Zoysiagrass Evaluation for Dna Content, Sting Nematode Response, Nitrogen Management, and Estimates of Heritability for Turfgrass Performance Traits

ZOYSIAGRASS EVALUATION FOR DNA CONTENT, STING NEMATODE RESPONSE, NITROGEN MANAGEMENT, AND ESTIMATES OF HERITABILITY FOR TURFGRASS PERFORMANCE TRAITS

BRIAN M. SCHWARTZ

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

2008

To Süz

ACKNOWLEDGMENTS

I would like to express my sincere appreciation to Dr. Kevin Kenworthy, the members of my advisory committee, the Agronomy Department, and the staff at the Plant Science Research and Education Unit for all of the guidance, support, and help. Thanks are also in order for the generous donation of laboratory equipment from Mark Kann and the Seven Rivers Golf Course

Superintendents Association. This research would not have been possible without these efforts.

TABLE OF CONTENTS

page

ACKNOWLEDGMENTS ...... 4

LIST OF TABLES...... 7

LIST OF FIGURES ...... 9

ABSTRACT...... 12

CHAPTER

1 INTRODUCTION ...... 14

2 VARIATION IN 2C NUCLEAR DNA CONTENT OF Zoysia spp. AS DETERMINED BY FLOW CYTOMETRY...... 17

Introduction...... 17 Materials and Methods ...... 20 Plant Materials...... 20 Flow Cytometry...... 20 Morphological Measurements...... 21 Statistical Analysis ...... 21 Results and Discussion ...... 22 Conclusions...... 24

3 EFFICIENT METHODOLOGY FOR SCREENING STING NEMATODE RESPONSE IN A TURFGRASS BREEDING PROGRAM ...... 28

Introduction...... 28 Materials and Methods ...... 30 Results...... 34 Discussion...... 36

4 VARIABLE RESPONSES OF ZOYSIAGRASS, ST. AUGUSTINEGRASS, AND BERMUDAGRASS GENOTYPES TO THE STING NEMATODE...... 46

Introduction...... 46 Materials and Methods ...... 48 Results and Discussion ...... 51 Conclusions...... 55

5 MOWING HEIGHT AND NITROGEN FERTILITY MANAGEMENT OF ZOYSIAGRASS IN FLORIDA ...... 59

Introduction...... 59 Materials and Methods ...... 61 Results and Discussion ...... 63 Zoysia japonica ...... 63 Zoysia matrella...... 66 Conclusions...... 69

6 HERITABILITY ESTIMATES FOR TURFGRASS PERFORMANCE AND STRESS RESPONSE IN Zoysia spp...... 97

Introduction...... 97 Materials and Methods ...... 100 Results and Discussion ...... 103 Conclusions...... 106

REFERENCES ...... 113

BIOGRAPHICAL SKETCH ...... 127

LIST OF TABLES

Table page

2-1 Means for 2C nuclear DNA content and leaf blade width of zoysiagrass genotypes for five species and five interspecific hybridizations ...... 26

3-1 Mean squares for Belonolaimus longicaudatus reproduction factor (Rf), final population density (Pf), population density on a total root length basis (Pf/TRL), population density on a total dry root basis (Pf/TDRW), total dry root weight (TDRW), total dry root weight percent reduction (TDRW % red.), total root length (TRL), total root length percent reduction (TRL % red.), fine root length (FRL), and fine root length percent reduction (FRL % red.) of TifEagle bermudagrass evaluated in three establishment (Est.) methods with different inoculation treatments (Inoc. TRT) in two experimental trials...... 41

3-2 Mean reproduction factor (Rf), final population density (Pf), population density on a total root length basis (Pf/TRL), population density on a total dry root basis (Pf/TDRW), of Belonolaimus longicaudatus on TifEagle bermudagrass 90 days after inoculation evaluated in three establishment methods with two inoculation treatments in two experimental trials...... 42

3-3 45-d and 90-d conetainer (above diagonal) and 90-d clay pot (below diagonal) correlation coefficients of Belonolaimus longicaudatus reproduction factor (Rf), final population density (Pf), population density on a total root length basis (Pf/TRL), population density on a total dry root weight basis (Pf/TDRW), total dry root weight (TDRW), total dry root weight percent reduction (TDRW % red.), total root length (TRL), total root length percent reduction (TRL % red.), fine root length (FRL), and fine root length percent reduction (FRL % red.) of TifEagle bermudagrass...... 43

3-4 Mean total dry root weight (TDRW), total dry root weight percent reduction (TDRW % red.), and total root length (TRL) of TifEagle bermudagrass 90 days after inoculation evaluated in three establishment methods with uninoculated and inoculated treatments in two experimental trials...... 44

3-5 Mean total root length percent reduction (TRL % red.), fine root length (FRL), and fine root length percent reduction (FRL % red.) of TifEagle bermudagrass 90 days after inoculation evaluated in three establishment methods with uninoculated and inoculated treatments in two experimental trials...... 45

4-1 Mean squares for Belonolaimus longicaudatus reproduction factor (Rf), total root length (TRL), and total dry root weight (TDRW) of six turfgrasses evaluated in two establishment (Est.) methods with uninoculated and inoculated treatments (Inoc. TRT) in two experimental trials...... 57

4-2 Mean reproduction factor (Rf) of Belonolaimus longicaudatus on six turfgrasses 90 days after inoculation with 50 B. longicaudatus, evaluated in 45-d conetainers and 90-d conetainers in two experimental trials...... 57

4-3 Mean total root length (TRL) and total dry root weight (TDRW) of six turfgrasses 90 days after inoculation evaluated in 45-d conetainers and 90-d conetainers with uninoculated (U) and inoculated (I) treatments in two experimental trials...... 58

5-1 Mean squares for turfgrass performance characteristics and thatch depth of four Zoysia japonica cultivars and four Zoysia matrella cultivars evaluated at two mowing heights and three nitrogen fertility rates near Gainesville, FL...... 71

5-2 Mean thatch depth of four Zoysia japonica cultivars when evaluated at two mowing heights and three nitrogen fertility rates in November of 2007 near Gainesville, FL...... 84

5-3 Mean thatch depth of four Zoysia matrella cultivars when evaluated at two mowing heights and three nitrogen fertility rates in November of 2007 near Gainesville, FL...... 84

6-1 Evaluation dates for turfgrass performance characteristics of zoysiagrass genotypes with very fine, fine, or coarse leaf texture visually rated during 2006, 2007, and 2008...... 108

6-2 Expected mean squares for turfgrass performance traits of zoysiagrass genotypes evaluated on one date...... 109

6-3 Expected mean squares for turfgrass performance traits of zoysiagrass genotypes evaluated on multiple dates...... 109

6-4 Variance component estimates, descriptive statistics, and broad-sense heritabilities (H2) for turfgrass performance characteristics of zoysiagrass genotypes with very fine, fine, or coarse leaf texture evaluated during 2006, 2007, and 2008...... 110

6-5 Variance component estimates, descriptive statistics, and broad-sense heritabilities (H2) for turfgrass color characteristics of zoysiagrass genotypes with very fine, fine, or coarse leaf texture evaluated during 2006 and 2007...... 111

6-6 Variance component estimates, descriptive statistics, and broad-sense heritabilities (H2) for turfgrass performance characteristics of zoysiagrass genotypes with very fine, fine, or coarse leaf texture evaluated during 2006 and 2007...... 112

LIST OF FIGURES

Figure page

5-1 Genetic color responses and significance of treatment effects in Empire zoysiagrass at two mowing heights (M) and three nitrogen levels (N) during 2007 and 2008. Fertilizer treatments were applied in March, May, August, and October of each year. Visual rating of five was considered to be the minimum acceptable value...... 72

5-2 Genetic color responses and significance of treatment effects in JaMur zoysiagrass at two mowing heights (M) and three nitrogen levels (N) during 2007 and 2008. Fertilizer treatments were applied in March, May, August, and October of each year. Visual rating of five was considered to be the minimum acceptable value...... 73

5-3 Genetic color responses and significance of treatment effects in Palisades zoysiagrass at two mowing heights (M) and three nitrogen levels (N) during 2007 and 2008. Fertilizer treatments were applied in March, May, August, and October of each year. Visual rating of five was considered to be the minimum acceptable value.....74

5-4 Genetic color responses and significance of treatment effects in Ultimate zoysiagrass at two mowing heights (M) and three nitrogen levels (N) during 2007 and 2008. Fertilizer treatments were applied in March, May, August, and October of each year. Visual rating of five was considered to be the minimum acceptable value...... 75

5-5 Turf density responses and significance of treatment effects in Empire zoysiagrass at two mowing heights (M) and three nitrogen levels (N) during 2007 and 2008. Fertilizer treatments were applied in March, May, August, and October of each year. Visual rating of five was considered to be the minimum acceptable value...... 76

5-6 Turf density responses and significance of treatment effects in JaMur zoysiagrass at two mowing heights (M) and three nitrogen levels (N) during 2007 and 2008. Fertilizer treatments were applied in March, May, August, and October of each year. Visual rating of five was considered to be the minimum acceptable value...... 77

5-7 Turf density responses and significance of treatment effects in Palisades zoysiagrass at two mowing heights (M) and three nitrogen levels (N) during 2007 and 2008. Fertilizer treatments were applied in March, May, August, and October of each year. Visual rating of five was considered to be the minimum acceptable value...... 78

5-8 Turf density responses and significance of treatment effects in Ultimate zoysiagrass at two mowing heights (M) and three nitrogen levels (N) during 2007 and 2008. Fertilizer treatments were applied in March, May, August, and October of each year. Visual rating of five was considered to be the minimum acceptable value...... 79

5-9 Turf quality responses and significance of treatment effects in Empire zoysiagrass at two mowing heights (M) and three nitrogen levels (N) during 2007 and 2008. Fertilizer treatments were applied in March, May, August, and October of each year. Visual rating of five was considered to be the minimum acceptable value...... 80

5-10 Turf quality responses and significance of treatment effects in JaMur zoysiagrass at two mowing heights (M) and three nitrogen levels (N) during 2007 and 2008. Fertilizer treatments were applied in March, May, August, and October of each year. Visual rating of five was considered to be the minimum acceptable value...... 81

5-11 Turf quality responses and significance of treatment effects in Palisades zoysiagrass at two mowing heights (M) and three nitrogen levels (N) during 2007 and 2008. Fertilizer treatments were applied in March, May, August, and October of each year. Visual rating of five was considered to be the minimum acceptable value...... 82

5-12 Turf quality responses and significance of treatment effects in Ultimate zoysiagrass at two mowing heights (M) and three nitrogen levels (N) during 2007 and 2008. Fertilizer treatments were applied in March, May, August, and October of each year. Visual rating of five was considered to be the minimum acceptable value...... 83

5-13 Genetic color responses and significance of treatment effects in Cavalier zoysiagrass at two mowing heights (M) and three nitrogen levels (N) during 2007 and 2008. Fertilizer treatments were applied in March, May, July, September, and November of each year. Visual rating of five was considered to be the minimum acceptable value.....85

5-14 Genetic color responses and significance of treatment effects in Diamond zoysiagrass at two mowing heights (M) and three nitrogen levels (N) during 2007 and 2008. Fertilizer treatments were applied in March, May, July, September, and November of each year. Visual rating of five was considered to be the minimum acceptable value...... 86

5-15 Genetic color responses and significance of treatment effects in Pristine zoysiagrass at two mowing heights (M) and three nitrogen levels (N) during 2007 and 2008. Fertilizer treatments were applied in March, May, July, September, and November of each year. Visual rating of five was considered to be the minimum acceptable value.....87

5-16 Genetic color responses and significance of treatment effects in Zeon zoysiagrass at two mowing heights (M) and three nitrogen levels (N) during 2007 and 2008. Fertilizer treatments were applied in March, May, July, September, and November of each year. Visual rating of five was considered to be the minimum acceptable value.....88

5-17 Turf density responses and significance of treatment effects in Cavalier zoysiagrass at two mowing heights (M) and three nitrogen levels (N) during 2007 and 2008. Fertilizer treatments were applied in March, May, July, September, and November of each year. Visual rating of five was considered to be the minimum acceptable value.....89

5-18 Turf density responses and significance of treatment effects in Diamond zoysiagrass at two mowing heights (M) and three nitrogen levels (N) during 2007 and 2008. Fertilizer treatments were applied in March, May, July, September, and November of each year. Visual rating of five was considered to be the minimum acceptable value.....90

5-19 Turf density responses and significance of treatment effects in Pristine zoysiagrass at two mowing heights (M) and three nitrogen levels (N) during 2007 and 2008. Fertilizer treatments were applied in March, May, July, September, and November of each year. Visual rating of five was considered to be the minimum acceptable value.....91

5-20 Turf density responses and significance of treatment effects in Zeon zoysiagrass at two mowing heights (M) and three nitrogen levels (N) during 2007 and 2008. Fertilizer treatments were applied in March, May, July, September, and November of each year. Visual rating of five was considered to be the minimum acceptable value.....92

5-21 Turf quality responses and significance of treatment effects in Cavalier zoysiagrass at two mowing heights (M) and three nitrogen levels (N) during 2007 and 2008. Fertilizer treatments were applied in March, May, July, September, and November of each year. Visual rating of five was considered to be the minimum acceptable value.....93

5-22 Turf quality responses and significance of treatment effects in Zeon zoysiagrass at two mowing heights (M) and three nitrogen levels (N) during 2007 and 2008. Fertilizer treatments were applied in March, May, July, September, and November of each year. Visual rating of five was considered to be the minimum acceptable value.....94

5-23 Turf quality responses and significance of treatment effects in Diamond zoysiagrass at two mowing heights (M) and three nitrogen levels (N) during 2007 and 2008. Fertilizer treatments were applied in March, May, July, September, and November of each year. Visual rating of five was considered to be the minimum acceptable value.....95

5-24 Turf quality responses and significance of treatment effects in Pristine zoysiagrass at two mowing heights (M) and three nitrogen levels (N) during 2007 and 2008. Fertilizer treatments were applied in March, May, July, September, and November of each year. Visual rating of five was considered to be the minimum acceptable value.....96

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy

ZOYSIAGRASS EVALUATION FOR DNA CONTENT, STING NEMATODE RESPONSE, NITROGEN MANAGEMENT, AND ESTIMATES OF HERITABILITY FOR TURFGRASS PERFORMANCE TRAITS

Brian M. Schwartz

December 2008

Chair: Kevin E. Kenworthy Major: Agronomy

Zoysiagrass (Zoysia spp.) use in landscapes and on golf courses has primarily occurred in

the transition zone of the United States. Variation associated with several abiotic, biotic, and

turfgrass performance characteristics is well documented. Laboratory, glasshouse, and field studies utilizing zoysiagrass germplasm were conducted from 2005 through 2008 to quantify 2C nuclear DNA content, sting nematode (Belonolaimus longicaudatus) response, effects of nitrogen fertilization rate and mowing height management, and broad-sense heritability estimates of turfgrass performance and stress related characteristics. All experiments were managed with supplemental irrigation. Genotypes from Z. minima and Z. matrella had the largest (0.96 pg) and smallest (0.77 pg) 2C nuclear DNA contents, respectively. The observed 0.19 pg spread between

zoysiagrass species was less than variation reported in other tetraploid warm-season grasses

within the same species. Total root lengths of ‘TifEagle’ bermudagrass (Cynodon dactylon [L.]

