Herbicides, weeds and endangered species: management of bitou bush (Chrysanthemoides monilifera subsp. rotundata)

Summary

Study 1

In Australia, weed invasion has caused the extinction of several endemic plant species and threatens others. Rare plants, which often have small population and range sizes, and are competitively inferior to invasives, are at particular risk. For example, Pimelea spicata is an endangered endemic shrub threatened by habitat loss and invasion by weeds, including the bitou bush Chrysanthemoides monilifera ssp. rotundata which infests about 70,000 ha of the New South Wales coastline. Bitou bush was first recorded in Australia near Newcastle, New South Wales around 1908, probably introduced through dumping of ships' ballast. From 1946 to 1968 it was planted along the coast to revegetate areas after mining. It is now regarded as one of the worst weeds in Australia because of its invasiveness and associated detrimental economic and environmental impacts.

The winter application of the herbicide glyphosate is used to control bitou bush, whilst other native plants such as sclerophyllous shrubs in the genera Acacia (wattles), Banksia (banksias) and Leptospermum (tea-trees), and large Lomandra (mat-rush) tussocks are resilient. However, no attempts at bitou bush control by glyphosphate application had been undertaken in areas with Pimelea as the effects of glyphosate on this shrub were poorly known. Observations at sites were wind had caused glyphosphate to drift over areas where Pimelea was present suggested that it suffered reduced growth and productivity. This study investigated the effect of glyphosate application and manual weed removal on Pimelea.

Study populations: Two large populations of Pimelea spicata to the south of Shellharbour, Illawarrawere were selected, such that only a proportion (20%) of the adult population would be affected by herbicide spraying. During September 1998, 40 flowering plants were selected throughout the known extent of each population. From these, 30 in each site were randomly selected, and the diameter of the tap root at its thickest point was measured. These plants were then grouped with their nearest neighbours in five groups of six plants. Typically, plants within a block and the blocks were separated by 2-3 m.

Seedlings: Pimelea seedlings were grown in glasshouses for 12 weeks. Each selected adult plant had two seedlings planted next to it, seedling leaf number and height were recorded, and the seedlings were left to settle for four weeks. After this time, all seedlings were alive and healthy.

Treatments: Treatments were applied to 1 m² plots around each plant (and encompassing the seedling). The herbicide glyphosate was applied at the recommended rate for bitou bush infestations (1:180) with a hand aerosol spray pack, with no application as a control. Weed removal was at one of three intensities:

1) Slash – weeds cut by hand and the biomass left in situ
2) Slash and remove – weeds cut by hand and the biomass removed
3) no cutting control

The experiment was fully factorial, giving 2 herbicide treatments x 3 weed removal treatments x 5 replicate blocks. Plots receiving both herbicide and slash treatments were sprayed first, the herbicide allowed to dry, and then cut. Treatments were applied on 'day 1' in late September 1998.

On days 13, 36, 54, 84, 126 and 169 (169 = late February), the sites were revisited and the plants allocated to one of five health categories:

1 = healthy plants
2 = 1-25% of above-ground biomass dead
3 = 26-50% dead
4 = 51-75% dead
5 = 76-100% dead

Also, on adult plants, the tap root diameter was remeasured, the number of reproductive shoots counted, and the number of green shoots counted on the first and last day.

Herbicide treatment: Of those Pimelea plants treated with herbicide, 57% had died by the end of the experiment, while all remaining plants suffered some damage. The application of glyphosate had a significant negative impact on the number of green shoots, the number of reproductive shoots, plant health and tap root diameter of adult plants. This negative effects became more pronounced towards the end of the experiment.

Weeding: There was no significant effect of weeding treatments on the number of green shoots, the number of reproductive shoots, plant health or tap root diameter of adult plants. However, when pooling the herbicide treatments (glyphosate and control) resprouting of plants was greater in plants in the slash and remove treatment (55% of plants), than the slash treatment (30%) and control (15%).

Seedlings: Only 17 of 60 (28%) of Pimelea seedlings survived until the end of the experiment. Furthermore, only one seedling survived the application of glyphosate.

Conclusions: Application of glyphosate killed Pimelea seedlings and half of the adult plants. Furthermore, reduction in tap root diameter suggests that further adverse affects might have been recorded if the experiment were longer-term. Consequently, the authors suggest that sustainable weed management is the best option for protecting this rare endemic from bitou bush and glyphosate application should be avoided in areas where Pimelea is present. (See also Case 381 for the efficacy of manual cutting in the control bitou bush).

