Previous Page Table of Contents Next Page


Rhodes grass breeding in Zimbabwe: aims, achievements, prospects and route to agricultural application

P.A. York1 and E. Nyamadzawo2

1Grasslands Research Station, P. Bag 3701, Marondera, Zimbabwe.
2Henderson Research Station, P. Bag 2004, Mazowe, Zimbabwe.


Introduction
Breeding habit
Germplasm assessments
Variety assessment
Breeding programme
Prospects
Route to agricultural application
References

Introduction

Sir George Stapledon (1940; 1942) founder of the Welsh Plant Breeding Station and William Davies (1952) advanced the concept of grass as a crop. The acceptance of grass as a crop with concommitant improved management are fundamental prerequisites to the development of grass through breeding effort. Put simply, improved seed has to be sown whether as long or short term leys or as permanent pasture to be of benefit.

The PANESA region has seen sporadic grass selection work in Zambia and Tanzania (Van Rensburg, 1969) and Kenyan breeding input (Bogdan, 1969) and more recently great improvement in Kenyan cultivars (Boonman, 1978). In developed countries pasture breeding has benefited from longer term involvement. The highly competitive nature of grass breeding in EEC countries for example is witness to the grass crop precept.

Grassland utilisation in Zimbabwe ranges from the grazing of unimproved veld through to intensive irrigated pastures based on Kikuyu grass. The most extensively sown pasture grass is Katambora Rhodes grass (Chloris gayana Kunth). The Rhodes grass breeding project in Zimbabwe arose as a response to the identification of important regional pasture research topics in an IDRC-Pasture workshop in Harare in 1984 (Clatworthy, 1985).

Katambora is grown in rotation with tobacco to control Meloidogyne javanica (root-knot nematode). A four year ley is recommended (Martin, 1967) to overcome the longevity of nematode eggs even in bare fallow soil. Potential area of use is in the region of 200,000 to 250,000 ha with annual reseeding of up to 60,000 ha.

Katambora, a diploid Rhodes grass, came into use in the nineteen fifties in Zimbabwe (West, 1952). It is less productive and palatable than tetraploid strains such as Zimbabwe Giant. The more productive cultivars are less effective in controlling root-knot nematode (Shepherd, 1968; York, 1989 a).

Primary aim of the breeding project is the inclusion of good nematode resistance in a variety with greater forage production. For a grass used in arable rotation requiring resowing every fifth year or more frequently, seed yield is also an area of concern. A valuable export market for Katambora seed magnifies this requirement.

Aimed initially at the commercial sector in Zimbabwe, the successful combination of yield and root-knot resistance could encourage the use of grassland/arable rotation in communal farming. Such a variety would serve a multiple role in controlling root-knot nematode, improving soil structure, reducing erosion as well as providing valuable forage and seed as a cash crop.

Aspects of the breeding project have included hybridization studies, variety and germplasm assessment agronomic, nematological and flowering behaviour. The breeding programme has two separate approaches; interploid hybridization between Katambora and tetraploid Rhodes grass, and enhancement of nematode resistance and other characters within tetraploid strains.

Breeding habit

Rhodes grass has been variously described as outcrossing Bogdan (1961) and apomictic (Hutton, 1961). Barnard (1971) described Katambora as 'probably apomictic'. However the stock for which Hutton claimed apomixis although diploid was not Katambora, but commercial (Pioneer) which according to Hutton was similar to Nzoia - this appeared uniform in Marondera nursery and had good nematode resistance (York, 1987; 1989 b). Later Australian work discounted apomixis in various material (Jones & Pritchard, 1971). Hybridization studies at Marondera (York, 1987) using cellophane bags to effect selfing and crossing only resulted in seed set when Katambora flowers of different plants were bagged together. Outcrossing was clearly demonstrated in tetraploids by Bogdan (1963).

Germplasm assessments

A range of diploid and tetraploid Rhodes grass cultivars and collections has been obtained from various sources. An observation nursery was established at Marondera in 1987. Plants were spaced at 1/5m and kept distinct by trimming. Accessions showed inter and intravarietal variation in vigour, habit, foliage type and flowering rate. Flowering date was amongst the most quantifiable characters. Table 1 shows the range of flowering behaviour observed.

The range of variation exhibited by most accessions supports the normal outcrossing mode of reproduction for this species. Boonman (1978) showed that early flowering resulted in improved seed yield and was related to vigour in Kenya. Earliness was sufficiently heritable for selection purposes and led to the creation of Elmba from Mbarara and Boma from Masaba.

