882 www.crops.org crop science, vol. 54, may–june 2014

RESEARCH

The causal agent of SLB, the ascomycete Cochliobolus het-erostrophus (Drechs.) Drechs. [anamorph = Bipolaris maydis

(Nisikado) Shoemaker], penetrates and colonizes the intercellular spaces of the maize leaf, ultimately causing necrotic lesions. The resulting disease lesion phenotype is easily visually assessed and highly heritable (Balint-Kurti et al., 2007; Kump et al., 2011; Zwonitzer et al., 2009).

In general, though not in every case (Belcher et al., 2011), quantitative disease resistance (QDR) in plants is controlled by many small-effect quantitative trait loci (QTL) (Poland et al., 2011; Wisser et al., 2006). Although the additive effects of these

Yield Effects of Two Southern Leaf Blight Resistance Loci in Maize Hybrids

Jose H. Santa-Cruz, Kristen L. Kump, Consuelo Arellano, Major M. Goodman, Matthew D. Krakowsky, James B. Holland, and Peter J. Balint-Kurti*

ABSTRACTIn this study we examined the effects of two quantitative trait loci (QTL) for southern leaf blight (SLB) resistance on several agronomic traits including disease resistance and yield. B73–3B and B73–6A are two near-isogenic lines (NILs) in the background of the maize (Zea mays L.) inbred B73, each carrying one introgression (called 3B and 6A respectively) encompassing a QTL for SLB resistance. Sets of isohybrid trip-lets were developed by crossing B73, B73–3B, and B73–6A to several inbred lines. A subset of these triplets for which the B73–3B and/or B73–6A hybrid was significantly more SLB resis-tant than the B73 check hybrid was selected and assessed in multi-environment yield trials with and without disease. In the presence of SLB, 3B was associated with an approximately 3% yield increase over B73. 6A was associated with a yield advantage in the presence of SLB in spe-cific pedigrees where the 6A resistance pheno-type was highly expressed. Results suggested that both introgressions might confer a yield cost in the absence of SLB, but only introgression 6A was associated with a statistically significant reduction. We present evidence to suggest that the yield cost is associated with the resistance phenotype rather than with linkage drag.

J.H. Santa-Cruz and P.J. Balint-Kurti, Dep. of Plant Pathology, N.C. State Univ., Raleigh, NC 27695; K.L. Kump, M.M. Goodman, M.D. Krakowsky, and J.B. Holland, Dep. of Crop Science, N.C. State Univ., Raleigh, NC 27695; C. Arellano, Dep. of Statistics, N.C. State Univ., Raleigh, NC 27695; M.D. Krakowsky, J.B. Holland, and P.J. Balint-Kurti, USDA-ARS, Plant Science Research Unit, Raleigh NC 27695. J.H. Santa-Cruz and K.L. Kump contributed equally to this work. Received 19 Aug. 2013. *Corresponding author (pjbalint@ncsu.edu ).

Abbreviations: CL, NCSU Central Crops Research Station in Clay-ton, NC; CTAB, cetyltrimethylammonium bromide; dQTL, disease-resistance quantitative trait locus (loci); DTA, days to anthesis; DTS, days to silking; KI, Cunningham Research Station in Kingston, NC; LE, Peanut Belt Research Station in Lewiston, NC; NIL, near isogenic line; PCR, polymerase chain reaction; SNP, single nucleotide polymor-phism; QDR, quantitative disease resistance; QTL, quantitative trait locus (loci); R, resistance; sAUDPC, standardized area under the dis-ease progress curve; SH, Sandhills Research Station in Jackson springs, NC; SLB, southern corn leaf blight; SSR, simple sequence repeat.

Published in Crop Sci. 54:882–894 (2014). doi: 10.2135/cropsci2013.08.0553 © Crop Science Society of America | 5585 Guilford Rd., Madison, WI 53711 USA

All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permission for printing and for reprinting the material contained herein has been obtained by the publisher.

Published March 21, 2014

crop science, vol. 54, may–june 2014 www.crops.org 883

disease-resistance QTL (dQTL) do create large pheno-typic differences, little is known about how much indi-vidual dQTL contribute to differences in agronomic traits, especially yield, in inbred lines and hybrids (Frey et al., 2011). Because most dQTL have small and some-what variable phenotypic effects, field experiments of a conventional scale would likely lack adequate power to detect their potentially correlated effects on agronomic characteristics. To detect effects of single QTL, it would be best to study QTL of larger magnitude in a pathosys-tem for which host resistance is highly heritable. Two SLB resistance loci, which we have called 6a and 3b, fit these criteria (Belcher et al., 2011; Zwonitzer et al., 2009).

B73–3B and B73–6A are two near isogenic lines (NILs) in the background of the commonly-used maize line B73. Each carries a single introgression, referred to as“3B” and “6A,” derived from the highly SLB-resistant line NC292; otherwise, the lines are almost entirely isogenic to B73. These introgressions carry, respectively, the SLB resistance alleles at the 3b and 6a bloci (Belcher et al., 2011) located in bins 3.04 and 6.01 of the maize genome (Davis et al., 1999).

Several studies have identified strong-effect SLB dQTL in bins 3.04 and 6.01. The 3.04 SLB dQTL has been identified at almost precisely the same locus in all these studies. In every case the susceptibility allele was derived from B73 but the resistance allele derived from a number of different lines, in particular NC292 or its progenitor NC250 (Bubeck, 1991; Zwonitzer et al., 2009) and Mo17 (Balint-Kurti et al., 2007; Kump et al., 2010). Zwonitzer et al. (2009) showed that the NC292 resistance alleles at both of these loci were recessive to the suscepti-bility alleles derived from B73.

Several studies have assessed the cost of disease resis-tance to the host, with most examining monogenic resis-tance conferred by major resistance (R) genes. Studies in Arabidopsis thaliana (L.) Heynh have demonstrated that, in the absence of disease, R genes are associated with a reduc-tion in fitness. For example, Tian et al. (2003) reported a 9% reduction in seed production due to the presence of the major R gene RPM1 in uninoculated A. thaliana plants. The RPW8 locus, which contains two genes that together confer broad-spectrum resistance in A. thaliana, was examined using a transgenic approach and determined to have a negative effect on fitness when plants were not infected (Orgil et al., 2007). The fitness costs of resistance genes can be high enough to reduce yield even under dis-ease pressure. A yield cost was associated with the pres-ence of the resistance gene RPS2 in A. thaliana, even in the presence of disease, when there was no intraspecific competition (Korves and Bergelson, 2004). In addition to studies of R genes, at least two studies have examined the costs of QDR. When lines of Brassica rapa L. were recur-rently selected for QDR against Leptosphaeria maculans and Peronospora parasitica, causal agents of blackleg disease and

downy mildew, respectively, and then assessed for fitness, the QDR against P. parasitica had a negative effect on host growth rate in the absence of disease (Mitchell-Olds and Bradley, 1996). However, QDR does not always carry a fitness cost. Working with near-isogenic hybrids, Frey et al. (2011) concluded that the maize major anthracnose stalk rot resistance QTL Rcg1 does not confer a yield cost under disease-free conditions in the field when compared to hybrids without Rcg1.

Here we report the results of our investigation into the agronomic effects of the 3B and 6A introgressions. Our objective was to estimate the effects of these two genomic introgressions on SLB resistance, yield, and other agro-nomic traits in inbred and hybrid material under infected and disease free conditions.

MATERIALS AND METHODSPlant MaterialNC250 is a highly SLB resistant, yellow dent, inbred line devel-oped by the NCSU maize breeding program from the cross (Nigeria Composite A-Rb × B37) × B37 (Thompson and Bergquist, 1984). B73 is an inbred line with good agronomic qualities derived by recurrent selection (Cycle 5) of the Iowa Stiff Stalk Synthetic (Russell, 1972), while B37 is an inbred line from an earlier selection cycle of the same synthetic. NC292 was obtained by crossing B73 to NC250P (the progenitor of NC250) and backcrossing three times to B73, followed by several cycles of ear-to-row selfing. Importantly, selection for SLB resistance was performed at each generation. Therefore, NC292 is agro-nomically similar to B73, but highly resistant to SLB. B73–3B and B73–6A are two B73-background NILs, each containing one introgression from NC292. NC292-specific introgressions were isolated by backcrossing NC292 to B73 twice and then using simple sequence repeats (SSR) marker-assisted selection to derive an NIL containing a single introgression (either intro-gression 3B or introgression 6A). Further details on the devel-opment of this material can be found elsewhere (Belcher et al., 2011; Zwonitzer et al., 2009). The 6A introgression comprises almost 11 Mb (out of a total genome size of ~2500 Mb), while the 3B introgression comprises ~5.6 Mb.

