EEA 225Operator: Liang Yi Dispatch: 3.11.04 PE: Diane Moody Journal Name Manuscript No. Proofreader: Zhang yeping No. of Pages: 10 Copy-editor: Howard Waller

Blackwell Publishing, Ltd. Influence of the larval environment on performance and adult body size of the wood-boring beetle Phoracantha semipunctata L.M. Hanks1,*, T.D. Paine2 & J.G. Millar 2 1Department of Entomology, University of Illinois at Urbana-Champaign, Urbana, IL 61801, USA 2Department of Entomology, University of California, Riverside, California 92521, USA Accepted: 27 September 2004

Key words: longhorned beetle, Eucalyptus, plant–insect interaction, developmental rate, intraspecific competition, cannibalism, Coleoptera, Cerambycidae

Abstract Large body size confers a reproductive advantage to adults of the wood-boring beetle Phoracantha semipunctata (F.) (Cerambycidae: Cerambycinae: Phoracanthini). Larvae of this species feed sub- cortically in stressed and dying eucalypt trees and logs. We evaluated the influence of the larval environ- ment on larval performance and adult body size by manipulating the post-felling age of host logs (from freshly cut to 2-weeks-old) and the density of colonizing neonates (low density with minimal competition for resources vs. high density with intense competition). Adult beetles emerged in greater numbers from logs that had been subjected to the aging treatment which reduced bark moisture content and favored colonization by neonates. Survival was greatest in larger logs having lower densities of neonates, but was greatly diminished in all treatments by mortality during pupation. Development time varied from 2 months to more than a year and was shortest in smaller logs having high densities of larvae. The size of adult beetles emerging from a log was not influenced by larval density, but was positively correlated with the age of logs when the neonates colonized, and log size. These findings suggest that the optimal developmental conditions for P. semipunctata larvae, in terms of larval performance and adult body size, are available in large, aged host logs having low densities of larvae. Manipulation of the larval environment in this study resulted in a considerable variation in adult body size, but large individuals were relatively more common in the wild popula- tion that was the source of neonates for the experiment. Potential body size may have been constrained by our use of only one host species and a narrow range of log dimensions.

is strongly influenced by the larval environment, in Introduction particular by the inherent nutritional and resistance Large body size confers a fitness advantage in many species qualities of host plant species, and the condition of host because larger individuals dominate in aggressive com- tissues (Andersen & Nilssen, 1983; Haack & Slansky, 1987). petition for mates, are more likely to be chosen as mates, The quality of host available to cerambycid larvae depends and are more fecund (Alcock, 1993). Dominance by large entirely on oviposition choice by the mother, because males in aggressive competition for mates has been larvae are apodous and cannot move between host plants. documented in several species of cerambycid beetles (e.g., Host quality can be greatly altered, however, by intra- and Zeh et al., 1992; Goldsmith & Alcock, 1993; Hanks et al., interspecific exploitative competition among wood-borers 1996; Ginzel et al., 2004). The adult body size of cerambycids for woody tissues (e.g., Ikeda, 1979; Ochi & Katagiri, 1979; 1 Khan & Maiti, 1982; Shibata, 1987; Iba, 1993; Hanks, 1999). Variation in host quality, and the inability to move between hosts, account for a broad variation in UNCORRECTED*Correspondence: Lawrence M. Hanks, Dept. of Entomology, PROOF University of Illinois, 320 Morrill Hall, 505 S. Goodwin Ave., Urbana, developmental rate and adult body size in cerambycids IL 61801, USA. Tel.: +1 217 333 8862; Fax: +1 217 244 3499; (Andersen & Nilssen, 1983; Banno & Yamagami, 1991). E-mail: [email protected] The nutrition of cerambycid larvae also can be influenced

© 2004 The Netherlands Entomological Society Entomologia Experimentalis et Applicata 113: 000–000, 2004 1