Table 1: Summary of flowering data of 81 C. gayana accessions at Marondera as days to first flower emergence from 1.1.87.

* Plants with first flower emergence later than 6/4 not recorded
e, 1 = earliest and latest plant of accession
r, 0 overall and mean range respectively n= number of accessions

Low night temperatures have been associated with poor seedset in Rhodes grass (Loch & Butler, 1987). In Zimbabwe low night temperatures (below 10º C) occur in many parts of the tobacco - hence Rhodes grass - areas by mid-April. Ideally flowering should be timed so that the majority of seed has been set and filled by this date, mainly hardening thereafter. Katambora is managed so that it flowers in a flush and is harvested late April/early May. Tetraploid varieties which may be quantitative short day plants (Dirven et al. 1979; Loch, 1984) tend to flower later and reach a peak more slowly. The data from the observation nursery showed as wide a range in date of first flower emergence amongst tetraploid plants as diploid; some could flower as early as Katambora. Photoperiod differences aside, it should be possible to select for earliness and improved seed yield as achieved in Kenya.

Variety assessment

Agronomic Characters: Field trials were established in December 1986 to compare the forage yield, flowering behaviour and seed production of eight tetraploid varieties with those of diploid Katambora. Sites were Chiredzi, which was irrigated to give total water similar to the other sites, Marondera and Mazowe. Forage yields for the year after establishment are given in Table 2. Tetraploids generally outyielded Katambora as expected. The better varieties overall were Elmba, Boma and Mt. Makulu 56. The greatest difference between tetraploids and Katambora was 80%; average superiority of tetraploids was in the order of 30%. The varieties retain a large degree of plant to plant variation. Taking individual plants during screening for nematode resistance would results in 30-40% yield gain over Katambora and possibly more. The forage productivity attained compares favourably with 1819 t DM/ha in more heavily fertilized trials (Rodel, 1969) with Giant in Zimbabwe. At a seasonal at N fertilisation rate of 120 kg/ha Chiredzi site gave over 20 t DM/ha from some tetraploids.

Table 2: Dry matter yield of 9 Rhodes grass varieties at 3 sites in Zimbabwe during the growing season 1987/88.


kg/plot


Mean t/ha

Variety/Site

Chiredzi

Mazowe

Marondera


Katambora

41

34

37

13.6

Mt. Makulu

60

42

43

19.6

Samford

54

39

39

17.6

Callide

54

41

37

17.4

Giant

53

36

37

16.8

Mbarara

57

43

42

18.9

Elmba

52

52

49

20.4

Masaba

53

43

43

18.5

Boma

59

42

43

19.2

S E 5%

3.5

2.9

2.7

-

Seed yield data are given in table 3 and flowering progress in Figure 1. The tetraploid cultivars and strains showed poorer seed content than Katambora throughout, but the potential seed yield is very high as shown by crude yield at Marondera. All heads were taken which tended to depress seed content unrealistically compared to farm practice. Given this, the true seed yield of Katambora and Callide were reasonable. Late harvesting favoured the tetraploids over Katambora. Germination of rubbed and extracted caryopses was high for all varieties. 1988 was a good year for tetraploid seed - one local farmer obtained PLSC (pure live seed content) in excess of 30% from his Giant Rhodes grass.

Table 3: Seed yield of 9 Rhodes grass varieties at Marondera


g/plot


Variety

Crude seed

*Caryopses

'PLS'

Katambora

516

106

20.5

Mt Makulu 56

474

21

4.4

Samford

509

42

8.0

Callide

785

106

15.3

Giant

551

61

10.9

Mbarara

250

32

12.9

Masaba

348

17

4.9

Elmba

288

10

3.5

Boma

491

17

3.2

S E 5%

78.0

9.8

-

*150 kg @ 25% PLSC - 94 g/plot
PLSC = Pure Live Seed Content

Similar flowering patterns were obtained at the three sites. The earlier flower flush of Callide at Marondera largely explains the better seed production of this compared with other tetraploid varieties. The locally adapted Giant also showed an earlier flower flush than other tetraploids. The varieties Elmba and Boma selected on the basis of their flowering pattern in Kenya (Boonman, 1978) were not different from their respective progenitors, Mbarara and Masaba. The similarity of the Kenyan cultivars may result from genetic drift since their release (Boonman, pers. comm.) or may be due to photoperiod differences between their selection site in Kenya and Zimbabwe. The Kenyan varieties were consistently later than Callide and Giant: selection on the basis of earliness in one environment need not lead to earliness in another. The performance of Callide gives some hope that even a small improvement in date of flower flush will give a boost to the seed content of other tetraploid lines in Zimbabwe.

