Introduction

Diamondback moth (DBM), Plutella xylostella, causes major damage to crucifer crops in tropical regions. In Réunion, these pests are currently managed by intensive chemical control methods, but this strategy generally does not sufficiently reduce crop damage levels and toxic pesticide residues can accumulate on consumed parts of the plants. We therefore conducted a field study to assess the impact of different chemical, biological and physical control methods on diamondback moth population levels.

Materials and methods

The following five field trials were carried out using Alta^® cabbages:

Trial 1

We compared populations of DBM and associated parasitoids on two adjacent 300 m² plots: one plot was treated weekly with broad-spectrum pesticides (pyrethroids, organophosphates) and the other was an untreated control plot in which natural organisms could develop for potential biological control. These test plots were located at Bassin Martin, at 250 m elevation, in a region where monocropped sugarcane predominates. The test plots were monitored between March and May (at the end of the hot humid season).

Trial 2

As in the first trial, moth and parasitoid population patterns in a plot managed with supervised chemical pesticide treatments and in a control plot were compared. The treatments included alternating sprays with Batik^® (Bacillus thuringiensis), Zolone^® (phosalone) and Rocky^® (endosulfan). The other experimental parameters were the same as in the first trial except that the monitoring period was between September and October (cool dry season).

Trial 3

DBM and associated parasitoid populations were monitored on a 400 m² plot located at Bras de Pontho, at 800 m elevation, in a region where vegetable and crucifer crops are mainly grown. No treatments were carried out on this plot, which was monitored from June to August (subtropical winter).

Trial 4

The test protocol was the same as the previous trial, but the plot was located at Piton Hyacinthe, at 1200 m elevation, with a monitoring period extending from April to June.

Trial 5

We assessed the efficacy of an insect net (Lutrasyl P17) to provide physical protection against adult DBM. Two plots planted with 10 cabbage plants from trial 1 were covered with the net immediately after planting.

Sampling

20 cabbage plants were sampled randomly once a week from each test plot, except in trial 5. II to IV instar larvae and cocoons were counted, isolated and reared in plastic boxes until parasitoid emergence after 15 days of incubation. Parasitism rates were calculated overall and separately for the main species. The dominant species of predators and hyperparasitoids were also recorded.

Results

Table 1 shows the auxiliary species detected in the different test plots. Three parasitoid species that are specific to P. xylostella and two Syrphidae (Diptera), including one undescribed species, were detected at the four sites. However, the two nonspecific species of hyperparasitoid identified were only observed in plot 4, which was located at the highest elevation.

Table 1. Beneficial insects and hyperparasitoids observed on DBM in cabbage trials on Réunion Island

Category	Order	Species		Trials
Parasitoids	Hymenoptera	Diadegma mollipla (Holmgren)		all
		Cotesia plutellae (Kurdjumov)		all
		Oomyzus sokolowskii (Kurdjumov)		all
		Tetrastichus howardi (Oliff)		1, 2
Predators	Diptera	Episyrphus sp. nov.		all
		Melanostoma annulipes (Macquart)		all
Hyperparasitoids	Hymenoptera	Notanisomorphella borborica (Giard)		4
		Trichomalopsis oryzae (Risbec)		4

Figures 1 to 6 show the population dynamics of P. xylostella and its two main parasitoids (Cotesia plutellae and Diadegma mollipla) in trials 1, 2, 3 and 4.

Figure 1. Dynamics of DBM and parasitoid populations in four trials on Réunion Island, a) Trial 1 - Control.

Figure 2. Dynamics of DBM and parasitoid populations in four trials on Réunion Island, b) Trial 1 - Chemical control.

Figure 3. Dynamics of DBM and parasitoid populations in four trials on Réunion Island, c) Trial 2 - Control.

Figure 4. Dynamics of DBM and parasitoid populations in four trials on Réunion Island, d) Trial 2 - Chemical control.

Figure 5. Dynamics of DBM and parasitoid populations in four trials on Réunion Island, e) Trial 3 - Control.

