- Email: [email protected]
Stalk rot fungi affect grain sorghum yield components in an inoculation stage-specific manner
Crop Protection 94 (2017) 97e105
Contents lists available at ScienceDirect
Crop Protection journal homepage: www.elsevier.com/locate/cropro
Stalk rot fungi affect grain sorghum yield components in an inoculation stage-specific manner Y.M.A.Y. Bandara a, D.K. Weerasooriya b, T.T. Tesso b, P.V.V. Prasad b, C.R. Little a, * a b
Department of Plant Pathology, 4024 Throckmorton Plant Sciences Center, Kansas State University, Manhattan, KS 66506, USA Department of Agronomy, 2004 Throckmorton Plant Sciences Center, Kansas State University, Manhattan, KS 66506, USA
a r t i c l e i n f o
a b s t r a c t
Article history: Received 22 June 2016 Received in revised form 22 November 2016 Accepted 16 December 2016
Although stalk rots are among the most prevalent and destructive diseases of sorghum, no systematic crop loss assessment has been reported for these diseases under controlled-inoculations. The objective of this study was to assess the impacts of Fusarium stalk rot and charcoal rot on grain sorghum yield components when plants were inoculated at two growth stages (GS). Four genotypes were field evaluated against three Fusarium spp. (F. thapsinum, F. proliferatum, F. andiyazi) and Macrophomina phaseolina. Inoculations were performed at GS1 (30 d after emergence) and GS3 (14 d after flowering). Panicles were harvested at physiological maturity and assessed for total seed weight (TSW), 100-seed weight (100-SW), and total seeds per panicle (TSP). The total number of reproductive sites and unfilled spikelets were counted per rachis (TRSR and NUSR, respectively) and panicle (TRSP, NUSP) bases. Length and nodes crossed by the lesion in split stems were measured to evaluate disease severity. Pathogens significantly reduced TSW in comparison to the control at both GS1 and GS3. The four pathogens, on average, caused greater TSW reductions when inoculated at GS1 (52%) than at GS3 (37%). All pathogens reduced TSP upon GS1 inoculation and 100-SW upon GS3 inoculations. All pathogens significantly reduced seed set percentage when plants were inoculated at GS1 while inoculations at GS3 did not have a significant impact. GS1 inoculations significantly decreased TRSP, demonstrating pathogen interference with head formation resulting in smaller heads than control. Although inoculations at GS1 had a greater impact on yield, inoculations at GS1 and GS3 did not significantly differ in disease severity. This study revealed inoculation stage-specific effects of stalk rot pathogens on yield parameters and provided insights into key yield traits to be emphasized in sorghum breeding programs to produce stalk rot tolerant sorghum genotypes. © 2016 Elsevier Ltd. All rights reserved.
Keywords: Sorghum bicolor Fusarium stalk rot Charcoal rot Yield components Resistance Tolerance
1. Introduction Stalk rots are among the most ubiquitous diseases of sorghum (Sorghum bicolor L. Moench) worldwide (Zummo, 1984; Tesso et al., 2010). Infected plants are characterized by injured cortical and vascular tissue in the root and stalk systems that leads to impaired nutrient and water absorption and translocation (Hundekar and Anahosur, 1994). Bandara et al. (2016a,b) have shown that stalk rot fungi affect grain sorghum leaf chlorophyll content in a genotype- and growth stage-dependent manner. Under extreme conditions, stalk rot diseases can cause lodging due to the poor stalk strength in the infected area (Zummo, 1984). Prolonged water
* Corresponding author. E-mail address: [email protected] (C.R. Little). http://dx.doi.org/10.1016/j.cropro.2016.12.018 0261-2194/© 2016 Elsevier Ltd. All rights reserved.
deficit stress, high temperature, and unbalanced mineral nutrition have been reported to predispose sorghum plants to stalk rot diseases (Dodd, 1980; Seetharama et al., 1987). There are two stalk rot diseases of sorghum, Fusarium stalk rot, caused by Fusarium spp., and charcoal rot, caused by Macrophomina phaseolina (Tassi) Goidanich (Tarr, 1962). Typically, infection by Fusarium spp. produces red colored lesions while M. phaseolina produces dark-colored lesions at the site of infection. Although the reaction of sorghum genotypes to stalk rot diseases can vary based on the causal organism (Tesso et al., 2005), Fusarium thapsinum Klittich, Leslie, Nelson & Marasas is considered as the most aggressive stalk rot Fusarium pathogen of sorghum (Tesso et al., 2005, 2010, Tesso and Ejeta, 2011; Leslie et al., 2005). Charcoal rot is one of the most widespread and destructive stalk rot diseases of sorghum (Mughogho and Pande, 1984). Therefore, screening sorghum germplasm with several stalk rot fungi is critical
98
Y.M.A.Y. Bandara et al. / Crop Protection 94 (2017) 97e105
for a better understanding of the pathogen dynamics involved and to recommend sorghum genotypes for multiple environments where different stalk rot fungi predominate. Bandara et al. (2016a,b) have reported the negative impacts of stalk rot diseases on important biofuel traits of sweet sorghum under controlled inoculations. Although no systematic quantitative crop loss assessment results have been described for stalk rot diseases under controlled-inoculations for grain sorghum, standability and grain weight have been recognized to be affected by stalk rot fungi, which contribute to yield losses in sorghum (Tesso et al., 2012). Zummo (1980) reported that stalk rots slow down or inhibit the grain-filling process and result in shriveled seeds. Anahosur and Patil (1983) have attributed charcoal rot-mediated sorghum seed weight losses to the varying levels of lodging caused by M. phaseolina. Seetharama et al. (1991) have reported that there were no simple correlations between sorghum yield or yield components with stalk rot incidence. However, their conclusions were based upon a study conducted under natural inoculation. Moreover, the observed stalk rot incidences were attributed to M. phaseolina, based on symptomatology. Moreover, a few publications have provided information concerning the plant growth stage at which stalk rot infections occur and possible impacts they could have on yield. For example, Reed et al. (1983) and Jardine and Leslie (1992) reported that most stalk rot pathogens colonize the stalk and incite disease during the “post-flowering” stages. Khune et al. (1984) indicated that stalk rot pathogens are found in host tissues at various growth stages. Eastin and Sullivan (1974) described a simple development stage terminology suitable for yield and yield components of grain sorghum according to the following growth stages: (i) the vegetative period from planting to panicle initiation (GS1); (ii) the reproductive period from panicle initiation to flowering (GS2); and (iii) grain filling from flowering to physiological maturity (GS3). The number of seeds per panicle is physiologically determined during the second growth stage (GS2) when floret number is set in the developing panicle (Eastin et al., 1999; Maiti and Bidinger, 1981). The yield components of sorghum include the number of panicles per square meter, number of seeds per panicle, and seed weight, which are defined as the first, second, and third yield components, respectively (Maman et al., 2004). As the second yield component directly relates to GS2, any biotic and/or abiotic stress that prevails before or at the onset of this stage could have serious second yield component impacts. Similarly, since seed filling is predominantly associated with GS3, stresses occurring at this stage can impact the third yield component. Therefore, the objective of this study was to determine the effects of multiple stalk rot pathogens and the stage of pathogen infection on the yield and yield components of selected grain sorghum genotypes. We hypothesize that different stalk rot fungi and the growth stage (GS1 and GS3) at which infection occurs have differential effects on sorghum yield, in general, and upon the second and third yield components, in particular. 2. Materials and methods 2.1. Sorghum genotypes and fungal isolates Sorghum genotypes, SC599R (Fusarium stalk rot and charcoal stalk rot resistant, breeding line), BTx3042 (Fusarium stalk rot and charcoal rot susceptible, breeding line), 84G62 (Dupont Pioneer, hybrid), and DKS37-07 (Dekalb, hybrid) were used. Pioneer 84G62 and DKS37-07 are classified as charcoal rot tolerant and are commonly grown hybrids by producers in the state of Kansas, USA. The three Fusarium spp. used in this study, F. thapsinum, F. proliferatum, and F. andiyazi, were previously isolated from infected stalks by the Row Crops Pathology Lab at Kansas State
University from local sorghum fields. A portion of the translation elongation factor (TEF-1a) gene of these isolates was PCR amplified with ef1 (forward) and ef2 (reverse) primers and subsequently sequenced (Geiser et al., 2004). Sequence information was used as a query for comparison with the NCBI and FUSARIUM-ID databases using BLAST algorithm to confirm the species identity. The M. phaseolina isolate (r144) was provided by Dr. Gary Odvody, Texas A&M AgriLife Research and Extension, Corpus Christi, Texas. 2.2. Field experiments Field experiments were carried out during the 2013 and 2014 growing seasons at the Kansas State University agronomy research farm in Manhattan (39.22 N, 96.60 W) and Ashland (39.13 N, 96.62 W), Kansas, respectively (hereafter referred to as the two environments) under rainfed conditions. Field preparation, planting, and crop maintenance were conducted according to standard procedures for sorghum. Seeds were treated with fungicide (ethyl mercaptan (Captan), two mL kg-1 seed) and planted in rows of 5 m length 0.75 m width (whole plot). The treatment structure was a 4 2 5 factorial where factors one, two, and three consisted of four sorghum genotypes, two stages of inoculation, and four pathogens and mock-inoculated control, respectively. The field study was arranged in a split-split-plot design with randomized complete blocks. Genotypes were assigned to the whole-plot unit and stage of inoculation to the sub-plot unit (2.5 m length 0.75 m width). Pathogen and control treatments were assigned to the sub-sub-plot unit (0.5 m length 0.75 m width). There were three subsample plants (¼ observational unit) in each pathogen or mock-inoculated control treatment. For each environment, there were three replicated blocks. 2.3. Inoculum preparation, artificial inoculation, and measuring yield parameters Inoculum preparation was performed according to the methods described by Bandara et al. (2015) and the protocol described therein for F. thapsinum inoculum preparation was also used to prepare F. proliferatum and F. andiyazi inocula. GS1 inoculations were performed at 30 days after seedling emergence for all genotypes when seedlings showed 7 to 10 leaf collars (ligules) (before flag leaf emergence). GS3 inoculations were performed at 14 days after flowering. At both stages, plants were inoculated by injecting 0.1 mL of inoculum at a concentration of 2 106 conidia or mycelium fragments per mL into the basal internode of the stalk. Conidia were used for F. thapsinum, F. proliferatum, and F. andiyazi inoculations. Hyphal fragments were used for M. phaseolina since this fungus does not typically produce spores. Phosphate-buffered saline (PBS) was used for the mock-inoculated control treatment. Inoculated plants were harvested at 110 days after planting (after all four genotypes attained physiological maturity). Panicles were separated from plants. The total number of rachis per panicle (TRP), total seeds per rachis (TSR), and the number of unfilled spikelets per rachis (NUSR) were counted for each panicle. The total number of reproductive sites per rachis (TRSR) in each panicle was computed by adding TSR and NUSR. These data were used to calculate the total seeds per panicle (TSP), number of unfilled spikelets per panicle (NUSP), total number of reproductive sites per panicle [TRSP ¼ NUSP þ TSP], and seed set percentage [SSP ¼ (TSP/ TRSP) 100]. Panicles were then dried for 10 d at 40 C and threshed to measure the total seed weight per panicle (TSW). 100seed weight (100-SW) was computed using the equation, (TSW 100)/TSP. Stems were split longitudinally to measure lesion length (cm) and the number of nodes crossed by the lesion.
