African Crop Science Journal African Crop Science Society ISSN: 1021-9730 EISSN: 2072-6589 Vol. 9, Num. 2, 2001, pp. 451-461

African Crop Science Journal, Vol. 9. No. 2, pp. 451-461

VARIATION IN YIELD AND RESISTANCE TO GROUNDNUT ROSETTE DISEASE IN EARLY- AND MEDIUM-MATURING GROUNDNUT GENOTYPES IN NIGERIA

B. R. NTARE* and P.E.OLORUNJU¹

ICRISAT-Bamako, BP 320, Bamako, Mali
¹Institute for Agricultural Research, Ahmadu Bello University, PMB 1044, Zaria, Nigeria

(Received 7 June, 2000; accepted 31 January, 2001)

Code Number: CS01027

INTRODUCTION

The groundnut (Arachis hypogaea L.) is an important oil, food and fodder crop, which plays a significant role in the agriculture economy of countries in the semi-arid tropics. Within West Africa region, Nigeria produces 41% of the region’s total production (Freeman et al., 1999). Over the last three decades the yield of this crop has suffered due to attack by groundnut rosette disease. The disease is a viral complex involving groundnut assistor luteovirus (GRAV), groundnut rosette umbravirus (GRV) (Murant , 1989; Tiliansky et al., 1996) and satellite RNA (sat RNA; Murant et al., 1988; Blok et al. 1994) of GRV which is transmitted by the aphid (Aphis craccivora Koch). All the three agents must occur together for transmission by the aphid vector and subsequent disease development (Naidu et al., 1999). Symptoms associated with the disease are variable and two types (chlorotic and green rosette) are known. These symptoms are largely due to sat RNA (Murant et al., 1988) and variants of sat RNA are responsible for different forms of the rosette disease (Murant and Kumar, 1990). Although rosette epidemics are sporadic, yield losses approach 100% whenever the disease occurs in epidemic proportions. For example, the rosette disease epidemic of 1975 in Nigeria destroyed an estimated 0.7 million hectares of groundnut incurring a loss of nearly US $ 250 million (Yayock et al., 1976). Recurrent epidemics have limited production since then.

Use of cultural practices such as spraying to control aphids was found to be effective in controlling GRV (Davies, 1975). Bringing forward sowing dates to allow the crop establish before aphid population build-up and using denser stands to discourage infestation were also effective (Booker, 1963; A’Brook, 1964; Farrel, 1976). But, small-scale farmers who grow groundnut often face difficulties in adopting such practices. Few can afford pesticides, which are usually unavailable. Most give priority to subsistence crops such as sorghum or pearl millet and have little labour to spare to sow groundnut earlier and at higher densities. Also, the commonly grown early maturing cultivars lack resistance to rosette.

The sporadic occurrence of the disease from year to year and the lack of adoption of cultural control measures, make it desirable to grow cultivars with genetic resistance. These have the greatest potential to minimise the risks of losses due to rosette.

Research on the development of groundnut cultivars with resistance to rosette was initiated in the early 1950s by the French Institut de Recherches pour les Huiles et Oléagineux (IRHO) in West Africa. Sources of resistance to rosette were first discovered in groundnut landraces of late-maturing Virginia (A. hypogaea L. subsp. hypogaea var. hypogaea ) from Burkina Faso (then Haut Volta) and Cote d ‘Ivoire in 1952 (Saunger and Catherinet, 1954). These sources formed the basis for the rosette resistance breeding programmes throughout Africa. These attempts resulted in the development of long-duration varieties such as 69-101 (130 days to maturity), RMP 12 , RMP 40 and RG 1 (140-150 days) and early-maturing (90 days) Spanish (A. hypogaea L subsp. fastigiata var. vulgaris) types such as KH 241 D (Bockelée-Morvan, 1960). Resistance among these cultivars was found effective against both chlorotic and green rosette and was governed by two independent recessive genes (Berchoux, 1960; Nigam and Bock, 1990; Olorunju et al., 1992). Unfortunately, the rosette resistant long-duration varieties developed are not adapted to the short growing seasons of the dry savannah of West Africa, where the bulk of the crop is grown. The few short-duration rosette-resistant varieties have poor agronomic characteristics and hence were not widely adopted by farmers.

