SEARCH

SEARCH BY CITATION

Keywords:

  • agronomic management;
  • conservation agriculture;
  • epidemiology;
  • Magnaporthe oryzae ;
  • rice blast;
  • upland rice

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

In response to the extensive development of upland rice on the hillsides of the Madagascan highlands, alternative cropping systems based on conservation agriculture have been recommended to halt declining soil fertility and to limit erosion. To assess the efficiency of these cropping systems in limiting rice blast disease and to measure their yield performance, an experiment was set up in 2003 at Andranomanelatra (1640 m a.s.l.) in the Madagascan highlands. The rice crop was planted every second year following oat (Avena sativa) after common bean (Phaseolus vulgaris), with both conventional tillage and no tillage. For each cropping system, two levels of fertilization were used: (i) organic and (ii) organic + mineral fertilization. The level of blast epidemic was measured on two different cultivars over a 5-year period. Disease severity was significantly lower in the no-tillage cropping system than in the conventional tillage system. Mineral fertilization increased the level of blast. A significant interaction between cropping system and fertilization indicated that the impact of fertilization differed with the cropping system. When the level of blast was low, yield was higher in the conventional cropping system but as soon as blast level increased, yield was better in the no-tillage cropping system. Possible interactions between cropping system and blast epidemics are explained and the problem of high performance but risky cropping systems is discussed.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Rice blast, caused by Magnaporthe oryzae, is the most serious fungal disease of rice crops in the world. In the field, it can develop on every part of the plant. During the vegetative phase, blast causes necrotic lesions on leaves. Severe attacks reduce the photosynthetic leaf area (Bastiaans, 1991) leading to a decrease in yield and even to the death of the plant when it occurs very early. After heading, the fungus can develop on the panicle stem and cause necrosis of the panicle node (neck blast) or of spikelets, which prevents grain filling and sometimes leads to complete sterility (white panicle). Panicle blast is the most damaging form of the disease and yield loss sometimes reaches 100% at the farmer field scale (Pennisi, 2010).

In Madagascar's central highlands, a densely populated area, increasing land pressure in lowland areas has made upland rice a great opportunity to improve rice production. Cold tolerant cultivars have been developed since the 1980s and are now extensively cropped (Dzido et al., 2004). The first set of varieties adapted to highland climatic conditions had a fairly good yield potential (over 6 t ha−1) but a narrow genetic base (Raboin et al., 2011). Blast fungus consequently rapidly adapted to these varieties, which became highly susceptible to infection. Moreover, the cool humid conditions of the highlands and the upland rice cropping systems favour blast epidemics (Lai et al., 1999). As farmers cannot afford chemical treatments, upland rice production now needs to be improved without increasing the risk of blast.

Integrated pest and disease management is a challenge worldwide. Some ways to increase the resilience of the cropping systems were reviewed in Ratnadass et al. (2011). These authors focused on introducing plant species diversification in agroecosystems at different scales. Diversification can modify the pathogen's life cycle with the introduction of non-host or repulsive plants in the cropping system or improve crop tolerance through better crop nutrition. Strategies for intraspecific diversification are also well documented (Mundt, 2002). In rice, they can cause a spectacular reduction in blast disease when a susceptible cultivar is cropped in association with a resistant one in irrigated conditions (Zhu et al., 2000) and in upland rainfed conditions (Castilla et al., 2010; Raboin et al., 2012). Other cropping practices have been shown to modify disease pressure including changing tillage intensity, which is often used to manage soilborne pathogens but also residue-borne pathogens (Fernandez et al., 1999), fertilization or specific nutrient inputs (nitrogen, silicon, phosphorus, etc.). These practices and their impact vary with the crop and the pathogen and need to be combined to maximize their potential ability to control the disease (Walters, 2009).

Conservation agriculture was introduced in Madagascar to limit erosion and to improve the resistance of the upland crops to climatic risks (Husson et al., 2010). The term ‘conservation agriculture’ covers a wide range of techniques all based on three principles defined by the FAO (2010; Scopel et al., 2013): minimum soil disturbance, permanent soil cover using crop residues or growing plants, and crop rotations. The conservation agriculture cropping systems introduced in the highlands of Madagascar for upland rice led to many modifications in cultural practices that could interact with the blast life cycle and with rice susceptibility to the disease. Permanent soil cover, no-tillage or intercropping can modify rice nutrition, the crop microclimate and plant density, which are determining factors in blast epidemics. A 5-year experimental study was conducted to assess the impact of these cropping systems on blast severity and its impact on upland rice yield.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Location