2 Hanks et al.

by cannibalism (e.g., Lieu, 1947; Solomon, 1972; Ochi & ability in male-male aggressive competition for mates, and Katagiri, 1979; Iba, 1993). fecundity (Hanks et al., 1996, 1998; L.M. Hanks, J.G. Millar We evaluated the influence of the larval environment and T.D. Paine, unpubl.). on larval performance and adult body size for the wood- Our goals were to evaluate how larval performance and boring beetle Phoracantha semipunctata (F.) (Cerambycidae: adult body size of P. semipunctata are influenced by larval Cerambycinae: Phoracanthini). This species is endemic to density and the post-felling age of host logs, testing Australia, and first appeared in southern California in 1984 hypotheses that were derived from earlier observations (Paine et al., 1995). The larvae develop in eucalypts and and experiments. These hypotheses were as follows: closely related tree species (Wang, 1995), and adults of 1 Larvae develop fastest at low densities because nutri- both sexes are attracted by plant volatiles to stressed and tional quality is maximal; dying trees, fallen branches, and cut logs (for general 2 Fewer neonates succeed in colonizing recently cut logs biology, see Chararas, 1969; Hanks et al., 1993a, 1998, 1999). because of high bark moisture content; The females deposit eggs in batches under loose bark, and 3 Colonization success of neonates is not influenced by usually oviposit on a new host for a period of less than their density because they avoid one another; 1 month in the field (Powell, 1982; Löyttyniemi, 1983; 4 Survivorship of older instars declines with density, Hanks et al., 1993a). If suitable hosts are not available, because of competition for food and cannibalism; and however, females will oviposit on the wrong host species, 5 Body size of adult beetles is greatest under low density resulting in greatly extended development times (Hanks conditions because of superior nutrition. et al., 1995). Females also occasionally oviposit on pre- sumably healthy and resistant eucalypts, resulting in the Materials and methods death of larvae during colonization or development (L.M. Hanks, pers. obs.). Source of host logs The colonization of hosts by neonate P. semipunctata Neonate larvae were transferred to cut logs of Eucalyptus is inhibited by high bark moisture content (Hanks trabutii, a natural hybrid of Eucalyptus camaldulensis et al., 1991b, 1999; Caldeira et al., 2002). Once established, Dehnhardt and Eucalyptus botryoides Sm. (Pryor, 1976). neonates feed within the cambium tissues and avoid one Logs of E. trabutii are high quality hosts for P. semipunctata another by maintaining a thin partition of wood between larvae (see Hanks et al., 1993a). On 14 July 1995, we felled feeding galleries. At high densities, competition among the four trees (mean ± SD circumference at base 31.2 ± 2.8 cm) larvae is intense, and they consume the entire cambial of the same age and provenance in a 13-year-old plantation zone, also furrowing the inner bark and outer sapwood at the Moreno Valley Field Station of the University of (Chararas, 1969; Powell, 1982; Hanks et al., 1991a, 1993a). California, Riverside (UCR). We bucked the trees into 39 Older instars are cannibalistic (see Powell, 1982; Haddan & logs, eliminating sections with branches and standardizing Fraval, 1988; L.M. Hanks, pers. obs.). Larvae may require the surface area of the logs to ca. 1700 cm2 by measuring 2 months to 1 year to complete development, and develop- their circumference and adjusting their lengths accordingly. ment time is strongly influenced by ambient temperature Therefore, logs from the base of trees were shorter and stouter (Chararas, 1969; Löyttyniemi, 1983; Mendel, 1985). Mature than apical logs (mean log circumference 26.0 ± 2.1 cm, larvae bore into the sapwood and excavate a pupation length 64.2 ± 5.7 cm). Within 4 h of felling, the cut ends of chamber, plugging the oval hole to the log surface (here- the logs were sealed with melted paraffin wax to retard after referred to as the ‘exit hole’) with coarse wood fibers. desiccation (see Hanks et al., 1993b). We prepared the Adult beetles exit the pupation chamber by chewing logs to receive larvae by cutting a narrow slit in the bark through this plug. The sex ratio of emerging adults is (ca. 4 cm deep, 7 cm long) parallel to the long axis and in approximately 1 : 1 (e.g., Hanks et al., 1993a,b). the middle of the log. Cutting a tight, V-shaped incision In southern California, adult P. semipunctata average ca. allowed the apodous larvae to gain purchase and begin 2.3 cm in total body length, with standard deviations of ca. boring. 0.4 cm, and females are slightly larger than males (<10%; Hanks et al., 1993a, 1998). We have reared adult P. semi- Transferring larvae to host logs punctata ranging in body length from 1.26 to 3.37 cm from Phoracantha semipunctata neonates for the experiment field-collected eucalypt logs (L.M. Hanks, J.G. Millar and were drawn from a continuous beetle rearing program at UNCORRECTEDT.D. Paine, unpubl.), a high degree of variability even for a UC Riverside [see Hanks PROOFet al. (1993b) for methods]. cerambycid species (see Andersen & Nilssen, 1983). Large Adult beetles were reared from eucalypt logs that had been body size confers a reproductive advantage in this species collected from the field in inland counties of southern because it maximizes powers of dispersal, competitive California during 1994 and 1995. Several hundred adults

Larval environment of a wood-boring beetle 3

were caged in the laboratory, and all eggs deposited during Development rate 2-day periods were pooled in one Petri dish. The Petri We checked the log envelopes for adult beetles every dish was covered and held under ambient laboratory 2 weeks until the first beetles emerged, then daily when conditions until the larvae hatched (ca. 5 days). Larvae less emergence rates were high, until emergence ceased (no than 2 days old were randomly selected and transferred new beetles for 2 years). Logs were retained and envelopes to the pre-cut incisions in the study logs with a damp were re-checked after another 2 years to confirm that no paintbrush (size 0000), and a triple layer of paper toweling more adults had emerged. We recorded the emergence was stapled over the slits to hold the larvae in place. date, size (see below), and sex of each beetle. The development time was calculated as the number of days Manipulating log age and density of larvae between the date that neonates were inserted into logs to The experiment included six treatments representing all the date that each adult was collected. Beetles that were combinations of three log ages and two larval densities, dead when discovered in the envelopes were eliminated with five study logs randomly assigned to each treatment from this data set. combination (n = 30 logs). We manipulated log age by transferring larvae to logs after varying periods of time: on Survivorship the same day that the trees were felled (0-day logs), 7 days Once emergence had ceased, we stripped the study logs of later, or 14 days later. These log ages were chosen in order their bark to measure survivorship. After penetrating the to simulate conditions of hosts that are available to bark, neonates radiated out from bark incisions, etching ovipositing females in the field, from living trees (0 days of characteristically narrow and shallow feeding galleries into aging), to freshly fallen branches (7 days of aging), to older the cambium. Older instars usually avoided feeding near and drier branches (14 days of aging). To confirm that the these concentrations of neonate galleries, perhaps because aging treatment was effective in modifying the quality of the tissues had desiccated more quickly. Thus, feeding logs as larval hosts, we estimated the bark moisture content galleries of neonates were often preserved, even in logs that of the nine logs that were not used in the experiment were riddled with the galleries of older larvae. We used the (three logs per age treatment) at the time the larvae were number of neonate galleries as a measure of colonization transferred to study logs. We sampled bark from each of success (number of galleries/number of neonates trans- these extra logs by cutting three 2-cm squares of bark ferred). Survivorship of later instars was calculated as from around the central circumference, measuring the the percentage of larvae that completed development wet weight of each square, lyophilizing them to a constant (indicated by numbers of exit holes) from the total number weight, and calculating a percentage water content. The of neonates that successfully colonized the cambium (sample water content of the study logs declined significantly with size reduced because neonate galleries were indistinct age, averaging 62 ± 2, 54 ± 1, and 39 ± 1% for the 0-, 7-, in some logs). Survival of pupation was calculated as the and 14-day logs, respectively (means significantly different, percentage of beetles emerging as adults (indicated by

