Insect–plant interactions among gall size determinants of adelgids

1. The adelgid Adelges japonicus induces shoot galls in ezo‐spruce Picea jezoensis. The relative importance of shoot characteristics and the number of adelgids that induce a gall in determining resource availability on a shoot was examined. Gall volume, which is highly correlated with the number of larvae that mature in a gall, was used as a measure of resources available for adelgids on a shoot. Gall volume was partitioned into two components: the number of chambers and the chamber volume. 2. Path analysis revealed that the number of chambers in a gall was not affected by the number of leaves (which is an indirect measure of shoot length), the whorl where the gall was attached, or the number of surrounding shoots arising from the 1‐year‐old parent shoot. Each of these shoot characteristics, however, affected chamber volume: chamber volume increased with increasing numbers of leaves and surrounding shoots and with increasing heights of the whorls. 3. Two generations of adelgids contributed to gall enlargement in different ways: the number of gall founders affected the number of chambers, and the number of gall inhabitants affected chamber volume. 4. In a path diagram that incorporated the effects of both shoot characteristics and the numbers of adelgids, the numbers of adelgids in the two generations together accounted for 30% of the variance in gall volume, whereas the three shoot characteristics explained only 3.7% of the variance. This suggests that the amount of resources available in a gall is affected mainly by the numbers of adelgids but only slightly by shoot characteristics.     

Intra‐plant heterogeneity influences the preference and performance of juveniles and adults of a gall midge

Abstract 1. Field studies were conducted to evaluate the preference and performance of a gall‐inducing midge (Harmandia tremulae) within the crown of trembling aspen (Populus tremuloides). Females did not select oviposition sites preferentially within leaves, but did lay preferentially on young leaves. 2. Larvae were the only life stage involved in gall site selection within leaves and in gall initiation and development. Gall size, which was positively related to survival, was highest for galls on mid veins that were located close to the petiole. However, one‐third of galls were located on lateral veins and most galls were not adjacent to the petiole, indicating that many larvae choose sub‐optimal gall initiation sites. 3. Gall density was positively associated with leaf length, and leaf length, was positively associated with gall size. However, gall density per leaf was not related to larval survival in galls. This latter result may be a result of an observed inverse relationship between gall size and gall density for similar‐sized leaves. 4. The results partially support the plant vigour and optimal plant module size hypotheses, which predict that galler fitness in successfully induced galls should be highest on large, fast‐growing plant modules. The lack of a strong preference‐performance link supports the confusion hypothesis, which predicts that oviposition and gall site selection may often be suboptimal in systems where galler lifespan is short. This study suggests that small‐scale variations in plant quality within leaves, can render gall site selection by juveniles as important as that previously reported for adult females.     

Oxycephalomyia, gen. nov., and life history strategy of O. styraci comb. nov. (Diptera: Cecidomyiidae) on Styrax japonicus (Styracaceae)

A new genus Oxycephalomyia is described to contain the gall midge that was previously known as Asteralobia styraci (Shinji). Oxycephalomyia styraci, comb. nov., produces leaf vein galls on Styrax japonicus (Styracaceae). The adult of O. styraci is redescribed, and its full‐grown larva and pupa are described for the first time. The annual life cycle of the gall midge in northern Kyushu was clarified; the first instars overwinter in the galls on the host plant. However, the galls of O. styraci mature much later in the season than those of other gall midges with a similar life history pattern, and the durations of second and third larval instars are remarkably short. Such a life history pattern is considered to have an adaptive significance in avoiding larval parasitism, particularly by early attackers. The number of host axillary buds as oviposition sites decreased in bearing years and increased in off years, but there was no sign of oviposition site shortage even in bearing years, probably due to the low population density of the gall midge. An unidentified lepidopteran that feeds on galled and ungalled host buds and a Torymus sp. that attacks pupae of O. styraci were recognized as mortality factors of the gall midge.     