Pers. var. dactylon × C. transvaalensis Burtt-Davy) were 57 %, 55 %, and 31 % greater for

uninoculated treatments when compared to an average of the two sting nematode inoculated

treatments in the 45-d conetainers, 90-d conetainers, and 90-d clay pots, respectively.

Quantifying root damage using 45-d conetainers inoculated with 50 sting nematodes provided

reproducible results characteristic of those reported in other greenhouse and field evaluations.

Sting nematode populations multiplied on the evaluated zoysiagrass and St. Augustinegrass

(Stenotaphrum secundatum [Walt.] Kuntze) cultivars with reproduction factors ranging from 2.2

to 11.0. The experimental Zoysia germplasm line UFTZ exhibited greater tolerance to sting

nematode injury than other turfgrass cultivars and exhibited no total root length reduction under

sting nematode pressure. Significant total root length percent reductions were observed between

uninoculated and inoculated treatments for ‘Empire’ (-24%), ‘Cavalier’ (-29%), ‘Emerald’ (-

29%), TifEagle (-32%), and ‘Floratam’ (-37%) in 45-d conetainers. In the management study,

nitrogen rate had the greatest influence on turfgrass performance, but mowing height was

important during colder periods or in the presence of Bipolaris disease pressure. Turfgrass

density was not maintained in Empire or ‘Palisades’ at the lowest N rate. Bipolaris incidence

was noted on ‘Cavalier’ and ‘Zeon’, and had the most detrimental effect on turf quality at the

lower mowing height and highest N rate. ‘JaMur’, ‘Ultimate’, ‘Diamond’, and ‘Pristine’ all had

acceptable density at the low nitrogen rate, but often did not have adequate color to sustain turf quality. Genotypic variance largely contributed to the wide range in expressed phenotypic response in a set of 324 zoysiagrass germplasm lines for establishment, turf density, turf quality, genetic color, and seedhead density which resulted in higher broad-sense heritability estimates

(0.62 ≤ H2 ≤ 0.94). Fall dormancy and spring greenup were influenced more by the environment

and had lower heritabilities (0.32 ≤ H2 ≤ 0.58). Turf quality was also rated considering the

effects of glufosinate herbicide application, Bipolaris incidence, and mole cricket damage.

Large error variances and low broad-sense heritabilities were typical for these stress related

traits. Overall, the potential exists for combining desirable traits in superior clonally propagated

F1 zoysiagrass hybrids.

CHAPTER 1 INTRODUCTION

Zoysia Willdenow are perennial grasses assigned to the family Gramineae (Poaceae),

subfamily Chloridoideae, and tribe Zoysieae. They are native to the Pacific Rim countries and

the center of origin is speculated to be near southeastern Asia and Indonesia (Engelke and

Anderson, 2003). Anderson (2000) classified available specimens into 11 species based on

morphological variation and nuclear DNA restriction fragment length polymorphism (RFLP)

fingerprints. Many of these species can be managed as turfgrasses, producing stolons and

rhizomes, having pointed leaves of various widths with rolled vernation, and characterized by

individual, raceme inflorescences which appear spike-like.

Seed and plant samples of Zoysia japonica and Z. matrella were first introduced into the

United States in the early 20th century by United States Department of Agriculture (USDA)

researchers Frank N. Meyer and C.V. Piper (Childers and White, 1947; Grau and Radko, 1951).

Other notable United States Department of Agriculture-Agricultural Research Service (USDA-

ARS) scientists who have made collection trips on the Eurasia continent include G.W. Burton,

D.R. Dewey, W.W. Hanna, D.A. Johnson, and J.J. Murray (M.C. Engelke, personal communication, 2006). Plant breeders in the private sector have collected in collaboration with

overseas scientists in efforts to expand the germplasm base necessary for improved seed yields

(Samudio, 1996). The two most extensive zoysiagrass collection trips were made by M.C.

Engelke and J.J. Murray in 1982 and by M.C. Engelke, M.P. Kenna, C.M. Taliaferro, and R.J.

Tyrl in 1993. In total, over 1000 unique Zoysia accessions have been brought back to the United

States, many of which are still maintained at Texas A&M University. Genotypes were collected from as far north as 43ºN latitude to as far south as 9ºN latitude, indicating the adaptability of

zoysiagrass to many climates. Considerable variation was noted among accessions for color,

winter dormancy, shade tolerance, growth habits, seed set, leaf texture, disease resistance, and

growing conditions. Samples were collected from areas under intensive grazing, prolonged snow

cover, and from the edge of the ocean where plants were growing under constant salt spray

(Engelke, 2000; Engelke and Anderson, 2003; Murray and Engelke, 1983). Further discovery

and collection of germplasm will broaden genetic diversity, thereby contributing to reduced

vulnerability of future cultivars (Brede and Sun, 1995; Busey, 1977; Loch et al., 2005).

In 1951, the USDA-ARS and the United States Golf Association (USGA) Green Section

released a finer textured Zoysia japonica named ‘Meyer’ (Grau and Radko, 1951). This cultivar is often credited as the first zoysiagrass, although an improved Z. matrella was released as

‘FC13521’ in 1930 by the Alabama Agricultural Experiment Station in Auburn (Ruemmele and

Engelke, 1990). Ian Forbes studied the cytological and morphological variation among many zoysiagrasses and recommended that Z. japonica, Z. matrella, and Z. tenuifolia be considered varieties of one species based on their cross compatibility. ‘Emerald’ was released in 1955 as a result of these efforts and was described as a Z. japonica × Z. tenuifolia hybrid (Forbes et al.,

1955; Forbes, 1952), although the results of RFLP fingerprint analysis later suggested that it arose from a hybridization of Z. matrella × Z. pacifica (Anderson, 2000).

Very few improved cultivars were developed in the 30 years following the release of

Emerald, likely because of a lack in variable germplasm resources needed for plant

improvement. ‘Midwest’ was released from the Indiana Agricultural Experiment Station in 1963

by W.H. Daniel. The breeding efforts J.J. Murray of the USDA-ARS in Beltsville, Maryland

and V.B. Youngner at the University of California, Riverside eventually led to the release of

‘Belair’ (1985) and ‘El Toro’ (1986), respectively (Diesburg, 2000; Ruemmele and Engelke,

1990). Approximately 27 vegetatively propagated cultivars have since been released or patented,

although many are no longer, or never have been, readily available. Many challenges have

impeded the successful development of seeded zoysiagrass cultivars. These include small seed

yields, seed dormancy, poor seedling vigor, and loss of stand uniformity and quality over time.

More recent synthetic releases ‘Compadre’ and ‘Zenith’ do show improvement. Roughly

another 17 seeded zoysiagrasses have been developed and tested in the United States, but have

had little market success (Engelke and Anderson, 2003; Morris, 2000; Morris, 2006; Morris and

Shearman, 1995; Samudio, 1996; Unruh et al., 2007). Haydu et al. (2005) reported that 1% of

Florida’s sod production hectareage was planted with Zoysia spp.

Variation for many abiotic, biotic, and turfgrass performance characteristics has been

quantified in zoysiagrass. Genotypic differences in drought and rooting characteristics (Marcum

et al., 1995), response to low temperatures (Patton and Reicher, 2007), salinity (Qian et al.,

2000), shade (White and Engelke, 1990), and wear tolerance (Youngner, 1961) have been

observed. Levels of insect (Reinert et al., 1997) and plant-parasitic nematode (Busey et al.,

1982) resistance have been identified. Some disadvantages of these turfgrasses are slow

establishment rate (Busey and Myers, 1979) and divot recovery (Karcher et al., 2006), delayed

spring greenup (Gibeault et al., 1997), thatch accumulation (Soper et al., 1988), and disease

susceptibility (Green et al., 1993).

Documentation of such extensive variation indicates the potential for continued

improvement of these species. Therefore, the objectives of the following chapters are to (i)

investigate the range between Zoysia spp. for 2C nuclear DNA content, (ii) develop procedures

for efficient screening of sting nematode response, (iii) identify base management guidelines for

nitrogen fertilization and mowing of zoysiagrass in Florida, and (iv) evaluate germplasm for

variation and heritability of turfgrass performance traits.

CHAPTER 2 VARIATION IN 2C NUCLEAR DNA CONTENT OF Zoysia spp. AS DETERMINED BY FLOW CYTOMETRY

Introduction

Anderson (2000) suggests that the genus Zoysia is comprised of 11 species that vary with

respect to morphology and nuclear DNA constitution. Speciation appears to have occurred through geographic isolation (Kim, 1983; Weng et al., 2007) rather than by genetic changes associated with an increase or decrease in ploidy level (Forbes, 1952). All cytological studies of

Zoysia spp. have determined that 2n = 40 (Arumuganathan et al., 1999; Chen and Hsu, 1962;

Christopher and Abraham, 1974; Forbes, 1952; Murray et al., 2005; Tateoka, 1955), except for an unexplained report of a diploid plant, 2n = 20, collected from Sri Lanka (Gould and

Soderstrom, 1974). Forbes (1952) theorized that Zoysia spp. were diploid in nature, but did not

disregard the possibility that the basic chromosome number could be 5 or 10. Chen and Hsu

(1962) suspected that the basic chromosome number was 10 based on cytological studies in other

eragrostoid grasses. This was later confirmed by Gould (1968). Zoysiagrasses are currently

described as allotetraploids based on restriction fragment length polymorphism (RFLP) linkage

mapping and analysis (Yaneshita et al., 1999). The absence of anaphase bridges, fragments,

lagging chromosomes, univalents or multivalents during meiosis I in conjunction with the

formation of 95% viable pollen (Forbes, 1952) supports the allotetraploid classification.

Taxonomic assignment in Zoysia has been predominately weighted towards

morphological and molecular marker variation rather than by reproductive isolation as suggested

by Forbes et al. (1955) in accordance with the biological species concept (Mayr, 1948). Genetic

variability has been characterized through random amplification of polymorphic DNA (RAPD)

(Choi et al., 1997; Weng et al., 2007), RFLP (Anderson, 2000; Yaneshita et al., 1999; Yaneshita

et al., 1997), amplified fragment length polymorphism (AFLP) (Cai et al., 2004), and simple

sequence repeat (SSR) (Cai et al., 2005; Tsuruta et al., 2005) marker analyses to discern relatedness among zoysiagrass species. Cross-compatibility does exist between a number of the

Zoysia species (Forbes, 1952; Hong and Yeam, 1985), although Engelke and Anderson (2003) observed low germination percentages between a few interspecific combinations.

Flow cytometry (FCM) was first developed for analyzing animal cells. The techniques and instrumentation were modified for lysis of the plant cell wall and separation of the nucleus from plastid and mitochondrial DNA (Galbraith, 1990). This methodology provides rapid and accurate ploidy determination and DNA content analysis for plant breeding programs

(Arumuganathan and Earle, 1991a; Dolezel et al., 1989; Laat et al., 1987). A major application of FCM has been the general characterization of plant species for their DNA contents

(Arumuganathan and Earle, 1991b; Bennett and Leitch, 1995), but other uses include analysis of plant cell cycles (Galbraith et al., 1983), sex identification in dioecious plants (Costich et al.,

1991), the estimation of C-banded constitutive heterochromatin (Rayburn et al., 1992), and use as a taxonomic marker when the gain or loss of DNA is correlated with evolutionary relationships between species (Ohri, 1998).

Reports of variation in DNA content from Zoysia spp. are limited. In an analysis of native New Zealand grasses, Murray et al. (2005) concluded that in general, tropical grasses have lower DNA contents than those of temperate origin. Zoysia pauciflora ranked the lowest of all grasses surveyed with a 2C-value of 0.97 pg. Zoysia minima was recorded to have a genome size of 0.99 pg. Arumuganathan et al. (1999) reported the 2C nuclear DNA content of ‘Zenith’, a seeded Z. japonica Steud. cultivar, to be 0.86 ± 0.00 pg.

Flow cytometry has become a useful tool for plant breeders to characterize the nuclear

DNA content and ploidy level in other warm-season turfgrass species. Taliaferro et al. (1997)

developed an accurate method to determine ploidy level and genome size in bermudagrass

(Cynodon spp.). This methodology was used by Wu et al. (2006) to study the genetic diversity

of Chinese bermudagrass germplasm. Triploid, tetraploid, pentaploid, and hexaploid genotypes were identified and further characterized with AFLPs. Johnson et al. (1998) used FCM to distinguish diploid, tetraploid, and hexaploid buffalograss (Buchloe dactyloides [Nutt.] Engelm.)

germplasm for breeding purposes. In doing so they discovered three previously un-described

pentaploid clones. Jarret et al. (1995) surveyed the nuclear DNA contents of 35 Paspalum species and determined that there was insufficient variation to discriminate between genotypes within a species, but that ploidy evaluation was accurate and repeatable. Vaio et al. (2007)

observed genome sizes in natural dallisgrass (Paspalum dilatatum Poir.) tetraploids that were less than twice the size of their diploid progenitors. This reduction in DNA content was not present in synthetically created allotetraploid dallisgrasses. This phenomenon, documented in other polyploid species, was credited to “genome downsizing” (Leitch and Bennett, 2004).

Genome analysis by FCM has been classified as a rapid and reliable method to identify cytotypes with varying ploidy levels in Kentucky bluegrass (Poa pratensis L.) (Barcaccia et al.,

1997; Huff and Bara, 1993). Eaton et al. (2004) were able to distinguish true hybrids derived

from both intra- and interspecific hybridizations, and Wieners et al. (2006) differentiated genotypes with the four major reproductive pathways in Kentucky bluegrass according to the presence and position of peaks. Other cool-season grasses in which ploidy level variation has been researched utilizing FCM include fine fescue (Festuca spp.) (Huff and Palazzo, 1998), ryegrass (Lolium spp.) (Barker et al., 2001), and bentgrass (Agrostis spp.) (Bonos et al., 2002).

Flow cytometry has been shown to reliably characterize nuclear DNA content and ploidy level in plants. Considering that these data are largely absent for members of the Zoysia genus,

research was initiated to (i) characterize the variation in 2C nuclear DNA content of available cultivars and experimental lines representing true, or interspecific hybrids between, Zoysia japonica, Z. macrantha, Z. matrella, Z. minima, Z. pacifica, and Z. pauciflora species through flow cytometry, and (ii) determine if sufficient variation exists between 2C nuclear DNA content within and among Zoysia species to statistically identify intermediate, mid-parent nuclear DNA contents from F1 hybrids.

