Susceptibility of great bindweed (Calystegia silvatica)to herbicides

Abstract

A ield trial was conducted in Palmerston North to compare autumn applications of several translocated herbicides for great bindweed (Calystegia silvatica) control in riparian zones. Regrowth in the following spring showed that a triclopyr/picloram/ aminopyralid mixture, a 2,4-D/dicamba mixture and aminopyralid by itself were the three most effective treatments, though none gave complete control. Glyphosate provided partial control whereas metsulfuron and clopyralid provided poor control. These and two other herbicides were further assessed in a glasshouse trial in which they were applied to leaves either on the upper or lower half of plants to compare eficacy. The relative effectiveness of these herbicides on great bindweed was similar to that found in the ield. Most herbicides had similar eficacy whether applied to upper or lower parts in autumn, except glyphosate, which was more effective applied to upper plant parts. Implications for control of great bindweed in riparian plantings are discussed.

Full text

94
Weed biology & control
New Zealand Plant Protection 66: 94-98 (2013)www.nzpps.org
T.L. Gawn, K.C. Harrington and C. Matthew
Institute of Agriculture and Environment, Massey University, Private Bag 11-222,
Palmerston North 4442, New Zealand
Corresponding author: K.Harrington@massey.ac.nz
Abstract A field trial was conducted in Palmerston North to compare autumn applications
of several translocated herbicides for great bindweed (Calystegia silvatica) control in
riparian zones. Regrowth in the following spring showed that a triclopyr/picloram/
aminopyralid mixture, a 2,4-D/dicamba mixture and aminopyralid by itself were the three
most effective treatments, though none gave complete control. Glyphosate provided partial
control whereas metsulfuron and clopyralid provided poor control. These and two other
herbicides were further assessed in a glasshouse trial in which they were applied to leaves
either on the upper or lower half of plants to compare efficacy. The relative effectiveness of
these herbicides on great bindweed was similar to that found in the field. Most herbicides
had similar efficacy whether applied to upper or lower parts in autumn, except glyphosate,
which was more effective applied to upper plant parts. Implications for control of great
bindweed in riparian plantings are discussed.
Keywords great bindweed, Calystegia silvatica, triclopyr, picloram, dicamba, aminopyralid,
glyphosate, riparian.
Susceptibility of great bindweed (Calystegia silvatica)
to herbicides
INTRODUCTION
In Taranaki and many other parts of New
Zealand, riparian zones next to pasture land are
being fenced off and planted in native shrubs and
trees to help reduce detrimental effects of farming
on waterways. Many of these riparian plantings
are now being threatened by severe competition
exerted by great bindweed (Calystegia silvatica),
which has become dense at some sites, often
totally covering planted species (Wilson-Davey
et al. 2009).
As this deciduous, perennial weed has seldom
caused much economic harm in the past, few
studies have been conducted on its biology and
control. However, much of its success is due to
an aggressive rhizome system (Williams 2009).
There had been uncertainty over whether it
produces much seed, but recent work by Gawn
(2013) has found that, on average, each flower
produces 1.5 seeds, with 71% viability. Seeds are
relatively large (43.4 mg) and only germinate
following scarification.
There has also been considerable confusion over
the taxonomy of bindweeds, especially differences
between Calystegia sepium (hedge bindweed),
Calystegia silvatica and their sub-species (Ogden
1978). A review of the taxonomy of these two
species followed by a brief field survey by Gawn
(2013) showed that most of the bindweed plants
in Manawatu and Hawke’s Bay are Calystegia
silvatica ssp. disjuncta.
The susceptibility of this weed to herbicides is
not known due to a lack of trial work, and thus
© 2013 New Zealand Plant Protection Society (Inc.) www.nzpps.org Refer to http://www.nzpps.org/terms_of_use.html

