You are here: Tussock Grasslands MIS Site contents Hieracium gateway Article

Sheep fancy hieracium flowers

Monitoring Grazing of Hieracium on Mid-Altitude Blocks at Tara Hills Research Station During 1996/97

C C Boswell and P R Espie

AgResearch, Invermay Agricultural Centre, Private Bag 50034, Mosgiel

Three species of hieracium (H. pilosella, H. praealtum, and H. lepidulum) grow on two mid-altitude blocks (750-1050 m a s l) on Tara Hills High Country Research Station. We report on the current relatively low incidence of the three species of Hieracium in the blocks, observations on flowering in the species, the extent of grazing the plants received from ewes and lambs during 1996-97, and discuss how grazing to remove flowers can reduce the rate of invasion of pastures by hieracium seed.

The grazed blocks

The distribution of the three Hieracium species was determined on two oversown tussock grassland blocks at Tara Hills. Block 65T (39.2 hectares) is a shady block with browntop the dominant plant; Block 66T (45.11 hectares) is a sunny block with cocksfoot the dominant species. Slopes vary from <5°to 32°; most are steep (>28 ^o) to moderately steep (15-28 ^o) with the steepest slopes found in the sunny block. Traditionally the blocks are periodically stocked with merino ewes and lambs at mean stocking rates of approximately 2.5 su/ha (stock unit per hectare). During 1996/97 the 2 blocks (84.27 ha) were grazed with a mob of 256 ewes and 267 lambs (i.e. a mean stocking density of about 3 su/ha for the period November to May). The actual stocking density of sheep on the shady block over the period December 1996 to February 1997, during which effects of grazing on hieracium flowers were measured, was 6.5su/ha.

Physical distribution of hieracium within blocks

The landscape of the two blocks was classified and mapped into a large number of landscape units. A study of physical properties of the units and their productivities also involved detailed analysis of the plants present within them. Hieracium pilosella was the main hieracium species growing on the sunny block and was widespread through the block. It also grew on sunnier sites within the shady block. It was especially associated with shallow soils overlying rock outcrops on dry sunny ridge tops, or on steep gully sides. Based on its occurrence within mapped land units we can say that 14.86 ha or 17.4% of the total area was affected by H. pilosella. However, H. pilosella grows in relatively small discrete patches within the affected area and actually covered 1-2% of the total area. H. pilosella usually grows alone or in association with H. praealtum..

H. praealtum normally grows where H. pilosella is also found. The actual area covered by H. praealtum plants is possibly half that occupied by H. pilosella.

The extent of H. lepidulum cover was larger than expected from previous vegetation surveys on other Tara Hills blocks (P R Espie pers. comm.); which may indicate that the species is actively invading the alternately grazed and spelled mid altitude pastures. It was found on the shady block. Its range was restricted to rocky, steep (approximately 28 ^o) shady faces; growing in association with either snowgrass (Chionochloa rigida), or H. praealtum and hard tussock (Festuca novae-zelandiae); and on a moderately steep (approximately 20 ^o) shady slope where it was a minor component in a browntop association. The land areas affected by H. lepidulum amounted to 6.4 ha or 16.3% of the shady block (7.6% of the total area). The total cover of H. lepidulum was estimated at approximately 2% of the total area.

Another Hieracium species H. auranticum was found as rare single plants on the shady block.

Timing of hieracium flowering

At 1000 m peak flowering of H. pilosella appeared to occur about the same time sheep were moved onto the shady block, i.e. 11 December 1996. The timing of flowering of H. praealtum was similar. The two species are easily recognised by a difference in colour of flowers (H. pilosella paler lemon yellow colour, H. praealtum darker buttercup yellow colour) and flower numbers (single vs multiple flowers per head, respectively). With H. pilosella and H. praealtum flowering had ceased about 2-3 weeks from the appearance of the first flower. H. auranticum, although seldom seen in the grassland, was also found flowering (orange-coloured flower) in mid-December. Although flowering was occurring in H. lepidulum at 11 December, its peak was at least 2 weeks later. Hieracium lepidulum also has lemon yellow flowers like H. pilosella) but has more than one flower per stalk. H. lepidulum was still flowering at 29 January 1997; thus its flowering may occur over a more extended period. Minor second flowering in H. pilosella was recorded at February 10. By February wind had dispersed most of seed from the 3 main species of hieracium.

