Oecologia (1993) 95:22-29 Oecologia �9 Springer-Verlag 1993

Host species preference and larval performance in the wood-boring beetle Phoracantha semipunctata F. Lawrence M. Hanks, Timothy D. Paine, Jocelyn G. Millar

Department of Entomology, University of California, Riverside, CA 92521, USA

Received: 23 October 1992/Accepted: 18 March 1993

Abstract. Adults of the wood-boring beetle Phoracantha semipunctata F. showed variability in their attractiveness to five varieties of Eucalyptus when presented with an array of logs in a natural setting. Logs of two host varieties (E. camaldulensis Dehnhardt and the hybrid E. trabutii) attracted two to three times more adult beetles than did logs of other host species (E. cladocalyx F., E. 9randis Hill ex Maiden and E. tereticornis Small). In the field, high oviposition rates by P. semipunctata adults resulted in severe competition among larvae. Larval survivorship was low in field logs of E. trabutii and high in E. cladocalyx logs, although these hosts were the most and least attrac- tive to the adult beetles, respectively. However, when logs were hand infested at low larval densities, survivorship of P. semipunctata larvae was highest in logs of both E. camaldulensis and E. trabutii. These findings suggest that adult beetles in the field were most attracted to those logs of Eucalyptus species that represented the highest quality hosts for their progeny under conditions of reduced larval competition.

Key words: Wood-boring insects - Host species preference - Host plant quality Eucalyptus - Phoracantha semi- punctata F.

The relationship between oviposition preference for cer- tain plant species and performance of offspring is key to our understanding of the evolution of host-plant associ- ations in herbivorous insects (Futuyma and Peterson 1985, Thompson 1988). Adult females confronted with a variety of potential hosts should select the host plant species that yield the greatest fitness in their progeny (Wiklund 1974; Chew 1975, 1977; Feeny 1975, Gilbert and Singer 1975). However, although a positive relationship between adult oviposition preference and larval per- formance has been documented in populations of some

Correspondence to." L.M. Hanks

herbivorous insects (e.g., Singer 1983; Myers et al. 1981; Williams 1983; Leather 1985, Denno et al. 1990), this may be absent in others (e.g., Wiklund 1975; Chew 1977; Rausher 1979; Courtney 1981, 1982; Roininen and Tahvanainen 1989). Poor correspondence between ovi- position preference and larval performance may indicate that natural selective factors other than host plant quality are influencing host plant choice by adults. For example, oviposition preference for one plant species over another may arise due to differences between these species in: 1) predation rates on larvae (e.g., Wiklund 1975, Smiley 1978, Price et al. 1980, Denno et al. 1990), 2) abundance within habitats (e.g., Wiklund 1974, 1981; Chew 1977, Williams 1983), and 3) food sources that are utilized by the adult insects (Courtney 1982).

Oviposition choice is critical for wood-boring insects because their larvae cannot compensate for poor host quality by moving to new hosts, as can some other kinds of insects (e.g., Chew 1977). Hosts that are suitable for wood- boring insects can be patchily distributed, requiring that adults have good dispersal abilities and effective host location mechanisms (Haack and Slansky 1987). If wood- boring larvae require fresh host material, adults are under further selective pressure to locate hosts rapidly while nutrient content remains satisfactory for the larvae (Haack and Slansky 1987).

The Eucalyptus Longhorned Borer (Phoracantha semi- punctata F.) is a Eucalyptus specialist that attacks stressed and downed host trees. Because the populations of this introduced pest have increased rapidly in California, large numbers of adult beetles are attracted to hosts. High rates of oviposition by P. sernipunctata result in intense competi- tion among larvae and reduced survivorship (Avidov and Harpaz 1969, Powell 1982, Mendel 1985, Gonzalez- Tirado 1987). In this paper, we examine the effect of larval competition on the relationship between attraction of adults to host species and subsequent larval performance by: 1) presenting a field population ofP. semipunctata with logs of five Eucalyptus species, capturing visiting adult beetles while allowing natural oviposition at high den- sities, and 2) hand infesting logs of the same host species with P. semipunctata larvae at a lower density.

23

Biology of P. semipunctata and its host plants

Phoracantha semipunctata (Coleoptera: Cerambycidae) is a severe pest of eucalyptus in many of the countries to which it has been introduced (e.g., Bytinski-Salz and Neumark 1952; Chararas 1969a; Drinkwater 1975; Ivory 1977; Cavalcaselle 1980; Cadahia 1986; Gonzfilez-Tirado 1986), although in its native Australia it is of only minor economic importance (Duffy 1963; Pook and Forrester 1984). The beetle was first reported in southern California in 1984 (Scriven et al. 1986), but it is uncertain how it was introduced or from where. The beetle has since spread throughout southern California and north to the San Francisco Bay Area.

Adult P. semipunctata average 1.4 cm in length, females being only slightly larger than males. The sex ratio of adult beetles emerging from logs usually approximates 1:1 (Hanks et al. 1993). The nocturnal adults are strong fliers (Drinkwater 1975), and may live in the field for one month feeding on the nectar and pollen of Eucalyptus flowers (Chararas 1969a; Scriven et al. 1986; Hanks et al. 1990). Adults of both sexes are strongly attracted to stressed Eucalyptus trees and logs on which they mate and oviposit (Chararas 1969b; Drinkwater 1975; Ivory 1977; Scriven et al. 1986; Gonz~ilez-Tirado 1986; Mendel 1987; LMH pets. obs.).