Materials and Methods

Plant Materials

Zoysiagrass materials consisted of 20 cultivars and 16 experimental lines from five species and five interspecific hybridizations (Table 2-1). Genotypes were assigned to their respective species based on morphology and not genotypic characterization. Each genotype was vegetatively propagated into three, five cm pots containing a 50% sand, 50% Metro-Mix 250

(Scotts-Sierra Horticultural Products Co., Marysville, OH) growing medium in January, 2008.

Plants were organized in a completely randomized design with three replications and grown in a glasshouse in Gainesville, FL with supplemental overhead irrigation. Fertilizer was applied monthly with Peters Professional 20-20-20 General Purpose Water Soluble Fertilizer (Scotts-

Sierra Horticultural Products Co., Marysville, OH) at a rate of 2.4 g m-2. Leaf canopies were trimmed twice monthly except for overhanging, aerial stolons.

Flow Cytometry

Flow cytometry analysis was initiated when at least one aerial stolon was present in each pot. The terminal node and tip from one stolon was removed from each experimental unit and stored on ice prior to FCM analysis. A CyStain® PI Absolute P (05-5002, Partec North

America, Inc., Mt. Laurel, NJ) nuclei extraction and DNA staining kit was used to prepare samples using a technique modified from Arumuganathan and Earle (1991a). The terminal end

of each stolon was chopped with a razor blade and scalpel in a glass petri dish with 500 µL of extraction buffer for approximately 60 s. The resulting solution, containing isolated nuclei, was pipetted into a five mL test tube through a 50 µm nylon mesh filter cap. One drop of chicken red blood cell (CRBC) standard (BioSure® Inc., Grass Valley, CA) was then added to the test tube followed by two mL of the propidium iodide (PI) based DNA staining solution. Samples were incubated at 4°C for 60 min in the dark. Flow cytometry analysis was conducted at the

University of Florida’s Interdisciplinary Center for Biotechnology Research on a LSR-II cytometer (BD Biosciences, San Jose, CA) using a 100 milliwatt solid-state laser emitting at 488

nm to excite the PI. DNA peak data, based on 10,000 scanned particles, were quantified using

FACS DiVa v5.2 software (BD Biosciences, San Jose, CA). The mean 2C nuclear DNA content of each sample, measured in picograms, was adjusted based on the peak of the CRBC internal standard by taking the ratio of the plant sample peak mean and the CRBC peak mean, then multiplying by the 2C nuclear DNA content of the internal standard, i.e., 2.5 pg (Rasch, 1985;

Tiersch et al., 1989). Genotypic 2C nuclear DNA contents are means of three replicates.

Morphological Measurements

Zoysiagrasses are commonly classified by their leaf width. A digital caliper was used to measure the widest section of three fully expanded and mature leaf blades in each experimental unit. The average value of the sub-samples in each experimental unit was used for further analysis.

Statistical Analysis

The distribution of data for both 2C nuclear DNA content and leaf blade width was assessed with a histogram and normal probability plot for normality. An analysis of variance was performed on each of the measured traits to test whether genotypes, species determined by morphology, and where appropriate, genotypes within species varied. Genotype and species

means were separated using a Waller-Duncan k-ratio LSD when the main effect was found to be

significant. Pearson correlation coefficients were computed to test whether 2C nuclear DNA

contents were associated with leaf blade widths.

Results and Discussion

Consistent extraction and staining of nuclei was not achieved using young, freshly

harvested leaves according to standard protocol (Arumuganathan and Earle, 1991a). This is

likely due to the high silica content in the leaves (Ruemmele and Engelke, 1990) which renders

physical chopping very difficult. Tissue from root and rhizome tips also resulted in inconsistent

FCM analysis. Only analysis of plant material from the terminal node and tip of a stolon yielded

consistent and repeatable results, producing peak CVs < 6.0% for all genotypes.

Genotypes varied (P = 0.001) for 2C nuclear DNA content. ‘Zenith’ (0.83 pg/2C) and

5194-5 (0.96 pg/2C), a representative of Z. minima, were both ~3% lower than found by

Arumuganathan et al. (1999) and Murray et al. (2005), respectively (Table 2-1). These small differences could be attributed to sample preparation or between-lab instrumental variation

(Dolezel et al., 1998). Zoysia matrella was the only species in which genotypes were significantly different (P = 0.01) for 2C nuclear DNA content. The low measured values for

Diamond and ‘Zorro’ were largely responsible for these differences.

The observed spread between Diamond, Z. matrella with 0.77 pg/2C nucleus, and 5194-

5, Z. minima with 0.96 pg/2C nucleus, was 0.19 pg. Similar ranges in 2C nuclear DNA content

have been reported in other grasses within the same species and ploidy level. Taliaferro et al.

(1997) and Wu et al. (2006) reported a range of 0.33 and 0.34 pg/2C nucleus, respectively in

tetraploid bermudagrasses. A 0.20 pg variation was reported for common dallisgrass (Jarret et

al., 1995), where a larger spread of 0.53 pg was observed in tetraploid buffalograss (Johnson et

al., 1998). 2C nuclear DNA contents are typically higher in temperate than tropical grasses

(Arumuganathan et al., 1999; Murray et al., 2005) resulting in proportionally larger ranges for this trait in cool season grasses of the same species and ploidy level (Barker et al., 2001; Bonos et al., 2002; Eaton et al., 2004; Huff and Palazzo, 1998).

Differences were found (P = 0.001) for 2C nuclear DNA content between Zoysia species and interspecific hybrids (Table 1). Only those between Z. minima and the rest of the group as a

whole may be meaningful, although others were statistically significant. The ability to detect true interspecific hybrids between Z. minima and other Zoysia spp. is supported by the differences between Z. minima, Z. matrella, and the interspecific hybrids of the two. A representative Z. pauciflora genotype was unfortunately not available for FCM analysis. Murray et al. (2005) reported Z. pauciflora to have 0.97 pg/2C nucleus. It may be inferred that differences between Z. pauciflora, Z. matrella, and interspecific hybrids of the two could be detected based on their respective 2C nuclear DNA contents.

Variation in leaf blade width was present between genotypes (P = 0.001) and species (P

= 0.001). Large leaf widths were typically associated with Z. macrantha and Z. japonica

genotypes, where those within Z. minima and Z. pacifica had the smallest leaves. This trait has been well characterized within, and between Zoysia species (Anderson, 2000; Choi et al., 1997;

Hong and Yeam, 1985; Kim, 1983; Yaneshita et al., 1997). Leaf blade width and 2C nuclear

DNA content were not correlated (r = -0.18, P > 0.05). Knight et al. (2005) suggested that leaf anatomical trait (i.e., length, width, area, and mass) correlation with 2C nuclear DNA content is dependent on the species sampled. Leaf width was not correlated with 2C nuclear DNA content in buffalograss (Johnson et al., 1998), bentgrass (Bonos et al., 2002), napiergrass (Pennisetum purpureum Schum.) (Taylor and Vasil, 1987), and switchgrass (Panicum virgatum L.) (Hultquist et al., 1997).

The speciation of zoysiagrass is partially linked to its dissemination over time throughout geographically isolated Pacific Rim countries (Weng et al., 2007). Greilhuber (1998) theorized that genome size variation between reproductively isolated populations will occur over time due to common chromosomal polymorphisms and spontaneous aberrations. Zoysia spp. have one of

the smallest genome sizes in the Poaceae family (Murray et al., 2005). A tentative association

has been made between species of increasing genome size and a decreased ability to adapt to

extreme environments where conditions quickly change (Grime and Mowforth, 1982; Knight et al., 2005). Conceivably, the vast variation in morphological characteristics and establishment

rates (Patton et al., 2007) in Zoysia spp. can be partially explained by its small genome size and the corresponding lack of evolutionary constraints associated with large amounts of repetitive

DNA such as longer cell cycles and non-coding regions that can buffer the effects of mutation.

This could also have a factor in the differential responses of zoysiagrasses to drought (Marcum et

al., 1995; White et al., 2001), shade (Morton et al., 1991), salinity (Marcum et al., 1998; Qian et

al., 2000), temperature (Patton and Reicher, 2007), disease (Green et al., 1994), insect (Braman

et al., 2000; Reinert and Engelke, 2001), and nematode (Busey et al., 1982) pressures.

Conclusions

Continuous variation for 2C nuclear DNA content across most of the Zoysia spp. studied

herein was present, but over a smaller range than in other warm-season grasses within the same

species and ploidy level. Further quantification of 2C nuclear DNA content variation in

additional Zoysia spp. and interspecific hybrids as they become available is needed before an

association between genome size and species is deemed relevant. The rapid and accurate method

for screening zoysiagrass germplasm using the terminal end of a stolon as described within will

allow for repeatable FCM analysis in the future.

An alternative viewpoint that merits re-introduction is the assignment of species based on

reproductive compatibility (Mayr, 1948) rather than by genotypic and phenotypic variation that

transverse multiple species within the genus. Murray (2005) emphasized the need to test for a reduction of fertility in hybrids between individuals that differed in C-value before the two are

recognized as different species. This premise could apply to any plant characteristic, whether it

is genome size or morphology. Leaf width was not correlated with 2C nuclear DNA content in the genotypes evaluated herein. Under the current taxonomic system, zoysiagrasses are often incorrectly distinguished as a species based on leaf texture alone. This confusion would be eliminated if the biological species concept was in use.

Table 2-1. Means for 2C nuclear DNA content and leaf blade width of zoysiagrass genotypes for five species and five interspecific hybridizations 2C DNA content Leaf morphology 2C DNA content Leaf morphology Developing Institution Genotype Genotypic mean SE Blade width Leaf texture¶ Zoysia species Species mean Blade width pg pg mm class pg mm Texas A&M University 5194-5 0.96 a§ 0.01 0.44 u§ Very fine Z. minima 0.96 a§ 0.44 e§

Texas A&M University 5504-6 0.94 ab 0.02 0.96 r Very fine Z. matrella × Z. minima 0.92 b 0.93 d Texas A&M University 5458-39 0.90 bc 0.00 0.90 rs Very fine

Texas A&M University 5335-3 0.89 cd 0.02 3.90 e Coarse Z. macrantha × Z. matrella 0.89 bc 3.90 a

Texas A&M University 5193-19 0.88 c-f 0.01 2.23 kl Medium Z. macrantha# 0.88 c 2.65 b Texas A&M University 5186-16 0.88 c-g 0.01 3.06 h Coarse

Texas A&M University 5463-9 0.88 c-g 0.01 0.83 s Very fine Z. matrella × Z. pauciflora# 0.87 cd 0.75 d Texas A&M University 5459-10 0.87 c-g 0.00 0.67 t Very fine

University of Florida BA433 0.85 d-i 0.03 0.46 u Very fine Z. pacifica 0.85 cde 0.46 e

Sod Solutions Empire 0.88 cd 0.02 4.48 b Coarse USDA and USGA† Meyer 0.87 c-i 0.01 3.52 g Coarse University of Florida UFTZ 0.86 c-i 0.01 3.11 h Coarse 26 University of Florida Ultimate 0.86 c-i 0.02 2.51 j Medium Texas A&M University Palisades 0.86 c-i 0.01 3.73 f Coarse Seed Research of Oregon Compadre 0.86 c-i 0.02 4.97 a Coarse Z. japonica# 0.85 cde 3.54 a University of California El Toro 0.85 d-k 0.02 3.61 fg Coarse Bladerunner Farms, Inc. JaMur 0.84 e-l 0.01 4.17 c Coarse University of California Victoria 0.84 f-l 0.01 2.57 j Medium Sod Solutions Empress 0.84 g-l 0.02 2.15 l Medium Texas A&M University Crowne 0.84 g-l 0.00 4.07 cd Coarse Patten Seed Company Zenith 0.83 h-m 0.01 3.59 fg Coarse

Table 2-1 Continued 2C DNA content Leaf morphology 2C DNA content Leaf morphology Developing Institution Genotype Genotypic mean SE Blade width Leaf texture¶ Zoysia species Species mean Blade width Texas A&M University Cavalier 0.87 c-h 0.01 1.80 m Fine Texas Tech University Shadow Turf 0.87 c-i 0.01 1.35 q Fine Texas A&M University Royal 0.86 d-i 0.01 1.60 no Fine Pursley Turf Farms Cashmere 0.85 d-i 0.02 1.66 n Fine Bladerunner Farms, Inc. Zeon 0.85 d-j 0.02 1.38 q Fine Z. matrella#†† 0.84 de 1.49 c USDA and USGA† Emerald 0.84 f-l 0.02 1.49 p Fine University of Florida Pristine 0.83 g-m 0.02 1.62 no Fine Texas A&M University Zorro 0.79 mn 0.00 1.54 op Fine Texas A&M University Diamond 0.77 n 0.01 0.97 r Very fine

Texas A&M University 5337-46 0.83 i-m 0.02 3.97 de Coarse Z. japonica × Z. macrantha 0.83 e 3.97 a

Texas A&M University 5343-52 0.86 c-i 0.00 2.18 l Medium Texas A&M University 5334-6 0.85 d-j 0.00 2.78 i Medium Texas A&M University 5332-52 0.81 j-n 0.02 2.22 kl Medium Z. japonica × Z. matrella# 0.83 e 2.33 b Texas A&M University 5283-5 0.81 k-n 0.03 2.14 l Medium Texas A&M University 5282-20 0.80 lmn 0.02 2.34 k Medium % CV 3.11 1.87 3.71 10.19 †United States Department of Agriculture and United States Golf Association. §Means within a column followed by the same letter are not different at K = 100 (approximates P = 0.05) according to Waller-Duncan LSD. ¶

27 Leaf texture classifications assigned based on leaf width measurements (Very fine < 1.0mm; 1.0mm < Fine < 2.0mm; 2.0mm < Medium < 3.0mm; 3.0mm < Coarse). #Leaf blade widths were different for genotypes within species at the 0.01 level of probability. ††2C nuclear DNA contents were different for genotypes within species at the 0.01 level of probability.

CHAPTER 3 EFFICIENT METHODOLOGY FOR SCREENING STING NEMATODE RESPONSE IN A TURFGRASS BREEDING PROGRAM

Introduction

Sting nematode (Belonolaimus longicaudatus) is a parasite of many warm-season turfgrasses. This nematode feeds primarily on the actively growing root tips of their hosts which over time results in shallow, necrotic roots. Visual symptoms associated with this root deterioration include chlorosis, wilting, reduced growth rates, and thin turf (Perry, 1974). Turf quality and drought tolerance can be reduced as a result of stunted root systems when sting nematode populations are not managed (Trenholm et al., 2005). Damaged turf may not respond to nitrogen fertility or irrigation over time when plant parasitic nematodes are present in the soil

(Luc et al., 2007). Nitrate leaching, as a result of decreased uptake capability in areas with weakened rootzones, has the potential to reduce water quality and increase management costs as water resources become limited (Luc et al., 2006).

