Selection for resistance to fungal diseases and other desirable traits in kikuyu grass (Cenchrus clandestinus)

While kikuyu (Cenchrus clandestinus) is an important grass for dairy and beef production in the subtropical region of Australia and the world, the most common cultivar, Whittet, is seriously affected by the fungal diseases, kikuyu yellows (Verrucalvus flavofaciens) and black spot (Bipolaris spp.). Thus resistance to these diseases is a priority in selecting a better kikuyu cultivar, along with higher herbage quality and yield and better winter growth. A study was conducted to identify suitable candidates from kikuyu ecotypes collected along the east coast of Australia plus lines obtained by subjecting Whittet to a mutagenic agent. Initial glasshouse studies identified 19 lines that were resistant to the KY1A strain of kikuyu yellows and 4 of these, with forage quality and yield superior to Whittet, were further evaluated in the field at 2 sites using Whittet as the control. At Site 1, line 12A demonstrated a much higher level of resistance to kikuyu yellows than Whittet, with 85% of plants resisting infection compared with only 15% of Whittet plants. At Site 2, the numbers of 12A and Whittet plants infected were similar. Further tests, using kikuyu yellows inoculum collected from 11 sites along the east coast of Australia, found that only 15% of 12A plants became infected compared with 61% of Whittet plants. Thus, kikuyu line 12A was resistant to most, but not all, strains of the kikuyu yellows pathogen. Annual yield of 12A (19,008 kg DM/ha) was 24% higher than that of Whittet and 12% higher than Acacia, but the difference was significant only for Whittet. During summer, 12A produced 10,212 kg DM/ha (24% higher than Whittet), was more active in early spring, had slightly higher dry organic matter digestibility (66.7 vs. 64.0%) and was resistant to black spot infection.


Introduction
Kikuyu grass (Cenchrus clandestinus) is a C4 summer grass, native to northeastern parts of Africa, predominantly in the Aberdare Mountain region of Kenya, which is inhabited by the Kikuyu tribe; hence the name, kikuyu. It is found at elevations between 1,950 and 2,700 m in the tropics, creating a niche subtropical climate with maximum temperature of 16-24 °C and a minimum of 2-8 °C (Morrison 1969), which led Pearson et al. (1985) to conclude that kikuyu had an optimal temperature range for growth of 8-21 °C. However, kikuyu is very adaptable, having become naturalized in areas with far more extreme temperatures, such as South Africa and Australia, where it grows successfully from the tropical highlands of the Atherton Tableland (North Queensland) to East Gippsland, Victoria (temperate climate) and the irrigated pastures of southwest Western Australia (Mediterranean climate).
'Common' kikuyu grass, introduced into Australia in 1919, was traditionally established by cuttings until Wilson (1968) identified a seeding line of kikuyu grass at Grafton Research Station in northern NSW. This line was primarily selected for its high level of free seeding and more vigorous growth than Common kikuyu grass, particularly in winter, and was registered as cultivar Whittet in 1970.
Three more lines were subsequently selected and registered: 1. 'Breakwell', selected for its ability to set seed but similar to Common kikuyu grass in other aspects, was registered in 1971. 2. 'Crofts' is claimed to have greater cold tolerance and to be slightly higher yielding than Whittet and was released in 1983.
3. 'Noonan' was selected for resistance to kikuyu yellows (Verrucalvus flavofaciens) (Wong and Wilson 1983) and was registered in 1983 but is now used mainly as a turf grass. 4. 'Acacia' selected from the Acacia Plateau at high elevation in the subtropical region was registered in 2013.
A survey in 1994 (Anonymous 1994) revealed that kikuyu grass was the basic forage for 30% of dairy pastures in NSW and it was estimated that 70% of milk production in summer came from kikuyu pastures. However, since then, the spread of the fungal diseases, kikuyu yellows and, to a lesser extent, black spot (Bipolaris spp.), inter alia, have reduced its contribution. When Mears (1970) reviewed kikuyu grass in 1970, the adverse impact of kikuyu yellows on kikuyu pastures was already evident. He concluded that "improvement (of kikuyu grass) should be directed towards disease resistance (e.g. kikuyu yellows), possibly higher digestibility and cold tolerance rather than vigour or free seeding". The dairy industry still considers these traits are top priority for selecting a better kikuyu cultivar.
Resistance/tolerance to fungal diseases Wong (1975) found that an undescribed fungus, subsequently described as a new species, Verrucalvus flavofaciens, by Dick et al. (1984), caused kikuyu yellows. It is spread by water-borne zoospores which infect the roots, resulting in death of the grass in patches. No commercial fungicides are currently available to control kikuyu yellows effectively.
With black spot disease, the tips of the more mature leaves of infected kikuyu plants often turn yellow and numerous dark brown to black spots appear on the leaf blades and sheaths, especially during periods of high humidity and rainfall. Normally, black spot does not kill the plant, but growth is slightly reduced, and more importantly, the leaves become chlorotic and unpalatable to cattle.