Cerambycids locate their hosts primarily by olfaction (Linsley 1963). The significance of olfaction (versus vision) for host location by P. semipunctata is indicated by the attraction of adults to individual stressed trees within stands of conspecifics and by adults flying to individual fresh logs when logs of the same species, but of various ages are stacked together (Hanks et al. 1991; LMH unpublished data). At close range, P. semipunctata adults apparently do not discriminate among logs of different Eucalyptus species by visual means or by bark texture (Chararas 1969b; Powell 1978).

The ovipositor of a female P. semipunctata can extend to a length of 2 cm and is specialized for laying eggs under loose bark and in bark cracks (Drinkwater 1975). Eggs are laid in batches of up to 40 eggs. Neonate larvae penetrate the bark and feed along the cambium and phloem until mature. These tissues have the highest nutrient content of the woody plant parts (Haack and Slansky 1987). Single larvae can survive in branches as small as 5 cm in diameter (LMH unpublished data).

Heavy infestations of P. semipunctata result in destruc- tion of the cambium layer, and the rapid death of stressed trees (Chararas 1969b; Drinkwater 1975: Scriven et al. 1986, pers. obs.). Extensive oviposition by adults com- monly results in reduced survivorship of larvae (Avidov and Harpaz 1969; Powell 1982; Mendel 1985; Gonzfilez- Tirado 1987). Low survivorship in P. semipunctata larvae is due to both cannibalism (interference competition) and to the earliest larvae to colonize consuming the most nutritious tissues, leaving low quality resources for later larvae (exploitative competition) (Ivory 1977; Powell 1978; Mendel et al. 1984; Mendel 1985; Haddan and Fraval 1988; LMH pers. obs.). Both of these forms of competition have been observed among larvae of other cerambycid species (Rose 1957; Grimble and Knight 1970: Ikeda 1979; Shibata 1987).

Larvae of P. semipunctata require approximately 2.5 mo to complete development. Mature larvae bore into the sapwood where they construct a pupal chamber, plugging the passage to the log surface with wood chips (the 'pupal plug'). During warm weather larvae rapidly pupate, but in colder weather they pass through a qui- escent prepupal stage before resuming development in the spring. Adults chew their way out of the log through the pupal plug. In southern California, adult emergence is not tightly synchronized, and beetles are present continuously from late spring until late fall. There are no effective natural enemies of P. semipunctata in California.

Phoracantha semipunctata attacks only Eucalyptus and a few species in closely related genera (Duffy 1963). In Australia, the beetle occurs wherever Eucalyptus grows (Fox and Curry 1980), but it is relatively uncommon and its attacks are restricted to dead or sickly trees and downed limbs or logs (Took 1935; Chararas 1969b). In other parts of the world where Eucalyptus species have been introduced, the beetle also attacks apparently healthy trees. This loss of resistance to borer attack is primarily due to poor adaptation to growing conditions (Chararas 1969b). Specifically, trees that are able to maintain bark turgidity under drought conditions can better resist beetle attack (Chararas 1969b; Hanks et al. 1991). Bark moisture will prevent larvae from penetrating to the cambium by drowning them in situ (Hanks et al. 1991). Eucalyptus species show variation in their susceptibilities to attack by P. semipunctata, but reduced vigor, especially through water deficits, render even resistant species vulnerable to attack (Chararas 1969b; Drinkwater 1975; Powell 1978; Cavalcaselle 1980; L6yttyniemi 1983; Mendel et al. 1984; Gonzfilez-Tirado 1986; Haddan and Fraval 1988). Felling trees destroys the defensive system, leaving logs highly receptive to colonization by larvae.

In this study, we examined the influence of larval competition on the relationship between attraction of adult P. semipunctata to host species and the subsequent performance of the larvae. We utilized logs of four Eucalyptus species that are common landscape trees in southern California: E. carnaldulensis Dehnhardt, E. cladocalyx F., E. grandis Hill ex Maiden and E. tereti- cornis Small. We also used the variety E. trabutii, a hybrid of E. camalduIensis and E. botryoides Small that occurs naturally in Australia (Pryor and Johnson t971) and that is extensively planted in the Mediterranean region (Pryor 1976). (For convenience, we will refer to E. trabutii as a species). E. camaldulensis and E. tereticornis are the most widely distributed Eucalyptus species on the Australian continent (Pryor 1976; Chippendale and Wolf 1981), and with E. grandis and E. cladocalyx, are among the most common eucalyptus planted outside of Australia (Zacharin 1978). As freshly cut logs, all five of these host species are attacked by P. semipunctata (Drinkwater 1975: Ivory 1977).

