|dc.description.abstract||Studies of the common grass grub (Costelytra zealandica (White)) covered two major aspects, population ecology and pest assessment. The object or these studies was to investigate the feasibility of developing models, predicting population density and estimating the associated losses in pasture productivity, as a basis for the formulation of a pest management programme.
I. Population Studies
Studies of natural populations of soil insects such as grass grub require the development of accurate and efficient mechanical methods of sampling and extraction as well as a statistically precise and efficient sampling plan.
1. Population Studies
a) Mechanical aspects.
A split barrelled manually operated corer was developed for sampling grass grub. This implement permitted 175 to 245 samples to be taken daily by one man. Sample sites were drawn randomly within strata by a computer as rectangular co-ordinates and converted to polar co-ordinates, originating from the centre of the 20, 20 x 20 m subplots in each study plot. Sites were then sorted and listed in numerical order based on the polar ordinate. The use of the computer allowed large savings in the time taken to draw and list random co-ordinates and made it possible to randomly draw samples from irregular shaped strata. A large compass wheel enabled sample sites to be located rapidly in the field.
Sampling times were based on beetle flights, seasons and, in the case of pupal sampling, pilot sampling.
Extraction or all the developmental stages of grass grub from soil was accomplished by means of a modified Ladell process which gave a 96 to 98% recovery rate and, dependent on the stage of the insect, a processing rate of between 4.4 and 6.5 man minutes per sample. The extraction process did not impair the insect’s viability.
b) Statistical aspects.
The main consideration in the development of a sampling plan was to obtain, for the lowest cost, an estimate of population density with a precision of ± 10% SE. The variance minimizing efficiencies of different methods of stratified sampling and sample allocation were assessed. The different strata used were based on the division of the study plots by; subplots, damaged and undamaged pasture and damaged and undamaged pasture within each subplot. The most efficient method of stratification was influenced by the accuracy with which damaged areas of pasture could be defined and the population level. Although optimal allocation of samples gave spectacular gains in efficiency, the use of this method of sample allocation in the field was not practical.
The adoption of a flexible rather than a rigid sampling plan in which the strata varied with population level and the ease with which pasture damage could be identified, enabled, except at very low population levels, the required level of precision to be attained with the resources available.
Sample size was inversely related to population density but was highest for the egg stage in which grass grub are most aggregated. The sample size required to attain the level of precision sought ranged between 131 sample units, for high third instar larval populations, and 2450 units for low egg populations.
2. Population Dynamics
Over the period of study at Takapau marked changes were not evident in the flight behaviour of female beetles. No parasites, or important invertebrate or vertebrate predators were found at Takapau and the effect of disease organisms on Takapau populations was considered unimportant. In the Waikato region, however, where the native milky disease may infect up to 40% of the third instar grass grub population, disease is considered to be an important factor in population regulation.
Analyses of age specific mortalities within generations indicated that mortality over the autumn and winter was strongly density dependent and above a certain threshold density compensated completely for change in population density. Laboratory experimentation showed that the major contribution to density dependent mortality arose from larval combat which increased as food supply declined. From this observation the hypothesis was proposed that weather conditions which influence the survival of damaged plants and therefore dispersal and aggregation of larvae, influences larval combat. From field data it was found that the autumn threshold for combat mortality over the autumn-winter period was linearly related to pasture production over this period.
At Takapau when summer soil moisture levels were in excess of wilting point larvae were found feeding close to the surface. Under these conditions combat mortality occurred and thus summer mortality was density dependent. When soil moisture levels approached and fell below wilting point summer mortality became linearly related to soil moisture. At Rukuhia as distinct to Takapau highest soil moisture levels under drought conditions were found close to the surface. As a result, larvae did not descend in the soil profile in response to drought and for this reason summer mortality at Rukuhia, under drought conditions, was attributed to the direct effect of the lethally high soil temperatures found near the surface.
It was concluded from these studies that grass grub populations at Takapau fluctuate in response to low summer or high spring soil moisture levels and are regulated in relation to food supply by larval combat. The effect of summer mortality on generation mortality is moderated by the density dependent nature of combat mortality of larvae in autumn-winter. Consequently, mortality in the summer only influences generation mortality if it decreases population density below the threshold density at which autumn-winter combat mortality occurs.
The knowledge gathered from these studies explains why grass grub populations are so difficult to control with transient insecticides and has suggested better ways of using insecticides.
A population model was developed for predicting population changes within and between generations at Takapau. The model was based on the relationships described above and gave encouragingly accurate results. However, it did highlight the need for more accurate definition of the relationships between the major mortalities identified by these studies and the important variables which affect them such as soil moisture, rainfall, pasture production and population density.
II. Pest Assessment Studies
A technique which does not involve the use of insecticides was developed for measuring losses in pasture production arising from grass grub damage. This technique is suited to areas of relatively low summer rainfall where the occurrence of grass grub can be located visibly by pasture damage. The method involves the division of a paddock into, and the measurement of herbage production from, three strata; undamaged areas, areas damaged by previous generations and areas damaged solely by the current generation. The sizes of the strata were measured by aerial photography or estimated from a growth curve of the area visibly damaged or, the relationship established between the area of visible damage and insect density. Given the size and herbage production from each stratum the total production for each plot or paddock can be estimated.
Pasture damaged by grass grub showed deterioration in botanical composition with an increase in litter and grass weeds and a decrease in white clover. Damaged pasture had a higher percentage of bare ground and in the autumn and winter was poorly utilized by livestock. Seasonal herbage losses were highest in autumn and winter, although the highest loss in monthly herbage production was recorded in late summer. In newly damaged areas seasonal losses in herbage production of 70% and 74% were recorded in autumn and winter respectively whereas the respective losses in the areas damaged by previous generations in autumn and winter were 54 and 34%. Recovery, in terms of pasture production, of damaged areas was complete in late spring.
Observations made in these studies showed that within the strata described the severity of damage caused by low and high populations of grass grub was not significantly different, and that increased losses in pasture production associated with higher populations resulted from an increase in the damaged area rather than an increase in the severity of damage in damaged areas. Based on this information losses in pasture productivity caused by different population levels were estimated from the relationship between population density and the area of visible damage and the mean pasture production from each stratum.
Two models were developed which simulated the growth, over successive generations, in area of visible damage under environmental conditions which favoured the increase or maintenance of grass grub populations. Both models describe the growth in area of damage up to a stage where damage was extensive, ill-defined and impossible to measure. One model is based on the relationships between, the size of individual patches of damage and the factor by which these grow over successive generations, and the rate of appearance of damaged areas in the following generation and proportion of the paddock currently damaged. The other model involves the growth curve of May larval populations under conditions which are favourable for population increase and the relationship between population density and pasture damage. The latter model tended to underestimate the actual growth of the area damaged by approximately 20%.
It was found that the establishment of accurate economic threshold levels for different classes of farms run at different intensities and in different regions is attended by many problems. In the light of current knowledge the translation of losses in herbage production into animal production and economic terms for different farming intensities under different climatic conditions will lead to such grossly inaccurate estimates that the worth of pest assessment studies would be lost. From these studies it was concluded that at best pest assessment studies of grass grub could provide the farmer with information that will allow him to predict population density and the associated losses in herbage production. Given this information the farmer is in the position to make more objective decisions, than would otherwise be possible, on the course of action to follow based on his own socio-economic circumstances.||en