Plant Protection Quarterly Vol.12(2) 1997 97 communication 1973) at a heavy stocking rate to graze a heavy infestation of H. Control of Hypericum perforatum L. (St. John’s wort) perforatum after a fire. The sheep control- by grazing management led the weed, although they had to be re- moved two weeks before dipping or shearing. Various flocks had to be used M.H. Campbell, NSW Agriculture, Forest Road, Orange, New South Wales and any one flock could only be grazed 2800, Australia. on the weed for a week during flower- ing. Set stocking on lightly infested (15% Summary ground cover) 80 ha non-arable paddocks If an infested pasture is left ungrazed 1922). H. perforatum is more toxic if in- south of Orange during the dry years of Hypericum perforatum (St. John’s wort) gested at flowering than when young or 1979–82 with Merino sheep at 5–8 dse ha-1 becomes dominant. If light infestations dry. Sheep and cattle prevented from in- virtually eliminated mature plants are heavily grazed it can be controlled. gesting the weed but allowed physical (Campbell and Dellow 1984). The sheep, Heavy infestations on arable land can be contact with it for three weeks in spring forced to graze H. perforatum because controlled by ploughing, sowing im- were not affected (Henry 1922). The weed other pasture was scarce, killed mature proved pastures and grazing heavily. The is still harmful when eaten in hay. The plants by eating the foliage as well as by most difficult situations for the control minimum toxic dose of foliage for cattle pulling the crowns and roots 10–20 cm out of the weed are heavy infestations in and sheep is approximately 1 and 4% of of the soil. Control was assisted in one large paddocks on non-arable land. Here live weight respectively (Jessop 1970), al- paddock by an infestation of larvae of the establishment of a Phalaris aquatica though there is variation in susceptibility Chrysolina quadrigemina Suffrain in 1979. based pasture by the aerial spray-sow within a herd. Sheep can begin to exhibit The sheep suffered only minor discomfort technique and heavy grazing has proved symptoms within two days of ingestion in from the effects of H. perforatum. Thus it successful on fertile soils but the high cost sunny weather but it may take >12 days in appears that light infestations can be con- and the risk of failure on infertile soils cloudy weather (Dodd 1920). Irritation trolled by grazing heavily with Merinos has limited its use. Aerial distribution of may disappear in dull weather or over- but heavy infestations require special Trifolium subterraneum seed and ferti- night only to reappear when the animal is grazing management to restrict the intake lizer and grazing only when there is re-exposed to the sun. Affected animals of H. perforatum to non-toxic levels. an abundance of pasture has reduced recover once ingestion of H. perforatum A management strategy employed by ground cover of H. perforatum by 75%. stops. Sheep usually take a month for M. Campbell, ‘Bonada’, Mumbil, New Heavy set stocking of T. subterraneum–H. symptoms to disappear, although scars, South Wales (NSW), embraced the aerial perforatum pastures with sheep (black or ragged fleece and poor condition remain. distribution of Trifolium subterraneum and white) has proved successful on light in- Cattle generally take three weeks to re- superphosphate onto hill paddocks in- festations but unsuccessful on heavy in- cover (Henry 1922). fested moderately with H. perforatum and festations. Cattle are more effective than The most important need for the future subsequent grazing, at 10–12 dse ha-1, only sheep and a mixture of goats and cattle control of H. perforatum in the most diffi- when there was an abundance of pasture has proved the most effective grazing cult situations, large paddocks on non-ar- available. Animals were removed once management system yet devised. Most of able land, is to devise grazing manage- feed quantity fell and they began to lose the reports on the effects of grazing man- ment strategies that encourage pasture condition. They were re-introduced when agement on the control of H. perforatum species at the expense of the weed which pasture again became abundant. Over a have come from observations by produc- allows livestock to graze without injury period of 10 years, stocking rate increased ers. There