Materials and methods

Study sites and host trees

The influence of host species on attraction of adult P. semipunctata and on performance of the larvae was studied using logs cut from

24

Table 1. Dimensions of Eucalyptus logs (field and laboratory logs combined; N = 27 logs per species). Means (_+ SEM) within columns with different letters are significantly different (P < 0.05, R E G W F means separation test)

Species Circumference Length Surface Area Bark Thickness (cm) (cm) (cm 2) (cm)

E. camaldulensis 42.6_+1.5" 54.9--1.6 ab 2320_+ 89 a 0.56_+0.014 c E. cladocalyx 39.2_+1.9" 58.3--1.8 a 2280_+130 a 0.64_+0.033 "b E. 9randis 43.9_+1.5 a 54.1--1.6 ab 2320_+ 41" 0.45• d E. tereticornis 39.8_+1.2 ~ 51.3--0.77 b 2036~_ 6P 0.65_+0.031" E. trabutii 41.7_+1.2 ~ 55.8_+1.6 "b 2290+ 59 ~ 0.56_+0.019 bc

trees from a research plantation of mixed Eucalyptus species at the Moreno Valley field station of the University of California in Riverside County, California. The plantation was flood-irrigated biweekly, and the natural incidence of attack by P. semipunctata was very low among standing trees. All trees were 8 years old. The field site for examining attractiveness of logs to adult beetles was a 6 ha stand composed primarily of E. cladocalyx at University Agricultural Operations Research Center, Riverside, California.

Host species effect on adult beetle attraction

The influence of host species on the attractiveness of logs to adult beetles was examined by arranging logs of the five Eucalyptus species in the field and attaching sticky cards to each log in order to capture visiting beetles. Three trees of each of the five Eucalyptus species were selected such that differences in trunk circumference were minimized. Trees were felled on 12 July 1990 and were cut into logs ~55 cm in length, yielding 9 logs per tree. Logs of the five Eucalyptus species did not differ significantly in circumference of surface area (Circum- ference: F4.130 = 1.75, P > 0.05; surface area : F4,130 = 2.16, P > 0.05; Table 1), although E. tereticornis logs were significantly shorter than E. cladocalyx logs by ~ 12% (E4,130 =2.95, P=0.023; Table 1). The percentage difference between the thickest bark (E. tereticornis and E. cIadocalyx) and the thinnest bark (E. 9randis) was ~ 30% (F4,130 =10.7, P<0.0001; Table 1).

Three logs from each tree were randomly assigned to each of 3 sets (5 species x 3 trees x 3 logs =45 logs per set). Two sets of logs (referred to as 'field logs') were moved to the field site where they were arrayed on the forest floor. Field logs within a set were separated by 1 m and were randomly assigned positions within a 5 by 9 grid. The two sets of field logs were separated by 100 m.

Paper sticky cards (7 cm x 18 cm) were constructed from green cardboard (Boise Cascade, Sheldon Springs, Vermont). Insect glue (Formula 4916, Seabright Enterprises, Emoryville, CA.) was applied hot to the cards. Sticky cards were stapled to the upper sides of the logs late in the afternoon on alternate days beginning on 12 July 1990 and continuing until beetles were no longer captured. By leaving sticky cards off the logs on alternate nights, we allowed adult P. semipunctata to visit logs for mating and oviposition.

Host effect on performance of P. semipunctata at natural densities

Exposing the two sets of field logs to ovipositing females in the field emulated the natural attack rate and allowed examination of per- formance under conditions of extreme intraspecific competition. It was not possible to count eggs on field logs due to the ability of the females to insert eggs into very tight bark crevices. Field logs were brought into a temperature-controlled facility ( ~ 3 0 + 5 ~ under natural daylight; ~ 4 0 % RH) on 15 September 1990 and were bagged in nylon mesh material to collect emerged adults, after which larval density and survivorship was determined (see Performance of P. semipunctata larvae section, below). We recorded the size and sex of adult beetles emerging from these field logs.

Host effect on performance of P. semipunctata at low densities

To investigate the effect of host species on P. semipunctata per- formance under low larval densities, the third set of 45 logs was brought into the laboratory. This set of logs (referred to as ' laborat- ory' logs) was infested with P. semipunctata larvae by hand one week after felling. Four shallow slits, 3 cm in length and ~0.2 mm in depth, were cut in the bark with a razor blade perpendicular to the long axis of the log. Slits were cut into one side of the log and were separated by ~ 8 cm. Thirty newly-eclosed first instars, reared from eggs from a laboratory colony, were placed into these slits with a damp paintbrush (see Hanks et al. 1993). These logs were not moved for 24 h to allow larvae to bore into the wood. Logs were then bagged with nylon mesh and were moved into the rearing facility. We measured size and sex ratio of adult P. semipunctata emerging from these logs. After adult emergence was complete, larval densities and performance were determined (see PerforMance of P. semipunctata larvae section, below).

Performance of P. semipunctata larvae

Once beetles stopped emerging from both field and laboratory logs, the remaining outer bark was removed to examine the effect of host species on the following variables: 1) Survivorship of neonate larvae penetrating the bark of laboratory logs (number of neonate galler- ies/ 30 neonates). The presence of neonates was indicated by narrow ( ~ 1 mm wide) feeding galleries etched into the sapwood. Although later feeding by larvae could obscure these galleries, some length of the neonate galleries remained. 2) Survival of larvae in both field and laboratory logs (number of pupal plugs/number of neonate galler- ies). 3) Survival of pupae (number of opened emergence plugs/total number of pupal plugs). 4) Total survivorship (number of opened pupal plugs/number of neonate galleries). 5) Proportion of the cambial area consumed by the beetle larvae, measured by wrapping over the log a 20x 10 cm sheet of transparent plastic having a grid of 1 cm squares, and recording the proportion of 20 randomly selected squares that did not contain a larval feeding gallery. The grid was placed over the log blindly, with two counts averaged for each log.

