Understanding the compensatory responses of crops to pest damage is important in developing pest thresholds. Compensation for pest damage in crops can occur at the plant level, where the architecture, growth dynamics and allocation patterns of damaged plants are altered, allowing them to recover or, at the crop level, where differential damage between plants may alter plant-to-plant interactions. We investigated growth and yield of cotton (Gossypium hirsutum L.) following non-uniform manual defoliation of seedlings. This partially replicates real pest damage and is valuable in understanding crop-level responses to damage because it can be inflicted precisely. Damage distributions included damaging 0, 25, 50, 75 or 100% of the plants. Damage intensity for the damaged plants was varied by removing 100 or 75% of each true leaf when plants had two, four and six true leaves. At the crop level, yield loss increased as the proportion of plants damaged and intensity of damage per damaged plant increased. Neighbour interactions occurred; undamaged plants with damaged neighbours grew larger and yielded better than undamaged plants with undamaged neighbours, while the converse applied for damaged plants with undamaged neighbours. Neighbour interactions were influenced by the intensity of damage and were stronger when 100% of the leaf area was removed than when 75% was removed. At the crop level, when compared with yield estimates based on yield of plants from uniformly damaged or undamaged plots, these interactions resulted in higher yield than expected (+8%). This suggests that damage distribution may have to be considered in studies where artificial or real pest damage is inflicted uniformly on plants.

A real-time PCR assay based on TaqMan® technology was developed and evaluated for the rapid detection of the B and Q biotypes of Bemisia tabaci (Gennadius) (Hemiptera: Aleyrodidae). A survey was conducted during 2005–2007 in order to identify the distribution and prevalence of B. tabaci biotypes in Cyprus using the real-time PCR assay. More than 700 adult whiteflies collected from 35 cultivated and weed plant species were individually haplotyped using TaqMan® PCR, and the results of the assay were validated by restriction fragment length polymorphism analysis and DNA sequencing of the mitochondrial cytochrome oxidase I (mtCOI) gene. Two biotypes, B and Q, were identified in the collected plant species on the island. The real-time PCR and RFLP assay consistently yielded the same results, although the real-time assay was more sensitive and less time consuming. Phylogenetic analysis of the mtCOI DNA sequences corroborated the identity of the B and Q biotypes 100% of the time and by phylogenetic analysis the haplotypes grouped, as expected, in the major North African-Mediterranean-Middle Eastern clade of the B. tabaci complex.

The cereal rust mite, Abacarus hystrix, is one of the most notable among mites causing losses in cultivated grasslands. It is one of a few eriophyoid species for which a broad host range has been reported. Recent studies, however, have shown that host specialization is very likely in this species. For two populations of A. hystrix (one inhabiting perennial ryegrass, the second inhabiting quackgrass), host-associated differences correlated with strong host fidelity, distinct phenotypes and reproductive barriers have been found. In the present study, the ability of wheat colonization by quackgrass- and ryegrass-adapted cereal rust mite was studied. The hypothesis that the potential for wheat colonization by the quackgrass strain is more likely was tested by comparing the colonization performance (assessed by female survival and fecundity) of quackgrass- and ryegrass-associated A. hystrix on their familiar hosts and on wheat. The ryegrass population had no success in wheat colonization (expressed by extremely low fecundity and female survival). Fecundity and survival of quackgrass strain were similar on wheat and the familiar host, or even higher on wheat. Phylogenetic similarity of quackgrass and wheat is discussed as a possible factor that might influence such patterns of host colonization. Since A. hystrix is the only vector of the ryegrass mosaic virus (RgMV), the presented results may be helpful in explaining the inability of RgMV to successfully infest wheat. The conclusions are that (i) quackgrass- and ryegrass-adapted strains of the cereal rust mite have different physiological host ranges and (ii) phylogenetic relationships between host plant species appear to be drivers for host specialization in this mite species.