Y.M.A.Y. Bandara et al. / Crop Protection 94 (2017) 97e105
2.4. Statistical analysis Data were analyzed for variance (ANOVA) using the PROC GLIMMIX procedure of SAS software version 9.2 (SAS Institute, 2008). The restricted maximum likelihood (REML) method was used to estimate variance components. Degrees of freedom for the denominator in F tests were computed using the Kenward-Roger option. Model assumptions were tested using studentized residual plots (for identical and independent distribution of residuals) and Q-Q plots (for normality of residuals). Whenever residuals were not homogeneously distributed, appropriate heterogeneous variance models were fitted to meet the model assumptions. Bayesian information criterion (BIC) was used to determine the most suitable model that best fit data after accounting for model assumptions. Data analyses were initially conducted for each environment considering the stage of inoculation, genotype, and pathogens as fixed factors. Block was regarded as a random factor. Initial analyses performed within each environment revealed similar significance levels and interaction effects among fixed factors. Therefore, the data were analyzed for combined environments by considering the stage of inoculation, genotype, and pathogens as fixed factors and environment and block as random factors. Blocks were considered as nested within environments for this analysis. Results and conclusions drawn from the combined analysis are presented here. Means separations were carried out using the PROC GLMMIX procedure of SAS. Main effects of factors were determined with adjustments for multiple comparisons using the TukeyKramer test. Simple effects of factors were determined using the Bonferroni adjustment when higher order interactions among factors were statistically significant for the trait concerned. Only relevant pairwise comparisons were considered for determining the critical comparison-wise error rate (aCER) and were subsequently compared against unadjusted P-values resulting from the Bonferroni correction to test the significance of treatment mean differences. Pearson correlation coefficients and significance levels among measured traits were computed using the PROC CORR procedure of SAS. 3. Results 3.1. Analysis of variance (ANOVA) Pathogen and genotype had significant main effects on lesion length while the main effect of genotype was significant on the total number of rachis per panicle. ANOVA did not reveal significant pathogen genotype stage interactions for most traits measured except for total seeds per panicle (Table 1). The two-way interaction between pathogen and stage was significant for total seed weight, 100-seed weight, total seeds per panicle, total number of
99
reproductive sites per panicle, seed set percentage, total seeds per rachis, and total number of reproductive sites per rachis while the interaction between pathogen and genotype was significant for total number of seeds per panicle, total number of reproductive sites per panicle, and total number of reproductive sites per rachis. The genotype stage interaction was significant for total seeds per rachis, and the total number of reproductive sites per rachis. 3.2. Effect of stalk rot pathogens and inoculation stage on disease severity All pathogens significantly increased lesion length and nodes crossed in comparison to control (P < 0.0001; Fig. 3E and F). No significant differences among pathogens were observed. Plant genotype also had significant main effect on lesion length (P ¼ 0.0465; Table 1). 84G62 and DKS37-07 had significantly higher lesion length (P < 0.05) than BTx3042 and SC599. 3.3. Effect of stalk rot pathogens and inoculation stage on seed weights and numbers The mean total seed weight per panicle of the control treatment was significantly greater than the F. andiyazi, F. proliferatum, F. thapsinum, and M. phaseolina treatments for inoculation at both GS1 and GS3 (P < 0.05; Fig. 2A). Inoculations reduced total seed weight per panicle by 42.2 to 57.2% (GS1 inoculation) and 29.4 to 41.2% (GS3 inoculation). The effects of the four pathogens on mean total seed weight per panicle were not significantly different from each other at either of the inoculation stages (Fig. 2A). 100-seed weight values were not significantly decreased from the control by three of the Fusarium pathogens upon GS1 inoculation (P < 0.001; Fig. 2B). M. phaseolina caused an 18.3% reduction in 100-seed weight. Its impact on 100-seed weight was also significantly higher compared to that of F. andiyazi and F. thapsinum (P < 0.001). At GS3 inoculation, all pathogens significantly reduced 100-seed weight compared to control (P < 0.001; Fig. 2B) although the effects of the four pathogens on mean 100-seed weight were not significantly different from each other. On average, the four pathogens reduced 100-seed weight by 16.1% after inoculation at GS3. When inoculated at GS1, all four pathogens significantly decreased the total seeds per panicle of 84G62 compared to the control inoculation (P < 0.001; Fig. 2C). A similar result was observed for DKS37-07 (P < 0.0001; Fig. 2E) for the other three pathogens but not for F. thapsinum inoculation as compared with the control. F. proliferatum caused a significant seed number reduction of SC599 compared to control and M. phaseolina (P < 0.001; Fig. 2F). Pathogens did not significantly reduce seed number of BTx3042 when this genotype was inoculated at GS1
Table 1 P-values of F-statistic from analysis of variance (ANOVA) for disease severity, yield, and yield related traits measured with four sorghum genotypes after inoculation with F. thapsinum, F. andiyazi, F. proliferatum, and M. phaseolina at two plant growth stages, GS1 (growth stage 1, 30 d after emergence); GS3 (growth stage 3, 14 d after flowering) across two field studies (a ¼ 0.05). Effect
LLa
NC
TSW
100-SW
TSP
NUSP
TRSP
SSP
TRP
TSR
NUSR
TRSR
Genotype (G) Stage (S) GS Pathogen (P) PG PS PGS
0.0465 0.6434 0.8008 <0.0001 0.0501 0.6727 0.1740
0.3060 0.6955 0.9957 <0.0001 0.2348 0.5776 0.2174
0.2188 0.2538 0.1600 0.0020 0.3035 0.0195 0.3529
0.0688 0.9180 0.1760 0.0003 0.1831 0.0031 0.0833
0.0379 <0.0001 0.2259 <0.0001 0.0138 0.0106 0.0187
0.1351 0.2448 0.1391 0.3393 0.1016 0.1357 0.1945
0.0022 0.0202 0.7292 0.0001 0.0126 0.0457 0.0592
0.8093 0.1380 0.4150 0.0003 0.3459 0.0043 0.1434
<0.0001 0.1216 0.1592 0.1305 0.1960 0.6211 0.1055
<0.0001 0.0006 0.0090 0.0143 0.0994 0.0185 0.1507
0.0623 0.0012 0.6314 0.4591 0.0760 0.1871 0.1366
<0.0001 0.0002 0.0013 0.0144 0.0071 0.0275 0.4138
a LL ¼ lesion length (cm); NC ¼ number of nodes crossed by the lesion; TSW ¼ total seed weight panicle1 (g); 100-SW ¼ 100-seed weight (g); TSP, NUSP, TRSP ¼ total number of seeds, unfilled spikelets, and total reproductive sites panicle1, respectively; SSP ¼ seed set percentage; TRP ¼ total number of rachis panicle1; TSR, NUSR, TRSR ¼ total seeds, unfilled spikelets, and total number of reproductive sites rachis1.
100
Y.M.A.Y. Bandara et al. / Crop Protection 94 (2017) 97e105
(Fig. 2D). Inoculations at GS3 did not result in seed number reductions for 84G62, BTx3042, or SC599 (Fig. 2C, D, F). However, F. thapsinum reduced total seeds per panicle of DKS37-07 compared to control (Fig. 2E). 3.4. Effect of stalk rot pathogens and inoculation stage on panicle traits The total number of reproductive sites per panicle was significantly decreased by all four pathogens at GS1 inoculation in comparison to control (P < 0.0001; Fig. 3A). No significant differences were observed among pathogens. On average, the four pathogens reduced the total number of reproductive sites per panicle by 32.4%. At GS3, F. thapsinum and M. phaseolina significantly reduced reproductive sites per panicle (P < 0.001; Fig. 3A), although no significant differences between pathogens were observed. F. thapsinum and M. phaseolina reduced reproductive sites per panicle by 24.0% compared to control. All pathogens significantly decreased seed set percentage (P < 0.0001; Fig. 3B) when inoculated at GS1, although pathogens did not significantly differ from one another. On average, the four pathogens caused a 12.0% reduction in seed set compared to control. Inoculations at GS3 did not significantly reduce seed set. The number of rachis per panicle and the number of unfilled spikelets per rachis were not significantly affected by pathogens (Table 1). When compared with control, all four pathogens significantly reduced the total number of reproductive sites per rachis (P < 0.01) and total seeds per rachis (P < 0.005) upon GS1 inoculations (Fig. 3C and D) while no such effect was evident with GS3 inoculations. The number of unfilled spikelets per rachis was significantly different between GS1 (6.75 ± 0.5) and GS3 (5.22 ± 0.5) inoculations (P < 0.005). 3.5. Correlation analysis among measured traits Pearson correlation coefficients were computed for all trait measurements within individual stages of inoculation. Total seeds per panicle, the number of unfilled spikelets per panicle, and seed set percentage had significant negative correlations with lesion length and diseased nodes crossed at GS1 inoculation (Table 2). Nodes crossed was also significantly and negatively correlated with total seed weight per panicle. Analysis of GS3 inoculations revealed significant and negative correlation of nodes crossed with total seed weight per panicle (r ¼ 0.20, P ¼ 0.03), 100-seed weight and total seeds per panicle (r ¼ 0.19, P ¼ 0.04), and total number of reproductive sites per panicle (r ¼ 0.20, P ¼ 0.03) (Table 2). Table 2 Pearson correlation coefficients and significance levels among eight traits measured from four sorghum genotypes inoculated with Fusarium thapsinum, F. andiyazi, F. proliferatum, and Macrophomina phaseolina at GS1 and GS3.
GS1 LL NC GS3 LL NC
TSW
100-SW
TSP
NUSP
TRSP
SSP
0.1567 0.0916 0.1912 0.0389a
0.0292 0.7545 0.0321 0.7312
0.2137 0.0207a 0.2558 0.0054a
0.3040 0.0011a 0.2990 0.0014a
0.0628 0.5126 0.1146 0.2307
0.3548 0.0001a 0.3758 <0.0001a
0.1169 0.2034 0.2006 0.0281a
0.0541 0.5593 0.1925 0.0359a
0.1410 0.1261 0.1925 0.0359a
0.0273 0.7679 0.1193 0.1962
0.1386 0.1324 0.2000 0.0292a
0.0866 0.3490 0.0071 0.9392
a LL ¼ lesion length (cm); NC ¼ nodes crossed by lesion; TSW ¼ total seed weight panicle1 (g); 100-SW ¼ 100-seed weight (g); TSP ¼ total number of seeds panicle1; NUSP ¼ number of unfilled spikelets panicle1; TRSP ¼ total number of reproductive sited panicle1; SSP ¼ seed set percentage. *Pearson correlation coefficients are significant at P < 0.05.