Small‐scale resource tracking in a population of a long‐lived insect

How plant‐feeding insects distribute themselves and utilize their host plant resources is still poorly understood. Several processes may be involved, and their relative roles may vary with the spatial scale considered. Herein, we investigate small‐scale patterns, namely how population density of a gall midge is affected by individual growth form, phenology, and microsite characteristics of its herb host. The long‐lived plant individuals vary much with regard to number of shoots, flower abundance, and flowering phenology. This variation is connected to site characteristics, primarily the degree of sun exposure. The monophagous insect galls the flowers of the host plant – an easily defined food resource. It is a poor disperser, but very long‐lived; diapausing larvae can stay in the soil for many years. Galls were censused on individual plants during 5 years; from a peak to a low in gall population density. Only a very small fraction of the flowers produced (<0.5%) were galled even in the peak year. Nevertheless, most plant individuals had galls at least 1 year. In a stepwise multiple regression, plant size (number of shoots) was found to be the most important predictor of gall density (galls/flower). However, gall density decreased more than one order of magnitude over the plant size range observed. There was also a weak effect of plant phenology. Early flowering plants had lower gall densities than those starting later. Sun exposure had no direct effect on gall density, but a path analysis revealed indirect effects via the timing of flowering. Gall population change was highly synchronous in different parts of the study area with no significant decrease in synchrony with distance.     

Nutrition as a facilitator of host‐race formation: the shift of a stem‐boring beetle to a gall host

1. The importance of host‐race formation to herbivorous insect diversity depends on the likelihood that successful populations can be established on a new plant host. A previously unexplored ecological aid to success on a novel host is better nutritional quality. The role of nutrition was examined in the shift of the stem‐boring beetle Mordellistena convicta to fly‐induced galls on goldenrod and the establishment there of a genetically distinct gall host race. 2. First, larvae of the host race inhabiting stems of Solidago gigantea were transplanted into stems and galls of greenhouse‐grown S. gigantea plants. At the end of larval development, the mean mass of larvae transplanted to galls was significantly greater than the mass of larvae transplanted to stems, indicating a likely nutritional benefit during the shift. This advantage was slightly but significantly diminished when the gall‐inducing fly feeding at the centre of the gall died early in the season. Additionally, there was a suggestion of a trade‐off in the increased mortality of smaller beetle larvae transplanted into galls. 3. In a companion experiment, S. gigantea gall‐race beetle larvae were likewise transplanted to S. gigantea stems and galls. Besides the expected greater mass in galls, the larvae also exhibited adaptations to the gall nutritional environment: larger inherent size, altered tunnelling behaviour, and no diminution of mass pursuant to gall‐inducer mortality. 4. In a third line of inquiry, chemical analyses of field‐collected S. gigantea plants revealed higher levels of mineral elements important to insect nutrition in galls as compared with stems.     

Gall distribution as a compromise between the optimal gall-site selection and the synchrony to host-plant phenology in the aphid Kaltenbachiella japonica

Gall-site selection by the aphid Kaltenbachiella japonica was evaluated in relation to leaf position in a shoot, and gall positions within a leaf. First-instar fundatrices induce closed galls on the midribs of host leaves, and several galls were often induced on one leaf. Leaves with many galls were often withered before emergence of sexuparae from the galls. Within a leaf, gall volume was positively correlated with the sum of lateral-vein length in the leaf segment at which the gall was induced. The observed pattern in gall volume among the leaf segments corresponded with that in the lateral-vein length. These results show that a foundatrix selects the most vigorous position within a leaf to produce more offspring. Although distal leaves grew faster than did basal leaves, gall density was highest on leaves at the middle order when a shoot has more than seven leaves. Optimal gall-site selection seems to be constrained by the asynchrony in timing between the hatching of fundatrices and leaf growth within a shoot. These results suggest that the observed gall distribution is affected by both the distribution of suitable galling sites within a leaf and the synchrony with leaf phenology of the host plant.     

Factors affecting refuge from parasitoid attack in a cynipid wasp, Aphelonyx glanduliferae

We examined seasonal patterns of gall morphology, growth, and survivorship of the agamic generation of a cynipid wasp, Aphelonyx glanduliferae, and discussed its mortality factors, especially from the point of view of refuge from parasitoid attack. Although the initiation period varied greatly among individual galls, the larvae of A. glanduliferae grew rapidly and reached their maximum size within 3 weeks before pupating in late September to early October. This growth period corresponded to the period when the gall walls became thinner. Parasitoid attack, which was the principal factor in the mortality of A. glanduliferae in the tree crown, was concentrated around the pupation period of the cynipid. Gall walls were significantly thinner in galls attacked by parasitoids than in those still containing a living cynipid. Therefore, the period available to parasitoids seems to be limited by both gall wall thickness and cynipid size. Thus, the growth pattern of A. glanduliferae larvae can have significance in that it narrows the window of vulnerability to parasitoids to a particular period. Although delaying gall initiation will also shorten the exposure period to parasitoid attacks, it was likely to increase the risk of death from gall abortion caused by seasonal degradation in the quality of host plant tissues. Although many cynipids were killed by disease in the galls that fell to the ground, the falling of mature galls to the ground may be another way to a parasitoid-free space. It is thus suggested that a trade-off among life history traits against multiple factors operates in the refuge of A. glanduliferae from parasitoid attack. Received: May 15, 2001 / Accepted: February 1, 2002     