Tolerance of low temperatures
In the subtropical dairy region of NSW, kikuyu pastures are typically over-sown with short-rotation ryegrass in autumn to provide winter feed, when kikuyu growth is low (Fulkerson et al. 2010). This over-sowing is costly and results can be variable, dependent on weather. Finding a kikuyu genotype more tolerant of winter cold would eliminate the need to sow ryegrass or shorten the period when it was needed.

Improved forage quality
Selection for improved forage quality, specifically digestibility, is also a common dairy industry request, as a means of achieving higher milk production. While ryegrass can produce 20 L milk/cow/day with good management on-farm, Reeves (1998) found cows grazing kikuyu during mid-lactation produce about 15 L milk/day, i.e. at a time when virtually all milk is coming from feed and not body reserves.
Based on these factors, the present study aimed to identify a kikuyu line with one or more of the following desirable traits: resistance to fungal diseases, improved herbage quality, greater tolerance of winter cold and higher yield than cv. Whittet, in that order of priority..

Materials and Methods
There were 4 stages in the evaluation of available genetic material of kikuyu as outlined below: Stage 1 -Yield, forage quality and resistance to kikuyu yellows of 115 kikuyu lines, as single plants, were evaluated.
Stage 2 -Kikuyu plants identified as resistant to kikuyu yellows in Stage 1 were tested in the field, using Whittet as the control.
Stage 3 -The kikuyu line, 12A, identified as resistant to kikuyu yellows in Stage 2, and the commercial kikuyu cultivar, Whittet, were further tested for resistance to kikuyu yellows by using inoculum from kikuyu yellowsinfected plants, collected over a wide geographical region.
Stage 4 -The 2 commercial cultivars, Whittet and Acacia, plus line 12A were evaluated against 6 lines found to be highest in other desirable traits (yield and forage quality) during Stage 1.

Stage 1 -Initial screening
Material tested. The study commenced in March 2015 with the evaluation of 115 kikuyu lines, selected from 1,600 lines held at the Animal Science Precinct, University of Queensland from a previous study on the genetic variation in kikuyu grass (Lowe et al. 2010 indigestible NDFom (iNDFom) (Harper and McNeil 2015) at the School of Agriculture and Food Sciences, University of Queensland, Gatton Campus, Queensland. The samples were analyzed for iNDFom using long-term (10 days) in vitro fermentation. This procedure is based on a fermentation procedure described by Goering and Van Soest (1970), adapted for use with Ankom filter bags using ANKOM Daisy incubators. Neutral detergent fiber concentration was determined using the procedure described by Goering and Van Soest (1970) and adapted for an Ankom fiber analyzer. • Dry organic matter digestibility (DOMD) by the rumen fluid fermentation method (Tilley and Terry 1963) at the NATA-accredited feed testing laboratory, Department of Primary Industries, Charles Sturt University, Wagga Wagga, NSW. The mean, standard deviation (SD) and range of the results for the 115 kikuyu lines were: NDFom 62.4% (±3.8) (range 54.3-71.9%); and iNDFom 20.7% (±4.6) (9.2-29.9%). Resistance to kikuyu yellows infection. The sensitivity of the kikuyu plants to kikuyu yellows was tested in a glasshouse at the Plant Breeding Institute, University of Sydney, Cobbitty, NSW, using kikuyu yellows-infected kikuyu leaves as described by Wong (2011). Where there was any doubt that the yellowing symptoms were due to the disease, yellow leaves were collected at random to reisolate the pathogen and to serve as a check on the method. This method was also used to confirm kikuyu yellows infection in the regional samples in Stage 3. Nineteen kikuyu lines showed resistance to the KY1A strain of kikuyu yellows, one of the 3 strains previously identified by Dr P. Wong and believed to be the most prevalent. Selected lines. Four of these kikuyu lines (Table 1), with forage quality (NDFom, iNDFom and DOMD) at least as good as Whittet, were chosen for further evaluation against Whittet, as the control, in Stage 2. The origins of these 5 lines were as follows: • Lines 15A, 11C and 12A were unidentified plants, which originated from the combined Queensland and Wollongbar collections. • Line 25D was the only identifiable plant and had shown some resistance to kikuyu yellows in the initial field evaluation at Wollongbar (Dairy Australiafunded project, DAN063).

