Abstract

A simple observational approach for studying the response of the diamondback moth (DBM), Plutella xylostella (L.), to environmental gradients including latitude, insecticide application and climate change, was developed in this study. Insecticide resistance and carboxylesterase frequency in DBM collected in 1979/1980 and 1986/1987 were selected as two ecological variables for analysing and defining the response. Variation in the correlation between larval carboxylesterase 9b and malathion resistance of this insect was also investigated during the period of 1987 to 1997. A clustering dendrogram based on frequencies of seven carboxylesterase isozymes in 16 Taiwan populations showed a decreasingly latitude-dependent distribution of three major clades orientated from north to south. A multidimensional scaling configuration based on the same data had a stress value of 0.18 indicating that these three clades are barely separated. Data on insecticide resistance of 18 populations to five insecticides in a study by Cheng (1981) were used for clustering and multidimensional scaling analysis. The cluster of insecticide resistance in DBM also showed a latitude-dependent distribution of three clades. Multidimensional scaling configuration of insecticide resistance had a stress value of 0.03 indicating that the whole operative taxonomic unit was unable to be separated except for the Tou-Cheng population. Nevertheless, these two clustering dendrograms were similar to each other in three aspects: 1. susceptible populations of Tou-Cheng and Jeo-Feng diamondback moth in northern Taiwan stood as an outgroup from the two different operative taxonomic units; 2. both clusters and both configurations showed that the Sheh-Tzu population in northern Taiwan was grouped together with three southern populations, indicating autopomorphy; 3. geographically related populations adjacent to one another were grouped together as three latitude-dependent clades. However, discordance of the morphometric gradient of the isozyme of DBM populations among adjacent regions was observed. Of seven carboxylesterase isozymes, frequency of EST 9b in DBM was found to be positively correlated with resistance to mevinphos, malathion, fenvalerate and permethrin in this insect. These findings suggest that: 1. the increasing frequency of EST 9b in these populations is roughly associated with decreasing latitude and the distribution of EST 9b in these populations is not random; 2. populations of higher insecticide-resistance have a higher frequency of EST 9b in the zymogram of DBM; 3. both frequencies of EST 9b and insecticide-resistance in DBM are temporarily sustainable and may have varied little during the period of 1981-1997. However, a series of studies on the correlation between EST 9b frequency in DBM and the susceptibility of DBM to a discriminating dose of malathion (0.12 mg per larva) during 1989-1997 revealed that the slopes of the linear regression lines increased in a tangent angle from 16.7^°, 20.4^° to 29.0^°. All three slopes intercepted approximately at zero frequency of EST 9b. This suggested that the association between the mechanism of malathion-resistance and EST 9b in 1989, 1991 and 1997 increased during this period of time. Frequency of EST 9b in 2001 DBM reached its highest titre of 78% suggesting that the association between EST 9b and the resistance mechanism may reach its maximum. Two factors may be involved in the rapid spread of EST 9b in Taiwan populations: 1. dispersal of EST 9b increased in susceptible populations due to migration of the gene coded for EST 9b; 2. a warm winter season due to El Niño in 1998/1999 enhanced the dispersal of EST 9b in all populations. Significance of variation in the correlation between frequency of EST 9b and malathion resistance of DBM is discussed.

Keywords

esterase 9b

Introduction

Synthetic pesticides are invaluable in suppressing damage to agricultural products. However, the side effects of these chemicals to organisms in the environment include the rapid development of resistance, especially to insecticides, in target insect pests. Investigation of the origin of a resistance gene acquired by insect pests and the way by which the resistance gene is dispersed are thus interesting topics for entomologists (Devonshire 1977, Georghiou & Taylor 1977, Campbell et al. 1997).

Studies on the resistance mechanism of diamondback moth to diazinon and methomyl were reported in Taiwan (Sun et al. 1978). Accordingly, a continuous gradient in increasing level of diazinon resistance in DBM populations was found from north to south. The level is low in the north, intermediate in the north-west and high in the west (Chi 1975). Nevertheless, levels of insecticide resistance in DBM populations which are adjacent to one another in southern, western and eastern counties of Taiwan have been occasionally found to be discordant (Cheng 1981). Population-dependent variations in insecticide resistance of DBM have also been reported in other countries (Miyata et al. 1986, Shelton et al. 1993, Yu 1993). Usually, a population of DBM is thought to be characterised by a particular geographical range or ecological range of tolerance to insecticides. These characteristics are more or less conservative and minor changes of insecticide-resistance in the moth occur in time (Tabashnik et al. 1992). However, how the combined temporary effects of latitude of habitat, weather conditions and insecticide applications affect the susceptibility of DBM to insecticide has not yet been explored in Taiwan.