4. Discussion This study was designed to investigate the impacts of stalk rot diseases on the second (number of seeds per panicle) and third (seed weight) yield components of grain sorghum. The number of fertile tillers produced by a plant is a key determinant of the first yield component. Tillering has been shown to be determined by various factors including plant density (Gerik and Neely, 1987; Lefarge et al., 2002), tiller fertility (Lefarge et al., 2002), environment (Kim et al., 2010; Downes, 1968), and seeding pattern (Bandaru et al., 2006). In this study, the first yield component (number of panicles per square meter) was not investigated since stalk rot pathogens do not directly affect this component unless stalk rot-mediated plant lodging is considered as a means of panicle loss under mechanistic harvesting. A primary objective of this study was to determine whether the growth stage at which pathogen infection occurs has a significant impact on grain yield and yield components. The two-way interaction effect between pathogen and stage of inoculation was found to be significant for most traits, indicating inoculation stagespecific impacts of pathogens. When plants were inoculated at GS1, reduced total seeds per panicle contributed more to the reduction of total seed weight than did reduction of 100-seed weight. However, 100-seed weight had a greater impact on total seed weight than the total seeds per panicle for GS3 inoculation. The number of reproductive sites per panicle is physiologically determined during the second growth stage (GS2) when floret number is set in the developing panicle (Eastin et al., 1999; Maiti and Bidinger, 1981). The effect of GS1 inoculations on the total seeds per panicle could be partitioned into two sub-effects. First, pathogens decreased seed set percentage (Fig. 3B) through an impeded grain filling process. Secondly, as pathogens significantly reduced the total number of reproductive sites per panicle (i.e., the total number of seeds and unfilled spikelets) (Fig. 3A), the effect of GS1 inoculations on physiological seed set is apparent. Therefore, decreased numbers of seeds per panicle can be attributed to the pathogens' effect on the physiological development of reproductive sites (spikelets) leading to a smaller sink in comparison to the potential sink size under normal growth conditions. Mu et al. (2005) showed that the size and activity of the early reproductive shoot apex (incipient panicle) determine the number of florets generated on the rice panicle, thus determining sink size. They found the initial size of the reproductive apex as the determinant of the primary branch (rachis) number while cell division activity in subsequent apex growth was the determinant of floret number per rachis. Although a similar study in sorghum has not been reported, this may be a common developmental pathway among most graminaceous crops. The present study provided evidence for the significant reduction of the total number of reproductive sites (florets) per rachis as a result of inoculation with stalk rot pathogens at GS1 while no such effect was evident after GS3 inoculation (Table 1, Fig. 3C). Therefore, our observations along with previous literature suggest the stalk rot pathogens’ capacity to reduce sink size via reduced floret numbers (through reduced apex cell division activity). We did not find evidence for reduced rachis numbers due to pathogen infection (Table 1). However, the size of the reproductive apex as indicated by rachis number appeared to be a characteristic of genotype. Taken together, these results suggest two possible ways that stalk rot pathogens can interfere with seed production in panicles. Since GS1 inoculations had a significant effect on seed set percentage, but a non-significant effect on 100-seed weight (Fig. 2B), the ability of stalk rot pathogens to completely inhibit the filling of some spikelets in the panicle was apparent. The opposite phenomenon was observed with GS3 inoculations and revealed that
Y.M.A.Y. Bandara et al. / Crop Protection 94 (2017) 97e105
pathogens mediated partial inhibition of filling across the spikelets in the panicle. Moreover, the significant correlations between disease severity traits (lesion length and nodes crossed) with percent seed set (negative correlation), no significant correlation with 100seed weight, and significant positive correlation with number of unfilled spikelets per panicle (Table 2) provided evidence for the complete inhibition of some spikelets in the panicle when inoculation was performed at GS1. On the other hand, non-significant correlations between disease severity traits (nodes crossed in this case) and seed set percentage/number of unfilled spikelets per
101
panicle, and significant and negative correlation with 100-seed weight indicated the partial inhibition of grain filling across panicles after GS3 inoculations. Pre-flowering floret abortion (Senanayake et al., 1991) is a major component of grain yield of many cereals (Kernich et al., 1997) and is the third determinant of panicle size, besides primary branch number and the number of florets on primary branches (Yamagishi et al., 2004). In the current study, more aborted florets were observed in GS1 inoculated panicles when compared with mock-inoculated controls (Fig. 1A and B), suggesting that the inoculation of stalk rot pathogens at GS1
Fig. 1. Effects of GS1 inoculation on pre-flowering floret abortion and seed set for (A) the mock-inoculated control treatment and (B) after inoculation with stalk rot pathogens. Arrows indicate aborted florets.
102
Y.M.A.Y. Bandara et al. / Crop Protection 94 (2017) 97e105
Fig. 2. Comparison of mean values (± standard error) among pathogen treatments for (A) total seed weight per panicle (g) and (B) 100-seed weight (g) across environments and genotypes at two inoculation stages (GS1 and GS3). Total number of seeds per panicle for genotypes (C) 84G62, (D) BTx3042, (E) DKS37-07, and (F) SC599 across environments and at GS1 and GS3 using Bonferroni adjustment. Lower case letters compare pathogen effects at GS1 while upper case letters compare effects at GS3. Means followed by the same letters within each inoculation stage and treatments without letter designations are not significantly different based on the adjusted P-value for multiple comparisons at a ¼ 0.05. CON ¼ phosphate-buffered saline mock-inoculated control, FA ¼ Fusarium andiyazi, FP ¼ F. proliferatum, FT ¼ F. thapsinum, and MP ¼ Macrophomina phaseolina.
Y.M.A.Y. Bandara et al. / Crop Protection 94 (2017) 97e105
103
Fig. 3. Comparison of mean values (± standard error) among pathogen treatments for (A) number of reproductive sites per panicle, (B) percent seed set, (C) number of reproductive sites per rachis, and (D) total seeds per rachis across environments and genotypes at two inoculation stages (GS1 and GS3). (E) Lesion length (cm) and (F) number of nodes crossed by the lesion across environments, genotypes, and inoculation stages. Comparisons made in panels A, B, C, and D were based on Bonferroni adjustments while those of E and F were based on the Tukey-Kramer's test (see Materials & Methods). Means followed by the same letters in each case and treatments without letter designations are not significantly different based on the adjusted P-value for multiple comparisons at a ¼ 0.05. CON ¼ phosphate-buffered saline control, FA ¼ F. andiyazi, FP ¼ F. proliferatum, FT ¼ F. thapsinum, and MP ¼ M. phaseolina.
104
Y.M.A.Y. Bandara et al. / Crop Protection 94 (2017) 97e105
leads to an increase in pre-flowering floret abortion. However, GS3 inoculations should not have any impact on pre-flowering floret abortion, as the inoculation was performed at 14 d after flowering. Hence, GS3 inoculation only affects grain filling. Fig. 4 is a generalized schematic representation of the yield determination process under stalk rot infection at GS1 and GS3. The effect of pathogens on lesion length and nodes crossed was not found to be inoculation stage-specific (Table 1; Fig. 3E and F). After GS1 inoculation, one would expect lengthier lesions or more nodes crossed since pathogens exist in plants for a longer duration (~75 d) than GS3 (~35 d). However, the data did not support that expectation. It could be that Fusarium pathogens grow endophytically and cause no visual symptoms during vegetative stages, but cause damage when seed number is being determined during panicle development, which leads to reduced total seed weight per panicle. In fact, some Fusarium species such as F. thapsinum and F. proliferatum can colonize sorghum symptomlessly as endophytes (Leslie, 2000). The two hybrids used in this study, 84G62 and DKS37-07, had significantly greater mean lesion lengths (P < 0.05) compared to the two inbred lines tested, BTx3042 and SC599. This observation agrees with the findings of Das et al. (2008) where they observed higher disease severity with high yielding genotypes. M. phaseolina inoculations resulted in shorter mean lesion lengths and fewer nodes crossed compared to the three Fusarium spp. (Fig. 3E and F). M. phaseolina is a necrotrophic pathogen (Islam et al., 2012) while Fusarium spp. are hemibiotrophic (Bacon and Yates, 2006). Therefore, Fusarium spp. may be more pathogenic than M. phaseolina during the early stages of plant growth while necrotrophs exhibit higher pathogenicity and virulence during the senescent phase. Charcoal rot is problematic under post-flowering stress conditions along with natural plant senescence, especially in the absence of the staygreen (delayed senescence) trait. Except for BTx3042, the other three genotypes tested in this study possess the staygreen trait. Therefore, the shorter lesions and fewer nodes crossed observed in staygreen genotypes in response to M. phaseolina inoculation were not surprising. However, despite the shorter lesion lengths and lower numbers of nodes crossed associated with M. phaseolina, it caused the highest reduction in total seed weight
compared to the three Fusarium spp. at both inoculation stages (GS1, 57.2%; GS3, 41.2%). This finding agrees with earlier reports that M. phaseolina is the most aggressive stalk rot pathogen in sorghum (Anahosur and Patil, 1982; Mughogho and Pande, 1984; Rosenow, 1984; Tesso et al., 2005). Analysis within each inoculation stage showed that there was no significant correlation between charcoal rot disease severity traits with percent yield reduction [(TSWcont e TSWinoc)/TSWcont] of genotypes under infection (GS1, r ¼ þ0.30, P ¼ 0.26; GS3, r ¼ þ0.19, P ¼ 0.47). Percent yield reduction compared to mock-inoculated control is a direct indicator of the genotype's ability to tolerate stalk rot or charcoal rot disease. This emphasizes the importance of incorporating both resistance (e.g., severity traits) and tolerance measures (e.g., yield traits) for screening sorghum genotypes against stalk rot diseases. 5. Conclusions The four stalk rot pathogens each significantly decreased total seed weight per panicle compared to the control treatment at two inoculation stages (GS1 and GS3). A greater yield reduction was observed after GS1 inoculations compared to those at GS3. The reduction of total seed weight per panicle due to GS1 inoculations was predominantly attributed to the reduction in the total number of seeds per panicle while 100-seed weight was found to be the key contributing yield component leading to reduced total seed weight per panicle with GS3 inoculations. Inoculations at GS1 significantly decreased the total number of reproductive sites per panicle, which indicated that the stalk rot pathogens had the ability to directly or indirectly interfere with panicle formation resulting in smaller heads compared to the control. Results further suggested the pathogens’ ability to induce/enhance pre-flowering floret abortion upon GS1 inoculation, contributing to reduced total seeds per panicle. Therefore, plant breeding and production research should consider both of these yield components to achieve enhanced sorghum production. Seed number merits greater attention where sorghum cultivation potentially undergoes pre-flowering stress conditions such as drought that predispose plants to natural infection by stalk rot pathogens. However, seed weight should be targeted where the crop is more vulnerable to post-flowering stress
Fig. 4. Schematic representation of yield determination for Sorghum bicolor L. Moench after stalk rot inoculations.