Stage 2 -Field evaluation of kikuyu lines for resistance to kikuyu yellows
Propagation and establishment. Cultivar Whittet plus the 4 kikuyu lines selected for the field evaluation were propagated vegetatively from the original 'mother' plants by means of rooted stolons. These were planted out on 20 December 2015 in 1 m 2 plots, forming a grid of 5 lines of kikuyu grass with 5 plants/line, replicated 4 times at 2 locations (Site 1 -Kyogle: 28° S, 153° E; 76 masl; and Site 2 -Lismore: 29° S, 153° E; 12 masl). This grid structure facilitated weeding and helped to identify lines and prevent incursion between lines through stolon growth. The soils were clay loams on a flood-plain aspect. The climate at both sites is subtropical with Site 1 having lower rainfall and humidity but higher summer temperature than Site 2. From planting to the completion of the study on 30 May 2016, Site 1 received 323 mm of rain and Site 2 had 406 mm.
At planting, the areas where the plots were located, showed obvious signs of current kikuyu yellows infestation (stunted growth, yellowing of the leaves, short stolon internodes and little root development, making the plants easy to pull from the soil). The soils at both sites were highly fertile with high original soil nitrate levels (>80 mg/kg). In order to stress the plants, in an effort to maximize active kikuyu yellows infestation, supplementary irrigation was applied only for the first 4 weeks after planting to enable the plants to establish effectively and no N fertilizer was applied. Confirmation of kikuyu yellows status. On 16 March 2016, stolon and leaf samples were taken from one 'yellow' plant of each kikuyu line in each replicate as well as from the 'green' 12A plants. All samples were tested for kikuyu yellows infection, to confirm or reject the visual symptoms, using the following methods: • Plating out of leaf pieces on water agar plates after surface sterilization (P. Wong, unpublished data); and • Floating leaf pieces of kikuyu in soil extract after surface sterilization (Wong 1975).

Stage 3 -Resistance to disease inoculum from the north coast of NSW and SE Queensland
In order to determine the geographical extent of the resistance of line 12A to kikuyu yellows relative to Whittet, kikuyu leaves from plants, which were obviously affected by kikuyu yellows, were collected from 11 dairy pastures located along the north coast of NSW and SE Queensland and used as a source of inoculum (Wong 2011). In this experiment, 3 replicates of each inoculum were tested against 12A and Whittet. The plants were kept in a glasshouse at 30:25 °C (day:night) temperature for about 3 months from mid-January. Kikuyu yellows infection usually became obvious after 5-7 weeks. One tiller or leaf from each plant, suspected to be infected, was tested to confirm kikuyu yellows, as outlined previously.