The main flowering on the Red Flat at Omarama (490 m a s l) was probably coming to an end at 11 December and seed was largely set by 18 December. When viewed from the grazing observation trial site (approximately 1000 m a s l) at 11 December, the flat appeared as a sea of yellow (the colour was reduced with the approach of evening as the petals closed); but the yellow had disappeared from the high view at 18 December. Detailed phenological studies of Hieracium spp. over the period 1994-1996 (P R Espie pers.comm.) show that within a species flowering occurs later as altitude increases. A difference of 7-10 days could be expected over a 500 m increase in altitude.

The shortness of the H. pilosella and H. praealtum flowering period and the timing of flowering with altitude has important implications for management of Hieracium spp. invasion by grazing.

Grazing observations/measurements

Effects of grazing on hieracium flowering were measured in summer 1996/97 from the shady block. Relatively small areas where patches of different Hieracium species were present were selected as sites to measure the effect of grazing on the plants. A series of paired sample sites on patches of H. pilosella (n=4), H. lepidulum and snowgrass (n=1), H. lepidulum/H. praealtum and hard tussock (n=1) were laid out on 11 December 1966 as the sheep were moved into the block. Paired sites were located at two 0.5 m² areas adjacent to each other; a cage (which excluded grazing) was placed on one area, and the other marked area was taken to represent grazed patches. Observations were made at subsequent visits to the trial area at 18 December 1996, 24 December 1996, 29 January 1997, and 10 February 1997.

At December 18 digital images were made from sample areas (each 50% of the plot areas) and used to determine the available floristic information as to numbers of flowers and buds present under the two grazing situations.

Biomass residues (especially flowering stalk tissue) from ungrazed (caged areas) and grazed areas, were measured from cores of turf taken at 29 January 1997.

Grazing of hieracium flowers

H. pilosella flowers:

Within the block Hieracium represented only a small proportion of the total herbage available to the ewes and lambs. Preferential grazing of the flowers of H. pilosella in particular was readily demonstrated given this circumstance. The effect of one week of grazing (11-18 December 1996) on populations of H. pilosella is shown in Figure 1. The figure records the mean numbers hieracium flowers, hieracium flower buds, white clover flowers, and buds of either white clover or hieracium flowers that could not be determined from the analyses of digital images of paired sites which were respectively grazed and protected from grazing.

Figure 1: December 18 1996; mean population density of H. pilosella and white clover flowers and buds from grazed areas and areas protected from grazing (no/m²).

There was also some grazing of hieracium leaves evident at 18 December, but this was minor compared with the flower removal. White clover flowers and seedheads were not grazed as preferentially as H. pilosella , H. praealtum, and H. lepidulum flowers. Clover flowers and florets with set seed remained when most H. pilosella flowers had been removed.

Almost all H. praealtum flowers were also grazed during the one week period. Flower removal from H. praealtum on a grazed area between 11 and 18 December was found by measuring the numbers of beheaded flower stalks within a single grazed quadrat sample (0.25 m² ). There were no flowers present at 18 December. However, it contained 27 obvious beheaded flower stems in the sample area and possibly a further 10 beheaded stems which were not so clearly defined in the photographic image. The recorded beheaded flower density was thus 108-152 stems/m². Hieracium praealtum cover in the plots was about 25% of total cover. The potential densities of beheaded flowers could reach 400-600 flowers/m² where the cover of is total.

Figure 2 shows the data from core samples taken from paired grazed and ungrazed sample sites at the completion of grazing in February1997. Flowering stalk was largely removed from the grazed samples, especially H. praealtum and H. pilosella. The headless stalk of H. praealtum present in grazed plots 1 week after grazing started in December was not recorded at the end of grazing (Figure 2). The amount of stalk in H. lepidulum plots reflects the upright growth habit of the plant.

Figure 2 : Comparison of the % stalk in core samples of three Hieracium species from grazed and ungrazed sites

Table 1 shows the Hieracium above-ground biomasses (stalk and vegetative tissue) in grazed and ungrazed plots at the completion of grazing. With H. lepidulum 79% of the plant was removed by grazing, with H. pilosella 55%, and with H. praealtum 29%.

Table 1: Biomass of hieracium species measured from grazed and ungrazed plots (kg DM/ha)

By this time it was clear that much of the available leaf tissue as well as the flowering stalk was grazed. In H. lepidulum 94% of the flower stems and 72 % of the leaf was removed by grazing, with H. pilosella 96% of the flowering stalk and 51% of the leaf, and with H. praealtum 100% of the flowering stalk and 21% of the leaf was removed. With grazed H. praealtum the hieracium represented only 38% of the total plant biomass in the plot; in all other plots non-hieracium species were much more restricted.