Hybrid bermudagrasses (Cynodon dactylon [L.] Pers. var. dactylon × C. transvaalensis

Burtt-Davy) have been widely utilized for turf purposes throughout the warmer environments of the world due to their broad adaptability and superior quality (Hanna, 2008). A limitation of these grasses has been their host suitability and susceptibility to the sting nematode in the sandy soils of the Southeastern United States (Good, 1959; Good et al., 1956; Holdeman and Graham,

1953; Robbins and Barker, 1973) and on artificially constructed, sand-based putting greens

(Crow, 2005a; Crow, 2005c). Damage to St. Augustinegrass (Stenotaphrum secundatum [Walt.]

Kuntze) roots, and subsequent top growth chlorosis, have been attributed to this nematode in the field (Kelsheimer and Overman, 1953) and observed in controlled greenhouse studies (Rhoades,

1962). Dynamics of the host-parasite relationship have been further studied (Giblin-Davis et al.,

1992a). Generally, diploid St. Augustinegrasses have been characterized as susceptible (Busey

et al., 1991), but resistance has been identified in some polyploid cultivars (Busey et al., 1993).

Good et al. (1956) described the sting nematode as one of the most antagonistic pests of

zoysiagrass (Zoysia spp.). Busey et al. (1982) identified germplasm with partial resistance, but

concluded that many breeding challenges regarding better adaptation and pest resistance needed

to be overcome before this species could be utilized in the sub-tropics. Seashore paspalum

(Paspalum vaginatum Swartz) is a reproductive host that is damaged by the sting nematode

(Hixson et al., 2004). Centipedegrass (Eremochloa ophiuroides [Munro] Hack) has been

reported to be a host of the sting nematode (Christie et al., 1954; Good, 1959), but parasitism on this species has not been well documented.

A variety of management techniques have been evaluated in an attempt to eliminate or suppress plant parasite nematodes where they are damaging. The most common, and often only, method of nematode management in susceptible cultivars has been the use of nematicides (Crow et al., 2003; Giblin-Davis et al., 1991; Johnson, 1970a; Perry et al., 1970). Heald and Burton

(1968) found that sting and root-knot nematode populations were reduced when organic sources of nitrogen were used instead of ammonium nitrate. White and Dickens (1984) observed the same trends, but reported that other management practices (i.e., topdressing, vertical mowing, and core aerification) had no effect on plant parasitic nematode populations. Both endo- and ectoparasitic nematodes can be eliminated from bermudagrass sprigs with a 15 min, 55ºC hot water bath without significantly increasing sprig mortality (Heald and Wells, 1967), but this procedure has limited practicality. Botanical nematicides, root biostimulants, and entomopathogenic nematodes have not shown consistent suppression of plant parasitic nematode populations as required from turfgrass managers (Crow, 2005b; Crow et al., 2006). Recent loss of fenamiphos, the most commonly used nematicide on turf in the United States, has heightened

the need for new nematode management strategies. This has renewed interest in the development of turfgrass cultivars with improved resistance and tolerance to the sting nematode.

The search for useful resistance or tolerance to the sting nematode has been limited to readily available cultivars and a limited number of germplasm accessions (Bekal and Becker,

2000; Giblin-Davis et al., 1992b; Johnson, 1970b; Tarjan and Busey, 1985). In bermudagrass, more effort has been dedicated to breeding and selecting for root-knot nematode resistance

(Burton et al., 1946; Riggs et al., 1962; Sledge, 1962). Burton (1974) stated that several radiation induced mutants of ‘Tifgreen’, ‘Tifway’ and ‘Tifdwarf’ were more resistant to root- knot nematodes than their respective parental clones. Root-knot nematode resistant bermudagrasses often lower the population of root-knot nematodes, but may serve as a host for other ectoparasitic nematode species (Good et al., 1965). The forage bermudagrass ‘Coastcross-

1’ exhibited sting nematode resistance (Burton, 1972). A greater effort to screen broader germplasm of susceptible turfgrass species would likely result in the discovery of more resistant or tolerant individuals. Improvement through breeding should be possible because genetic variability for these characteristics has been demonstrated (Giblin-Davis et al., 1992b). As regulations regarding the use of soil fumigants and nematicides become stricter, it would be valuable to identify a genetic source for effective control of this pest.

This research was initiated to investigate evaluation methods to identify a high throughput, accurate and repeatable greenhouse screen useful to turfgrass breeding programs in comparing response to an ectoparasitic nematode, B. longicaudatus.

Materials and Methods

Two experimental trials were conducted sequentially during the 2007 growing season in a glasshouse at the University of Florida Turfgrass Envirotron in Gainesville, FL. Planting materials were nematode-free, aerial stolons of ‘TifEagle’ hybrid bermudagrass (Hanna and

Elsner, 1999). Three establishment protocols were evaluated using (3.8 cm diameter × 21 cm deep) UV stabilized Ray Leach “Cone-tainers”™ (SC10, Stuewe & Sons, Inc., Tangent, OR) and

(10 cm diameter × nine cm deep) tapered clay pots. Establishment methods were classified as conetainers grown in for 45 days (45-d conetainers), conetainers grown in for 90 days (90-d conetainers), and clay pots grown in for 90 days (90-d clay pots) before inoculation with sting nematodes, respectively. Two inoculation rates, 50 and 100 mixed life-stages of B. longicaudatus/100 cm3 of soil, were compared to an uninoculated control within each establishment method. The experimental design was a split-plot with establishment methods arranged as whole-plots and inoculation treatments as sub-plots with six replications. Upon inoculation, conetainers were placed in (60 × 35 × 15 cm) Beaver Plastics Styroblock™ containers (77/170, Stuewe & Sons, Inc., Tangent, OR) to simulate the insulation provided by the clay pot treatment. The daily average high and low air temperatures in the glasshouse were

33.6 ± 2.9ºC and 23.9 ± 1.3ºC, respectively over the course of both trials.

Both conetainers and clay pots were filled with autoclaved United States Golf

Association (USGA) root-zone specification sand (Anonymous, 1993) with Poly-fil (Fairfield

Processing Corporation, Danbury, CT) placed in the bottom of each to prevent sand from escaping from the drainage holes. Conetainers were planted with one terminal aerial stolon approximately five cm long and clay pots were planted with seven equivalent stolons. One minute of overhead mist was applied eight times daily for one week to allow the rootless sprigs to establish. The frequency was reduced to four times daily during the second week of growth.

Beyond the second week, a single application of two minutes/day of irrigation was scheduled.

Peters Professional 20-20-20 General Purpose Water Soluble Fertilizer (Scotts-Sierra

Horticultural Products Co., Marysville, OH) was applied weekly at a rate of 2.4 g/m2 for the first

month after planting. Subsequent applications were made at the time of inoculation, and again

45 days later. Leaf canopies were trimmed weekly for the duration of the experiments.

Nematode inoculum was extracted using a modified Baermann funnel method (McSorley and Frederick, 1991) from a pure population of B. longicaudatus which was maintained on ‘FX-

313’ St. Augustinegrass. The population density of nematodes in the extract was estimated by

quantifying the number of sting nematodes in one mL aliquots on a counting slide (Hawksley

and Sons Limited, Lancing, Sussex, UK). Nematode counts were replicated five times with 83 ±

6 and 85 ± 4 sting nematodes/mL present in the extracted solutions for the first and second

experiments, respectively. The solutions were diluted to deliver aliquots of 50 sting

nematodes/three mL of solution. Experimental units within establishment methods were sorted

by canopy density, and groups of three with similar densities were assigned to the same replication. Inoculation proceeded sequentially according to replication. An aliquot of inoculum

for conetainers receiving the 50 sting nematode/100 cm3 soil treatment was pipetted into a single hole (one cm diameter × three cm deep), which was then pressed closed. Those receiving the

100 sting nematode/100 cm3 soil treatment were inoculated with two aliquots in separate holes.

Aliquot size was adjusted for the larger soil volume in the clay pots. Inoculum was delivered into two and four holes in the 50 and 100 sting nematodes/100 cm3 soil treatments, respectively.

Experiments were terminated after 90 days and brought to the laboratory for destructive

analysis. The plant and corresponding soil volume were isolated from each conetainer and clay

pot. Shoots were trimmed off at the soil level and the Poly-fil was removed. Roots and nematodes were extracted from the entire experimental unit in the conetainer treatments, and from a 100 cm3 soil core serving as a representative sample from each clay pot.

Nematodes were extracted from inoculated treatments using a centrifugal-sugar flotation

technique (Jenkins, 1964), modified by adding five cm3 of clay to keep the sediment plug intact

as the supernatant liquid was removed after the first centrifugation. Nematode counts were made to determine final sting nematode population densities (Pf) on an inverted light microscope at

×40 magnification and to verify nematode pressure was present.

Roots were collected from uninoculated and inoculated treatments, submersed underwater in 50 mL plastic centrifuge tubes, and stored at -23ºC for later analysis. The root samples were thawed after sting nematode counts had been completed. Individual root systems were placed into a clear acrylic glass tray where a digital image was created using an Epson

Perfection V700 Photo scanner (Epson America, Inc., Long Beach, CA). Lengths of five root diameter classifications (< 0.125 mm, 0.125 to 0.250 mm, 0.250 to 0.500 mm, 0.500 to 1.000 mm , and > 1.000 mm) were individually quantified using WinRHIZO Pro v2007d software

(Regent Instruments, Inc., Quebec, QC) and then summed to determine total root length (TRL) of each sample. Roots were later dried at 75ºC for 48 h to obtain total dry root weights (TDRW).

Reproduction factor (Rf), sting nematode numbers on a total root length basis (Pf/TRL), sting nematode numbers on a total dry root weight basis (Pf/TDRW), total root dry weight percent reduction (TDRW % red.), total root length percent reduction (TRL % red.), and fine root

(diameter < 0.125 mm) length percent reduction (FRL % red.) were calculated with the measured observations.

Pf Rf = (3-1) # of sting nematodes inoculated

⎡⎤()TDRW of inoculated− TDRW of uninoculated TDRW%. red =×⎢⎥100 (3-2) ⎣⎦TDRW of uninoculated

⎡⎤()TRL of inoculated− TRL of uninoculated TRL%. red =×⎢⎥100 (3-3) ⎣⎦TRL of uninoculated

⎡⎤(FRL of inoculated− FRL of uninoculated ) FRL%. red =×⎢⎥100 (3-4) ⎣⎦FRL of uninoculated

The distribution of data for each characteristic was assessed with a histogram and normal probability plot for normality. Transformed datasets were utilized where conditions of normality were not met. An analysis of variance was performed on each trait to test whether establishment methods and inoculation treatments varied. To further study the precision of each method, data were analyzed separately even where establishment method interactions were not significant.

Where appropriate, differences between the inoculated treatments and uninoculated controls were tested with orthogonal coefficients. Pearson correlation coefficients were computed using the CORR procedure in SAS software (SAS, 2008) for 90-d clay pots individually, and both 45-d and 90-d conetainer establishment methods collectively, to test whether any traits were associated with, or could be predictors of, other characteristics in the two dissimilar pots.

Results

Variation between establishment methods was detected (P ≤ 0.05) in all four sting nematode population characteristics (Pf, Rf, Pf/TRL, and Pf/TDRW), but inoculation treatment × establishment method interactions were significant (P ≤ 0.05) for final nematode populations and sting nematode numbers on a total dry root weight basis (Table 3-1). Nematode populations increased in every treatment except for clay pots inoculated with 100 sting nematodes/100 cm3 soil. Reproduction factors were approximately two times larger (P ≤ 0.01) in the low inoculation treatment than in the higher treatment for both conetainer establishment methods. Final nematode populations and sting nematode numbers on a total root length and total dry root weight basis were not different for inoculation treatments within establishment methods. Results

for population characteristics were less variable with conetainer treatments than when evaluated in clay pots (Table 3-2). Sting nematode reproduction and final population density were correlated (P ≤ 0.05) with all root characteristics, excluding total dry root weights, in both conetainer establishment methods. Alternatively, there were no associations detected between nematode count data and root measurements in the clay pots (Table 3-3).

Total dry root weights were different (P ≤ 0.01) for establishment methods and inoculation treatments (Table 3-1), but no treatment differences were found for total dry root weight percent reduction due to variable observations in the clay pot establishment method.

Total dry root weights were 38 % and 28 % larger (P ≤ 0.01) for uninoculated treatments when compared to an average of the two inoculated treatments in the 45-d and 90-d conetainers, respectively. Root weights were not significantly reduced by sting nematode pressure in the clay pots (Table 3-4). Total dry root weights were good predictors of total (r = 0.92, P ≤ 0.01) and fine (r = 0.85, P ≤ 0.01) root lengths in the conetainer establishment methods. Correlation between total dry root weight percent reduction and total root length percent reduction (r = 0.85,

P ≤ 0.01) was also observed. Associations between these traits were present (P ≤ 0.01) in the clay pots, but the relationships were not as strong (Table 3-3).

Total root lengths and fine root lengths were highly correlated (r ≈ 0.98, P ≤ 0.01), as were the corresponding percent reductions (r ≈ 0.97, P ≤ 0.01), in all establishment methods

(Table 3-3). Analysis of variance indicated a significant interaction (P ≤ 0.01) between inoculation treatments and establishment methods for total and fine root lengths. Total and fine root length percent reductions did not vary with establishment method, but inoculation treatments differed (P ≤ 0.05) for both characteristics in the combined analysis (Table 3-1).

Total root lengths of uninoculated treatments were 57 %, 55 %, and 31 % greater (P ≤ 0.01) than

an average of the two inoculated treatments in the 45-d conetainers, 90-d conetainers, and 90-d clay pots, respectively (Table 3-4). Similar results were found for fine root lengths, except in trial two where these lengths were not reduced as greatly in the higher inoculation treatment within the 90-d conetainer method. Related inoculation treatment × trial interactions for total and fine root length percent reductions were also evident (P ≤ 0.05) within 90-d conetainers.

Differences in root length reduction between inoculation treatments were only significant (P ≤

0.05) in the 90-d conetainer method at the conclusion of trial one, but results were inconsistent between trials. Greater numerical differences in root length percent reduction were found between the higher and lower inoculation treatments in the clay pots than with either conetainer method, but within treatment variability prevented significant detection of these differences

(Table 3-5).

Discussion

Identifying turfgrasses that are resistant to, or tolerant of, plant-parasitic nematodes will become necessary as currently used chemical control agents are restricted or removed entirely from use. Methods used to manage these pests in annual cropping systems, including crop rotation and fallowing fields, are not economically or logistically feasible for most perennials.

Sources of genetic resistance may be the only way perennial grasses can persist through nematode pressure for a longer duration in the absence of alternative control methods.