The effects of Bt transgenic cottons (Bt-I expressing cry1Ac and Bt-II expressing cry1Ab and cry2Ab or cry1Ab and cry1Fa) and non-Bt cottons on feeding, oviposition and longevity of adults, and development and survival of Liriomyza trifolii larvae were studied under laboratory conditions; and infestation on four Bt and two non-Bt cotton traits were investigated under field conditions. Laboratory choice and no-choice tests showed that L. trifolii adults were capable of distinguishing between Bt cottons and non-Bt cottons. In a choice test on younger plants (4–5 leaves), the adults were found more often and made more feeding punctures (FP) on non-Bt cottons than on Bt cottons. On older plants (8–9 leaves), adults made the most FP on non-Bt cotton followed by those on Bt-II cottons and the least on Bt-I cotton. The females oviposited more eggs (6.7 eggs per leaf) on non-Bt cotton than on Bt-I (1.7 eggs per leaf) and Bt-II (0.8 eggs per leaf) cottons on younger plants and oviposited similar numbers of eggs (0.7–1.3 eggs per leaf) on non-Bt and Bt cottons on older plants. In a no-choice test, the females also fed more FP on non-Bt cottons than on Bt cottons on both younger and older plants. The females oviposited more eggs (15.6 eggs per leaf) on non-Bt cotton than on Bt-I (8.2 eggs per leaf) and Bt-II (6.5 eggs per leaf) cottons on younger plants and similar numbers of eggs (2.5–3.3 eggs per leaf) on non-Bt and Bt cottons on older plants. Larval and puparial survivals were not different among Bt and non-Bt cottons. The occurrence and damage of leafminers on cottons in the field showed that L. trifolii infested more plants and leaves and had more mines on non-Bt cotton than on Bt cottons.

The wheat curl mite (WCM), Aceria tosichella Keifer, is an eriophyoid pest of cereals, and the vector responsible for transmitting wheat streak mosaic virus. Several authors have suggested cryptic species of this mite identified through morphological variation, but this has never been conclusively demonstrated. Here, we use the mitochondrial 16S rRNA gene and two nuclear markers (internal transcribed spacer 1 and adenine nucleotide translocase) to show that WCM from Australia consists of at least two separate lineages that may represent putative species. In our study, both WCM variants were widespread and the only eriophyoids found on wheat varieties. The WCM variants were also found on alternate host plants, including some plants not known to host WCM. These results have implications for the control of this pest within Australian cereal crops.

Culicoides imicola is the main vector for bluetongue (BT) and African horse sickness (AHS) viruses in the Mediterranean basin and in southern Europe. In this study, we analysed partial mitochondrial cytochrome c oxidase subunit I (COI) gene to characterize and confirm population expansion of Culicoides imicola across Spain. The data were analysed at two hierarchical levels to test the relationship between C. imicola haplotypes in Spain (n=215 from 58 different locations) and worldwide (n=277). We found nineteen different haplotypes within the Spanish population, including 11 new haplotypes. No matrilineal subdivision was found within the Spanish population, while western and eastern Mediterranean C. imicola populations were very structured. These findings were further supported by median networks and mismatch haplotype distributions. Median networks demonstrated that the haplotypes we observed in the western Mediterranean region were closely related with one another, creating a clear star-like phylogeny separated only by a single mutation from eastern haplotypes. The two, genetically distinct, sources of C. imicola in the Mediterranean basin, thus, were confirmed. This type of star-like population structure centred around the most frequent haplotype is best explained by rapid expansion. Furthermore, the proposed northern expansion was also supported by the statistically negative Tajima's D and Fu's Fs values, as well as predicted mismatch distributions of sudden and spatially expanding populations. Our results thus indicated that C. imicola population expansion was a rapid and recent phenomenon.

A resistance monitoring program conducted for the polyphagous whitefly, Bemisia tabaci (Gennadius), in Imperial Valley, CA, USA generated a large set of LC50s for adults collected from broccoli, cantaloupe and cotton crops over a four-year period. A vial bioassay and, subsequently, a yellow-sticky card bioassay produced similar temporal profiles of relative susceptibilities to the pyrethroid insecticide bifenthrin. Both bioassays revealed that whiteflies collected from broccoli were significantly less susceptible to bifenthrin compared to the other two crops. A similar finding was observed for endosulfan and the mixture of bifenthrin+endosulfan in the yellow-sticky card bioassay. The possibility that seasonal differences contributed to the observed differences in susceptibility provided the impetus to conduct a reciprocal transfer experiment using broccoli (or kale) and cantaloupe grown simultaneously in the field and greenhouse. Whitefly adults collected from an organic farm over three consecutive weeks had significantly higher LC50s on kale than those collected the same day on cantaloupe. After culturing in the greenhouse on broccoli or cantaloupe and testing again, LC50s remained significantly higher on broccoli after one week and again at the F1 generation. In contrast, whiteflies originating on kale in the field and transferred to cantaloupes in the greenhouse had significantly reduced LC50s at the F1 generation. When tested against the bifenthrin+endosulfan mixture, significantly higher LC50s were generated for whiteflies reared on broccoli in the greenhouse at one week and the F1 compared to the field source from cantaloupes. The consistently higher LC50s for whiteflies on broccoli and other Brassica spp. crops, compared to cantaloupes or cotton, point to statistically significant host-plant influences that are expressed in both field-collected and greenhouse-reared populations of whiteflies.