These observations give added support to the grazing management recommendations for hieracium espoused by Peter Espie (2001) in his book "Hieracium in New Zealand: ecology and management", and in Espie (1994).

It could be argued that the high utilisation of the hieracium reflects the small amount of the plant in the total pasture, its confinement to small discrete areas of plants surrounded by taller vegetation, and a tendency for it to be forced to grow more erect than in hieracium-dominant pastures. However, results in the following section suggest that even in hieracium-dominant pastures there can be high utilisation of hieracium tissues by grazing animals.

Hieracium pilosella eaten when under heavy grazing pressure

In June 1997 a mob of 720 ewes was grazed for 4 days on a 2.6 ha experimental block where H. pilosella represented over 50% of the pasture cover. At the end of grazing, replicated cores of turf (10 cm diameter) were taken from sites representative of the grazed area and a small fenced plot which was protected from grazing. Herbage was removed from the cores to ground level by scalpel and hand sorted into hieracium leaf, or stalk, and other species. Mean herbage on offer (ungrazed plot) amounted to 2650kg DM/ha of which approximately 86% was hieracium leaf and 6% stalk and 8% other species. After grazing 970 kg DM/ha remained; i.e. 63% of the herbage available to the sheep was eaten. The proportions of leaf, stalk and other species after grazing were similar to before grazing. Hieracium. pilosella represented about 90% of the herbage eaten. There were instances where hieracium leaves were grazed so hard that only the white petiole of the leaf remained at ground level.

A high degree of utilisation of hieracium tissue was an unexpected finding. In previous articles we have suggested that H. pilosella in its normal prostrate habit is not likely to be subjected to hard defoliation. However, the results reported confirm observations of close grazing of hieracium during the 1996-97 season made by the farm manager at Tara Hills Mr Grant Wardell.

In a different study, also undertaken during 1996-97, diet selection by rabbits in semi-arid grasslands of the Mackenzie Basin showed the proportion of Hieracium species eaten was directly proportional to the amount of hieracium available in the diet (Reddiex 1998). Hieracium present in the diet at three sites was 31-39% of total available plant material; the hieracium in the diets selected ranged from 32% to 39%. While this does not demonstrate selective grazing of hieracium by rabbits, it does show that the plants can be an important food for these animals.

Comparative chemical composition of a range of grasses and herbs from the grazed blocks

In May 1997 samples of a range of plants including grasses and herbs were taken from Blocks 65 and 66 and chemical analyses conducted on them (Tables 2a and b). Plant nitrogen levels for herbs were mid table compared with grasses, but H. pilosella had comparatively low N contents. Although plant P levels of several herbs were generally high compared with grasses, H. pilosella was low on the list. Herbs had relatively low S contents. The reverse was true of plant aluminium contents. Herbs in general, and H. pilosella in particular, had high Al contents.

Herbs, including H. pilosella, had high cation contents, especially magnesium and calcium (Table 2b).

Other chemical analyses of different ranges of species were made on other occasions. The above patterns of herbs containing high concentrations of cations and aluminium, and low S contents relative to grasses, was consistent at different samplings. Of particular interest was a sampling in November 1997 when both H. pilosella and H. lepidulum were analysed. Hieracium lepidulum had a higher content of N and K than H. pilosella and a lower content of P and Mg ; Ca levels were similarly high and S levels similarly low compared with grasses analysed at the same time.

Table 2: Comparative chemical composition of a range of grasses and herbs (including H. pilosella) from the grazed blocks (sampled May 1997)

(a) nitrogen and anions

(b) cations

Table 3 shows an example of plant chemical analyses of different tissues of H. pilosella sampled from a Block in the same general location as reported above. Separate analyses of different plant components of H. pilosella have shown that flowers generally have higher concentrations of P and K, and possibly N, than leaf tissues and top samples; leaves and tops have greater Ca and Mg contents. Old flowers and their stems have much reduced nutrient levels than new flowers. While roots are not available to grazing sheep, their contents are shown for comparison.