Only a small percentage of available turfgrass genetic resources have been evaluated for response to ectoparasitic nematodes. For breeding programs, current methods have limited the efficient screening of broader germplasm pools due to the space constraints associated with larger clay pots (Crow and Welch, 2004; Hixson et al., 2004; Winchester and Burt, 1964) or the extended cycle times needed to observe root reductions (Giblin-Davis et al., 1992b; Tarjan and

Busey, 1985). Damage induced by plant-parasitic nematodes becomes more apparent on golf

courses when secondary abiotic stresses are present (Crow, 2005a). Simulating field conditions by reducing the frequency of fertilizer application and limiting irrigation may result in more effective screening. Schwartz et al. (2006) found that weekly fertilization and twice daily irrigation hindered the detection of genetic potential for resistance or tolerance to sting nematodes when zoysiagrasses were grown in shorter, uninsulated conetainers.

Many turfgrass species are suitable hosts for sting nematode reproduction (Bekal and

Becker, 2000; Robbins and Barker, 1973). Identifying an appropriate plant standard for use as a control when evaluating sting nematode response may increase the accuracy of screening large numbers of germplasm lines. Plant standards will need to respond to sting nematode populations in the greenhouse in a manner characteristic of the host-pathogen relationship seen in the field.

Damaging numbers of sting nematodes have been associated with subjectively measured declines in turf quality and root lengths on ultradwarf bermudagrasses in Florida (Crow, 2005c).

Reductions in total and fine root lengths of the ultradwarf bermudagrass, TifEagle, also corresponded to increasing sting nematode population densities when evaluated in the greenhouse (Crow and Welch, 2004).

Nematode and root characteristics reported for turfgrasses in the literature were evaluated to select the most informative and repeatable combinations of establishment method and inoculation treatment used in this study. Dry root weights were the most widely used measure of plant-parasitic inflicted root damage before the advent of digital scanners and root length analysis software. Plant breeders should first focus on the primary symptom of nematode damage, reduction of root lengths, and not the correlated decrease in root weight. Root weight can only be used to indirectly measure the problem if associations between root weight and root length are strong. Qualitative turfgrass quality ratings and clipping weights have not been

consistently associated with root damage in bermudagrass (Giblin-Davis et al., 1992b; Hixson et al., 2004; Tarjan and Busey, 1985) and would probably not make effective selection criteria for estimating plant-parasitic nematode damage.

There were sufficient sting nematodes in both inoculation treatments to establish reproducing populations. Lautz (1959) found that inoculum treatments of only 10 sting nematodes/pot resulted in no population increases even though good reproduction was observed when 40 sting nematodes/pot were artificially inoculated onto the same host. Final population densities within the 45-d and 90-d conetainers were equal for both low and high inoculation treatments even though the reproduction factors were approximately two times larger in the lower treatment. This indicates that sting nematode carrying capacities may have been met for the respective size of the root systems in each establishment method. Total root lengths in the clay pots were reduced by the high inoculation treatment when compared to the uninoculated control despite a reproduction factor below one, suggesting that the carrying capacity of the root systems were exceeded and a resulting nematode maximum population density occurred.

Therefore, if identifying differences in nematode reproduction (resistance) is the primary objective of the research, an inoculation rate of 50 B. longicaudatus/100 cm3 of soil and an evaluation period of 90 days appear adequate.

Differences in the number of nematodes on a total root length and total dry root weight basis between establishment methods may not be meaningful when the only treatments are within the same cultivar. Standardized population descriptors such as these could be more useful when making comparisons between different grasses within, or among species. Correlation between root length percent reductions and sting nematode population density would be expected on a putting green under stress. These traits were significantly correlated in both conetainer

establishment methods with no detectable association found in the clay pots. This may indicate that conditions in the conetainers were more representative of field conditions.

Differences were not found between total dry root weights of uninoculated and inoculated treatments in the clay pot establishment method even though total root lengths were reduced by the sting nematode pressure. Johnson (1970b) noted that root systems of uninoculated bermudagrass controls were more dense and fibrous than those of inoculated plants. He concluded that visual symptoms were much more apparent than indicated by root weight differences. Results found in the conetainer establishment methods contrasted those from the clay pots. Identification and selection of superior genotypes in conetainers using only total dry root weight reductions could be possible due to significant differences found between uninoculated and inoculated treatments and the high correlation of root weights and lengths in these establishment methods.

Total and fine root lengths of TifEagle bermudagrass were reported to be 57 % and 69% greater, respectively, in uninoculated treatments than inoculated when evaluated in large (15 cm) clay pots (Crow and Welch, 2004). These measurements and the corresponding root length percent reductions were very similar to the results found in the 45-d conetainer method evaluated herein. When fine root length is highly associated with total root length it may not need to be reported. However, no additional effort is required for its measurement and it may be informative when explaining root damage caused by plant-parasitic nematodes in other turfgrass species with variable root architecture. Results across trials were not consistent for fine root length or total and fine root length percent reductions in the 90-d conetainers suggesting a potential lack of stability for this establishment method.

Plant breeding methodologies are purposefully designed to only be as comprehensive as required to gather enough information to make genetic gains from selection. The balance between greater investments in time and complexity must be maintained with screening larger population sizes in the search for genes of interest. Plant-parasitic nematode populations and root systems are inherently variable. Therefore it is particularly necessary to reduce outside variation when experimenting with nematodes in the greenhouse so that the genetic potential of each turfgrass being evaluated can be reached and quantified. Conetainer treatments were not root-bound at the completion of the experiments. Sting nematode and bermudagrass measurements were characterized from the entire experimental unit rather than from a sample core as necessary in the root-bound clay pots. Sampling error may have attributed to the variability observed in characteristics measured within the clay pot establishment method.

Based on this research, TifEagle bermudagrass can be effectively used as a susceptible plant standard to verify pathogenicity of sting nematode incoculum and quantify root injury of untested genotypes during intial screening of large sets of turfgrass germplasm. Use of conetainers will increase the number of germplasm lines that can be evaluated in a single trial because they require less bench space in the greenhouse and the extraction of roots and nematodes does not require sampling as with larger clay pots. Inoculation levels should be no higher than required to establish reproducing populations of nematodes because availability of inoculum can be a limiting factor to the number of genotypes that can be screened in a particular trial. Our results support establishing turfgrasses for 45 days in deep, insulated conetainers before inoculation with 50 B. longicaudatus/100 cm3 soil to determine initial plant response.

Further testing of plant response/tolerance may require higher inoculum levels or longer evaluation intervals.

Table 3-1. Mean squares for Belonolaimus longicaudatus reproduction factor (Rf), final population density (Pf), population density on a total root length basis (Pf/TRL), population density on a total dry root basis (Pf/TDRW), total dry root weight (TDRW), total dry root weight percent reduction (TDRW % red.), total root length (TRL), total root length percent reduction (TRL % red.), fine root length (FRL), and fine root length percent reduction (FRL % red.) of TifEagle bermudagrass evaluated in three establishment (Est.) methods with different inoculation treatments (Inoc. TRT) in two experimental trials. Mean squares TDRW TRL FRL Source df Rf Pf Pf/TRL Pf/TDRW TDRW† % red. TRL† % red. FRL† % red. Trial (T) 1 1.488** 95.4** 0.4960** 11789** 0.08664** 7.91 5729209** 1.03 711290** 4.77 Error A: Rep (T) 10 0.034 7.4 0.0135 163 0.00285 2.99 349298 1.44 53090 2.02 Est. Method (E) 2 1.874** 137.1** 0.1377* 1292* 0.03145** 13.08 6132352** 6.03 1358829** 7.67 E × T 2 0.499 31.5 0.0449 1144 0.01405** 5.93 1623045** 0.29 250491** 0.85 Error B: Rep × E (T) 20 0.177 23.3 0.0276 333 0.00112 3.86 86198 2.15 15915 2.34 Inoc. TRT (I) 1 6.778** 66.2 0.0094 237 0.00584** 2.58 3946964** 7.29* 874549** 9.30** I × T 1 0.101 0.0 0.0000 1 0.00007 0.92 265678* 0.00 89535** 0.10 I × E 2 0.290 91.1* 0.0563 946* 0.00154 0.18 283354** 0.74 63174** 0.84 I × T × E 2 0.178 12.5 0.0201 77 0.00098 1.44 63695 2.02 22437 2.22 41 Error C: MSE 30 0.141 23.5 0.0230 265 0.00079 2.07 58618 1.19 12990 1.05 % CV 26.3 35.0 34.8 32.9 26.0 16.0 17.9 13.2 18.4 12.4 *, **Significant at the 0.05 and 0.01 probability levels, respectively. †Analysis included comparison to un-inoculated controls, therefore df for I, I × T, I × E, I × T × E, and Error C were 2, 2, 4, 4, and 60, respectively.

Table 3-2. Mean reproduction factor (Rf), final population density (Pf), population density on a total root length basis (Pf/TRL), population density on a total dry root basis (Pf/TDRW), of Belonolaimus longicaudatus on TifEagle bermudagrass 90 days after inoculation evaluated in three establishment methods with two inoculation treatments in two experimental trials. Rf Pf Pf/TRL Pf/TDRW Treatment† (nematodes) (nematodes) (nematodes/cm) (nematodes/g) 45-d conetainers I (50) 2.7 ± 1.5**‡ 134 ± 75 0.17 ± 0.10 2790 ± 1976 I (100) 1.4 ± 0.6 138 ± 58 0.22 ± 0.14 3569 ± 3203 90-d conetainers I (50) 4.2 ± 1.6** 210 ± 80 0.14 ± 0.07 1750 ± 863 I (100) 2.2 ± 1.0 223 ± 101 0.15 ± 0.07 1850 ± 789 90-d clay pots I (50) 2.8 ± 1.5** 391 ± 213 0.36 ± 0.17 4518 ± 2597 I (100) 0.8 ± 0.7 221 ± 193 0.29 ± 0.39 3154 ± 4088 **Inoculated treatments significantly different at the 0.01 probability level, according to orthogonal coefficient analysis. †Inoculated (I) with 50 or 100 B. longicaudatus/100 cm3 soil. ‡Data are means of two trials with six replications each ± standard deviations.

Table 3-3. 45-d and 90-d conetainer (above diagonal) and 90-d clay pot (below diagonal) correlation coefficients of Belonolaimus longicaudatus reproduction factor (Rf), final population density (Pf), population density on a total root length basis (Pf/TRL), population density on a total dry root weight basis (Pf/TDRW), total dry root weight (TDRW), total dry root weight percent reduction (TDRW % red.), total root length (TRL), total root length percent reduction (TRL % red.), fine root length (FRL), and fine root length percent reduction (FRL % red.) of TifEagle bermudagrass. Nematode and TDRW TRL FRL root data Rf Pf Pf/TRL Pf/TDRW TDRW % red. TRL % red. FRL‡ % red. count count count/cm count/g g % cm % cm % Rf 0.89** 0.66** 0.52** -0.17 ns -0.28* -0.27* -0.48** -0.25* -0.41** Pf 0.94** 0.76** 0.59** -0.20 ns -0.34** -0.32** -0.58** -0.32** -0.53** Pf/TRL 0.64** 0.78** 0.95** -0.57** -0.59** -0.64** -0.66** -0.63** -0.59** Pf/TDRW 0.67** 0.77** 0.97** -0.60** -0.58** -0.63** -0.56** -0.60** -0.48** TDRW 0.03 ns† 0.02 ns -0.29 ns -0.38* 0.58** 0.92** 0.43** 0.85** 0.37** TDRW % red. 0.13 ns 0.07 ns -0.22 ns -0.25 ns 0.51** 0.59** 0.85** 0.56** 0.74** TRL 0.01 ns -0.05 ns -0.36* -0.39* 0.79** 0.30 ns 0.60** 0.98** 0.57** TRL % red. 0.10 ns -0.01 ns -0.28 ns -0.19 ns 0.08 ns 0.55** 0.32 ns 0.62** 0.97**

43 FRL 0.05 ns -0.01 ns -0.32 ns -0.33* 0.69** 0.25 ns 0.97** 0.37* 0.61** FRL % red. 0.16 ns 0.06 ns -0.23 ns -0.13 ns 0.06 ns 0.45** 0.34* 0.96** 0.41* *, **Significant at the 0.05 and 0.01 probability levels, respectively. †Not significant at the 0.05 probability level. ‡Fine root (diameter < 0.125 mm) length.

Table 3-4. Mean total dry root weight (TDRW), total dry root weight percent reduction (TDRW % red.), and total root length (TRL) of TifEagle bermudagrass 90 days after inoculation evaluated in three establishment methods with uninoculated and inoculated treatments in two experimental trials. Treatment† TDRW (g) TDRW % red. (g) TRL (cm)

45-d conetainers U 0.09 ± 0.04**‡ ─ 1394 ± 523** I (50) 0.07 ± 0.06 -24 ± 23 950 ± 455 I (100) 0.06 ± 0.04 -33 ± 24 828 ± 338

90-d conetainers U 0.16 ± 0.03** ─ 2393 ± 260** I (50) 0.13 ± 0.03 -19 ± 27 1608 ± 352 I (100) 0.12 ± 0.02 -23 ± 19 1478 ± 260

90-d clay pots U 0.11 ± 0.05 ─ 1401 ± 609** I (50) 0.10 ± 0.05 4 ± 46 1153 ± 472 I (100) 0.11 ± 0.07 -3 ± 45 994 ± 483 **Uninoculated controls significantly different from both inoculated treatments at the 0.01 probability level, according to orthogonal coefficient analysis. †Uninoculated (U); Inoculated (I) with 50 or 100 B. longicaudatus/100 cm3 soil. ‡Data are means of two trials with six replications each ± standard deviations.

Table 3-5. Mean total root length percent reduction (TRL % red.), fine root length (FRL), and fine root length percent reduction (FRL % red.) of TifEagle bermudagrass 90 days after inoculation evaluated in three establishment methods with uninoculated and inoculated treatments in two experimental trials. Treatment† TRL % red. (cm) FRL (cm) FRL % red. (cm)

45-d conetainers U ─ 670 ± 224** ─ I (50) -31 ± 16‡ 438 ± 170 -32 ± 19 I (100) -41 ± 13 383 ± 122 -42 ± 13 90-d conetainers (Trial 1) U ─ 1130 ± 162** ─ I (50) -31 ± 11§ 787 ± 118§§ -30 ± 11§§ I (100) -46 ± 10 573 ± 90 -49 ± 9 90-d conetainers (Trial 2) U ─ 1074 ± 107** ─ I (50) -34 ± 19 721 ± 142 -32 ± 16 I (100) -30 ± 10 776 ± 103 -28 ± 6

90-d clay pots U ─ 610 ± 250* ─ I (50) -9 ± 39 518 ± 193 -6 ± 39 I (100) -28 ± 25 425 ± 184 -28 ± 27 *, **Uninoculated controls significantly different from both inoculated treatments at the 0.05 and 0.01 probability levels, respectively, according to orthogonal coefficient analysis. †Uninoculated (U); Inoculated (I) with 50 or 100 B. longicaudatus/100 cm3 soil. ‡Data are means of two trials with six replications each ± standard deviations for 45-d conetainers and 90-d clay pots, and means of one trial with six replications ± standard deviations for 90-d conetainers. §, §§Inoculated treatments significantly different at the 0.05 and 0.01 probability levels, respectively, according to orthogonal coefficient analysis.