Host plant response and phenotypic plasticity of a galling weevil (Collabismus clitellae: Curculionidae)

The response of a host plant to gall‐inducing insects varies both among and within plants, so that different levels of resources are available to the insects. The weevil Collabismus clitellae Boheman induces galls on the shoots of Solanum lycocarpum St Hil. in south‐east Brazil. Galls are found on a range of parts within an individual plant and are more abundant on smaller plants. In the present study, the host plant response as a possible influence on the performance of C. clitellae both between and within plants was tested. Gall abortion increased with plant height. Within plants, gall size was positively related to shoot diameter and number of chambers within the gall. The increase in gall larval density (number of individuals per gall volume unit) resulted in smaller adults and reduced developmental rates, probably because of resource limitation within the gall. The number of eggs laid by females increased with shoot diameter. Females laid more eggs on thicker shoots, where there are fewer chances to form galls with high larval density. However, this relationship was weak and a large variation was found for adult sizes. The availability of high quality sites is limited to smaller plants and thicker shoots located on the basal region of the plant. The phenotypic plasticity of this insect species in adult size and development time allows individuals growing on low quality sites to reach maturity, thus enhancing exploitation of the host plant.     

Post‐gall induction performance of Adelges Abietis (L.) (Homoptera: Adelgidae) is influenced by clone,shoot length,and density of colonising gallicolae

1. We evaluated the effect of clone (one susceptible and one resistant clone), shoot length, crown level, and gallicola density on post‐gall induction performance of Adelges abietis. Galls that had been successfully induced by one fundatrix on a range of shoot sizes were selected, and the number of gallicolae that could colonise the gall was manipulated. 2. Post‐induction gall development success was inversely related to shoot length and was higher on the susceptible clone than on the resistant clone. As gallicola density did not influence the proportion of galls that successfully completed development, reduced post‐induction gall development on large shoots was not likely to be result of an insufficient stimulus from gallicolae. 3. Clone was the only factor that significantly influenced gall volume and galls were larger on the susceptible clone than on the resistant clone. As gall volume did not increase when more gallicolae attempted to colonise a gall, competition within a gall increased. Gallicola survival was inversely related to the number of colonising gallicolae. Our results suggest that gall size may be limiting at natural densities. 4. Previous studies report positive relationships between gall induction success and fundatrix density, and between gall size and fundatrix density. As each fundatrix produces one egg mass of gallicolae, this study suggests that there may be a trade‐off between the successful induction of a large gall and subsequent survival of gallicolae. 5. In the present study, clone influenced all measures of post‐gall induction performance. Performance was always higher on the susceptible than on the resistant clone.     

Life history strategy and taxonomic position of gall midges (Diptera: Cecidomyiidae) inducing leaf galls on Styrax japonicus (Styracaceae)

Four gall midge species (Diptera: Cecidomyiidae) that induce leaf galls on Styrax japonicus (Styracaceae) were identified to generic level based on larval morphology. Three of these gall midges, which induce whitish hemiglobular galls, flattened subglobular galls, and purple globular galls, respectively, were identified as three genetically distinct species of Contarinia, and the remaining species, which induces globular galls with dense whitish hairs, was identified as a species of Dasineura. Field surveys in Fukuoka, Japan, revealed that adults of these gall midges emerged and oviposited in late March to mid‐April at Mount Tachibana (approximately 200 m a.s.l.) and in late April to early May at Mount Sefuri (about 1050 m a.s.l.), coinciding with the leaf‐opening season of S. japonicus. Larvae of these gall midges mostly developed into third instars by June and then left their galls and dropped to the ground. These species therefore have a life history strategy that differs from that of another S. japonicus‐associated gall midge, Oxycephalomyia styraci, which overwinters as the first instar in ovate swellings, matures rapidly in spring, and emerges directly from the galls.     

