You are on page 1of 4

We suggest that foliar spraying will be most efficient for

A pilot project testing the combating large infestations of lantana, while hand pull-
effectiveness of three weed ing techniques are recommended for creating firebreaks
control methods on the or when minimizing damage to native species is para-
mount.
removal of Lantana camara
in Forty Mile Scrub National Introduction
Park, Queensland, Australia Lantana (Lantana camara L.) poses a serious risk to Aus-
tralian ecosystems through its reduction in biodiversity
(Humphries et al. 1991), loss of agricultural land (Sinden
Justin Dohn1,2, Z. Carter Berry1,3 and Timothy J. Curran1,4
et al. 2004) and toxicity to native animals and livestock
(1School for Field Studies, Centre for Rainforest Studies,
(Oosterhout 2004). Its high resilience following distur-
Yungaburra, Qld, Australia; Email: justin.dohn@
bance and tolerance of a variety of environmental condi-
colostate.edu; 2Department of Biology, Wake Forest
tions enables this weed to cover a variety of Australian
University, 137 Winston Hall, Winston Salem, NC, 27109,
habitats (Oosterhout 2004). Lantana has also been impli-
USA; 3Department of Biology, Wake Forest University,
cated with an increase in the intensity of forest fires by sig-
135A Winston Hall, Winston Salem, NC, 27103, USA;
4 nificantly increasing the biomass available to fuel fires
Department of Ecology, Lincoln University, PO Box 84,
(Fensham et al. 1994; Berry et al. 2011). These fires
Canterbury, 7647, New Zealand).
provide space for further lantana colonization, creating a
fire-lantana-positive feedback cycle (Fensham et al.
Key words: Australia, dry rainforest, Lantana camara,
1994). Australian dry rainforest ecosystems are particularly
semi-evergreen vine thicket, weed control.
vulnerable to lantana invasion, which can inhibit native
seedling growth, suppress native flora, negatively impact
Summary fauna (Gentle & Duggin 1997) and modify fuel loads
(Fensham et al. 1994; Berry et al. 2011).
Lantana (Lantana camara) poses a serious threat to the Weed control to manage existing infestations of lantana
biodiversity of the dry rainforest vegetation at Forty Mile has varied considerably in form, function and success.
Scrub National Park, Queensland, Australia, by outcom- Manual techniques, including pulling, grubbing and cut-
peting native species and increasing vulnerability to fire. ting, have been successful, but are extremely time and
This pilot study tests the effectiveness of three weed con- labour intensive (Tu et al. 2001). Herbicides have been
trol methods (hand pulling, a glyphosphate-based foliar used to treat large thickets, but depending on active con-
spray herbicide and a picloram- and triclopyr-based basal stituents, application rates and methods, costs and climate
bark herbicide) in removing lantana and their success in conditions, these differ in their effectiveness (e.g. Somer-
reducing lantana fuel loads. The foliar spray herbicide ville et al. 2011). We report a pilot project comparing
was the most effective in killing lantana, while manual effectiveness of three lantana control techniques (hand
pulling resulted in the largest decrease in fuel height. pulling, a foliar spray herbicide and a basal bark herbi-
SHORT
REPORTS

cide) on the heavily invaded dry rainforest vegetation at One month later (April 2008), measurements were
Forty Mile Scrub National Park, Queensland, Australia. taken from four randomly selected 5 9 5 m subplots
We evaluated these techniques for their effectiveness in within each plot. The height of the highest lantana individ-
removing lantana and reducing fuel loads. We also pro- ual was recorded at 1-m intervals around the perimeter
pose a lantana control programme for Forty Mile Scrub and along the diagonals of the subplot. Next, all twig bio-
NP by integrating these results with a detailed cost–benefit mass of lantana was collected from within 1 m3 frames
analysis. constructed at two corners of the subplot. All live lantana
plants within the plot were counted and placed into height
Methods categories: 5–30, 30–100 and 100 cm and above. Finally,
the number of dead lantana plants in each herbicide plot
Study area
was recorded and organized into the same size designa-
The study was conducted at Forty Mile Scrub National tions.
Park (18°08′S 144°50′E), a 6300 ha park conserving exten-
Analysis
sive tracts of dry rainforest. In 1993, 73% of the dry rainfor-
est in the park had densities of lantana greater than 5000 One-way ANOVA was used to evaluate differences
plants per ha (Fensham et al. 1994). between treatments, and Tukey’s HSD pairwise post hoc
tests were used to investigate significant relationships.
Treatment procedure
All data were tested for normality and homogeneity of
Queensland Parks and Wildlife Service (QPWS) applied all variances, and Kruskal–Wallis tests were used when data
three treatments in March 2008 (wet season) on violated these assumptions (Table S1). Analyses were con-
50 9 50 m plots, with a fourth plot left untreated. Hand ducted using the computer program JMP 7.0 SAS Institute
pulling (Treatment 1) was conducted by pulling every Inc., Cary, NC, USA.
identifiable lantana plant in the designated plot. Foliar her- It is important to acknowledge the existence of ‘pseu-
bicide (Treatment 2), with the active constituent glyphos- doreplication’ inherent in this study (Hurlbert 1984). The
phate at 9% concentration, was administered in the plot to replicated treatment plots in the experimental design
the foliage of lantana with an effort to target only lantana. are, in fact, replicated samples of the same treatment plot
Basal bark spray (Treatment 3) was administered to the and thus are pseudoreplicates. However, this study still
base of lantana plants from the ground to 30 cm up the provides useful insights on the effectiveness of these dif-
stem (active ingredients: picloram and triclopyr, 17% in ferent lantana control measures and can inform future
diesel; Clark et al. 2006). management within Forty Mile Scrub National Park.

