Gossypium arboreum Accessions Resistant to Rotylenchulus reniformis

Abstract

In the southeastern United States, reniform nematode (Rotylenchulus reniformis) is a serious pest of upland cotton (Gossypium hirsutum), a species which has no naturally occurring resistance against this nematode. To identify sources of reniform nematode resistance in species closely related to upland cotton, 222 G. arboreum accessions from the U.S. germplasm collection were evaluated in repeated growth chamber experiments. In initial screenings, root infection was measured 4 wks after inoculation. The 15 accessions supporting the fewest infections (PI 529992, PI 615755, PI 615766, PI 615788, PI 615848, PI 615856, PI 615950, PI 615977, PI 615991, PI 616008, PI 616016, PI 616062, PI 616126, PI 616159, and A2 553) were evaluated again in confirmation tests lasting 8 wk. The combined totals of nematodes extracted from soil and eggs extracted from roots were analyzed. All 15 accessions tested supported significantly smaller reniform nematode populations than the susceptible controls (G. hirsutum cultivar Deltapine 16 and G. arboreum accession PI 529729). Nine accessions (PI 529992, PI 615755, PI 615766, PI 615788, PI 615856, PI 615950, PI 615991, PI 616008, and PI 616159) supported reniform nematode populations comparable to the resistant control (G. arboreum accession PI 615699), and accession PI 615848 had significantly fewer reniform nematodes than the resistant control. Cotton breeders would benefit from introgressing the newly identified resistance from these accessions into their upland cotton improvement programs.

Cotton (Gossypium hirsutum L.) farmers from Texas to the Atlantic seaboard experience yield losses as a result of the reniform nematode (Rotylenchulus reniformis Linford and Oliveira) on an annual basis. Losses to reniform nematode for the 2013, 2014, and 2015 growing seasons averaged 3.3%, 6.1%, and 4.0% for cotton in Louisiana, Mississippi, and Alabama, respectively (Lawrence et al., 2014, 2015, 2016). A number of factors including lack of resistance within commercially available cultivars (Robinson et al., 1999; Usery et al., 2005; Starr et al., 2007), loss of effective soil-applied fumigants and nematicides from the market (Starr et al., 2007; Mueller, 2011), and grower preference for cotton monoculture over crop rotation (Robinson, 2007; Starr et al., 2007) allow nematode survival and reproduction resulting in population densities at or above damaging thresholds at planting and throughout the cropping season.

Host plant resistance would be highly advantageous to cotton growers because it is cost effective, environmentally friendly, simple to deploy, and it persists throughout the entire growing season. The primary reason for the lack of reniform nematode resistant cultivars is the lack of high levels of resistance to this nematode in G. hirsutum. Robinson et al. (2004) surveyed more than 1,800 primitive G. hirsutum accessions obtained from the U.S. National Plant Germplasm System (NPGS) cotton collection and found only six that were moderately resistant.

Germplasm lines have been released with resistance to reniform nematode derived from relatives of G. hirsutum. The tetraploid species Gossypium barbadense L. is the source of resistance in several germplasm lines released within the past decade. In 2010, two breeding lines of cotton, TAM RKRNR-9 (PI 662039) and TAM RKRNR-12 (PI 662040), with reniform nematode resistance derived from G. barbadense TX 110 (PI 163608) were released (Starr et al., 2011). Gossypium barbadense accession GB 713 (PI 608139) was the source of reniform nematode resistance in four other germplasm lines released in 2012. Three lines, M713 Ren1 (PI 665928), M713 Ren2 (PI 665929), and M713 Ren5 (PI 665930), were developed from a cross between G. barbadense GB 713 and the G. hirsutum cultivar SureGrow 747 (McCarty et al., 2013). The fourth germplasm line, BARBREN-713 (PI 671965), was developed by crossing G. barbadense GB 713 with the cultivar Acala NemX, followed by several backcrosses to G. hirsutum lines (Bell et al., 2015); this line has resistance to Meloidogyne incognita (Kofoid and White) Chitwood in addition to reniform nematode resistance. To date, no commercial cultivars have been released that have these germplasm lines in their pedigrees.