ANOVA; F2,8 = 54.8, P<0.0001). open exit holes) from the total number that had at least Densities of larvae were chosen to represent extremes started to make pupation chambers (including beetles that under field conditions, with minimal competition for failed to emerge, indicated by plugged exit holes). Total cambium tissues at low densities (0.0047 larvae/cm2 of survivorship was calculated as the percentage of beetles bark surface area, or eight larvae per log), and intense com- that emerged as adults from the total number of neonates petition at fourfold higher densities (0.019 larvae/cm2 of that had been transferred to logs. bark, or 32 larvae per log; see Hanks et al., 1993a). After Because the condition of the logs might influence the transferring the neonates, study logs were held under survival of males and females differently, we also tested the laboratory conditions (ca. 20 °C, 40% r.h.) for 24 h, with influence of treatment effects on the sex ratios of adult the bark incision facing upwards, to allow the larvae to beetles emerging from logs. Sex ratio was calculated as the penetrate the bark. The logs then were moved to a green- percentage of adult females per log, including only logs house (30 ± 5 °C, ca. 40% r.h., natural daylight) where they that produced at least four beetles. were randomly positioned on wooden benches, with each log leaning against a 10-cm high lip to minimize contact Body size with the bench that might hinder emergence of adults. Two The index of body size was the length of the right elytron UNCORRECTEDmonths later, logs were enclosed in envelopes of aluminum (in cm) which is tightly correlatedPROOF with total body length window screen with the edges double folded and stapled, (best fit regression for n = 24 individuals of each sex, and one end closed with binder clips to allow access for Y = 0.63X − 0.023, r2 = 0.96, P<0.0001). To evaluate the removing adult beetles. influence of the experimental treatments on variation in

4 Hanks et al.

body size, we compared the distribution of elytron lengths dependent variable). For models that lacked covariates we for beetles in this study (sexes combined) with data for 550 used analysis of variance (ANOVA). Percentages were adult P. semipunctata collected from host trees in Riverside arcsine transformed to meet the model assumptions. We Co., California, in 1990 for a different study (see Hanks confirmed that the data met these assumptions with the

et al., 1998). We tested any differences between the two Fmax test (homogeneity of variances in error terms) and the samples in body size variance with a two-tailed F-test G-test (error terms normally distributed; Sokal & Rohlf, (Sokal & Rohlf, 1995) and tested for normality in the 1995). Differences between means were tested with full distribution of body sizes with the Shapiro-Wilk test  models, but insignificant interactions and random (Analytical Software, 2000). effects were eliminated step-wise from the models, starting with highest order interactions, and leaving only signifi- Statistical considerations cant terms (see Milliken & Johnson, 1984). There were no The performance of P. semipunctata larvae might be significant interactions in any of the analyses (see Results). influenced by several characteristics of the host logs, Differences between pairs of means were tested with including girth, length, and bark thickness (e.g., see Starzyk the REGWQ means-separation test to control maximum & Witkowski, 1986). Log girth and length were strongly experiment-wise error rates (SAS Institute, 2001), and and positively autocorrelated (r2 = 0.69, P<0.0001) as a the tests were ‘protected’ (i.e., means-separation tests were consequence of cutting the logs to a standard surface area. contingent on a significant overall F; Day & Quinn, 1989). Log girth and bark thickness also were autocorrelated Linear associations between variables were tested by (r2 = 0.27, P<0.0031) because bark thickness decreased regression analysis (PROC REG, SAS Institute, 2001). In with the vertical position of the logs within trees. The accuracy all analyses, we confirmed that there were patterns in the of statistical models was not improved by including dispersion of residuals that indicated non-linear relation- multiple autocorrelated variables, so we chose log girth as ships between variables. We present untransformed means a model term to represent these various log characteristics. ± 1 SEM throughout, except where stated otherwise. To assess the impact of the larvae on their food source, we estimated the proportion of cambium tissues of each Results study log that was lost to feeding. This was accomplished by subsampling logs with a transparent plastic sheet Developmental rate marked with a 17 × 23 grid of 1-cm squares, with one The emergence of adult beetles from study logs began ca. square in each of the 23 rows randomly selected for 60 days after starting the experiment (Figure 1), and a total sampling. (This grid size covered about one-quarter of the of 173 beetles emerged. Many beetles (ca. 40%) emerged bark surface of an average study log). We wrapped the grid during a single initial peak (Figure 1), emergence then around one side of the log (conducted blindly), counted declined during December 1995, resumed in January 1996, the sample squares that contained feeding galleries, then and continued through the summer, for a total emergence rotated the log 180° and re-sampled. The total number of period of 360 days. All the logs had at least one larva that squares containing damaged cambium was then divided reached pupation, but there were six logs that yielded no by 46 (× $100) to yield an estimate of the percentage of adult beetles, which reduced the sample size in analyses of cambium consumed. This value averaged 64.3 ± 23% (± SD) developmental rate. across treatments, similar to estimates of feeding damage to E. trabutii logs in an earlier experiment (Hanks et al., 1993a). As would be expected, the percentage of cambium consumed was strongly and positively autocorrelated with the density of mature larvae per log (r2 = 0.60, P<0.0001). We therefore did not include a feeding damage term in the statistical models, but assumed that the damage was strongly associated with larval density. Treatment effects on developmental rate, survivorship, and body size of adult P. semipunctata were tested by analysis of covariance (ANCOVA; PROC MIXED, SAS UNCORRECTEDInstitute, 2001), and fixed effects included log age, number PROOF of neonates transferred, and beetle sex, whereas covariates Figure 1 Phenology of emergence of adult Phoracantha included the density of mature larvae, elytron length, and semipunctata from study logs. Time 0 was the date that the larvae log circumference (choice of effects contingent on the were transferred to the 0-day old study logs.