Study 2

In Australia, weed invasion has caused the extinction of several endemic plant species and threatens others. Rare plants with small population and range sizes, and which are competitively inferior to invasives, are at particular risk. For example, Pimelea spicata is an endangered endemic shrub threatened by habitat loss and invasion by weeds, including the bitou bush Chrysanthemoides monilifera ssp. rotundata which infests about 70,000 ha of the New South Wales coastline. Bitou was first recorded in New South Wales around 1908, probably introduced through dumping of ships' ballast. From 1946 to 1968 it was planted along the coast to revegetate areas after mining. It is now regarded as one of the worst weeds in Australia because of its invasiveness and associated detrimental economic and environmental impacts.

The winter application of the herbicide glyphosate is used to control bitou, whilst other native plants such as sclerophyllous shrubs e.g. Acacia (wattles), Banksia (banksias) and Leptospermum (tea-trees), and large Lomandra (mat-rush) tussocks are resilient. However, no attempts at bitou control by glyphosphate application had been undertaken in areas with Pimelea as the effects of glyphosate on this shrub were poorly known. Observations at sites where wind had caused glyphosphate to drift over areas where Pimelea was present suggested that it induced reduced growth and productivity. This study investigates the effect of glyphosate application and two different watering regimes on P.spicata, two native plant species and of two weed species under glasshouse conditions.

The experiment, undertaken in a glasshouse, used a fully-factorial block design, with three factors:

1) Species – seedlings of five co-occurring species were used: the endangered endemic shrub Pimelea spicata; the native kangaroo grass Themeda triandra; Kikuyu grass Pennisetum clandestinum (native to East Africa); the invasive bitou C.m.rotundata; and the coastal wattle Acacia sophorae (commom shrub of coastal dunes growing to 1-4 m high);

2) Watering regime – high water or low water;

3) Herbicide – glyphosate or untreated control.

There were six blocks, each containing 20 plants (5 x species, 2 x watering, 2 x herbicide).

Seedlings: Experimental plants were germinated from seed sown onto potting mixture. P.spicata seeds were collected from plants grown outdoors at the Australian National Botanic Gardens, Canberra. The parental stock of these plants originated from the Cumberland Plain and the Illawarra coast, the seeds were pooled to ensure a sufficient sample size. Seeds of the other species came from commercial suppliers (A.sophorae and P.clandestinum) or collected from the field (T.triandra and C.m.rotundata). Seedlings 10-15 cm high were transplanted into 8 cm diameter pots, which were randomly allocated within the experimental design.

Watering: 'High watered' plants were watered every day with 100 ml of tap water; 'low watered' plants were watered every second day with 50 ml of tap water. Individual leaves on each plant in three randomly selected blocks were tagged to monitor stomatal conductance. Plants suffering water stress have lower stomatal conductance than non-stressed plants, because their stomata close to prevent water loss. Stomatal conductance of the tagged leaf was measured every three days from 07:00 hr using a Porometer (ΔT^TM) to record the time taken for the relative humidity to increase by 10% within a small chamber clamped onto a leaf.

Herbicide: Glyphosate herbicide (1:180) was applied four weeks after the experiment began (after it was revealed that the stomatal conductance of water-limited plants is greater than that of regularly watered plants). The experiment was run for another seven weeks after herbicide application.

Seedling growth: At the end of the 11 week experiment the oven dried shoot and root mass was determined, each plant was allocated to one of five health categories:

1 = healthy plant
2 = 1-25% of above-ground biomass dead
3 = 26-50% dead
4 = 51-75% dead
5 = 76-100% dead

Survival: All herbicide free plants survived until the end of the experiment, but the application of glyphosate resulted in clear differences. Irrespective of watering regime, the survival of herbicide-treated Pimelea was very low (17%) and all Kikuyu grass survived. For glyphosate treated kangaroo grass, bitou and coastal wattle, survival of well-watered individuals was lower than for 'water-stressed' individuals. Indeed, all coastal wattle (a xeric species) plants died when well-watered and all survived when water-stressed.

Growth: The application of glyphosate reduced plant mass, though the magnitude was affected by water regime and species. Pivotally, the mass of Pimelea and bitou was limited by glyphosate regardless of water treatment. In contrast, the mass of kikuyu and kangaroo grasses, and coastal wattle, was not affected by herbicide when water-limited, but was significantly reduced when well-watered.

Health: Plant health was negatively affected by herbicide application. This negative affect was particularly severe for kikuyu grass, kangaroo grass and coastal wattle under well-watered conditions.

Conclusions: Kikuyu grass, kangaroo grass and coastal wattle plant weight was not affected by glyphosate application when water-limited but it  was significantly reduced when these plants were well-watered. However, both Pimelea and bitou were intolerant to glyphosate irrespective of watering regime. Consequently, the authors suggest that weeding rather than herbicide application is the best option for protecting Pimelea from bitou bush encroachment.


Note: If using or referring to this published study please read and quote the original paper, this is available at:

http://www.environmental-expert.com/magazine/elsevier/biocon/index.htm

Please do not quote as a www.conservationevidence.com case as this is for previously unpublished work only.