Statistics

Proportions were arcsine transformed and counts were square root transformed before analysis to meet the assumptions of the ANOVA (Sokal and Rohlf 1981). Differences between treatment means were tested as unplanned comparisons, and tests were "protected" (that is, tests were applied only when overall ANOVA F test was significant; Day and Quinn 1989). To control the maximum experimentwise error rate under any complete or partial null hypothesis, the REGWQ means separation test was used when sample sizes were equal, while the Tukey-Kramer method was used when samples sizes were not equal (SAS 1988).

Results

Host species effect on adult beetle attraction

Adult P. semipunctata were captured on sticky cards for a period of 30 days. Logs of E. trabutii and E. camaldulensis were the most attractive to beetles (77 and 62 beetles captured, respectively) while E. cladocalyx was the least attractive (22 beetles captured; Fig. 1) (means significantly different, species term F4,8o=7.08, P<0.0001; set and species x set terms not significant; Fig. 2).

The sex ratio of adult beetles captured on logs did not differ significantly among the host species (overall F4,59 =0.88, P>0.05), with means ( • of 0.46_+0.09, 0.62 _+0.12, 0.61 _+0.076, 0.48_+0.13 and 0.47_+0.076 (females/ total) for E. camaldulensis, E. cladocalyx, E. 9randis, E. tereticornis and E. trabutii, respectively.

o

5 Z I r

80 -

70 -

6 0 *

5 0 '

4 0 '

3 0 '

2 0 '

10'

0

Eutr

-~~ . . . . Euca

.Is // ~ - - - - Eugr . - - " . . . . . ,-~.,,-~. ~ - . . . . . . . . . . Eute

J I" ~ . . . . . . . . . . . EUC[

5 10 15 2 0 2 5 3 0

Days After Felling

Fig. 1. Cumulative numbers of adult P. semipunctata captured on sticky cards stapled to logs of five Eucalyptus species (Euca=E. camaldulensis, Eucl=E. cladocalyx, Eugr=E. 9randis, Eute = E. tereticornis and Eutr = E. trabutii)

O . J

qD

h, ,-1

Q.

O

5 Z

r

=Z

Euca Eucl Eugr Eute Eutr

Host Species

Fig. 2. Mean number of adult beetles captured per log for five Eucalyptus species (species abbreviations as in Fig. 1). Means (_+ SEM) marked with different letters are significantly different (REGWQ means separation test, P <0.05)

25

Table 2. Mean (_+ SEM) number of P. semipunctata neonate gal- leries per tog for five Eucalyptus species. Means (+ SEM) within columns with different letters are significantly different (P < 0.05, REGWF means separation test)

Mean no. neonate galleries

Host Species Field logs Laboratory togs

E. camaldulensis 59.2_+ 8.05" 17.3_+2.56 a E. cladocalyx 45.4-+ 5.1 a 13.8+_2.19 a E, grandis 68.7 _+ 14.1" 20.2 -t-_ 4.7" E. tereticornis 54.8-+ 7.8 a 16.6 + 3.02" E. trabutii 68.9_+ 6.9 a 22.8 _+ 2.3 a

Host effect on performance of P. semipunctata at natural densities

Although the number of neonate galleries in field logs was 51% higher in E. trabutii than in E. cladocalyx (the most and least attractive species to adults, respectively), there were no statistically significant differences in numbers of neonate galleries between species (species term F4.80 = 1.43, P >0.05; Table 2). The number of neonate galleries per log was influenced by log set (F1, 8o = 13.9, P =0.0003), but the species x set term was not significant (F4, 8o = 1.48, P > 0.05).

Survivorship of P. semipunctata larvae in the two sets of field logs was significantly higher in E. camaldulensis than in E. trabutii which in turn showed a mean sur- vivorship that was significantly higher than E. tereticornis (species term F4, 8o = 8.42, P < 0.0001, set and species x set terms not significant; Fig. 3A). Pupal survivorship in the field logs showed a pattern among host species that was nearly the inverse of larval survivorship, with pupal sur- vivorship higher in E. cladocalyx than in E. camaldulensis, E. tereticornis or E. trabutii, and higher in E. 9randis than E. trabutii (species t e rm F4,65 =8.23, P<0.0001, set and species x set terms not significant; Fig. 3B). Total sur- vivorship (first instar to adult) for field logs was higher in E. camaldulensis and E. cladocalyx than in E. tereticornis (species term F4, 8o = 4.78, P = 0.0017, set and species x set terms not significant; Fig. 3C).

Female P. semipunctata emerging from the field logs of all five host species were larger than males (females: 1.48 +0.013 cm; males: 1.38 +0.018; means significantly differ- ent, ANOVA F1,178=22.52, P<0.0001). Because of this sex effect on body size, we included the sex ratio of beetles emerging from logs as a covariate in the ANOVA model to examine the host species effect on body size. Elytra length of adult P. semipunctata emerging from the field logs was significantly higher for E. camaldulensis and E. trabutii than for E. grandis (F4,43 =4.28, P=0.0053; Fig. 3D). The sex ratio effect on body size was also significant Fa,43 = 6.11, P = 0.018; sex ratio x species interaction not signi- ficant). The lower mean body size of adults emerging from E. 9randis ( ~ 8% smaller than beetles from E. camaldulen- sis and E. trabutii) may be due to the greatly skewed sex ratio of beetles emerging from logs of this species. The proportion of females emerging from logs of E. 9randis was significantly lower than from logs of the 4 other host species (ANOVA for all three log sets together, species t e rm F4,67 = 10.4, P<0.0001, set and species • set terms

26

o >

(0 >

0 ch

P> ._>

U)

o I-

A 0.80-

0.60"

0 . 4 0 -

0 . 2 0 -

0,00

C 0 . 5 0

0 . 4 0

0 . 3 0

0 . 2 0

0.10

0,00

Euca Eur Eugr Eute Eutr

e

Euca Eur Eugr Eute

Host Species

Eutr

B 1.00

. v 0 , 8 0 Q.