A
160 Total
>100 cm
140 30–100 cm
5–30 cm
120
Number of live lantana

A
100
AB

80

AB
60
AB
B
40 B
A

AB AB
20 A
B B B
B B
0
Basal Bark Control Foliar Spray Pulling
Treatment
Figure 1. Mean number of live lantana growing in each treatment plot organized into three size categories: 5–30, 30–100 and >100 cm; and total
number of live lantana. Bars within the same size category that do not share the same letter are significantly different (Tukey’s pairwise HSD, P < 0.05,
Error bars = +1 standard error of mean).
SHORT
REPORTS

Results lantana biomass, justifying fuel height as an appropriate


There was a significantly higher total number of live lan- surrogate measure of lantana biomass in the plots. There-
tana plants counted in the control plots compared to both fore, we suggest that all three treatments considerably
the pulling and foliar spray plots (Fig. 1; Table S1). No dif- decrease the biomass of lantana, with the largest of such
ference was found between the number of lantana plants decreases coming with the pulling method. This is an
in the basal bark plots and any of the other plots. The important effect in an ecosystem where, in the absence
foliar spray plots showed significantly more dead lantana of lantana, fires are rare due to a lack of fuel at the ground
plants than the basal bark plots (Fig. 1; Table S1). level (Fensham et al. 1994) and effective fire risk manage-
There was a significant difference in the average fuel ment involves the reduction in the fuel load in this
load height between the control and each of the treatment stratum.
methods, as well as between the herbicide treatments and
Forty mile scrub NP Lantana camara management plan
the pulling plots (Table S1; Fig. S1). Regression analyses
found a significant positive correlation between the mean A principle limiting factor in the control of harmful weeds
corner fuel height of the plots and the corresponding total in ecosystems is the financial investment that land manag-
corner biomass (Fig. S2). ers are willing to make; inadequate funding is often the
primary contributing factor to the rapid dispersal of weeds
Discussion (Sinden et al. 2004). The estimated material cost of the
basal bark herbicide treatment used here was $0.67/L,
Effectiveness of lantana removal
while the foliar spray herbicide was $0.08/L. Costs of man-
While all three treatments reduced the counts of lantana, ual weed removal were calculated using pulling rates and
only the foliar spray and pulling plots had significantly 2008 wages of the QPWS staff; it is estimated to cost $1,920
fewer lantana than the control, perhaps due to the delayed for one person to remove a hectare of heavily infested lan-
response of the basal bark spray. This treatment tana. To clear the same amount of lantana, it would cost
requires an estimated 9–12 weeks for the death of the QPWS $128 to use the foliar spray and $840 to use the
plant following application (Clark et al. 2006), and thus, basal bark method. In addition to financial considerations,
plants counted as live during data collection may have preserving biological diversity is vital at Forty Mile Scrub
died in subsequent weeks. NP due to the magnitude and intensity of the invasion in
The significantly higher number of smaller lantana the park. QPWS employees counted over 10,200 plants
plants in the pulling plot compared to the foliar spray plot pulled in the hand pulling plot (Simon Thompson, QPWS,
may reflect the difficulty of correctly locating, identifying personal communication), which corresponds with a den-
and removing very small lantana seedlings. Alternatively, sity of over 40,000 plants per hectare. As such, large
the weed control treatments applied in this study may amounts of lantana need to be removed at an affordable
have differential effects on regrowth, potentially affecting cost to minimize its impact on the environment while min-
plant counts in our post-treatment survey. A foliar spray imizing the financial burden on park managers.
herbicide leaves targeted plants rooted and maintains can- Our preliminary results suggest that the foliar spray is
opy cover and enters the seed bank inhibiting the growth the most effective and economical, saving QPWS $712
of seedlings (Tu et al. 2001). In the pulling treatment, per ha compared to basal barking. If all of the approxi-
however, plants are removed completely and the dense mately 1,990 ha infested were treated, this would save
subcanopy of lantana is entirely eliminated. This leaves QPWS up to $1,416,880 in operating costs compared to
behind a patch of soil that has ample sunlight, an abun- basal barking.
dant lantana seed bank and few other competing species, The pulling technique was very effective at removing
potentially stimulating fast regrowth of newly germinated lantana and reducing fuel loads. However, the large
seedlings and thereby increasing the small lantana counts amount of time required for this process (64 person
in the pulling plot. hours per ha of heavily infested terrain) significantly
inflates the financial cost for large-scale projects. We pro-
Effect on the fuel load
pose that hand pulling methods be focused primarily
All three treatments resulted in significantly decreased fuel along scenic paths and edge areas vulnerable to fire inva-
height, with the pulling method the most effective. The sion to create a fire buffer zone from surrounding ecosys-
two herbicide treatments showed no significant differ- tems (Fensham et al. 1994). Follow-up control must
ences to each other, as these treatments do not include include the close monitoring of the treated areas to ensure
any mechanical means to remove the dead plants from native species re-establishment instead of lantana recolon-
the ground. ization. Our proposal mirrors the guidelines laid out by
Regression analyses found a significant positive correla- the Lantana Best Practice Manual and Decision Support
tion between the mean fuel height of lantana and nearby Tool (Stock et al. 2009), which calls for prioritization of
SHORT
REPORTS