For a pilot study, families were used that derived the 6A introgression from B73–6A or the 3B introgression from Mo17 (rather than from NC292). Inbred line Mo17 was developed by M.S. Zuber at the University of Missouri as part of C.O. Grogan’s thesis project. It was released in 1964 (Troyer, 1999; Zuber, 1973) and it is most known for its use in the very popular hybrid B73 × Mo17. B73-Mo17.3B is an NIL containing the Mo17 3B introgression and was derived by backcrossing the Mo17 x B73 F1 to B73 four times and using marker-assisted selection to ensure maintenance of 3B introgression and purg-ing of other Mo17 alleles. For the pilot study, F2:3 families with the 3B introgression were derived by crossing the B73-Mo17.3B NIL to B73, and then selfing to the F3 generation; F2:3 families with the 6A introgression were generated by self-pollinating the resulting F2 individuals from the B73–6A × B73 cross.

884 www.crops.org crop science, vol. 54, may–june 2014

visually to obtain the measurement of standardized area under the disease progress curve (sAUDPC), as explained below.

Preliminary SLB Evaluation and Selection of Isogenic Triplet-Tester CombinationsIsogenic hybrids were developed by crossing B73, B73–3B, and B73–6A to fifteen tester lines or pedigrees (Table 1). Since the resistance conferred by the introgressions is known to be recessive with respect to the corresponding B73 allele, and thus may be recessive to the corresponding alleles in some of the tester lines, the F1 progenies (isohybrids) were evaluated for SLB resistance. Seed for this disease trial was generated during the summer of 2009 in Clayton, NC. Isohybrid sets from the 15 pedigrees were evaluated for SLB resistance during the winter of 2009 to 2010 in Homestead, FL. Fifteen seeds of each of the three isogenic hybrids, as well as their parent lines, were planted in three replications of a randomized incomplete block design. Blocks consisted of B73, the tester inbred, B73–3B, B73–6A, and three isohybrids: B73 × tester, B73–3B × tester, and B73–6A × tester. Plots were individual rows 3.6 m in length sown with 15 seeds with row spacing of 0.91 m and an alley after each plot of approximately 0.91 m. Disease pressure was applied to all plots via the artificial inoculation method described below.

Triplets of isogenic hybrids showing significant differences in visual appearance of disease symptoms between the B73 × tester control hybrid and either the B73–3B × tester or B73–6A × tester hybrids (or both) at the p < 0.05 level were selected for additional disease and agronomic trait evaluation. A single control tester that produced no significant difference between isohybrids was also selected. Ten testers were thus represented in the agronomic evaluation (Table 1).

Disease and Agronomic Evaluation of Selected Isogenic Triplet-Tester CombinationsDuring the summers of 2010 and 2011, the selected isohybrid triplets were evaluated in two-row plots in a strip-split plot design (Steel et al., 1997). The main plot factor was treatment by artificial inoculation or fungicide control (to prohibit spread of the disease). Main plots were represented by two adjacent fields, separated by four rows of border maize to avoid spread of the disease from the inoculated plot to the fungicide-con-trol plot. Each main plot was divided into four replications of ten sets of isogenic hybrids. The three isogenic hybrids (B73 × tester, B73–3B × tester, B73–6A × tester) were randomized within the replication × tester subplots. To summarize, the main plot factor (inoculation vs. spray) tests the effect of disease development; the subplot factor (pedigree) tests the effect of the inbred tester; the sub-subplot factor (introgression) tests the effect of having the B73 allele vs. containing NC292 alleles at either the 3B or 6A introgressions.

Experiments were planted at NCDA/NCSU research sta-tions in three locations in North Carolina during both 2010 and 2011. In 2010, experiments were planted at the Sandhills Research Station in Jackson Springs (SH), the Peanut Belt Research Station in Lewiston (LE), and CL. In 2011, experiments were planted at LE, CL, and the Cunningham Research Station in Kinston (KI).Thus, a total of six North Carolina environments were repre-sented in this study: SH10, LE10, CL10, LE11, CL11 and KI11. Experiments were planted on 6, 8, and 12 Apr. 2010 for SH,

Isogenic hybrids were developed by crossing B73–3B and B73–6A to several inbred testers (Table 1). B73 was also crossed to the testers to create a relatively susceptible comparison hybrid.

Field TrialsPilot Study on Near-Isogenic Inbred LinesA pilot study on near-isogenic inbred lines was performed at the NCSU Central Crops Research Station in Clayton, NC (CL). The 955 F2:3 families from the B73 × B73-Mo17.3B NIL cross, and 520 F2:3 families from the B73 × B73–6A cross, were planted in two independent experiments. Each experiment was treated as a large block, where different F2:3 family entries were planted according to an incomplete block design. Each single-row plot of eight to twelve plants represented a different F2:3 family that was either homozygous for the B73 allele, homozygous for either the 3B introgression or the 6A introgression (depending on the experiment), or segregating. The experiments from this pilot study were artificially inoculated with C. heterostrophus and scored as described below. Using molecular markers, we identi-fied instances where a family homozygous for the B73 allele and a family homozygous for the introgression were planted in adja-cent or almost adjacent (up to two rows away) plots. In each case, that pair of families was analyzed for yield and disease symptoms. Forty-nine pairs of 3B introgression families were hand-harvested during the summer of 2009 and 50 pairs of 6A introgression fami-lies were hand-harvested during summer 2009 and summer 2010. Yield was estimated by measuring the average weight per ear (g) and 50-Kernel weight (g). Disease symptoms were assessed

Table 1. Preliminary SLB evaluation and selection of differ-ent isogenic triplet-tester combinations. Least squares (LS) means differences of standardized area under the disease progress curve values for isohybrids with and without intro-gressions 3B and 6A, significance level, and standard error are indicated. A significant positive value for LS Means Dif-ference represents higher disease resistance in the isohybrid with the 3B or 6A introgression, versus the allele. The ten ped-igrees selected for the agronomic evaluation are underlined.

Tester line

Introgression

3B 6A

LS means difference SE

LS means difference SE

B97 -0.61 0.32 -0.50 0.32

CML103 0.33 0.39 -0.07 0.32

CML333 1.26** 0.45 -0.35 0.45

CML69 0.80 0.45 -0.61 0.45

H95 1.74*** 0.32 0.41 0.32

H95rhm 1.33*** 0.32 1.75*** 0.32

Ky21 -0.13 0.45 -0.03 0.45

LH85 0.22 0.32 -0.03 0.33

M162W 0.27 0.39 0.17 0.32

Mo17 1.25*** 0.32 0.45 0.32

NC250 2.66*** 0.45 2.12*** 0.39

NC350 1.16** 0.39 0.89** 0.32

Oh43 1.32*** 0.32 0.85** 0.32

Va35 1.24*** 0.32 -0.15 0.32

Va35rhm 1.77*** 0.31 2.34*** 0.32

** Significant at 0.01 probability level.

*** Significant at the 0.001 probability level.

crop science, vol. 54, may–june 2014 www.crops.org 885

CL and LE locations, respectively; and on 14, 15, and 19 Apr. 2011 for LE, CL and KI, respectively. Double-row plots were planted in every case, 4.88 m in length, with spacing of 0.965 m between rows at all locations except LE, where row spacing was 0.914 m. Plots were planted with 44 seeds plot−1 with a popula-tion density of approximately 45,000 plants ha−1 for all locations, except for LE where density was of approximately 50,000 plants ha−1. In 2011, stand counts were taken at the four-leaf stage, and approximately 10% of the plots at LE, CL, and KI exhibited poor germination. To foster inter-row competition, all plots with ger-mination rates lower than 50% were replanted with the hybrid Pioneer 3394 in these environments. In the analysis, data from these plots were considered missing.

Measurements on several agronomic traits were recorded on each sub-subplot. Days to silking (DTS) and days to anthesis (DTA ), estimated as the number of days from planting to when half of the plants in the plot were silking or shedding pollen, respectively, were recorded at CL only. All other traits were measured at all locations. After grain fill, plant and ear heights of four plants from the middle of the plot were recorded to the nearest 5 cm. Plant height was measured to the height of the flag leaf, and ear height was measured at the node from which the ear emerged. The four measurements were averaged for the analysis of the traits. Immediately before harvesting, the num-ber of lodged plants per plot was counted, and percentage of erect plants was calculated. Grain was mechanically harvested and measured at approximately 4 mo after planting. Grain mass and moisture readings were recorded for each plot. Yield mea-surements were converted to Mg per hectare according to the plot size and planting density.