In biological control programmes introduced natural enemies compete with indigenous enemies for hosts and may also engage in intraguild predation when two species competing for the same prey attack and consume one another. The large pine weevil, Hylobius abietis L. (Coleoptera: Curculionidae), is an important pest of coniferous reforestation in Europe. Among its natural enemies, the parasitoid Bracon hylobii Ratz. (Hymenoptera: Braconidae) and entomopathogenic nematodes have potential as biological control agents. Both parasitoid and nematodes target the weevil larvae and, hence, there is potential for competition or intraguild predation.

In this study, we examine the interaction of B. hylobii with the nematode Heterorhabditis downesi Stock, Griffin and Burnell (Nematode: Heterorhabditidae), testing the susceptibility of larvae, pupae and adults of B. hylobii to H. downesi and whether female parasitoids discriminate between nematode-infected and uninfected weevils for oviposition. In choice tests, when weevils were exposed to nematodes 1–7 days previously, no B. hylobii oviposited on nematode-infected weevil larvae. Up to 24 h, healthy weevils were twice as likely as nematode-infected ones to be used for oviposition. Bracon hylobii females did not adjust clutch size; nematode-infected hosts were either rejected or the parasitoid laid a full clutch of eggs on them.

When nematodes were applied to the parasitoid feeding on weevil larvae, the nematodes parasitized the parasitoid larvae, there was a reduction in cocoon formation and fewer cocoons eclosed. Eclosion rate was not reduced when nematodes were applied to fully formed cocoons, but nearly all of the emerging adults were killed by nematodes.

House-resting Anopheles mosquitoes are targeted for vector control interventions; however, without proper species identification, the importance of these Anopheles to malaria transmission is unknown. Anopheles longipalpis, a non-vector species, has been found in significant numbers resting indoors in houses in southern Zambia, potentially impacting on the utilization of scarce resources for vector control. The identification of An. longipalpis is currently based on classical morphology using minor characteristics in the adult stage and major ones in the larval stage. The close similarity to the major malaria vector An. funestus led to investigations into the development of a molecular assay for identification of An. longipalpis. Molecular analysis of An. longipalpis from South Africa and Zambia revealed marked differences in size and nucleotide sequence in the second internal transcribed spacer (ITS2) region of ribosomal DNA between these two populations, leading to the conclusion that more than one species was being analysed. Phylogenetic analysis showed the Zambian samples aligned with An. funestus, An. vaneedeni and An. parensis, whereas the South African sample aligned with An. leesoni, a species that is considered to be more closely related to the Asian An. minimus subgroup than to the African An. funestus subgroup. Species-specific primers were designed to be used in a multiplex PCR assay to distinguish between these two cryptic species and members of the An. funestus subgroup for which there is already a multiplex PCR assay.

Analysis of DNA sequence data has proven invaluable for defining the relationships among taxa, as well as resolving their evolutionary histories. Here, we analyzed DNA sequence variation of one mitochondrial gene (COI) and two nuclear regions (ITSI and II) to clarify the phylogenetic position of Culicoides dewulfi, a midge species widely spread in Europe and a suspected vector for bluetongue virus. Various authors have described C. dewulfi either as part of the Culicoides obsoletus sensu lato complex or as a separate taxonomic group. A maximum likelihood phylogeny, based upon an optimal model of sequence evolution, placed C. dewulfi outwith the C. obsoletus s.l. complex. Shimodaira-Hasegawa test highlighted that this topology was significantly more likely than any topology that placed C. dewulfi anywhere else in the phylogeny. As such, C. dewulfi should not be considered part of the C. obsoletus s.l. complex and instead be treated as a separate group, phylogenetically close to the classical Old World vector C. imicola.

