Distribution of Panonychus citri (McGregor) and Euseius tularensis Congdon on central California orange trees with implications for binomial sampling

Agriculture, Ecosystems and Environment, 14 (1985) 119--129 Elsevier Science Publishers B.V., Amsterdam -- Printed in The Netherlands

119

DISTRIBUTION OF PANONYCHUS CITRI (McGREGOR) AND EUSEIUS TULARENSIS CONGDON ON CENTRAL CALIFORNIA ORANGE TREES WITH IMPLICATIONS FOR BINOMIAL SAMPLING

FRANK G. ZALOM, CHARLES E. KENNETT, NEIL V. O'CONNELL, DONALD FLAHERTY, JOSEPH G. MORSE and LLOYD T. WILSON

IPM Implementation Group, University o f California, Davis, CA 95616 (U.S.A.) (Accepted for publication 24 May 1985)

ABSTRACT Zalom, F.G., Kennett, C.E., O'Connell, N.V., Flaherty, D., Mome, J.G. and Wilson, L.T., 1985. Distribution of Panonychus cirri (McGregor) and Euseius tularensis Congdon on central Californiaorange trees with implicationsfor binomial sampling. Agric. Ecosystems Environ., 14: 119--129. Adults and active stages of the citrus red mite, Panonychus citri (McGregor) and active stages of the predatory mite Euseius tularensis Congdon were sampled in a central California orange grove for the period 1968--1976. In 1983, 10 groves were sampled for P. cirri adult females. Analysis of variance showed no difference (P > 0.05) between sampling locations within a tree. Clumping pattern per leaf for each species and age class sampled were expressed as the relationship between proportion of leaves infeeted with mites, mean number of mites per leaf and variance of mites per leaf. The dumping patterns of both P. cirri classes were fairly constant between years and between orchards and were more clumped than with E. tularensis. The relationship between proportion of leaves infested and mean number of mites per leaf was determined for both species and age clamei and incorporated into a binomial sampling program. Binomial sequential sampling schemes were developed for P. citri using thresholds proposed in the literature.

INTRODUCTION

California ranks second among states in sweet orange (Citrus sinensis) Osbeck production, with most of its harvested acreage being grown in the San Joaquin Valley (Anonymous, 1984). The most important mite pest of California citrus is the citrus red mite, Panonychus cirri (McGregor) (Jeppson et al., 1975). Populations increase in the late spring, late summer and early autumn in response to flushes of new leaves (McMurtry, 1969; .Jeppson et ah, 1975). They feed on both leaves and fruit, causing stippling and in severe infestations m a y cause leaf and fruit drop and twig die back. Various p r e d a c e o u s m i t e s and insects help t o regulate P. citri p o p u l a t i o n s , t h e m o s t i m p o r t a n t b e i n g p r e d a c e o u s mites o f t h e genus Euseius ( M c M u r t r y ,

0167-8809/85/$03.30

© 1985 Elsevier Science Publishers B.V.

120

1969). Recent taxonomic studies have shown that several Euseius species are present on citrus. The most c o m m o n species on central valley citrus and the species probably observed in this study is Euseius tularensis n.sp. (Congdon and McMurtry, 1985). Little has been published on the distribution or sampling of P. citri and E. tularensis on orange trees. The most complete study of P. citri sampling was conducted on southern California lemons, Citrus limon Burmann, by Jones and Parrella (1984a). This paper analyzes the within-tree distributions and clumping patterns of P. cirri adult females and active stages and E. tularensis active stages on San Joaquin Valley orange trees during an 8-year period. The clumping patterns of adult female P. cirri were also examined for 10 orchards in a separate study. Binomial models which incorporate clumping pattern as a function of population density (Wilson and R o o m , 1983) were generated and later used to develop sequential sampling decision rules. MATERIALS AND METHODS