Larval environment of a wood-boring beetle 5

Developmental time (neonate to adult) averaged 173 ± 97 days (mean ± SD) across treatments and was significantly influenced by the experimental treatments

(overall ANCOVA F4,23 = 5.39, P = 0.0045), with signifi- cant effects being the covariates ‘density of mature larvae’

(F1,23 = 10.4, P = 0.0045) and ‘log circumference’ (F1,23 = 9.6, P = 0.0059). Contrary to Hypothesis 1, development was fastest under high density conditions: mean development time of beetles in logs having the greatest number of mature larvae was about half that of beetles from logs having the fewest larvae (Figure 2). Development time was 2 positively correlated with log circumference (r = 0.24, Figure 2 Relationship between development time of Phoracantha best fit regression line Y = 23.3X – 384). Log age did not semipunctata larvae and their density in study logs. Best fit 2 have a significant effect on development time (F2,23 = 0.08, regression equation: Y = −6.11X + 279 (d.f. = 23; r = 0.27). P = 0.92) which averaged 216 ± 39, 235 ± 29, and 188 ± 24 days for the 0-, 7-, and 14-day treatments, respectively. measure of colonization success, averaged 61 ± 31% (mean ± SD) across treatments, but as expected, there were Survivorship significant differences among treatments. The experiment Feeding galleries of individual neonates could be counted supported Hypothesis 2, that survival of colonization in 22 of the 30 study logs. The survivorship of neonates, a would be lowest in the freshest logs (Figure 3A): 38 ± 9%

UNCORRECTEDFigure 3 Survivorship of Phoracantha semipunctata larvae in study logs: (A) Mean (± SEM) survivorship PROOF of bark colonization by neonates in logs of three ages (post-felling) and two densities of neonates (n = 4, 4, 3, 5, 3, and 3 logs, left to right); (B) Relationship between larval survivorship (established neonates to pupation) and neonate density in study logs; best fit regression equation: Y = −1.46X + 91.8 (d.f. = 21; r2 = 0.29); and (C) Mean (± SEM) total survivorship (from colonizing neonate to emerged adult; n = 5 logs for each mean). 6 Hanks et al.

of neonates colonized 0-day logs, whereas nearly twice as many colonized 7-day logs (64 ± 11%), and even greater numbers colonized 14-day logs (87 ± 5%; overall ANCOVA

F4,21 = 3.84, P = 0.021; log age F2,21 = 6.79, P = 0.0068). Consistent with Hypothesis 3, successful colonization of logs by neonates was not significantly influenced by the number of neonates that had been transferred to the logs, nor was it influenced by log circumference (ANCOVA P>0.05). Survivorship of older larvae (from established neonate to pupation) averaged 71 ± 28% (mean ± SD) across treatments and was also significantly influenced by experi- ± mental treatments (overall F4,21 = 4.85, P = 0.0085). The Figure 4 Relationship between mean elytron length ( SEM) of experiment supported Hypothesis 4, that larval survivor- Phoracantha semipunctata adults and circumference of host log ship would decline with density (Figure 3B; neonate den- (n = 8, 18, 12, 18, 11, 29, and 20 beetles, left to right). Best fit regression equation: Y = 0.027X + 0.65 (d.f. = 117; r2 = 0.18). sity covariate F1,21 = 5.44, P = 0.032). Larval survivorship also was significantly, but positively, correlated with log 2 circumference (F1,21 = 7.24, P = 0.015; r = 0.32; best fit regression line Y = 7.5X – 125). The log age term was not 1.33 ± 0.02 cm for 0-, 7-, and 14-day logs, respectively; significant (ANCOVA P>0.05). REGWQ P<0.05). Body size was also significantly and The percentage of larvae surviving pupation averaged positively correlated with log circumference (Figure 4; ± ± 45 32% (mean SD) across treatments, and was not elytron length F1,117 = 37.2, P<0.0001), with larger beetles significantly influenced by log age, density of mature emerging from stouter logs.

larvae, or log circumference (overall ANCOVA F4,29 = 2.34, Adult P. semipunctata emerging from study logs varied P = 0.083). Total survivorship, from colonizing neonate to considerably in body size (Figure 5), and elytron length emerged adult, averaged 21 ± 16% (mean ± SD) across was normally distributed (Shapiro-Wilk W = 0.99, P = 0.23). treatments and was influenced significantly only by log Mean elytron length (see above) was 9% smaller that that

age (Figure 3C; overall ANCOVA F4,29 = 3.21, P = 0.029; of the 550 adult P. semipunctata that were field collected ± ± log age F2,29 = 4.56, P = 0.020), averaging 10.2 3.2, in the 1990 study (mean 1.44 0.0098, Hanks et al., 1998, ± ± 19.7 5.0, and 30.1 5.0% for log ages 0, 7, and 14 days, means significantly different; ANOVA F1,720 = 47.5, P< respectively. Total survivorship was not significantly influ- 0.0001). Variance in elytron length was 62% lower in the enced by neonate density or log circumference (ANCOVA present study than in the earlier study (variances 0.02 and P>0.05). Sex ratios of adult beetles per log averaged 0.053, respectively; significantly different, F-test statistic = 57 ± 16% (mean ± SD) female and were not significantly 2.65, P<0.001). Thus, adult P. semipunctata in the field influenced by either log age or neonate density (overall tended to be larger on average, resulting in a greater vari-