95
Weed biology & control
extrapolation has often been necessary from
work done overseas on Convolvulus arvensis
(field bindweed) and Ca. sepium (Wilson-Davey
et al. 2009). Work by Rahman & Sanders (1992)
on great bindweed emerging from rhizomes
within New Zealand asparagus crops showed
few herbicides had much effect on it apart from
imazapyr, which is not suitable for use among
young native shrubs beside waterways. A 2,4-D/
dicamba mixture has often been recommended
for great bindweed control in waste places
(O’Connor 1984).
The objective of this work was to determine
the relative efficacy of several translocated
herbicides on great bindweed when applied in
autumn just prior to winter dormancy of the
weed, and also to determine if these herbicides
would work effectively if only applied to part of
each plant as might be necessary when spraying
near native plants.
MATERIALS AND METHODS
Field trial
A field trial was conducted in ungrazed areas
near the Turitea Stream and adjacent to the
Massey University No. 1 Dairy Farm on Poultry
Farm Road, Palmerston North. Plots were set
up around discrete colonies of established
great bindweed plants, identified as Calystegia
silvatica ssp disjuncta. The plants were growing
over a range of different self-planted tree, shrub,
perennial grass and herbaceous weed species,
such as willow (Salix spp.), elder (Sambucus
nigra), poroporo (Solanum aviculare), Yorkshire
fog (Holcus lanatus), hemlock (Conium
maculatum) and wandering Jew (Tradescantia
fluminensis). There were 28 plots in total and the
great bindweed plants varied in size.
On 16 April 2012, six herbicide treatments
(Table 1) were applied to the great bindweed
plants and compared with untreated plants
within a randomised complete block design
(blocking based on plant size) with four replicates.
Each herbicide was mixed with Done That
(triarylmethane dyestuff, Farmers Industries Ltd)
marker dye (but no surfactants) and was applied
only to great bindweed foliage using a 15-litre Solo
backpack sprayer, thoroughly wetting leaves but
trying to avoid run-off to plants beneath the great
bindweed plants. A 10% portion of each great
bindweed plant was left unsprayed to simulate
avoiding spraying near wanted native plants.
The maximum and minimum temperatures for
the 2 weeks following spraying (taken from the
nearby AgResearch Grasslands weather station)
were 17.8°C and 5.3°C respectively.
The health of the great bindweed plants in each
plot was visually scored each week until the plants
had died back for winter. Scores were assigned
from 0 (for plants in which all foliage had died) to
10 (where foliage was very healthy). When plants
began regrowing again in the following spring,
the number of shoots per m2 emerging within
each plot was recorded. All data were subjected to
analyses of variance using SAS and least significant
differences were calculated where treatment
means were significantly different.
Glasshouse trial
Rhizomes of great bindweed plants were
collected from the Massey University orchard on
17 February 2012 and segments averaging 35 cm
in length were planted 3 cm deep within planter
bags each containing 1.5 litres of potting mix.
These were kept in a glasshouse with automated
irrigation at the Massey University Plant Growth
Unit, Palmerston North, and rhizomes or roots
were prevented from growing out of the bottom
of the bags by placing each bag on an inverted
plastic dish. The foliage was trained up strings
suspended from overhead wires so that all foliage
was kept to just one string and trimmed when
they reached a height of 2 m.
On 21 April 2012, all eight herbicide treatments
(Table 1) were applied to the upper half of
plants and these were compared with the same
treatments applied to the lower half of plants.
These treatments plus two untreated controls
were replicated four times and organised within
a randomised complete block design, with initial
health and vigour of plants used for blocking.
Only 1.0 ml of each herbicide treatment was
applied per plant using a small paint brush, with
0.1% organosilicone surfactant (Boost Penetrant)
© 2013 New Zealand Plant Protection Society (Inc.) www.nzpps.org Refer to http://www.nzpps.org/terms_of_use.html