To ensure availability of groundnut genotypes with rosette disease resistance, early maturing (90-110 days), and suitable for small-scale farmers in short season agro-ecological zones, a joint breeding programme was initiated by the International Crops Research Institute for Semi-Arid Tropics (ICRISAT) and the Institute for Agricultural Research (IAR) at Samaru, Nigeria in 1992. The objective of this study was to evaluate new advanced breeding lines for resistance to groundnut rosette disease over several field locations and compare the disease incidence with pod yield.

Selection of resistant genotypes. Breeding lines used in this study were derived from crosses involving rosette-resistant parents; RG1 and RMP 40 (long-duration Virginia type) from Malawi and Cote d’ Ivoire, respectively, and KH 241D (short-duration Spanish type) from Burkina Faso. These were crossed with early-maturing susceptible Spanish types.

Crosses were made at Chitedze in Malawi during the 1988/89 growing season. The F₁ generation was grown in the screen house under disease-free conditions. F₂ population was evaluated in a rosette disease nursery in 1990/91 and selected plants were bulked and sent to Nigeria for further selection. The F₃ generation was grown in the dry season for generation advance. The F₄ generation was grown in a rosette nursery at Samaru in Nigeria and resistant plants were harvested individually. Promising and homozygous F_4:5 and F_4:6 progenies were bulked and seed increased for further purification and elimination of late-maturing plants.

Evaluations of genotypes in a rosette disease nursery. The F₇ lines tracing back to a single F₄ plant were stratified in three groups. Group 1 contained 49 lines mainly Spanish types with a growing cycle of <100 days, group 2 had 49 Virginia bunch types that matured around 110 days and group 3 contained 36 Virginia types that matured in 115 to 120 days. These were evaluated in rosette disease nursery at Samaru and Bagauda in 1996 and 1997. Entries in each group were arranged in a simple lattice design with two replications. Plots consisted of single 4 m rows. An infector-row technique developed by Bock and Nigam (1988) was used to induce artificial epidemic. This technique results in a disease incidence of 99% in susceptible entries. At sowing time, one infector row of susceptible cultivar 55-437 was sown after every two contiguous rows of test lines, such that every test row was adjacent to one infector row. Six weeks before planting on the field, a large number of seedlings of 55-437, were raised in the green house and inoculated with GRV, using a greenhouse culture of viruliferous aphids, which had been reared on GRV-infected plants. These were transplanted into the infector rows at 1.5-2 m spacing. In both years sowing was at the end of June to early July.

Detection of GRAV. All genotypes grown at Samaru in the rosette nursery in 1997 were tested for the presence of GRAV. The triple-enzyme linked immunosorbent assay (TAS-ELISA) as described by Rajeshwari et al. (1987) was used. The genotypes were not tested for GRV and its sat RNA as previous results showed a good correlation between symptoms and the presence of GRV and its sat RNA in either rosette susceptible or resistant germplasm (Blok et al., 1995).

Disease evaluation. Each entry was assessed for disease incidence 60 days after sowing. The total number of plants in each plot and the number of plants showing rosette symptoms with severe stunting were counted and percentage of disease incidence computed. Plants showing severe symptoms were stunted and bushy in appearance due to reduced internodes length. Leaves of the infected plants were reduced in size and the plants did not produce pods. Lines were considered resistant when no susceptible plants were found within the complete entry (0% incidence), highly susceptible when no resistant plants were present (100% incidence) and moderately resistant when at least one plant within the entry has mild symptoms (< 10% incidence). No yield data was recorded in these nurseries.