Optimal clutch size of the chestnut gall-wasp,Dryocosmus kuriphilus yasumatsu (Hymenoptera: Cynipidae)

Optimal clutch size of the chestnut gall-wasp, Dryocosmus kuriphilus Yasumatsu (Hymenoptera: Cynipidae), was examined in galls on wild and resistant chestnut trees in 1988 and 1989. The rate of escape success of newly-emerged adults from galls was an average of 60%, irrespective of cell numbers per gall. Dry mass per cell of a gall (as an index of nutritive condition) decreased with increasing cell number per gall, but was proportional to the mean number of mature eggs of new adults per gall. The number of cells per gall that occurred most frequently did not agree with that attained by the maximum survival rate from young larva to adult emergence of the gall-wasp. This discrepancy was examined from the viewpoint of three factors: 1) quality of offspring, 2) defensive response of the host plant causing mortality of the gall-wasp before cell formation, and 3) fitness per gall vs. fitness per egg. It is concluded that the third factor is most likely to be the one best in explaining the discrepancy.     

Host range and host preference of a flea weevil, Orchestes hustachei, parasitizing aphid galls

Although larvae of flea weevils (Curculioninae: Rhamphini) have been known to be leaf miners, larvae of the rhamphine weevil Orchestes hustachei have been found in aphid galls of four Tetraneura species on Ulmus davidiana and in galls of Paracolopha morissoni on Zelkova serrata. This study clarified the feeding habits of O. hustachei larvae and evaluated gall selection by ovipositing females to test the hypothesis of host race formation on their hosts, Tetraneura and Paracolopha galls. When weevil larvae were placed in half‐cut galls, they always fed on aphids rather than on gall tissue. When given gall tissue only, all larvae failed to reach adulthood. The number of aphids surviving in a parasitized gall decreased significantly with the development of the weevil larvae. These results suggest that O. hustachei larvae use aphids as their major food source. In the field, ovipositing females did not discern between four Tetraneura species on U. davidiana, in spite of a large difference in suitability as food. Paracolopha morrisoni was introduced into Hokkaido approximately 100 years ago, together with the host plant Z. serrata. It is probable that P. morrisoni has recently come to be used as a host by O. hustachei in Hokkaido. Host choice experiments using Tetraneura sp. O and P. morrisoni galls indicated that female weevils from Z. serrata preferred P. morrisoni to Tetraneura sp. O galls, while females from U. davidiana selected the two types of gall randomly. On Z. serrata, female weevils selected larger P. morrisoni galls, while on U. davidiana, females did not show a preference for gall size. These results suggest that a host shift to P. morrisoni galls may have led to an initial stage of host race formation between the weevil population using Tetraneura galls on U. davidiana, and that using P. morrisoni galls on Z. serrata.     

The problem of optimal clutch size in a tritrophic system: the oviposition strategy of the thistle gallfly Urophora cardui (Diptera,Tephritidae)

The problem of optimal clutch sizes is a central theme in life history theory. Optimal allocation of eggs is especially complicated for insects in tritrophic systems. In this study we analyze some of the processes determining clutch sizes of the thistle gallfly Urophora cardui, a monophagous tephritid fly associated with Cirsium arvense. U. cardui forms multilocular shoot galls, which vary broadly in their size and number of their gall cells. We investigate various fitness consequences of gall size. An analysis of the number of cells per gall (which is correlated with gall diameter and gall weight) showed that in U. cardui there is mutual facilitation rather than larval competition. Increasing numbers of larvae per gall led to a decreasing mortality and increasing larval weight. Larval weight in turn was positively correlated with the probability of survival to adulthood and with adult weight and fecundity. Thus, all fitness parameters measured favoured large galls. Clutch sizes in oviposition experiments were distinctly larger than the number of gall cells of field populations and in cage experiments, suggesting high mortality of eggs and/or early larval instars. There was a significant relationship between the internal structure (i.e., the size of the growing point) of the bud and clutch size, suggesting that U. cardui females are able to measure bud quality and adapt clutch sizes accordingly. Clutch size was positively correlated with the female's age at first oviposition and negatively with the number of previous ovipositions and previously laid eggs. Since the potential egg capacity per female is higher than the average number of larvae it is likely to produce during its short adult lifespan, U. cardui females tend to be time-limited rather than egglimited, which might favour large clutches once an appropriate oviposition site has been located. As the development of the gall and hence the fate of a clutch depends on a number of unpredictable factors, exclusive concentration of eggs in a few large clusters would involve risks which could be avoided by increasing the number of clutches. Therefore we interpret the high variation of clutch sizes in U. cardui as a mixed strategy of bet hedging and gambling.     

