Preamble

The key objectives of the 1981-82 studies were to assess the actual crop loss attributable to S. incertulas and the potential for controlling this and other borer species. While crop loss could be observed directly as deadhearts and whiteheads (both denoting death of the apical internode, the latter term being used for unfilled panicles), the loss caused by the feeding activity of borer larvae in parts of the stem other than the terminal, panicle-bearing internode was not felt by the 1979 review team to have been properly resolved.

In agreeing with van Emden and the other members of the review team, I was running counter to the view propagated by David Catling.

In addition to the short summary notes submitted to the IRRI Newsletter (Taylor, 1984, Taylor & Islam, 1984), which I have commented on earlier as having been altered in the editing process without my knowledge, the comments and conclusions which follow began life (other than in internal reports, that is) in my 1988 paper. After the publication by Catling and his co-workers of what I regarded as misleading reports, I had further considerations, including references to more recent evidence, published in 1996. What follows is a pulling together of those various submissions. Perhaps, I could emphasize that I had access to all the actual field records and data analyses from the first phase, so my comments are based on a careful examination of data and not reliance on considering simple summaries.

The "infestation must equal damage concept"

This particularly stems from H.D. Catling who wrote, in 1979, "The yellow rice borer was consistently found to be a severe pest, especially in the stem elongation phase when very high levels of stem infestation were recorded. Larval feeding damages and destroys main stems and basal tillers which then stimulates compensatory branching. Notwithstanding this, stem density is lowered and much yield is lost" (Catling, 1979).

After further experiments, he added that "There was some evidence that early borer attacks (in June and July) cause greater yield loss. In the pot experiments and the 1978 field experiment most yield loss was due to early attack, whereas in the two 1979 field experiments, where the loss was considerably less, the outbreaks occurred in August and September" (Catling, 1981).

At the 1981 Deepwater Rice Conference, Catling and his co-authors held to the basic precept that larval feeding causes 'damage' and as a consequence high numbers of larvae lead to considerable 'damage'; and asserted that there is a damaged stems : yield loss ratio of about 1 : 1 (Catling et al. 1982, Catling & Islam 1982a, Catling & Islam 1982b) .

Subsequently, this was modified by Catling et al. (1987), who claimed that, assuming a linear relationship, there is an approximate ratio of 1% yield loss from every 2% increment of stem damage in deepwater rice, and thus, in Bangladesh and Thailand, from the average rates of 38-44% damaged stems at maturity it follows that a yield loss >20% is common in many fields.

Investigation of S. incertulas control with insecticides

Catling's earliest claim seems largely to have been based on the results of field experiments in 1978 and 1979 in which attempts were made to secure an insect-free crop and so to determine crop loss. I suspected, however, that the use of insecticide treatment at regular intervals throughout the season must have precluded the determination of when borer attacks had the most effect on the crop. An examination of the 1978-79 field data, moreover, showed that, at harvest in both years, some 75% of the infestations were at the top of the stems.

Population dynamics studies had shown that the S. incertulas populations in Bangladesh displayed the characteristic succession of distinct broods throughout the season (Catling 1979, 1981; Torii 1971). This offered a possibility for control by insecticide application at distinct points in time, and the 1981 study at Bastia, therefore, was designed to evaluate the effect of pre-flood use of insecticides. The last application of the series, on 24-25 June, was just before the first floodwater entered the fields. Although some reduction in infestation appeared to be achieved by use of certain insecticides, the yield data showed no related significant differences between treatments.

The study, together with an experiment in a deepwater pond, facilitated the construction of a population model. What emerged also was that, because there are successive broods, stem dissections over a period of time may show an apparent decline in the borer larval population, but this could be simply a consequence of the increasing tiller numbers so that there are more tillers but no more larvae.

Thus, in 1982, a repetition of the pre-flood application of insecticides was planned as a factor in multi-factorial trials at several sites. Light trap results and stem dissections both showed very low S. incertulas populations. The adult numbers crashed after a late April peak and almost no live larvae were found from mid-May to late July (Z. Islam & B. Taylor, in Research work conducted by the ODA/BRRI Deepwater Rice Project. Unpublished report to the BRRI internal review meeting, 1983). High temperatures (up to 37.9°C) appeared to have been the main factor in this crash. In mid-August, a considerable increase in S. incertulas moths in light traps gave the opportunity to study the impact of insecticide application on a specific borer brood.