Leaf sampling, to determine the abundance of P. citri and E. tularensis on mature orange trees, was conducted at Lindcove in Tulare County, California from spring 1968 to spring 1976. The sampling interval was approximately every 2 weeks. During the 1968--1969 through 1971-1972 seasons, eight leaves were selected from t w o groups of four trees each. One leaf was selected from the upper and lower portions of the tree at each compass direction. During the 1972--1973 through 1975--1976 seasons, 16 leaves were selected from each of t w o groups of four trees each. T w o leaves were selected from the upper and lower portion of the tree at each compass direction. Two different groups of four trees were sampled in 1973--1974 and 1974--1975. Leaves sampled in a given season were those resulting from the spring flush o f growth that year. Each leaf was examined individually for the numbers of P. cirri adult females, P. cirri active stages including females and E. tularensis active stages. Mite abundance for each location within the four trees was compared on each sampling date by one-way analysis o f variance. Proportion of leaves infested with mites versus mean number of mites per leaf was determined after combining the eight leaves per tree and four trees (32 leaves total) on each sampling date from the 1968--1969 through 1971--1972 seasons and the 16 leaves per tree and two of the trees (32 leaves total) on each sampling date from the 1972--1973 through 1975--1976 seasons. The clumping pattern per leaf for each species and/or age class was expressed as the relationship between the proportion o f leaves infested with mites (P(I)), the mean number of mites per leaf (~-), and the variance o f mites per leaf ($2). Taylor's coefficients (Taylor, 1961, 1971) describe the relationship between variance and mean.

121 S~ = a ' ~ b

(1)

Wilson et al. (1983b) related P(I) to mean by incorporating the " a " and " b " coefficients in the equation

P(I)=I-

(a'-~b-z)

e - ~ " l°ge

"( a ' ¥ b - z -

1) -1

(2)

Estimates of Taylor's coefficients were derived in this way by using an iterative regression procedure (Wilson et al., 1983a). Validation of the 1968--1976 P. cirri adult female data was conducted in 10 orchards representing all citrus growing areas of Tulare Co. Twenty leaves were sampled from the perimeter o f each tree on each week beginning 15 March and ending 21 July. Leaves sampled on and after 28 April were from the orchards located at Ivanhoe, Orosi and Terra Bella; five trees from the orchards located at Lindsay, Orange Cove, Porterville, S. Porterville and Richgrove and four trees from the orchards located at E. Lindsay and Lindcove. Each leaf was examined for the number o f P . citri adult females. Analysis o f the clumping pattern per leaf was described using the same m e t h o d as t h a t used for the 1968--1976 data. Standard error about each regression curve and, therefore, for any point on the curve, will vary depending upon proportion of leaves infested ( P ( I ) ) , proportion of leaves not infested ( q ( I ) ) , and sample size (n) for a given level of certainty (t) as follows: SE = t(p,n_ 1)

/P(I) . q(I)

n

(3)

Binomial sequential sampling plans were developed for P. cirri adult females at two densities and for P. cirri active stages using published thresholds. Equations used to generate the treat and no treat decision lines were first presented by Wilson et al. (1983b). nv = T~2 " (P - Ti) -2 " P " q

(4)

nl

= r ~ 2" (P - Ti) -2 " P ' q

(5)

nv

= sample size required to estimate with a given error rate (~) that the

proportion of infested leaves (p) is above the control action threshold (T) at any point in time (i). nl

= sample size required to estimate with a given error rate (a) that the

proportion of infested leaves is below the threshold. r

= the standard normal variate.

We have arbitrarily chosen a and /3 error rates of 0.10, for example, as these rates have been used elsewhere with mites as a target organism (Wilson et al., 1983a).

122 RESULTS AND DISCUSSION

Within-tree distribution No statistical difference ( P > 0.05) was detected between sampling locations on the tree or between samples from the various compass directions for P. citri adult females and active stages, or for E. tularensis active stages. However, on some of the sampling dates, slightly greater densities of each species were present on the north and east quadrants. The lack of significant difference between sampling location in a tree is important as it then becomes possible to select randomly leaves from around the perimeter of a tree without biasing the estimate of P. citri or E. tularensis density. Other mites have been shown to become randomly distributed among quadrants on almond and apple, including Tetranychus spp. and its predator Typhlodromus occidentalis (Nesbitt) (Wilson et al., 1984) and Panonychus ulmi (Koch) and its predator Zetzelia mall (Ewing) (Herbert and Butler, 1973).