SLB Inoculation and Fungicide ApplicationAll experiments that were artificially inoculated with SLB fol-lowed procedures previously described (Carson, 1998; Carson et al., 2004). Approximately 20 C. heterostrophus Race O-infected sorghum [Sorghum bicolor (L.) Moench] grains were placed in the whorl of plants at the 4 to 6 leaf stage. Sorghum grains were infected with a mixed isolate population. Following inocula-tion, overhead irrigation was briefly applied to provide mois-ture for fungal growth.

For the pilot evaluation of the 3B and 6A introgressions in near-isogenic lines, plots were inoculated on 4 June 2009 and 14 June 2010. For the preliminary disease evaluation of the iso-hybrids, all plots in the Homestead, Florida environment were inoculated on 29 Oct. 2009.

For the disease and agronomic evaluation of isohybrids, one of the main factor plots was inoculated, while the other main plot was sprayed with fungicide in each of the six envi-ronments. Plots were SLB inoculated on 12, 13, and 19 May 2010 for SH, CL and LE, respectively. In 2011, CL experiments were inoculated on 17 May and LE and KI experiments were inoculated on 18 May. The remaining main plots were sprayed with fungicide approximately every 3 wk, starting on the day their corresponding main plots were artificially inoculated. Headline (BASF, a.i. Pyraclostrobin-23.3%) was applied at a rate of 630.5 g ha−1 (a.i. 146.9 g ha−1) for the first two applica-tions, and Folicur (Bayer Crop Science, a.i. Tebuconazole 430 g L−1) at a rate of 420 g ha−1 (a.i. 180.6 g ha−1)was used for the third application in both years.

Disease RatingAll plots were scored at least three different times after flower-ing, on a scale of 1 to 9, where 9 is highly resistant and 1 is dead (see Figure S1). sAUDPC values were calculated as previously documented (Shaner and Finney, 1977). Briefly, the average value of two consecutive ratings was obtained and multiplied by the number of days between the ratings. Values were then summed over all intervals, and divided by the number of days of evaluation to determine the weighted average.

For the preliminary disease evaluation of the isohybrids, SLB was scored on 22 Nov. 2009, 19 Dec. 2009, and 5 Jan. 2010. For the disease and agronomic evaluation, isohybrids were scored for SLB symptoms approximately every 7 d in 2010 and 2011. In 2010, disease assessments were conducted on 2, 9, and 16 July for CL and SH (with a fourth assessment on 23 July for CL); and on 8, 16, and 23 July for LE. In 2011, SLB was assessed on 12, 19, 25 July for CL and 13, 20, 27 July for LE and KI. These ratings were used to calculate sAUDPC on a per-plot basis as described above. In the blocks sprayed with fungicide there was no disease development and so no SLB evaluation was performed.

Statistical AnalysesFor the pilot study on near-isogenic lines, each pair of families (one family homozygous for B73 allele and one family homo-zygous for the introgression) was considered a dependent pair. Therefore, paired t tests were used to analyze average weight per ear (g ear−1) and 50-kernel weight (g 50K−1) using SAS v. 9.1.3 (SAS Institute, 2004).

To evaluate the statistical significance of the differences in disease resistance between the isogenic hybrids in the preliminary evaluation, sAUDPC scores were analyzed using PROC MIXED in SAS v. 9.1.3 (SAS Institute, 2004). The sAUDPC score of an isohybrid plot was modeled as the sum of the fixed effects of its introgression (or lack thereof ), its tester alleles, and the interac-tion between its introgression and tester, as well as random effects due to block and residual. Least squares mean differences between isohybrids within a tester combination were evaluated for statisti-cal significance. Only these pre-planned comparisons were con-ducted; thus, no experimentwise error correction was used.

In the disease and agronomic evaluation of the isohybrids, disease scores were only available for the inoculated plots. The sAUDPC scores were analyzed using PROC MIXED, using a split plot design with pedigree as the main plot and introgression as the subplot. Statistical analysis of the strip-split plot experiment was performed using the MIXED procedures in SAS version 9.2 (SAS Institute, 2004). The main plot (inoculated vs. sprayed), pedigree (or tester), treatment (presence or absence of the 3B or 6A introgression), and the pedigree–introgression interaction effects were treated as fixed. All other effects (i.e., year–location combinations [environments], replications, and their interactions with fixed effect terms) were considered random. Percent stand out of 40 plants was included as a fixed covariate. Statistical mod-els were considered for each trait separately.

GenotypingTo verify that families chosen for the pilot study were homo-zygous for the 3B introgression from Mo17, or for the native B73 allele, leaf tissue from F2:3 families was genotyped at DNA markers within the 3B introgression. The DNA was extracted

886 www.crops.org crop science, vol. 54, may–june 2014

the pedigrees showing a difference for 6A were amongst the nine showing a difference for 3B. These nine isogenic triplet-tester combinations were selected for additional disease and agronomic trait evaluation. Isohybrids devel-oped from B97 all had similar levels of disease resistance, and were included as a negative control (Table 1).

Disease and Agronomic Evaluation of Selected Isogenic Triplet-Tester CombinationsThe selected isohybrid triplets were evaluated in two-row plots according to a strip-split plot design in six year–environment combinations. Parameters of disease suscep-tibility and agronomic traits were measured. Trials took place during the summers of 2010 and 2011 at four loca-tions in North Carolina. These summers were relatively hot and dry in North Carolina, which were not ideal conditions for SLB (e.g., average temperatures in Raleigh in June and July 2010–2011 were 3.5°F higher than in 2012–2013 and average humidity was 5.5% lower accord-ing to the State Climate Office).

Disease EvaluationFor analysis of disease resistance, only inoculated plots were taken into account. Thus in this case, the main plot factor (pedigree) tested the effect of the inbred tester haplotype and the subplot factor (introgression) tested the effect of having B73 alleles at both loci vs. containing NC292 alleles at either the 3B or 6A introgression. Environment was the single biggest factor driving variation in sAUDPC, but the effects associated with replication by environment and rep-lication by pedigree by environment interactions were neg-ligible (Table 2A). Pedigree and introgression were both significant sources of variation in sAUDPC and there was a significant pedigree by introgression interaction (Table 2B).

according to a modified cetyltrimethylammonium bromide (CTAB) protocol (Doyle and Doyle, 1987). Tissue was sent to DuPont Pioneer and DuPont Crop Genetics for genotyp-ing. Single nucleotide polymorphism (SNP) markers PZB7039, PZA3524, PZA13887, PZA12576, PZA17767, PZA4145, and PZA5816 were genotyped at DuPont Pioneer using polymerase chain reaction (PCR) and a probe-based detection method. The SSR marker UMC2000 and indel marker IDP6793 were genotyped at DuPont Crop Genetics via polymerase chain reaction (PCR) and visualization on agarose gel. Homozygos-ity for the introgression 6A or B73 allele in the F2:3 families selected for the pilot study was verified by genotyping total nucleic acid using a SSR marker bnlg2243 located in intro-gression 6A. Total nucleic acid was extracted using a modi-fied CTAB extraction procedure (Saghai-Maroof et al., 1984; Zwonitzer et al., 2009). The SSR genotyping was performed following procedures described elsewhere (Kirigwi et al., 2008; Schuelke, 2000) and visualized on agarose gel.

RESULTSPilot Study on Near-Isogenic Inbred LinesIn this experiment the yield effect of the presence of the 3B and 6A introgressions was evaluated in a B73 inbred line background. For this experiment the 3B introgression derived from Mo17 (designated Mo17.3B), whereas in the rest of the work documented in this paper the 3B intro-gression derived from NC292. As noted above, the SLB resistance QTL in bin 3.04 derived from both NC292 and Mo17 map to the same locus and confer equivalent levels of SLB resistance (Kump et al., 2010; Zwonitzer et al., 2009).

Families carrying the 3B introgression had an average yield advantage of 12% over families lacking the intro-gression. Similarly, families carrying the 6A introgression had an average yield advantage of approximately 8% over families lacking this introgression over 2 yr (Table S1). The differences in SLB disease resistance between fami-lies carrying the 3B and 6A introgression and the families without any introgression was 1.4 and 3 points, respec-tively, on the 1 to 9 scale (data not shown).

