasive species. Ann Rev Ecol Syst 32, 305 332. 162. Saksirirat, W. and Hoppe, H. H. 1991. Teliospore germination of soybean rust fungus (Phakopsora pachyrhizi Syd.). J Phytopathol. 32:339 342. 163. S angawongse, P., Kittisin, S., and Nunthapun, M. 1977. Chemical control of soybean rust in Thailand. Thai. J. Agric. Sci. 10:1 8. 164. Scherm, H. and van Bruggen, A. H. C. 1993. Sensitivity of simulated dew duration to meteorologic al variations in different climatic regions of California. Agric. For. Meteorol. 66:229 245. 165. Scherm, H. and van Bruggen, A. H. C. 1994. Effects of fluctuating temperatures on the latent period of lettuce downy mildew ( Bremia lactucae ). Phytopathology 84:8 53 859. 166. Scherm, H. and van Bruggan, A. H. C. 1994. Global warming and nonlinear growth: how important are changes in average temperature? Phytopathology 84:1380 1384. 167. Schneider R W Hollier C A and Whitam H K. 2005. First report of soybean rust ca used by Phakopsora pachyrhizi in the continental United States. Plant Dis 89:774 168. Schneider, R. W., Robertson, C. A., Giles, C. G., and Mumma, E. P. 2005. Evaluation of fungicides, registered to control Asian soybean rust, for the control of Cercospora le af blight F & N Tests 61:FC015. 169. Sconyers, L. E., Kermerait, R. C., Brock, J., Phillips, D. V., Jost, P. H., Sikora, E. J., Gutierrez Estrada, A., Mueller, J. D., Marois, J. J., Wright, D. L., and Harmon,
160 C. L. 2006. Asian soybean rust development in 2005 : A perspective from the Southeastern United States. APSnet Feature. Online publication. 170. Sconyers, L. E., Kemerait, Jr. R. C., Brock, J. H., Gitaitis, R. D., Sanders, F. H., Phillips, D. V., and Jost, P. H. 2006. First report of Phakopsora pachyrhizi, the causal agent of Asian soybean rust, on Florida beggar weed in the United States. Plant Dis. 90:972. 171. Sharma, N. D., and Mehta, S. K. 1996. Soybean rust in Madhya Pradesh. Acta Botanica Indica 24:1150116. 172. Shibles, R., Anderson, I. C., and Gibson, A. H. 19 75. Soybean. Pages 151 189 in: Crop physiology: Some case histories, L. T. Evans, ed. Cambridge University Press, NY. 173. Sierotzki, H., Wullschleger, J., and Gisi, U. 2000. Point mutation in cytochrome b gene conferring resistance to strobilurin fungicides i n Erysiphe graminis f. sp. tritici field isolates. Pesticide Biochemistry and Physiology 68:107 112. 174. Sikora, E. J., Delaney, D. P., Delaney, M. A., Lawrence, K. S., and Pegues, M. 2009. Evaluation of sequential fungicide spray programs for control of soyb ean rust. Online. Plant Health Progress doi: 10.1094/PHP 2009 0402 01 RS. 175. Silva, D. C. G, Yamanaka, N., Brogin, R. L., Arias, C. A. A., Nepomuceno, A. L., Mauro, A. D., Pereira, S. S., Noguiera, L. M., Passianotto, A. L. L., and Abdelnoor, R. V. 2008. Molecular mappi ng of two loci that confer resistance to Asian rust in soybean. Theor. Appl. Genet. 117:57 63. 176. Sinclair, J. B. and Hartman, G. L. eds. 1996. Soybean rust workshop. College of Agricultural, Consumer, and Environmental Sciences. National Soybean Research La boratory Pub. n. 1. 177. Skoss, J. D. 1955. Structure and composition of the plant cuticle in relation to environmental factors and permeability. Bot. Gaz .117:55 72 178. Slaminko, T. L., Miles, M. R., Frederick, R. D., Bonde, M. R., and Hartman, G. L. 2008. New l egume hosts of Phakopsora pachyrhizi based on greenhouse evaluations. Plant Dis. 92:767 771. 179. Smits, M. M., Slaminko, T. L., Miles, M. R., Frederick, R. D., Bonde, M. R., and Hartman, G. L. 2008. New legume hosts of Phakopsora pachyrhizi based on greenhou s e evaluations. Plant Dis. 92 :767 771. 180. Soares, R. M. 2007. Balano da ferrugem asia tica na safra 2006/07 com base nos nu meros do Sistema de Alerta. In: Anais do Simpo sio Brasileiro de