Y.M.A.Y. Bandara et al. / Crop Protection 94 (2017) 97e105
conditions. In general, Fusarium stalk rot was more severe compared to charcoal rot, although charcoal rot caused greater yield reduction at both inoculation stages. Significant correlations were evident between disease severity traits and yield related traits although correlation coefficients were not strong. There were no significant correlations between disease severity traits such as lesion length and % yield reduction. These finding suggested that disease severity measurements are inadequate as sole criteria in screening germplasm for stalk rot disease resistance. Our results emphasize the importance of a stalk rot screening approach in which “tolerance” should be an integral component. Acknowledgments The Kansas Grain Sorghum Commission is gratefully acknowledged for their financial support of this research. The authors also wish to thank Kyle Shroyer, Chad Brady, and Dr. Kraig Roozeboom for their technical assistance and advice during this study. This paper is Contribution No. 14-344-J from the Kansas Agricultural Experiment Station, Manhattan, Kansas. References Anahosur, K.H., Patil, S.H., 1982. Some promising sources of resistance to charcoal rot of sorghum. Sorghum Newsl. 25, 118. Anahosur, K.H., Patil, S.H., 1983. Assessment of losses in sorghum seed weight due to charcoal rot. Indian Phytopathol. 36, 85e88. Bacon, C.W., Yates, I.E., 2006. Endophytic Root colonization by Fusarium species: histology, plant interactions, and toxicity. In: Schulz, B.J.E., Boyle, C.J.C., Sieber, T.N. (Eds.), Microbial Root Endophytes. Springer, pp. 133e152. Bandara, Y.M.A.Y., Weerasooriya, D.K., Tesso, T.T., Little, C.R., 2016a. Stalk rot diseases impact sweet sorghum biofuel traits. BioEnergy Res. http://dx.doi.org/ 10.1007/s12155-016-9775-6. Bandara, Y.M.A.Y., Weerasooriya, D.K., Tesso, T.T., Little, C.R., 2016b. Stalk rot fungi affect leaf greenness (SPAD) of grain sorghum in a genotype- and growth stagespecific manner. Plant Dis. 100 (10), 2062e2068. Bandara, Y.M.A.Y., Perumal, R., Little, C.R., 2015. Integrating resistance and tolerance for improved evaluation of sorghum lines against Fusarium stalk rot and charcoal rot. Phytoparasitica 43, 485e499. Bandaru, V., Stewart, B.A., Baumhardt, R.L., Ambati, S., Robinson, C.A., Schlegel, A., 2006. Growing dryland grain sorghum in clumps to reduce vegetative growth and increase yield. Agron. J. 98 (4), 1109e1120. Das, I.K., Prabhakar, Indira, S., 2008. Role of stalk-anatomy and yield parameters in development of charcoal rot caused by Macrophomina phaseolina in winter sorghum. Phytoparasitica 36, 199e208. Dodd, J.L., 1980. The photosynthetic stress translocation balance concept of sorghum stalk rot, 11-15 December 1978. In: Proceedings of International Workshop of Sorghum Diseases. Texas A&M University/ICRISAT Press, Hyderabad, India, pp. 300e305. Downes, R.W., 1968. The effect of temperature on tillering of grain sorghum seedlings. Crop Pasture Sci. 19 (1), 59e64. Eastin, J.D., Sullivan, C.Y., 1974. Yield consideration in selected cereals. In: Bielski, R.L., et al. (Eds.), Mechanisms of Regulation of Plant Growth. Bull. 12. Royal Soc. of New Zealand, Wellington, New Zealand, pp. 871e877. Eastin, J.D., Petersen, C.L., Zavala-Garcia, F., Dhopte, A., Verma, P.K., Ounguela, V.B., Wit, M.W., Hernandez, V.G., Munoz, M.L., Gerik, T.J., Gandoul, G.I., Hovney, M.R.A., Onofre, L.M., 1999. Potential heterosis associated with developmental and metabolic processes in sorghum and maize. In: Coors, J.G., Pandey, S. (Eds.), The Genetics and Exploitation of Heterosis in Crops. ASA, CSSA, and SSSA, Madison, WI, pp. 205e229. Geiser, D.M., Gasco, M.D.M.J., Kang, S., Makalowska, I., Veeraraghavan, N., Ward, T.J., Zhang, N., Kuldau1, G.A., O'Donnell, K., 2004. FUSARIUM-ID v. 1.0: a DNA sequence database for identifying Fusarium. Eur. J. Plant Pathol. 110, 473e479. Gerik, T.J., Neely, C.L., 1987. Plant density effects on main culm and tiller development of grain sorghum. Crop Sci. 27 (6), 1225e1230. Hundekar, A.R., Anahosur, K.H., 1994. Pathogenicity of fungi associated with sorghum stalk rot. Karnataka J. Agric. Sci. 7, 291e295. Islam, M.S., Haque, M.S., Islam, M.M., Mannan, E., Emdad, E.M., Halim, A., Hossen1, Q.M.M., Hossain, M.Z., Ahmed, B., Rahim, S., Rahman, M.S., Alam, M.M., Hou, S., Wan, X., Saito, J.A., Alam, M., 2012. Tools to kill: genome of one of the most destructive plant pathogenic fungi Macrophomina phaseolina. BMC Genomics 13, 493. http://dx.doi.org/10.1186/1471-2164-13-493.
105
Jardine, D.J., Leslie, J.F., 1992. Aggressiveness of Gibberella fujikuroi (Fusarium monliforme) isolates to grain sorghum under greenhouse conditions. Plant Dis. 76, 897e900. Kernich, G.C., Halloran, G.N., Flood, R.G., 1997. Variation in duration of pre-anthesis phases of development in barley (Hordeum vulgare). Aust. J. Agric. Res. 48, 59e66. Khune, N.N., Kurhekar, D.E., Raut, J.G., Wangikar, P.D., 1984. Stalk rot of sorghum caused by Fusarium moniliforme. Indian Phytopathol. 37, 316e319. Kim, H.K., van Oosterom, E., Dingkuhn, M., Luquet, D., Hammer, G., 2010. Regulation of tillering in sorghum: environmental effects. Ann. Bot. 106 (1), 57e67. Lefarge, T.A., Broad, I.J., Hammer, G.L., 2002. Tillering in grain sorghum over a wide range of population densities: identification of a common hierarchy for tiller emergence, leaf area development and fertility. Ann. Bot. 90 (1), 87e98. Leslie, J.F., 2000. Fusarium species associated with sorghum. In: Fredericksen, R.A., Odvody, G.N. (Eds.), Compendium of Sorghum Diseases, 2nd Ed. APS Press, St Paul, USA, pp. 30e31. Leslie, J.F., Zeller, K.A., Lamprecht, S.C., Rheeder, J.P., Marasas, F.O., 2005. Toxicity, pathogenicity and genetic differentiation of five species of Fusarium from sorghum and millet. Phytopathology 95, 275e283. Maiti, R.K., Bidinger, F.R., 1981. Growth and development of pearl millet plant (Hyderabad, India). ICRISAT Res. Bull. 6. Maman, N., Mason, S.C., Lyon, D.J., Dhungana, P., 2004. Yield components of pearl millet and grain sorghum across environments in the central Great Plains. Crop Sci. 44, 2138e2145. Mu, C., Nemoto, K., You, Z., Yamagishi, J., 2005. Size and activity of shoot apical meristems as determinants of floret number in rice panicles. Plant Prod. Sci. 8, 51e59. Mughogho, L.K., Pande, S., 1984. Charcoal rot of sorghum. In: Rosenberg, G. (Ed.), Sorghum Root and Stalk Rots, a Critical Review. Proceedings of the Consultative Group Discussion on Research Needs and Strategies for Control of Root and Stalk Rot Diseases. 27 Nov - 2 Dec 1983. ICRISAT Press, Bellagio, Italy. Patancheru, A.P. 502324, India, pp. 11e24. Reed, J.E., Partridge, J.E., Nordquist, P.T., 1983. Fungal colonization of stalks and roots of grain sorghum during the growing season. Plant Dis. 64, 417e420. Rosenow, D.T., 1984. Breeding for resistance to root and stalk rots in Texas. In: Rosenberg, G. (Ed.), Sorghum Root and Stalk Rots, a Critical Review. Proceedings of the Consultative Group Discussion on Research Needs and Strategies for Control of Root and Stalk Rot Diseases. 27 Nov - 2 Dec 1983. ICRISAT Press, Bellagio, Italy. Patancheru, A.P. 502324, India, pp. 209e217. SAS Institute, 2008. The SAS Users Guide, Version 9.2. SAS Inst., Cary, NC. Seetharama, N., Bidinger, F.R., Rao, K.N., Gill, K.S., Mulgund, M., 1987. Effect of pattern and severity of moisture deficit stress on stalk rot incidence in sorghum. I. Use of line source irrigation technique, and the effect of time of inoculation. Field Crops Res. 15, 289e308. Seetharama, N., Sachan, R.C., Huda, A.K.S., Gill, K.S., Rao, K.N., Bidinger, F.R., Reddy, D.M., 1991. Effect of pattern and severity of moisture-deficit stress on stalk-rot incidence in sorghum. II. Effect of source/sink relationships. Field Crops Res. 26, 355e374. Senanayake, N., De Datta, S.K., Naylor, R.E.L., Tomson, W.J., 1991. Lowland rice apical development: stages and cultivar differences detected by electron microscopy. Agron J. 83, 1013e1023. Tarr, S.A.J., 1962. Root and Stalk Diseases: Red Stalk Rot, Colletotrichum Rot, Anthracnose, and Red Leaf Spot, in: Diseases of Sorghum, Sudan Grass and Brown Corn. Commonwealth Mycological Institute, Kew, Surrey, UK, pp. 58e73. Tesso, T.T., Claflin, L.E., Tuinstra, M.R., 2005. Analysis of stalk rot resistance and genetic diversity among drought tolerant sorghum genotypes. Crop Sci. 45, 645e652. Tesso, T.T., Ochanda, N., Little, C.R., Claflin, L., Tuinstra, M.R., 2010. Analysis of host plant resistance to multiple Fusarium species associated with stalk rot disease in sorghum [Sorghum bicolor (L.) Moench]. Field Crops Res. 118, 177e182. Tesso, T., Ejeta, G., 2011. Stalk strength and reaction to infection by Macrophomina phaseolina of brown midrib maize (Zea mays) and sorghum (Sorghum bicolor). Field Crops Res. 120, 271e275. Tesso, T., Perumal, R., Little, C.R., Adeyanju, A., Radwan, G.L., Prom, L.K., Magill, C.W., 2012. Sorghum pathology and biotechnology - a fungal disease perspective: Part II. Anthracnose, stalk rot, and downy mildew. Eur. J. Plant Sci. Biotechnol. 6, 31e44. Yamagishi, J., Miyamoto, N., Hirotsu, S., Laza, R.C., Nemoto, K., 2004. QTLs for branching, floret formation, and pre-flowering floret abortion of rice panicle in a temperate japonica tropical japonica cross. Theor. Appl. Genet. 109, 1555e1561. Zummo, N., 1980. Fusarium disease complex of sorghum in West Africa. In: Proceedings of the International Workshop on Sorghum Diseases. ICRISAT Press, Hyderabad, India, pp. 11e15. Zummo, N., 1984. Fusarium root and stalk disease complex. In: Root, Sorghum, Stalk Rots, A. (Eds.), Critical Review, Proceedings of the Consultative Group Discussion on Research Needs and Strategies for Control of Sorghum Root and Stalk Rot Diseases. ICRISAT Press, Bellagio, Italy, pp. 25e29.