A previous study (Maa & Liao 2000) revealed that all resistant cultures of Sheh-Tzu (ST) DBM which had slow-moving esterase isozymes (EST 8b or 9b) were associated with malathion resistance. On the other hand, susceptible cultures which had the fast-moving, paraoxon-tolerant, esterase isozyme (EST 4b) were positively related to a low level of malathion resistance. These esterases are likely coded by recessive genes and were suggested to be monitoring proteins, especially EST 9b, for malathion susceptibility in DBM.

Tabashnik et al. (1992) indicated that DBM is highly mobile. Thus, differentiation between individual populations in larger regional population units is a matter of time. Carboxylesterase isozymes which are related to malathion resistance in DBM should be used as morphological traits for quantitative gradation in resistance in this study. Variation of esterase isozymes, especially esterase 9b, of DBM populations found in different counties of Taiwan was thus examined during 1988, 1991 and 1997 in order to determine how this insect was affected by the complex effects of insecticide application, weather conditions and time.

Taiwan DBM populations were sampled along a 300 km north to south transect and assayed for malathion susceptibility in 1988/1989. The data were compared with those of insecticide resistant populations assayed by Cheng (1981) in order to determine the temporal variations in insecticide resistance and variation in EST 9b frequency in DBM.

Materials and methods

(a) Insects

Individual populations of DBM were collected from vegetable farms in different counties of Taiwan and were reared in the laboratory according to the method of Maa and Liao (2000). In this study, 11, 24, 18, 20 and 18 field populations were sampled respectively in the years 1987, 1989, 1991, 1997 and 2001. The IV instar 84 h old larvae were used for the susceptibility test, synergistic test and esterase zymogram study. The stocks were initiated by mating 30 pairs of DBM; the offspring of the first generation were used.

(b) Chemicals

All of the chemicals and reagents were of analytical grade or reagent grade. Diazoblue, lauryl sulfate, Fast blue RR, 1-naphthyl acetate (1-NA) and eserine were purchased from Sigma Chem. Co., MO, USA. All of the reagents for electrophoresis were purchased from Bio-Rad Lab., CA, USA. Paraoxon; O, O-diethyl-o-p-nitro-phenylphosphate; malathion; O, O-dimethyl-S-(1,2-dicarboethoxtethyl) phosphorodithioate; piperonyl butoxide (PB); diethyl maleate (DEM); and S, S, S-tributylphosphoro trithioate (TBPT) were purchased from Chemical Service, PA, USA.

(c) Zymogram study and frequency of esterase isozyme

Larvae of 16 populations from 1986/1987 were used. Samples for the zymogram study were prepared and run through PAGE according to Davis (1964). Esterase isozymes were stained with 1-NA, and 10–⁵ M paraoxon was used to characterise the isozyme in the gel according to Ogita and Kasai (1965). The migrating distance (Rf) and the frequency of seven carboxylesterase isozymes (Fq) were used as two measurements for morphometric analysis by clustering and multidimensional scaling. Isozymes in which staining was dim or presence of paraoxon was feint were not explored in this study.

(d) Bioassay and discriminating dose

Batches of 30 larvae were treated topically with malathion in two series of concentrations: 0.5, 1.6, 4.0, 8.3, 16.5 and 33.0 µg/larva for the susceptible larvae and 16.5, 33, 66, 132 and 176 µg/larva for the resistant larvae. The larvae were checked every 12 h until they emerged as adults. Results were calculated using Abbott’s formula. Dosage-mortality curves of LD₅₀ were calculated using the probit analysis method proposed by Finney (1971). Data of LD₅₀ were transformed into log values.

Correlation between the log LD₅₀ and the frequency of esterase isozyme in DBM was tested for linear regression. A discriminating dose of 120 µg malathion per larva was also chosen for testing the correlation between frequency of EST 9b and malathion resistance of DBM collected in 1986/1987, 1988/1989, 1990/1991 and 1996/1997. It tested how long the correlation between EST 9b and resistance of DBM can last. Data of log LD₅₀ and that of frequency of EST 9b obtained by our laboratory was tested for correlation with the log LC₅₀ of DBM resistance to five insecticides obtained by Cheng (1981). It determined if there is cross resistance between malathion and the other five insecticides. It also determined how far the correlation between EST 9b and resistance of DBM can be extended backwards. The purpose of both of the above-mentioned tests was to determine whether EST 9b frequency can be used as a constant parameter to monitor insecticide resistance in DBM.

