Interaction of population processes in ragwort (Senecio jacobaea L.) and ragwort flea beetle (Longitarsus jacobaeae Waterhouse) : a thesis presented in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Ecology at Massey University
Loading...
Date
2000
DOI
Open Access Location
Journal Title
Journal ISSN
Volume Title
Publisher
Massey University
Rights
The Author
Abstract
The primary goal of this study was to improve understanding of biological control of
weeds by investigating how population processes in ragwort and herbivorous insect
interact.
Specific aims were to measure the consumption rates of the three larval instars of
ragwort flea beetle (Longitarsus jacobaeae), to investigate how the process of
herbivory by ragwort flea beetle affects the population density of ragwort, and to
investigate how soil moisture influences the population densities of ragwort flea
beetle and ragwort.
An extraction apparatus was constructed to obtain L. jacobaeae larvae from ragwort
roots and root crowns. This apparatus was 84% efficient .
A preliminary survey of ragwort flea beetle numbers included ragwort plants from
B allantrae, Turakina, and Pahiatua (Southern North Island, New Zealand). The larval
population was highest at Ballantrae but the adul t population was highest at Turakina.
Data were collected from Ballantrae from 1996 to 1998 to develop the interaction
model between L. jacobaeae and ragwort. The interaction depended on the effect that
soil water content had on the populations of both L. jacobaeae and ragwort, the effect
that larval density has on larval mortality, and the effect of ragwort density on the
population of L. jacobaeae larvae. Soil water content was positively correlated with
the increase in numbers of L. jacobaeae. L. jacobaeae larval mortality was dependent
on larval density. High numbers of larvae per plant resulted in a reduction in the
number of larvae over time ( 1 3 .6 larvae/plant on November 1 997 to 1 .8 larvae/plant
in December 1997). The average number of larvae extracted at Ballantrae was lower
in October and November 1996 (4.4 and 4.6 larvae/plant) than in October and
November 1 997 ( 1 3.4 and 1 3. 6 larvae/plant). However, the average numbers of
rosettes was higher in October and November 1 996 (7.6 and 5 .78m -2) than in October
and November 1 997 (2.8 and 2 .7 m -\ There was a significant inverse correlation
between the numbers of L. jacobaeae larvae and ragwort rosettes (-0.4608). When
0.8983 in 1 5 day old larvae, 0.926 1 in 30 day old larvae, and 0.9454 in 45 day old
larvae. The lowest percentage survival (0.9067 in 15 day old larvae) was found at the
highest larval density (40 larvae per plant). Finally, the same experiment was tested
in a field and the data from this was used to construct an interaction model for L.
jacobaeae and its food, ragwort. This model was based on the correlation between
soil water and populations of L. jacobaeae and ragwort; the effect of larval density on
the mortality of larvae and on the weight loss o f ragwort; and on the effect that
ragwort density has on the mortality of L. jacobaeae larvae. Mean soil water was 1 2
± 0.29 to 7 6 ± l.8 1 % over the first 1 5 days, then 3 6 ± 1 . 1 0 to 8 2 ± 0.99% up to 30
days, and 35 ± 0.76 to 65 ± 1 .78% up to 45 days of larval life. These were the soil
water contents that occurred during the field experiment. The model showed that the
highest larval survival again occurred when few larvae were introduced to ragwort
plants ( 1 7.5% survival from 0- 1 5 days, 1 4.33% from 1 6-30 days, and 1 8 .5% from 3 1 -45 days). High larval densities also produced the lowest survival (8.4% survival over
0- 15 days, 5 .87% over 1 6-30 days, and 6.7% over 3 1 -45 days).
The effect of plant density on larval survival was also tested in the field. The highest
larval survival (10.76%) occurred when there were on 1 6 plants m-2, and the larvae
were 0 to IS -days old. The lowest larval survival (6. 6 1 %) occurred with 1 6-30 day
old larvae on plants at a density of 4 plantsm-2. A cohort life-table was constructed
for predicting population fluctuations of L. jacobaeae. Values from this life table
were used to model populations of L. jacobaeae, ragwort and the interactions between
these species using "STELLA" software. Data for the ragwort model was obtained
from published papers. Additional data from the experimental determination of
feeding rates of L. jacobaeae larvae were used when both the L. jacobaeae and
ragwort models were combined to examine the interactions between these species.
This latter model was used to estimate population fluctuations of L. jacobaeae and its
food over two years. It indicated that L. jacobaeae is a very effective control agent
for ragwort, and that it can cause ragwort populations to decline to extinction within
two years.
Description
Keywords
Weed control, Biological control