Contents lists available at ScienceDirect
Crop Protection journal homepage: www.elsevier.com/locate/cropro
Stalk rot fungi affect grain sorghum yield components in an inoculation stage-specific manner Y.M.A.Y. Bandara a, D.K. Weerasooriya b, T.T. Tesso b, P.V.V. Prasad b, C.R. Little a, * a b
Department of Plant Pathology, 4024 Throckmorton Plant Sciences Center, Kansas State University, Manhattan, KS 66506, USA Department of Agronomy, 2004 Throckmorton Plant Sciences Center, Kansas State University, Manhattan, KS 66506, USA
a r t i c l e i n f o
a b s t r a c t
Article history: Received 22 June 2016 Received in revised form 22 November 2016 Accepted 16 December 2016
Although stalk rots are among the most prevalent and destructive diseases of sorghum, no systematic crop loss assessment has been reported for these diseases under controlled-inoculations. The objective of this study was to assess the impacts of Fusarium stalk rot and charcoal rot on grain sorghum yield components when plants were inoculated at two growth stages (GS). Four genotypes were field evaluated against three Fusarium spp. (F. thapsinum, F. proliferatum, F. andiyazi) and Macrophomina phaseolina. Inoculations were performed at GS1 (30 d after emergence) and GS3 (14 d after flowering). Panicles were harvested at physiological maturity and assessed for total seed weight (TSW), 100-seed weight (100-SW), and total seeds per panicle (TSP). The total number of reproductive sites and unfilled spikelets were counted per rachis (TRSR and NUSR, respectively) and panicle (TRSP, NUSP) bases. Length and nodes crossed by the lesion in split stems were measured to evaluate disease severity. Pathogens significantly reduced TSW in comparison to the control at both GS1 and GS3. The four pathogens, on average, caused greater TSW reductions when inoculated at GS1 (52%) than at GS3 (37%). All pathogens reduced TSP upon GS1 inoculation and 100-SW upon GS3 inoculations. All pathogens significantly reduced seed set percentage when plants were inoculated at GS1 while inoculations at GS3 did not have a significant impact. GS1 inoculations significantly decreased TRSP, demonstrating pathogen interference with head formation resulting in smaller heads than control. Although inoculations at GS1 had a greater impact on yield, inoculations at GS1 and GS3 did not significantly differ in disease severity. This study revealed inoculation stage-specific effects of stalk rot pathogens on yield parameters and provided insights into key yield traits to be emphasized in sorghum breeding programs to produce stalk rot tolerant sorghum genotypes. © 2016 Elsevier Ltd. All rights reserved.
Keywords: Sorghum bicolor Fusarium stalk rot Charcoal rot Yield components Resistance Tolerance
1. Introduction Stalk rots are among the most ubiquitous diseases of sorghum (Sorghum bicolor L. Moench) worldwide (Zummo, 1984; Tesso et al., 2010). Infected plants are characterized by injured cortical and vascular tissue in the root and stalk systems that leads to impaired nutrient and water absorption and translocation (Hundekar and Anahosur, 1994). Bandara et al. (2016a,b) have shown that stalk rot fungi affect grain sorghum leaf chlorophyll content in a genotype- and growth stage-dependent manner. Under extreme conditions, stalk rot diseases can cause lodging due to the poor stalk strength in the infected area (Zummo, 1984). Prolonged water
* Corresponding author. E-mail address: [email protected] (C.R. Little). http://dx.doi.org/10.1016/j.cropro.2016.12.018 0261-2194/© 2016 Elsevier Ltd. All rights reserved.
deficit stress, high temperature, and unbalanced mineral nutrition have been reported to predispose sorghum plants to stalk rot diseases (Dodd, 1980; Seetharama et al., 1987). There are two stalk rot diseases of sorghum, Fusarium stalk rot, caused by Fusarium spp., and charcoal rot, caused by Macrophomina phaseolina (Tassi) Goidanich (Tarr, 1962). Typically, infection by Fusarium spp. produces red colored lesions while M. phaseolina produces dark-colored lesions at the site of infection. Although the reaction of sorghum genotypes to stalk rot diseases can vary based on the causal organism (Tesso et al., 2005), Fusarium thapsinum Klittich, Leslie, Nelson & Marasas is considered as the most aggressive stalk rot Fusarium pathogen of sorghum (Tesso et al., 2005, 2010, Tesso and Ejeta, 2011; Leslie et al., 2005). Charcoal rot is one of the most widespread and destructive stalk rot diseases of sorghum (Mughogho and Pande, 1984). Therefore, screening sorghum germplasm with several stalk rot fungi is critical
98
Y.M.A.Y. Bandara et al. / Crop Protection 94 (2017) 97e105
for a better understanding of the pathogen dynamics involved and to recommend sorghum genotypes for multiple environments where different stalk rot fungi predominate. Bandara et al. (2016a,b) have reported the negative impacts of stalk rot diseases on important biofuel traits of sweet sorghum under controlled inoculations. Although no systematic quantitative crop loss assessment results have been described for stalk rot diseases under controlled-inoculations for grain sorghum, standability and grain weight have been recognized to be affected by stalk rot fungi, which contribute to yield losses in sorghum (Tesso et al., 2012). Zummo (1980) reported that stalk rots slow down or inhibit the grain-filling process and result in shriveled seeds. Anahosur and Patil (1983) have attributed charcoal rot-mediated sorghum seed weight losses to the varying levels of lodging caused by M. phaseolina. Seetharama et al. (1991) have reported that there were no simple correlations between sorghum yield or yield components with stalk rot incidence. However, their conclusions were based upon a study conducted under natural inoculation. Moreover, the observed stalk rot incidences were attributed to M. phaseolina, based on symptomatology. Moreover, a few publications have provided information concerning the plant growth stage at which stalk rot infections occur and possible impacts they could have on yield. For example, Reed et al. (1983) and Jardine and Leslie (1992) reported that most stalk rot pathogens colonize the stalk and incite disease during the “post-flowering” stages. Khune et al. (1984) indicated that stalk rot pathogens are found in host tissues at various growth stages. Eastin and Sullivan (1974) described a simple development stage terminology suitable for yield and yield components of grain sorghum according to the following growth stages: (i) the vegetative period from planting to panicle initiation (GS1); (ii) the reproductive period from panicle initiation to flowering (GS2); and (iii) grain filling from flowering to physiological maturity (GS3). The number of seeds per panicle is physiologically determined during the second growth stage (GS2) when floret number is set in the developing panicle (Eastin et al., 1999; Maiti and Bidinger, 1981). The yield components of sorghum include the number of panicles per square meter, number of seeds per panicle, and seed weight, which are defined as the first, second, and third yield components, respectively (Maman et al., 2004). As the second yield component directly relates to GS2, any biotic and/or abiotic stress that prevails before or at the onset of this stage could have serious second yield component impacts. Similarly, since seed filling is predominantly associated with GS3, stresses occurring at this stage can impact the third yield component. Therefore, the objective of this study was to determine the effects of multiple stalk rot pathogens and the stage of pathogen infection on the yield and yield components of selected grain sorghum genotypes. We hypothesize that different stalk rot fungi and the growth stage (GS1 and GS3) at which infection occurs have differential effects on sorghum yield, in general, and upon the second and third yield components, in particular. 2. Materials and methods 2.1. Sorghum genotypes and fungal isolates Sorghum genotypes, SC599R (Fusarium stalk rot and charcoal stalk rot resistant, breeding line), BTx3042 (Fusarium stalk rot and charcoal rot susceptible, breeding line), 84G62 (Dupont Pioneer, hybrid), and DKS37-07 (Dekalb, hybrid) were used. Pioneer 84G62 and DKS37-07 are classified as charcoal rot tolerant and are commonly grown hybrids by producers in the state of Kansas, USA. The three Fusarium spp. used in this study, F. thapsinum, F. proliferatum, and F. andiyazi, were previously isolated from infected stalks by the Row Crops Pathology Lab at Kansas State
University from local sorghum fields. A portion of the translation elongation factor (TEF-1a) gene of these isolates was PCR amplified with ef1 (forward) and ef2 (reverse) primers and subsequently sequenced (Geiser et al., 2004). Sequence information was used as a query for comparison with the NCBI and FUSARIUM-ID databases using BLAST algorithm to confirm the species identity. The M. phaseolina isolate (r144) was provided by Dr. Gary Odvody, Texas A&M AgriLife Research and Extension, Corpus Christi, Texas. 2.2. Field experiments Field experiments were carried out during the 2013 and 2014 growing seasons at the Kansas State University agronomy research farm in Manhattan (39.22 N, 96.60 W) and Ashland (39.13 N, 96.62 W), Kansas, respectively (hereafter referred to as the two environments) under rainfed conditions. Field preparation, planting, and crop maintenance were conducted according to standard procedures for sorghum. Seeds were treated with fungicide (ethyl mercaptan (Captan), two mL kg-1 seed) and planted in rows of 5 m length 0.75 m width (whole plot). The treatment structure was a 4 2 5 factorial where factors one, two, and three consisted of four sorghum genotypes, two stages of inoculation, and four pathogens and mock-inoculated control, respectively. The field study was arranged in a split-split-plot design with randomized complete blocks. Genotypes were assigned to the whole-plot unit and stage of inoculation to the sub-plot unit (2.5 m length 0.75 m width). Pathogen and control treatments were assigned to the sub-sub-plot unit (0.5 m length 0.75 m width). There were three subsample plants (¼ observational unit) in each pathogen or mock-inoculated control treatment. For each environment, there were three replicated blocks. 2.3. Inoculum preparation, artificial inoculation, and measuring yield parameters Inoculum preparation was performed according to the methods described by Bandara et al. (2015) and the protocol described therein for F. thapsinum inoculum preparation was also used to prepare F. proliferatum and F. andiyazi inocula. GS1 inoculations were performed at 30 days after seedling emergence for all genotypes when seedlings showed 7 to 10 leaf collars (ligules) (before flag leaf emergence). GS3 inoculations were performed at 14 days after flowering. At both stages, plants were inoculated by injecting 0.1 mL of inoculum at a concentration of 2 106 conidia or mycelium fragments per mL into the basal internode of the stalk. Conidia were used for F. thapsinum, F. proliferatum, and F. andiyazi inoculations. Hyphal fragments were used for M. phaseolina since this fungus does not typically produce spores. Phosphate-buffered saline (PBS) was used for the mock-inoculated control treatment. Inoculated plants were harvested at 110 days after planting (after all four genotypes attained physiological maturity). Panicles were separated from plants. The total number of rachis per panicle (TRP), total seeds per rachis (TSR), and the number of unfilled spikelets per rachis (NUSR) were counted for each panicle. The total number of reproductive sites per rachis (TRSR) in each panicle was computed by adding TSR and NUSR. These data were used to calculate the total seeds per panicle (TSP), number of unfilled spikelets per panicle (NUSP), total number of reproductive sites per panicle [TRSP ¼ NUSP þ TSP], and seed set percentage [SSP ¼ (TSP/ TRSP) 100]. Panicles were then dried for 10 d at 40 C and threshed to measure the total seed weight per panicle (TSW). 100seed weight (100-SW) was computed using the equation, (TSW 100)/TSP. Stems were split longitudinally to measure lesion length (cm) and the number of nodes crossed by the lesion.