(e) Effect of temperature

Frequency of EST 9b in DBM collected in 2000/2001 was compared with that of DBM collected in previous years in order to determine the influence of temperature on DBM when the global climate phenomenon of EI Niño affected Taiwan during 1998 to 1999. The correlation between survival rate of DBM and temperature for bioassay was determined.

(f) Data analysis

Frequencies of seven carboxylesterases in 18 populations were analysed by clustering dendrogram and multidimensional scaling (MDS) configuration using the PRIME 5 software developed by Clarke and Gorley (2001). Lethal concentrations of mevinphos, carbofuran, fenvalerate, cartap and permethrin administered to the larvae of the DBM populations from Cheng (1981) were used for clustering dendrogram, MDS configuration and linear assay with either a frequency of 9b or LD₅₀ of malathion. These two sets of clustering dendrograms and MDS configurations were used as two different ecological variables for determining if there is a common ground to determine the response of DBM to the influence of environmental states during 1980 to 1987.

Results and discussion

(g) Geographical features of collection sites

Taiwan is located to the south of mainland China. The physical map of Taiwan (Figure 1) shows mountainous regions (natural area in white), boundaries of agricultural areas (in shadow), urban areas (in dark), counties and collection sites. Collection sites were located in a 300 km section, mainly in west and north-west Taiwan. Distance between sites was approximately 15 to 35 km.

Figure 1. Geographical distribution of 25 DBM populations in Taiwan.

Name Index of Sample Locations:
JF:	Jeo-Fen	ML:	Miao-Li	PT:	Pu-Tzu	IL:	I-Lan	MH:	Min-Hsiung
ST:	Sheh-Tzu	TC:	Ta-Chia	LC:	Lu-Chu	FS:	Feng-Shan	MT:	Ma-Tou
SH:	Shan-Hsia	TT:	Tsao-Tun	LY:	Lin-Yuan	HH:	His-Hu	PC:	Pao-Chung
TY:	Ta-Yuan	YC:	Yuan-Ching	PN:	Pi-Nan	HY:	Hsin-Ying	PL:	Pu-Li
CP:	Chu-Pei	HL:	His-Lo	CA:	Chi-An	KS:	Kao-Shu	TCh:	Tou-Cheng

(h) Clustering and MDS analysis on carboxylesterase isozymes

Taiwan DBM are thought to be descended from a common ancestry. DBM populations that departed from each other geographically retained their own differentiated isozyme pattern. Minor variations in the composition of esterase isozymes of DBM in the same region is to be expected. However, the clustering analysis dendrogram of frequency of the isozyme (Figure 2) shows that Jeo-Feng (JF) is independently branched off as an outgroup at the similarity distance of 0.65 from the main operating taxonomic unit. This population has a higher frequency of EST 3b. Although JF and I-Lan (IL) are adjacent to one another geographically, DBM in these sites showed different isozyme patterns. Lu-Chu (LC) and IL, separated from the south clade and north clade respectively, are counted as secondary outgroups at the distance of 0.85. The rest of the DBM populations are clustered together as a monophyletic group with three clades separated off at a distance of 0.85 (Figure 2). The southern clade (clade A) includes three southern populations, Lin-Yuan (LY), His-Lo (HL) and Pu-Tzu (PT), and a northern population, Sheh-Tzu (ST). The central clade (clade B) includes four western populations: Yung-Jin (YJ), Miao-Li (ML), Ta-Chia (TC) and Chu-Pei (CP). The northern clade (clade C) includes two northern populations, Ta-Yuan (TY) and Shan-Hsia (SH); one west-central population, Tan-Tzu (TT); and two eastern populations, Chi-An (CA) and Pi-Nan (PN). CA and PN are geographically far away from the rest of the major unit.

Figure 2. Dendrogram for hierarchical clustering (using group-average) showing relationships among 16 diamondback moth populations based on the percentage of seven carboxylesterase isozymes. Similarity matrix was calculated using the Bray-Curtis similarity index.

Name Index of Sample Locations:
JF:	Jeo-Fen	ML:	Miao-Li	PT:	Pu-Tzu	IL:	I-Lan	MH:	Min-Hsiung
ST:	Sheh-Tzu	TC:	Ta-Chia	LC:	Lu-Chu	FS:	Feng-Shan	MT:	Ma-Tou
SH:	Shan-Hsia	TT:	Tsao-Tun	LY:	Lin-Yuan	HH:	His-Hu	PC:	Pao-Chung
TY:	Ta-Yuan	YC:	Yuan-Ching	PN:	Pi-Nan	HY:	Hsin-Ying	PL:	Pu-Li
CP:	Chu-Pei	HL:	His-Lo	CA:	Chi-An	KS:	Kao-Shu	TCh:	Tou-Cheng

The cluster shows that distribution of these taxa follows an increasing latitude gradient, accompanied by a decreasing frequency of EST 9b, or an increasing frequency of EST 3b in the zymogram of the DBM, in orientation from south to north.