Preliminary SLB Evaluation and Selection of Isogenic Triplet-Tester CombinationsSince disease resistance conferred by the 3B and 6A intro-gressions acts recessively with respect to the B73 allele, and possibly to other alleles that may be present in the inbred tester lines, we first needed to identify a set of hybrids in which their phenotypic effects were expressed. With that objective, isogenic hybrids were produced by crossing B73, B73–3B, and B73–6A, to 15 inbred testers (Table 1). The resulting isohybrid sets were evaluated for SLB resistance. Isohybrids carrying the 3B and 6A intro-gressions were significantly more SLB resistant than the corresponding B73-background hybrid control in 9 and 5 out of the 15 pedigrees (NIL-tester combinations), respectively, at the a = 0.05 level (Table 1). All five of

Table 2. Covariance parameters estimates of random effects (A) and type 3 tests of significance of fixed effects (B) for stan-dardized area under the disease progress curve disease rating values related to the effects of introgressions 3B and 6A on iso-hybrid comparisons under southern corn leaf blight pressure.

A.

Covariance parameter Estimate Standard error Ratio†

Environment 0.3894 0.2486 0.609

rep(Environment) 0 – 0

rep*Pedigree(Environment) 0.0928 0.0146 0.145

Residual 0.1573 0.0109 0.246

B.

Effect Num DF‡ Den DF§ F value Pr > F

Pedigree 9 225 61.90 <0.0001

Introgression 2 427 196.84 <0.0001

Pedigree*Introgression 18 426 12.63 <0.0001† Ratio, covariance estimate/total covariance.‡ Num DF, Numerator degrees of freedom.§ Den DF, Denominator degrees of freedom.

crop science, vol. 54, may–june 2014 www.crops.org 887

Results largely agreed with the preliminary SLB eval-uation (see above) but there were some differences (Table 3). Across pedigrees, the mean sAUDPC value for the iso-hybrids without either introgression was 6.8, significantly lower than either mean value for the isohybrids carrying 3B or 6A introgressions (7.6 and 7.4, respectively; Table 3). Averaged across introgression treatments, the pedigrees’ mean sAUDPC values ranged from 6.2 for the B97 back-ground to 8.1 for the NC350 background (Table 3). The significant pedigree by introgression interaction (Table 3) indicated that the presence of either introgression reduced the amount of disease by different degrees across the dif-ferent testers. Just as in the preliminary experiment, the effect of introgression 3B was highly significant at a = 0.01 for all nine pedigrees tested, except control B97 (Table S2, Fig. 1). The effect of 6A on SLB resistance was significant in 6 crosses out of 10 at a = 0.05 (4 out of 10 at a = 0.01 Table S2, Fig. 1). Surprisingly, no significant resistance effects were detected for introgression 6A in crosses with Oh43 and NC350, whereas there had been a significant effect in the preliminary experiment. In con-trast, significant effects associated with the 6A introgres-sion were detected in crosses with Va35, CML333, and H95, where no such effects had been identified previously. The effect of 6A on SLB resistance was highly signifi-cant at the a = 0.01 level in crosses with H95, H95rhm, NC250, and Va35rhm (Table S2).

Agronomic EvaluationWe next determined the effect of the 3B and 6A introgres-sions on yield and other agronomic traits in hybrids, both in the presence and absence of SLB disease. Agronomic traits evaluated included yield, moisture content, plant height, ear height, lodging (percent erect plants at harvest), and flowering time (DTS and DTA). For each agronomic trait considered, pedigree main effect (effect of a different tester haplotype complementing the NIL) was significant (Tables 4 and S3). Significant effects were not observed for the main plot (inoculated vs. sprayed) for any of the agronomic traits at the 0.05 significance level. For yield, significant effects (p < 0.05) were found for the interaction between the main plot factor (inoculated vs. sprayed) and introgression factors (presence or absence of the 3B or 6A introgression; Table 4). For moisture content, main intro-gression effect and the interaction effect between pedigree and introgression were significant at a 0.10 significance level (Table S3). There was a significant (p < 0.01) main plot treatment by pedigree by introgression effect on the percentage of erect plants (Table S3). Covariance param-eters for the random effects are also presented in Table S3.

Across all pedigrees and introgression treatments, the mean yield of the fungicide-sprayed main plot was 0.28 Mg ha-1 (approximately 5%) greater than the mean yield of the SLB-infected plot, but the difference was not

Table 3. Least-square means for standardized area under the disease progress curve (sAUDPC) related to the effect of pedigree and introgressions. Estimates of southern corn leaf blight-resistance levels observed in the different pedigrees over all introgressions, in the different introgression lines over all pedigrees and in all the pedigree by introgression com-binations. Mean estimates are presented on a 1–9 scale of sAUDPC, in which 9 is a highly resistant plant and 1 is dead.

EffectTester

pedigreeIntro-

gression

Estimate(sAUDPC

units)Standard

error

Pedigrees across all introgressionsPedigree B97 6.1834 0.2664

Pedigree CML333 7.1505 0.2666

Pedigree H95 7.5440 0.2679

Pedigree H95rhm 8.0277 0.2670

Pedigree Mo17 6.7466 0.2664

Pedigree NC250 6.7838 0.2676

Pedigree NC350 8.1424 0.2669

Pedigree Oh43 7.2671 0.2670

Pedigree Va35 6.9991 0.2664

Pedigree Va35rhm 7.8045 0.2664

Introgressions across all pedigrees

Introgression – 6.8452 0.2569

Introgression 3B 7.5927 0.2571

Introgression 6A 7.3569 0.2568

Individual pedigree * introgression combinations

Pedigree*Introgression B97 – 6.0938 0.2744

Pedigree*Introgression B97 3B 6.2229 0.2744

Pedigree*Introgression B97 6A 6.2337 0.2744

Pedigree*Introgression CML333 – 6.8287 0.2751

Pedigree*Introgression CML333 3B 7.5864 0.2758

Pedigree*Introgression CML333 6A 7.0365 0.2744

Pedigree*Introgression H95 – 7.0205 0.2744

Pedigree*Introgression H95 3B 7.9176 0.2837

Pedigree*Introgression H95 6A 7.6939 0.2868

Pedigree*Introgression H95rhm – 7.4418 0.2784

Pedigree*Introgression H95rhm 3B 8.1330 0.2758

Pedigree*Introgression H95rhm 6A 8.5083 0.2744

Pedigree*Introgression Mo17 – 6.3830 0.2744

Pedigree*Introgression Mo17 3B 7.3613 0.2751

Pedigree*Introgression Mo17 6A 6.4955 0.2744

Pedigree*Introgression NC250 – 6.1125 0.2786

Pedigree*Introgression NC250 3B 7.1440 0.2786

Pedigree*Introgression NC250 6A 7.0948 0.2753

Pedigree*Introgression NC350 – 7.9295 0.2744

Pedigree*Introgression NC350 3B 8.4703 0.2795

Pedigree*Introgression NC350 6A 8.0274 0.2744

Pedigree*Introgression Oh43 – 7.0141 0.2758

Pedigree*Introgression Oh43 3B 7.6314 0.2776

Pedigree*Introgression Oh43 6A 7.1559 0.2752

Pedigree*Introgression Va35 – 6.6972 0.2744

Pedigree*Introgression Va35 3B 7.4108 0.2744

Pedigree*Introgression Va35 6A 6.8892 0.2744

Pedigree*Introgression Va35rhm – 6.9309 0.2744

Pedigree*Introgression Va35rhm 3B 8.0490 0.2751

Pedigree*Introgression Va35rhm 6A 8.4337 0.2744

888 www.crops.org crop science, vol. 54, may–june 2014

significant. Averaged across all treatments, the hybrid means for the different pedigrees ranged from 5.4 Mg ha-1 for H95rhm isohybrids to 6.54 Mg ha-1 for Mo17 isohy-brids. Least-Square means for yield and other agronomic traits are presented in Table S4.

In the absence of SLB and averaged across pedigrees, mean yield for the hybrids carrying introgression 6A was approximately 0.17 Mg ha-1 (about 3%) less than the con-trol B73-background hybrids without any introgression. This difference was significant at a 0.05% level. The 3B introgression was associated with a 0.13 Mg ha-1 reduc-tion in yield in the absence of SLB, but this difference was not significant (Fig. 2, Table 5).