Y.M.A.Y. Bandara et al. / Crop Protection 94 (2017) 97e105
2.4. Statistical analysis Data were analyzed for variance (ANOVA) using the PROC GLIMMIX procedure of SAS software version 9.2 (SAS Institute, 2008). The restricted maximum likelihood (REML) method was used to estimate variance components. Degrees of freedom for the denominator in F tests were computed using the Kenward-Roger option. Model assumptions were tested using studentized residual plots (for identical and independent distribution of residuals) and Q-Q plots (for normality of residuals). Whenever residuals were not homogeneously distributed, appropriate heterogeneous variance models were fitted to meet the model assumptions. Bayesian information criterion (BIC) was used to determine the most suitable model that best fit data after accounting for model assumptions. Data analyses were initially conducted for each environment considering the stage of inoculation, genotype, and pathogens as fixed factors. Block was regarded as a random factor. Initial analyses performed within each environment revealed similar significance levels and interaction effects among fixed factors. Therefore, the data were analyzed for combined environments by considering the stage of inoculation, genotype, and pathogens as fixed factors and environment and block as random factors. Blocks were considered as nested within environments for this analysis. Results and conclusions drawn from the combined analysis are presented here. Means separations were carried out using the PROC GLMMIX procedure of SAS. Main effects of factors were determined with adjustments for multiple comparisons using the TukeyKramer test. Simple effects of factors were determined using the Bonferroni adjustment when higher order interactions among factors were statistically significant for the trait concerned. Only relevant pairwise comparisons were considered for determining the critical comparison-wise error rate (aCER) and were subsequently compared against unadjusted P-values resulting from the Bonferroni correction to test the significance of treatment mean differences. Pearson correlation coefficients and significance levels among measured traits were computed using the PROC CORR procedure of SAS. 3. Results 3.1. Analysis of variance (ANOVA) Pathogen and genotype had significant main effects on lesion length while the main effect of genotype was significant on the total number of rachis per panicle. ANOVA did not reveal significant pathogen genotype stage interactions for most traits measured except for total seeds per panicle (Table 1). The two-way interaction between pathogen and stage was significant for total seed weight, 100-seed weight, total seeds per panicle, total number of
99
reproductive sites per panicle, seed set percentage, total seeds per rachis, and total number of reproductive sites per rachis while the interaction between pathogen and genotype was significant for total number of seeds per panicle, total number of reproductive sites per panicle, and total number of reproductive sites per rachis. The genotype stage interaction was significant for total seeds per rachis, and the total number of reproductive sites per rachis. 3.2. Effect of stalk rot pathogens and inoculation stage on disease severity All pathogens significantly increased lesion length and nodes crossed in comparison to control (P < 0.0001; Fig. 3E and F). No significant differences among pathogens were observed. Plant genotype also had significant main effect on lesion length (P ¼ 0.0465; Table 1). 84G62 and DKS37-07 had significantly higher lesion length (P < 0.05) than BTx3042 and SC599. 3.3. Effect of stalk rot pathogens and inoculation stage on seed weights and numbers The mean total seed weight per panicle of the control treatment was significantly greater than the F. andiyazi, F. proliferatum, F. thapsinum, and M. phaseolina treatments for inoculation at both GS1 and GS3 (P < 0.05; Fig. 2A). Inoculations reduced total seed weight per panicle by 42.2 to 57.2% (GS1 inoculation) and 29.4 to 41.2% (GS3 inoculation). The effects of the four pathogens on mean total seed weight per panicle were not significantly different from each other at either of the inoculation stages (Fig. 2A). 100-seed weight values were not significantly decreased from the control by three of the Fusarium pathogens upon GS1 inoculation (P < 0.001; Fig. 2B). M. phaseolina caused an 18.3% reduction in 100-seed weight. Its impact on 100-seed weight was also significantly higher compared to that of F. andiyazi and F. thapsinum (P < 0.001). At GS3 inoculation, all pathogens significantly reduced 100-seed weight compared to control (P < 0.001; Fig. 2B) although the effects of the four pathogens on mean 100-seed weight were not significantly different from each other. On average, the four pathogens reduced 100-seed weight by 16.1% after inoculation at GS3. When inoculated at GS1, all four pathogens significantly decreased the total seeds per panicle of 84G62 compared to the control inoculation (P < 0.001; Fig. 2C). A similar result was observed for DKS37-07 (P < 0.0001; Fig. 2E) for the other three pathogens but not for F. thapsinum inoculation as compared with the control. F. proliferatum caused a significant seed number reduction of SC599 compared to control and M. phaseolina (P < 0.001; Fig. 2F). Pathogens did not significantly reduce seed number of BTx3042 when this genotype was inoculated at GS1
Table 1 P-values of F-statistic from analysis of variance (ANOVA) for disease severity, yield, and yield related traits measured with four sorghum genotypes after inoculation with F. thapsinum, F. andiyazi, F. proliferatum, and M. phaseolina at two plant growth stages, GS1 (growth stage 1, 30 d after emergence); GS3 (growth stage 3, 14 d after flowering) across two field studies (a ¼ 0.05). Effect
LLa
NC
TSW
100-SW
TSP
NUSP
TRSP
SSP
TRP
TSR
NUSR
TRSR
Genotype (G) Stage (S) GS Pathogen (P) PG PS PGS
0.0465 0.6434 0.8008 <0.0001 0.0501 0.6727 0.1740
0.3060 0.6955 0.9957 <0.0001 0.2348 0.5776 0.2174
0.2188 0.2538 0.1600 0.0020 0.3035 0.0195 0.3529
0.0688 0.9180 0.1760 0.0003 0.1831 0.0031 0.0833
0.0379 <0.0001 0.2259 <0.0001 0.0138 0.0106 0.0187
0.1351 0.2448 0.1391 0.3393 0.1016 0.1357 0.1945
0.0022 0.0202 0.7292 0.0001 0.0126 0.0457 0.0592
0.8093 0.1380 0.4150 0.0003 0.3459 0.0043 0.1434
<0.0001 0.1216 0.1592 0.1305 0.1960 0.6211 0.1055
<0.0001 0.0006 0.0090 0.0143 0.0994 0.0185 0.1507
0.0623 0.0012 0.6314 0.4591 0.0760 0.1871 0.1366
<0.0001 0.0002 0.0013 0.0144 0.0071 0.0275 0.4138
a LL ¼ lesion length (cm); NC ¼ number of nodes crossed by the lesion; TSW ¼ total seed weight panicle1 (g); 100-SW ¼ 100-seed weight (g); TSP, NUSP, TRSP ¼ total number of seeds, unfilled spikelets, and total reproductive sites panicle1, respectively; SSP ¼ seed set percentage; TRP ¼ total number of rachis panicle1; TSR, NUSR, TRSR ¼ total seeds, unfilled spikelets, and total number of reproductive sites rachis1.
100
Y.M.A.Y. Bandara et al. / Crop Protection 94 (2017) 97e105
(Fig. 2D). Inoculations at GS3 did not result in seed number reductions for 84G62, BTx3042, or SC599 (Fig. 2C, D, F). However, F. thapsinum reduced total seeds per panicle of DKS37-07 compared to control (Fig. 2E). 3.4. Effect of stalk rot pathogens and inoculation stage on panicle traits The total number of reproductive sites per panicle was significantly decreased by all four pathogens at GS1 inoculation in comparison to control (P < 0.0001; Fig. 3A). No significant differences were observed among pathogens. On average, the four pathogens reduced the total number of reproductive sites per panicle by 32.4%. At GS3, F. thapsinum and M. phaseolina significantly reduced reproductive sites per panicle (P < 0.001; Fig. 3A), although no significant differences between pathogens were observed. F. thapsinum and M. phaseolina reduced reproductive sites per panicle by 24.0% compared to control. All pathogens significantly decreased seed set percentage (P < 0.0001; Fig. 3B) when inoculated at GS1, although pathogens did not significantly differ from one another. On average, the four pathogens caused a 12.0% reduction in seed set compared to control. Inoculations at GS3 did not significantly reduce seed set. The number of rachis per panicle and the number of unfilled spikelets per rachis were not significantly affected by pathogens (Table 1). When compared with control, all four pathogens significantly reduced the total number of reproductive sites per rachis (P < 0.01) and total seeds per rachis (P < 0.005) upon GS1 inoculations (Fig. 3C and D) while no such effect was evident with GS3 inoculations. The number of unfilled spikelets per rachis was significantly different between GS1 (6.75 ± 0.5) and GS3 (5.22 ± 0.5) inoculations (P < 0.005). 3.5. Correlation analysis among measured traits Pearson correlation coefficients were computed for all trait measurements within individual stages of inoculation. Total seeds per panicle, the number of unfilled spikelets per panicle, and seed set percentage had significant negative correlations with lesion length and diseased nodes crossed at GS1 inoculation (Table 2). Nodes crossed was also significantly and negatively correlated with total seed weight per panicle. Analysis of GS3 inoculations revealed significant and negative correlation of nodes crossed with total seed weight per panicle (r ¼ 0.20, P ¼ 0.03), 100-seed weight and total seeds per panicle (r ¼ 0.19, P ¼ 0.04), and total number of reproductive sites per panicle (r ¼ 0.20, P ¼ 0.03) (Table 2). Table 2 Pearson correlation coefficients and significance levels among eight traits measured from four sorghum genotypes inoculated with Fusarium thapsinum, F. andiyazi, F. proliferatum, and Macrophomina phaseolina at GS1 and GS3.
GS1 LL NC GS3 LL NC
TSW
100-SW
TSP
NUSP
TRSP
SSP
0.1567 0.0916 0.1912 0.0389a
0.0292 0.7545 0.0321 0.7312
0.2137 0.0207a 0.2558 0.0054a
0.3040 0.0011a 0.2990 0.0014a
0.0628 0.5126 0.1146 0.2307
0.3548 0.0001a 0.3758 <0.0001a
0.1169 0.2034 0.2006 0.0281a
0.0541 0.5593 0.1925 0.0359a
0.1410 0.1261 0.1925 0.0359a
0.0273 0.7679 0.1193 0.1962
0.1386 0.1324 0.2000 0.0292a
0.0866 0.3490 0.0071 0.9392
a LL ¼ lesion length (cm); NC ¼ nodes crossed by lesion; TSW ¼ total seed weight panicle1 (g); 100-SW ¼ 100-seed weight (g); TSP ¼ total number of seeds panicle1; NUSP ¼ number of unfilled spikelets panicle1; TRSP ¼ total number of reproductive sited panicle1; SSP ¼ seed set percentage. *Pearson correlation coefficients are significant at P < 0.05.