Clumping pattern per leaf Taylor's coefficients describe the relationship between the variance and the mean (Taylor, 1961) of samples from a population and provide estimates of the aggregation pattern of the organism. When the "a" coefficient is equal to or significantly greater than 1, a "b" value significantly greater than 1 indicates a clumped distribution; "b" equal to or approaching 1 indicates a random distribution; and "b" less than 1 indicates a regular distribution. 1.00 o

.

-

0.5'0

ILl

~- 0 . 8 0 09

•

,

i

,.,.

::-;...

2~ >

o. o

a_ ~ "

0.40 0.30

U_o 0 . 2 0

,'

hus c i t r i A d u l t

~

1968-1976 (8 years)

~:'

0,10 0.00

I

o

I

6

I

,o

I

,i

MEAN NUMBER OF M I T E S / L E A F

Fig. 1. Proportion of leaves infested w i t h Panonychus citri adult females P. citri active stages and E. tularensis active stages as a f u n c t i o n o f d e n s i t y for t h e years 1968--1969 to 1975--1976 combined and for P. citri adult females for 10 Tulare County, California orchards during 1983. Curves were drawn using eqn. (2).

123 1.00 a

-

0.90-

t

I.LI ~- 0 . 8 0 -

_0,, 0 . 7 0 I--- Z OU~ a.w

0.60 0.50-

o~ o4o a.

030 IJ.

0

0.20

tive Stages

O. I O 0.00

-

I

0

2

4

6

8

I0

12

14

MEAN NUMBER OF M I T E S / L E A F 1.00

0.90

~ o~o 0.70 z

06o

O n ~Lu 0,50 0 > rr < 0.40 n ~ 0.30

i'

'

s tularensis Active Stages

0,20

1 9 6 8 - 1 9 7 6 (8 years)

0.10

t

0.00

I

0

1,00 I

I

2

I

...............

09°

I

i

:

~ o.~ol g~ o.7o 1

• ..,.~.~ ~-~.. :::,.-7'.::'-:...,

.

~_ o6ol

:Z..'. " '

''

:

o._, o3o -1~~"'~*"" o,o

o2o

15~

I

I

I

4 6 8 I0 12 MEAN NUMBER OF M I T E S / L E A F

.,._,,,

,9a3c10

14

~

c,.i A~.,t ,.,..,,

oo

~

o~,J~o,,,b,'ne,J)

0.10 1~ o.oo

o

~

~

~

~

,~

,~

,~

MEAN NUMBER OF M I T E S / L E A F

Figure 1 shows the relationship of proportion of leaves infested to the mean number of mites per leaf for P. cirri adult females, P. citri active stages and E. tularensis active stages for the years 1 9 6 8 - - 1 9 6 9 through 1 9 7 5 - - 1 9 7 6 . Figure 1 also shows the proportion of infested leaves as a

124

function of density for P. citri adult females from 10 orchards in 1983. P. citri adult females exhibited very little variation in distribution (x+ SD) between the eight years ("b ' ' = 1.252+ 0.148; R 2 = 0.954; n = 278) (Table I) or between the 10 sampling sites ("b ' ' = 1.346 + 0.068; R 2 = 0.940; n = 710) (Table II) in the study. Such stability becomes an important factor in developing a robust sampling strategy. When all stages of P. cirri were combined (Table III), the resulting distribution was slightly more clumped than for the adult females alone. This might be expected TABLE I Taylor's coefficients for Panonychus cirri adult females sampled during an 8-year period. Tw o plots were sampled in 1973--1974 and 1974--1975 and are designated (A) and

(B) Year

1968--1969 1969--1970 1970--1971 1971--1972 1972--1973 1973--1974 1973--1974 1974--1975 1974--1975 1975--1976 Combined

a

(A) (B) (A) (B)

2.173 1.628 1.502 1.437 2.604 2.255 0.967 1.737 1.345 1.466 1.785

b

1.423 1.291 1.325 1.060 1.396 1.358 0.961 1.240 1.292 1.178 1.298

R~

0.924 0.981 0.989 0.932 0.975 0.941 0.947 0.910 0.949 0.967 0.954

Number of observations with ~- > 0

P(I)

Maximum

47 30 13 23 9 36 21 50 35 14 278

0.97 0.97 1.00 0.38 0.28 0.78 0.19 0.84 0.75 0.91 1.00

TABLE II Taylor's coefficients for Panonychus cirri adult females sampled from 10 orchards in 1983 Location