Stage 4 -Field evaluation of kikuyu lines for yield, forage quality, stolon vigor, black spot susceptibility and cow preference
This plot experiment was undertaken at the same location as used in Stage 2, i.e. Site 2 (Lismore). The climatic data during Stage 4 are shown in Figure 1. Six kikuyu lines selected in Stage 1 for high quality (DOMD, NDFom and iNDFom) or yield, as individual plants were selected for evaluation in this stage (Table 2).  The commercial cultivars, Whittet and Acacia, and kikuyu line 12A, which was shown to be resistant to kikuyu yellows in Stages 2 and 3, were also included in the experiment (see Table 1 for quality and yield data).
Parameters measured were: seasonal yield, forage quality, stolon vigor, level of black spot infestation, winter growth and cow preference, in 2 × 1 m field plots, randomly allocated within 3 replicates. The evaluation continued for 11 months (24 November 2016-24 October 2017). The plants were vegetatively propagated from the original 'mother' plants in August 2016 and transplanted into the plots on 6 October. Recording of yield and other traits commenced with the first recorded defoliation on 8 December 2016.
Site preparation. Lime and a complete fertilizer (Rubisca plus at 350 kg/ha containing N:P:K:S at 12:5:14:9.7 mg/kg plus trace elements) were applied to the total plot area before planting to ensure that soil pH and soil nutrients [Colwell P -110; exchangeable K -227; S -30 in mg/kg; pH (CaCl2) -5.1] were not limiting kikuyu growth. Urea was then applied after each defoliation at a rate equivalent to a per-month recommendation of 100 kg/ha.
Methodology and chemical analysis. The plots were cut to 6 cm stubble height with a Massport hand push mower each time the plants had regrown 4-4.5 new leaves/tiller (after approximately 14 days in 'summer' extending out to 35 days or more in 'winter'), i.e. the stage when herbage quality and yield are optimized (Reeves 1998). The 'green' plot yield and weight of a 'green' subsample were recorded before the green subsample was dried at 65 °C for 48 h in a forced-draft oven.
At the defoliations on 27 February, 24 April and 2 December 2017, herbage samples were taken for analyses of DOMD, NDFom and iNDFom (for details see Stage 1) and the number of stolons protruding from each plot was recorded as an indication of stolon vigor.
The presence of the fungal infection, black spot, was scored as: 0 (no infection), 1 (slight infection -4-6 yellow leaf tips/plot), 2 (moderate -more than half the leaf tips yellow) and 3 (severe -nearly all leaf tips yellow), prior to each defoliation from 4 December 2016 to 9 May 2017.
A cow preference test (Horadagoda et al. 2009) was performed 14 days after the final harvest of the experiment on 16 December 2017 (when 4 new leaves/tiller had regrown and at a time when there was no black spot infection). The test involved 2 observers recording which plots were being grazed every 10 seconds for 40 minutes by 3 Jersey milking cows, accustomed to grazing kikuyu pastures.

Stage 2 -Field evaluation of kikuyu lines for resistance to kikuyu yellows
The status of the kikuyu plants at Site 1 (Kyogle) on 20 April (130 days after planting) and at Site 2 (Lismore) on 31 May (171 days after planting) is shown in Table 3.  Observations on plots at Lismore were continued until May as a number of plants were seen to be recovering in April, whereas most plants that were infected with kikuyu yellows at Kyogle had died by 20 April. The photos below (Figure 2) show Replicate 2 at Site 1, 38 days after planting and at 130 days after planting, illustrating the impact of kikuyu yellows on the kikuyu lines. Table 3 shows the very marked differences at Site 1 between line 12A and the other 4 kikuyu lines which had all, apart from Whittet, shown resistance to the kikuyu yellows strain KY1A in the initial glasshouse tests in Stage 1. At Site 2, the proportion of plants infected with kikuyu yellows was substantially lower than at Site 1 and all lines recorded similar levels of infection, but a high percentage of infected plants recovered in late autumn. In addition, the onset of infection was later, which suggests the strain of kikuyu yellows at Site 1 was different from that at Site 2.
At Site 1, kikuyu yellows was isolated from 19 leaf/stolon samples that displayed symptoms of kikuyu yellows infection, confirming the presence of the disease, while 2 plants of 12A, identified as 'green' or symptomless, were confirmed as free from kikuyu yellows infection. At Site 2, 15 of the 18 'yellow' leaf/stolon samples were confirmed as infected and 2 leaves/stolons identified as 'green' were also infected with kikuyu yellows.