It is interesting to note that ST, a northern population, was clustered with three southern populations. These four populations all have a high frequency of EST 9b (Table 1) and retain high resistance to insecticides (Table 2). This suggests that heavy insecticide application in the local cultivated land like ST (Cheng 1981) may result in quantitative or qualitative differentiation of esterase isozymes in this insect. Discordance of the morphometric gradient of the isozyme among adjacent DBM populations is possibly related to two environmental gradients: selective pressure of insecticide and latitude/altitude. In other words, zymogram patterns of esterase isozymes of any two adjacent DBM populations may be alike or unlike if one of the populations was exposed to an extreme selection pressure such as insecticide application.

Table 1. Distribution of carboxylesterase isozymes of the larval homogenate of 16 diamondback moth populations in Taiwan (1987) (28 larvae were used for the zymogram analysis)

Population Rf of isozymes	JF	ST	SH	TY	CP	ML	TC	TT	YC	HL	PT	LC	LY	PN	CA	IL
	Frequency of isozyme in percentage
0.13 (EST 12b)	100	100	100	100	100	100	100	100	100	100	100	100	100	100	100	100
0.39 (EST 10b)	0	4	0	7	36	14	29	0	0	0	14	0	0	7	7	29
0.41 (EST 9b)	14	97	50	36	86	93	86	50	100	79	83	86	64	36	50	86
0.43 (EST 8b)	0	7	7	7	7	0	0	7	0	21	7	57	7	4	7	0
0.47 (EST 7b)	0	0	7	7	7	0	7	0	0	14	15	50	7	0	0	0
0.66 (EST 4b)	0	22	0	0	0	0	0	0	0	14	22	0	14	4	0	29
0.72 (EST 3b)	43	0	0	0	0	0	0	0	0	0	0	7	7	0	0	64

Table 2. Correlation between the frequency of Est 9b of DBM larvae and LC₅₀ or LD₅₀ of the larvae from different DBM populations in Taiwan

DBM population	Est 9b freqency	Log values of LC50 of five insecticides			Log LD50
		fenvalerate	carbofuran	mevinphos	permethrin	cartap	malathion
ST	0.92	3.88	3.14	2.52	3.67	2.73	2.29
SH	0.70	2.72	2.45	1.95	2.52	2.73	2.24
TY	0.25	3.07	2.33	2.04	2.75	2.66	1.71
CP	0.92	3.52	2.70	2.41	2.96	2.61	1.94
ML	0.65	3.28	2.37	2.45	2.88	2.72	2.20
TC	0.54	3.66	2.64	2.48	3.10	2.58	2.31
TT	0.86	3.39	2.40	2.43	3.06	2.83	2.31
HL	0.71	3.72	2.67	2.61	3.60	2.75	2.32
PT	0.96	3.99	2.53	2.52	3.41	2.72	2.34
LC	1.00	3.76	2.92	2.84	3.58	2.86	2.42
LY	0.83	4.00	2.87	2.68	3.67	2.84	2.60
PC	0.70	3.64	2.21	2.40	3.13	2.68	2.47
HH	0.83	3.58	2.41	2.51	3.70	2.73	2.40
HY	0.79	3.98	2.55	2.29	3.35	2.81	2.34
KS	0.71	3.50	2.54	2.35	3.09	2.62	2.31
PL	0.85	3.45	2.45	2.42	2.95	2.57	2.27
MH	0.88	3.66	2.32	2.28	3.13	2.73	2.23
IL	0.79	-	-	-	-	-	2.23
TCh	-	2.13	1.62	1.81	2.02	2.40	-

LC data were transcribed from Cheng (1981). Linear regression assay:

1. between insecticides and Est 9b:		2. between malathion and other insecticides:
malath./Est 9b:	slope, 0.577; r, 0.5225 (<P.05);
mevinp./Est 9b:	slope, 0.651; r, 0.5546 (<P.05);	malath./mevinp.:	slope, 0.629; r, 0.5904 (<P.05)
fenval./Est 9b:	slope, 0.966; r, 0.5184 (<P.05);	malath./fenval.:	slope, 0.928; r, 0.5487 (<P.05)
permet./Est 9b:	slope, 0.981; r, 0.5017 (<P.05);	malath./permet.:	slope, 1.070; r, 0.6029 (<P.05)
carbof./Est 9b:	slope, 0.566; r, 0.4211;	malath./carbof.:	slope, 0.293; r, 0.2404
cartap/Est 9b:	slope, 0.183; r, 0.3743;	malath./cartap:	slope, 0.199; r, 0.4492