Abundance of Gall Midges on Poulsenia armata (Moraceae): Importance of Host Plant Size and Light Environment in Tropical Rain Forests1

We studied the effects of contrasting light environments on the relationship between the host plant size of Poulsenia armata and the abundance of two gall midges in a tropical rain forest in Veracruz, Mexico. The number and density of two gall morphs (i.e., laminar and vein‐petiole galls) were positively correlated with plant size only in trees found in the forest but not in gaps. The availability of foliar area of P. armata trees was greater in forest gaps than in the forest. The foliar area was positively correlated with the abundance of laminar galls in trees in forest sites, but not with vein‐petiole galls. We concluded that the abundance of two morphs of gall midges on P. armata was associated with host plant size only in the forest trees. Larger plants had more galls than small plants, although this relationship was affected by local light environments.     

Very long diapause and extreme resistance to population disturbance in a galling insect

1. Some insects have a prolonged diapause – a dormancy that extends over more than 1 year. In most species prolonged diapause involves one or a few extra years, but in extreme cases diapause may surpass 10 years. Few cases of very long diapause have been described, and very little is known about the population consequences of the temporal refuge formed by the diapausing individuals. 2. The gall midge Contarinia vincetoxici Kieffer galls the flowers of a long‐lived herb Vincetoxicum hirundinaria Med. After completing development, larvae leave the galls for the ground where they enter diapause. Extending an earlier published inoculation experiment, we show that the diapause may last up to at least 13 years, with a median duration of at least 6 years. 3. The gall midge is attacked by two parasitoid species. Dissections of gall midge larvae for presence of parasitoids revealed that Omphale salicis Haliday had a maximum 2 year diapause and Synopeas acuminatus Kieffer a maximum 4 years. The very long diapause of the gall midge may thus provide a temporal refuge from these enemies. 4. In a 15‐year field experiment all galls were removed every year from six isolated habitat patches. Density changes in experimental populations were not statistically different from control populations for over a decade. After 14–15 years a modest decline could be observed. This slow response illustrates that prolonged diapause in C. vincetoxici provides a very strong population buffer against mortality during the galling stage.     

The chemical composition of plant galls: are levels of nutrients and secondary compounds controlled by the gall-former?

The chemical composition of galled and ungalled plant tissue was compared in a series of experiments. Gall and adjacent plant tissue was analysed for 20 species of gall-former on 11 different plant species. There were clear differences between galled and ungalled tissue in levels of nutrients and secondary compounds. Gall tissue generally contained lower levels of nitrogen and higher levels of phenolic compounds than ungalled plant tissue. The gall tissue produced by the same plant in response to different species of gall-former differed in chemical composition, as did the gall-tissue from young and mature galls of the same species. The chemical differences between gall and plant tissues were studied in more detail in two field manipulations. Firstly, the seasonal changes in phenolic biosynthesis in Pontania proxima and P. pedunculi (Hymenoptera: Tenthredinidae) gall tissue were compared to those of their host plants, Salix alba and S. caprea. In both types of gall tissue, phenolic levels declined as the season progressed, but levels in the surrounding plant tissue increased. When the gall insects were killed with insecticide, phenolic levels in the galled tissue dropped to the same level as those in adjacent plant tissue. Secondly, the density of Cynips divisa (Hymenoptera: Cynipidae) galls on Quercus robur leaves was reduced by removing half the galls present, either those from the central region of the leaf or those from the edge. Decreasing gall density increased the size of the remaining galls and the weight of the insects, but these effects were most marked when the galls remaining were growing centrally on the leaf, i.e. when the galls from the edge had been removed. Decreasing gall density increased the nitrogen content of the remaining galls, again to a greater extent in galls growing centrally on the leaf. The results of these studies suggest that the levels of nutrients and secondary compounds in gall tissue are usually markedly different to those of surrounding plant tissue, and that gall-formers may produce species-specific and temporally variable changes in the chemical composition of gall tissue. Received: 7 July 1997 / Accepted: 29 September 1997     

Bottom-up trophic cascades caused by moose browsing on a natural enemy of a galling insect on balsam fir

To elucidate the impact of browsing by moose on the indigenous insect community, we investigated the abundance of a specialist insect herbivore (galler; Paradiplosis tumifex) and its natural enemy (inquiline; Dasineura balsamicola) on balsam fir saplings in the field using exclosures in Newfoundland, Canada. The number of current-year shoots on each sapling, but not the density of balsam fir saplings, was significantly higher inside exclosures, where moose had been excluded for 9–12 years, than outside exclosures. Current-year shoots were longer and contained more, but smaller, needles inside than outside exclosures. Moose browsing had a negative impact on the abundance of both the galler and its natural enemy. Gall density was 85.6% and 60.4% lower in 2006 and 2007, respectively, outside than inside exclosures. Similarly, the attack rate of D. balsamicola was 76.5% lower outside than inside exclosures in 2007. The apparent survival rate of P. tumifex larvae was similar outside and inside exclosures due to higher rates of unexplained mortality in galls outside exclosures. Survival of P. tumifex larvae was positively related to shoot length and inversely related to gall density. Only gall density was negatively related to D. balsamicola survival. Our results demonstrate that a large mammal can cause strong bottom-up trophic cascades.     