Not only were significant reductions in infestation and in whitehead numbers achieved (Table III) but also yield savings of some 8-6% in the variety Boron Bawalia. In two fields of the variety Khama, a significant reduction in whiteheads (around 7%) resulted from a single insecticide application. Of the four rice varieties treated, the variety Sonna Digha had no reduction in whiteheads although a significant reduction in stem infestation was recorded. A possible explanation is that this variety matured earlier and, thus, the panicle development was more advanced, making it susceptible earlier than the other varieties to damage to the terminal internode.

The level of yield saving in this trial, involving one to three applications of monocrotophos, was

a. - very similar to that achieved by the 1979 series of 20 applications of diazinon (Catling 1981, Catling et al. 1987), and,

b. - of the same order as the reduction in whiteheads.

More recent insecticide studies

Islam, Catling & Pojananuwong (1988) gave a summary of trials in which several methods of application and timing of insecticides in farmers' fields were tested in Bangladesh and Thailand. With regard to whole-season spraying, they went back to the early field experiments in Bangladesh (in 1978 and 1979) in which an insecticide, diazinon, was applied at regular intervals throughout the season in the expectation that an insect-free crop would result, and so a direct determination of crop loss could be achieved (Catling 1981; Catling, Islam & Pattrasudhi 1987).

In 1983, the late-season tests were repeated by Islam, this time using five fields of different genotypes. Again, one or two sprays significantly reduced the percentage of stem damage, whiteheads and grain sterility, and increased the panicle density, grains per panicle, and yield. One spray was associated with a reduction in stem damage from 48 to 37%, a yield increase of 8%, and a cost:benefit ratio (C:B) of 1.85; whereas with two sprays the stem damage was reduced to 22% and there was a C:B of only 1.30.

The results led Islam, Catling & Pojananuwong (1988) to conclude that, 'It appears that one well-timed spray was better than two or three poorly timed ones. Thus, in Bangladesh, a single spray in the critical period from late August to early September at the time of the fifth seasonal moth peak can be tried.' Their report is confusing, however, as in relation to various unsuccessful small-scale experiments they had carried in Bangladesh and Thailand, they described the results as 'unexpectedly poor'. Curiously, this negative view appears to contradict that expressed previously by Catling, Islam & Pattrasudhi (1987).

My examination of the original data (in Catling, 1981) shows that a yield saving of 21.1% was achieved in 1978, and 2.8 and 13.8% in 1979. This yield saving was of the same order as that achieved by spraying solely in the late-season (see below).

I favour the term 'yield saving' as being more realistic than 'yield increase' when the results of crop protection measures are expressed. For instance, while the use of a fertilizer may well promote grain production; with very few exceptions, insecticides do not act to stimulate yields but only prevent grain loss due to the action of a pest. Thus, results are better expressed from the formula:

(yield in insecticide treated yield in untreated)/ yield in treated) X 100 = yield saving %

Early-season spraying was further tried in Bangladesh in 1983, when one half of each of two fields in a stem borer hot spot area was sprayed with monocrotophos at 250 g ai/ha with a knapsack sprayer. The unsprayed half of each field served as a check treatment. Stem population density and deadhearts were assessed by quadrat counts, and stems were dissected for borer damage. Field 1 was sprayed at basal tillering stage, on 8 June at the time of peak moth activity. One week later, the spray had reduced deadhearts from 9.2 to 1.8/m², but the number of damaged stems was not affected (21 and 23%). Field 2 was sprayed on 15 June, one week before flooding, when deadheart density was 12-13/m². Three weeks later, at a water depth of 100 cm, the two halves of the field did not differ in stem population or damage (Islam, Catling & Pojananuwong, 1988).

Overall, my suspicion is that Islam, Catling & Pojananuwong (1988) did not have appreciate the real significance of the studies on critical timing. For instance, they described the results from the season-long spraying trials as 'unexpectedly poor', but this remark seems to have been made in hindsight, as it was not what was reported originally (see above). Actually, the yields from the sprayed fields were better than the unsprayed but by no more than was achieved with the critically-timed sprays (Taylor 1988).

The very fact that the critically-timed spraying yielded results which could not be bettered by season-long applications is strong evidence that the mid-season, elongation phase is not affected in any crucial manner by S. incertulas larvae.

continue

href="\dwr\ysb11.htm"