The experiment was conducted at Andranomanelatra (19°47′ S, 47°06′ E, 1640 m a.s.l.) in the Vakinankaratra region of the highlands of Madagascar. The tropical altitude climate is characterized by cool dry winters and warm humid summers. Average annual rainfall is 1460 mm. The wettest months are December and January. Mean temperatures range from 18°C in October, the beginning of the rice sowing period, to 20°C during the reproductive stage. Minimum temperatures can be under 10°C during the early vegetative stage and are below 15°C during the reproductive and grain filling stages. The night/day thermal amplitude is high (10–12°C) throughout the rice growing season. The soil is a ferralsol (Razafimbelo et al., 2006) according to the FAO classification (FAO, 2006).

Experimental design

A long-term experiment comparing conservation agriculture systems with conventional ones has been underway since 2003. As part of this experiment, two cropping systems were compared (Table 1): a 2-year rotation under conventional tillage (CT) and no-tillage soil management (NT). The rotation was upland rice (Oryza sativa) the first year, followed by common bean (Phaseolus vulgaris) the second year. After the common bean was harvested, oat (Avena sativa) was sown. In the conventional tillage system, most of the rotation crop residues were removed, whereas in the no-tillage system, a mulch made of the rotation crop residues was left on top of the soil. Each season, the two crops in the rotation, i.e. rice and common bean, were present in the experiment. Under each cropping system, two levels of fertilization were used: ‘OF’, organic fertilizer (cattle manure at 5 t ha−1), which corresponds to the conventional fertilization level used by farmers; and ‘MF’, organic fertilizer plus NPK mineral fertilizer (11% N, 22% P2O5, 16% K2O) at 300 kg ha−1, dolomite (CaMg (CO3)2) at 500 kg ha−1, cattle manure at 5 t ha−1 and two top-dressings with urea (46% N) at 50 kg ha−1 applied 30 and 70 days after sowing. The second fertilization is the recommended fertilization level for upland rice by Malagasy Research Institute of Rural Development (FOFIFA), but it is still rarely applied by farmers.

Table 1. Description of experimental treatments showing crop sequences and soil management methods
PlotCrop in year nCrop in year + 1Soil management
Novembera–MayNovember–May
  1. a

    Sowing dates of rice: 8–11 November 2004, 7–11 November 2005, 13–17 November 2006, 11–19 November 2008 and 15–23 November 2010.

Conventional tillage (CT)RiceBean + oatTillage, residues removed
No tillage (NT)RiceBean + oatNo tillage, residues left on the soil
Organic fertilization (OF)Cattle manure at 5 t ha−1Cattle manure at 5 t ha−1 
Mineral fertilization (MF)Cattle manure at 5 t ha−1, NPK at 300 kg ha−1, dolomite at 500 kg ha−1Cattle manure at 5 t ha−1, NPK at 300 kg ha−1, dolomite at 500 kg ha−1 

Six to eight rice seeds were sown manually in hills at a spacing of 20 × 20 cm (25 hills m−2). Final plant populations, counted at harvest, varied from 70 to 160 plants m−2. Weeds were controlled by hand, and protection against insects and fungi was provided with a seed treatment (Gaucho: 35% imidacloprid + 10% thiram at 2·5 g kg−1 of seeds). The experimental design comprised a split plot with four replications to compare the two cropping systems as main plots, and the two fertilization levels as subplots, with a cultivar susceptible to blast disease in the local conditions (Fofifa 154) and a partially resistant cultivar (Fofifa 161). The study took place over the period 2005–2011. The two cultivars were cropped in 2005, 2006 and 2007. In the following years, only the partially resistant cultivar was cropped, except in 2010 when it was replaced by a resistant cultivar. Blast was measured on both cultivars in 2005 and 2006, but only on the susceptible cultivar in 2007 and only on the partially resistant cultivar in 2009 and 2011.

Blast assessment

Blast infection occurred naturally in the experiments. Disease severity was estimated on leaves and panicles on 10 hills per plot. Leaf blast severity was assessed as follows: on each hill, the total number of stems and the number of stems with at least one susceptible symptom were counted. On three diseased stems, the leaf area affected by blast symptoms on the four upper leaves was estimated. In 2005 and 2006, the same estimation was made by using the IRRI scale (IRRI International Rice Research, 2009) and each score was transformed with the corresponding mean percentage of area. After 2006, the percentage of leaf area affected by blast was estimated directly.