East Lindsay Ivanhoe Lindcove Lindsay Orangecove Orosi Porterville Richgrove South Porterville Terra Bella Combined

a

2.072 2.100 1.901 2.218 1.925 1.858 2.076 2.294 1.916 2.462 2.087

b

1.308 1.413 1.423 1.334 1.225 1.262 1.386 1.416 1.366 1.325 1.378

R~

0.917 0.935 0.914 0.932 0.957 0.963 0.949 0.917 0.977 0.967 0.940

Number of observations with ~" > 0

P(1)

Maximum

59 101 73 64 93 86 45 41 69 79 710

1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00

125

as colonies form a b o u t adult females, increasing the total n u m b e r of mites in each restricted area. The relative distributions of P. citri adult females and active stages on orange trees in our study were similar, but slightly less clumped than P. cirri in a study b y Jones and Parrella (1984b) (1.298 and 1.378 vs. 1.40 and 1.456 vs. 1.52, respectively) in southern California lemons. This could TABLE III Taylor's coefficients for Panonychus cirri active stages sampled during an 8-year period. Two plots were sampled in 1973--1974 and 1974--1975 and are designated (A) and

(B) Year

1968--1969 1969--1970 1970--1971 1971--1972 1972--1973 1973--1974 1973--1974 1974--1975 1974--1975 1975--1976 Combined

a

(A) (B) (A) (B)

2.751 2.432 4.149 2.834 1.987 3.516 2.758 2.586 2.182 2.955 2.839

b

1.654 1.011 1.282 1.274 1.326 1.692 1.300 1.525 1.301 1.408 1.456

R~

0.776 0.983 0.980 0.876 0.753 0.870 0.795 0.911 0.950 0.960 0.940

Number of observations with ~" > 0

P( I)

Maximum

48 40 16 35 11 38 27 50 36 15 316

1.00 1.00 0.66 0.84 0.59 0.53 0.47 0.94 0.78 1.00 1.00

TABLE IV Taylor's coefficients for Euseius tularensis active stages sampled during an 8-year period. Two plots were sampled in 1973--1974 and 1974--1975 and are designated (A) and

(B) Year

a

b

R2

Number of observations with ~ > 0

Maximum P(I)

1968--1969 1969--1970 1970--1971 1971--1972 1972--1973 1973--1974 1973--1974 1974--1975 1974--1975 1975--1976 Combined

1.139 1.865 1.740 1.396 1.011 1.979 1.672 1.258 1.364 1.687 1.458

0.986 1.304 1.295 1.239 0.941 1.350 1.184 0.986 1.159 1.342 1.161

0.982 0.965 0.960 0.968 0.946 0.974 0.958 0.983 0.984 0.965 0.967

30 29 43 50 39 52 42 47 35 20 387

0.84 0.63 0.66 0.88 0.72 0.53 0.41 0.94 0.78 0.97 0.97

(A) (B) (A) (B)

126

arise from several factors, including basic differences in the growth habit of navels and lemons and differences in pruning practices in the orchards. Lemons have more growth flushes, which might lead to greater clumping on each flush. Further, more severe pruning of lemon trees might destabilize the P. citri population to some extent, resulting in greater clumping. The distribution of E. tularensis active stages (1.161 + 0.054; R 2 = 0.967; n = 387) was less clumped (Table IV) than that of P. cirri active stages (1.456 + 0.128; R 2 = 0.940; n = 316). This is expected of predatory mites (Wilson et al., 1984) which have higher searching and dispersal rates than their prey.