161 Ferrugem Asia tica da Soja. Se rie Documentos 281, Embrapa. Lond rina, Brazil, pp. 9 13. 181. Sojka, R. E., and Parsons, J. E. 1983. Soybean w ater and c anopy m icroclimate r elationships of f our r ow s pacings. Agron. J. 75:961 968 182. Southerst, R. W. and Maywald, G. F. 1985. A computerized system for matching climates in ecolog y. Agriculture, Ecosystems and Environment 13:281 299. 183. Soybean and Corn Advisor. 2011. Soybean free period in Brazil Starts Online publication. 184. Soy Stats. 2010. Soy Statistics and U.S. Soy Crop Statistics. Online publication 185. Srivastava, P ., George, S ., Marois, J J. Wright, D L. and Walker D R. 2011. Saccharin induced systemic acquired resistance against rust ( Phakopsora pachyrhizi ) infection in soybean: Effects on growth and development Crop Prot. 30:726 732. 186. Steinmuller, D., and Tevini, M., 19 85. Action of ultraviolet radiation (UV B) upon cuticular waxes in some crop plants. Planta 164, 557 564. 187. St ovold, G. E., and Smith, H. J. P. 1991. The prevalence and severity of diseases in the coastal soybean crop of New South Wales. Aust. J. Exp. Agric. 31:545 550. 188. Sun, C. N. 1955. Growth and development of primary tissues in aerated and non aerated roots of soybean. Bull. Torr. Bot. Club 82:491 502. 189. Sun, J. H., Li, Z. C., Jewett, D. K., Britton, K. O., Ye, W. H., and Ge, X. J. 2005. Genetic diversity of Pueraria lobata (kudzu) and closely related taxa as revealed by inter simple sequence repeat analysis. Eu r. Weed Res. Soc. Weed Res. 45:255 260. 190. Tameling, W. I. Elzinga, S. D. Darmin, P. S., Vossen, J. H. Takken, F. L., Haring, M. A., and Cornelissen, B. J. 2002. The tomato R gene products I 2 and MI 1 are functional ATP binding proteins with ATPase activity Plant Cell 14:2929 2939. 191. Tamural, K., Peterson, D., Peterson, N., Stecher, G., Nei, M., and Kumar, S. 2011. MEGA5: molecular evolutionary genetics analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods. Mol. Biol. Evol. 28:2731 2739. 192. Tan, Y. J., Yu, Z. L., and Yang, C. Y. 1996. Soybean Rust. China Agriculture Press, Beijing, China.
162 193. Teng, P. S. 1985. A comparison of simulation approaches to epidemic modeling. Annual Review of Phytopathology 23:351 379. 194. Thompson, J. D., H iggins, D. G., and Gibson, T. J. 1994. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position specific gap penalties and weight matrix choice. Nucleic Acids Res. 22:4673 4680. 195. Thornton, M. R. 20 01. Arthropod fauna associated with kudzu ( Pueraria lobata Willd.) in North Carolina. M.S. thesis, North Carolina State University, Raleigh, North Carolina, USA. 196. Traut, T. W. 1994. The functions and consensus motifs ofnine types of peptide3 segments that f orm different types of nucleotide binding sites. Eur J Biochem. 222:9 19. 197. Tschanz, A. T. and Tsai, B. 1982. Effect of maturity on soybean rust development. Soybean Rust Newsletter 5:38 41. 