4. Discussion This study was designed to investigate the impacts of stalk rot diseases on the second (number of seeds per panicle) and third (seed weight) yield components of grain sorghum. The number of fertile tillers produced by a plant is a key determinant of the first yield component. Tillering has been shown to be determined by various factors including plant density (Gerik and Neely, 1987; Lefarge et al., 2002), tiller fertility (Lefarge et al., 2002), environment (Kim et al., 2010; Downes, 1968), and seeding pattern (Bandaru et al., 2006). In this study, the first yield component (number of panicles per square meter) was not investigated since stalk rot pathogens do not directly affect this component unless stalk rot-mediated plant lodging is considered as a means of panicle loss under mechanistic harvesting. A primary objective of this study was to determine whether the growth stage at which pathogen infection occurs has a significant impact on grain yield and yield components. The two-way interaction effect between pathogen and stage of inoculation was found to be significant for most traits, indicating inoculation stagespecific impacts of pathogens. When plants were inoculated at GS1, reduced total seeds per panicle contributed more to the reduction of total seed weight than did reduction of 100-seed weight. However, 100-seed weight had a greater impact on total seed weight than the total seeds per panicle for GS3 inoculation. The number of reproductive sites per panicle is physiologically determined during the second growth stage (GS2) when floret number is set in the developing panicle (Eastin et al., 1999; Maiti and Bidinger, 1981). The effect of GS1 inoculations on the total seeds per panicle could be partitioned into two sub-effects. First, pathogens decreased seed set percentage (Fig. 3B) through an impeded grain filling process. Secondly, as pathogens significantly reduced the total number of reproductive sites per panicle (i.e., the total number of seeds and unfilled spikelets) (Fig. 3A), the effect of GS1 inoculations on physiological seed set is apparent. Therefore, decreased numbers of seeds per panicle can be attributed to the pathogens' effect on the physiological development of reproductive sites (spikelets) leading to a smaller sink in comparison to the potential sink size under normal growth conditions. Mu et al. (2005) showed that the size and activity of the early reproductive shoot apex (incipient panicle) determine the number of florets generated on the rice panicle, thus determining sink size. They found the initial size of the reproductive apex as the determinant of the primary branch (rachis) number while cell division activity in subsequent apex growth was the determinant of floret number per rachis. Although a similar study in sorghum has not been reported, this may be a common developmental pathway among most graminaceous crops. The present study provided evidence for the significant reduction of the total number of reproductive sites (florets) per rachis as a result of inoculation with stalk rot pathogens at GS1 while no such effect was evident after GS3 inoculation (Table 1, Fig. 3C). Therefore, our observations along with previous literature suggest the stalk rot pathogens’ capacity to reduce sink size via reduced floret numbers (through reduced apex cell division activity). We did not find evidence for reduced rachis numbers due to pathogen infection (Table 1). However, the size of the reproductive apex as indicated by rachis number appeared to be a characteristic of genotype. Taken together, these results suggest two possible ways that stalk rot pathogens can interfere with seed production in panicles. Since GS1 inoculations had a significant effect on seed set percentage, but a non-significant effect on 100-seed weight (Fig. 2B), the ability of stalk rot pathogens to completely inhibit the filling of some spikelets in the panicle was apparent. The opposite phenomenon was observed with GS3 inoculations and revealed that
Y.M.A.Y. Bandara et al. / Crop Protection 94 (2017) 97e105
pathogens mediated partial inhibition of filling across the spikelets in the panicle. Moreover, the significant correlations between disease severity traits (lesion length and nodes crossed) with percent seed set (negative correlation), no significant correlation with 100seed weight, and significant positive correlation with number of unfilled spikelets per panicle (Table 2) provided evidence for the complete inhibition of some spikelets in the panicle when inoculation was performed at GS1. On the other hand, non-significant correlations between disease severity traits (nodes crossed in this case) and seed set percentage/number of unfilled spikelets per
101
panicle, and significant and negative correlation with 100-seed weight indicated the partial inhibition of grain filling across panicles after GS3 inoculations. Pre-flowering floret abortion (Senanayake et al., 1991) is a major component of grain yield of many cereals (Kernich et al., 1997) and is the third determinant of panicle size, besides primary branch number and the number of florets on primary branches (Yamagishi et al., 2004). In the current study, more aborted florets were observed in GS1 inoculated panicles when compared with mock-inoculated controls (Fig. 1A and B), suggesting that the inoculation of stalk rot pathogens at GS1
Fig. 1. Effects of GS1 inoculation on pre-flowering floret abortion and seed set for (A) the mock-inoculated control treatment and (B) after inoculation with stalk rot pathogens. Arrows indicate aborted florets.
102
Y.M.A.Y. Bandara et al. / Crop Protection 94 (2017) 97e105
Fig. 2. Comparison of mean values (± standard error) among pathogen treatments for (A) total seed weight per panicle (g) and (B) 100-seed weight (g) across environments and genotypes at two inoculation stages (GS1 and GS3). Total number of seeds per panicle for genotypes (C) 84G62, (D) BTx3042, (E) DKS37-07, and (F) SC599 across environments and at GS1 and GS3 using Bonferroni adjustment. Lower case letters compare pathogen effects at GS1 while upper case letters compare effects at GS3. Means followed by the same letters within each inoculation stage and treatments without letter designations are not significantly different based on the adjusted P-value for multiple comparisons at a ¼ 0.05. CON ¼ phosphate-buffered saline mock-inoculated control, FA ¼ Fusarium andiyazi, FP ¼ F. proliferatum, FT ¼ F. thapsinum, and MP ¼ Macrophomina phaseolina.
Y.M.A.Y. Bandara et al. / Crop Protection 94 (2017) 97e105
103
Fig. 3. Comparison of mean values (± standard error) among pathogen treatments for (A) number of reproductive sites per panicle, (B) percent seed set, (C) number of reproductive sites per rachis, and (D) total seeds per rachis across environments and genotypes at two inoculation stages (GS1 and GS3). (E) Lesion length (cm) and (F) number of nodes crossed by the lesion across environments, genotypes, and inoculation stages. Comparisons made in panels A, B, C, and D were based on Bonferroni adjustments while those of E and F were based on the Tukey-Kramer's test (see Materials & Methods). Means followed by the same letters in each case and treatments without letter designations are not significantly different based on the adjusted P-value for multiple comparisons at a ¼ 0.05. CON ¼ phosphate-buffered saline control, FA ¼ F. andiyazi, FP ¼ F. proliferatum, FT ¼ F. thapsinum, and MP ¼ M. phaseolina.
104
Y.M.A.Y. Bandara et al. / Crop Protection 94 (2017) 97e105
leads to an increase in pre-flowering floret abortion. However, GS3 inoculations should not have any impact on pre-flowering floret abortion, as the inoculation was performed at 14 d after flowering. Hence, GS3 inoculation only affects grain filling. Fig. 4 is a generalized schematic representation of the yield determination process under stalk rot infection at GS1 and GS3. The effect of pathogens on lesion length and nodes crossed was not found to be inoculation stage-specific (Table 1; Fig. 3E and F). After GS1 inoculation, one would expect lengthier lesions or more nodes crossed since pathogens exist in plants for a longer duration (~75 d) than GS3 (~35 d). However, the data did not support that expectation. It could be that Fusarium pathogens grow endophytically and cause no visual symptoms during vegetative stages, but cause damage when seed number is being determined during panicle development, which leads to reduced total seed weight per panicle. In fact, some Fusarium species such as F. thapsinum and F. proliferatum can colonize sorghum symptomlessly as endophytes (Leslie, 2000). The two hybrids used in this study, 84G62 and DKS37-07, had significantly greater mean lesion lengths (P < 0.05) compared to the two inbred lines tested, BTx3042 and SC599. This observation agrees with the findings of Das et al. (2008) where they observed higher disease severity with high yielding genotypes. M. phaseolina inoculations resulted in shorter mean lesion lengths and fewer nodes crossed compared to the three Fusarium spp. (Fig. 3E and F). M. phaseolina is a necrotrophic pathogen (Islam et al., 2012) while Fusarium spp. are hemibiotrophic (Bacon and Yates, 2006). Therefore, Fusarium spp. may be more pathogenic than M. phaseolina during the early stages of plant growth while necrotrophs exhibit higher pathogenicity and virulence during the senescent phase. Charcoal rot is problematic under post-flowering stress conditions along with natural plant senescence, especially in the absence of the staygreen (delayed senescence) trait. Except for BTx3042, the other three genotypes tested in this study possess the staygreen trait. Therefore, the shorter lesions and fewer nodes crossed observed in staygreen genotypes in response to M. phaseolina inoculation were not surprising. However, despite the shorter lesion lengths and lower numbers of nodes crossed associated with M. phaseolina, it caused the highest reduction in total seed weight
compared to the three Fusarium spp. at both inoculation stages (GS1, 57.2%; GS3, 41.2%). This finding agrees with earlier reports that M. phaseolina is the most aggressive stalk rot pathogen in sorghum (Anahosur and Patil, 1982; Mughogho and Pande, 1984; Rosenow, 1984; Tesso et al., 2005). Analysis within each inoculation stage showed that there was no significant correlation between charcoal rot disease severity traits with percent yield reduction [(TSWcont e TSWinoc)/TSWcont] of genotypes under infection (GS1, r ¼ þ0.30, P ¼ 0.26; GS3, r ¼ þ0.19, P ¼ 0.47). Percent yield reduction compared to mock-inoculated control is a direct indicator of the genotype's ability to tolerate stalk rot or charcoal rot disease. This emphasizes the importance of incorporating both resistance (e.g., severity traits) and tolerance measures (e.g., yield traits) for screening sorghum genotypes against stalk rot diseases. 5. Conclusions The four stalk rot pathogens each significantly decreased total seed weight per panicle compared to the control treatment at two inoculation stages (GS1 and GS3). A greater yield reduction was observed after GS1 inoculations compared to those at GS3. The reduction of total seed weight per panicle due to GS1 inoculations was predominantly attributed to the reduction in the total number of seeds per panicle while 100-seed weight was found to be the key contributing yield component leading to reduced total seed weight per panicle with GS3 inoculations. Inoculations at GS1 significantly decreased the total number of reproductive sites per panicle, which indicated that the stalk rot pathogens had the ability to directly or indirectly interfere with panicle formation resulting in smaller heads compared to the control. Results further suggested the pathogens’ ability to induce/enhance pre-flowering floret abortion upon GS1 inoculation, contributing to reduced total seeds per panicle. Therefore, plant breeding and production research should consider both of these yield components to achieve enhanced sorghum production. Seed number merits greater attention where sorghum cultivation potentially undergoes pre-flowering stress conditions such as drought that predispose plants to natural infection by stalk rot pathogens. However, seed weight should be targeted where the crop is more vulnerable to post-flowering stress
Fig. 4. Schematic representation of yield determination for Sorghum bicolor L. Moench after stalk rot inoculations.
