African Crop Science Journal African Crop Science Society ISSN: 1021-9730 EISSN: 2072-6589 Vol. 7, Num. 1, 1999, pp. 47-57

E.N. NUKENINE, A.G.O. DIXON, A.T. HASSAN¹ and J.A.N. ASIWE

International Institute of Tropical Agriculture (IITA), PMB 5320, Ibadan, Nigeria
¹Department of Zoology, University of Ibadan, Ibadan, Nigeria
*Current address for correspondence: Department of Biological Sciences, Faculty of Science, University of Ngaoundere, B.P. 454, Ngaoundere, Cameroon

Three field trials were conducted at IITA, Ibadan, Nigeria between 1993 and 1995 to identify a rapid method of screening cassava (Manihot esculenta Crantz) cultivars for canopy retention and to determine the association between canopy retention and resistance to green spider mite (Mononychellus tanajoa Bondar) in cassava. Three methods (I, II and III) were used to assess canopy retention in 70 cultivars. Method I involves visual estimation and it takes the longest time; method II involves visual estimation and it takes the shortest time; and method III involves taking measurements and it takes a longer time than method II, but with a much shorter time than method I. Method II was the best method for screening cassava cultivars for canopy retention during dry periods. The cultivars showed significant (P < 0.01) differences for canopy retention, stay green ability, mite population density and damage scores, but not for relative water content. Canopy retention was positively correlated (P < 0.01) with stay green ability at the peak of the dry season (January). Generally, canopy retention and stay green ability were inversely associated (P < 0.01) with mite density in March 1994 and 1995, and with damage during the dry season (December to March 1994 and 1995, respectively). It is proposed that cassava cultivars which are tolerant to drought may also be resistant to M. tanajoa and that the genetic potential of cassava to retain many green leaves during the dry season may be a major factor of resistance to M. tanajoa.

Key Words: Manihot esculenta, Mononychellus tanajoa, relative water content

Trois essais en champs ont été menés à l=IITA à Ibadan au Nigeria entre 1993 et 1995 pour identifier une méthode rapide de criblage des cultivars de manioc (Manihot esculenta Crantz) pour la rétention foliaire et la capacité de rester vert, et de déterminer la relation entre la rétention foliaire et la resistance à l=acarien vert du manioc (Mononychellus tanajoa Bondar). Trois méthodes (I, II et III) ont été utilisées pour évaluer la rétention foliaire chez 70 cultivars. La méthode I implique l=estimation visuelle et elle prend le temps plus long; la méthod II implique l=estimation visuelle et elle prend le temps plus court; et la méthode III implique les mesures et elle prend un temps plus long que celui de la méthode II, mais beaucoup plus court que celui de la méthode I. La méthode II a été la meilleure méthode de criblage des cultivars de manioc pour la rétention foliaire pendant les périodes sèches. Il y avait des différences significatives (P < 0,01) entre les cultivars pour la rétention foliaire, la capacité de rester vert, la densité des acariens et des dégâts causés par des acariens mais pas pour la teneur relative en eau. Des corrélations positives ont été notées entre la rétention foliare et la capacité de rester vert au pic de la saison seche (Janvier). En général, la rétention foliaire et la capacité de rester vert étaient inversément associées (P < 0,01) à la densité des acariens en Mars 1994 et 1995, et avec des dégâts causés par l=acarien pendant la saison sèche (de Décembre à Mars 1994 et 1995). Il ressort de ces résultats que les cultivars de manioc qui sont tolérants à la sécheresse semblent aussi résistants à M. tanajoa et que le potentiel génétique du manioc à retenir ses feuilles pendant la saison sèche peut être un facteur majeur de résistance au M. tanajoa.

Mots Clés: Capacité de rester vert, manioc, Mononychellus tanajoa, teneur relative en eau

Cassava (Manihot esculenta Crantz [Euphorbiaceae]) is a species native to South America that was introduced to Africa in the 1600s where it quickly became an important food. This crop now provides more than half of the dietary calories for over 200 million people of sub-Saharan Africa (IITA, 1992). Cassava has long been considered a hardy crop resistant to pests and diseases. In 1971, the green spider mite (Mononychellus tanajoa Bondar), new to Africa, was discovered on cassava in Uganda (Nyiira, 1972). The mite has since spread to 27 countries in the cassava belt, causing serious damage to cassava by feeding on its leaves (Yaninek et al., 1989). Production losses caused by M. tanajoa have been estimated to range from 13 to 80% (Lyon, 1974; Nyiira, 1976; Shukla, 1976; Ndayiragije, 1984).