Figure 1. Flowering progress of 9 Rhodes grass cultivars at Marondera in 1988 based on semilogarithmic assessment of numbers of flowers. means of 4 replicates.

Nematode Host Status Assessments: The susceptibility of a range of diploid and tetraploid Rhodes grass accessions was compared using nematode eggs in suspension as inoculum (York, 1989 b). The level of nematode eggs and range of response are presented in Table 4. Generally tetraploid accessions comprised more susceptible plants with susceptibles being more extreme than among diploids. Some tetraploids were less susceptible than others but all allowed more nematode reproduction than Katambora. The better forage yielder, Elmba was the most nematode susceptible. The distribution of host status suggested that susceptibility at least was under polygenic control. The number of egg-masses developed was the major determinant of susceptibility, and size of root system was not a factor in number of egg-masses produced. Although diploids supported fewer egg-masses per plant on average and more plants were completely resistant than with tetraploids, this generalization was not strong enough to recommend diploid varieties for root-knot control.

Breeding programme

Interploid Hybridization: The basis for this approach is the genetic affinity of diploid and tetraploid Rhodes grass. Japanese cytological studies indicate a degree of autoploidy (Nakagawa and Sato, 1981) hinted at by earlier workers (Moffett 1944). The occurrence of a natural triploid is also encouraging (Pritchard & Gould, 1964). Interploid hybridization has proved possible with other autoploids, e.g. cocksfoot (Dactylis glomerata) (Caroll & Borill, 1965). The sufficient homology of genomes and the 'spotaneous' non-reduction or restitution of diploidy in gametes from diploid plants allow the formation of tetraploid zygotes by fusion with normally reduced gametes from the tetraploid. Other configurations may occur especially triploids which would require backcrossing to restore the tetraploid condition.

The usefulness of this approach depends on the existence of a recessive genetic marker stock anthocyanin-free (Bogdan, 1963) which will facilitate the identification of hybrids without recourse to labourious cytological techniques. The anthocynanin-free flowers are distinctively 'yellow-heads'.

Seedling bases are also free of purple pigmentation in this stock, but Katambora is normally purple pigmented, so that purple based seedlings from a 'yellow head' stock has been isolated at Marondera. Hybridization in cellophane bags between Katambora and anthocyanin-free tetraploid plants suggests that 1% of caryopses formed on the tetraploid may be interploid (York, 1987).

To be of value, clearly large numbers of crosses have to be attempted. This is being accomplished on a field scale this season. There are two advantages to this approach; i. if the cross is successful a resistant yet anthocynanin-free line may be selected from the heterozygous stock - in the absence of adverse linkage. This would permit the ready identification of resistant material in seed certification schemes; ii. The 'yellow head', stock was produced by selfing and although itself vigorous, a small proportion of the progeny when the stock is intercrossed are albino. Inbreeding depression would result rapidly if selection for resistance within this were attempted, especially as the 'yellow heads' are very nematode susceptible (York, 1989 c). Interploid hybridity would release some variation and vigour.

Selection for Root-knot Resistance in Tetraploid Material: Screening tests at the Tobacco Research Board of Zimbabwe showed that the Zambian strain Mt. Makulu 56 (Van Rensburg, 1969) may be less susceptible than some tetraploid varieties (Way, pers. comm). This variety was assessed during 1987 to determine the proportion of resistant plants and degree of susceptibility of individual plants. Figure 2 shows the frequency distribution obtained. 14% of the parent stock were devoid of egg-masses 12 weeks after inoculation with 5900 eggs of M. javanica (York, 1989a). 37% of the population would allow maintenance or increase the nematode population. The resistant plants have formed the basis of a mass and recurrent selection programme to increase the proportion of resistant plants in this tetraploid background.

After two selection cycles the proportion of resistant plants as measured by mf (ratio egg-masses/inoculum egg-mass equivalents) is as shown in Figure 3. There has been a significant increase in the proportion egg-mass free plants. Some plants still support more than maintenance level of nematode reproduction. The rapid increase in proportion of resistant plants suggests that resistance itself is governed by fewer genes.

The selected susceptible stock shows a greater degree and wider range of susceptibility than the parent stock. Already the resistant selection is comparable with Katambora. Further selection will be undertaken to increase and fix resistance. Progeny of paired hybridizations of plants of known nematode susceptibility/resistance will be screened to investigate genetics of resistance.