Stage 3 -Resistance to disease inoculum from the north coast of NSW and SE Queensland
In this test, only 1 or 2 plants of the 3 became infected, probably because some of the leaves originally collected and showing obvious symptoms of kikuyu yellows may not have had viable inoculum at the time of sampling. Therefore, it was not possible to say definitively at which sites 12A would be resistant to kikuyu yellows. However, the results support the proposition that 12A is far more resistant to kikuyu yellows than Whittet with only 5 plants (of 33 treated) becoming infected and 2 recovering, compared with 20 Whittet plants becoming infected, of which 6 recovered.

Stage 4 -Field evaluation of kikuyu lines for forage quality and other desirable traits
The forage quality of the 7 kikuyu lines and the 2 commercial cultivars used in this stage are shown in Table 4.
In 'summer', DOMD of kikuyu line 12A was 2.7% units higher than that of Whittet but the differences were not significant. The DOMD of line 18A was significantly higher than that of any other line except 3A. There were no significant differences in DOMD for the samples taken in 'autumn' and 'winter-spring'.   Within columns and parameters means without a common letter are significantly different at the levels indicated.
2 Analysis of kikuyu lines 15A (low quality and yield) and 7A and 14B (both highly affected by black spot fungus) were discontinued after the first analysis in summer owing to the issues indicated.
The NDFom of Acacia was significantly higher than that of all other kikuyu lines except Whittet in 'summer', while NDFom values of all kikuyu lines in 'autumn' and 'winter-spring' were similar.
The iNDFom in 'summer' was highest for Whittet and lowest for 18A and the difference was significant but, by contrast, the iNDFom of Acacia in 'autumn' was significantly higher than that of any other line, except 18A. There was no difference in iNDFom between lines in 'winter-spring'.
Dry matter yield. Forage yields for the 11-month period from 8 December 2016 to 24 October 2017 are shown in Table 5. Clearly, 12A was higher-yielding than both Whittet and Acacia during the main 'summer' growth period but not different from lines 2A, 7A, 18A and 14B. Line 12A also produced higher yields than Whittet in the 'winter-spring' period but the difference was not significant (P>0.05). Total yield for 12A was significantly greater than that for Whittet but not for Acacia nor lines 2A, 7A and 18A. Other desirable traits. The incidence of black spot and indications of stolon vigor for 12A, Whittet and Acacia were recorded during the 'summer' and 'autumn' periods from 22 December 2016 to 9 May 2017, while the cow preference test was recorded in early December 2017, when there were no symptoms of black spot infection and none was expected owing to the low relative humidity at that time (Table 6). Black spot infestation on 12A was zero and on Acacia was low but the mean score for Whittet was 1.6, being between moderate and severe at each defoliation over the 'summer' growth period and this difference was significant.
Whittet showed the highest stolon vigor, being more than 70% greater than 12A and nearly 3 times as great as Acacia but these differences were not significant (P>0.05). While line 12A was preferred by cattle over both Whittet and Acacia, differences were again not significant, but did indicate that 12A was not less palatable than the cultivars at this stage of growth.
The germination rate of line 12A and the cultivars, Whittet and Acacia, were all over 80% and seedling vigor was similar. Recorded grazings in 40 minutes at 10-second intervals; at the end of the experiment.