Leaf severity was calculated by estimating the leaf area presenting blast symptoms in one plot. The value of leaf severity of the plot is the mean of the values of the 10 hills:

  • display math

After flowering, only panicle blast symptoms were evaluated. On 10 hills, the numbers of healthy and diseased panicles were counted. On five diseased panicles, the percentage of diseased spikelets (i.e. those with a black stem) was visually estimated. When the neck nodes are infected by the fungus, the panicle dries and becomes white (counted as 100% of spikelets affected).

Panicle blast severity was the estimated percentage of grains attacked by blast in one plot. It was obtained by calculating the mean of the severity values for the 10 hills of the plot:

Panicle severity = mean (% diseased spikelets on diseased panicles × the number of diseased panicles/total number of panicles).

In 2005, the susceptible cultivar was checked for blast weekly to identify the pattern of the dynamics of the epidemics. This included five checks for leaf symptoms and five checks for panicle blast. In the other cases, checks were made once or twice for leaf blast and once or twice for panicle blast. Here the results of one notation at each stage of each year are presented.

Grain yield and yield components

In all 5 years, grain yield (using unhulled seeds, after drying at 60°C for 3 days), yield components, and straw dry weight were measured at maturity. Grain and straw yield were measured at harvest in a 4 m2 area located in the centre of each plot. The number of plants, panicles, and spikelets per panicle, the number of filled and unfilled spikelets, and 1000-grain weight were determined on a subplot of eight hills located at the edge of the 4 m² area. All plant samples were separated into straw and panicles. Panicles were counted and hand threshed, and filled and unfilled spikelets were separated. Dry weights of filled spikelets and unfilled spikelets were determined after oven drying. Subsamples of 200 filled spikelets and 200 unfilled spikelets were weighed to calculate the total number of filled and unfilled spikelets and 1000-grain weight. The number of spikelets per panicle and the grain filling percentage (100 × number of filled spikelets/total number of spikelets) were calculated.

Analysis

Statistical analyses were performed with sas v. 9.2 (SAS Institute Inc.). Means and standard errors were generated with proc means. Disease severity values were transformed using the arcsin of the square root transformation. The mixed procedure (with Random statement) was used for analysis of variance. Stepwise regression was performed with proc reg. Analysis of correlations was performed with SigmaPlot v. 11.0 (Systat Software, Inc.).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Initial study of blast dynamics (2005)

Leaf blast and panicle blast dynamics were studied from the appearance of the very first lesions to the end of the cropping season on the susceptible cultivar Fofifa 154 (F154) in 2005. Leaf blast dynamics differed between cropping systems (Fig. 1a). From the first observation on, the blast level was higher in the conventional tillage system with mineral fertilization than in the other treatments. The effect of the cropping system was significant on the five dates, and fertilization had a significant effect on the four first dates. The highest level of blast was found in the conventional tillage system with mineral fertilization. The no-tillage system with mineral fertilization (NT–MF) was subject to the same level of attack as the conventional tillage system without mineral fertilization (CT–OF). Blast level sometimes appeared to decrease as a result of the appearance of new leaves. Based on leaf blast dynamics, it was determined that the best dates for blast assessment were between 85 and 92 days after sowing (3 and 10 February 2005) because blast was at its highest level, before the decrease that was observed at the last observation date.

image

Figure 1. (a) Variations in leaf blast severity (mean % leaf area with symptoms) and (b) comparison of panicle blast severity in the four treatments: conventional tillage (CT) or no-tillage (NT), and with organic + mineral fertilization (MF) or organic fertilization only (OF) in 2005. Letters indicate significant impacts (CS, cropping system; F, fertilization) according to the statistical analysis, date by date.

Download figure to PowerPoint

Panicle blast dynamics (Fig. 1b) showed a significant effect of fertilization only at the second date. In the other cases, only the effect of the cropping system was significant, with a decrease of between 3 and 29% in blast severity in the no-tillage system compared to the conventional system.

Multiple year validation

Cropping systems were compared for three or four years, depending on the variety. The cropping system and fertilization had significant effects on the susceptible cultivar (F154) each year (Fig. 2; Table 2), with a decrease in the severity of leaf and panicle blast in the no-tillage without mineral fertilization. In 2006, this cropping system had the most spectacular impact, with a decrease from 86% in the CT–MF condition to 20% in the NT–MF condition and from 49% in the CT–OF condition to 16% in the NT–OF condition. In 2007, blast pressure appeared to be lower on the susceptible cultivar but the effect of the cropping system was still significant on leaves and panicles. In 2006 and 2007, the effect of the interaction between cropping system and fertilization was also significant, suggesting that the impact of fertilization, which is known to increase blast severity (Ou, 1985), was modified in the no-tillage cropping system.