A greater research challenge is the exploitation of the reniform nematode resistance found in diploid Gossypium species. Transferring resistance from diploid Gossypium species into tetraploid cotton is difficult. Barriers to hybridization between the different species include mechanisms that prevent fertilization or inhibit development of viable seed from successful fertilizations (Brubaker et al., 1999; Mehetre et al., 2003; Mehetre and Aher, 2004; Ganesh Ram et al., 2008). Techniques such as bridging lines (Brubaker et al., 1999; Romano et al., 2009), induced polyploidy (Mehetre et al., 2003), in vitro interspecific fertilization (Liu et al., 1992), protoplast fusion (Sun et al., 2006), and ovule culture (Stewart and Hsu, 1977, 1978; Gill and Bajaj, 1984, 1987) have been used to overcome these breeding limitations.

Immunity to reniform nematode in G. longicalyx Hutch. & Lee (Yik and Birchfield, 1984; Stewart and Robbins, 1996); resistance in G. arboreum L. (Carter, 1981; Stewart and Robbins, 1995; Sacks and Robinson, 2009), G. somalense (Gurke) Hutch. (Yik and Birchfield, 1984), and G. stocksii Mast. Ex. Hook. (Yik and Birchfield, 1984); and moderate levels of resistance in G. aridum (Sacks and Robinson, 2009), G. herbaceum (Yik and Birchfield, 1984), and G. raimondii Ulbr. (Yik and Birchfield, 1984), have been reported. With the exception of G. longicalyx, in which all accessions tested to date have exhibited immunity, variability in resistance to reniform nematode exists within the diploid Gossypium species.

To date, the only germplasm lines released with resistance from a diploid species are LONREN-1 and LONREN-2, with resistance that had been introgressed from G. longicalyx (Bell et al., 2014). However, this resistance has been linked to intolerance (Sikkens et al., 2011; Weaver et al., 2013), with plants exhibiting stunting when challenged with high inoculum levels of the nematode. Because of this problem, nearly all breeding programs have stopped using this source of resistance. Gossypium hirsutum lines with reniform nematode resistance introgressed from G. arboreum accession A2-190 (PI 615699) (Sacks and Robinson, 2009) and G. arboreum accession A2-19 (PI 129723) (Avila et al., 2005) have been developed, though no germplasm lines from these programs have been released to date.

Because reniform nematode resistance has just recently become available in upland cotton, no data are available with respect to the durability of any one source of resistance. Variability within reniform nematode has been well documented on a genetic, morphological, and physiological basis (Dasgupta and Seshadri, 1971; Nakasono, 2004; Agudelo et al., 2005b; Arias et al., 2009; McGawley et al., 2010; Leach et al., 2012). Over time, reniform nematode may adapt to one or more resistance sources, as has been documented with development of races in pathogens such as Phytophthora infestans (Mont.) de Bary and Heterodera glycines Ichinohe. Use of a single source of resistance over time may result in development of nematode biotypes that can reproduce on the resistant cultivar (Young, 1998), so rotation among different resistance sources may be necessary to reduce selection pressure on the nematodes (Starr and Roberts, 2004). If different resistance genes can be identified, they could be combined (“pyramided”) into the same plant to make resistance more durable.

The objectives of this research were to evaluate a selection of Gossypium arboreum accessions for their reaction to the reniform nematode, and to identify sources of host plant resistance that could be introgressed into upland cotton and used to manage this pathogen.

Materials and Methods

Identification of resistant lines:

A total of 222 G. arboreum accessions were evaluated in growth chamber tests for resistance to infection by reniform nematode. The specific accessions tested are listed in Tables 1, 2, and 3. Seeds not already in the authors’ research collections were obtained from the NPGS (College Station, TX).

Table 1.

Infection of Gossypium roots by Rotylenchulus reniformis females 4 wk after inoculation in growth chamber Test 1. All accessions are Gossypium arboreum except for susceptible control Gossypium hirsutum cultivar Deltapine 16.

Table 2.

Infection of Gossypium roots by Rotylenchulus reniformis females 4 wk after inoculation in growth chamber Test 2. All accessions are Gossypium arboreum except for susceptible control Gossypium hirsutum cultivar Deltapine 16.

Table 3.

Infection of Gossypium roots by Rotylenchulus reniformis females 4 wk after inoculation in growth chamber Test 3. All accessions are Gossypium arboreum except for susceptible control Gossypium hirsutum cultivar Deltapine 16.