is a major need for detailed re- (Campbell et al. 1996). Although no de- from 1.9 to 5.6 dse ha-1 and ground cover search into grazing systems to control the tailed research has been done on this sub- of H. perforatum decreased by 75%; Merino weed, particularly on non-arable land. ject there have been a number of practical ewes and lambs were not affected by H. experiences by landholders and others perforatum nor were Aberdeen Angus cat- Introduction that could provide the basis for future re- tle but Herefords were. This grazing tech- When animals are forced to graze Hyperi- search. nique provides animals with pasture cum perforatum (St. John’s wort) they suf- which could allow them to consume some fer photosensitization and nervous disor- Grazing experiences H. perforatum without injury, unless they ders. Symptoms observed in sheep (Dodd If left ungrazed or lightly grazed H. are extremely sensitive to the weed. 1920) include, head shaking and general perforatum becomes dominant. For exam- Before 1900 it was known that black or restlessness followed by reddening and ple, lightly grazed water catchment re- stained sheep could graze H. perforatum swelling of the lighter parts of the skin and serves around Burrenjuck, Blowering, with no obvious ill effects (Dodd 1920). then exudation of fluid that coagulates in Burrendong and Wyangala dams have be- Later, Daley (1937) found that fat lambs masses often matting the wool. Owing to come heavily infested with the weed. Al- could be raised on H. perforatum infested the irritation, the animal bites or rubs itself ternatively, in heavily grazed native pas- pastures provided: sufficient improved until the wool is removed and the skin be- tures, H. perforatum, although present, is pasture was available to allow rotational comes raw. The eyes, ears and nostrils only a minor constituent. If the grazing grazing; Romney Marsh × Merino ewes, swell and become covered with scabs. In pressure on such pastures is reduced then selected for dark muzzles, were crossed severe cases the ears may be torn to the weed often becomes dominant. with Suffolk rams; sheep were blade shorn shreds and/or the animal rendered blind. As it generally takes 1–2 weeks for the to leave a protective layer of wool; trees The weed also causes frequent and pro- effects of H. perforatum to appear and 3–6 were available; and earmarking of spring fuse urination, convulsions and enteritis, weeks for severely affected animals to re- lambs was delayed until autumn. Border generally fatal in lambs and sometimes in cover (Dodd 1920, Henry 1922), a two Leicester × Merino ewes and Southdown adults. Affected animals lose condition, weeks on and five weeks off rotational rams were not effective because neither stop producing milk and suffer abortions. system could allow H. perforatum to be transmitted a sufficiently dark colour to Symptoms for cattle and horses are simi- grazed by adult sheep. Merino wethers the lambs. Others have observed poor lar to those of sheep (Rogers 1914, Henry were used by A.M. Traill (personal performance of both white and black 98 Plant Protection Quarterly Vol.12(2) 1997 lambs suckling their mothers, suggesting with the Santa Gertrudis cattle which tol- subterraneum and Phalaris aquatica and that the toxin from H. perforatum is trans- erate the detrimental effects of H. ‘grazing rotationally’, and on non-arable mitted to the lamb in the milk or that the perforatum better than cattle breeds with land by distributing T. subterraneum and mother’s condition was adversely affected white markings. superphosphate and ‘grazing continu- (personal communications, A.M. Traill, By 1986 the goats had controlled the ously’. They believed that the main factor Dargo, Victoria 1973; R. Arnott, Coolah, blackberries, briars and thistles and by in the control of the weed was the reduc- NSW 1979). 