0.60

> '~, 0 . 4 0

0 , 2 0

O. O.O0

J: ==

LU

Euca Eucl Eugr Eute Eutr

D

1.70 1 1.60

a 1.50 ]

1 . 4 0 ]

1.30]

1.20 " Euca Euc! Eugr Eute Eutr

Host Species

Fig. 3A-D. Mean survivorship (A-larval, B-pupal, C-Total) of P. semipunctata in field logs of five Eucalyptus species (abbreviations as in Fig. 1) when logs were naturally infested by beetle oviposition in the field. Means ( i SEM) marked with different letters are significantl5 different (REGWQ means separation test, P<0.05). D. Mean elytra length of adult P. semipunctata emerging from logs of five Eucalyptus species. Means (_+ SEM) marked with different letters are significantly different (Tukey-Kramer means separation test, P < 0.05)

A (~

0

u) 4)

m

E U.

0

n -

X 4)

( / )

1 . 0 0

0 . 8 0

0 . 6 0

0 . 4 0 '

0 . 2 0

0 . 0 0 Euca Eucl Eugr Eute Eutr

Host Spec ies

Fig. 4. Mean sex ratio (number of females/total number) of adult P. semipunctata emerging from logs of five Eucalyptus species (abbreviations as in Fig. 1). Means (--SEM) marked with different letters under the bar are significantly different (REGWQ means separation test, P <0.05)

not significant; Fig. 4). Powell (1982) reported a similar male-biased sex ratio in adult P. semipunctata emerging from logs of E. 9randis.

There were significant differences among host species in the proportion of cambium consumed by P. semi- punctata larvae under field conditions (species term F4, so =2.85, P=0.029, set term F1,so=5.9, P=0.017, species x set terms not significant). E. trabutii and E. cladocalyx

logs showed significantly greater levels of proportional consumption (means 0.93_+0.026 and 0.93+_0.022, re- spectively) than E. grandis (0.78-4-0.062), while E. camal- dulensis and E. tereticornis showed intermediate values (0.90 4- 0.035 and 0.87 • 0.035, respectively). These differ-

ences in consumption did not correlate with larval performance (compare with Fig. 3A), suggesting that per- formance was not determined by consumption rates alone.

Host effect on performance of P. semipunctata at low densities

There were no significant differences among tree species in the number of P. semipunctata larvae colonizing the bark of laboratory logs (F4,40 = 1.55, P>0.05; Table 2). These initial larval densities for laboratory logs were about one third the densities of the field logs (Table 2). Larval survivorship (first instar to pupation) was significantly higher in E. trabutii than in E. tereticornis and E. 9randis (F4, 4o = 3.12, P = 0.025; Fig. 5A), with E. camaldulensis and E. cIadocalyx being intermediate. These patterns of larval survivorship were not explained by the initial densities of the larvae which were similar for E. trabutii and E. 9randis (Table 2). In the laboratory, there were no significant differences among host species in pupal survivorship (F4,37 = 1.97, P>0.05; Fig. 5B). The patterns of total sur- vivorship (first instar to adult) matched those of larval survivorship, but with both E. trabutii and E. camaldulen- sis showing significantly higher survivorship than E. 9randis (F4,4o=4.67, P=0.0035; Fig. 5C) and E. clado- calyx and E. tereticornis being intermediate. Body sizes (as indicated by elytra lengths) of adult beetles emerging from the laboratory logs did not differ significantly among host species (F4, 29 = 1.60, P > 0.05; Fig. 5D) and body size was not significantly effected by the sex ratio of beetles emerg- ing from logs (F1 ,29=0 .09 , P > 0 . 0 5 ) .

The proportion of the cambial area of laboratory logs that was consumed by P. semipunctata larvae did not differ significantly (F4,4o=0.92, P>0.05) with means of 0.7 +0.08, 0.71 _+0.074, 0.57_+0.07, 0.61 +0.08 and 0.74_+0.08 for E. camaldulensis, E. cladocaIyx, E. 9randis, E. tereticor- his and E. trabutii, respectively.

Q. o

o

U:,

A 0.80"

0.601

0.401

0.201

0.00

C 0 Q. 0.50

e~ 0.40

0.30

0.20

0 . ~ 0

o 0.00 p.

a

Euca Eucl Eugr Eute Eutr

a

EuCa Eucl Eugr Eute Eutr

Hos t Species

B ,~. CL 1.00"

i 0.80"

"~ 0.60 -

'~. 0,40 -

o~ 0.20 "

0.00 O.

D 1,70"

i 1,60"

1.50"

- - I 1 , 4 0 -

r

1.30- LIJ

1.20

a

a a

Euca Eucl Eugr Eute Eutr

Euca Eucl Eugr Eute Eulr

Hos t Species

27

Fig. 5A-D. Mean survivorship (A-larval, B-pupal, C-Total) of P. semipunctata in laboratory logs of five Eucalyptus species (abbreviations as in Fig. 1) when logs were infested at low density by hand in the laboratory. Means (+ SEM) marked with different letters are significantly different (REGWQ means separation test, P<0.05). D Mean elytra length of adult P. sernipunctata emerging from logs of five Eucalyptus species. Means (_+ SEM) marked with different letters are significantly different (Tukey- Kramer means separation test, P < 0.05)