Accessions were arbitrarily divided into three screening tests of approximately 75 entries each due to growth chamber space limitations. The susceptible controls Gossypium hirsutum cultivar Deltapine 16 (Yik and Birchfield, 1984; Robinson and Percival, 1997) and G. arboreum accession PI 529729 (Sacks and Robinson, 2009; Erpelding and Stetina, 2013), and the resistant control G. arboreum accession PI 615699 (Sacks and Robinson, 2009) were included in each test. The experimental design for each screening was a completely randomized design with three replications, and each test was repeated. The growth chamber temperature was maintained at 28°C and the daylength was set at 16 hours. Soil moisture was maintained using an automated watering system, with the timing adjusted periodically during the experiment to supply additional water as plants grew.

Screening test protocols were similar to those described by Stetina et al. (2014). Briefly, single plants of each accession were established in conical plastic pots (Ray Leach SL-10 Cone-tainer, Stuewe & Sons, Inc., Tangent, OR) containing 120 cm³ of a steam-sterilized soil mixture consisting of one part sandy loam soil mixed with two parts sand. Approximately 7 days after planting, soil in each pot was infested with 1,000 reniform nematodes (mixed vermiform life stages) suspended in 1 ml water. Mississippi reniform nematode population MSRR04 (Arias et al., 2009), originally isolated from upland cotton and maintained in a greenhouse on tomato (Solanum lycopersicon cultivar Rutgers), was used for all experiments. Plants were harvested 4 wk after inoculation. Shoots were removed at the soil line and discarded. Roots were separated from soil, stained with red food coloring using standard protocols (Thies et al., 2002), and the number of swollen females attached to the roots were counted at ×50 magnification. After counting, roots were allowed to drain briefly on paper towels to remove excess water and fresh weights were recorded. Counts were expressed as females per gram of fresh root tissue to compensate for differences in root sizes.

In addition to statistically comparing root infection levels, accessions within each test were classified based on a nematode index, following that described by Schmitt and Shannon (1992) for soybean cyst nematode. Infection on an accession is expressed as a percentage of the average number of females that developed on susceptible G. hirsutum cultivar Deltapine 16. Based on the nematode index, accessions were classified as resistant (nematode index <10%), moderately resistant (10% to 30%), moderately susceptible (31% to 60%), or susceptible (>60%).

Confirmation of reaction to reniform nematode:

A subset consisting of 15 of the most resistant accessions identified in the initial screening tests was further evaluated in a longer-duration test that measured reniform nematode reproduction. As in the screening tests, the susceptible controls Gossypium hirsutum cultivar Deltapine 16 and G. arboreum accession PI 529729, and the resistant control G. arboreum accession PI 615699 were included. To monitor survival of the nematode with no roots present, a fallow treatment also was included.

Test establishment and inoculation procedures were the same as described for the initial screenings. The experimental design was a completely randomized design with five replications, and the test was repeated. The test duration was extended to 8 wk. At the end of the test, standard elutriation (Byrd et al., 1976) and sucrose centrifugation (Jenkins, 1964) protocols were used to extract vermiform stages of nematodes from all of the soil in each pot. In addition, eggs were extracted from the root system by cutting the roots into 2.5-cm segments, stirring for 10 min in a 0.6% NaOCl solution (Hussey and Barker, 1973), and collecting eggs on a standard 25-µm-pore sieve. Egg and vermiform counts were added together, and total numbers were analyzed.

In addition to statistically comparing reniform nematode population sizes, a reproduction factor was determined for each of the accessions. The reproduction factor is calculated by dividing the number of nematodes per pot at the end of test by the initial inoculum level of 1,000 nematodes. Reproduction factor values of 1.0 or more indicate that the plant is a good host for the nematode; poor hosts have values smaller than 1.0 (Walters et al., 1996).

Statistical analysis:

Prior to analysis of variance (ANOVA), nematode counts were subjected to log₁₀(x+1) transformation to normalize data; backtransformed means are presented. Initial data analyses identified no significant differences between trials, and no significant interactions between trial and accession. Therefore, data from both trials of each identification and confirmation test were combined for final analysis, and trials and their interactions were modeled as random effects. Where significant differences among genotypes were found using ANOVA, differences of least squares means (P ≤ 0.05) were used to compare means. SAS statistical software (PROC MIXED; SAS Institute, Cary, NC) was used for analysis.

Results

The reactions to reniform nematode for all 222 G. arboreum accessions evaluated are presented in Tables 1, 2, and 3. The susceptible controls were significantly different from the resistant control in each of the three tests based on the number of females infecting the roots, although the number of infections on the resistant control was higher than expected in Test 2. These initial screening experiments identified 19 susceptible, 96 moderately susceptible, 100 moderately resistant, and 7 resistant accessions in total.