In the presence of SLB and averaged across pedigrees, mean yield for the hybrids carrying introgression 3B was approximately 0.19 Mg ha-1 (about 3%) more than the control B73-background hybrids without any introgres-sion, which was significant at a 0.05% level. In contrast, introgression 6A did not show any significant overall effect on yield compared to the control (Fig. 2, Table 5) under disease pressure. However, this lack of overall yield effect is misleading as the 6A introgression, unlike the 3B introgression, did not confer significant levels of resistance in all the different selected crosses (see Table S2, Fig. 1). Introgression 6A conferred the highest amount of resistance in crosses with NC250, Va35rhm, and H95rhm (Fig. 3). In these crosses introgression 6A was associ-ated with yield increases of 0.53, 0.57, and 0.34 Mg ha-1 (about 10%, 10%, and 6%), respectively. For NC250 and

Va35rhm, these increases in yield were statistically signifi-cant at a 0.05 significance level (Table 5, Fig. 3).

For introgression 3B, while all the hybrid combina-tions with B73–3B were significantly more resistant to SLB compared to the B73 hybrid control (Fig. 1), and even though under inoculation the overall yield effect of 3B was significant, within specific pedigrees, only hybrid Va35rhm x B73–3B showed significantly more yield than the corresponding hybrid control, Va35rhm x B73 under inoculation (Table 5, Fig. 3).

DISCUSSIONSLB resistance is mostly quantitative in nature and most corn grown in the central to southern United States has some level of resistance. Several studies have reported quan-titative resistance against SLB in U.S. germplasm (Balint-Kurti et al., 2006; Burnette and White, 1985; Holley and Goodman, 1989; Hooker et al., 1970; Kump et al., 2011; Lim, 1975; Lim and Hooker, 1976; Pate and Harvey, 1954; Thompson and Bergquist, 1984). Although QDR against SLB has been used in corn breeding for a long time, little is known about how much individual QTLs may contribute to differences in agronomic traits, such as yield, in inbred lines and hybrids, and in the presence and absence of dis-ease pressure. To investigate effects of two single QTLs in both inbred and hybrid backgrounds, we chose QTLs with large additive effects in a pathosystem for which host resis-tance is known to be highly heritable. The lines B73–3B and B73–6A each contain one introgression from highly

Figure 1. Effect of different pedigree haplotypes on southern leaf blight resistance in isogenic hybrid comparisons with and without introgressions 3B and 6A. Effects are presented as the standardized area under the disease progress curve (sAUDPC) mean difference between isohybrids carrying introgressions 3B or 6A and isohybrids with no introgression. A scale of 1–9 is employed on which 9 is a highly-resistant plant and 1 is dead. * Significant at p < 0.10. ** Significant at p < 0.01.

crop science, vol. 54, may–june 2014 www.crops.org 889

SLB-resistant line NC292 in a B73 background. Each of these introgressions carries an SLB dQTL with a relatively large effect. To study the effects of introgressions 3B and 6A on agronomic traits, we used isogenic hybrids, differ-ing only in the presence of one of our introgressions.

We discerned a yield effect associated with each introgression under disease pressure during a pilot study comparing F2:3 families homozygous for either introgres-sion to F2:3 families without either introgression (essen-tially inbred isolines). In that study, we found a significant yield advantage when isolines carrying either introgres-sion were infected with SLB compared to isolines with-out the introgression. This result suggested that the SLB resistance conferred by these introgressions may cause an increase in grain weight in inbred lines under disease pres-sure. It should be noted that this preliminary study used the Mo17 allele at 3B as the source of resistance while the other experiments reported here used the NC292 allele.

For our purposes we are assuming that the genes underly-ing the resistance at 3B are identical in Mo17 and NC292, based on that fact that they precisely co-localize, that they are both recessive to the B73 allele, and that they have similar effect estimates (Kump et al. ,2010 and unpub-lished data; Zwonitzer et al., 2009).

A more relevant measure of the economic impact of these introgressions is their effect on yield in hybrids. However, the resistances conferred by introgressions 3B and 6A are recessive relative to their respective B73 alleles. So we first needed to identify a set of hybrids in which their phenotypic effects were expressed. Therefore, we developed and challenged with SLB 15 sets of isohybrids triplets (carrying one copy of introgression 3B, 6A, or no introgression) from a diverse set of pedigrees, to determine whether we could detect resistance effects of 3B and 6A across different backgrounds. In our preliminary experi-ments we observed significant resistance effects in 5 out of 15 backgrounds for 6A and 9 out of 15 backgrounds for 3B (Table 1). All five backgrounds in which resistance associ-ated with 6A was expressed were among the nine in which 3B resistance was expressed. In the main experiment assess-ing agronomic effects, resistance effects associated with 3B were again detected in the same nine backgrounds (Fig. 1, Table 3) but there were some differences in the backgrounds

Table 4. Covariance Parameters Estimates for the random effects (A) and Type 3 tests of significance of fixed effects (B) for yield for the effects of introgressions 6A and 3B on isohybrid comparisons with and without disease pressure. Yield measurements were standardized based on a stand of 40 plants at 14.5% grain moisture and then converted to Mg ha-1 according to the plot size and planting density. Stand counts (ST) was used as a covariate for yield.

A.

Covariance parameter estimates

Ratio‡Covariance parameter Estimate

Standard error Pr > Z†

Environment 3.5233 2.3201 0.0644 0.670Environment*Main 0.1712 0.1623 0.1458 0.033

rep(Environment*Main) 0.2629 0.07915 0.0004 0.050

Environment*Pedigree 0.1480 0.05214 0.0023 0.028

r ep*Pedigree (Environment*Main)

0.4644 0.05491 <0.0001 0.088

E nvironment*Pedigree* Introgression

0.0144 0.01643 0.1910 0.003

Residual 0.6778 0.03604 <0.0001 0.129

B.

Type 3 tests of significance of fixed effects on Yield

EffectNum DF§

Den DF¶ F value Pr > F

ST 1 551 218.68 <0.0001

Main (inoculated vs. sprayed) 1 5.01 0.93 0.3802

Pedigree 9 45.1 3.28 0.0038

Main*Pedigree 9 364 1.07 0.3871

Introgression 2 92.7 1.07 0.3485

Main*Introgression 2 756 4.19 0.0155

Pedigree* Introgression 18 91.6 0.76 0.7365

Main*Pedigree* Introgression 18 753 0.84 0.6492† Wald test for Ho: Covariance parameter = 0.‡ Ratio = Covariance estimate/Total covariance.§ Num DF, Numerator degrees of freedom.¶ Den DF, Denominator degrees of freedom.

Figure 2. Least squares mean yield estimates of isogenic hybrids with and without introgression 3B or 6A across pedigrees under presence and absence of southern corn leaf blight (SLB). Effect of carrying introgression 3B or 6A on yield in t ha-1 across all pedigree treatments under adequate SLB disease pressure (inoculated) and no SLB (sprayed) treatments. Mean estimates for yield were standardized for stand and moisture and then converted to Mg ha-1 according to the plot size and planting density. Significance is based on comparisons against the control isohybrids (No introgression). ** Significant at 5%. NS Nonsignificant.

890 www.crops.org crop science, vol. 54, may–june 2014

in which resistance associated with 6A was expressed. Sig-nificant resistance effects were detected for 6A in crosses with Oh43 and NC350 in the preliminary but not in the main experiment, whereas in crosses with Va35, CML333, and H95, 6A effects were detected in the main but not the preliminary experiment. The backgrounds in which 6A was associated with the largest effects on SLB resistance,

namely H95rhm, NC250, and Va35rhm, were consistent across the two experiments. It is not clear why we observed these differences, but environment was likely a contribut-ing factor. These two experiments were performed in quite different environments, with the preliminary experiment in Homestead, FL, during the winter of 2009 to 2010 and the main experiment in the summers of 2010 and 2011 in

Table 5. Least-squares mean contrasts for yield related to the effects of the presence or absence of southern corn leaf blight on isogenic hybrids comparisons with and without introgressions 3B or 6A. Contrast estimate is the difference between iso-hybrids without any introgression (B73) and isohybrids carrying either 3B or 6A introgression. Contrast estimate is presented in Mg ha-1.