Discussion

Eucalyptus species differed in their attractiveness to adult P. semipunctata even though logs were separated by only 1 m. Because the adult beetles locate their hosts primarily by olfaction, differences between host species in relative attractiveness likely reflect variation in production of volatile compounds. Preliminary studies of volatiles re- leased from freshly cut logs of E. camaldulensis, E. grandis and E. tereticornis have revealed a complex of mono- terpenes including ~-pinene, [3-pinene, cineole, cymene, limonene and myrcene as well as a variety of unidentified sesquiterpenes (LMH, JGM and TDP, unpublished data). Differences in the ratios of these compounds emitted by logs, and other unidentified monoterpenes that are present in one Eucalyptus species but absent in another, suggest that adult beetles may be discriminating among host logs by means of volatile cues. Powell (1978) also attributed variation in P. semipunctata oviposition on 12 species of Eucalyptus to differences in the olfactory response of adults.

Because P. semipunctata has only recently colonized California, the host preferences of adults are unlikely to have evolved here. The preference of adults for certain host species most likely reflects adaptation to Eucalyptus in Australia. However, because P. semipunctata is relatively uncommon in Australia, its host species preferences on that continent have not been studied. Outside of Australia, P. semipunctata is known to attack logs and downed trees of all five of our host species (Drinkwater 1975; Ivory 1977; Powell 1982). Where trees are standing and apparently healthy, E. camaldulensis is known to be relatively suscept- ible to borer attack when compared to the highly resistant E. cIadocalyx (Bytinski-Salz and Neumark 1952; Brown 1968; Chararas 1969b; Chararas et al. 1969; Ivory 1977; Mendel 1985; Scriven et al. 1985; Montoya-Oliver et al. 1983). (Little is known about the susceptibility of E. trabutii due to its limited distribution). In our study, the

high attractiveness of E. camaldulensis logs compared to the low attractiveness of E. cladocalyx might reflect adaptation of the borer in response to tree resistance encountered elsewhere in the world.

Our experiment to compare the attractiveness of host species to P. semipunctata adults was conducted in a stand of E. cladocalyx. It is likely that most of the local beetle population were reared on downed branches of this host, yet E. cladocalyx remained the least attractive host to these adult beetles. The lack of a preference in the adults for their natal host species argues against larval conditio- ning. This finding is consistent with other work that suggests that larval conditioning is not common in cerambycids (see discussion in Linsley 1963; Walsh and Linit 1985).

Adult P. semipunctata were most attracted to logs of E. camalduIensis and E. trabutii, the species that were the highest quality hosts for larvae when larval densities were low. Adult P. semipunctata may select hosts for their larvae using criteria that are appropriate only under the condi- tions of their native country (Australia) where host plant vigor and natural enemies apparently maintain popula- tions at much lower densities (Tooke 1935; Chararas 1969b). By this reasoning, adult host choice is effective only when competition among the larvae is low. However, in the habitat of the California landscape, where P. sernipunctata and its host plants are exotic and there are no effective natural enemies, intense larval competition changes the ordering of quality for the five host species. For example, adults were least attracted to E. cladocalyx in the field, but survivorship was relatively high on this host under larval densities that are typical in the field. In contrast, the host quality of E. trabutii was very high at low larval densities, but low when competition among the larvae was severe.

Thompson (1988) concluded that a lack of correspond- ence between oviposition preference and larval per- formance may be due to insufficient time for adaptation. In

28

Cal i fornia , P. semipunctata has had little t ime to a d a p t to new hab i t a t character is t ics . Moreover , genet ic var iab i l i ty in the p o p u l a t i o n is l ikely to be very low due to the founde r effect, h i nde r ing any e v o l u t i o n a r y response to selective pressures exerted by its hos t p l an t c o m m u n i t y . It therefore seems un l ike ly tha t adu l t hos t preference in P. semi- punctata will easily be ad jus ted to op t imize la rva l per- f o rmance in the face of in tense compe t i t i on .

Acknowledgements. We are grateful to Richard Redak and two anonymous reviewers for their comments on an earlier draft of this manuscript. We thank Kristin Kund and Ta Promlee for their technical assistance. We also thank the Rancho Santa Fe Associ- ation, the Elvenia Slosson Fund for Ornamental Horticulture, the California Department of Forestry and Fire Protection, and the California Department of Food and Agriculture for their financial support.