Larval distributions of the ectoparasitoid wasp Eurytoma robusta relative to the host tephritid gall fly Urophora cardui

Parasitism may explain the patchy distributions of host populations. The present paper is a study of larval distributions of the parasitoid Eurytoma robusta in galls of the tephritid gall fly Urophora cardui. It focuses on E. robusta's choice of U. cardui gall and whether this changes relative to the rate of parasitism. Oviposition patterns were inferred by direct counts of larvae in galls and genetically, for both species, using indirect relatedness estimates between gall‐members. Furthermore, rates of parasitism in four populations were monitored for 4 years. The modal distribution of E. robusta larvae per gall was one and independent of the level of parasitism. The mean number of E. robusta per gall did not differ from Poisson distributions at different parasitism rates. We were not able to demonstrate a parasitoid preference for gall size. In contrast, parasitoids may have a negative effect on gall growth. Relatedness estimates showed that E. robusta gall members were often unrelated, whereas U. cardui were siblings. Thus, larval distributions of E. robusta suggest that oviposition behaviour is generally constrained and density independent. In four populations monitored over 4 years, parasitism was initially high (up to 70%), but suddenly declined with no apparent effect on fly (gall) abundance.     

Life history strategy and adult and larval behavior of Macrodiplosis selenis (Diptera: Cecidomyiidae), a species that induces leaf‐margin fold galls on deciduous Quercus (Fagaceae)

Life historical, behavioral and ecological traits of Macrodiplosis selenis, which induces leaf‐margin fold galls on Quercus serrata, Q. mongolica and Q. dentata (Fagaceae) in Japan and South Korea, were studied. Daily activity and larval development indicate that M. selenis is a diurnal and univoltine gall midge. In April, females lay their eggs both on upper and under surfaces of fresh leaves. The duration of the egg stage varies from 5 to 9 days, depending on daily temperatures. Hatched larvae crawl to the upper surface of the leaf margin, where they start to induce galls. Larvae become full‐grown in October, drop to the ground in November and overwinter in cocoons on the ground, while larvae of congeners mature in May and drop to the ground in June. A relatively long period of the second larval stadium from July to October on the host trees seems to be effective for M. selenis in avoiding summer mortalities caused by predation and aridity on the ground and by ectoparasitoids that attack mature larvae or pupae on the host leaves. The spatial distribution pattern of M. selenis leaf galls is contagious and the mean gall density per leaf is significantly correlated with the mean crowding. This study adds new insights of life history strategy and adult and larval behavioral pattern to the ecological knowledge of gall midges, and these kinds of information are essential for further studies of M. selenis population dynamics and interactions with other Quercus‐associated herbivores.     

Density-dependent egg mortality in early stages of gall induction by the fruit gall midge <Emphasis Type="Italic">Asphondylia aucubae</Emphasis> Yukawa and Ohsaki

Density-dependent mortality has been considered a symptom of intraspecific competition. We examined the occurrence of such mortality in the early stages of gall induction by the gall midge Asphondylia aucubae Yukawa et Ohsaki (Diptera: Cecidomyiidae). Female midges deposit eggs into young fruit of the dioecious shrub Aucuba japonica Thunberg to induce gall formation. Each host fruit received 0–67 eggs (mean 18.5 eggs) from multiple females, whereas established galls each contained one to ten larvae. Midges suffered intense mortality (65–90%) at the egg stage. Egg mortality occurred even in fruit in which no larvae had hatched, suggesting that this mortality cannot be wholly attributed to larval interference. Egg mortality was affected by fruit size, i.e., resource capacity. Midges distributed more eggs in larger fruit. Egg mortality increased as the per-fruit density increased relative to fruit size. In contrast, the mortality of hatchlings was density-independent. Our results suggest that A. aucubae intensely compete for gall-induction substrates, which are spatiotemporally rare resources.