MDS configuration of esterase isozymes in DBM (Figure 3) shows that most of the populations lie on the left part of the scale. Populations with higher frequencies of EST 9b are on the left side; those with lower frequencies of EST 9b are in the middle zone; JF, with the lowest frequency of EST 9b, is on the right side of the scale. JF and TCh, with higher frequencies of EST 3b, are in the upper right of the scale. The wide dispersal of taxa in the left and middle parts of the model may suggest that JF is different from all other populations. In fact, analysis of variance (ANOVA) in Tukey Studentized Range test revealed that all other DBM populations in Taiwan were somehow significantly differentiated from the JF population in their frequency of EST 9b.

Figure 3. Multidimensional scale model of 16 diamondback moth populations based on the percentage of seven carboxylesterase isozymes. Similarity matrix was calculated using the Bray-Curtis similarity index.

Name Index of Sample Locations:
JF:	Jeo-Fen	ML:	Miao-Li	PT:	Pu-Tzu	IL:	I-Lan	MH:	Min-Hsiung
ST:	Sheh-Tzu	TC:	Ta-Chia	LC:	Lu-Chu	FS:	Feng-Shan	MT:	Ma-Tou
SH:	Shan-Hsia	TT:	Tsao-Tun	LY:	Lin-Yuan	HH:	His-Hu	PC:	Pao-Chung
TY:	Ta-Yuan	YC:	Yuan-Ching	PN:	Pi-Nan	HY:	Hsin-Ying	PL:	Pu-Li
CP:	Chu-Pei	HL:	His-Lo	CA:	Chi-An	KS:	Kao-Shu	TCh:	Tou-Cheng

However, the difference in EST 9b frequency between JF and other populations of DBM disappeared gradually as time elapsed (Table 3). Meanwhile, DBM populations with high EST 9b frequencies remained, regardless of whether the populations were resistant or not. In fact, the average EST 9b frequency of Taiwan populations was maintained at a level below 75% throughout the years of 1987 to 1997 and the frequency climbed to a high peak of 78% in the year 2000, although the frequency of EST 9b in most populations fluctuated from year to year (Table 3). It seems that dispersal of EST 9b, possibly associated with a resistant gene, is not avoidable in time in any population found in Taiwan.

Table 3. Frequency of EST 9b and malathion resistance of the diamondback moth collected in Taiwan during 1987–2001

DBM population	1987	1989		1991		1997		2001
	EST 9b frequency	EST 9b frequency	Survival rate	EST 9b frequency	Survival rate	EST 9b frequency	Survival rate	EST 9b frequency
IL	0.86	0.79	0.36	0.77	0.78	-	-	0.81
JF	0.14	0.25	0.30	0.22	0.10	0.54	0.28	0.69
ST	0.97	0.92	0.40	0.77	0.55	1.00	0.58	0.80
SH	0.50	0.70	0.50	0.65	0.36	0.54	0.36	0.57
TY	0.36	0.25	0.20	-	0.20	0.39	0.57	0.77
CP	0.86	0.92	-	0.59	0.68	0.81	0.89	0.73
ML	0.93	0.65	0.39	0.76	0.63	0.46	0.51	0.68
TC	0.86	0.54	0.54	0.87	0.41	0.77	-	0.66
TT	0.50	0.86	0.47	-	0.40	-	-	0.87
PL	-	0.85	0.24	-	-	-	-	-
HH	-	0.83	0.55	1.00	0.41	-	-	-
YC	1.00	-	-	-	-	0.69	0.48	0.93
HL	0.79	0.71	0.55	0.88	0.56	0.83	0.64	0.92
MT	-	-	-	0.53	-	-	-	-
PT	0.83	0.96	0.49	0.77	0.56	0.85	0.71	0.87
HY	-	0.79	0.57	-	-	-	-	0.68
KS	-	0.71	0.38	0.59	-	0.58	0.61	0.73
LC	0.86	1.00	0.64	0.53	0.43	0.94	0.75	0.93
FS	-	-	-	-	-	0.77	0.90	0.68
LY	0.64	0.83	0.62	0.76	0.40	0.50	0.31	0.93
M+S.D.	0.72+0.26	0.74+0.22	0.45+0.13	0.69+0.19	0.46+0.18	0.69+0.19	0.58+0.20	0.78+0.11

Twenty-five (25) larvae were used for each treatment; each larva was treated topically with 120 μg malathion.