1994 the ground cover of H. perforatum had tion of light under the pasture, the decum- Although black coloration can reduce been reduced from 90% dense wort at bent winter growth of the weed being sus- photosensitization by 95% (Daley 1937), commencement, to 20% dense wort, 50% ceptible to shading. Sowing a P. aquatica ingestion of H. perforatum still adversely scattered wort and 30% nil wort. The based pasture and ‘grazing heavily’ has affects the behaviour of black sheep dur- stocking rate had been increased form 2.5 been successful in controlling H. perforatum ing herding, shearing, dipping, (C.A. dse ha-1 pre-1979 to 8.2 dse ha-1 in 1994. even on non-arable land (Campbell et al. Bourke, Veterinary Research Officer, Or- The paddocks now turn off fat cattle. By 1975, Arnott and Campbell 1994, ange, personal communication 1983) and grazing at 8.2 dse ha-1 the quality of the Campbell and Watson 1994). However ear marking (intense irritation of the cut pasture has improved, the animals now sowing T. subterraneum alone on non- ear for up to six weeks, the unmarked ear grazing mainly T. subterraneum and native arable land has not always been success- being unaffected (Daley 1937)). Surpris- grasses, along with the wort. The feed ful, the weed growing up through dense ingly, Daley (1937) found no ill effects value of the leaves of the weed is high (Ta- clover in spring (Campbell et al. 1975). from tailing or castration of black lambs ble 1) which could contribute to the in- Casual observations suggest this tends to grazing H. perforatum. creased animal production. Despite some occur more in central and northern NSW Arnott found that cattle were more ef- wort still remaining in the grazed pad- where H. perforatum may be more difficult fective than Suffolk sheep because cattle docks, surprisingly, the animals have suf- to control than in southern NSW due to its survived better, had less severe effects on fered little photosensitization; i.e. animals response to the hotter spring/summer pe- associated pasture and caused more physi- with white skin, Hereford × Santa riods in the former regions. Another rea- cal damage to the weed often pulling out Gertrudis cattle and white goats, were son for the failure of T. subterraneum to whole plants. A disadvantage of using cat- only slightly affected. It is possible that control the weed may be ineffective graz- tle in large paddocks was their limited continual exposure to low levels of H. ing management. Research by Clark ranging from water. perforatum has resulted in animals devel- (1933) has shown that H. perforatum is sus- A further advance in grazing manage- oping resistance to its effects (R. Arnott ceptible to defoliation in spring. However ment of H. perforatum in large paddocks is personal communication 1996). the weed is most toxic in spring due to the use of a mixture of feral goats and Control of H. perforatum by a mixture of high levels of hypericin (Southwell and Santa Gertrudis cattle (Arnott and goats and cattle is probably limited to hill Campbell 1991), long sunny days and high Campbell 1994). Grazing began in 1979 on paddocks of less than 80 ha because in big- biomass. The most practical time to graze heavily infested hill country that had been ger paddocks it is difficult to get cattle to the weed heavily is in winter when there is aerially top dressed with T. subterraneum graze the whole area as they tend to stay minimum biomass of the weed (Clark and and fertilizer. The goats grazed the black- near water. By providing more watering Clark 1952), low levels of hypericin berries, briars and thistles and reduced the points and using Bos indicus cattle which (Southwell and Campbell 1991) and mini- height of the grass allowing the cattle to range more widely than British breeds, mum sunlight. However no grazing man- range freely. The stocking rate used was control in paddocks larger than 80 ha agement program has been devised to 8.2 dse ha-1; in the beginning it comprised 4 could be achieved. Good fencing and take advantage of these factors. dse ha-1 goats and 4.2 dse ha-1 cattle, but eliminating rogue goats that lead mobs Grazing tactics could play a role in the after two years, the stocking rate of goats out of their home paddock simplifies goat control of H. perforatum. For example, was progressively reduced and that of cat- management. spelling a pasture containing barley grass tle increased until now the consistent rates Moore and Cashmore (1942) showed (Hordeum leporinum) in winter encouraged are 2.2 dse ha-1 goats and 6 dse ha-1 cattle. that H. perforatum could be controlled: on more upright growth making it more vul- Feral goats of various colours were used arable land by ploughing, sowing T. nerable to removal by sheep (Kemp et al. 1995). Spelling or grazing H. perforatum at Table 1. Feed