96
Weed biology & control
added to help retain herbicide on treated foliage.
All leaves and stems in the treated half of the
plant were covered in herbicide by this method.
The maximum and minimum temperatures for
the 2 weeks following treatment were 21.7°C and
12.9°C respectively.
The plants were scored weekly following
treatment for state of health as described for the
field trial until they had died back for winter.
The shoot material was then cut back and the
pots were placed in an unheated shadehouse for
the remainder of the trial. Once regrowth began
in spring, shoots regrowing from each pot were
counted. All shoot material was removed on
two occasions (6 December 2012 and 8 January
2013), dried at 75°C and weighed. Roots and
rhizomes were also removed from pots on 9
January, dried and weighed. Data were analysed
as for the field trial.
RESULTS
Field trial
The triclopyr-based mixture caused the most
rapid chlorosis of great bindweed foliage. By 4
weeks after treatment (WAT), all treatments had
adversely affected the plants noticeably, with the
triclopyr-based mixture and glyphosate causing
most damage (Table 2). All great bindweed
plants started to die back for winter dormancy
following the 4 week assessment, making it
difficult to separate treatment effects from
natural senescence. Very little damage could be
found 6 WAT from herbicide treatments on other
plants within each plot.
In the following spring, great bindweed plants
regrew for all treatments, although there was
least regrowth from plots that had been treated
with the triclopyr-based mixture, 2,4-D/dicamba
or aminopyralid. In the absence of follow-up
treatment, the great bindweed re-established
eventually from all herbicide treatments so that
by early December, many plants had recovered
to similar densities to those present prior to
spraying 8 months earlier.
Table 2 Health scores (0 = dead; 10 = healthy)
of great bindweed foliage 2 and 4 weeks after
treatment (WAT), and the density of great
bindweed shoots that emerged in plots after
winter dormancy.
Health score Shoots m2
Treatment 2 WAT 4 WAT 22 WAT
2,4-D/dicamba 4.3 2.5 0.6
aminopyralid 4.3 3.0 1.0
clopyralid 4.5 2.3 3.8
glyphosate 3.5 1.5 2.7
metsulfuron 5.5 3.8 5.3
triclopyr/picloram/
aminopyralid
2.5 2.0 0.3
untreated 7.0 5.3 3.5
LSD (P=0.05) 1.4 2.1 3.4
Glasshouse trial
As with the field trial, the triclopyr-based mixture
produced a rapid response in the glasshouse
trial, but this time so did metsulfuron and 2,4-D
(Table 3). With plants dying back for winter
Table 1 Details of herbicides used in both trials (2,4-D and fluroxypyr were used only in the glasshouse
trial).
Active ingredient Trade name Formulation g ai/100 litres
2,4-D Pasture Kleen ethyhexyl ester 780
2,4-D/dicamba Banvine amine salt 240/120
aminopyralid Tordon Max tri-isopropylamine salt 18
clopyralid Versatill amine salt 90
fluroxypyr Starane 200 methylheptyl ester 100
glyphosate Roundup 360 Pro isopropylamine salt 540
metsulfuron Answer methyl ester 15
triclopyr/picloram/
aminopyralid
Tordon Brushkiller XT amine salt/butoxyethyl ester/
amine salt
90/30/2.4
© 2013 New Zealand Plant Protection Society (Inc.) www.nzpps.org Refer to http://www.nzpps.org/terms_of_use.html

97
Weed biology & control
soon after the 4-week assessment, it was not until
the spring regrowth that the true effect of the
treatments on plants could be seen.
At the first harvest of shoot material on 6
December 2012 (32 WAT), the 2,4-D/dicamba,
aminopyralid and the triclopyr-based mixture,
had only small regrowth (Table 3). The second
harvest on 8 January 2013 (37 WAT) showed
similar results. The metsulfuron and 2,4-D
treatments, which gave rapid knockdown initially,
had regrowth similar to untreated plants.
When underground material was dug up on
9 January 2013 and healthy rhizome tissue was
separated from necrotic tissue and fibrous roots, it
was found that all treatments still had some healthy
rhizome present (Table 3). However, the triclopyr-
based mixture and 2,4-D/dicamba treatments had
the least rhizome material, whereas aminopyralid-
treated pots still had a lot of rhizome. There
was generally little difference in results between
applications to lower or upper parts of the plant
(Table 3), except glyphosate where applications to
upper parts were more effective.
DISCUSSION
Results from both experiments show that the
2,4-D/dicamba mix currently recommended
for control of great bindweed (O’Connor 1984)
is still the best herbicide for controlling it.
However, the triclopyr/picloram/aminopyralid
mixture gave similar levels of control and the
aminopyralid by itself also appeared promising.
Table 3 Health score (0 = dead; 10 = healthy) at 4 WAT, shoot regrowth in spring (32 and 37 WAT), and
dry weight (DW) of healthy rhizome at 37 WAT for great bindweed foliage to which herbicides has been
applied either to the upper (U) or lower (L) parts of the plant.
Treatment
Score
4 WAT
Shoot no./pot
32 WAT
Shoot DW
(g/pot)
32 WAT
Shoot DW
(g/pot)
37 WAT
Rhizome DW
(g/pot)
37 WAT
2,4-D (U) 1.8 5.5 0.73 0.54 5.09
2,4-D (L) 1.0 6.8 0.93 0.54 4.25
2,4-D/dicamba (U) 1.8 0.3 0.03 0.03 1.49
2,4-D/dicamba (L) 3.5 0.0 0.00 0.05 3.25
aminopyralid (U) 3.0 0.3 0.00 0.06 9.67
aminopyralid (L) 3.8 0.5 0.03 0.15 6.48
clopyralid (U) 5.3 2.8 0.43 0.60 15.21
clopyralid (L) 4.8 3.8 0.35 0.78 9.33
fluroxypyr (U) 2.3 5.5 0.58 0.94 8.83
fluroxypyr (L) 2.3 3.3 0.33 0.77 7.43
glyphosate (U) 2.8 1.3 0.03 0.40 8.34
glyphosate (L) 4.0 8.0 0.45 1.27 12.95
metsulfuron (U) 1.8 5.3 0.73 0.64 4.64
metsulfuron (L) 1.0 3.3 0.58 0.43 2.71
triclopyr/picloram/
aminopyralid (U)
1.0 1.5 0.20 0.24 2.30
triclopyr/picloram/
aminopyralid (L)
1.0 0.8 0.00 0.01 0.18
untreated control A 4.0 5.0 0.63 0.74 10.69
untreated control B 5.0 4.0 0.53 1.04 10.80
LSD (P=0.05) 1.4 2.0 0.32 0.39 4.52
© 2013 New Zealand Plant Protection Society (Inc.) www.nzpps.org Refer to http://www.nzpps.org/terms_of_use.html