References

Avidov Z, Harpaz 1 (1969) Plant pests of Israel. Israel University Press, Jerusalem

Brown FG (1968) Pests and diseases of forest plantation trees. Clarendon Press, Oxford

Bytinski-Salz H, Neumark S (1952) The eucalyptus borer (Phoracan- tha semipunctata) in Israel. In: Transactions IX International Congress of Entomology, Amsterdam, 1951, vol 1, pp 696 699

Cadahia D (1986) Importance des insects ravageurs de l'eucalyptus en r6gion m6diterran6enne. Bull European and Mediterranean Plant Prot Org 16:265-283

Cavalcaselle B (1980) Osservazioni sulla diffusione di Phoracantha semipunctata F in Sardegna, Sicilia e hell'Italia CentroMeridion- ale. In: Atti, XII Congresso Nazionale Italiano di Entomologia, Roma (1980) Tip ografia Compositori, Bologna, pp 455 464

Chararas C (1969a) Etude biologique de Phoracantha semipunctata F (Col6opt6re Cerambycidae xylophage) specifique des Eucalyptus en Tunisie et recherches sur la vitalit6 et l'adaptation de ces essences. CR Acad Agric France 55:47 57

Chararas C (1969b) Biologie et +cologie de Phoracantha semi- punctata F (Col6opt6re Cerambycidae xylophage) ravageur des Eucalyptus en Tunisie, et methodes de protection des peuple- ments. Ann Inst Nat Rech For Tunisie 2 :1-37

Chararas C, Schoenenberger A, Poupon H (1969) Variations de la vitalit6 et de la pression osmotique de divers Eucalyptus, en fonction des conditions ~cologiques et r61e de Phoracantha semipunctata Fabr, Col6optere Cerambycidae xylophage. CR Acad Sci Paris 268:2697 2700

Chippendale GM, Wolf L (1981) The natural distribution of Eucalyptus in Australia. Austr Natl Parks Wildl Serv Spec Publ 6 :1 192

Chew FS (1975) Coevolution of pierid butterflies and their crucife- rous foodplants. I. The relative quality of available resources. Oecologia 20:117-127

Chew FS (1977) Coevolution of pierid butterflies and their cruci- ferous foodplants. II. The distribution of eggs on potential foodplants. Evolution 31 : 568-579

Courtney SP (1981) Coevolution of pierid butterflies and their cruciferous foodplants. III. Anthocharis cardamines (L.). Survival, development and oviposition on different host plants. Oecologia 51 : 91-96

Courtney SP (1982) Coevotution of pierid butterflies and their cruciferous foodplants. IV. Crucifer apparency and Anthocaris cardamines (L.) oviposition. Oecologia 52:258 265

Day RW, Quinn GP (1989) Comparisons of treatments after an analysis of variance in ecology, Ecol Mono 59 : 433 463

Denno RF, Larsson S, Olmstead KL (1990) Role of enemy-free space and plant quality in host-plant selection by willow beetles. Ecology 71 : 124 137

Drinkwater TW (1975) The present pest status of eucalyptus borers Phoracantha spp. in South Africa. In: Proc. 1st Congr Entomol Soc S Afr (1975). Entomological Society of Southern Africa, Pretoria, pp 119-129

Duffy EAJ (1963) A monograph of the immature stages of Australian timber beetles (Cerambycidae). British Museum (Natural History), London

Feeny P (1975) Biochemical coevolution between plants and their insect herbivores. In: Gilbert LE, Raven PH (eds) Coevolution of animals and plants. University of Texas Press, Austin, Texas, pp 3-19

Fox JED, Curry SJ (1980) Notes on the tuart tree (Eucalyptus 9omphocephala) in the Perth area. Western Aust Nat 14:174 186

Futuyma D J, Peterson SC (1985) Genetic variation in the use of resources by insects. Ann Rev Entomol 30:217 238

Gilbert LE, Singer MC (1975) Butterfly ecology. Ann Rev Ecol Syst 6 : 365-397

Gonz~ilez Tirado L (1986) Phoracantha semipunctata F.: Dafios ocasionados en la provincia de Huelva durante 1983 y 1984. Valoracion econ6mica. Boletin de Sanidad Vegetal, Plagas 12:147-162

Gonzfi.lez Tirado L (1987) Tabla de vida para Phoracantha semi- punctata Fab. (Col. Cerambycidae). Perforador de los eucaliptos, en el sudoeste espafiol. Boletin de Sanidad Vegetal, Plagas 13: 283-301

Grimble DG, Knight FB (1970) Life table and mortality factors for Saperda inornata (Coleoptera: Cerambycidae). Ann Entomol Soc Am 63:1309-1319

Haack RA, Slansky F Jr (1987) Nutritional ecology of wood- feeding Coleoptera, Lepidoptera, and Hymenoptera. In: Slansky F Jr, Rodriguez JG (eds) Nutritional ecology of insects, mites, spiders, and related invertebrates. Wiley, New York, pp 449 485

Haddan M, Fraval A (1988) Estimation a posteriori, ~ partir des caract6ristiques des galeries de la mortalit6 de Phoracantha semipunctata (F.) (Col. Cerambycidae), ravageur xylophage des eucalyptus au Maroc. Actes Inst Agron V6t 8 : 27 34

Hanks LM, Millar JG, Paine TD (1990) Biology and ecology of the eucalyptus longhorned borer (Phoracantha semipunctata F.) in southern California. In: Adams D, Rios J (eds) Proceedings of the 39th California forest pest council, 14 15 Nov. 1990, Sacramento, Calif. Calif Dept Forest Fire Prot, Sacramento California, pp 12 16

Hanks LM, Paine TD, Millar JG (1991) Mechanisms of resistance in Eucalyptus against larvae of the eucalyptus longhorned borer (Coleoptera: Cerambycidae). Environ Entomol 20:1583-1588

Hanks LM, McElfresh JS, Millar JG, Paine TD (1993) Phoracantha semipunctata (Coleoptera: Cerambycidae), a Serious Pest of Eucalyptus in California: Biology and Laboratory-Rearing Procedures. Ann Entomol Soc Am 86:96 102