management areas and an integration of multiple control Fensham R. J., Fairfax R. J. and Cannell R. J. (1994) The invasion of Lantana
camara L in Forty Mile Scrub National Park, north Queensland. Austra-
methods and follow-up monitoring. lian Journal of Ecology 19, 297–305.
Lantana poses a threat to the biodiversity of a range of Gentle C. B. and Duggin J. A. (1997) Lantana camara L. invasions in dry
Australian ecosystems, and appropriate action must be rainforest open forest ecotones: the role of disturbances associ-
ated with fire and cattle grazing. Australian Journal of Ecology 22,
taken to ensure that this threat is reduced. We suggest that 298–306.
foliar spray herbicides most effectively remove lantana Humphries S. E., Groves R. H., Mitchell D. S., Hallegraef T. G. M. and Clark
and thus may be most efficient in combating large-scale J. (1991) Plant Invasions Kowari 2. Australian National Parks and Wild-
life Service, Canberra.
invasions in dry rainforests, while hand pulling techniques Hurlbert S. H. (1984) Pseudoreplication and the design of ecological field
may be most beneficial in ecologically sensitive areas or to experiments. Ecological Monographs 54, 187–211.
establish firebreaks around areas threatened by changed fire Oosterhout E. V. (2004) Lantana: Current Management and Control Options
for Lantana (Lantana camara) in Australia. Department of Natural
regimes. Following effective weed removal, cost-effective Resources, Mines and Energy, Brisbane.
management of dry rainforest will depend on close monitor- Sinden J., Jones R., Hester S. et al. (2004) The economic impact of weeds
ing of susceptible areas and quick treatment responses to in Australia. CRC for Australian Weed Management, Technical Series
8, 55.
lantana along park borders to minimize the magnitude of Somerville S., Somerville W. and Coyle R. (2011) Regenerating native forest
invasions. Ultimately, frequent monitoring and quick treat- using splatter gun techniques to remove Lantana. Ecological Manage-
ment will be more cost-effective than allowing areas to ment and Restoration 12, 164–174.
Stock D., Johnson K., Clark A. and van Oosterhout E. (2009) Lantana Best
become infested at the densities of our study area. Practice Manual and Decision Support Tool. The State of Queensland,
Department of Employment, Economic Development and Innovation, Ye-
erongpilly, Australia.
Acknowledgements Tu M. C., Hurd J. and Randall M. (2001) Weed Control Methods Handbook:
We thank Brett Addis, Colin Venner, Christian Skorik, Erin Tools and Techniques for Use in Natural Areas. The Nature Conservancy:
Wildland Invasive Species Team. [Accessed October 10, 2012] Available
Kenison and Erica Rubin for help with fieldwork. Simon from URL: http://www.invasive.org/gist/handbook.html.
Thompson and other QPWS staff applied the treatments.
Funding for the experimental treatments was provided Supporting Information
by QPWS, while School for Field Studies covered other Additional Supporting Information may be found in the
costs. This study was conducted under research permits online version of this article:
WISP04234607 and WITK04234707 to TJC. We thank three Figure S1. Mean lantana fuel height found in each treat-
referees for their valuable feedback on an earlier draft of ment plot.
this manuscript. Figure S2. Regression analysis of mean total lantana
biomass taken from cubic meter corner sections in each
References treatment plot against mean corner fuel height
Berry Z. C., Wevill K. and Curran T. J. (2011) The invasive weed Lantana
camara increases fire risk in dry rainforest by altering fuel beds. Weed (y = 3.1355x + 159.37, r2 = 0.3218, P < 0.001).
Research 51, 525–533. Table S1. Results of one-way analysis of variance compar-
Clark A., Raven C. and Stock D. (2006) Using Herbicides on Lantana: A ing Lantana camara treatment methods to a range of
Guide to Best Management Practices. Department of Natural
Resources, Mines and Water, Queensland. variables.

You might also like