Table 2. Statistical effects of cropping system (CS), fertilization (Fert) and interactions on leaf and panicle blast. Values are the results of the mixed analysis (P > F). Values in bold italics represent significant effects (< 0·05). Statistical analysis was performed for each cultivar separately (a), or with both cultivars when possible (b)
YearCultivarDateLeaf/panicle blastCSFertCS × Fert
(a) Cultivar by cultivar analysis
2004–2005F15410 February 2005Leaf 0·0023 0·0242 0·1585
17 March 2005Panicle 0·0006 0·07460·9803
2005–2006F15409 February 2006Leaf 0·0039 0·0365 0·0314
30 March 2006Panicle <0.0001 0·0065 0·0224
2006–2007F15422 February 2007Leaf 0·0447 0·0009 0·5797
27 March 2007Panicle 0·0177 <0·0001 0·0220
2004–2005F16118 February 05Leaf0·0755 0·0070 0·4769
30 March 05Panicle0·1756 0·0336 0·8171
2005–2006F16120 February 2006Leaf0·38460·80390·6355
30 March 2006Panicle0·0517 0·0042 0·2149
2008–2009F16103 February 2009Leaf0·23260·20230·5197
26 March 2009Panicle 0·0076 <0·0001 0·0005
2010–2011F16117 February 2011Leaf 0·0439 0·17310·5027
27 April 2011Panicle 0·0015 0·0083 0·0401
YearLeaf/panicle blastCSFertCultivarCS × Fert
(b) Two-cultivars analysis
2004–2005Leaf 0·0006 0·0037 <0·0001 0·1627
Panicle 0·0002 0·1542 <0·0001 0·7032
2005–2006Leaf 0·0103 0·0979 <0·0001 0·1445
Panicle <0·0001 0·0159 <0·0001 0·0963
image

Figure 2. Leaf (a) and panicle (b) blast severity of the susceptible cultivar Fofifa 154 in 2005, 2006 and 2007 in the two cropping systems (CT, conventional tillage; NT, no-tillage) and two fertilization treatments (MF, organic + mineral fertilization; OF, organic fertilization only). Letters indicate if the cropping system (CS), the fertilization (F) and the interaction (CS × F) have a significant effect at < 0·05 each year. Bars represent the means (± SE) of four plots.

Download figure to PowerPoint

The different treatments had little effect on leaf blast in the partially resistant cultivar F161 (Fig. 3; Table 2). In 2005, mineral fertilization significantly increased leaf blast. In 2011, no-tillage systems were less attacked than conventional systems. Panicle blast severity was much lower in the partially resistant cultivar F161 in 2005 and 2006 than in the F154 cultivar (<8% of diseased spikelets) but severity increased in 2009 and 2011. In these 2 years, the effects of cropping system, fertilization and their interaction were significant (Table 2), with a higher severity in the conventional tillage system with mineral fertilization.

image

Figure 3. Leaf (a) and panicle (b) blast severity for the partially resistant cultivar Fofifa 161 in the two cropping systems (CT, conventional tillage; NT, no-tillage) and two fertilization treatments (MF, organic + mineral fertilization; OF, organic fertilization only). Letters indicate if the cropping system (CS), the fertilization (F) and the interaction (CS × F) have a significant impact at < 0·05 each year. Bars represent the means (± SE) of four plots.

Download figure to PowerPoint

Blast level and grain filling

Combined for all years and all cultivars, grain filling and panicle blast severity were closely correlated with an exponential decay equation (Fig. 4; f = 83·734 exp (−0·027x), R² = 0·892). This relationship confirmed that the measure of blast was a good indicator of disease severity and confirmed the direct impact of panicle blast on grain filling. The relationship gave a percentage of 83·7% of filled grain for zero blast attack corresponding to around 12% (between 4 and 40%) grain sterility due to other causes. A one-for-one correlation (one diseased spikelet induces one unfilled grain) would be represented by a linear regression. The type of curve obtained here indicated that grain filling was usually under the rate predicted by a simple linear equation. Panicle blast thus directly and negatively impacted grain filling but was not the only cause of the reduction in grain filling. N fertilization has already been shown to decrease cold tolerance and increase grain sterility (Gunawardena & Fukai, 2005) and may be one reason for the decrease in fertility in the current study.

image

Figure 4. Relationship between panicle blast and grain filling. Dots represent the mean values per cropping system, fertilization, cultivar and year. The curve corresponds to the exponential decay equation f = 83·734 exp (−0·027x). R² = 0·892.