From the first selection cycle, plants showed a range of flowering rates as demonstrated in Figure 4. 10% of these plants were not significantly different from Katambora in date of first flower emergence. Earlier plants had significantly more flowering tillers at 90 days (from 1.1.88) than later plants. The proportion of resistant plants among progeny of maternal plants selected for earliness and vigour showed some variation, but overall the level of resistance was comparable with that of the bulk selection. There was no apparent adverse correlation between nematode resistance and earliness. The results indicate that root-knot resistance and early flowering can be combined in a tetraploid background.

Figure 2. % Frequency distribution of host status of Mt. Makulu 56 plants expressed as egg-masses per plant.

Figure 3. % Frequency distribution of susceptibility of parent stock (P) and resistant (R) and susceptible selection (S) from Mt. Makulu 56 a mf (ration eff-masses/inoculum).

Figure 4. Flowering of selected root-knot resistant lines from Mt. Makulu 56.

Prospects

The desired combination of root-knot resistance, at least as effective as that of Katambora, improved seed quality through earlier flowering and a yield superiority of a minimum 40-50% over Katambora seem quite achievable. What is required is sufficient time to take selection for all traits to the optimum level and for fixation of these characters.

Route to agricultural application

The keenness of commercial farmers in Zimbabwe to improve on the forage production of grass in the tobacco rotation without sacrificing root-knot control will ensure the popularity of a new cultivar with the above features. There is legislation on Plant Variety Rights and on seed Certification in Zimbabwe. However, the most widely sown pasture cultivar, Katambora is subject to an 'informal' scheme. Control of seed quality is excellent, but the view that Katambora was apomictic led to relaxed standards of seed increase on the assumption that all stocks were identical. It is pretty uniform but was never stabilized for root-knot resistance; stocks may differ to some degree in resistance, although Katambora generally is more resistant than tetraploid cultivars.

The successful product of this breeding programme will demand more careful management. Mother plants of a resistant tetraploid will be established to give Breeder's Seed, some of which will go to long term storage. Mother plants will be propagated vegetatively periodically and give rise to successive sequences of Prebasic, Basic and Certified seed as with temperate grasses (Anon, 1978). An improved outcrossing variety must be multiplied through as few generations as practicable from its basis to seed for farm use. This is more critical in a variety incorporating specific pest resistance: Seed certification and documented provenance of seed will be essential. To capitalise on breeding gains tight control will remain necessary throughout the useful lifespan of the variety. To ensure resistance is not lost periodic monitoring of host status of different generations of seed stocks to M javanica is advisable. This should be coupled with virulence tests of different M. javanica populations to detect resistance breaking races as soon as possible.

Table 4. Host status of C. gayana to M. javanica.



Diploids



Tetraploids

Accession

n

Mean egg-masses/ plants

% plants over 5 egg masses

Accession

n

Mean egg-masses/ plant

% plants over 5 egg masses

2

24

36

63

34

23

33

74

12

25

14

60

38

23

17

61

21

19

7

21

40

26

43

69

29

20

3

20

Giant

7

86

86

52

26

16

39

Mbarara

27

46

89

53

28

20

43

Masaba

23

43

87

54

24

20

50

Elmba

23

90

92

55

23

20

61

Boma

23

55

74

56

24

17

50

Callide

25

49

80

57

10

15

60

Samford

22

27

50

74

25

2

41

Mt Makulu 56

245

45.8

76

Short term funding of such projects can result in outright loss of material - few of van Rensburg's selections are available now-or reduced value of material e.g. through inadequate seed isolation after varietal release as occurred in Kenya (Boonman, pers. comm.). Sustained breeding effort with pasture species could be done on a co-operative regional basis with longer term joint funding or external support.

References

Anon, 1978. Principles of herbage seed production. Welsh Plant Breeding Station, University College of Wales, Aberystwth, Wales, U.K.

Barnard, C. 1972. Register of Australian herbage plant cultivars. pp 99-100. Division of Plant Industry, CSIRO, Canberra, Australia.

Bogdan, A.V. 1969. Herb. Abstr. 39:1-3. Bogdan, A.V. 1963. Chloris gayana without anthocyanin colouration. Heredity, London, 18:364-368.

Bogdan, A.V. 1961. Intravariety variation in Rhodes grass (Chloris gayana Kunth) in Kenya. J. Brit. Grassld. Soc. 16:238-239.