Table 3 : Mean concentrations of major plant nutrients in different H. pilosella tissues summer 1998 (% of DM; standard deviations from means in brackets)

It is possible that sheep select the herbs like hieracium for their high cation content. Other plants are selected for their high N content (e.g. plants at camp sites; goosegrass for example and yarrow were regularly present at the higher fertility camp sites within the blocks, had high N and K contents, and both were readily eaten when vegetative; Table 2). Within hieracium tissues flowers offer an easily accessible short-term source of higher P and K than leaf tissue, and with a different taste, which no doubt encourages their being subjected to highly selective browsing.

Controlling the rate of invasion by hieracium with grazing sheep

The potential effect of invasion by hieracium plants through the spread of seeds is enormous. From studies at the Red Flat at Omarama (490 m) the potential seed production calculated for H. pilosella is approximately 40,000 seeds/m² (400 million seeds per hectare) and H. praealtum up to 30,000 seeds/m² (300 million seeds per hectare) (Makepiece 1985).

Dispersal of the seed by wind provides a high probability that a proportion of the seed will alight at bare areas in the grassland and allow germination and establishment without immediate plant competition. Even areas bare of vegetation are not necessarily the best for seedling hieracium establishment; Rose and Frampton (1999) found that establishment was greater for H. pilosella where germination occurred within the protection of tall and short tussock.

Grazing to control seed production by eating flowers of Hieracium species offers a practical management tool especially in blocks where the weeds are present at low densities (i.e. in the phase of weed invasion of a pasture or grassland). The shortness of the H. pilosella and H. praealtum flowering period allows managers a relatively short time to use sheep to remove flower heads within a specified block and so lessen the huge potential for seed dispersal from the plant which allow it to colonise bare areas within the pasture. However, the pattern of delayed flowering with altitude allows the chance to use a mob of sheep (probably dry sheep such as wethers or even ewe hoggets at this time of year) as selective flower removers by moving them quickly progressively uphill as hieracium flowering occurs (i.e. in blocks not already stocked with ewes and lambs).

Three characteristics of H. lepidulum should make its invasion of grassland or pasture more controllable by grazing sheep. First its flowering is more prolonged and thus it offers sheep a greater chance of browsing the plant; second, its erect growth habit especially when compared with H. pilosella, means a greater proportion of the plant tissue is made available for removal by grazing; and third, H. lepidulum is typically at an earlier stage of invasion than H. pilosella or H. praealtum in most New Zealand grassland situations and therefore has lower plant populations.

Rather surprisingly, anecdotal evidence [from fellow scientists based at Invermay] of blocks "full of H. lepidulum" has been reported in the Otago Lakes District during the summer 2001-2002, which was a very productive growing season. Runholders practicing sustainable management carry numbers of stock commensurate with plant productivity levels typical of mean or lean years; thus in years such as 2001-2002 there is a surplus of growth. Some blocks are not required for grazing and in these circumstances the opportunities for seed production in affected blocks is unchecked. This highlights the need to be aware of the need for light grazing of blocks even in years of bountiful growth.

Conclusion

In the early stages of invasion of grassland by Hieracium species, grazing by sheep can restrict the rate of invasion of the weeds through their selective grazing of flowers. In contrast to some of our earlier thinking regarding sheep grazing of vegetative plant tissues, it appears that sheep can remove over half of the above-ground leaf biomass, both where the plants are sufficiently few to be a novelty in the diet, and where there is heavy stocking of hieracium-dominated grassland.

References

Espie P.R. 1994. Integrated pastoral management strategies for Hieracium control. Proceedings of the New Zealand Grassland Association 56: 243-247.

Espie P.R 2001. Hieracium in New Zealand: ecology and management. AgResearch Limited, Invermay Agricultural Centre, Private Bag 50034, Mosgiel.

Makepeace W. 1985. Some establishment characteristics of mouse-ear and king devil hawkweeds. New Zealand Journal of Botany 23:91-100.

Reddiex B. 1998. Diet selection of European rabbits (Oryctolagus cuniculus) in the semi-arid grasslands of the Mackenzie Basin, New Zealand. M.Sc. thesis Lincoln University.

Rose A.B., Frampton C.M. 1999. Effects of microsite characteristics on Hieracium seedling establishment in tall- and short-tussock grasslands, Marlborough, New Zealand. New Zealand Journal of Botany 37:469-477.

This paper has been developed for this site and cleared for reproduction here by AgResearch. Your comments are welcomed and should be forwarded to Colin Boswell colin.boswell@agresearch.co.nz; or Peter Espie peter.espie@agresearch.co.nz

Page last updated on Monday, August 20, 2002 Site hosting by Landcare Research Copyright and Disclaimer