CHAPTER 4 VARIABLE RESPONSES OF ZOYSIAGRASS, ST. AUGUSTINEGRASS, AND BERMUDAGRASS GENOTYPES TO THE STING NEMATODE

Introduction

Zoysiagrass (Zoyisa spp.) can be utilized successfully as an alternative turf in warmer environments (Patton and Reicher, 2007) with various insect pressures (Braman et al., 2000;

Reinert and Engelke, 2001). A limitation preventing more widespread use of Zoysia is their susceptibility to damage by the plant parasitic sting nematode (Belonolaimus longicaudatus) on sandy, well drained soils (Christie et al., 1954; Good, 1959; Good et al., 1956). Rapid sting nematode population increases are possible in favorable environments where available food is not limiting. Huang and Becker (1999) reported that reproductive females can lay approximately three eggs every two days and that their life cycle can be completed in as few as 24 days. Sting nematodes are ectoparasites that feed on host root tips through an oral stylet. Damage to root tip meristematic tissue ceases growth and serves as an infection point for secondary pathogens.

Visual symptoms associated with this root deterioration include chlorosis, wilting, reduced growth rates, and thin turf (Perry, 1974). Stunted root systems of sting nematode damaged turf likely contribute to greater nitrogen leaching (Luc et al., 2006) when applications of irrigation and fertilizer are increased to sustain growth. Zoysiagrass was the least effective in limiting nitrate leaching of six warm-season turfgrasses (Bowman et al., 2002). Therefore, there are several factors that may contribute to reduced groundwater quality and increased fertilization costs when growing zoysiagrass in areas suitable for sting nematode population growth.

Sting nematodes can parasitize many warm-season turfgrasses. Turf quality and drought survival of hybrid bermudagrass (Cynodon dactylon [L.] Pers. var. dactylon × C. transvaalensis

Burtt-Davy) can be reduced as a result of stunted root systems when sting nematode populations are not controlled (Trenholm et al., 2005). The search for useful resistance or tolerance to the

sting nematode in Cynodon spp. has been limited to readily available cultivars and a limited number of germplasm accessions (Giblin-Davis et al., 1992b). Damage to St. Augustinegrass

(Stenotaphrum secundatum [Walt.] Kuntze) roots, and subsequent top growth chlorosis, have been attributed to this nematode in the field (Kelsheimer and Overman, 1953) and observed in controlled greenhouse studies (Rhoades, 1962). Generally, diploid St. Augustinegrasses have been characterized as susceptible (Busey et al., 1991), but resistance was identified in some polyploid cultivars (Busey et al., 1993). Seashore paspalum (Paspalum vaginatum Swartz) is a reproductive host of the sting nematode and it may not be as tolerant as bermudagrass in some situations (Hixson et al., 2004). Centipedegrass (Eremochloa ophiuroides [Munro] Hack) is reportedly a host of the sting nematode (Christie et al., 1954; Good, 1959), but parasitism on this species has not been well documented.

Several management techniques have been evaluated in attempts to eliminate or suppress damaging plant parasitic nematodes. The most common, and often only, method of control has been the use of nematicides (Crow, 2005b; Perry et al., 1970). Heald and Burton (1968) found that sting and root-knot nematode populations were reduced when organic sources of nitrogen were used instead of ammonium nitrate on bermudagrass. White and Dickens (1984) observed the same trends, but reported that additional management practices (i.e., topdressing, vertical mowing, and core aerification) had no effect. However, Dunn et al. (1981) reported that mowing height and thatch removal did affect nematode populations in ‘Meyer’ zoysiagrass. Greater numbers of spiral (Helicotylenchus spp.) and dagger nematodes (Xiphinema spp.) were associated with higher mowing heights. Thatch removal contributed to higher populations of dagger nematodes, but neither the spiral or stunt nematode numbers were significantly affected.

Botanical nematicides, root biostimulants, and entomopathogenic nematodes have not shown

consistent suppression of plant parasitic nematode populations as required from turfgrass managers (Crow, 2005a; Crow et al., 2006).

The search for useful resistance or tolerance to the sting nematode in Zoysia spp. has been limited to a small number of cultivars and germplasm accessions. Busey et al. (1982) identified germplasm with partial resistance, but determined that breeding challenges regarding better adaptation and pest resistance needed addressing prior to its utilization in the sub-tropics.

Greater efforts to screen diverse germplasm will likely result in the discovery of additional sources of resistance or tolerance. Because zoysiagrass is slow to recover from injury, any source of sting nematode resistance will be valuable because remedial nematicide applications are less effective in species that can not rapidly re-establish root and shoot densities while plant parasitic nematode populations in the soil are low (Busey et al., 1982; Juska, 1972; Perry, 1974).

This research was initiated to evaluate the host status and relative tolerance of six turfgrass genotypes to the sting nematode as estimated by two different screening methodologies.

Materials and Methods

2002), ‘Emerald’ (Forbes et al., 1955), ‘Empire’, and an experimental germplasm line designated

UFTZ. ‘Floratam’ St. Augustinegrass (Horn et al., 1973) was also evaluated and ‘TifEagle’ bermudagrass (Hanna and Elsner, 1999) was included as a susceptible control. Two establishment protocols were evaluated using (3.8 cm diameter × 21 cm deep) UV stabilized Ray

Leach “Cone-tainers”™ (SC10, Stuewe & Sons, Inc., Tangent, OR). Establishment methods were classified as conetainers grown in for 45 days (45-d conetainers) and 90 days (90-d conetainers) respectively, before inoculation with sting nematodes. An inoculation treatment of

50 sting nematodes in mixed life-stages per 100 cm3 of soil was compared to an uninoculated control within each establishment method. The experimental design was a split-split-plot with establishment methods arranged as whole-plots, genotypes as sub-plots, and inoculation treatments as split-sub-plots with six replications. Upon inoculation, conetainers were placed in

(60 × 35 × 15 cm) Beaver Plastics Styroblock™ containers (77/170, Stuewe & Sons, Inc.,

Tangent, OR) to provide insulation from extreme temperature changes.

Conetainers were filled with autoclaved United States Golf Association (USGA) root- zone specification sand (Anonymous, 1993) with Poly-fil (Fairfield Processing Corporation,

Danbury, CT) placed in the bottom of each to prevent sand from escaping. Conetainers were planted with one terminal aerial stolon approximately five cm long. One minute of overhead mist was applied eight times daily for one week to allow the rootless sprigs to establish. The frequency was reduced to four times daily during the second week of growth. Beyond the second week, a single application of two minutes day-1 was scheduled. Peters Professional 20-

20-20 General Purpose Water Soluble Fertilizer (Scotts-Sierra Horticultural Products Co.,

Marysville, OH) was applied weekly at a rate of 2.4 g/m2 for the first month after planting.

Subsequent applications were made at the time of inoculation, and again 45 days later. Leaf canopies were trimmed weekly for the duration of the experiments.

Sting nematode inoculum was extracted using a modified Baermann funnel method

(McSorley and Frederick, 1991) from a pure population of B. longicaudatus which was maintained on ‘FX-313’ St. Augustinegrass. The population density of nematodes in the extract was estimated by quantifying the number of sting nematodes in one mL aliquots on a counting slide (Hawksley and Sons Limited, Lancing, Sussex, UK). Nematode counts were replicated five times with 83 ± 6 and 85 ± 4 sting nematodes mL-1 present in the extracted solutions for the

first and second experiments, respectively. The solutions were diluted to deliver aliquots of 50 sting nematodes per three mL of solution. Experimental units within establishment methods were sorted by canopy density, and groups of two with similar densities were assigned to the same replication. Inoculation proceeded sequentially according to replication. An aliquot of inoculum for conetainers receiving the 50 sting nematodes was pipetted into a single hole (one cm diameter × three cm deep), which was then pressed closed.

Experiments were terminated after 90 days and brought to the laboratory for destructive analysis. The plant and corresponding soil volume were isolated from each conetainer. Shoots were trimmed off at the soil level and the Poly-fil was removed. Nematodes were extracted from inoculated treatments using a centrifugal-sugar flotation technique (Jenkins, 1964), modified by adding five cm3 of clay to keep the sediment plug intact as the supernatant liquid was removed after the first centrifugation. Nematode counts were made on an inverted light microscope at

×40 magnification to verify nematode pressure and to determine final sting nematode population densities (Pf).

(Regent Instruments, Inc., Quebec, QC) and then summed to determine total root length (TRL) of each sample. Roots were later dried at 75ºC for 48 h to obtain total dry root weights (TDRW).

Reproduction factor (Rf), mean total root length percent reduction (TRL % red.), and mean total root dry weight percent reduction (TDRW % red.) were then calculated:

Pf Rf = (4-1) # of sting nematodes inoculated

⎡⎤()TDRW of inoculated− TDRW of uninoculated TDRW%. red =×⎢⎥100 (4-2) ⎣⎦TDRW of uninoculated

⎡⎤()TRL of inoculated− TRL of uninoculated TRL%. red =×⎢⎥100 (4-3) ⎣⎦TRL of uninoculated

The distribution of data for reproduction factors, total root lengths, and total dry root weights were assessed with a histogram and normal probability plot for normality. Transformed datasets were utilized where conditions of normality were not met. An analysis of variance was performed on these traits to test whether establishment methods, genotypes, and inoculation treatments varied. Data were analyzed separately within each establishment method to further study the effects of these treatments. Genotype means were separated using a Waller-Duncan k- ratio LSD. Differences between the inoculated treatments and uninoculated controls within each genotype were tested with orthogonal coefficients.

Results and Discussion

Sting nematode Rfs were not consistent between trials, establishment methods, and genotypes as noted by the significant interactions in the overall analysis of variance (Table 4-1).

All turfgrasses were suitable hosts for nematode reproduction, but definitive trends within establishment methods or on individual grasses were not apparent. UFTZ generally supported more sting nematodes with final populations five to 11 times greater than initial inoculation levels. Less reproduction usually occurred in Floratam and TifEagle (Table 4-2). Zoysiagrasses with resistance to, or tolerance of the sting nematode have not yet been identified in controlled

greenhouse evaluations, although Emerald and other Zoysia cultivars are known reproductive hosts of this plant parasite (Bekal and Becker, 2000). Busey et al. (1993) detected partial antibiosis to the sting nematode in Floratam and other polyploidy St. Augustinegrasses, but complete resistance to nematode reproduction was not observed. Sting nematode reproduction factors reached 25 on TifEagle bermudagrass when evaluated in large clay pots (Crow and

Welch, 2004). Our results suggest that TifEagle is a good host, but smaller soil and root volumes from the conetainers may limit total nematode populations that can be supported.

Differences were detected (P ≤ 0.01) between the two conetainer establishment methods, among the six genotypes, and between uninoculated and inoculated treatments for TRL and

TDRW. Genotype × establishment method interactions were also significant (P ≤ 0.01) for these two characteristics (Table 4-1). Data were analyzed separately within each establishment method and genotype to evaluate root damage and the effect of establishment time on detecting tolerance or susceptibility to the sting nematode in each turfgrass. Total root length and TDRW of genotypes in the inoculated treatment were evaluated to determine which turfgrasses maintained larger root systems in the presence of sting nematode pressure (Table 4-3).

Identifying tolerance in perennial grasses could enhance long term survival by facilitating water and nutrient uptake, especially in the presence of other abiotic and biotic stresses.

Total root lengths were correlated (r = 0.91, P ≤ 0.01) with fine (diameter < 0.125 mm) root lengths. Data for fine root lengths did not offer additional explanation of the variation observed between establishment methods and among genotypes, and therefore was not reported.

UFTZ was the only genotype in which nematode inoculated plants did not exhibit a significant decline in TRLs relative to the uninoculated controls indicating a level of tolerance when evaluated in 45-d conetainers. Reductions (P ≤ 0.05) were observed for the remaining inoculated

cultivars and ranged from 24% for Empire to 37% in Floratam (Table 4-3). Percent reductions are useful to determine the relative susceptibility of a genotype, but reveal little concerning the root systems overall ability to remain functional even when injured. Total root lengths of UFTZ were greater (P ≤ 0.05) than those of the other turfgrasses under sting nematode pressure, but

Empire did demonstrate moderate TRLs when inoculated. TifEagle and Floratam had the smallest TRLs and greatest TRL % reductions of the genotypes evaluated in 45-d conetainers

(Table 4-3).

Conetainers established for 90 days before inoculation generally had longer TRLs for all genotypes. Sting nematode response in UFTZ, Empire, and TifEagle was similar in both establishment methods, but significant reductions in TRLs were not found between inoculated and uninoculated treatments for both Emerald and Cavalier in the 90-d conetainers (Table 4-3).

This suggests tolerance to sting nematode feeding may be attributed to root maturation prior to the onset of pest pressure for some genotypes. Thicker primary roots were credited for resistance observed in polyploid St. Augustinegrasses (Busey et al., 1993), but our results indicate (data not shown) that coarse (diameter > 1 mm) root length was reduced by the sting nematodes.

Maintenance of fine root length primarily contributed to the differences in estimated tolerance between establishment methods for Emerald and Cavalier. In the 90-d conetainers UFTZ,

Empire, and Emerald had longer (P ≤ 0.05) TRLs than Cavalier, Floratam, and TifEagle when inoculated with sting nematodes (Table 4-3).

Estimating plant parasitic nematode damage with dry root weight reductions was the most practical approach before digital scanners and root length analysis software became readily available. Root injury can only be accurately diagnosed with TDRWs when they are correlated to the primary symptom of sting nematode damage, reduction of TRL. Associations between

these two characteristics in this study were significant (P ≤ 0.01), but not strong in either the 45- d conetainers (r = 0.60) or the 90-d conetainers (r = 0.38). Alternatively, TDRWs were good predictors (r = 0.91, P ≤ 0.01) of coarse root length. Total dry root weights of all genotypes were reduced (P ≤ 0.05) when inoculated in the 45-d conetainers. The 30% decrease in TDRW of UFTZ in the inoculated treatments indicates for this genotype that heavier, coarse roots may have been stunted, though TRL was maintained. Possibly, this genotype increases its lateral fine root production when primary roots are limited due to biotic stress. This would be a highly desirable characteristic and would allow for greater soil volume exploration. Total dry root weight % reductions categorized Empire, Cavalier, and Emerald similarly to TRL % reductions with regards to sting nematode susceptibility when given 45 days to establish. Injury to fine roots and immature coarse roots in Floratam when inoculated in 45-d conetainers was likely responsible for both the TDRW and TRL % reductions in these trials. TifEagle TDRWs were reduced the least and had the smallest weights compared to other genotypes under nematode pressure, suggesting the root architecture of this bermudagrass is composed of more fine than coarse roots (Table 4-3).