Boonman, J.G. 1978. Rhodes grass breeding in Zimbabwe. III Seed and herbage yield in selections of four maturity classes based on intravariety variation. Euphytica 27. 649-656.

Carroll, C.P. and Borrill, M. 1965. Tetraploid hybrids from crosses between diploid and tetraploid Dactylis and their significance. Genetica 36:65-82.

Clatworthy, J.N. 1985. Pasture research in Zimbabwe, 1964-1984. In Pasture Improvement Research in Eastern and Southern Africa. Ed. J.A. Kategile. Proc. IDRC Workshop Harare, Zimbabwe 17-21, Sept. 1984. pp. 25-58.

Daulton, R.A.C. 1963. Controlling M. javanica in Southern Rhodesia Rhod. Agric. J. 60:150-152.

Davies, W. 1952. The grass crop: its development use and maintenance. E & F.N. Spon Ltd. London, U.K.

Dirven, J.G.P., van Soest, L.J.M., and Wind, K. 1979. The influence of photoperiod on head formation in some Brachiaria species and Chloris gayana cv. Masaba. Neth. J. Agric. Sci. 27:48-49.

Hutton, E.M. 1961. Intervariety variation in Rhodes grass (Chloris gayana Kunth). J. Brit. Grassld. Soc. 16:23-29.

Jones, R.J. and Pritchard, A.S. 1971. The method of reproduction in Rhodes grass (Chloris gayana Kunth). Trop. Agric. (Trinidad) 48, (4), 301-307.

Loch, D.S. 1984. Constraints on seed production of Chloris gayana cultivar, Ph D. Thesis, University of Queensland, Australia.

Loch, D.S. and Butler, J.E. 1987. Effects of low night temperatures on seed set and seed quality in Chloris gayana. Seed Sci. & Technol. 15. (3). 593-597.

Martin, G.C. 1967. Longevity of Meloidogyne javanica under conditions of bare fallow in Rhodesia. Rhod. Agric. J. 64:112-114.

Moffett, A.A. 1944. Note on the cytology of Rhodes grass. Rhod. J. Agric. 41:11-13.

Nakagawa, H. and Sato, H. 1981. Cytological studies on tropical grasses 1. Meiosis of pollen mother cells and the formation of pollens of Rhodes grass (Chloris gayana Kunth). Bull. Kyushu Nat. Agric. Expt. Sta. 21:317-331.

Pritchard, A.J. and Gould, K.F. 1964. Chromosone numbers in some introduced and indigenous legumes and grasses. CSIRO Div. Tropical Pastures Tech. paper. No. 2, CSIRO, Australia.

van Rensburg, H.J. 1969. Selection of productive strains of Chloris gayana in Zambia. Gvt. Printer, Lusaka, Republic of Zambia.

Rodel, M.G.W. 1969. The effect of applying nitrogen in various ways on the herbage yield of Giant Rhodes grass (Chloris gayana Kunth). Rhod. agric. J. 66:43-45.

Shepherd, J.A. 1968. A nematode survey of tobacco soils in Rhodesia and Zambia and the effects of grass-tobacco rotations on nematode populations. Rhod. J. agric. Res. 6. (1). 19-26.

Stapledon, R.G. 1940. Regrassing an essential part of food production. (1942) Ley farming. War food production advisory bulletins No 1 and 2 - University College of Wales, Aberystwth, Welsh Plant Breeding Station. Cambrian News Ltd. Aberystwth, Wales, U.K.

West, O. 1952. Promising new grasses for selected pastures in Southern Rhodesia, Rhod. Agric. J. 49:89-95.

York, P.A. 1987. Rhodes grass breeding. Ann. Rep. Div. Livestock & Pastures, Dept. Res. 8 Specialist Services, Zimbabwe.

York, P.A. (in press, a). Resistance to Meloidogyne javanica (root-knot nematode) in Chloris gayana (Rhodes grass). Nematologica 1989 in print.

York, P.A. (in press, b). Range of susceptibility within and between deploid and tetraploid strains of Chloris gayana (Rhodes grass) to Meloidogyne javanica (root-knot nematode). Revice de nematologie.

York, P.A. (in press, c). Progress and prospects for combining resistance to Meloidogyne javanica (root-knot nematode) with improved forage yield in Chloris gayana (Rhodes grass) strains. Proc. XVIth Int. Grassld. Cong. Nice 1989.


Previous Page Top of Page Next Page