value of Hypericum perforatum. specific times of the year could assist in its control. However there have been no in- Part of the plant Crude protein Acid detergent Dry matter Metabolizable vestigations into the use of grazing tactics (%) fibre (%) digestibility energy (megajoules for the control of H. perforatum. -1 (%) kg dry matter) In Victoria new grazing systems (‘Pro- Leaves 20.9 15.3 79.7 12.0 ductive Pasture Packages’) that make the Flowers, top stems most of soil fertility, pasture quality and and top leaves 20.0 23.6 72.6 10.9 quantity and a carrying capacity timed in Side stems 7.0 49.5 45.7 6.9 unison with pasture production (Saul 1994, Main stems 4.3 59.8 36.1 5.4 de Fegley 1994) have greatly increased production and controlled weeds (Table Table 2. The effect of rate of phosphatic fertilizer over 16 years applied in a 2). Productive Pasture Package on pasture composition at Hamilton, Victoria, in In the above case the control of R. rosea was due to increased soil fertility and pas- mid-spring (Saul 1994). ture production/competition from the ad- Pasture species Green forage (%) on plots treated with the following rates dition of phosphatic fertilizer; stocking of fertilizer (kg P ha-1 yr-1) rate had no effect on its control (Saul 1994). Thus grazing management alone may not 137142123 be sufficient to control St. John’s wort Romulea rosea (onion grass) 35 31 9 2 1 0 whereas a productive pasture package as T. subterraneum 61114272824 part of an integrated control program Lolium perenne 12 15 32 34 39 37 may be successful. Plant Protection Quarterly Vol.12(2) 1997 99 References Campbell, M.H. and Watson, R.W. (1994). Agricultural Gazette of NSW 33, 205-7. Arnott, R. and Campbell, M.H. (1994). St. John’s wort. NSW Agriculture Jessep, C.T. (1970). St. John’s wort in New Control of St. John’s wort (Hypericum Agfact P7.6.1. Zealand. Tussock Grassland and Moun- perforatum L.) in large paddocks on hill Clark, L.N. and Clark, N. (1952). A study tain Lands Institute Review 20, 76-83. country by the aerial spray-sow of the effect of Chrysomela hyperici Forst, Kemp, D., Michalk, D.L. and Dowling, technique and by grazing manage- on St. John’s wort in the Mannus Valley, P.M. (1995). Managing pastures to im- ment. Proceedings 9th Conference NSW. Australian Journal of Agricultural prove composition and quality. Pro- Grassland Society NSW, Queanbeyan, Research 3, 29-59. ceedings 36th Conference Grassland pp. 98-9. Clark, N. (1953). The biology of Hypericum Society Victoria, pp. 113-23. Campbell, M.H., Briese, D.T. and Delfosse, perforatum L. var. angustifolium DC. (St. Moore, R.M. and Cashmore, A.B. (1942). E.S. (1996). Hypericum perforatum L. In John’s wort) in Ovens Valley, Victoria, The control of St. John’s wort by com- ‘The Biology of Australian Weeds’, eds. with particular reference to entomo- peting pasture plants. Council of Scien- R.H. Groves, R.C.H. Shepherd and R.G. logical control. Australian Journal of tific and Industrial Research, Bulletin Richardson, pp. 149-67. (R.G. and F.J. Botany 1, 95-120. No. 151, pp. 23. Richardson, Melbourne). Daly, C.J. (1937). St. John’s wort control by Rogers, T.B. (1914). On the action of St. Campbell, M.H. and Dellow, J.J. (1984). pasture improvement and grazing. Ag- John’s wort as a sensitizing agent for Control of St. John’s wort (Hypericum ricultural Gazette NSW 48, 301-18. non-pigmented skin. American Veteri- perforatum) by insects and/or grazing de Fegley, C. (1994). Putting it all together— nary Review 46, 145-62. animals. Proceedings 7th Australian a productive perennial pasture system. Saul, G. (1994). Productive and profitable Weeds Conference, Perth, Volume 1, Proceedings 9th Conference Grassland grazing systems in western Victoria. pp. 109-17. Society NSW, Queanbeyan, pp 74-6. Proceedings 9th Conference Grassland Campbell, M.H., Flemons, K.F. and Dodd, S. (1920). St. John’s wort and its ef- Society NSW, Queanbeyan, pp. 67-73. Dellow, J.J. (1975). Control of St. John’s fects on livestock. Agricultural Gazette of Southwell, I.A. and Campbell, M.H. (1991). wort on non-arable land. Australian NSW 31, 265-72. Hypericin content variation in Hyperi- Journal of Experimental Agriculture 15, Henry, M. (1922). Feeding and contact cum perforatum in Australia. Phytochem- 812-7. experiments with St. John’s wort. istry 30, 465-8.