Chemical control is not a viable option because the crop is of long duration (8-24 months in the field). In addition, chemical control is expensive and leads to environmental pollution. Host plant resistance and classical biological control were identified as environmentally sound control options for M. tanajoa in the early eighties (IITA, 1992). Many cassava cultivars have shown some level of resistance to M. tanajoa (Hahn et al., 1989; IITA, 1993). However, research on the mechanisms of resistance is scanty. In Nigeria, population and damage peaks of M. tanajoa occur in January (the middle of the dry season) and cultivars which retain fewer leaves appear to sustain heavy mite damage, whereas those which retain many leaves sustain little mite damage (Dixon and Nukenine, personal observation). There is, thus, a need to screen cassava cultivars for canopy retention and to investigate the relationship between canopy retention and resistance to M. tanajoa. The objectives of this study were therefore to devise a rapid, simple screening method to estimate canopy retention in cassava and to determine the association between canopy retention and resistance to M. tanajoa.

Field trials. One and two field trials were established in 1993 and 1994, respectively, in a research farm at the International Institute of Tropical Agriculture (IITA), Ibadan, Nigeria. Planting was done in May for both years. The 1993, first 1994 and second 1994 trials were composed of 38 (27 improved and 11 local), 25 (24 improved and 1 local) and 25 (24 improved and 1 local) entries, respectively. Eighteen cultivars from the 1993 trial were repeated in the 1994 trials. Three of them (91/02319, 30572 and TME 1) were repeated in the two trials, while 15 of them (91/00453, 91/00455, 91/00458, 91/00459, 91/01730, 91/02312, 91/02316, 91/02317, 91/02325, 91/02327, 92/0057, 92/0067, 92/0342, 92/0402 and 92/0427) were repeated in one of the two trials.

Each trial was arranged in a randomised complete block design. In the 1993 trial, some cultivars were replicated two times, while others were replicated three times. In the 1994 trials each cultivar was replicated four times. A single 10 m row (ridges 30 cm high) plot spaced 1 m apart was used for the 1993 trial, while in the 1994 trials, each plot consisted of four rows of the same length and distance apart. Plant spacing was 1 m x 1 m for all the trials, giving a population of 10,000 plants ha^-1.

Estimation of canopy retention (CR) using three methods (CR methods I, II and III). Three methods (CR methods I, II and III) were used to assess cassava cultivars for canopy retention, with the aim of identifying the simplest and fastest method. CR methods I and II were used in the 1993 trial, while CR methods II and III were used in the 1994 trials. Method II was common to both years, making comparison among the three methods possible. In each plot, two plants (1993 trial) and five plants (1994 trials) were randomly selected for data collection.

For CR method I, the portion of the main stem with leaves was visually estimated as a percentage of the total length of the main stem (CA). Scoring was done in September (baseline data) and December 1993, as well as January and March 1994. CR method I was estimated as the CA at a given time during the dry period (CA_t) as a function of the maximum CA during the wet period (baseline) (CA_b) for each cultivar. The amount of canopy lost as a result of drought (the dry season) is given as

The amount of canopy lost in percentage is given as:

Therefore the percentage of canopy retained or CR method I (%) is:

Since CR method I involves a long period of estimation (wet and dry season), CR method II was introduced with the aim of reducing the time by not necessitating the collection of data in the wet season. Thus, with CR method II, the CA at any given period (usually in the dry season) was taken as canopy retention. Therefore,

If CR method I correlates with CR method II, then it will be unnecessary to use CR method I with a much longer estimation time. However, since the two methods involve visual estimation, they may not be accurate enough. In the second year, CR method III involving measurements was introduced. It was thought that since CR III involves measurements, it could be more accurate, and could also be used to test the accuracy of CR method II. There was a strong positive correlation between CR methods I and II, hence, only CR method II was estimated in the second year. CR method III was estimated in December 1994, and January and March 1995. The distances from the ground level to the tip of a mainstem of a plant (C_u) and from the ground level to the level of insertion of the first leaf from the bottom of the same stem (C_l) were measured using a graduated wooden rod. CR method III was then calculated as follows:

Estimation of stay green ability. Stay green ability (SGA) which is the number of leaves retained on the mainstem as a ratio of the total number of leaves present was also assessed only in the 1993 trial. SGA was estimated in September and December 1993, as well as January and March 1994.