The differences between the seasonal occurrence and likelihood of being infected by FD phytoplasmas, of male and female Scaphoideus titanus Ball, were investigated. Sex ratio (male: female) was calculated by counting males and females sampled by means of yellow sticky traps and sweep-nets and from adults derived from hatched eggs in field-collected grapevine wood. PCR essays were performed to test differences in infection between genders. Overall, the sex ratio on sticky traps was significantly more male biased (1.99:1) if compared to net sweeping (0.62:1) and laboratory rearing (0.60:1). The peak of male presence was recorded in the middle of July in laboratory rearing and sweep net, and in the middle of August on sticky traps; the maximum presence of females was detected at the end of July in laboratory rearing, and at the end of August in sweep net samplings and on sticky traps. The seasonal sex ratio was more male biased at the beginning in laboratory rearing (1.50:1) and sticky traps (9:1), and then decreased in favor of females at the end of the sampling period, both in laboratory rearing (0.17:1) and in sticky traps (0.07:1). This trend was significantly less skewed, although similar, in sweep net samplings that recorded a sex ratio of 1:1 and 0.16:1 at the beginning and at the end of the sampling period, respectively. Concerning phytoplasma detection, an interaction between gender and sampling period was observed, the males showing a peak of infected individuals later in the season (35%). Some possible behavioral explanations of the data obtained are given.

The seed fly Mesoclanis polana (Diptera: Tephritidae) was released in Australia in 1996. Its impact on seed production of bitou bush (Chrysanthemoides monilifera subsp. rotundata) was monitored at eight sites along the New South Wales coast from 1996 to 2004. Peak flowerhead production occurred in autumn (March to May); therefore, samples collected in May of each year were used to compare abundance and impact of M. polana across sites and between years. Latitude had a significant effect on abundance and impact of M. polana. By May 2004, 99.6% of flowerheads at the five most northern sites contained at least one egg, while 64% of flowerheads from the two most southern sites contained at least one egg. In May 2004, mean numbers of M. polana eggs per flowerhead were between 13 and 17 at four of the six northern sites and below two at the two southern sites. In May 2004, average seed destruction by M. polana was 58% at the four most northern sites and 11% at the two most southern sites. The highest level of seed destruction recorded was 86% at Kingscliff in May 2003. Percentage seed destruction in May provided a reasonable estimate of seed destruction for the whole year. Parasitoids were found throughout the range of M. polana in Australia. Their attack rate on M. polana was less than 10% at all sites, except at the two most northern sites where parasitism of up to 27% was recorded. Parasitism results are compared with predictions made following an earlier study. The release of Mesoclanis magnipalpis, a species suited to cooler environments, is recommended for the southern range of C. monilifera subsp. rotundata.

All aphid species studied so far share the same sex pheromone components, nepetalactol and nepetalactone. Variation by different enantiomers and blends of the two components released by different aphid species are limited and can only partially explain species-specific attraction of males to females. While some host-plant odours are known to enhance specific attraction of aphid species, herbivore-induced plant volatiles that synergise attractiveness to the sex pheromone are unknown. Here, we demonstrate that for the host-alternating rosy apple aphid (Dysaphis plantaginea (Passerini)) specificity of attraction of males to females is triggered by female-induced tree odours in combination with a 1:8 ratio of (4aS,7S,7aR)-nepetalactone and (1R,4aS,7S,7aR)-nepetalactol. Female aphid infestation induces increased release of four esters (hexyl butyrate, (E)-2-hexenyl butyrate, (Z)-3-hexenyl 3-methylbutyrate and hexyl 2-methylbutyrate) from apple leaves. Two different combinations of three esters applied in a 1:1:1 ratio increase the number of male D. plantaginea and decrease the number of other aphid species caught in water traps in the presence of the pheromone components. The ester blend alone was not attractive. Combination of the pheromone blend with each single ester was not increasing attraction of male D. plantaginea. The demonstration that sexual aphid species use herbivore-induced plant volatiles as a species-specific attractant for mate finding adds a new dimension to our understanding of insect species using or manipulating chemical cues of host plants for orientation.

Aphids are, arguably, the single most damaging group of agricultural insect pests throughout the world. Plant tolerance, which is a plant response to an insect pest, is viewed as an excellent management strategy. Developing testable hypotheses based on genome-wide and more focused methods will help in understanding the molecular underpinnings of plant tolerance to aphid herbivory. As a first step in this process, we undertook transcript profiling with Affymetrix GeneChip Barley Genome arrays using RNA extracted from tissues of tolerant and susceptible genotypes collected at three hours, three days and six days after Diuraphis noxia introduction. Acquired data were compared to identify changes unique to the tolerant barley at each harvest date. Transcript abundance of 4086 genes was differentially changed over the three harvest dates in tolerant and susceptible barley in response to D. noxia feeding. Across the three harvest dates, the greatest number of genes was differentially expressed in both barleys at three days after aphid introduction. A total of 909 genes showed significant levels of change in the tolerant barley in response to D. noxia feeding as compared to susceptible plants infested with aphids. Many of these genes could be assigned to specific metabolic categories, including several associated with plant defense and scavenging of reactive oxygen species (ROS). Interestingly, two peroxidase genes, designated HvPRXA1 and HvPRXA2, were up-regulated to a greater degree in response to D. noxia feeding on tolerant barley plants, indicating that specific peroxidases could be important for the tolerance process. These findings suggest that the ability to elevate and sustain levels of ROS-scavenging enzymes could play an important role in the tolerant response.