Yield performance. The same lines evaluated in the disease nursery were evaluated for yield performance at three locations representing three important agro-ecological zones for groundnut production in West Africa in 1996, 1997 and 1998. The first was Minjibir (12°8’N, 8° 40’E, 500 m above sea level) with an average annual rainfall of 700 mm and a growing season length of about 100 days. The soil is well drained with 0-1% slopes and is classified as hypothermic, ustic Plinthic Quartzipsamment (USDA taxonomy). The second was Bagauda (11° 40’N, 8° 30’E.) with an average annual rainfall of 900 mm with an average growing season length of 120 days with moderately well drained clay-loam soils. The third was Samaru (11° 8’N, long 7°E) with an average annual rainfall of 1200 mm and a growing length of 140-150 days with well-drained leached luvisols described as ferruginous tropical soils. Minjibir and Samaru are experimental research farms of IAR, while Bagauda was the ICRISAT research station in the country. Sowing by hand was usually done in late June to early July at all locations. Individual plots were 4 rows, 4 m long and 0.75 m apart. Within row spacing was about 10 cm. A basal dose of 100 kg ha^-1 of single super phosphate was incorporated into the soil by broadcasting during land preparation. The experimental design used in all yield trials was a lattice depending on the number of entries with three replications. Fields were kept weed-free by regular manual weeding. The trials were rainfed and no fungicides were used to control foliar diseases. In 1997, plots at Bagauda were artificially infested with viliruferous aphids raised on a rosette susceptible cultivar using an infector row technique described above. Disease incidence in each plot was assessed as described in the disease nursery. At harvest all plants in a plot were hand-lifted. Pods were separated from haulms and dried in the sun. The pods were weighed after cleaning and removal of soil and plant debris. Shelling percentage was determined from a 200-g sample of pods and seed weight was taken by weighing 100 sound mature kernels from each plot. The data were subjected to standard analysis of variance using GENSTAT statistical procedures.

Rosette disease reaction. In the rosette disease nurseries of 1996 and 1997, all susceptible checks were fully susceptible reaching 100 % disease incidence (Table 1). This indicated a high disease inoculum and the effectiveness of the infector-row technique. Both chlorotic and green rosettes were observed at all locations. Green rosette symptoms were, however, predominant at Samaru while chlorotic rosette symptoms were predominant at Bagauda and Minjibir. Under natural infection, rosette incidence varied among locations and years (data not shown). In the early maturity group disease incidence varied from 0 to 90% while in the medium group, the range was 0-30 %.

Among group 1 genotypes ICGV-IS 96859, ICIAR 12 AR, ICIAR 7B and ICIAR 10B were consistently free of rosette symptoms in all the nurseries at both locations and years, unlike the resistant parent KH241D that showed mild symptoms on some plants. In group 2 genotypes, eight genotypes did not show symptoms in the rosette nurseries at Samaru and Bagauda in 1996 and 1997. In group three all selected genotypes were resistant (0% disease incidence) at Samaru in both years. Some resistant genotypes showed mild symptoms especially of chlorotic rosette at Bagauda.

There were significant (P < 0.01) correlations between the number of plants with rosette symptoms (disease incidence) at Samaru in 1996 and 1997, which ranged from 0.82 to 0.92. At Bagauda the correlations ranged from 0.88 to 0. 96. Correlations of rosette incidence between Samaru and Bagauda were also significant (P = < 0.01) and ranged from 0.79 to 0.82 in 1996 and 0.72 to 0.84 in 1997.

GRAV detection. TAS-ELISA results revealed the presence of the luteovirus GRAV in both susceptible and resistant genotypes (data not shown). This suggested that all rosette resistant genotypes were infected by GRAV.

Yield performance. Since not all the genotypes were tested for three years at the three locations, the data set was not balanced to conduct a year- location analysis. Thus, individual years and locations are presented. Weather conditions were favourable in all the years at the three locations except at Minjibir where rains ended nearly one month earlier (early September) in each year compared to end of October at Samaru and Bagauda.

Several genotypes in each maturity group yielded significantly higher than the check cultivars under both induced and natural rosette epidemics. Yield of the susceptible checks under induced epidemic at Bagauda in 1997 demonstrated the destructive nature of the rosette disease (Table 2). For example, in group 1 genotypes, the pod yield of ICGV-IS 96900 was 1171 % higher than 55-437 and 424 % higher than RRB. Averaged over the years the yield of the top three genotypes (i.e., ICGV-IS 96894, ICGV-IS 96900 and ICGV-IS 96896) ranged from 19-92 % higher than the check cultivars. Among the group 2 genotypes, ICGV-IS 96801, ICGV-IS 96848, ICGV-IS 96826 and ICGV-IS 96808, produced the highest average yields (1.83 t ha^-1) (Table 3). These lines averaged 54 %, 59 % and 120 % higher yield than KH 241D, RRB and 55-437, respectively. In group three, some lines were clearly higher yielding than the resistant checks (Table 4) and were 10 days earlier maturing.