Fewer differences in TDRWs were seen between uninoculated controls and inoculated treatments at the conclusion of the study in the 90-d conetainers. Only Emerald and TifEagle behaved similarly for this trait in both establishment methods, exhibiting significant (P ≤ 0.05)

TDRW reductions (Table 4-3). Longer establishment times should allow uninterrupted growth of larger, heavier roots that may be less available for sting nematode feeding after inoculation.

In this case, longer experimental duration would be necessary to see TDRW reductions in susceptible genotypes. Giblin-Davis et al. (1992a) did not see root weight reductions in sting nematode inoculated treatments of a susceptible St. Augustinegrass cultivar until 84 days after

inoculation. Floratam, known for producing thicker diameter roots, had the heaviest TDRWs, where the zoysiagrasses UFTZ, Empire, and Emerald had root systems of similar weights when grown in the presence of sting nematodes in 90-d conetainers (Table 4-3).

Conclusions

Trends of increasing regulation and restriction of nematicides highlight the importance of identifying turfgrasses with improved genetic resistance to plant parasitic nematodes. Variability was detected for sting nematode response in the zoysiagrasses studied indicating that potential exists for improving resistance or tolerance to the sting nematode. Inherent nematode resistance is especially important in Zoysia because the genus lacks the recuperative potential of other faster growing grasses (Busey and Myers, 1979).

The long term responses of perennial crops must often be estimated by abbreviated selection procedures in order to make progress in a timely manner. Simulating field conditions in greenhouse experiments can help increase realized genetic gains. The two methods evaluated reproduce conditions where grasses are sprigged into fumigated soil and can establish root systems before soil nematode populations return. Unfortunately, it is difficult to simulate growth into plant parasitic nematode infested soils because absence of an initial food source limits the pathogen in vitro. Pathogenicity of the sting nematode inoculum was verified by the response of the susceptible TifEagle bermudagrass. Total root length measurements suggest that the degree of tolerance exhibited by genotypes to the sting nematode in 90-d conetainers can be quantified further when less time is allowed for root system development. UFTZ showed tolerance to root length reductions in these trials but was not resistant to the sting nematode as indicated by its high Rfs. Cavalier and Emerald demonstrated tolerance when their root systems were more mature, but not when inoculated while still establishing. In this study, the use of 45-d conetainers produced the most conservative results by identifying only one tolerant genotype.

UFTZ maintained its root system under sting nematode pressure prior to complete establishment, which is perhaps the most desired form of tolerance. Remaining questions are how long can

UFTZ continue to sustain root and shoot development and what impacts could secondary biotic and abiotic stress have in the presence of increased sting nematode populations. These answers require further research. Empire zoysiagrass had significant TRL damage when compared to its uninoculated control but maintained longer TRLs than most of the other turfgrasses in the presence of sting nematode feeding. Selection based on percent reduction data and persistent root growth will reduce the risk of overlooking viable sources of tolerance to plant parasitic nematodes in a plant breeding program. Total dry root weights were associated with larger, more coarse roots and not necessarily the fine roots capable of uptaking more water and nutrients. This research supports the need for continued efforts to screen turfgrass germplasm for greater resistance and tolerance to the sting nematode.

Table 4-1. Mean squares for Belonolaimus longicaudatus reproduction factor (Rf), total root length (TRL), and total dry root weight (TDRW) of six turfgrasses evaluated in two establishment (Est.) methods with uninoculated and inoculated treatments (Inoc. TRT) in two experimental trials. Mean squares Source df Rf† TRL TDRW Trial (T) 1 3.259* 7842180** 0.37681** Error A: Rep (T) 10 0.331 776626 0.03657 Est. Method (E) 1 0.053 27610011** 0.73430** E × T 1 4.037* 434739 0.08067** Error B: Rep × E (T) 10 0.361 186288 0.00576 Genotype (G) 5 2.642** 7158096** 0.38374** G × T 5 1.186** 298312 0.00253 G × E 5 0.936* 500769** 0.01588** G × T × E 5 0.897* 388716* 0.01021* Error C: Rep × G (T × E) 100 0.303 133142 0.00364 Inoc. TRT (I) 1 13807287** 0.31893** I × T 1 31168 0.00026 I × E 1 8600 0.02563** I × G 5 228443 0.00609 I × T × E 1 376504 0.00272 I × T × G 5 97970 0.00690 I × E × G 5 164457 0.00400 I × T × E × G 5 117750 0.00749 Error D: MSE 120 133109 0.00346 % CV 27.5 19.4 12.8 *, **Significant at the 0.05 and 0.01 probability levels, respectively. †Reproduction factor mean squares apply only to inoculated treatments.

Table 4-2. Mean reproduction factor (Rf) of Belonolaimus longicaudatus on six turfgrasses 90 days after inoculation with 50 B. longicaudatus, evaluated in 45-d conetainers and 90-d conetainers in two experimental trials. Rf 45-d conetainers 90-d conetainers Genotype Trial 1 Trial 2 Trial 1 Trial 2 –––––––– nematodes –––––––– UFTZ 5.3 a† 11.0 a 5.0 ab 7.4 a Empire 2.4 bc 7.9 a 4.2 ab 3.3 a Cavalier 3.0 bc 8.4 a 5.9 a 3.9 a Emerald 3.4 ab 3.4 b 6.1 a 3.4 a Floratam 2.3 bc 2.2 b 2.7 b 3.2 a TifEagle 1.7 c 3.7 b 3.0 b 5.4 a % CV 23.9 28.2 24.0 31.4 †Means within a column followed by the same letter are not different at K = 100 (approximates P = 0.05) according to Waller-Duncan LSD.

Table 4-3. Mean total root length (TRL) and total dry root weight (TDRW) of six turfgrasses 90 days after inoculation evaluated in 45-d conetainers and 90-d conetainers with uninoculated (U) and inoculated (I) treatments in two experimental trials. TRL TDRW 45-d conetainers 90-d conetainers 45-d conetainers 90-d conetainers Both trials Both trials Both trials Both trials Genotype U I† % red. U I % red. U I % red. U I % red. –––––––––––––––––––––– cm –––––––––––––––––––––– ––––––––––––––––––––––– g ––––––––––––––––––––––– UFTZ 2343 2142 a‡ -8 2755 2445 a -11 0.30** 0.22 a -30 0.34 0.28 b -16 Empire 2275* 1729 b -24 2837** 2231 a -21 0.22* 0.17 b -25 0.28 0.26 b -7 Cavalier 1631** 1161 cd -29 1944 1752 b -10 0.17** 0.11 c -34 0.23 0.19 c -18 Emerald 1715* 1210 c -29 2491 2142 a -14 0.17** 0.11 c -38 0.31** 0.24 b -23 Floratam 1412** 883 d -37 2000** 1679 b -16 0.36** 0.21 ab -43 0.46 0.40 a -14 TifEagle 1394** 950 cd -32 2393** 1608 b -33 0.09* 0.07 d -21 0.16* 0.13 d -22 *, **Uninoculated controls significantly different from inoculated treatments within establishment method at the 0.05 and 0.01 probability levels, respectively, according to orthogonal coefficient analysis. †Inoculated with 50 Belonolaimus longicaudatus. ‡Means within a column followed by the same letter are not different at K = 100 (approximates P = 0.05) according to Waller-Duncan LSD.

CHAPTER 5 MOWING HEIGHT AND NITROGEN FERTILITY MANAGEMENT OF ZOYSIAGRASS IN FLORIDA

Introduction

Zoysiagrasses (Zoysia spp.) are warm-season perennial grasses adapted to a wide range of environments. Most cultivars are vegetatively propagated to preserve the hybrid vigor created in the initial cross pollination (Forbes et al., 1955; Grau and Radko, 1951; Hanson, 1966).

Maintaining conditions for optimal growth is very important in these species because their slow spread can limit production (Juska, 1959). Topdressing and the use of pre- and postemergence herbicides have been shown to quicken zoysiagrass grow-in by improving sprig survival and reducing weed competition. The effects of nitrogen have generally been found to be minimal during the establishment phase of both plugs and sprigs (Carroll et al., 1996; Fry and Dernoeden,

1987; Richardson and Boyd, 2001). Fall applications of nitrogen often promote weed infestations in dormant zoysiagrass which can result in decreased spring turf quality due to competition for light interception, water, and nutrients (Busey, 2003; Dunn et al., 1993; Henry et al., 1989). Dunn et al. (1995) recommended only fertilizing ‘Meyer’ enough to maintain acceptable turf quality and density so that shoot growth is not encouraged over root development as excess nitrogen is transported to the canopy.

Differential responses of Z. japonica to varying mowing heights have been documented.

Joo et al. (1999) reported size and silica content of leafs were reduced in lower mowed turf, and recommend a mowing height of 3.8 cm or greater to maintain turf quality, shoot density, and root mass under low maintenance conditions. Meyer zoysiagrass maintained acceptable summer turf quality under 71% continuous shade when mowed at 1.6 and 2.5 cm (Bunnell et al., 2005).

Engelke et al. (1992) described the effects of three nitrogen fertilization rates (49, 98, and

293 kg ha-1) and two mowing heights (1.6 and 2.5 cm) on 10 zoysiagrass genotypes near Dallas,

TX. Plots fertilized with N at 49 kg ha-1 retained green color during winter months and greened- up quicker in the spring compared to those treated with higher nitrogen rates. Turfgrass quality during summer was improved with higher N rates, especially at the lower mowing height, but there were no discernable effects of N and mowing height on winter and spring turf quality.

Greater incidence of large patch (Rhizoctonia solani) was observed on 12 Zoysia cultivars when mowed at 1.2 and 2.5 cm rather than at 4.5 or 5.1 cm, and was not affected by nitrogen source or rate (Green et al., 1994). Centipedegrass (Eremochloa ophiuroides [Munro] Hack) (Toler et al.,

2007), buffalograss (Buchloe dactyloides [Nutt.] Engelm.) (Frank et al., 2004), and carpetgrass

(Axonopus affinis Chase) (Bush et al., 2000) have been evaluated for response to mowing height and N fertility combinations to find regimes that sustain acceptable quality for lower maintenance locations. Influence of mowing height and nitrogen on characteristics important for intensely managed areas of hybrid bermudagrasses (Cynodon dactylon [L.] Pers. var. dactylon ×

C. transvaalensis Burtt-Davy) (Guertal and Evans, 2006; Johnson et al., 1987; Tucker et al.,

2006) and seashore paspalum (Paspalum vaginatum Swartz) (Kopec et al., 2007) has also been researched.

Additional cultural practices have been studied to prevent or reduce the accumulation of thatch in zoysiagrasses. Lower mowing heights, verticutting, and aerification have been associated with decreased thatch layers, but mechanical injury from increased maintenance

(verticutting + aerifying) often caused reduced turf quality and weed invasion (Cockerham et al.,

1997; Weston and Dunn, 1985). Hollingsworth et al. (2005) observed similar reductions in quality of ultradwarf bermudagrasses when cultural management was too intense. Kenworthy and Engelke (1999) reported that two aerifications per year and 97 to 195 kg ha-1 N during the growing season produced the highest quality turf when Cavalier, ‘Crowne’, Palisades, and ‘El

Toro’ were maintained at 1.2 cm. Reducing N fertilization has been shown to lower thatch and tiller density in zoysiagrass, but clipping removal and growth regulator application had no effects on its accumulation (Soper et al., 1988).

Little research has been performed regarding the management of Zoysia spp. in Florida to date. Several newer cultivars of zoysiagrass are now available for use in the Southeastern United

States. This research was initiated to (i) characterize a general response (color, density, turf quality, thatch accumulation, and disease incidence) to nitrogen fertilization, mowing, and their interactions among zoysiagrass cultivars, and (ii) establish appropriate mowing height and fertility recommendations for each of the cultivars studied.

Materials and Methods

Separate fields of three Zoysia japonica cultivars (‘Empire’, ‘Palisades’ (Engelke et al.,

2002c), and ‘Ultimate’) and three Z. matrella cultivars (‘Cavalier’ (Engelke et al., 2002b),

‘Diamond’ (Engelke et al., 2002a), and ‘Pristine’) were planted from sprigs during the summer and fall of 2005. All plots were established by late summer of 2006. ‘JaMur’ (Z. japonica) and

‘Zeon’ (Z. matrella) were planted from sod in the fall of 2006 and had successfully rooted and begun spreading by the spring of 2007. Applications of irrigation, topdressing, fertilizer, pre- and postemergence herbicides, and fungicides were uniform between all cultivars before experiments were initiated. Mowing heights were gradually lowered to designated treatment levels beginning in January of 2007 to avoid scalping. Plots were verticut annually in March prior to fertilizer applications. All research was conducted on irrigated plots at the University of

Florida G.C. Horn Turfgrass Research Facility near Citra, FL in Candler sand (hyperthermic, uncoated Lamellic Quartzipsamments) during 2007 and 2008.

Coarse textured Z. japonica cultivars were evaluated independently of the fine textured Z. matrella cultivars because of differences in their required mowing heights and potential uses.

The experiments were split-plot designs nested within cultivars, with mowing heights as whole plots (4.5 × 1.5 m), and N fertilization levels as the subplots (1.5 × 1.5 m). Treatments were replicated three times. Empire, JaMur, Palisades, and Ultimate were mowed weekly at 2.5 and

5.0 cm with a mulching rotory mower. Cavalier, Diamond, Pristine, and Zeon were walk mowed twice weekly at 0.6 and 1.2 cm with a reel mower. Clippings were not collected. Total annual nitrogen treatments were 73, 171, and 268 kg ha-1 yr-1 for all cultivars, but split between four or five application dates in the Z. japonica and Z. matrella experiments, respectively. A rotary fertilizer spreader was used to deliver each initial yearly application uniformly across all plots from a balanced fertilizer (15N-5P-15K) on 1 March at the N rate of 24 kg ha-1. Subsequent applications were made by hand using Uflexx® (J. R. Simplot Company, Boise, ID) fertilizer

(46N-0P-0K), with the remaining yearly rates split equally on May 1, August 1, and October 1 for the Z. japonica study, and on May 1, July 1, September 1, and November 1 for the Z. matrella experiment. Plots were irrigated with approximately 1.3 cm of water immediately following fertilizer applications. All treatments on Pristine were inadvertently fertilized with an additional 48 kg N ha-1 from a balanced fertilizer (15N-5P-15K) in June of 2007.

Turfgrass quality, density, and genetic color were rated visually on a monthly basis.

Quality was rated on a scale of 1 to 9 with 1 = poor, 5 = acceptable, and 9 = excellent. The rating scale for density was 1 = least and 9 = most. For genetic color 1 = yellow, 5 = acceptable, and 9 = dark green. Bipolaris disease incidence was noted as symptoms occurred. Turfgrass quality ratings were adjusted as warranted by severity of the disease symptoms. Thatch depth was estimated fall 2007 on each plot by measuring the organic layer that accumulated between the soil surface and the turf canopy with a ruler from two 10.8 cm diameter cores.