Sequential sampling When a pest species is small and abundant, as with mites, scouting for the pest with a certain degree of reliability is often time consuming. The relationship between proportion of sampling units infested and the mean number of the pest species per leaf used in a binomial (presence--absence) sampling plan can provide savings in such a case. This t y p e of sampling strategy has been used successfully to estimate P. ulmi populations on apple (Mowery et al., 1980) and Tetranychus spp. on c o t t o n (Wilson et al., 1981) and for P. citri on lemons by Jones and ParreUa (1984b). When the density of a pest is considerably greater than or considerably less than a control action threshold, additional scouting time can be saved by combining binomial sampling with sequential sampling (Wald, 1947). Several control action thresholds have been proposed for P. cirri on oranges. In central California, one threshold in use is t w o P. citri adult females per leaf in spring (Kennett, 1984). Below the threshold, no treat200 u~ LU

Panonychus c i r r i

adult females

p-

-~ 160

Threshold = 2 m i t e s / l e a f

I--

120

TREAT - ~

'"I I..U >

~ u_ 80 O

~ ~00LU40 I ~ ~ . ~

_ NO TREAT

'*~

Z OIP

0

''-

1

30

[

i

i

i

i

60 90 120 150 IS0 NUMBEROF LEAVESSAMPLED

210

Fig. 2. S e q u e n t i a l s a m p l i n g d e c i s i o n lines f o r Panonychus citri a d u l t females i n c o r p o r a t i n g a control action threshold o f t w o p e r leaf a n d a a n d # o f 0.10.

127

ment would be recommended if the P. citri : E. tularensis ratio is < 0.2. Above the threshold, the ratio would have to be much greater. The proportion of leaves infested equivalent to two P. citri adult females per leaf is 0.73 (from Fig. 1) (SE = 0.086; P = 0.05; n = 100). Sequential sampling rules (Fig. 2) using this threshold were calculated. It is also possible to determine a ratio o f the number of P. citri adult females per leaf from a binomial sample with the number o f E. tularensis from the same binomial sample (Fig. 1) and then to use the resulting ratio in the decision-making process. A second control action threshold for P. cirri in central California 200 P a n o n y c h u s cltri

w i-

adult f e m a l e s

~

~

TREA~

Threshold = 3 . 5 m i t e s / l e a f

,6o -i-

u') 120 w > w "LL 80 O tr

W

40

f

D

~AERT

I

i

i

30

60

90

120

NO I

T

i

Z

0

0

150

180

210

NUMBER OF LEAVES SAMPLED

Fig. 3. Sequential sampling decision lines for Panonychus citri adult females incorporating a control action threshold o f 3.5% and a and ~ 0.10. 200 w I-~ leo -1I-

Panonychus citri

active stages

Threshold = 2 mites/leaf

(,9 12o w TREAT

i.u J 0 t'r" l,U

m

80

40

A

z

0 0

I . 50

I,

I~

60

90

I 120

NOI TRE i 150

180

210

NUMBER OF LEAVES SAMPLED

Fig. 4. Sequential sampling decision lines for Panonychus citri active stages incorporating a control action threshold of two per leaf.

128 is t h r e e o r f o u r a d u l t f e m a l e s p e r l e a f in t h e spring ( K e n n e t t , 1984). In this case, t h e P. citri : E. tularensis ratio s h o u l d b e >t 1.0. T h e p r o p o r t i o n i n f e s t e d e q u i v a l e n t t o 3.5 P. citri a d u l t f e m a l e s p e r l e a f is 0 . 8 5 ( f r o m Fig. 1) (SE = 0 . 0 6 9 ; P = 0.05; n = 100). S e q u e n t i a l s a m p l i n g rules (Fig. 3) using this t h r e s h o l d w e r e c a l c u l a t e d as well. In Sardinia, I t a l y , Delrio et al. ( 1 9 7 9 ) p r o p o s e d a d a m a g e t h r e s h o l d o f t w o active P. citri p e r l e a f o n oranges a n d l e m o n s . T h e y suggested t h a t in f u t u r e this t h r e s h o l d s h o u l d b e m o d ified t o give a p e r c e n t a g e o f i n f e s t e d leaves. A s s u m i n g t h e d i s t r i b u t i o n o f P. citri active stages in California is similar t o t h a t in Sardinia, t h e prop o r t i o n o f leaves i n f e s t e d e q u i v a l e n t t o t w o active stages p e r l e a f w o u l d b e 0.62 ( f r o m Fig. 1) (SE = 0 . 0 9 4 ; P = 0.05; n = 100). Sequential sampling rules f o r this t h r e s h o l d are p r o v i d e d in Fig. 4. CONCLUSION T h e d i s t r i b u t i o n o f P. citri a d u l t f e m a l e s a n d active stages a p p e a r s t o r e m a i n fairly c o n s t a n t b e t w e e n y e a r s and b e t w e e n o r c h a r d s a n d is m o r e c l u m p e d t h a n active stages o f t h e p r e d a t o r y m i t e E. tularensis, w h i c h m a y also be p r e s e n t o n o r a n g e trees. T h e r a p i d i t y o f b i n o m i a l s a m p l i n g and t h e significant r e l a t i o n s h i p b e t w e e n p r o p o r t i o n o f leaves i n f e s t e d a n d m e a n n u m b e r o f m i t e s p e r l e a f f o r b o t h g r o u p s o f P. citri and f o r E. tularensis m a k e s b i n o m i a l s a m p l i n g a highly p r a c t i c a l m e t h o d .