Download figure to PowerPoint

Yield

Due to the severe damage caused by blast, yields of the susceptible cultivar F154 were very low each year (maximum of 1·37 t ha−1 in 2007; Table 3a). Variation in grain yields was closely associated with variations in grain filling (for the ‘grain filling’ component, stepwise regression performed on yield gave a partial R² = 0·84 in 2005 and 0·85 in 2006). Nevertheless, the significant effects of the cropping system and of the fertilization on blast severity were not the same on yield or on the percentage of grain filling, perhaps as a result of the unexplained unfilled grains mentioned above (Fig. 5). In F161 (Table 3b), the grain yield was much higher the first year. Significantly higher yields were obtained with conventional tillage in 2005 and 2006. In 2009, the percentage of grain filling was better without mineral fertilization and lower in the conventional tillage system with mineral fertilization (significant effects of fertilization and of the interaction between cropping system and fertilization). In 2006 and 2011, only fertilization and the interaction with the cropping system had a significant effect on the percentage of grain filling, with the lowest level in the conventional tillage system without mineral fertilization. Variations in the percentage of grain filling explained variations in yield better in 2011 than in 2009 (partial R² = 0·61 and 0·27, respectively). When blast level was low, yield appeared to be higher in the conventional cropping system with mineral fertilization. However, blast was higher in the partially resistant cultivar in 2009 and 2011, after which it had more impact on grain filling and final yield.

Table 3. Yield and percentage of grain filling for each year and results of the statistical tests. Means are given per cropping system (CS), per fertilization treatment (Fert) and per cropping system × fertilization. An analysis combining all years was performed for each cultivar. Results of the statistical analysis of variance are given. Values in bold show significant results (< 0·05)
Treatmenta2005 2006 2007 All years
Rice yield (t ha−1)Grain filling (%)Rice yield (t ha−1)Grain filling (%)Rice yield (t ha−1)Grain filling (%)Rice yield (t ha−1)Grain filling (%)
(a) Cultivar F154
Cropping system
CT0·2314·00·4319·61·1653·30·6129·0
NT0·9232·40·6734·71·3760·60·9942·5
Fertilization
MF0·3520·30·5626·71·2349·40·7132·1
OF0·8026·10·5527·61·30 64·5 0·8839·4
Source of variation
CS0·26510·27050·29390·07690·35200·37620·23900·1787
Fert0·05440·35760·95850·92800·8261 0·0325 0·37590·2598
CS × Fert0·26170·61440·32570·67320·30440·43310·53690·7593
Stepwise regression on yield
Partial R2 0·8360 0·8525 0·3940 0·7273
Treatmenta20052006 20092011 All years
Rice yield (t ha−1)Grain filling (%)Rice yield (t ha−1)Grain filling (%)Rice yield (t ha−1)Grain filling (%)Rice yield (t ha−1)Grain filling (%)Rice Yield (t ha−1)Grain filling (%)
  1. a

    CT, conventional tillage; NT, no-tillage; MF, organic + mineral fertilization; OF, organic fertilization only.

(b) Cultivar F161
Cropping system
CT 5·82 90·6 3·51 85·02·7766·42·2156·63·5874·7
NT4·9491·82·1188·6 3·68 76·0 3·5772·43·5783·1
Fertilization
MF6·0390·42·9885·23·2360·12·6753·53·7372·3
OF4·7392·12·64 88·4 3·22 82·3 3·12 79·2 3·43 85·5
CS × Fert
CT MF6·5489·73·6981·42·6348·91·7537·33·6564·3
CT OF5·0991·53·3488·62·91 83·9 2·68 75·8 3·51 85·0
NT MF5·5191·02·2888·93·84 71·2 3·58 69·8 3·80 80·2
NT OF4·3692·61·9488·33·53 80·7 3·55 82·6 3·34 86·0
Source of variation
CS 0·0179 0·3265 0·0259 0·11610·0053 0·0491 0·05190·10970·97550 .0683
Fert0·08230·06670·4378 0·0294 0·9753 0·0049 0·4383 0·0020 0·4449 0·0010
CS × Fert0·82320·92530·9901 0·0138 0·5432 0·0480 0·4123 0·0410 0·6925 0·0554
Stepwise regression on yield
Partial R2   0·2734 0·6067 0·2359
image