Effect Main TesterIntrogression

differenceContrast estimate DF t value Pr > |t|

Main*Introgression Inoculated B73 B73–3B -0.19 292 -2.25 0.0249

Main*Introgression Inoculated B73 B73–6A -0.06 277 -0.68 0.4985

Main*Introgression sprayed B73 B73–3B 0.13 293 1.57 0.1186

Main*Introgression sprayed B73 B73–6A 0.17 280 2.08 0.0381

Main*Pedigree*Introgression Inoculated B97 B73 B73–3B -0.13 268 -0.51 0.6081

Main*Pedigree*Introgression Inoculated B97 B73 B73–6A 0.14 258 0.55 0.5814

Main*Pedigree*Introgression Inoculated CML333 B73 B73–3B -0.15 287 -0.58 0.5639

Main*Pedigree*Introgression Inoculated CML333 B73 B73–6A 0.34 268 1.34 0.1802

Main*Pedigree*Introgression Inoculated H95 B73 B73–3B -0.01 343 -0.02 0.9858

Main*Pedigree*Introgression Inoculated H95 B73 B73–6A 0.53 299 2.01 0.0456

Main*Pedigree*Introgression Inoculated H95rhm B73 B73–3B -0.10 318 -0.37 0.7153

Main*Pedigree*Introgression Inoculated H95rhm B73 B73–6A -0.34 317 -1.26 0.2094

Main*Pedigree*Introgression Inoculated Mo17 B73 B73–3B -0.34 269 -1.35 0.1786

Main*Pedigree*Introgression Inoculated Mo17 B73 B73–6A 0.02 260 0.09 0.9288

Main*Pedigree*Introgression Inoculated NC250 B73 B73–3B -0.32 271 -1.13 0.258

Main*Pedigree*Introgression Inoculated NC250 B73 B73–6A -0.53 261 -1.98 0.0483

Main*Pedigree*Introgression Inoculated NC350 B73 B73–3B -0.20 310 -0.72 0.4727

Main*Pedigree*Introgression Inoculated NC350 B73 B73–6A -0.22 258 -0.88 0.3787

ain*Pedigree*Introgression Inoculated Oh43 B73 B73–3B -0.02 297 -0.08 0.9337

Main*Pedigree*Introgression Inoculated Oh43 B73 B73–6A 0.07 303 0.27 0.7881

ain*Pedigree*Introgression Inoculated Va35 B73 B73–3B -0.07 277 -0.27 0.7864

Main*Pedigree*Introgression Inoculated Va35 B73 B73–6A 0.01 278 0.02 0.984

Main*Pedigree*Introgression Inoculated Va35rhm B73 B73–3B -0.57 258 -2.25 0.0254

Main*Pedigree*Introgression Inoculated Va35rhm B73 B73–6A -0.57 291 -2.31 0.0216

Main*Pedigree*Introgression sprayed B97 B73 B73–3B 0.03 278 0.11 0.9137

Main*Pedigree*Introgression sprayed B97 B73 B73–6A -0.02 288 -0.1 0.9232

Main*Pedigree*Introgression sprayed CML333 B73 B73–3B 0.45 268 1.73 0.0851

Main*Pedigree*Introgression sprayed CML333 B73 B73–6A 0.22 294 0.87 0.3826

Main*Pedigree*Introgression sprayed H95 B73 B73–3B 0.14 275 0.51 0.6138

Main*Pedigree*Introgression sprayed H95 B73 B73–6A 0.37 301 1.46 0.1467

Main*Pedigree*Introgression sprayed H95rhm B73 B73–3B 0.31 297 1.12 0.2626

Main*Pedigree*Introgression sprayed H95rhm B73 B73–6A 0.23 295 0.84 0.4024

Main*Pedigree*Introgression sprayed Mo17 B73 B73–3B 0.11 280 0.41 0.6816

Main*Pedigree*Introgression sprayed Mo17 B73 B73–6A 0.16 293 0.63 0.5277

Main*Pedigree*Introgression sprayed NC250 B73 B73–3B 0.25 277 0.9 0.367

Main*Pedigree*Introgression sprayed NC250 B73 B73–6A 0.44 272 1.64 0.1028

Main*Pedigree*Introgression sprayed NC350 B73 B73–3B -0.21 259 -0.82 0.411

Main*Pedigree*Introgression sprayed NC350 B73 B73–6A -0.09 348 -0.38 0.7048

Main*Pedigree*Introgression sprayed Oh43 B73 B73–3B 0.05 343 0.17 0.8685

Main*Pedigree*Introgression sprayed Oh43 B73 B73–6A 0.17 280 0.59 0.5544

Main*Pedigree*Introgression sprayed Va35 B73 B73–3B 0.03 280 0.13 0.8956

Main*Pedigree*Introgression sprayed Va35 B73 B73–6A 0.21 276 0.8 0.4227

Main*Pedigree*Introgression sprayed Va35rhm B73 B73–3B 0.17 258 0.66 0.5093

Main*Pedigree*Introgression sprayed Va35rhm B73 B73–6A 0.03 280 0.14 0.8895

crop science, vol. 54, may–june 2014 www.crops.org 891

several different locations in North Carolina. Disease pres-sure in Homestead was substantially higher than in North Carolina, and the corresponding effects of both introgres-sions were on average substantially higher for the prelimi-nary compared to the main experiment (compare the least squares mean differences values in Table 1 and the estimate of contrast values in Table S2). In addition, the preliminary experiment had three replications for each set of germplasm whereas the main experiment was much more highly repli-cated (24 replications per entry).

The NC250 tester carries both of the alleles (3b, 6a) tested in this study (Zwonitzer et al., 2009). Hybrids derived from a cross of NC250 with either B73–3B or B73–6A should therefore be homozygous for their respec-tive introgressions and, as expected, were significantly more resistant than the NC250 × B73 hybrid. It is also worth noting that the hybrids derived from crosses of B73–6A to testers carrying the SLB resistance gene rhm (H95rhm and Va35rhm) showed high levels of resis-tance. Not only were Va35rhm × B73–6A and H95rhm × B73–6A significantly more resistant than their respec-tive isohybrids without introgression 6A, but the hybrids Va35rhm × B73–6A and H95rhm × B73–6A were more resistant than Va35 × B73–6A and H95 × B73–6A respec-tively (1.6 and 0.8 units on the 1–9 scale). Both rhm and introgression 6A map to the same distal region of the short arm of chromosome 6. Since introgression 6A is recessive to the corresponding allele in B73, this suggests that the tester allele is somehow complementing or allowing the

expression of 6A. In this sense, rhm and introgression 6A could represent alternate versions of the same resistance gene or could indeed be the same gene (Belcher, 2009).

Heterosis, a phenomenon common in outcrossing spe-cies, causes hybrid plants to be taller, higher-yielding, and more vigorous than inbred plants. There are modest but highly significant correlations between the sAUDPC values under disease pressure and yield under both SLB pressure (~37%) and no disease (~27%) indicating that higher resis-tance was correlated with higher yield (Table S5). The cor-relation under disease-free conditions suggests that differen-tial levels of heterosis in the different hybrids are influencing both yield and disease resistance. Physiological differences due to heterosis, such as tougher cell wall structures, greater accumulation of disease-fighting metabolites, etc., could have accounted for resistance against SLB in hybrids.

Overall in this study a ~5% yield difference was observed between the SLB-infected and fungicide-sprayed blocks. This difference was not statistically significant (Table 4B). The relatively small effect of SLB infection on yield was somewhat surprising to us since previous litera-ture had suggested that SLB was responsible for substan-tially higher yield losses. However, the 38% and 46% yield losses cited for the Byrnes and Pataky (1989) and Fisher et al. (1976) studies respectively were maximum values for specific pedigrees and/or locations. Most of the yield losses reported in those studies were in the 10 to 20% ranges. In this study, looking at individual pedigrees, the maximum yield loss associated with SLB infection was 14.7% for the

Figure 3. Effect of different pedigree haplotypes on yield for isogenic hybrid comparisons with and without introgressions 3B or 6A under southern corn leaf blight presence (inoculated) and absence (sprayed). Yield mean differences are presented in t ha-1. Effects are presented as the yield mean difference between isohybrids carrying introgressions 3B or 6A and isohybrids with no introgression (background hybrid control). * Significant at p < 0.10. ** Significant at p < 0.05. All others are nonsignificant.