Within locations, greater rosette incidence was significantly (P = < 0.01) correlated with reduced pod yield in all genotypes. Correlations ranged from -0.25 to -0.89 in group 1 genotypes, -0.30 to -0.91, in group 2 and –0.34 to -0.81 in group 3. Yield and rosette incidence were more negatively correlated at Bagauda and Minjibir with greater mean and range of rosette incidence. The lowest correlations were observed at Samaru.

Shelling percentages of the new breeding lines were comparable to the checks but had larger kernels seeds than the checks (Table 5). The medium-maturing genotypes had the largest kernels. Most of the selected genotypes have the preferred tan colour compared to the red colour of the early (KH 241 D) and late (RG1) maturing sources of resistance to rosette disease.

The present study showed that resistance to rosette symptoms was not absolute since small portions of plants or a few branches of plants in resistant lines had rosette symptoms. All the genotypes resistant to GRV were susceptible to GRAV indicating lack of resistance to this component of the rosette complex. The results indicated variability of the virus complex and probably the behaviour of transmission efficiency of A. craccivora. Thus resistance to GRV could be overcome under high inoculum pressure or adverse environmental conditions (Naidu et al., 1999). These results along with earlier reports (Bock et al., 1990; Olorunju et al., 1991) suggest that distinct mechanisms of resistance might operate against the three agents (GRV and its satellite RNA, and GRAV) in the resistant material. An understanding of these mechanisms would enable the development of better strategies for incorporating resistance to all agents of rosette disease.

A high level of resistance to both green and chlorotic rosette was demonstrated in both maturity groups. Even under severe rosette conditions some genotypes showed less than 10 % disease incidence. Not only was the percentage of plants with disease symptoms lower in these selections, but also when symptoms did occur, they were of a mild nature to cause yield loss compared to the severe symptoms on susceptible checks. Selection under rigorous selection should result in lines with increased resistance. The high correlations (P = < 0.01) between rosette incidence within years and between locations indicated that genotypes reaction to rosette disease was similar. Thus, screening of genotypes at one site is sufficient to obtain estimates of resistance that are predictive of performance at other environments.

The impact of groundnut rosette on yield was demonstrated at Bagauda in 1997 where the susceptible checks produced negligible pod yields under induced epidemic. Even under natural conditions, greater rosette disease incidence was significantly correlated with reduced pod yield at all sites. Not only was rosette incidence less with several of the genotypes but also their yield potential was better than the commonly grown early-maturing cultivars such as RRB and 55-437. Our results suggest that resistance to rosette disease in the genotypes tested is the result of physiological resistance.

The reasonable pod yields, acceptable seed colour and size of the resistant genotypes give an indication of the potential benefit to groundnut producers in rosette endemic areas in Nigeria and West Africa as a whole. The resistant genotypes are being tested in regional trials in Ghana, Benin, Burkina Faso, Mali and Cameroon where rosette is a serious constraint to groundnut production.

Novel sources of resistance have been identified in wild Arachis (Subrahmanyam et al., 1998). This sets the stage where useful germplasm within the wild species can be utilised to develop more stable sources of resistance to groundnut rosette virus. Genotypes resistant to GRV identified in this study will contribute to such a programme.