Discussion
In this study, a line of kikuyu grass, 12A, was identified as being more resistant to the fungal diseases kikuyu yellows and black spot and higher yielding than the cultivars Whittet and Acacia and similar to the two commercial cultivars for stolon vigor, herbage quality and cow preference. The most important attribute, where 12A was superior to the other lines, was resistance to the fungal disease, kikuyu yellows, in most, but not all cases. The resistance to kikuyu yellows differed markedly between the 2 evaluation sites. At Site 1, the resistance was outstanding, with only 15% of 12A plants infected with kikuyu yellows, whereas over 85% of plants in the other lines, including Whittet, were infected and all infected plants died. This was despite the fact that all lines, except Whittet, were found to be resistant to the KY1A strain of kikuyu yellows, when individual plants were tested under glasshouse conditions in Stage 1 of the study.
In contrast, at Site 2, line 12A showed no advantage over the other kikuyu lines under test in terms of resistance to kikuyu yellows with 85% of plants infected. However, 49% of all plants infected with kikuyu yellows in summer recovered in the autumn, compared with only 8% at Site 1. This was despite the fact that the minimum temperature remained above 15 °C longer than normal and well into April (Table 7), a temperature conducive to kikuyu yellows activity. In addition, the onset of visual symptoms of kikuyu yellows infection was later at Site 2 than at Site 1. This difference between sites was presumably due to strain differences of the pathogen at each site with the strain at Site 1 being more pathogenic.
The regional study also supports the contention that line 12A has greater resistance to kikuyu yellows than Whittet with only 16% of 12A plants becoming infected compared with 64% for Whittet and nearly all inoculum that infected 12A plants came from 2 properties on the lower north coast of NSW. In addition to displaying resistance to kikuyu yellows, 12A also had significantly higher 'summer' and total yields than Whittet plus earlier onset of growth in early spring with dry matter yields of 12A and Acacia being twice that of Whittet at the first spring defoliation. The results also indicate that forage quality, stolon vigor and cow preference for 12A are at least as good as those for the 2 commerciallyavailable kikuyu cultivars. The complete absence of black spot infection in 12A, compared with a very high level of infection in Whittet and other lines tested in Stage 4, is also very important. Unlike kikuyu yellows, black spot does not kill the kikuyu plant and scarcely affects its growth, as only the tips of the mature leaves are affected, but palatability of the infected herbage for grazing cattle is very low. Thus, dairy farmers consider that black spot is just as important a problem as kikuyu yellows.
Interestingly, the DOMD for samples of 12A were not significantly different from those of Whittet in any season. In this study, DOMD was assessed by the Tilley and Terry method (Tilley and Terry 1963), which is considered to most accurately reflect the availability of nutrients to the ruminant animal. One can be confident that forage from 12A was not inferior to that of Whittet.
We used 4.5 new leaves/tiller as the optimal time to defoliate, as digestibility is considered to be at a maximum at this point, after which the oldest leaves sequentially begin to senesce and the proportion of stolon:leaf growth increases (Reeves et al. 1996). In this environment, 4.5 leaves is usually reached at about 2-weeks regrowth in 'summer' and 3-4-weeks regrowth in late 'spring and early 'autumn' and is almost entirely dependent on ambient temperature.
While the DM yield of 12A over the main 'summer' period was greater than those of both Whittet and Acacia plus 15A and 3A, total DM yield of 12A over the 11 months was greater than those of Whittet, 15A, 3A and 14B only. The observation that leaves of Whittet were yellow during 'winter', while 12A and Acacia remained green, was in keeping with the higher yields of 12A and Acacia at the first defoliation after 'winter' on 6 October. This was not unexpected for Acacia, as it was selected from the Acacia plateau at a higher elevation and in a colder climate. This 'winter' activity is a highly desirable trait for a dairy pasture and more particularly for beef farms, where over-sowing of kikuyu with ryegrass in autumn, to provide winter feed, is not practiced often, primarily due to cost.

Conclusions
This study has identified a kikuyu line, 12A, which appears superior to the commercial cultivars Whittet and Acacia, in terms of resistance to fungal infection, i.e. kikuyu yellows, in most but not all situations, and Bipolaris spp., both of which have been responsible for a major decline in the area of kikuyu pastures grown in the subtropical dairy region of Australia and a hesitancy by farmers to re-establish kikuyu grass with the existing cultivars. This successful selection of 12A has been achieved along with an increase in forage yield over that of the most common commercial cultivar, Whittet, and, most importantly, without a decline in forage quality and cow preference. The study has also given indications of the genetic diversity in kikuyu grass for the traits tested. Evaluation of this line under grazing to assess the benefits of improved disease resistance and yields in terms of increased livestock production would provide further evidence for recommending widespread plantings.