Estimation of relative water content. Relative water content (RWC) was determined in December 1993 and January 1994 for all test cultivars, and in March 1994 for 14 cultivars with varying levels of resistance, following the procedures described by Smart and Bingham (1974). On each sampling date, leaf 3 (leaf 1 being the first fully expanded leaf from the tip of the main stem) was plucked from two plants in a plot between 0700 and 0800 hrs. Ten leaf discs (ca. 3.14 cm²) were taken from each leaf using a number 15 cock borer. The fresh weight of the 20 leaf discs from each plot was determined using an electronic balance, model GT 210. The leaf discs were floated in distilled water in petri-dishes and placed in a dark growth chamber at 26± 2 ° C for 3.5 hr. Floated leaf discs were surface-dried using tissue paper, and turgid weights were immediately taken. Leaf discs were then oven-dried at 85 ° C for 24 hr. The dry weights of the discs were measured. The RWC, an indicator of the plant water status, was determined as the ratio of the amount of water in the leaf tissue at sampling to that present when fully turgid, using the formula:

Estimation of the population and damage of M. tanajoa. Cassava green mite population was estimated once in January and March of 1994 (1993 trial) and of 1995 (1994 trials) .

From each plot, leaf 3 was plucked from two and five randomly selected plants for the 1993 trial and 1994 trials, respectively, on each sampling date. The leaves were carried to the laboratory in paper bags and refrigerated for at least two hours to immobilize the mites. The actives (sum of all developmental stages less the eggs) of M. tanajoa on the abaxial leaf surface were counted under a stereomicroscope (model WILD M5A; WILD HEEBUGG, Switzerland). Leaf area was determined using an electronic leaf area meter (model LI-3000A; LI COR Inc., U.S.A.). Mite density cm-2 of leaf surface was calculated as:

Data analysis. Data were subjected to analysis of variance using SAS software (SAS Institute, 1993). Mite population data were transformed [(square root (x + 0.5)] prior to the analysis. The standard errors of the mean were calculated for all the parameters. Correlation analysis (across cultivars: N = 38 (38 cultivars) for 1993 trial; N = 25 (25 cultivars) for each of the 1994 trials) was conducted to determine linear relationships among M. tanajoa population density and damage, canopy retention methods (CR methods I, II and III), RWC and SGA.

Evaluation of cassava cultivars for canopy retention. Canopy retention (CR methods I, II and III) varied significantly (P < 0.01) among the cultivars on all dates, irrespective of trial (Tables 1, 2 and 3). At the peak of the dry season (January 1994), the three cultivars with the highest CR method I for the 1993 trial were 92/0401, 92/0067 and 92/0427, whereas those with the lowest were ISU, 30001 and 30572 (Table 1). For CR method II, MS-20, 92/0067 and 92/0427 had the highest values while 30572, 2ND AGRIC and ISU had the lowest values.

In the first 1994 trial, cultivars 92/0427, 92/0342 and TME 1 had the largest canopy retention assessed by CR methods II and III, irrespective of sampling period. However, in March 1995, the cultivar 92/02312 replaced TME 1 as one of the first three cultivars for CR method II, while in December 1994, the cultivar 91/02322 replaced 92/0342 for CR method III. Except for March 1995, the cultivars 91/0427, 91/02163 and 91/01363 were among the cultivars with the smallest canopy retention, regardless of canopy retention method or sampling period (Table 2). In the second 1994 trial, cultivars TME 1, 91/02325 and 91/02319 recorded the largest values for canopy retention in January 1995, irrespective of the CR method (Table 3). Cultivars 82/00661, 90/02030 and 90/01718, and 82/00661, 90/02030 and 88/02555 had the smallest values for CR method II and CR method III, respectively.