198. Tschanz, A. T. and Shanmugasundaram, S. 1985. Soybean rust. Page s 562 567 in: III Proceedings of World Soybean Research Conference, R. Shibles ed. Westview Press. Boulder, CO. 199. Tsutsui N D Suarez A V Holway D A and Case T J 2000 Reduced genetic variation and the success of an invasive species. Proceedings of the National Academy of Sciences of the USA 97, 5948 5953. 200. Twizeyimana, M. Ojiambo, P. S., Ikotun, T., Paul, C., Hartman, G. L., and Bandyopadhyay, R. 2007. Comparison of field, greenhouse, and detached leaf evaluations of soybean germplasm for resistance to Phakopso ra pachyrhizi Plant Dis. 91:1161 1169. 201. U.S. Soy Production. American Soybean Association World Initiate for Soy in Human Health. 2010. Online publication. 202. USDA. 2010. United State Soybean Industry: Background Statistics and Information. Online publica tion. 203. van der Maesen L. J. G. 1985 Revision of the genus Pueraria DC. With some notes on Teyleria Backer. Agricultural University of Wagenigen Paper 85 1. 204. van de Mortel M Recknor J C Graham M A Nettleton D Dittman J D Nelson R T God oy C V Abdelnoor R V Almeida A M R and Baum T J 2007 Distinct biphasic mRNA changes in response to Asian soybean rust infection. Mol Plant Microbe In 20: 887 899
163 205. V idhyasekaran, P. 2004. Concise encyclopedia of plant pathology Food Produ cts Press and The Haworth Reference Press, Binghamton, N.Y. 206. Vincelli, P. 2002. QoI (Strobilurin) fungicides: benefits and risks. The Plant Heath Instructor DOI: 10.1094/PHI I 2002 0809 02 207. Walker, D. R., Boerma, H. R., Phillips, D. V., Schneider, R. W., Buck ly, J. B., Shipe, E. R., Mueller, J. D., Weaver, D. B., Sikora, E. J., Moore, S. H., Hartman, G. L., Miles, M. R., Harris, D. K., Wright, D. L., Marois, J. J., and Nelson, R. L. 2011. Evaluation of USDA soybean germplasm accessions for resistance to soybean rust in th e southern United States. Crop Sci. 51:678 693. 208. Ward, E. W. B., and Buzzell, R. I. 1983. Influence of light, temperature and wounding on the expression of soybean genes for resistance to Phytophthora megasperma f. sp. glycinea Physiol. Plant Pathol. 23:4 01 409. 209. Ward, D. B. 1998 Pueraria montana: the correct scientific name of the kudzu. Castanea 63 : 76 77. 210. Ward, N. A., Schneider, R. W., and Aime, M. C. 2011. Colonization of soybean rust sori by Simplicillium lanosoniveum Fungal Eco. 4:303 308. 211. Ward, N. A. 2011. Simplicillium lanosoniveum a mycoparasite of Phakopsora pachyrhizi and its use as a biological control agent. Dissertation. Louisiana State University. 212. Watson S. R., Giesler L. J. Ziems A. D. and Brovont T. E 2008. Report of Soybean Rust Caused by Phakopsora pachyrhizi in Nebraska. Plant Dis 92: 1588. 213. Webster, J. and Weber, R. Introduction to Fungi. 3 rd Ed. Cambridge, UK: Cambridge University Pr ess, 2007. 609 635. 214. Welles, J. M., and Cohen, S. 1996. Canopy structure measuring by gap fraction analysis using commercial instrumentation. J. Exp. Bot. 47:1335 1342 215. Williams, L. F. 1950. Structure and genetic characteristics of the soybean. Pages 11 13 4 in: Soybeans and soybean products, K. S. Markley, ed. Interscience Publishers, NY. 216. Wrather, J. A., Shannon, J. G., Stevens, W. E., Sleper, D. A., and Arelli, A. P. 2004. Soybean cultivar and foliar fungicide effects on Phomopsis sp. seed infections Pla nt Dis. 88:721 723. 217. Wright, D. L., Rich, J. R., Marois, J. J., Sprenkel, R. K., and Ferrell, J. A. 2002. Soybean Production in Florida. Online Publication: Electronic Data Information Source of UF/IFAS Extension (EDIS). SS AGR 182.