Though not statistically distinguishable from the control, four accessions classified as resistant had lower infection indices than the resistant control: PI 529992, PI 615950, PI 615977, and PI 616008 (Table 3). At the other end of the spectrum, five accessions classified as susceptible had higher infection indices than the susceptible controls: PI 183202, PI 129742, PI 408772, PI 529806 (Table 1); and PI 616101 (Table 3).

The 15 most resistant accessions identified in the initial screenings were tested again in longer experiments to confirm their reaction to the reniform nematode (Table 4). All accessions tested reduced reniform nematode populations compared to the susceptible controls. Nine accessions were comparable to the resistant control with respect to final population sizes, and accession PI 615848 supported significantly smaller reniform nematode populations than the resistant control. However, none of the accessions suppressed the populations to the same level as the fallow treatment. A comparison of the reproduction factors showed that 14 accessions and the fallow treatment were comparable to the resistant control, though only PI 615848 and the fallow treatment had reproduction factors less than 1.0, indicative of poor host status.

Table 4.

Comparison of reniform nematode population development on 17 Gossypium arboreum accessions, the susceptible control Gossypium hirsutum cultivar Deltapine 16, and one fallow treatment in a growth chamber.

Discussion

Ten G. arboreum accessions were identified as resistant to reniform nematode in both initial screening and subsequent confirmation tests. This conclusion was based on the number of females infecting the roots and on the nematode population development in growth chamber tests as compared to the resistant control G. arboreum accession PI 615699. The nine accessions that were comparable to the resistant control in supporting reniform nematode population development were PI 529992, PI 615755, PI 615766, PI 615788, PI 615856, PI 615950, PI 615991, PI 616008, and PI 616159. One accession, PI 615848, was more effective than the resistant control at suppressing reniform nematode population development, and had a reproduction factor of 0.8, indicative of poor host status. All of these sources supported 3% or less of the reniform nematode population development that was observed on the susceptible G. hirsutum control cultivar Deltapine 16. As such, any of them would be excellent candidates for inclusion in a germplasm improvement program.

Results from this study indicate that a reduced number of infections and smaller population sizes are associated with the 10 resistant accessions identified. However, specific mechanisms governing the successful establishment and maintenance of a feeding site, the rate of nematode development, and the number of eggs produced by each female were not evaluated (Agudelo et al., 2005a; Starr et al., 2011; Stetina, 2015), though any or all of these factors could be contributing to the observed resistance. Discerning the mechanism(s) behind the observed reniform nematode population suppression could be the subject of future research.

Within the subset of 222 accessions that were tested from the G. arboreum germplasm collection, the plants were divided fairly evenly between the resistant and susceptible ends of the reniform nematode resistance spectrum. Most of the accessions tested were classified as either moderately resistant or moderately susceptible based on root infection levels, with only a few lines initially identified as resistant. The subset of accessions tested represents only about 12% of the G. arboreum collection. A significant time investment will have to be made to screen the remainder of the accessions using the methods employed in this study. To facilitate discovery of new sources of resistance in this germplasm collection, molecular markers associated with the resistance already documented are needed. The markers could be used to rapidly evaluate the remaining accessions to identify accessions having similar DNA banding patterns as resistant accessions so that future screening efforts could be directed toward identifying putatively unique types of resistance.

In the screening experiments, 19 accessions susceptible to the reniform nematode were identified. Of these, PI 129742, PI 183202, PI 408772, PI 529806, and PI 616101 had higher female indices than the susceptible controls. While these accessions are not useful for developing cultivars resistant to reniform nematode, they do have utility in understanding how resistance is controlled. Populations from crosses between the susceptible and resistant accessions can be studied to determine how resistance is inherited, to identify molecular markers for resistance, and to map the location of the gene(s) conferring resistance.

A limitation of this study is that the accessions were screened using a single isolate of reniform nematode. There are reports in the literature documenting cotton (Agudelo et al., 2005b; Arias et al., 2009; McGawley et al., 2010) and soybean (Agudelo et al., 2005b; McGawley et al., 2011) lines responding differently to unique geographic populations of reniform nematode. Therefore, the accessions identified as resistant in this study could show a different level of resistance if challenged with different populations of the nematode.

In summary, this research provides new phenotypic information on 222 G. arboreum accessions, including the identification of 10 accessions with useful levels of reniform nematode resistance. Public and private cotton breeding programs could benefit from using these resistant accessions as parents, although there may be challenges related to the introgression of the resistance that were not evaluated in this study.