(i) Clustering and MDS analysis of resistance of DBM to five insecticides

The map of insecticide resistance in Taiwan DBM populations was first completed by Cheng (1981). The sampled populations were scattered in a section of 300 km along the western side of Taiwan. TCh population was the only one collected in north-eastern Taiwan. Cheng (1981) found that Kao-Shu (KS) population in southern Taiwan is slightly resistant to insecticides, and TCh and TY populations in northern Taiwan are susceptible to insecticides. The remaining Taiwan populations are all resistant to insecticides including carbofuran, cartap, fenvalerate, permethrin and mevinphos. The data sets of insecticide resistance were transformed into log values and analysed by clustering dendrogram and MDS model.

The clustering dendrogram (Figure 4) shows that TCh is an outgroup, at a distance of 65% similarity, from the taxonomic unit of insecticide resistance. The operational taxonomic unit also has three clades separated at the distance of 0.75 in the similarity scale of the cluster. LC, in this case, was grouped with the southern clade. It is interesting to note that coincidences were found between the cluster of insecticide resistance and that of esterase isozyme: 1. susceptible DBM populations, TCh and JF in northern Taiwan stood respectively as outgroups from two different taxonomic units; 2. the ST population in northern Taiwan was grouped with the southern clade in both clusters; 3. the southern, central and northern clades in both clusters followed a decreasing gradient of temperature from south to north. However, geographical discordance among adjacent populations of DBM was also found in the clustering dendrogram of insecticide resistance. For example, KS DBM should retain a comparatively high level of resistance to insecticides compared with the northern populations since this population is located in southern Taiwan. The KS DBM population is, however, clustered with two other susceptible populations as a member of the north clade. On the other hand, the ST DBM population in northern Taiwan was clustered as a member of the southern clade. All members of the southern clade had a higher frequency of EST 9b and a higher level of insecticide resistance, while all members of the northern clade had a low frequency of EST 9b and a lower level of insecticide resistance. The distribution of a DBM population in the cluster system seems to be dependent on three factors: 1. selection pressure of insecticide application; 2. latitude of the niche for the insect; 3. speed of dispersal of the resistant gene in the population.

Figure 4. Dendrogram for hierarchical clustering (using group-average) showing relationships among 18 diamondback moth populations based on the insecticide-resistance of DBM to five insecticides (fenvalerate, carbofuran, mevinphos, permethrin and cartap). Similarity matrix was calculated using the Bray-Curtis similarity index. Data on insecticide resistance of DBM (Cheng 1981) were used.

Name Index of Sample Locations:
JF:	Jeo-Fen	ML:	Miao-Li	PT:	Pu-Tzu	IL:	I-Lan	MH:	Min-Hsiung
ST:	Sheh-Tzu	TC:	Ta-Chia	LC:	Lu-Chu	FS:	Feng-Shan	MT:	Ma-Tou
SH:	Shan-Hsia	TT:	Tsao-Tun	LY:	Lin-Yuan	HH:	His-Hu	PC:	Pao-Chung
TY:	Ta-Yuan	YC:	Yuan-Ching	PN:	Pi-Nan	HY:	Hsin-Ying	PL:	Pu-Li
CP:	Chu-Pei	HL:	His-Lo	CA:	Chi-An	KS:	Kao-Shu	TCh:	Tou-Cheng

Shelton et al. (1993) suggested that insecticide resistance in DBM in North America originated from southern states. Similarly, we believe that the resistant gene in Taiwan population likely originated and subsequently spread from a southern population since the northern population, JF, has a lower frequency of this isozyme and retains its susceptibility to insecticide screening. In other words, the susceptible population found in northern Taiwan should be a better choice for determining the original development of the resistant gene. We created a culture of DBM that had EST 9a, instead of EST 9b, in the esterase zymogram, by sequentially interbreeding siblings of a single female-male pairing (Maa & Liao 2000). It is expected that a DBM population without EST 9b is possibly originally one that had EST 9b because of insecticide resistance. Unfortunately, the author was unable to find any wild DBM population with a zero frequency of EST 9b by as early as 1982. As Cheng (1981) indicated, all Taiwan populations are resistant to insecticides, since development of resistance of this insect species to various pesticides were noticed early in the sixties (Tao 1973).