Y.M.A.Y. Bandara et al. / Crop Protection 94 (2017) 97e105
conditions. In general, Fusarium stalk rot was more severe compared to charcoal rot, although charcoal rot caused greater yield reduction at both inoculation stages. Significant correlations were evident between disease severity traits and yield related traits although correlation coefficients were not strong. There were no significant correlations between disease severity traits such as lesion length and % yield reduction. These finding suggested that disease severity measurements are inadequate as sole criteria in screening germplasm for stalk rot disease resistance. Our results emphasize the importance of a stalk rot screening approach in which “tolerance” should be an integral component. Acknowledgments The Kansas Grain Sorghum Commission is gratefully acknowledged for their financial support of this research. The authors also wish to thank Kyle Shroyer, Chad Brady, and Dr. Kraig Roozeboom for their technical assistance and advice during this study. This paper is Contribution No. 14-344-J from the Kansas Agricultural Experiment Station, Manhattan, Kansas. References Anahosur, K.H., Patil, S.H., 1982. Some promising sources of resistance to charcoal rot of sorghum. Sorghum Newsl. 25, 118. Anahosur, K.H., Patil, S.H., 1983. Assessment of losses in sorghum seed weight due to charcoal rot. Indian Phytopathol. 36, 85e88. Bacon, C.W., Yates, I.E., 2006. Endophytic Root colonization by Fusarium species: histology, plant interactions, and toxicity. In: Schulz, B.J.E., Boyle, C.J.C., Sieber, T.N. (Eds.), Microbial Root Endophytes. Springer, pp. 133e152. Bandara, Y.M.A.Y., Weerasooriya, D.K., Tesso, T.T., Little, C.R., 2016a. Stalk rot diseases impact sweet sorghum biofuel traits. BioEnergy Res. http://dx.doi.org/ 10.1007/s12155-016-9775-6. Bandara, Y.M.A.Y., Weerasooriya, D.K., Tesso, T.T., Little, C.R., 2016b. Stalk rot fungi affect leaf greenness (SPAD) of grain sorghum in a genotype- and growth stagespecific manner. Plant Dis. 100 (10), 2062e2068. Bandara, Y.M.A.Y., Perumal, R., Little, C.R., 2015. Integrating resistance and tolerance for improved evaluation of sorghum lines against Fusarium stalk rot and charcoal rot. Phytoparasitica 43, 485e499. Bandaru, V., Stewart, B.A., Baumhardt, R.L., Ambati, S., Robinson, C.A., Schlegel, A., 2006. Growing dryland grain sorghum in clumps to reduce vegetative growth and increase yield. Agron. J. 98 (4), 1109e1120. Das, I.K., Prabhakar, Indira, S., 2008. Role of stalk-anatomy and yield parameters in development of charcoal rot caused by Macrophomina phaseolina in winter sorghum. Phytoparasitica 36, 199e208. Dodd, J.L., 1980. The photosynthetic stress translocation balance concept of sorghum stalk rot, 11-15 December 1978. In: Proceedings of International Workshop of Sorghum Diseases. Texas A&M University/ICRISAT Press, Hyderabad, India, pp. 300e305. Downes, R.W., 1968. The effect of temperature on tillering of grain sorghum seedlings. Crop Pasture Sci. 19 (1), 59e64. Eastin, J.D., Sullivan, C.Y., 1974. Yield consideration in selected cereals. In: Bielski, R.L., et al. (Eds.), Mechanisms of Regulation of Plant Growth. Bull. 12. Royal Soc. of New Zealand, Wellington, New Zealand, pp. 871e877. Eastin, J.D., Petersen, C.L., Zavala-Garcia, F., Dhopte, A., Verma, P.K., Ounguela, V.B., Wit, M.W., Hernandez, V.G., Munoz, M.L., Gerik, T.J., Gandoul, G.I., Hovney, M.R.A., Onofre, L.M., 1999. Potential heterosis associated with developmental and metabolic processes in sorghum and maize. In: Coors, J.G., Pandey, S. (Eds.), The Genetics and Exploitation of Heterosis in Crops. ASA, CSSA, and SSSA, Madison, WI, pp. 205e229. Geiser, D.M., Gasco, M.D.M.J., Kang, S., Makalowska, I., Veeraraghavan, N., Ward, T.J., Zhang, N., Kuldau1, G.A., O'Donnell, K., 2004. FUSARIUM-ID v. 1.0: a DNA sequence database for identifying Fusarium. Eur. J. Plant Pathol. 110, 473e479. Gerik, T.J., Neely, C.L., 1987. Plant density effects on main culm and tiller development of grain sorghum. Crop Sci. 27 (6), 1225e1230. Hundekar, A.R., Anahosur, K.H., 1994. Pathogenicity of fungi associated with sorghum stalk rot. Karnataka J. Agric. Sci. 7, 291e295. Islam, M.S., Haque, M.S., Islam, M.M., Mannan, E., Emdad, E.M., Halim, A., Hossen1, Q.M.M., Hossain, M.Z., Ahmed, B., Rahim, S., Rahman, M.S., Alam, M.M., Hou, S., Wan, X., Saito, J.A., Alam, M., 2012. Tools to kill: genome of one of the most destructive plant pathogenic fungi Macrophomina phaseolina. BMC Genomics 13, 493. http://dx.doi.org/10.1186/1471-2164-13-493.
105
Jardine, D.J., Leslie, J.F., 1992. Aggressiveness of Gibberella fujikuroi (Fusarium monliforme) isolates to grain sorghum under greenhouse conditions. Plant Dis. 76, 897e900. Kernich, G.C., Halloran, G.N., Flood, R.G., 1997. Variation in duration of pre-anthesis phases of development in barley (Hordeum vulgare). Aust. J. Agric. Res. 48, 59e66. Khune, N.N., Kurhekar, D.E., Raut, J.G., Wangikar, P.D., 1984. Stalk rot of sorghum caused by Fusarium moniliforme. Indian Phytopathol. 37, 316e319. Kim, H.K., van Oosterom, E., Dingkuhn, M., Luquet, D., Hammer, G., 2010. Regulation of tillering in sorghum: environmental effects. Ann. Bot. 106 (1), 57e67. Lefarge, T.A., Broad, I.J., Hammer, G.L., 2002. Tillering in grain sorghum over a wide range of population densities: identification of a common hierarchy for tiller emergence, leaf area development and fertility. Ann. Bot. 90 (1), 87e98. Leslie, J.F., 2000. Fusarium species associated with sorghum. In: Fredericksen, R.A., Odvody, G.N. (Eds.), Compendium of Sorghum Diseases, 2nd Ed. APS Press, St Paul, USA, pp. 30e31. Leslie, J.F., Zeller, K.A., Lamprecht, S.C., Rheeder, J.P., Marasas, F.O., 2005. Toxicity, pathogenicity and genetic differentiation of five species of Fusarium from sorghum and millet. Phytopathology 95, 275e283. Maiti, R.K., Bidinger, F.R., 1981. Growth and development of pearl millet plant (Hyderabad, India). ICRISAT Res. Bull. 6. Maman, N., Mason, S.C., Lyon, D.J., Dhungana, P., 2004. Yield components of pearl millet and grain sorghum across environments in the central Great Plains. Crop Sci. 44, 2138e2145. Mu, C., Nemoto, K., You, Z., Yamagishi, J., 2005. Size and activity of shoot apical meristems as determinants of floret number in rice panicles. Plant Prod. Sci. 8, 51e59. Mughogho, L.K., Pande, S., 1984. Charcoal rot of sorghum. In: Rosenberg, G. (Ed.), Sorghum Root and Stalk Rots, a Critical Review. Proceedings of the Consultative Group Discussion on Research Needs and Strategies for Control of Root and Stalk Rot Diseases. 27 Nov - 2 Dec 1983. ICRISAT Press, Bellagio, Italy. Patancheru, A.P. 502324, India, pp. 11e24. Reed, J.E., Partridge, J.E., Nordquist, P.T., 1983. Fungal colonization of stalks and roots of grain sorghum during the growing season. Plant Dis. 64, 417e420. Rosenow, D.T., 1984. Breeding for resistance to root and stalk rots in Texas. In: Rosenberg, G. (Ed.), Sorghum Root and Stalk Rots, a Critical Review. Proceedings of the Consultative Group Discussion on Research Needs and Strategies for Control of Root and Stalk Rot Diseases. 27 Nov - 2 Dec 1983. ICRISAT Press, Bellagio, Italy. Patancheru, A.P. 502324, India, pp. 209e217. SAS Institute, 2008. The SAS Users Guide, Version 9.2. SAS Inst., Cary, NC. Seetharama, N., Bidinger, F.R., Rao, K.N., Gill, K.S., Mulgund, M., 1987. Effect of pattern and severity of moisture deficit stress on stalk rot incidence in sorghum. I. Use of line source irrigation technique, and the effect of time of inoculation. Field Crops Res. 15, 289e308. Seetharama, N., Sachan, R.C., Huda, A.K.S., Gill, K.S., Rao, K.N., Bidinger, F.R., Reddy, D.M., 1991. Effect of pattern and severity of moisture-deficit stress on stalk-rot incidence in sorghum. II. Effect of source/sink relationships. Field Crops Res. 26, 355e374. Senanayake, N., De Datta, S.K., Naylor, R.E.L., Tomson, W.J., 1991. Lowland rice apical development: stages and cultivar differences detected by electron microscopy. Agron J. 83, 1013e1023. Tarr, S.A.J., 1962. Root and Stalk Diseases: Red Stalk Rot, Colletotrichum Rot, Anthracnose, and Red Leaf Spot, in: Diseases of Sorghum, Sudan Grass and Brown Corn. Commonwealth Mycological Institute, Kew, Surrey, UK, pp. 58e73. Tesso, T.T., Claflin, L.E., Tuinstra, M.R., 2005. Analysis of stalk rot resistance and genetic diversity among drought tolerant sorghum genotypes. Crop Sci. 45, 645e652. Tesso, T.T., Ochanda, N., Little, C.R., Claflin, L., Tuinstra, M.R., 2010. Analysis of host plant resistance to multiple Fusarium species associated with stalk rot disease in sorghum [Sorghum bicolor (L.) Moench]. Field Crops Res. 118, 177e182. Tesso, T., Ejeta, G., 2011. Stalk strength and reaction to infection by Macrophomina phaseolina of brown midrib maize (Zea mays) and sorghum (Sorghum bicolor). Field Crops Res. 120, 271e275. Tesso, T., Perumal, R., Little, C.R., Adeyanju, A., Radwan, G.L., Prom, L.K., Magill, C.W., 2012. Sorghum pathology and biotechnology - a fungal disease perspective: Part II. Anthracnose, stalk rot, and downy mildew. Eur. J. Plant Sci. Biotechnol. 6, 31e44. Yamagishi, J., Miyamoto, N., Hirotsu, S., Laza, R.C., Nemoto, K., 2004. QTLs for branching, floret formation, and pre-flowering floret abortion of rice panicle in a temperate japonica tropical japonica cross. Theor. Appl. Genet. 109, 1555e1561. Zummo, N., 1980. Fusarium disease complex of sorghum in West Africa. In: Proceedings of the International Workshop on Sorghum Diseases. ICRISAT Press, Hyderabad, India, pp. 11e15. Zummo, N., 1984. Fusarium root and stalk disease complex. In: Root, Sorghum, Stalk Rots, A. (Eds.), Critical Review, Proceedings of the Consultative Group Discussion on Research Needs and Strategies for Control of Sorghum Root and Stalk Rot Diseases. ICRISAT Press, Bellagio, Italy, pp. 25e29.