Omnivory is common among arthropods, but little is known about how availability of plant resources and prey affects interactions between species operating at the third and fourth trophic level. We used laboratory and field cage experiments to investigate how the provision of flowers affects an omnivorous lacewing, Micromus tasmaniae (Hemerobiidae) and its parasitoid Anacharis zealandica (Figitidae). The adult lacewing is a true omnivore that feeds on both floral resources and aphids, whereas the parasitoid is a life-history omnivore, feeding on lacewing larvae in the larval stage and floral nectar as an adult. We showed that the effect of floral resources (buckwheat) on lacewing oviposition depends on prey (aphid) density, having a positive effect only at low prey density and that buckwheat substantially increases the longevity of the adult parasitoid. In field cages, we tested how provision of flowering buckwheat affects the dynamics of a four trophic level system, comprising parasitoids, lacewings, pea aphids and alfalfa. We found that provision of buckwheat decreased the density of lacewings in the first phase of the experiment when the density of aphids was high. This effect was probably caused by increased rate of parasitism by the parasitoid, which benefits from the presence of buckwheat. Towards the end of the experiment when the aphid populations had declined to low levels, the effect of buckwheat on lacewing density became positive, probably because lacewings were starving in the no-buckwheat treatment. Although presence of buckwheat flowers did not affect aphid populations in the field cages, these findings highlight the need to consider multitrophic interactions when proposing provision of floral resources as a technique for sustainable pest management.

The effects of different prey regimes on the performance and digestive physiology of the spined soldier bug, Podisus maculiventris (Say) (Hemiptera: Pentatomidae), were assessed. Specifically, P. maculiventris nymphs were fed on Colorado potato beetle (CPB), Leptinotarsa decemlineata Say (Coleoptera: Chrysomelidae), larvae; Egyptian cotton leafworm (ECW); Spodoptera littoralis (Boisduval) (Lepidoptera: Noctuidae); larvae; Calliphora spp. (CAL) (Diptera: Calliphoridae) pupae or a mixture of the three prey. No differences in development and weight gain were observed when P. maculiventris nymphs were fed different prey species (CPB, ECW or CAL). However, an increase in weight gain and a reduction in the duration of the stadia were observed for nymphs fed with a mixture of the three prey. To investigate the physiological background, biochemical analysis were carried out on insects dissected at the end of the feeding assay. We have found that the proteolytic activity in the salivary glands of P. maculiventris nymphs was not affected by prey species, whereas the relative activity of these proteases in the midgut depends on the prey. Moreover, gel assays proved that the proteolytic profiles of midguts from P. maculiventris nymphs feeding on CPB, ECW and CPB closely resembled those of their prey. All together, these results suggest that P. maculiventris may utilize enzymes from the prey they consume that may facilitate the process of digestion.

An investigation of the diversity and distribution of wild crucifer species and their importance for cultivated crucifers was conducted during 2005 and 2006 in the highland and mid-altitude semi-arid areas of Kenya. Thirteen species of wild crucifers in nine genera were recorded: Raphanus raphanistrum, Erucastrum arabicum, Sisymbrium officinale, Crambe kilimandscharica, Capsella bursa-pastoris, Rorippa nudiuscula, Ro. micrantha, Ro. microphylla, Lepidium bonariense, Coronopus didymus, Brassica rapa, B. juncea and an unidentified Brassica species. Highland areas had significantly higher species diversity and species richness than mid-altitude semi-arid areas. Species richness, diversity and evenness varied with season and location. Raphanus raphanistrum was the dominant non-cultivated species in the highlands followed by E. arabicum, which was also present and dominant in the semi-arid study sites. Diamondback moth (DBM) was recorded from ten wild crucifer species and R. raphanistrum and E. arabicum were the preferred host plant species. Overall, four larval, one larval-pupal and one pupal parasitoid of DBM were recorded: Diadegma semiclausum, D. mollipla, Apanteles sp., Cotesia plutellae, Oomyzus sokolowskii and Brachymeria species, respectively. Diadegma semiclausum was the most dominant species on all crucifers. We conclude that wild crucifers act as alternative hosts for DBM and provide refugia for DBM parasitoids, which risk local extinction through pesticide application or competition from introduced exotic parasitoid species. The wild crucifers also act as recolonization sites for DBM parasitoids.