892 www.crops.org crop science, vol. 54, may–june 2014

Mo17 × B73 hybrid. Furthermore, determining yield loss associated with SLB infection was not the primary goal of this study and it was not designed to capture these differ-ences optimally. For instance, a more susceptible hybrid or set of hybrids might well have shown larger yield effects associated with infection. Hot dry summers during 2010 and 2011 could have suppressed the disease, reducing the overall difference in yield between SLB-infected and dis-ease-free conditions. With respect to the lack of signifi-cance, the primary disadvantage in a strip-split plot design is the loss in power to detect whole plot factor effects because of the small number of replications for the main plot treatment (inoculated vs. sprayed) and the consequent low degrees of freedom associated with the error term used to test the whole plot factor (Steel et al., 1997).

Our analysis was able to detect an interaction between introgression and main plot treatment. In the presence of the SLB disease pressure, mean yield for the hybrids carry-ing introgression 3B across all pedigrees was approximately 3% more than the control B73-background hybrids, a sig-nificant difference (Fig. 2, Tables 5, S4) implying that the resistance conferred by introgression 3B may be respon-sible for this reduction in yield loss. Looking at specific pedigrees, even though all the hybrid combinations with B73–3B were significantly more resistant to SLB compared to the B73 hybrid control, only hybrid Va35rhm × B73–3B showed significantly more yield than the corresponding hybrid control, Va35rhm × B73 (Fig. 3).

No significant advantage in yield was detected with introgression 6A across pedigrees in the presence of SLB (Fig. 2, Tables 5, S4). However, only 4 out of the 10 pedi-grees tested in this study showed increases in resistance asso-ciated with introgression 6A that were significant at a = 0.01 (as compared to nine for 3B, see Fig. 1). So, we would not necessarily expect yield benefits from having the introgres-sion 6A in the presence of SLB in the other six testers. Of the four pedigrees that did show a highly significant resistance advantage associated with 6A, two of them, crosses with NC250 and Va35rhm, showed a significant yield gain asso-ciated with the presence of 6A under infected conditions. This result suggests that the resistance conferred by intro-gression 6A can indeed protect yield during SLB infection.

Yield penalties associated with the presence of genetic disease resistance have been documented in several studies, especially in the absence of pressure from the correspond-ing disease (Korves and Bergelson, 2004; Orgil et al., 2007; Tian et al., 2003). The prevalence of susceptibility alleles of certain genes in natural populations has suggested that the corresponding resistance alleles may confer a fitness cost (Bergelson and Purrington, 1996; Parker, 1992). While most studies of this type have addressed major gene effects on fitness (Dietrich et al., 2005; Heidel et al., 2004; Korves and Bergelson, 2004; Orgil et al., 2007; Tian et al., 2003), two studies have worked with QDR. In one of these cases

a yield cost associated with QDR was observed (Mitchell-Olds and Bradley, 1996) and in the other case no yield cost was observed (Frey et al., 2011). In the absence of SLB, mean yield for the hybrids carrying introgression 6A was signifi-cantly (3%) less than the control B73-background hybrids without any introgression suggesting a cost of resistance, while the 3B introgression was associated with a smaller (~2%) yield reduction which was not statistically significant.

One explanation for the yield penalty associated with the introgressions is linkage drag, the introduction of deleterious alleles distinct from but linked to the disease resistance genes on 3B and 6A. We have genotyped both B73–3B and B73–6A with a set of > 50,000 SNP markers (data not shown) and consequently have a good measure of the size of the two introgressions. The 6A introgres-sion is almost 11 Mb (out of a total genome size of ~2500 Mb) while the 3B introgression is ~5.6 Mb. Both these introgressions likely carry a large number of genes, some of which may be deleterious to yield.

The other explanation for the yield penalty is that a physiological cost is conferred by the actual disease resis-tance gene. A constitutively active defense response or high basal defense level has been associated with yield penalties in several different studies (Korves and Bergel-son, 2004; Orgil et al., 2007; Tian et al., 2003). To address this possibility, it is instructive to inspect the individual pedigrees with respect to resistance and yield. If expres-sion of resistance itself were the reason for the reduction in yield in the absence of disease then one would expect that hybrids in which 6A was associated with high lev-els of SLB resistance (corresponding presumably to robust phenotypic expression of the resistance gene) might have correspondingly high yield penalties associated with the presence of 6A in the absence of disease pressure (compare Fig. 1 and 3). While there is not a completely consistent pattern, the four pedigrees in which 6A has no signifi-cant effect on resistance: B97, Mo17, NC350 and Oh43 (see Fig. 1) have relatively low (or no) yield penalties asso-ciated with 6A in the absence of disease pressure (0.02, -0.16, 0.09, and -0.17 Mg ha-1 respectively), while for the 4 pedigrees in which 6A has a highly significant effect on resistance (H95, H95rhm, NC250, and Va35rhm), the yield penalties associated with 6A are somewhat larger (-0.37, -0.23, -0.44, and -0.03 t ha-1, respectively). The Spearman correlation between resistance associated with 6A and yield penalty associated with 6A under no disease pressure amongst the 10 hybrid pedigrees is -0.60, which is significant at a = 0.1. The corresponding figure for introgression 3B is -0.59, again significant at α = 0.1. These results suggest that the modest yield penalties asso-ciated with 6A and possibly 3B in disease-free conditions are associated with expression of the disease resistance mechanisms themselves rather than with linkage drag.

crop science, vol. 54, may–june 2014 www.crops.org 893

In summary, our results suggest that both the 3B and 6A introgressions can protect yield during southern leaf blight epidemics in specific genetic backgrounds, though they may confer modest yield penalties in the absence of the disease. If the yield penalties are due in whole or in part to linkage drag rather than expression of the resis-tance itself, then smaller introgressions at these loci could be used that still confer the SLB resistance phenotype but have smaller associated yield penalties. However, we have some evidence suggesting that the yield penalties may be associated with the expression of resistance and so may not be separable from it.

AcknowledgmentsThe authors would like to thank the following people who helped with various aspects of the research: William Hill, Wayne Dillard, Dale Dowden, Abbey Sutton, David Rhyne, and Shannon Sermons. We thank Cathy Herring and the staff of Central Crops Research Station, the NCSU Sandhills Research Station, the Peanut Belt Research Station, the Cunningham Research Station, and the staff of 27 Farms in Homestead, FL. for their expert help. We acknowledge genotyping help from Pioneer Hi-Bred, in particular Petra Wolters and Mark Jung. This work was funded by the USDA-ARS and NCSU. J. Santa-Cruz’s fellowship was funded by Monsanto, Inc.

ReferencesBalint-Kurti, P.J., M.D. Krakowsky, M.P. Jines, L.A. Robertson,

T.L. Molnár, M.M. Goodman, and J.B. Holland. 2006. Iden-tification of quantitative trait loci for resistance to southern leaf blight and days to anthesis in a maize recombinant inbred line population. Phytopathology 96:1067–1071. doi:10.1094/PHYTO-96-1067

Balint-Kurti, P.J., J.C. Zwonitzer, R.J. Wisser, M.L. Carson, M.A. Oropeza-Rosas, J.B. Holland, and S.J. Szalma. 2007. Precise mapping of quantitative trait loci for resistance to southern leaf blight, caused by Cochliobolus heterostrophus race O, and flowering time using advanced intercross maize lines. Genet-ics 176:645–657. doi:10.1534/genetics.106.067892

Belcher, A.R. 2009. The physiology and host genetics of quantitative resistance in maize to the fungal pathogen cochliobolus heter-ostrophus. M.S. thesis, North Carolina State Univ., Raleigh.

Belcher, A.R., J. Zwonitzer, J.H. Santa-Cruz, M. Krakowsky, C.-L. Chung, R. Nelson, C. Arellano, and P.J. Balint-Kurti. 2011. Analysis of quantitative disease resistance to south-ern leaf blight and of multiple disease resistance in maize, using near-isogenic lines. Theor. Appl. Genet. 124:433–445. doi:10.1007/s00122-011-1718-1

Bergelson, J., and C.B. Purrington. 1996. Surveying patterns in the cost of resistance in plants. Am. Nat. 148:536–558. doi:10.1086/285938

Bubeck, D.M. 1991. Molecular and biometric evaluation of gray leaf spot and southern corn leaf blight resistance in maize. Ph.D. diss., North Carolina State Univ., Raleigh.