Figure 6. Dynamics of DBM and parasitoid populations in four trials on Réunion Island, f) Trial 4 - Control

Trial 1

D. mollipla populations were not taken into account in this trial. There was an increase in P. xylostella populations between the beginning of the cropping period until harvest, but with a more substantial increase in the chemically treated plot (Figure 1). This phenomenon could have been the result of the activity of auxiliary organisms unaffected by the pesticides. The C. plutellae population increased to a higher level in the treated plot, indicating that this parasitoid could be resistant to the broad-spectrum pesticides applied in this trial. Individuals derived from the control plot would nevertheless have been able to lay eggs between two pesticide sprays, therefore maintaining a high level of parasitism.

Trial 2

DBM populations gradually increased in the treated plot to reach a high level at the end of the cropping period (around 1200 larvae and cocoons on 20 plants). Only the first treatment with Rocky^® seemed to temporarily reduce the populations, but this effect was not confirmed after the second treatment just prior to harvest. In the untreated plot, populations first increased and then sharply dropped at the end of the cropping period. Parasitism also increased in the two plots during the cropping period, but did not explain the difference in patterns between these plots. We noted that C. plutellae accounted for most of the parasitism at this elevation (250 m).

Trial 3

On this plot, DBM and associated parasitoid populations steadily increased over the time course of the trial, but the impact of the latter was not sufficient to hamper crop damage. D. mollipla was the dominant parasitoid in this case.

Trial 4

Pest populations also generally increased in this trial, but there was a temporary decrease 1.5 months after planting. The parasitoid population patterns were similar, and D. mollipla was again the dominant species.

Trial 5

The results of this test were very disappointing, i.e. DBM was detected under the insect net (it was possibly not properly sealed, or eggs might have been laid on leaves in contact with the net) and the cabbages were substantially damaged (small size, yellow leaves, etoliated appearance that was probably due to poor lighting, temperature and humidity conditions).

Table 2 summarizes the number of parasitoids observed and parasitism rates calculated for the main species per plot, along with the overall rate. The parasitism rate ranged from 8 to 45% and increased with elevation.

Table 2. Parasitism rates (%) of DBM observed in four cabbage field trials on Réunion Island

Trial (altitude)	Chemical control	No of parasitoids	Diadegma mollipla (%)	Cotesia plutellae (%)	Global rate (%)
1. (250 m)	Yes (hard)	211	0.01	9.99	10
	No	116	0.01	7.99	8
2. (250 m)	Yes (soft)	235	1	9	10
	No	123	2	7	9
3. (800 m)	No	788	19	3	22
4. (1200 m)	No	901	33	13	44

Discussion

A relatively high number of auxiliary organisms were detected on P. xylostella in our study, whereas many other insect groups are scarce in Réunion due to the remoteness of this island. The hymenopteran C. plutellae was more common at low elevations whereas D. mollipla populations dominated at higher sites, as also confirmed elsewhere (in litteris).

In all trials, a high proportion of cabbages was not marketable due to damage caused by DBM, regardless of the pest management conditions (intensive chemical control, supervised control or natural biological control). These results demonstrate the inefficacy of the tested registered active ingredients against DBM populations in Réunion Island, indicating that these pests are likely resistant to several families of chemicals as a result of the longstanding unmanaged massive application of these compounds. Natural biological control is also not efficient enough, a situation that could worsen with the introduction of new beneficial insects. It was noteworthy that the chemical treatments seemed to have no impact on the parasitoids, indicating that these species could have also built up resistance.

The increase in the parasitism rate with elevation is perhaps associated with the fact that the cabbage cropping area is greater in the highlands (more suitable climate) and hence there would be a greater reservoir of parasitoids.

Insect nets are currently expensive, so their use on cabbage crops to hamper attacks by DBM would not be cost-effective. Moreover, the use of mini-tunnels to hinder moth egg laying and limit climatic constraints is also still too expensive to implement.