ANCOVA F3,11 = 0.06, P = 0.98). ance in body size compared to those that were reared under the experimental manipulations of larval environment. Body size Elytron length of P. semipunctata adults averaged Discussion 1.31 ± 0.14 (mean ± SD) cm across treatments and was strongly influenced by the experimental treatments (overall The majority of the beetles completed development in

ANCOVA F5,117 = 10.8, P<0.0001). There was no support, study logs within 3 months (Figure 1), but some indi- however, for Hypothesis 5, that logs with low densities of viduals required more than a year to emerge as adults. larvae would produce the largest adults (density covariate A prolonged emergence period of adult P. semipunctata P>0.05). Consistent with sexual dimorphism in body size results from extended development time due to poor host (see Introduction), females were ca. 6% larger than males quality (Chararas, 1969; Löyttyniemi, 1983; Mendel, 1985; (elytron lengths 1.36 ± 0.016 and 1.28 ± 0.021 cm, respec- Hanks et al., 1993b, 1995), as is true for other cerambycid

tively; beetle sex term significant, F1,117 = 12.0, P = 0.0007). species (Khan & Maiti, 1982; Banno & Yamagami, 1991). The log age term also was significant (F = 5.47, P = Nutritional constraints on the developmental rate of UNCORRECTED2,117 PROOF 0.0054), with logs that received larvae on day 0 producing cerambycid larvae are best illustrated by the greatly adults that were ca. 4.5% smaller than those emerging accelerated growth rates in beetles reared on artificial diet from older logs (means 1.28 ± 0.028, 1.35 ± 0.023, and (e.g., Ivanovic et al., 1989). Larval environment of a wood-boring beetle 7

Figure 5 Frequency distributions of elytron lengths of adult Phoracantha semipunctata emerging from study logs (‘Laboratory reared’) and collected in the field from host trees for an earlier study (‘Field captured’; see Methods).

The inverse relationship between development time and The survivorship of neonates to mature larvae (ca. 70%) density of P. semipunctata larvae (Figure 2) seems counter- was very similar to that in E. trabutii logs in an earlier study intuitive and, in fact, contradicts an earlier study (Powell, (65%; Hanks et al., 1993a), and in logs of a different species 1982). Shorter development time at high densities could in another study (73%; Hanks et al., 1993b). The inverse arise if competition for resources resulted in a shortage of relationship between the survival of older larvae (subsequent food and earlier pupation. However, this relationship to colonization of the log) and densities of neonates (Figure 3B) would result in a negative correlation between density and was consistent with aggressive competition among larvae, adult body size, which was not evident in this study. Develop- and cannibalism. Survivorship was only high in logs that ment time may have been reduced at high densities contained few larvae. The reduced survivorship of P. semi- because feeding by larvae altered the physical and nutritive punctata larvae at high densities has been reported previously qualities of logs in ways that promoted development, such (Paine et al., 2001), and is a common occurrence in other as by reducing moisture content (see below). The increase cerambycid species (e.g., Ikeda, 1979; Togashi, 1986; Iba, in development time of P. semipunctata larvae with log 1993). The positive relationship between survivorship of circumference, however, was consistent with the larger body P. semipunctata larvae and log size has also been reported for sizes of beetles emerging from stouter logs (see below). another cerambycid species (Banno & Yamagami, 1991). Colonization by P. semipunctata neonates was not influ- More than half of the beetles that constructed pupation enced by density, as is true in another cerambycid species chambers in the study logs died of unknown causes. (Togashi, 1990), but was strongly influenced by log age, Mortality rates were similarly high in E. trabutii logs in an being lowest in freshly cut logs (Figure 3A). In our earlier earlier study (Hanks et al., 1993a), and many beetles died studies, the inhibition of colonization in fresh logs was in pupation chambers in a variety of developmental stages, attributed to a high bark moisture content, with moisture from prepupae to sclerotized adults, in other studies content greater than 60% presenting a barrier to coloni- (Hanks et al., 1991a, 1993b). High mortality rates during zation (Hanks et al., 1991b, 1999). Consistent with that this life stage have been reported for P. semipunctata and threshold effect, survivorship in the present experiment other cerambycid species, also due to unknown causes was lowest in 0-day logs (bark moisture content 62%; see (Powell, 1982; Togashi, 1990; Banno & Yamagami, 1991). Methods), but higher in the drier 7- and 14-day logs Old eucalypt logs in the field commonly have similarly (moisture content 54 and 39%, respectively). The coloni- high proportions of plugged exit holes (L.M. Hanks, pers. zation success of neonate P. semipunctata in 7-day logs obs.), confirming that death during the last instar and (averaging ca. 64%) was similar to that in logs of the same pupation is an important mortality factor for P. semipunctata. species and age in an earlier study (77%; Hanks et al., This high mortality rate could be due to several factors, 1993a). High rates of neonate mortality have been attri- including unsuitable microclimate and microbial UNCORRECTEDbuted to host plant effects in several cerambycid species infection. Larvae may also PROOFfail to complete development (Solomon, 1972; Roonwal, 1978; Solomon & Donley, because of the energy required to excavate a tunnel and 1983; Srot, 1983; Shibata, 1987), but to unknown causes in pupation chamber into the desiccated, iron-hard, and another species (Grimble & Knight, 1970). nutrient-poor sapwood of dead eucalypts. This effort 8 Hanks et al.