A disrupted allele (r1) of a cadherin gene (Ha_BtR) is genetically associated with incompletely recessive resistance to Bacillus thuringiensis toxin Cry1Ac in a Cry1Ac-selected strain (GYBT) of Helicoverpa armigera. The r1 allele of Ha_BtR was introgressed into a susceptible SCD strain by crossing the GYBT strain to the SCD strain, followed by repeated backcrossing to the SCD strain and molecular marker assisted family selection. The introgressed strain (designated as SCD-r1, carrying homozygous r1 allele) obtained 438-fold resistance to Cry1Ac, >41-fold resistance to Cry1Aa and 31-fold resistance Cry1Ab compared with the SCD strain; however, there was no significant difference in susceptibility to Cry2Aa between the integrated and parent strains. It confirms that the loss of function mutation of Ha_BtR alone can confer medium to high levels of resistance to the three Cry1A toxins in H. armigera. Reciprocal crosses between the SCD and SCD-r1 strains showed that resistance to Cry1Ac in the SCD-r1 strain was completely recessive. Life tables of the SCD and SCD-r1 strains on artificial diet in the laboratory were constructed, and results showed that the net replacement rate (R0) did not differ between the strains. The toxicity of two chemical insecticides, fenvalerate and monocrotophos, against the SCD-r1 strain was not significantly different from that to the SCD strain. However, larval development time of the SCD-r1 strain was significantly longer than that of the SCD strain, indicating a fitness cost of slower larval growth is associated with Ha_BtR disruption in H. armigera.

Predictions of survivorship are critical to quantify the probability of establishment by an alien invasive species, but survival curves rarely distinguish between the effects of temperature on development versus senescence. We report chronological and physiological age-based survival curves for a potentially invasive noctuid, recently described as Copitarsia corruda Pogue & Simmons, collected from Peru and reared on asparagus at six constant temperatures between 9.7 and 34.5°C. Copitarsia spp. are not known to occur in the United States but are routinely intercepted at ports of entry. Chronological age survival curves differ significantly among temperatures. Survivorship at early age after hatch is greatest at lower temperatures and declines as temperature increases. Mean longevity was 220 (±13 SEM) days at 9.7°C. Physiological age survival curves constructed with developmental base temperature (7.2°C) did not correspond to those constructed with a senescence base temperature (5.9°C). A single degree day survival curve with an appropriate temperature threshold based on senescence adequately describes survivorship under non-stress temperature conditions (5.9–24.9°C).

We tested the hypothesis that host tree age may act as a selective factor and lead to local adaptation of the hemlock looper (Lambdina fiscellaria), a geometrid Lepidoptera that has a wide geographical distribution and has evolved in different eco-zones characterized by different levels of floristic composition, age structure and fragmentation level. Considering that hemlock looper outbreaks mainly occurred in old forests, we compared the biological performances of two populations. The first population was collected in the northern virgin boreal forest, which is dominated by mature and overmature coniferous stands that have not suffered from human disturbance. The other population was collected in the southern mixed-wood forest, which is more diversified and has been modified by forest harvesting. Larvae were reared under controlled conditions on foliage from three age classes of balsam fir trees: juvenile, mature and overmature. Although we measured significant variations of biological performances between the two populations, no significant effect of the age of the balsam fir trees could be detected for males from both populations or for females from the southern population. However, northern females were strongly affected by the age of balsam fir trees on which they fed, as their pupal weight was 12% higher and their fecundity increased by 27% on overmature trees compared with juvenile ones. These results indicate that under the same selective pressure, females adapt their strategy to maximize their fitness, and thus they appear as the driving force of evolution through the local adaptation concept. Furthermore, the two populations evolved in distinct habitats and their adaptation reflects selective pressures occurring inside their original environment. This is the first report on local adaptation of an herbivore that is mediated by host tree age. Changes in forest age structure may have a considerable impact on insect local adaptation and presumably on their population dynamics.