Variation in stay green ability, relative water content and mite responses among cultivars. Except in September, SGA varied significantly (P < 0.01) among the cultivars (Table 4). There were no significant (P > 0.05) differences in RWC among cultivars in December 1993, as well as in January and March 1994. Mite population density and damage varied (P < 0.01) among the cultivars in all field trials during the dry season (Tables 1, 2 and 3).

TABLE 1. Canopy retention (CR) methods I and II of 38 cassava cultivars at different periods in the 1993/94 trial

TABLE 2. Canopy retention (CR) methods II and III, and M. tanajoa density (Pop) and damage (dam) of 25 cassava cultivars at different periods in the first of the 1994/95 trials

TABLE 3. Canopy retention (CR) methods II and III of 25 cassava cultivars in December (Dec) 1994, January (Jan) and March (Mar) 1995 as well as M. tanajoa density (Pop) and damage (Dam) in the second of the 1994/95 field trials

Relationships among canopy retention methods (CR methods I, II and III), SGA, RWC, M. tanajoa population density and damage. In the 1993 trial, CR method I was positively correlated with CR method II on all sampling dates (Table 5). The CR method II in September was positively but weakly correlated with CR method I in December 1993 (r = 0.35; P < 0.05); it, however, was not correlated (P > 0.05) with either that of January or March 1994.

Stay green ability was positively correlated with CR method I and CR method II on all sampling dates, except in December 1993 and September 1993, respectively. RWC was not correlated with CR methods and SGA. In the two 1994 trials, CR method II was positively and closely associated with CR method III in all the sampling dates.

In January 1994, the cultivars (92/0427, 92/0067 and MS-20) with highest CR methods I (37-54%) and II (25-33%) had low mite densities (0.60-2.07 actives cm-2) and low mite damage scores (1.36-1.54) (Tables 1 and 4) for the 1993 trial. In contrast, the cultivars (30572, 2ND AGRIC and ISU) with the lowest canopy retention (CR methods I (5-11%) and II (4-7%) ) had high mite densities (1.06-5.81 actives cm-2) and damage scores (2.5-4). In general, cultivars with high canopy retention values had low mite densities and damage scores, whereas cultivars with low canopy retention values had high mite densities and damage scores, irrespective of the canopy retention method and sampling period. Generally, in the first 1994 trial, cultivars with high canopy retention assessed by both CR methods II and III recorded relatively low mite density and significantly lower mite damage scores, while cultivars with small CR values had high mite density and damage scores (Table 2). Apart from TME 1, the cultivar with the overall highest CR value did not show a consistent trend with either low mite density or damage score in the second 1994 trial (Table 3). A similar observation was made for cultivars with low CR values and mite density and damage.

In the 1993 trial, mite damage was inversely associated with CR methods I and II in all sampling dates. Except for December 1993, mite damage was also inversely associated with SGA on all the other dates. Significant correlations in December 1993 were generally weaker than those for January and March 1994. Mite density was negatively correlated with SGA, CR methods I and II, and SGA in March 1994, but there were no significant relationships in January 1994 (Table 5). RWC did not correlate with mite density and damage. In the first 1994 trial, CR method II was negatively correlated with mite damage in all the sampling dates whereas CR method III was negatively correlated with mite damage in January and March 1995. In March 1995, mite density was positively and negatively correlated with CR method II and CR method III, respectively. For the second 1994 trial, neither CR method II nor CR method III was correlated with mite damage in either December 1994 or March 1995. Mite density did not correlate with either CR method II or CR method III.