might deplete larval resources to such an extent that the neonates used in the experiment (Figure 5). Because body larvae cannot successfully complete development. Adults size was correlated with log size (see above), maximum size must then bore their way to freedom through a tunnel may have been constrained by the limited range of log sizes packed solidly with wood fibers. Thus, the protection and used in the experiment. Body size of adult P. semipunctata safety afforded by the pupation chamber may exact a heavy is also strongly influenced by larval host species (Hanks cost from the effort expended for its construction and, et al., 1993a, 1995; Paine et al., 2000), as is true for other later, for the adult to escape its chamber (e.g., Togashi, cerambycid species (e.g., Rice, 1995). Therefore, the use of 1990). The lack of a density effect on the survivorship of a single host species in the present study may also have pupation has been reported for another cerambycid constrained any potential variation in body size. Further- species (Ochi & Katagiri, 1979; Togashi, 1990). more, under natural conditions, P. semipunctata colonizes Total survivorship (neonate to adult; Figure 3C) averaged a broad range of host ages and sizes, from branches only a ca. 20% and reflected the cumulative influences of the few centimeters in diameter with very thin bark, to mature environmental variables. For example, in 0-day logs, about trees with diameters greater than 1 m and bark thicker than three out of the eight neonates successfully colonized the 1 cm (L.M. Hanks, pers. obs.). Thus, the much broader low density logs, whereas about 14 out of 32 larvae estab- spectrum of host quality in the field could produce a lished in the high density logs (from Figure 3A), survivor- greater variation in body size of adult beetles than did the ship of older instars at densities of three and eight neonates standardized logs used in our experiments. However, our was 88 and 81% (from Figure 3B), and about half of these experiments did produce small beetles in proportions mature larvae died during pupation, yielding about one similar to those of the wild population (Figure 5), suggesting adult from low density logs and three adults from high that our experimental treatments were able to simulate density logs. For 14-day logs, almost all the neonates estab- relatively poor quality larval environments. lished in low density logs, whereas 81% established in the In summary, adult P. semipunctata emerged in greater high density logs, survivorship of mature larvae at these numbers from logs that had been subjected to the aging densities averaged ca. 80% and 54%, respectively, and treatment which reduced bark moisture content and about half died during pupation, yielding ca. three adults favored colonization by neonates. Survival was greatest in from low density logs and six adults from high density logs. larger logs having lower densities of neonates, but was Overall, an individual neonate had the greatest odds of greatly diminished in all treatments by mortality during reaching adulthood in the 14-day logs at low larval densities pupation. Development time varied from 2 months to (36% of neonates survived to adult emergence, Figure 3C). more than a year, and was shortest in smaller logs having The sexes did not differ in how they were influenced by high densities of larvae. The size of adult beetles emerging environmental factors, and the lack of a density effect from a log was not influenced by larval density, but was on sex ratio has also been reported in earlier studies on positively correlated with the age of logs when the neonates P. semipunctata (Powell, 1982; Hanks et al., 1991a). colonized, as well as log size. These findings suggest that The body size of adult P. semipunctata increased signifi- the optimal developmental conditions for P. semipunctata cantly with log circumference (Figure 4), but it should be larvae, in terms of larval performance and adult body size, kept in mind that this variable was autocorrelated with log are available in large, aged host logs having low densities of length and bark thickness. Thus, larger body size may have larvae. However, intense competition among adult P. semi- resulted from the improved nutrition provided by thicker punctata for larval hosts in urban landscapes of California bark tissues, or the opportunity for larvae to develop to a results in oviposition in hosts of varying quality and high larger size in logs having thicker subcortical layers, or some densities of larvae (Hanks et al., 1993a), generating the combination of these factors. The physical dimensions of wide range of body sizes characteristic of wild populations. host logs did not influence the adult body size of a different cerambycid species (Ikeda, 1979). It is significant that the Acknowledgements body size of P. semipunctata was not influenced by the den- sity of larvae, consistent with an earlier study (Hanks et al., We thank J. Arredondo, C. Campbell, T. Promlee, and 1991a). One explanation for this finding is that larvae X. Vazguez for their technical assistance, and J.M. Hanks feeding under high density conditions, in which cambial for help with the data. This publication is based upon work tissues are quickly consumed and/or degraded, may com- supported by the Elvenia J. Slosson Endowment Fund, UNCORRECTEDpensate for poorer nutrition by cannibalism. University of California IntegratedPROOF Pest Management Our experiment yielded beetles that were relatively Project, and the Cooperative State Research, Education, small and with a lower variance in body size compared to and Extension Service, US Department of Agriculture, beetles of the wild population that was the source of the under Agreement no. 95-37312-1633. Larval environment of a wood-boring beetle 9