Ikeda K (1979) Consumption and food utilization by individual larvae and the population of a wood borer Phymatodes maaki Kraatz (Coleoptera: Cerambycidae). Oecologia 40:287 298

Ivory MH (1977) Preliminary investigations of the pests of exotic forest trees in Zambia. Commonwealth For Rev 56:47 56

Leather SR (1985) Oviposition preferences in relation to larval growth rates and survival in the pine beauty moth, Panolis flammea. Ecol Entomol 10:213 217

Linsley EG (1963) The Cerambycidae of North America. University of California Publications in Entomology vol. 18

L6yttyniemi K (1983) Flight pattern and voltinism of Phoracantha beetles (Coleoptera, Cerambycidae) in a semihumid tropical climate in Zambia. Ann Entomol Fennici 49:49 53

Mendel Z (1985) Seasonal development of the eucalypt borer, Phoracantha semipunctata, in Israel. Phytoparasitica 13:85 93

Mendel Z (1987) Major pests of man-made forests in Israel: origin, biology, damage and control. Phytoparasitica 15:131 t37

Mendel Z, Golan Y, Madar Z (1984) Studies on the phenology and some mortality factors of the eucalyptus borer Phoracantha semipunctata in Israel. La-Yaaran 34:41-44

Montoya-Oliver JM, Rizani B, Meni-Mahzoum A, Allue-Camacho M (1983) Especies de eucaliptos aptas para la production forestal

29

en zonas atacadas por Phoracantha: estudio comparativo de Eucalyptus cladocalyx y Eucalyptus camaldulensis. Ann Rech Forestiere Maroc 23 : 425-444

Myers JH, Monro J, Murry N (1981) Egg clumping, host plant selection and population regulation in Cactoblastis cactorum (Lepidoptera). Oecologia 51:7-13

Powell W (1978) Colonization of twelve species of Eucalyptus by Phoracantha semipunctata (F.) (Coleoptera: Cerambycidae) in Malawi. Bull Entomol Res 68:621-626

Powell W (1982) Age-specific life table data for the Eucalyptus boring beetle, Phoracantha semipunctata (F.) (Coteoptera: Cerambyci- dae), in Malawi. Bull Entomol Res 72:645 653

Price PW, Bouton CE, Gross P, McPheron BA, Thompson JN, Weis AE (1980) Interactions among three trophic levels: influence of plants on interactions between insect herbivores and natural enemies. Ann Rev Ecol Syst 11 : 41 65

Pryor LD (1976) Biology of the eucalypts. Camelot, Southampton, England

Pryor LD, Johnson LAS (1971) A classification of the eucalypts. The Australian National University Press, Canberra, Australia

Rausher MD (1979) Larval habitat suitability and oviposition preference in three related butterflies. Ecology 60:503-511

Roininen H, Tahvanainen J (1989) Host selection and larval per- formance of two willow-feeding sawflies. Ecology 70:129 136

Rose AH (1957) Some notes on the biology of Monochamus scutellatus (Say) (Coleoptera: Cerambycidae). Can Entomol 89:547 553

SAS Institute (1988) SAS/STAT user's guide. Release 6.03 edition. SAS Institute Inc., Cary, North Carolina, USA

Scriven GT, Reeves EL, Luck RF (1986) Beetle from Australia threatens eucalyptus. Cal Agric 40:4-6

Shibata E (1987) Oviposition schedules, survivorship curves, and mortality factors within trees of two cerambycid beetles (Coleop-

tera: Cerambycidae), the Japanese pine sawyer, Monochamus alternatus Hope, and sugi bark borer, Semanotusjaponicus Lacor- daire. Res Popul Ecol 29:347-367

Singer MC (1983) Determinants of multiple host use by a phytophag- ous insect population. Evolution 37:389 403

Smiley J (1978) Plant chemistry and evolution of host specificity: new evidence from Heliconius and Passiflora. Science 201 : 745-747

Sokal RR, Rohlf FJ (1981) Biometry. The principles and practice of statistics in biological research Second edition. Freeman and Co., New York, USA

Thompson JN (1988) Evolutionary ecology of the relationship between oviposition preference and performance of offspring in phytophagous insects. Entomol Exp Appl 47:3-14

Tooke FGC (1935) Insects injurious to forest and shade trees. Bull Dept Agric For S Aft. The Government Printer, Pretoria. 142: 33 39

Walsh KD, Linit MJ (1985) Feeding preferences of the adult pine sawyer, Monocharnus carolinensis (Coleoptera: Cerambycidae), for four pine species. Environ Entomol 13:1164 1166

Wiklund C (1974) The concept of oligophagy and the natural habitat and host plants of Papilio machaon in Fennoscandia. Entomolog- ica Scandinavica 5:151-160

Wiklund C (1975) The evolutionary relationship between adult oviposition preference and larval host plant range in Papilio machaon L. Oecologia 18:185 197

Wiklund C (1981) Generalist vs. specialist oviposition behavior in Papilio machaon (Lepidoptera) and functional aspects on the hierarchy of oviposition preferences. Oikos 36:163-170

Williams KS (1983) The coevolution of Euphydryas chalcedona butterflies and their larval host plants. III. Oviposition behavior and host plant quality. Oecologia 56:336 340

Zacharin RF (1978) Emigrant eucalypts. Gum trees as exotics. Melbourne University Press, Carlton, Australia