Analysis of variance was performed to test whether cultivars, mowing heights, and N fertilization levels varied by date. Data were analyzed separately by date, and within each cultivar to further study the effects of the treatments. Mowing height and N fertilization level means were separated using a Fisher’s protected LSD where statistically significant.

Results and Discussion

Zoysia japonica

Cultivars responded differently (P ≤ 0.01) for turf density, turf quality, and thatch depth in the overall analysis of variance. Mowing height contributed to the observed variation (P ≤

0.01) in genetic color, turf quality, and thatch depth. Nitrogen fertilization rates were significant

(P ≤ 0.01) in all four measured traits and interactions between mowing height and N rate were found for turf density and quality. Turfgrass performance characteristics did not respond consistently over dates as indicated by the significant (P ≤ 0.01) interactions when analyzed as a split-plot in time (Table 5-1). Therefore, responses on each date within individual cultivars were evaluated to determine trends associated with mowing height and N rate treatments.

Generalities regarding genetic color were apparent in all four cultivars. N rate accounted for most of the observed variation across all dates, but mowing height was important in the late winter and early spring. Many of the mowing height × N rate interactions were due to the response of mowing height at the lowest fertility level when compared to either the medium or high rate. Acceptable genetic color ratings (≥ 5) were rarely found at the lowest N rate. The highest N fertility typically resulted in the greenest turf, but acceptable color was maintained at the medium N rate for most of the year outside of the winter months. Only the low mowed plots with the highest annual N fertility rated ≥ 5 during the winter. October fertilizer applications consistently improved genetic color in all treatments, and a negative response to the May

fertilizer application at the low N rate was observed in both years (Figure 5-1, Figure 5-2, Figure

5-3, Figure 5-4).

Empire zoysiagrass lost its color dramatically in the winter when mowed and fertilized at the highest levels (Figure 5-1). Palisades responded similarly with Empire, although mowing heights had a greater impact on genetic color throughout the year (Figure 5-3). Acceptable color ratings were observed at times on Ultimate at the lowest N rate. Beginning in July of 2008, scalping resulted in reduced genetic color and overall plot health at the higher mowing height

(Figure 5-4).

Density was primarily influenced by nitrogen fertility rate in all Z. japonica cultivars.

Interactions between mowing height and N rate for this characteristic were more common in the fall and winter months as illustrated in Empire, Palisades, and Ultimate where the relative differences in winter density between low and high mowing heights were greater at low nitrogen fertility compared to those at the medium and high N rates. JaMur (2007 and early 2008) and

Ultimate (2007 and 2008) maintained acceptable (≥ 5) density at low N fertility, and were generally denser than Empire and Palisades under the same treatments. A trend of decreasing density during the summer months at the low N rate was present in all cultivars. The late fertilizer application in October resulted in a favorable response in all treatments and cultivars, and may help these grasses compete with winter weeds in environments where the turf never completely goes dormant (Figure 5-5, Figure 5-6, Figure 5-7, Figure 5-8).

Nitrogen fertility was largely responsible for the differences in turf quality in all cultivars.

Mowing height had a significant effect in the fall before the October fertilization date and in later winter months. At the low N rate, turfgrass quality was generally greater when mowed low than high, and at times during the year resulted in acceptable (≥ 5) ratings for JaMur and Ultimate.

Turf quality was not adequately maintained at either mowing height under low fertility in Empire and Palisades. Fall and winter quality declines were more pronounced at the high than low mowing height when grown under high fertility, except in Palisades. Results indicate that acceptable quality in these cultivars may be sustained throughout the winter at either the medium or high fertilizer treatment when mowed lower. Negative responses were observed in both years to the May fertilizer application at the low N rate (Figure 5-9, Figure 5-10, Figure 5-11, Figure 5-

12). Dunn et al. (1995) stressed that zoysiagrass should only be fertilized with enough nitrogen to maintain acceptable density and turf quality, but that higher N rates may be necessary on sandier soils with low organic matter content.

The initial decline in fall turfgrass quality observed for most cultivars as the season began to change was less prominent in the high mowing height of Empire, suggesting that its prostrate growth habit may facilitate longer duration of quality until the first hard freeze (Figure 5-9).

Ultimate showed a dramatic response to the first fall (October 2007) fertilization application, although scalping at the higher mowing height in the late summer months of 2008 indicate that its density and thatch accumulation may influence long term turf quality (Figure 5-12). Fall and winter fertilizer applications in zoysiagrass have typically resulted in higher weed infestation, especially when the turf is dormant (Busey, 2003; Dunn et al., 1993; Henry et al., 1989).

There were no differences in thatch depth between N rates within individual cultivars when measured at the end of 2007. Thatch depth was numerically greater at the high mowing height in all cultivars, but significant differences (P ≤ 0.05) were only found at the low and medium fertility levels in Empire, and at the high nitrogen rate for Palisades and Ultimate.

Cockerham et al. (1997) observed less thatch at lower mowing heights in ‘DeAnza’ and

‘Victoria’ zoysiagrass. Generally, thatch depths were comparable for all cultivars at the low

mowing height, but at the high mowing height Empire and Ultimate performed similarly as did

JaMur and Palisades (Table 5-2).

Zoysia matrella

Much of the observed variation (P ≤ 0.01) for each turfgrass performance characteristic was due to cultivar and N rate in the overall analysis of variance, although various significant (P

≤ 0.05) interactions between cultivar and other main effects suggested that important seasonal information unique to each cultivar could be gathered if they were analyzed separately by date.

Thatch depths varied (P ≤ 0.05) with mowing heights. Differences in thatch were not attributed to the effects of cultivar and N rate (Table 5-1), but were further explained in an analysis of their interaction (Table 5-3).

Nitrogen fertility was primarily responsible for the observed responses in genetic color for all Z. matrella cultivars on most dates. Maintaining acceptable color at the low N rate may not be possible over time, but ratings ≥ 5 were sustained at the medium and high fertility rates in both mowing heights over the entire year. November fertilizer applications provided drastic improvement to winter color in all treatments, and except for Pristine in 2008, there was generally no response from the May fertilizer application at the low N rate in either year (Figure

5-13, Figure 5-14, Figure 5-15, Figure 5-16).

Mowing Cavalier at the higher height during the summer months typically resulted in greener genetic color at both the medium and high nitrogen treatments, although similar color ratings were observed at both mowing heights at low N fertility. Peaks and valleys in the genetic color of Cavalier demonstrated instability through time (Figure 5-13). Diamond (Figure 5-14) and Pristine (Figure 5-15) had more consistent genetic color and responded similarly to comparable treatments beginning in January of 2008. Mowing height × N rate interactions for genetic color in Zeon were often observed in the month immediately following fertilizer

application, and were mostly due to large changes in color at low N fertility in the higher mowing height when compared to more consistent responses for the medium and high nitrogen rates at both mowing heights (Figure 5-16).

Diamond and Pristine have greater turf density and finer leaf-texture than either Cavalier or Zeon. Scalping was observed at the high nitrogen rate for both of these cultivars as noted by reductions in density during October of 2007. Hale (2006) described the same trend in three Z. matrella cultivars when fertilized with annual rates of 146 and 195 kg N ha-1. Results indicate that it may be possible to maintain acceptable density at the lowest nitrogen fertilization level in

Diamond and Pristine while avoiding the scalping associated with higher fertility. Leaf spot

(Bipolaris spp.) damage severely reduced turf density in Cavalier during the late summer and fall of 2007. Although the injury was relatively uniform across all treatments, plots at the low mowing height and highest N rate were the least dense. Bipolaris disease incidence was also noted on Zeon at this time, but the damage was not as prominent. When Bipolaris was active on

Cavalier and Zeon their density was influenced by mowing height and nitrogen fertility, but when Bipolaris was not limiting, nitrogen fertilization rates were more important (Figure 5-17,

Figure 5-18, Figure 5-19, Figure 5-20). Green et al. (1994) found that large patch (Rhizoctonia solani) more severely blighted 12 zoysiagrass cultivars at lower mowing heights, and that neither nitrogen rate or source influenced disease severity.

Trends in turfgrass quality were not evident for the entire set of Z. matrella cultivars because of genetic differences, confounding results from disease incidence, and fertilizer application error. Cavalier zoysiagrass generally remained unacceptable due to lasting effects from the initial July 2007 Bipolaris damage, but symptom severity appeared to decrease over time at the higher mowing height possibly as a result of reduced stress (Figure 5-21).

Establishment of mowing treatments and early disease incidence in Zeon impeded the detection of clear treatment differences on turf quality until the winter of 2007. Nitrogen fertility significantly influenced the expression of quality when the turf was not under diseases pressure, with the highest N rates resulting in the best turf. Bipolaris symptoms in August and September of 2008 had less impact on turfgrass quality at the high mowing height under medium or low N fertility (Figure 5-22). Diamond typically performed better at the low mowing height, especially in the late fall and early spring when the majority of seedheads were not removed at the high mowing height. It may be possible to maintain turf quality during the fall and winter months at the low mowing height and nitrogen rate, but not when the turf is actively growing or has been recently injured by mechanical cultivation. N rate had a greater influence on turf quality during the growing season, but severe scalping was observed in the high fertility treatment at both mowing heights in September and October of 2007 (Figure 5-23). Effects of fertility and mowing height became more apparent on Pristine in 2008 when the response from an incorrect fertilization in June of 2007 had subsided. Various degrees of scalping occurred at the high mowing height periodically throughout the study. Slow recovery from mechanical injury following verticutting was noted at the low N rate. Weston and Dunn (1985) described similar injury and increased weed encroachment in Meyer zoysiagrass when vertical mowing and core aerification were used to manage thatch. Overall turfgrass quality was consistent during 2008 at the low mowing height and medium N rate, which may be the appropriate management combination to avoid the above mentioned scalp damage (Figure 5-24).

Trends were not clear regarding the effect of mowing height and N rate on thatch depth in the Z. matrella cultivars. Significant (P ≤ 0.05) differences were observed, but none were meaningful. Thatch depth at the lower mowing height was similar among all cultivars, although

Pristine appeared to have more thatch at the high mowing height (Table 5-3). This occurrence could have directly resulted from the inadvertent fertilizer application. Soper et al. (1988) found a direct relationship between increasing N fertilization and thatch accumulation in Meyer zoysiagrass. They also concluded that stolons are less easily degraded than leaf tissue, and were therefore more influential on thatch depth than returned mower clippings.

Conclusions

Understanding how cultivars respond to basic mowing and fertility treatments is critical in developing more comprehensive management practices. Genetic differences and varying levels of adaptation to the environment in Florida were evident among this set of genotypes, although a few trends were common among both Zoysia species. Nitrogen rate had a greater impact on most of the observed characteristics when the turf was actively growing, but mowing height was important during the winter and in times of stress. Our results and observations suggest that increasing nitrogen fertility to temporarily maximize genetic potential for color and density may have lasting detrimental effects on turfgrass quality. Problems concerning scalping, thatch accumulation, disease incidence, and drought injury may be lessened if consideration is given to maintaining acceptable, rather than exceptional, conditions. Preventing damage to zoysiagrasses is especially important because they lack the recuperative potential of other faster growing grasses (Busey and Myers, 1979).

Each cultivar responded uniquely to the environment and imposed treatments. Turfgrass density was not maintained in Empire or Palisades at the lowest N rate, but for different reasons than observed in Cavalier and Zeon. Simply increasing nitrogen fertility could improve turf density to acceptable levels in the Z. japonica cultivars, but higher N fertility may require a preventative fungicide program for the two Z. matrella genotypes. JaMur, Ultimate, Diamond, and Pristine all had acceptable density at the low nitrogen rate, but often did not have adequate

color to sustain turf quality. In this case additional iron supplementation could augment genetic color without the risk of incurring detrimental effects from excess N fertilization. More than one season of growth may be required to detect the effects of mowing height and N rate treatments on thatch accumulation.

This research should serve as the framework on which additional practices are tested for each cultivar. The results herein suggest that lowering mowing heights during the winter months, and raising them during times of stress will improve overall turf quality. Likewise, nitrogen fertility should be adjusted accordingly with incidence of disease or mechanical injury.

Only through further testing of N source and rate, additional nutrients, timing of application, thatch management (grooming, verticutting, aerifying, and topdressing), and fungicide programs in the presence or absence of biotic and abiotic stresses will the best management practices be formulated.

Table 5-1. Mean squares for turfgrass performance characteristics and thatch depth of four Zoysia japonica cultivars and four Zoysia matrella cultivars evaluated at two mowing heights and three nitrogen fertility rates near Gainesville, FL. Mean squares –––––––––– Zoysia japonica –––––––––– –––––––––– Zoysia matrella –––––––––– Genetic Turf Turf Thatch Genetic Turf Turf Thatch Source df color density quality depth† color density quality depth† Cultivar (C) 3 24.98 277.40** 60.35** 54.68** 86.48** 685.74** 351.72** 38.85 Rep (C) 8 3.82 3.25 5.21 0.63 8.76 8.86 13.04 15.44 Mowing height (M) 1 77.05** 7.87 40.67** 308.88** 21.54* 5.06 7.72 47.69* M × C 3 1.24 3.53 3.88 7.88 4.59 39.19* 46.61* 8.22 Rep × M (C) 8 1.09 1.14 0.71 3.04 1.04 6.32 7.04 7.18 N rate (N) 2 693.03** 443.48** 520.56** 13.60* 754.13** 124.31** 260.40** 0.42 N × C 6 2.97 2.66 3.41 2.24 12.85** 12.34* 21.40** 4.38* N × M 2 3.15 9.82** 10.79** 6.17 12.57** 3.69 16.50 0.34 N × C × M 6 0.53 1.88 1.62 3.20 0.75 10.44* 8.72 1.72 Rep × N (C × M) 32 0.83 0.90 1.44 3.78 1.61 4.00 4.48 1.54 Date (D) 17 44.45** 10.84 25.78** 32.66 15.22 19.15

71 D × C 51 5.16** 5.72** 3.82** 10.11** 10.59** 13.29** D × M 17 3.88** 1.14 2.94** 2.30 3.16 3.40 D × N 34 10.40** 4.50** 5.53** 9.61** 6.32** 11.70** D × C × M 51 1.04** 0.70** 1.11** 0.81** 3.52** 3.63** D × C × N 102 0.79** 0.68** 0.88** 1.19** 1.39* 2.46** D × M × N 34 0.73** 0.36 0.49 0.95** 1.38 2.73** D × C × M × N 102 0.35 0.25 0.41 0.41 0.98* 1.41** MSE 816 0.28 0.32 0.46 0.38 0.77 1.00 % CV 10.0 9.3 12.0 11.0 10.3 14.9 18.8 9.6 *, **Significant at the 0.05 and 0.01 probability levels, respectively. †Thatch depth only evaluated on one date.

4 Genetic Color 3