Hanks LM, Millar JG & Paine TD (1995) Biological constraints References on host range expansion by the wood-boring beetle Phora- Alcock J (1993) Animal Behavior: An Evolutionary Approach. cantha semipunctata (Coleoptera: Cerambycidae). Annals of Sinauer, Sunderland, MA. the Entomological Society of America 88: 183–188. Andersen J & Nillsen AC (1983) Intrapopulation size variation of Hanks LM, Millar JG & Paine TD (1996) Body size influences free-living and tree-boring Coleoptera. Canadian Entomologist mating success of the eucalyptus longhorned borer (Coleoptera: 115: 1453–1464. Cerambycidae). Journal of Insect Behavior 9: 369–382. 2 Analytical Software (2000) Statistix 7 User’s Manual. Analytical Hanks LM, Millar JG & Paine TD (1998) Dispersal of the Software, Tallahassee, FL. eucalyptus longhorned borer (Coleoptera: Cerambycidae) in Banno H & Yamagami A (1991) Life cycle and larval survival rate urban landscapes. Environmental Entomology 27: 1418–1424. of the redspotted longicorn beetle, Eupromus ruber (Dalman) Hanks LM, Paine TD & Millar JG (1991b) Mechanisms of resist- (Coleoptera: Cerambycidae). Applied Entomology and Zoology ance in Eucalyptus against larvae of the eucalyptus longhorned 26: 195–204. borer (Coleoptera: Cerambycidae). Environmental Entomo- Caldeira M, Fernandéz V, Tomé J & Pereira JS (2002) Positive logy 20: 1583–1588. effects of drought on longicorn borer larval survival and growth Hanks LM, Paine TD & Millar JG (1993a) Host species preference on eucalyptus trunks. Annals of Forest Science 59: 99–106. and larval performance in the wood-boring beetle Phoracantha Chararas C (1969) Biologie et écologie de Phoracantha semipunc- semipunctata F. Oecologia 95: 22–29. tata F. (Coléoptère Cerambycidae xylophage) ravageur des Hanks LM, Paine TD, Millar JG, Campbell CD & Schuch UK Eucalyptus en Tunisie, et methodes de protection des peuple- (1999) Water relations of host trees and resistance to the ments. Annales de l’Institut National de Recherches Forestières phloem-boring beetle Phoracantha semipunctata F. (Coleo- de Tunisie 2: 1–37. ptera: Cerambycidae). Oecologia 119: 400–407. Day RW & Quinn GP (1989) Comparisons of treatments after an Iba M (1993) Studies on the ecology and control of the yellow- analysis of variance in ecology. Ecological Monographs 59: spotted longicorn beetle, Psacothea hilaris Pascoe (Coleoptera: 433–463. Cerambycidae), in the mulberry fields. Bulletin of the National Ginzel MD, Ray AM & Hanks LM (2004) Mating strategy and Institute of Sericulture and Entomological Science 8: 1–119. sexual dimorphism in male Megacyllene robiniae (Förster) Ikeda K (1979) Consumption and food utilization by individual (Coleoptera: Cerambycidae). Journal of Insect Science, larvae and the population of a wood borer, Phymatodes maaki submitted. Kraatz (Coleoptera: Cerambycidae). Oecologia 40: 287–298. 3 Goldsmith SK & Alcock J (1993) The mating chances of small Ivanovic J, Jankovic-Hladni M, Stanic V, Nenadovic V & Frusic M males of the cerambycid beetle Trachyderes mandibularis differ (1989) The role of neurosecretion and metabolism in develop- in different environments (Coleoptera: Cerambycidae). Journal ment of an oligophagous feeding habit in Morimus funereus of Insect Behavior 6: 351–360. larvae (Col., Cerambycidae). Comparative Biochemistry and Grimble DG & Knight FB (1970) Life tables and mortality factors Physiology Part A: Physiology 94: 167–171. for Saperda inornata (Coleoptera: Cerambycidae). Annals of Khan TN & Maiti PK (1982) The bionomics of the round-head the Entomological Society of America 63: 1309–1319. borer, Olenecamptus bilobus (Fabricius) (Coleoptera: Ceram- Haack RA & Slansky F Jr (1987) Nutritional Ecology of bycidae). Proceedings of the Zoological Society of Calcutta 33: Wood-Feeding Coleoptera, Lepidoptera, and Hymenoptera. 71–85. Nutritional Ecology of Insects, Mites, Spiders, and Related Lieu KOV (1947) The study of wood borers in China II: Biology Invertebrates (ed. by F Slansky Jr & JG Rodriguez), pp. 449– and control of the citrus-trunk-cerambycids, Nadezhdiella 487. John Wiley, NY. cantori (Hope) (Coleoptera). Musée Heude Notes D’entomo- Haddan M & Fraval A (1988) Estimation a posteriori, à partir des logie Chinoise 11: 69–117. caractéristiques des galeries de la mortalité larvaire de Phora- Löyttyniemi K (1983) Flight pattern and voltinism of Phoracantha cantha semipunctata (F.) (Col. Cerambycidae), ravageur xylo- beetles (Coleoptera, Cerambycidae) in a semihumid tropical phage des eucalyptus au Maroc. Actes Institut Agronomique et climate in Zambia. Annales Entomologici Fennici 49: 49–53. Vétérinaire (Maroc) 8: 27–34. Mendel Z (1985) Seasonal development of the eucalypt borer, Hanks LM (1999) Influence of the larval host plant on reproduc- Phoracantha semipunctata, in Israel. Phytoparasitica 13: 85– tive strategies of cerambycid beetles. Annual Review of 93. Entomology 44: 483–505. Milliken GA & Johnson DE (1984) Analysis of Messy Data, Vol. Hanks LM, McElfresh JS, Millar JG & Paine TD (1993b) 1: Designed Experiments. Van Nostrand Reinhold, New York. Phoracantha semipunctata (Coleoptera: Cerambycidae), a serious Ochi K & Katagiri K (1979) Survival curves of Monochamus pest of Eucalyptus. California: biology and laboratory-rearing alternatus Hope (Coleoptera: Cerambycidae) population on procedures. Annals of the Entomological Society of America the pine dead trees. Bulletin of the Forestry and Forest Products 86: 96–102. Research Institute 303: 125–152 (in Japanese). UNCORRECTEDHanks LM, Millar JG & Paine TD (1991a) Evaluation of cold Paine TD, Hanks LM, Millar JGPROOF & Paine EO (2000) Attractiveness temperatures and density as mortality factors of the eucalyptus and suitability of host tree species for colonization and survival longhorned borer (Coleoptera: Cerambycidae) in California. of Phoracantha semipunctata (Coleoptera: Cerambycidae). Environmental Entomology 20: 1653–1658. Canadian Entomologist 132: 907–913. 10 Hanks et al.