TABLE 4. Stay green ability (SGA), relative water content (RWC) and mite density (Pop), and damage (Dam) of 38 cassava cultivars at different periods in the 1993/94 field trial

TABLE 5. Correlation coefficients among stay green ability (SGA), canopy retention (methods I, II and III), relative water content (RWC), and mite density (pop) and damage (dam) at different periods in the 1993 and 1994 trials

Plant water deficit (drought stress) causes reduction in leaf area, increase in leaf abscission, low stomatal conductance and reduced transpiration rate (Hsiao, 1973). Baker et al. (1989) noticed increased leaf senescence and leaf fall, and reduced leaf production in cassava during drought stress. It is implicit that cultivars that lose more leaves and with a higher rate of leaf senescence may be more susceptible to drought stress than cultivars that lose less leaves and have a lower rate of leaf senescence. Cultivars with higher CR and SGA may be less affected by drought stress than those with lower CR and SGA. Thus, cultivars with higher CR and SGA may maintain a higher water status during dry periods than those with lower CR and SGA.

The positive and close relationships obtained between the CR methods I and II, and methods II and III, show an agreement between these methods in estimating canopy retention, indicating that each method could be a reliable indicator of the other. Although method I appeared to have the highest accuracy, it may not be the most efficient since it takes the longest time (wet to dry season) to measure. Both CR methods II and III can equally be estimated just within the dry season. However, method II is better than method III because it is less tedious, rapid and simple to measure without necessarily sacrificing accuracy. In terms of resource allocation, the methods could be ranked in increasing order of importance for screening for CR thus: method I < method III < method II. The use of these methods does not require use of sophisticated and expensive equipment.

The consistently close, but inverse associations obtained over 2 years between mite damage and canopy retention during the peak of the dry periods (January) indicates that the genetic potential of cassava to retain green leaves during these periods may be a major factor of resistance to M. tanajoa. The absence or presence of a relatively low correlation between mite damage and canopy retention in December was not surprising because a cultivar with a genetically conditioned large canopy may not necessarily retain many leaves during dry periods. This is supported by the absence of a significant linear relationship between the baseline canopy retention (September) and canopy retention in January and March. January (mid-dry season) is the most appropriate period for screening for canopy retention, since correlation coefficients between mite damage and CR were consistently high in January of both years, regardless of the trial. March would have also been appropriate if there were no rainy days that reduced the correlations between CR and mite damage. Rain boosts canopy development and dislodges mites (Yaninek et al., 1989). The close, positive association between CR and SGA and between CR and mite damage in January implies that cultivars which are tolerant to drought are resistant to M. tanajoa. The results also show that there is no need for estimation of SGA along with CR, owing to the high degree of correlations obtained between SGA and CR in January and March.

In all trials, mite density was not correlated with CR methods I, II and III in December and January. Except for the second of the 1994/95 field trials, mite population density was not correlated with CR only in March. The absence of a relationship between CR and mite density in January was not unexpected since susceptible cultivars dropped most or all of their leaves during this period, while some resistant cultivars retained many green leaves. The mites were, thus in a no-choice situation of migrating to resistant cultivars, resulting in a higher mite population on some resistant than susceptible cultivars during relatively severe dry periods, especially in January 1994.

Owing to the differential effects of drought on resistant and susceptible cultivars, the correlation of mite population and CR during dry periods, for the evaluation of cassava cultivars for resistance to M. tanajoa cannot be relied upon. Instead, mite damage and CR should be correlated during such periods. Since there is a positive relationship between drought tolerance (as measured by canopy retention) and mite resistance, future research should focus on other indicators of drought tolerance such as rooting depth, bark thickness and stomatal sensitivity among cassava cultivars; these could provide a more quantitative and physiological basis of resistance to M. tanajoa.

During drought stress, the plant suffers a reduction in turgor pressure that causes a decline in cell expansion and cell wall synthesis (Hsiao, 1973). In addition, reduced transpiration during drought stress (Hsiao, 1973) causes leaf temperature to increase and relative humidity (RH) to decrease (Holtzer et al., 1988). It is possible that the resistance to M. tanajoa in cultivars with high CR and SGA may be caused by high turgor pressure and RH and low temperature. Higher turgor pressure leads to thicker cell walls (Hsiao, 1973), which makes it difficult for mites to feed. Higher RH and lower temperatures are not conducive for the survival of M. tanajoa and other tetranychid mites (Holtzer et al., 1988; Yaninek et al., 1989).