It is noted that the MDS configuration of insecticide resistance (Figure 5) shows a different trend from that of esterase isozyme. It shows that the sampling populations are aggregated in patches along a full tangent line (450) in the lower-left part of this two-dimensional model. In addition, the resistance MDS has a stress value of 0.03 indicating that the whole taxonomic unit was unable to be separated except for the TCh population.

Figure 5. Multidimensional scale model of 18 diamondback moth populations based on the insecticide resistance of DBM to five insecticides (fenvalerate, carbofuran, mevinphos, permethrin and cartap). Similarity matrix was calculated using the Bray-Curtis similarity. Data on insecticide resistance of DBM (Cheng 1981) were used.

Name Index of Sample Locations:
JF:	Jeo-Fen	ML:	Miao-Li	PT:	Pu-Tzu	IL:	I-Lan	MH:	Min-Hsiung
ST:	Sheh-Tzu	TC:	Ta-Chia	LC:	Lu-Chu	FS:	Feng-Shan	MT:	Ma-Tou
SH:	Shan-Hsia	TT:	Tsao-Tun	LY:	Lin-Yuan	HH:	His-Hu	PC:	Pao-Chung
TY:	Ta-Yuan	YC:	Yuan-Ching	PN:	Pi-Nan	HY:	Hsin-Ying	PL:	Pu-Li
CP:	Chu-Pei	HL:	His-Lo	CA:	Chi-An	KS:	Kao-Shu	TCh:	Tou-Cheng

Meanwhile, the esterase MDS (Figure 3) has a stress value of 0.18, suggesting that these three clades are separated and populations of the same clade (Figure 2) can be grouped together in this model. This difference between the two MDS reflect that esterase isozymes, or a detoxification mechanism associated with EST 9b of DBM is possibly playing a partial role in detoxification of insecticides. Meanwhile, the majority of the sampled DBM populations assayed by Cheng (1981) possibly had a common resistance mechanism.

It is accepted that resistance of DBM to different categories of insecticide varies depending on the detoxification mechanism of the insect. A broad spectrum of insecticide resistance observed in field populations is due to multiple resistance mechanisms, including detoxification of insecticides by microsomal oxidase, enhanced carboxylesterase, glutathion-S-transferase and target site insensitivity such as insensitive acetyl cholinesterase (Cheng 1986, Maa et al. 1997, Miyata et al. 1986, Sun et al. 1986, Yu & Nguyen 1992). Although Motoyama et al. (1992) suggested that permethrin shared no cross resistance with malathion or mevinphos in DBM, our synergistic test showed that fenvalerate was strongly synergised by Pb and malathion was depressed by TBPT (data not shown). Miyata et al. (1986) suggested that different populations have different resistance mechanisms. Cheng (1986) suggested that partial cross resistance was found between organophosphates and synthetic pyrethroids. Motoyama et al. (1992) found that the fenvalerate resistance of the revertant larvae could be restored by just one selection event with fenvalerate or even with malathion. It was concluded that there was an unknown factor(s) necessary to maintain the insecticide resistance in DBM, which cannot be explained by the conventional preadaptation theory.

The low stress value of the insecticide resistance MDS may hint that most Taiwan DBM had a common mechanism responsible for resisting different categories of insecticides. We found that correlation between EST 9b frequency of 1988/1989, or 1996/1997 populations and mevinphos/permethrin/fenvalerate resistance of 1979/1980 populations were all significant (refer to Tables 2 and 3) throughout this time interval. These results suggested that the titre or mechanism of insecticide resistance associated with DBM varied little during this period of time. The resistance level of any population may decrease when the pressure of selection was released. Nevertheless, a resistant gene in a heterogeneous form will be well-retained in the wild population.

Halpern and Morton (1987) found that malathion-selected Drosophila melanogaster produced fewer offspring and had defective larval development. Similarly, we found that an EST 9b homogeneous resistant pair, in a sibling interbreeding bioassay, produced fewer offspring when the parent male DBM had been either selected for highest frequency of EST 9b or had been selected under highest lethal dose of malathion (Maa & Liao 2000). It is easy to maintain an EST 9a homogeneous susceptible culture, but difficult to maintain an EST 9b homogeneous resistant culture in the laboratory when the culture is selected.