164 218. Wu, B. M., van Bruggen A. H. C., Subbarao, K. V., and Pennings, G. G. H. 2001. Validation of weather and leaf wetness forecasts for a lettuce downy mildew warning system. Can. J. Plant Pathol. 23:371 383. 219. Yang, X. B., Dowler, W. M., and Tschanz, A T. 1991. A simulation model for assessing soybean rust epidemics. Journal of Phytopathology 133:187 200. 220. Yang, X. B., Dowler, W. M., and Royer M. H 1991. Assessing the risk and potential impact of an exotic plant disease. Plant Dis 75:976 982 221. Yang, X. B., Royer, M. H., Tschanz, A. T., and Tsai, B. Y. 1990. Analysis and quantification of soybean rust epidemics from seventy three sequential planting experiments. Phytopathology 80:1421 1427. 222. Yang, X. B., Tschanz, A. T., Dowler, W. M., and Wang, T. C. 1991. Development of yield los s models in relation to reductions of components of soybean infected with Phakopsora pachyrhizi Phytopathology. 81:1420 1426. 223. Yeh, C. C., Tschanz, A. T., and Sinclair, J. B. 1981. Induced teliospore formation by Phakopsora pachyrhizi on soybeans and othe r hosts. Phytopathology 71:1111 1112. 224. Yorinori, J. T., Paiva, W. M., Frederick, R. D., Costamilan, L. M., Bertagnolli, P. F., Hartman, G. L., Godoy, C. V., and Nunes Jr, J. 2005. Epidemics of soybean rust ( Phakopsora pachyrhizi ) in Brazil and Paraguay fro m 2001 to 2003. Plant Dis 89:675 677. 225. Young, H. M., Marois, J. J, Wright, D. L., Narvez, D. Epidemiology of Soybean Rust in Soybean Sentinel Plots in Florida. Plant Dis.95:744 750. 226. Youssef, B., Subramaniam, R., and Dangl, J. L. 2004 Plant disease resistance protein signaling: NBS LRR proteins and their partners. Current Opinion in Plant Biology 7:391 399. 227. Zavala, J. A, Scopel, A. L., and Ballar, C. L. 2001. Effects of ambient UV B radiation on soybean crops: Impact on leaf herbivo ry by Anticarsia gemmatalis Plant Ecology 156:121 130. 228. Zhang, L. X., Kyei Boahen, S., Zhang, J., Zhang, M. H., Freeland, T. B., Watson, C. E., Jr., and Liu, X. M. 2007. Modifications of optimum adaptation zones for soybean maturity groups in the USA. On l ine. Crop Management. Sept. 2. doi:10.1094/CM 2007 0927 01 RS.
165 229. Zhou, T., Wang, Y., Chen, J. Q., Araki, H., Jing, Z., Jiang, K., Shen, J., and Tian, D. 2004. Genome wide identification of NBS genes in japonica rice reveals significant expansion of divergen t non TIR NBS LRR genes. Mol. Gen. Genomics. 271:402 415. 230. Zidak, N. K. and Backman, P. A. 1996. Biological control of kudzu ( Pueraria lobata ) with the plant pathogen Pseudomonas syringae pv. phaseolicola Weed Sci. 44:645 649.
166 BIOGRAPHICAL SKETCH Heather Marie Young, daughter of Dennis and Linda Young, was born in Orlando, Florida in 1984. Heather attended Auburndale high school and graduated valedictorian in 2003. While still in high school she attended Polk Community College and in 2004 was awarded and Associate of Arts in Business. She attended Florida State University where she majored in biological sciences and was awarded a Bachelor of Science degree in April 2007 and a Master of Science in science teaching in April 2008. In May 2008, Heather bega n working at University of Florida as a biological scientist at the North Florida Research and Education Center. Within a year of working in agricultural research she decided to obtain her doctorate and began her doctorate of philosophy in Plant Pathology in the fall of 2009. Under the guidance of Dr. James Marois and Dr. David Wright she conducted research on the biology of the obligate fungal parasite, Phakopsora pachyrhizi the casual agent of soybean rust. Heather was awarded multiple travel grants t o present her research at the American Phytopathological Society Annual Meeting in Oahu, Hawaii in summer 2011, at the CropWorld Global International Meeting in London, United Kingdom in f all 2011, and at the American Phytopathological Society Southern Div ision Meeting in Birmingham, Alabama in spring 2012 She was also awarded first place for her paper and oral presentation at the Southern Soybean Disease Workers conference in Pensacola, Florida in spring 2012. After her PhD, Heather plans on continuing her agricultural research career in academics.