98
Weed biology & control
Only one application rate was used for each
herbicide, based on recommended rates for
similar species. Some of the other herbicides may
have also performed well if higher rates were
applied. Also, if a surfactant was used in the field
trial, or the entire plant had been treated in the
glasshouse experiment, improved efficacy could
have resulted. Complete control was achieved in
some replicates, but the main objective of this
work was to rank the herbicides for efficacy.
As great bindweed is deciduous, it is assumed
that materials from most leaves are translocated
down to the rhizome system in late autumn. In
spring and summer, leaves in upper parts of the
stem send most sugars towards the shoot tip as
part of the sink source relationship (Willeke et al.
2012). Thus it was expected that herbicide applied
to these leaves in late autumn would move to
the rhizomes as effectively as herbicide applied
to lower leaves, which was confirmed with the
present work. This would be an advantage where
great bindweed plants are growing around and
over newly planted native species in riparian
zones, as herbicides such as 2,4-D/dicamba and
aminopyralid would presumably need to be kept
off most broad-leaved native species (Harrington
& Gregory 2009).
Results from the field trial suggest that it would
not be wise to wait until the following autumn to
respray the regrowth as plants could recover fully
by then. However, respraying as soon as shoots
appear may not be effective as sugar flow would
be mainly upwards from rhizomes to new shoots
initially (Willeke et al. 2012). It may be better to
wait until translocation of sugars back down to
the rhizomes has begun again so herbicides might
be moved into the rhizomes, although only basal
leaves would be expected to send herbicides to
the rhizomes at this stage.
ACKNOWLEDGEMENTS
The authors wish to thank the staff of Massey
University’s Plant Growth Unit and Mark
Osborne for technical assistance, and are also
grateful for funding assistance from the C Alma
Baker Trust, George Mason Sustainable Land
Use Scholarship, Taranaki Regional Council and
Putaruru Veterinarian Club Education Trust.
REFERENCES
Gawn TL 2013. Aspects of the biology, taxonomy
and control of Calystegia silvatica. Masters
thesis, Massey University, Palmerston North,
New Zealand. 119 p.
Harrington KC, Gregory SJ 2009. Field assessment
of herbicides to release native plants from
weeds. New Zealand Plant Protection 62:
368-373.
O’Connor BP 1984. New Zealand agrichemical
manual. Agpress, Masterton, New Zealand.
228 p.
Ogden H 1978. Variation in Calystegia R.Br.
(Convolvulaceae). New Zealand Journal of
Botany 16: 123-140.
Rahman A, Sanders P 1992. Herbicides for
control of two bindweed species (Calystegia
silvaticum and Convolvulus arvensis) in
asparagus. Proceedings of the 45th New
Zealand Plant Protection Conference: 27-30.
Willeke L, Kraehmer H, Gerhards R, Claupein
W 2012. Seasonal variation of the
sprouting ability of rhizome/root buds and
concentrations of storage compounds in
Calystegia sepium (L.)R.Br. and Convolvulus
arvensis L. Proceedings of the 25th German
Conference on Weed Biology and Weed
Control: 694-701.
Williams PA 2009. Biological summaries of
bindweeds (Calystegia) and other climbers.
Landcare Research, Nelson, New Zealand. 17 p.
Wilson-Davey J, James T, Rahman A 2009.
Management and control of greater bindweed
(Calystegia silvatica) in riparian margins in
New Zealand. Landcare Research, Lincoln,
New Zealand. 53 p.
© 2013 New Zealand Plant Protection Society (Inc.) www.nzpps.org Refer to http://www.nzpps.org/terms_of_use.html