REFERENCES Anonymous, 1984. California agriculture -- 1983. Calif. Dep. Food and Agric. Report, 17 pp. Congdon, B.D. and McMurtry, J.A., 1984. Biosystematics of the genus Euseius on California citrus and avocado (Acari : Pnytoseiidae). Int. J. Acarol., in press. Delrio, G., Ortu, S. and Prota, R., 1979. Aspects of integrated control in the citrus cultures of Sardinia. Studi Sassaresi, 27:205--222. Herbert, H.J. and Butler, K.G., 1973. Distribution of phytophagous and predaceous mites on apple trees in Nova Scotia. Can. Entomol., 105: 271--276. Jeppson, L.R., Keifer, H.H. and Baker, E.W., 1975. Mites Injurious to Economic Plants. University of California Press, Berkeley, CA, 614 pp. Jones, V.P. and Parrella, M.P., 1984a. Dispersion indices and sequential sampling plans for the citrus red mite (Acari : Tetranychidae). J. Econ. Entomol., 77: 15--79. Jones, V.P. and Parrella, M.P., 1984b. Intra-tree regression sampling plans for the citrus red mite (Acari : Tetranychidae) on lemons in southern California. J. Econ. Entomol., 77: 810--813. Kennett, C.E., 1984. Citrus red mite. In: J.E. Pehrson, D.L. Flaherty, N.V. O'Connell and P.A. Phillips (Editors), Div. Agric. Sci. Publ. 3303, pp. 90--94. McMurtry, J.A., 1969. Biological control of citrus red mite in California. Proc. First Int. Citrus Symp., 2: 855--862. Mowery, P.D., Hull, L.A. and Asquith, D., 1980. Two new sampling plans for European red mite surveys on apple utilizing the negative binomial distribution. Environ. Entomol., 9: 159--163.

]29 Taylor, L.R., 1961. Aggregation, variance and the mean. Nature (London), 189: 732-735. Taylor, L.R., 1971. Aggregation as a species characteristic. In: G.P. Patil, E.C. Pielou and W.E. Waters (Editors), Statistical Ecology. Vol. 1. Penn. State University Press, Philadelphia, PA, pp. 357--377. Wald, A., 1947. Sequential analysis. Wiley, New York, 212 pp. Wilson, L.T. and Room, P.M., 1983. Clumping patterns of fruit and arthropods in cotton, with implications for binomial sampling. Environ. Entomol., 12: 50---54. Wilson, L.T., Leigh, T.F. and Maggi, V., 1981. Presence--absence sampling of spider mites densities on cotton. Calif. Agric., 35: 10. Wilson, L.T., Gonzalez, D., Leigh, T.F., Maggi, V., Foristiere, C. and Goodell, P., 1983a. Within-plant distribution of spider mites (Acari : Tetranychidae) on cotton: a developing implementable monitoring program. Environ. Entomol., 12: 128--134. Wilson, L.T., Pickel, C., Mount, R.C. and Zalom, F.G., 1983b. Presence---absence sequential sampling for cabbage aphid and green peach aphid (Homoptera : Aphididae) on brussels sprouts. J. Econ. Entomol., 76: 476--479. Wilson, L.T., Hoy, M.A., Zalom, F.G. and Smilanick, J.M., 1984. Sampling mites in almonds: I. Within-tree distribution and clumping patterns of mites with comments on predatory--prey interactions. Hilgardia, 52(7): 1--24.