Figure 5. Relationship between panicle blast and yield according to cultivar (F154, F161), cropping system and fertilization level. Dots represent the mean values per system, fertilization, cultivar and year. CT, conventional tillage; NT, no-tillage; MF, organic + mineral fertilization; OF, organic fertilization only.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

In these experiments, blast pressure was never a limiting factor. The levels of blast severity obtained were sometimes the highest possible, close to 90% of grain sterility due to the disease. This confirms that blast is the most dangerous disease of rice crops in the region and can sometimes cause major yield losses in farmers' fields. Nevertheless, the principle of these experiments, based on a randomized design with four replications and rice fields, sometimes with a high level of nitrogen fertilization, and with or without soil tillage, could maintain artificial blast pressure. Farmers rarely use mineral fertilizer (the use of mineral fertilizer is uniformly low in rural Madagascar because of its excessive cost; Minten et al., 2007) and they rarely leave rice residues in their fields as they are used to feed cattle. So, although still causing important yield losses, blast incidence and severity in farmers' fields may be lower. However, the high level of blast in this trial made it possible to measure slight differences in blast, which is difficult when intrafield variations are high, and to forecast the situation that could appear in the case of intensification of upland rice cropping systems.

The analysis of blast dynamics revealed a decrease in blast severity in the no-tillage system compared to the conventional cropping system and an increase in severity with the addition of mineral fertilization whatever the measurement date. The same results were obtained in the multiple year validation in both the susceptible and the partially resistant cultivars. Both the cropping system and fertilization had a significant impact on leaf and panicle blast in the susceptible cultivar each year. Blast severity decreased in the no-tillage cropping system and with lower inputs (organic fertilization only). The effects on leaf blast were rarely significant in the partially resistant cultivar, F161, but fertilization and the cropping system had a significant impact on panicle blast the last two years, when blast pressure was highest. In the two last years, blast was more severe in the conventional system than in the no-tillage system. A highly susceptible cultivar with a high pressure of blast was not protected by any agronomic management but when blast pressure was lower or when the cultivar presented a partial resistance, the cropping system and level of fertilization had more impact. Partial resistance appears as a potential of the cultivar that can be stimulated in certain cropping systems or fertilization levels.

Differences were observed between leaf and panicle blast severity. For example, from 2009 to 2011 leaf blast severity decreased in the partially resistant cultivar F161 with a significant effect of cropping system only in 2011, while panicle blast severity increased over the same period, with a significant effect of the cropping system, of fertilization, and of their interaction in 2009 and 2011. Leaf and panicle blast appeared to be independent of each other. This is consistent with the concept of two subsystems proposed by Teng et al. (1991) and confirms the importance of working on both leaf and panicle blast. The two subsystems are certainly not disconnected because in the majority of cases similar results were observed on leaves and panicles, but certain mechanisms of resistance are probably specific to one or other stage of development (Bonman, 1992).

The two cultivars displayed very different variations in blast severity in successive years. Because of the differences in notation in 2005 and 2006, leaf blast levels cannot be compared with those in the other years, but panicle blast evolution is informative. Blast severity appeared to decrease in 2006 and became low in 2007 in the more susceptible cultivar, F154. This may be as a result of the reduced use of this cultivar in the region. F154 was one of the first upland rice cultivars adapted to high altitude that was proposed to farmers in the 1990s but was progressively abandoned between 2000 and 2010 because of its high susceptibility to blast. In contrast, the partially resistant cultivar F161 was subject to progressively more attacks, especially on the panicle, and blast populations present in the vicinity of the trial plots appeared to adapt to this cultivar. With the increase in blast pressure on F161, differences between cropping systems became significant. The improvement of yield due to the increase of fertilization level was also reduced in the last years for F161 because of the simultaneous increase of blast severity. The use of two cultivars was also important because the primary inoculum was not controlled and because severity can differ depending on the specific climate of each year and on the prevalence of the cultivar in the region.

When the level of blast was low, yields of the partially resistant cultivar F161 were much better in the conventional cropping system. This is explained by the fact that the conservation agriculture rice cropping systems that have been tested in the highlands of Madagascar up to now have not been able to improve rice yields, particularly because of the early growth lag which could be due to restricted root growth and N immobilization (Dusserre et al., 2012). The better performance of the cropping system only becomes apparent under a certain level of blast pressure, but the yield increase is not great. In the last two years, the no-tillage cropping system kept blast at an acceptable level and compensated for the difference in yield in the partially resistant cultivar. When blast pressure was very high, the cropping system was no longer sufficient to protect a very susceptible rice cultivar against the disease. The no-tillage cropping system needs to be used in combination with other practices, such as a mixture of cultivars, to efficiently reduce the blast pressure on a susceptible cultivar in the environmental conditions of this study area (Raboin et al., 2012).