1. A’Brook, J. 1964. The effect of planting date and spacing on the incidence of groundnut rosette disease and of the vector, Aphis craccivora Koch, at Mokwa, Northern Nigeria. Annals of Applied Biology 54:199-208
2. Berchoux, C. 1960. La rosette de l?arachide en Haute-Volta. Oléagineux 15:229-223.
3. Blok, V.C., Ziegler, A., Robinson, D.J. and Murant, A.F. 1994. Sequence of 10 variants of satellite RNA-3 of groundnut rosette virus. Virology 202:25-32.
4. Blok, V.C., Ziegler, A., Scott, K., Dangora, D.B., Robinson, D.J. and Murant, A.F. 1995. Detection of groundnut rosette umbravirus infection with radioactive and non-radioactive probes to its satellite RNA. Annals of Applied Biology 127:321-328.
5. Bock, K.R. and Nigam, S.N. 1988. Methodology of groundnut rosette resistance screening and vector ecology studies in Malawi. In: Proceedings Collaborative Research Groundnut Rosette Virus. 8-10 March 1987, Lilongwe, Malawi. International Crops Research Institute for the Semi-Arid Tropics, Patancheru, Andhra Pradesh 502324, India. pp. 7-10.
6. Bock, K.R., Murant, A.F. and Rajeshwari, R. 1990. The nature of resistance in groundnut to rosette disease. Annals of Applied Biology 117:379-384
7. Bockelée-Morvan, A. 1960. Les variétés d’arachide CN 94 C et Q h 243 C. Oléagineux 43:421-426.
8. Booker, R. H. 1963. The effects of sowing and spacing on rosette disease of groundnut in northern Nigerai, with observations on the vector, Aphis craccivora. Annals of Applied Biology 52:125-131.
9. Davies, J.C. 1975. Use of menazon insecticide for the control of rosette in Uganda. Tropical Agriculture (Trinidad) 52:359-367.
10. Farrell, J.A.K. 1976. Effects of groundnut crop density on the population dynamics of Aphis craccivora Koch (Hemiptera, Aphididae) in Malawi. Bulletin of Entomological Research 66:317-329.
11. Freeman, H.A., Nigan, S.N., Kelly, T.G., Ntare, B.R., Subrahmanyam, P. and Boughton, D. 1999. The world groundnut economy, facts, trends, and outlook. International Crops Research Institute for Semi-Arid Tropics, Patacheru 502 324 Andrha Pradesh, India.
12. Murant, A.F., Rajeshwari, R., Robinson, D.J. and Raschke, J.H. 1988. A satellite RNA of groundnut rosette virus that is largely responsible for symptoms of groundnut rosette diseases. Journal of General Virology 69: 1479-1486.
13. Murant, A.F. 1989. Groundnut assistor virus. CMI/AAB Description of Plant Viruses, No 345.
14. Murant, A.F. 1990. Dependence of groundnut rosette virus on its satellite RNA as well as on groundnut assistor luteovirus for transmission by Aphis craccivora. Journal of General Virology 71: 2163-2166.
15. Murant, A.F. and Kumar, I.K. 1990. Different variants of satellite RNA of groundnut rosette virus one responsible for the chlorotic and green form of groundnut rosette disease. Annals of Applied Biology 117:85-92.
16. Nigam, S.N. and Bock, K.R. 1990. Inheritance of resistance to groundnut rosette virus in groundnut (Arachis hypogea L.). Annals of Applied Biology 117:553-560.
17. Naidu, R.A., Kammins, F.M., Deom, C.M., Subrahmanyam, P., Chiyembekeza, A.J. and van de Merwe, P.J.A. 1999. Groundnut rosette: A virus disease affecting groundnut production in sub-Saharan Africa. Plant Disease 83:700-709.
18. Olorunju, P.E., Kuhn, C.W., Demski, J.W., Misari, S.M. and Ansa, O.A. 1991. Disease reactions and yield performance of peanut genotypes grown under groundnut rosette and rosette-free field environments. Plant Disease 75: 1269-1273
19. Olorunju, P.E., Kuhn, C.W., Demski, J.W., Ansa, O.A. and Misari, S.M. 1992. Inheritance of resistance in the peanut to mixed infection of groundnut rosette virus and groundnut assistor virus and a single infection of GRV. Plant Disease 76:95-100.
20. Rajeshwari, R., Murant, A.F. and Massalki, P.R. 1987. Use of monoclonal antibody to potato leaf roll virus for detecting groundnut rosette assistor virus by ELISA. Annals of Applied Biology 111:353-358.
21. Sauger, L. and Catharinet, M. 1954. La rosette chlorotique de l’arachide et les lignées sélectionnées. Bulletin Agronomie Ministere France Outremer 13:163-180.
22. Subrahmanyam, P., Hildebrand, G.L., Naidu, R. A. and Reddy, J.L. 1998. Sources of resistance to groundnut rosette disease in global groundnut germplasm. Annals of Applied Biology 132:473-485.
23. Taliansky, M.E., Rybov, E.V. and Robinson, D.J. 1996. Two distinct mechanisms of transgeneic resistance mediated by groundnut rosette virus Satellite RNA sequences. Molecular Plant Microbe Interaction 11:367-374.
24. Yayock, J.Y., Rossel, H.W. and Harkness, C. 1976. A review of the 1975 groundnut rosette epidemic in Nigeria. Samaru Conference Paper 9. Institute for Agricultural Research (Samaru). Ahmadu Bello University, Zaria, Nigeria, 12pp.