Failure to detect variations in RWC among the cultivars may be because RWC is a rather insensitive indicator of water status when water deficit is not severe (Hsiao, 1973). Hsiao (1973) explained that in nearly saturated tissues, a small change in water content corresponds to a change in RWC which is about the random error in RWC measurements in several studies. However, Pillai and Palaniswami (1990) obtained a strong negative correlation between RWC in cassava and damage by a different species of spider mite in India.

In conclusion, canopy retention method II is a simple, rapid and cheap method of screening cassava cultivars for canopy retention. Breeding of cassava cultivars with the ability to retain large canopy during dry periods may reduce cassava leaf damage caused by M. tanajoa.

This is a contribution for the International Institute of Tropical Agriculture (IITA/JA/97/56). It is part of a PhD Thesis that was submitted to the Department of Zoology, University of Ibadan, Nigeria by the first author.

1. Baker, G.R., Fukai, S. and Wilson, G.L. 1989. The response of cassava to water deficits at various stages of growth in the subtropics. Australian Journal of Agricultural Research 40:517-528.
2. Hahn, S.K., Isoba, J.C.G. and Ikotun, T. 1989. Resistance breeding in root and tuber crops at the International Institute of Tropical Agriculture, Ibadan, Nigeria. Crop Protection 8:147-168.
3. Holtzer, T.O., Norman, J.M., Perring, T.M., Berry, J.S. and Heintz, J.C. 1988. Effects of microenvironment on the dynamics of spider mite population. Experimental and Applied Acarology 4:247-264.
4. Hsiao, T.C. 1973. Plant responses to water stress. Annual Review of Plant Physiology 24:519-570.
5. IITA. 1992. Sustainable Food Production in Sub-Saharan Africa 1: IITA=s contribution. International Institute of Tropical Agriculture, Ibadan, Nigeria. 208 pp.
6. IITA. 1993. Crop Improvement Division: Activity Report and Workplan for 1993. International Institute of Tropical Agriculture, Ibadan, Nigeria. 119 pp.
7. Lyon, W.F. 1974. A green cassava mite recently found in Africa. Plant Protection Bulletin 22: 11-13.
8. Ndayiragije, P. 1984. Cassava green mite (Mononychellus tanajoa (Bondar)) in Burundi. In: Integrated Pest Management of Cassava Green Mite. Greathead, A.H., Markham, R.H., Murphy, S.T. and Robertson, I.A.D. (Eds.), pp. 67-73. Proceedings of a Regional Training Workshop in East Africa, 30 April-4 May, 1984.
9. Nyiira, Z.M. 1972. Report of investigation of cassava mite, Mononychellus tanajoa Bondar. Kampala, Uganda, Kwanda Res Sta. 14pp.
10. Nyiira, Z.M. 1976. Advances in research on the economic significance of the green cassava mite Mononychellus tanajoa in Uganda. In: The International Exchange and Testing of Cassava Germplasm in Africa. Terry, E. and Mclntyre, R. (Eds), pp. 22-29. Proceedings of an interdisciplinary workshop held at IITA, Ibadan, Nigeria, 17-21 November, 1975.
11. Pillai, K.S. and Palaniswami, M.S. 1990. Evaluation of cassava accessions resistant to spider mite and factors governing resistance in India. In: Tropical Root Crop and Tuber Crops Changing Role in a Modern World. Howeler, R.H. (Ed.), pp. 341-350. Proceedings of the 8th Symposium of the International Society for Tropical Root Crops. Bangkok, Thailand, 30 October-5 November, 1988.
12. SAS Institute. 1993. SAS Companion for Microsoft Windows Environment, Version 6 1st ed. SAS Institute, Cary, North Carolina, U.S.A.
13. Shukla, P.T. 1976. Preliminary report on green mite (Mononychellus tanajoa Bondar) resistance in Tanzania local cassava varieties. East African Agriculture and Forestry Journal 42:55-59.
14. Smart, R.E. and Bingham, G.E. 1974. Rapid estimation of relative water content. Plant Physiology 58:258-260.
15. Yaninek, J.S., de Moraes, G.J. and Markham, R.H. 1989. Handbook on the Cassava Green Mite Mononychellus tanajoa in Africa. IITA, Ibadan, Nigeria. 210 pp.

Copyright 1999, African Crop Science Society