The many differences between the no-tillage and the conventional systems may explain the impact of the system on blast epidemics. First, there is a difference in plant nutrition. Interactions were frequently observed between the effect of the cropping system and fertilization. Nitrogen fertilization is known to increase blast severity, especially excessive application of nitrogen (Long et al., 2000). In the present experiments, the effect of the organic + mineral fertilization may be because the global amount of fertilization was higher and also that nitrogen was applied in a mineral form. The increase in blast as a result of fertilization was also lower in the no-tillage cropping system. This suggests that the dynamics of plant nitrogen assimilation may have been changed in the no-tillage system or that the cropping system allows the plant to keep some mechanisms of partial resistance which are usually susceptible to nitrogen intake. Dordas (2008) reviewed the role of nutrients in the control of plant disease in sustainable agriculture. The plant mineral balance may explain the interaction between the cropping system effect and the type of fertilization observed in these experiments. The dynamics of nitrogen release, a cause of high susceptibility (Kürschner et al., 1992), is very often modified in conservation agriculture cropping systems (Scopel et al., 2013). The presence of a soil cover has been shown to change soil nitrate and ammonium content (Xu et al., 2012), and the degradation of crop residues left on the soil in the no-tillage cropping system and exported in the conventional tillage system may be a source of silicon. Silicon has been shown to have a high impact on plant–pathogen interactions and a silicon input improves rice tolerance against blast (Seebold et al., 2000).

Crop growth may also be important. It determines the period when crop density is highest and when the duration of leaf wetness is longest. In low temperature environments, a short wetness period may limit penetration by the pathogen (Bonman, 1992). A delay in vegetative development, as observed in the no-tillage system, may be an advantage for better management of blast if the period of high susceptibility of the crop is not the same as the period of higher blast pressure. The experiments conducted here did not allow determination of the relative part of the cropping system on rice nutritional balance and on crop development. Further investigations based on the analysis of plant nutrient content and different sowing densities could validate one or other option.

The conservation agriculture cropping systems tested in these experiments have clearly been shown to modify the level of blast disease and the impact of fertilization on rice susceptibility. These systems were not as productive as conventional systems when blast level was low, but in cases of higher blast pressure, they at least produced the same yield. The question is then: should one accept a small drop in yield to improve the resilience and security associated with the cropping system? That question is not easy to answer in developing countries. Systems should be optimized to ensure at least the same yield in each case and to increase the protection against blast. Improved performance cropping systems will now be tested and more measurements will be made to explain whether the effects of the system depend more on crop development, fertilization, type of partial resistance involved or another cause.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

This work was partly funded by the GSDM (Groupement Semis Direct Madagascar) and by the French Agence Nationale de la Recherche under the Systerra Programme: ANR-09-STRA-03 (Agronomic management of rice blast resistance).