Burnette, D.C., and D.G. White. 1985. Inheritance of resistance to Bipolaris maydis race O in crosses derived from nine resis-tant inbred lines of maize. Phytopathology 75:1195–1200. doi:10.1094/Phyto-75-1195

Byrnes, K.J., and J.K. Pataky. 1989. Relationships between yield of three maize hybrids and severity of southern leaf blight caused by race O of Bipolaris maydis. Plant Dis. 73:834–840. doi:10.1094/PD-73-0834

Carson, M.L. 1998. Aggressiveness and perennation of isolates of Cochliobolus heterostrophus from North Carolina. Plant Dis. 82:1043–1047. doi:10.1094/PDIS.1998.82.9.1043

Carson, M.L., C.W. Stuber, and M.L. Senior. 2004. Identification and mapping of quantitative trait loci conditioning resistance to southern leaf blight of maize caused by Cochliobolus het-erostrophus race O. Phytopathology 94:862–867. doi:10.1094/PHYTO.2004.94.8.862

Davis, G.L., M.D. McMullen, C. Baysdorfer, T. Musket, D. Grant, M. Staebell, G. Xu, M. Polacco, L. Koster, S. Melia-Hancock, K. Houchins, S. Chao, and E.-H. Coe, Jr. 1999. A maize map standard with sequenced core markers, grass genome refer-ence points and 932 expressed sequence tagged sites (ESTs) in a 1736-locus map. Genetics 152:1137–1172.

Dietrich, R., K. Ploss, and M. Heil. 2005. Growth responses and fitness costs after induction of pathogen resistance depend on environmental conditions. Plant Cell Environ. 28:211–222. doi:10.1111/j.1365-3040.2004.01265.x

Doyle, J.J., and J.L. Doyle. 1987. A rapid DNA isolation procedure for small quantities of fresh leaf tissue. Phytochem. Bull. 19:11–15.

Fisher, D.E., A.L. Hooker, S.M. Lim, and D.R. Smith. 1976. Leaf infection and yield loss caused by 4 Helminthosporium leaf diseases of corn. Phytopathology 66:942–944. doi:10.1094/Phyto-66-942

Frey, T.J., T. Weldekidan, T. Colbert, P.J.C.C. Wolters, and J.A. Hawk. 2011. Fitness evaluation of Rcg1, a locus that con-fers resistance to Colletotrichum graminicola (Ces.) G.W. Wils. using near-isogenic maize hybrids. Crop Sci. 51:1551–1563. doi:10.2135/cropsci2010.10.0613

Heidel, A.J., J.D. Clarke, J. Antonovics, and X. Dong. 2004. Fit-ness costs of mutations affecting the systemic acquired resis-tance pathway in Arabidopsis thaliana. Genetics 168:2197–2206. doi:10.1534/genetics.104.032193

Holley, R.N., and M.M. Goodman. 1989. New sources of resis-tance to southern corn leaf blight from tropical hybrid maize derivatives. Plant Dis. 73:562–564. doi:10.1094/PD-73-0562

Hooker, A.L., D.R. Smith, S.M. Lim, and J.B. Beckett. 1970. Reaction of corn seedlings with male-sterile cytoplasm to Helminthosporium maydis. Plant Dis. Rep. 54:708–712.

Kirigwi, F., J. Zwonitzer, M. Mian, Z.Y. Wang, and M. Saha. 2008. Microsatellite markers and genetic diversity assessment in Lolium temulentum. Genet. Resour. Crop Evol. 55:105–114. doi:10.1007/s10722-007-9218-5

Korves, T., and J. Bergelson. 2004. A novel cost of R gene resis-tance in the presence of disease. Am. Nat. 163:489–504. doi:10.1086/382552

Kump, K.L., P.J. Bradbury, R.J. Wisser, E.S. Buckler, A.R. Belcher, M.A. Oropeza-Rosas, J.C. Zwonitzer, S. Kreso-vich, M.D. McMullen, D. Ware, P.J. Balint-Kurti, and J.B. Holland. 2011. Genome-wide association study of quanti-tative resistance to southern leaf blight in the maize nested association mapping population. Nat. Genet. 43:163–168. doi:10.1038/ng.747

Kump, K.L., J.B. Holland, M.T. Jung, P. Wolters, and P.J. Balint-Kurti. 2010. Joint analysis of near-isogenic and recombinant inbred line populations yields precise positional estimates for quantitative trait loci. Plant Gen. 3:142–153. doi:10.3835/plantgenome2010.05.0011

894 www.crops.org crop science, vol. 54, may–june 2014

Lim, S.M. 1975. Heterotic effects of resistance in maize to Hel-minthosporium maydis race O. Phytopathology 65:1117–1120. doi:10.1094/Phyto-65-1117

Lim, S.M., and A.L. Hooker. 1976. Estimates of combining ability for resistance to Helminthosporium maydis Race O in a maize population. Maydica 21:121–128.

Mitchell-Olds, T., and D. Bradley. 1996. Genetics of Brassica rapa. 3. Costs of disease resistance to three fungal pathogens. Evo-lution 50:1859–1865. doi:10.2307/2410743

Orgil, U., H. Araki, S. Tangchaiburana, R. Berkey, and S. Xiao. 2007. Intraspecific genetic variations, fitness cost and benefit of RPW8, a disease resistance locus in Arabidopsis thaliana. Genetics 176:2317–2333. doi:10.1534/genetics.107.070565

Parker, M.A. 1992. Constraints on the evolution of resistance to pests and pathogens. In: P.G. Ayres, editor, Pest and pathogens:plants responses to foliar attack. Bios Scientific, Oxford. p. 181–197.

Pate, J.B., and P.H. Harvey. 1954. Studies on the inheritance of resistance in corn to Helminthosporium maydis leaf spot. Agron. J. 46:442–445. doi:10.2134/agronj1954.00021962004600100002x

Poland, J.A., P.J. Bradbury, E.S. Buckler, and R.J. Nelson. 2011. Genome-wide nested association mapping of quantitative resistance to northern leaf blight in maize. Proc. Natl. Acad. Sci. USA 108:6893–6898. doi:10.1073/pnas.1010894108

Russell, W.A. 1972. Registration of B70 and B73 parental lines of maize (Reg. Nos. PL16 and PL17). Crop Sci. 12:721. doi:10.2135/cropsci1972.0011183X001200050085x

Saghai-Maroof, M.A., K.M. Soliman, R.A. Jorgensen, and R.W. Allard. 1984. Ribosomal DNA spacer-length polymorphisms in barley: Mendelian inheritance, chromosomal location, and population dynamics. Proc. Natl. Acad. Sci. USA 81:8014–8018. doi:10.1073/pnas.81.24.8014

SAS Institute. 2004. SAS/STAT® 9.2 User’s guide. 2nd ed. SAS Inst., Cary, NC

Schuelke, M. 2000. An economic method for the fluorescent labeling of PCR fragments. Nat. Biotechnol. 18:233–234. doi:10.1038/72708

Shaner, G., and P.E. Finney. 1977. The effect of nitrogen fertilizer on expression of slow mildewing resistance in Knox wheat. Phytopathology 67:1051–1056. doi:10.1094/Phyto-67-1051

Steel, R.G.D., J.H. Torrie, and D.A. Dickey. 1997. Principles and procedures of Statistics. A biometrical Approach. 3rd ed. McGraw Hill, New York.

Thompson, D.L., and R.R. Bergquist. 1984. Inheritance of mature plant resistance to Helminthosporium maydis race 0 in maize. Crop Sci. 24:807–811. doi:10.2135/cropsci1984.0011183X002400040042x

Tian, D., M.B. Traw, J.Q. Chen, M. Kreitman, and J. Bergelson. 2003. Fitness costs of R-gene-mediated resistance in Arabidop-sis thaliana. Nature 423:74–77. doi:10.1038/nature01588

Troyer, A.F. 1999. Background of U.S. hybrid corn. Crop Sci. 39:601–626. doi:10.2135/cropsci1999.0011183X003900020001x

Wisser, R.J., P.J. Balint-Kurti, and R.J. Nelson. 2006. The genetic architecture of disease resistance in maize: A synthesis of published studies. Phytopathology 96:120–129. doi:10.1094/PHYTO-96-0120

Zuber, M.S. 1973. Registration of 20 maize parental lines. Crop Sci. 13:779–780. doi:10.2135/cropsci1973.0011183X001300060085x

Zwonitzer, J., D.M. Bubeck, D. Bhattramakki, M.M. Good-man, C. Arellano, and P.J. Balint-Kurti. 2009. Use of selec-tion with recurrent backcrossing and QTL mapping to iden-tify loci contributing to southern leaf blight resistance in a highly resistant maize line. Theor. Appl. Genet. 118:911–925. doi:10.1007/s00122-008-0949-2