(j) EST 9b for monitoring protein and 120 µg for diagnostic dose

Roush and Daly (1990) suggested that when two resistance genes are involved in resistance of an insect, a dose that kills 75% of the backcross is useful, since the most likely genotypes to be killed are the 75% that are most susceptible. In this case, a 99% kill of a susceptible population is necessary for a discrimination test. ST population selected with malathion for three to four generations produced defective offspring which either died before emergence or had only a few adults emerge. A routine rule for genetic analysis on the resistance gene was therefore not used in this study. Reports on the resistance mechanism of DBM revealed that it is a matter of multiple resistance. Tolerance of acetylcholinesterase found in malathion-resistant populations of Taiwan tend to be codominant to dominant in their inheritance (Maa et al. 1997). Our previous study revealed that malathion resistance of ST culture tended to be incompletely dominant in inheritance (Maa & Liao 2000). We also found cross resistance between malathion, an organophosphorous compound, and permethrin, a synthetic pyrethroid (Table 1). Yu and Nguyen (1992) indicated that DBM in Florida had high oxidative activity against various kinds of insecticides. Yu (1993) indicated that resistance to permethrin was inherited in an incompletely dominant, autosomal manner. We therefore chose 120 µg malathion per larva as the diagnostic dose for topical treatment. This dose is high enough to kill 95% of an intrabreeding susceptible culture of ST population (Maa & Liao 2000). In addition, nearly 75% of Taiwan populations had a LD₅₀ lower than 120 µg/larva. This dose was also used for monitoring the population of 1990/1991 and 1996/1997.

Results of the study on the correlation between EST 9b frequency in the population and the malathion resistance of 1988, 1991 and 1997 populations (Figure 6) show that the slopes of the linear regression lines shift in tangent angle from 16.7^°, 20.4^° and 29.0^°. All three slopes of 1988, 1991, 1997 intercepted approximately at zero frequency of EST 9b. We suggest that the association between malathion resistance and EST 9b frequency somehow increased gradually during 1987~1997. We expect that frequency of EST 9b would be a good indicating protein for monitoring malathion resistance in DBM in the field. The mechanism of malathion resistance associated with EST 9b will be studied with molecular techniques. However, the defect of a resistant homogeneous population is a disadvantage for survival in the long run.

Figure 6. Linear correlation between frequency of EST 9b and survival rate of diamondback moth treated with 120 μg malathion per larva. 1989: y = 0.3001x + 0.2192; 1991: y = 0.3718x + 0.2062; 1997: y = 0.5535x + 0.2031.

(k) Temperature effect

A rising temperature due to El Niño during 1998/1999 made the mean of EST 9b frequency increase from 69% in the 1997 population to 78% in the 2001 population. Correlation between EST 9b frequency and temperature is positively related. Correlation between survival rate of DBM to malathion is also positively related. This indicates that temperature is an important factor affecting the survival rate of the insect against malathion.

(l) Frequency, resistance, environmental gradients and time factor

Results of morphometric analysis on divergence of carboxyl esterase, frequency of the esterases and titre of insecticide resistance of DBM populations reveal that it is the sum of all selecting pressures including insecticide application, altitude, latitude, climate and time that makes the resistant population what it is. We assume that discontinuous geographical distribution of esterase isozyme patterns in adjacent populations of DBM was due to adaptation capability, survival rate and the ability of DBM to adapt to selection pressure from the environmental gradients. The dominant esterase isozymes, ESTs 3, 4, 8 and 9 are widely distributed in field DBM (Maa et al. 2000). These isozymes, unlike what was found in aphids (Devonshire & Moores 1982), are limited in quantity and are hardly able to function as major metabolic enzymes for degrading malathion or as major binding proteins for rendering malathion or its derivatives nontoxic. Doichuanngam and Thornhill (1989) suggested that there was a positive correlation between activity of carboxylesterase and malathion resistance. We argue that it is possible to simplify the often complex arguments of resistance mechanisms and detoxification systems using desirable general results and offer as proof, but without a powerful explanation, the actual patterns in nature. However, these isozymes are good monitoring proteins for malathion resistance for the following reasons: first, these isozymes are easily detected by 1-NA staining with a combination of paraoxon inhibition and second, ESTs 3b/9b are coded by incompletely dominant genes. It is a good quality for a monitoring protein since the associated resistance gene linked with EST 9b might be protected and thus be heritable even under extreme environmental stress. It was once expected that southern populations would obtain insecticide resistance faster than the northern ones. Although the high selective pressure of insecticide application against DBM would drive the adaptation of this insect to be highly resistant to insecticides, the south DBM, the heavily selected DBM, would adapt rapidly to pressure caused by insecticide application since there is higher temperature and less precipitation all year round in the niche of DBM. Nevertheless, this study revealed that the spread of EST 9b, or the resistance gene, in Taiwan populations is only a matter of time.

Acknowledgements

This research was supported by grants from the Institute of Zoology, Academia Sinica and the National Science Council, Taiwan, ROC.

References