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • Bastiaans L, 1991. Rice leaf blast photosynthesis effect. Phytopathology 81, 6115.
  • Bonman J, 1992. Durable resistance to rice blast disease – environmental influences. Euphytica 63, 11523.
  • Castilla NP, Willocquet L, Suwarno S et al., 2010. Assessing the effect of resistant–susceptible associations and determining thresholds for associations in suppressing leaf and neck blast of rice. Crop Protection 29, 390400.
  • Dordas C, 2008. Role of nutrients in controlling plant diseases in sustainable agriculture. A review. Agronomy for Sustainable Development 28, 114.
  • Dusserre J, Chopart JL, Douzet JM, Rakotoarisoa J, Scopel E, 2012. Upland rice production under conservation agriculture cropping systems in cold conditions of tropical highlands. Field Crops Research 138, 3341.
  • Dzido JL, Vales M, Rakotoarisoa J, Chabanne A, Ahmadi N, 2004. Upland rice for the highlands: new varieties and sustainable cropping systems to face food security. Promising prospects for the global challenges of rice production the world will face in the coming years? FAO Rice Conference. Rome, Italy: FAO, 12–3.
  • FAO, 2006. World Reference Base for Soil Resources, 2nd edn. World Soil Resources Reports no. 103. Rome, Italy: Food and Agriculture Organization of the United Nations, International Society of Soil Science, 145.
  • FAO, 2010. Conservation agriculture. [http://www.fao.org/ag/ca/]. Accessed 5 March 2013.
  • Fernandez MR, Mcconkey BG, Zentner RP, 1999. Effects of tillage method and fallow frequency on leaf spotting diseases of spring wheat in the semiarid Canadian prairies. Soil and Tillage Research 50, 25969.
  • Gunawardena TA, Fukai S, 2005. The interaction of nitrogen application and temperature during reproductive stage on spikelet sterility in field-grown rice. Australian Journal of Agricultural Research 56, 62536.
  • Husson O, Charpentier H, Naudin K et al., 2010. Les systèmes SCV adaptés aux différentes zones agroécologiques de Madagascar. In: Manuel Pratique du Semis Direct à Madagascar. Antananarivo, Madagascar: GSDM/CIRAD, 124.
  • IRRI International Rice Research, 2009. Rice standard evaluation system for rice. [http://www.knowledgebank.irri.org/extension/ses.html]. Accessed 5 March 2013.
  • Kürschner E, Bonman J, Garrity DP, Tamisin MM, Pabale D, Estrada BA, 1992. Effects of nitrogen timing and split application on blast disease in upland rice. Plant Disease 76, 3849.
  • Lai X, Marchetti M, Petersen HD, 1999. Comparative slow-blasting in rice grown under upland and flooded blast nursery culture. Plant Disease 83, 6814.
  • Long DH, Lee FN, TeBeest DO, 2000. Effect of nitrogen fertilization on disease progress of rice blast on susceptible and resistant cultivars. Plant Disease 84, 4039.
  • Minten B, Randrianarisoa JC, Barrett CB, 2007. Productivity in Malagasy rice systems: wealth-differentiated constraints and priorities. Agricultural Economics 37, 22537.
  • Mundt CC, 2002. Use of multiline cultivars and cultivar mixtures for disease management. Annual Review of Phytopathology 40, 381410.
  • Ou SH, 1985. Rice Diseases. Slough, UK: Commonwealth Mycological Institute.
  • Pennisi E, 2010. Armed and dangerous. Science 327, 8045.
  • Raboin LM, Ramanantsoanirina A, Dzido JL et al., 2011. Upland (aerobic) rice breeding for the harsh environment of the High Plateau of Madagascar. In: Kiepe P, Diatta M, Millar D, eds. Innovation and Partnerships to Realize Africa's Rice Potential. Bamako, Mali: Africa Rice Congress 2010, 267. [http://www.warda.cgiar.org/workshop/ARC/1.13%20Raboin%20ed2.pdf]. Accessed 5 March 2013.
  • Raboin LM, Ramanantsoanirina A, Dusserre J et al., 2012. Two-component cultivar mixtures reduce rice blast epidemics in an upland agrosystem. Plant Pathology 61, 110311.
  • Ratnadass A, Fernandes P, Avelino J, Habib R, 2011. Plant species diversity for sustainable management of crop pests and diseases in agroecosystems: a review. Agronomy for Sustainable Development 32, 273303.
  • Razafimbelo TM, Albrecht A, Basile I et al., 2006. Effet de différents systèmes de culture à couverture végétale sur le stockage du carbone dans un sol argileux des Hautes Terres de Madagascar. Etude et Gestion des Sols 13, 11327.
  • Scopel E, Triomphe B, Affholder F et al., 2013. Conservation agriculture cropping systems in temperate and tropical conditions, performances and impacts. A review. Agronomy for Sustainable Development 33, 11330.
  • Seebold KW, Datnoff LE, Correa-Victoria F, 2000. Effect of silicon rate and host resistance on blast, scald, and yield of upland rice. Plant Disease 84, 8716.
  • Teng PS, Klein-Gebbinck HW, Pinnschmidt H, 1991. An analysis of the blast pathosystem to guide modeling and forecasting. In: Rice Blast Modeling and Forecasting. Selected Papers from the International Rice Research Conference. Los Banos, The Philippines: IRRI, 130.
  • Walters D, 2009. Managing crop disease through cultural practices. In: Walters D, ed. Disease Control in Crops: Biological and Environmentally Friendly Approaches. Oxford, UK: Wiley-Blackwell, 721.
  • Xu H, Qin F, Xu Q, Ma G, Li F, Li J, 2012. Paddy rice can be cultivated in upland conditions by film mulching to create anaerobic soil conditions. Journal of Food, Agriculture and Environment 10, 695702.
  • Zhu Y, Chen H, Fan J et al., 2000. Genetic diversity and disease control in rice. Nature 406, 71822.