St. John’s wort experiences at Birnam Wood

Richard Arnott, Birnam Wood, Coolah, New South Wales 2843, Australia.

The property The property Birnam Wood has an annual week using the old type metal knapsack. It when a patch of St. John’s wort has been rainfall of 750 mm on lower parts and up was later thought a great step forward eaten out. to 800 mm on the high tableland. The soil when misters were introduced, and then In the 1960s Tordon 50D came on the is heavy basalt and it is steep country with plastic back sprays. As the weed spread, market and it seemed to do some good. an altitude of 610–1320 metres. It has been more country was given up as uncontrol- Thousands of dollars were spent control- fertilized with gypsum, sulphur and sul- lable. The battle was lost paddock by pad- ling St. John’s wort where easy access phur fortified fertilizer (45% sulphur) since dock. could be gained and combined with even 1953 and so conditions are perfect for stocking pressure. However, on the blackberry, briar and timber regrowth, The herbicide steeper sides, in spite of walking every which is in addition to a heavy infestation 2,4,5-T was used, mixtures of this and 2,4 D, foot of the hills two or three times in a of St. John’s wort. combined with distillate was also used. summer with back sprays or misters, St. In desperation polyborchlorate and John’s wort still got away. The weed The problem borascue were used as a soil sterilent. It seemed to be coming under control and An ever increasing number of farmers are was carried on horses in wheat bags hang- then following a wet spring the St. John’s becoming aware of St. John’s wort devas- ing down over the seat of saddle, as it was wort would reappear. The paddock would tation to hilly country. Attempts at Birnam very heavy. This poison only had the ef- have been heavily grazed for some years Wood to control and eliminate St. John’s fect of killing everything on steep sides, and carrying stock well with very little wort started in the late 1940s. It was first causing some washing and the first plant sign of St. John’s wort, then it would ex- noticed after hay had been purchased to return was the St. John’s wort. There plode with patches appearing the size of from the Mudgee district during the 1946 were still a few patches where nothing else tennis courts. drought, and seed germinated and was was growing except the St. John’s wort In paddocks where spraying had been spread by cattle. Salt was thrown over the and so it was given up. stopped, and could not run sheep, even patches which could be seen and it was In about 1958 Chrysolina beetles were though there was an excellent cover of hoped this would kill it. made available from Michael Kidd, north white and sub clover, the St. John’s wort There was very little spreading until the of Tamworth. They took off and it seemed would come through the clover and by wet years of the 1950s when fertilizer had they were the answer as whole hillsides December the paddock would be a solid been spread and suddenly the St. John’s were eaten out, however this was not the yellow mass of St. John’s wort flowers. wort ‘took off’. At that time 2,4-D amine case. The beetles are very seasonal, only Of course, without sheep the black- was the only chemical available that was effective in larger areas, and only appear berry and briar increased very rapidly on known of, and it had very limited success. in about one year in three. However, they the steep sides, (a drop of 80 metres). Thousands of acres in steep hill country are very useful as they give other grasses, The cattle grazing area available be- were hand sprayed two or three times a especially natives, a chance to regenerate came more limited until it was regarded