Paine TD, Millar JG & Hanks LM (1995) Integrated program japonicus Lacordaire. Researches on Population Ecology 29: protects trees from eucalyptus longhorned borer. California 347–367. Agriculture 49: 34–37. Sokal RR & Rohlf FJ (1995) Biometry, 3rd edn. W.H. Freeman, Paine TD, Millar JG, Paine EO & Hanks LM (2001) Influence of New York. host log age and refuge from natural enemies on colonization Solomon JD (1972) Biology and habits of the living beech borer and survival of Phoracantha semipunctata. Entomologia in red oaks. Journal of Economic Entomology 65: 1307–1310. Experimentalis et Applicata 98: 157–163. Solomon JD & Donley DE (1983) Bionomics and control of the Powell W (1982) Age-specific life-table data for the Eucalyptus white oak borer. USDA Forest Service Research Paper SO-198. boring beetle, Phoracantha semipunctata F. (Coleoptera: Srot M (1983) Factors reducing the population density of large poplar Cerambycidae), in Malawi. Bulletin of Entomological Research borer (Saperda carcharias L.). Lesnictví 29: 785–803 (in Czech). 72: 645–653. Starzyk JR & Witkowski Z (1986) Dependence of the sex ratio of Pryor LD (1976) The Biology of Eucalypts. Edward Arnold, cerambycid beetles (Col., Cerambycidae) on the size of their London. host trees. Journal of Applied Entomology 101: 140–146. Rice ME (1995) Branch girdling by Oncideres cingulata (Coleo- Togashi K (1986) Effects of the initial density and natural enemies ptera: Cerambycidae) and relative host quality of persimmon, on the survival rate of the Japanese pine sawyer, Monochamus hickory, and elm. Annals of the Entomological Society of alternatus Hope (Coleoptera: Cerambycidae), in pine logs. America 88: 451–455. Applied Entomology and Zoology 21: 244–251. Roonwal ML (1978) The biology, ecology, and control of the sal Togashi K (1990) Life table for Monochamus alternatus heartwood borer, Hoplocerambyx spinicornis: a review of recent (Coleoptera, Cerambycidae) with dead trees of Pinus thunbergii. work. Indian Journal of Forestry 1: 21–34. Japanese Journal of Entomology 58: 217–230. SAS Institute (2001) SAS/STAT User’s Guide for Personal Wang Q (1995) A taxonomic revision of the Australian genus Computers, release 8.01. SAS Institute, Cary, NC. Phoracantha (Coleoptera: Cerambycidae). Invertebrate Taxonomy Shibata E (1987) Oviposition schedules, survivorship curves, and 9: 865–958. mortality factors within trees of two cerambycid species Zeh DW, Zeh JA & Tavakilian G (1992) Sexual selection and sexual (Coleoptera: Cerambycidae), the Japanese pine sawyer, dimorphism in the harlequin beetle Acrocinus longimanus. Monochamus alternatus Hope, and sugi bark borer, Semanotus Biotropica 24: 86–96.

UNCORRECTED PROOF Author Query Form

Journal: Entomologia Experimentalis et Applicata

Article: eea_225.fm

Dear Author,

During the copy-editing of your paper, the following queries arose. Please respond to these by marking up your proofs with the necessary changes/additions. Please write your answers on the query sheet if there is insufficient space on the page proofs. Please write clearly and follow the conventions shown on the attached corrections sheet. If returning the proof by fax do not write too close to the paper’s edge. Please remember that illegible mark-ups may delay publication.

Many thanks for your assistance.

No. Query Remarks 1 Ginzel & Hanks 2004 has been changed to Ginzel et al. 2004 so that this citation matches the list 2 ‘Andersen & Nillsen’ has been changed to ‘Andersen & Nilssen’ to match the text citations 3More details available for Ginzel et al. 2004? MARKED PROOF ÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐ Please correct and return this set ÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐÐ

Please use the proof correction marks shown below for all alterations and corrections. If you wish to return your proof by fax you should ensure that all amendments are written clearly in dark ink and are made well within the page margins.

Instruction to printer Textual mark Marginal mark Leaveunchanged under matter to remain Stet Insert in text the matter New matter followed by indicated in the margin Delete through matter to be deleted Delete and close up through matter to be deleted Substitute character or through letter or through New letter or new word substitute part of one or word more word(s) Change to italics under matter to be changed Change to capitals under matter to be changed Change to small capitals under matter to be changed Change to bold type under matter to be changed Change to bold italic under matter to be changed Change to lower case Encircle matter to be changed Change italic to upright type (As above) Insert `superior' character through character or where under character required e.g. Insert `inferior' character (As above) over character e.g. Insert full stop (As above) Insert comma (As above) Insert single quotation marks (As above) and/or Insert double quotation (As above) and/or marks Insert hyphen (As above) Start new paragraph No new paragraph Transpose Close up linking letters Insert space between letters between letters affected Insert space between words between words affected Reduce space between letters between letters affected Reduce space between words between words affected