Study sites

The study was carried out in the Baïnem arboretum, located at 15 km west of Algiers. This arboretum was set up in 1956, in the center of the Baïnem forest, where it occupied 50 ha of the 500 ha of this forest (longitude 2° 57′ 20″ to 2° 57′ 57″ and latitude 36° 47′ 16″ to 36° 48′ 06″). The soil was clay on schist with many coarse elements. The altitude varies from 195 to 215 meters, with an average of 200 meters. The climate is of the Mediterranean type, sub-wet with hot summers characterized by dry and a hot period extending from May to September and a rainy period from October to May. Three cypress species were examined: C. sempervirens, C. arizonica arizonica, and C. arizonica glabra.

The C. arizonica arizonica stand was started in 1959 on a declivity from 20 to 30 % with geographical coordinates: 2° 57′ 46″ of longitude and 36° 48′ of latitude, with 132 trees and the general exposure is Northwest.

The C. arizonica glabra stand was started on March 31 1959 in the Southern part of the forest of Baïnem (longitude 2° 57′ 30″ and latitude 36° 47′ 35″) with a total of 86 trees, on a slope of 5%.

The C. sempervirens artificial stand (45 trees) was situated on the arboretum Bert trench in the Northern part of the Baïnem forest (longitude 2° 57′ 36″ and latitude 36° 47′ 49″). A natural C. sempervirens stand appeared in the southeast part of Baïnem forest after the disappearance of the bald cypress, Taxodium disticum. It consisted of 67 trees on a slope of 5 to 10 % with a Northwest exposure (longitude 2° 57′ 40″ and latitude 36° 48′ 2″). A hedge of C. sempervirens that separated the Pinus and Eucalyptus stands was also included in this study. They grew on a 5 to 10% slope exposed to the Southeast.

In March 1995 20 trees bearing seed cones, yet unpollinated, were selected throughout the stand of C. sempervirens. Three seed-cone-bearing branches were tagged on each tree. The position of the cones on each branche was mapped. This cohort of seeds cones was monitored periodically until seeds were mature, in June 1996. During this period the sample branches were examined eight times (three each in 1995 and 1996, and twice in 1997), at approximately four month intervals. During each visit, the possible cause of injury or death was recorded. A sample of the damaged cones was collected at each visit and dissected to confirm the cause of mortality.

Species richness and damage assessment

Standardized cone collections were carried out every 15 days between December 1994 and August 1996 in the 4 stands in the Baïnem Arboretum. In each stand, two 40 cm-branches (one from the lower crown and another from the upper-crown) were selected randomly from each of 10 trees. First, the branches were beaten immediately over a net to collect the insects present on the cone surface as well as from the surrounding foliage. All the 1-year-old cones (i.e., green cones that are in the growth process), 2-year-old cones (i.e., ash-grey cones that completed the growth period the year before, and that will release seeds the following winter), and 3-year-old cones (i.e., overmature cone) were counted, and removed from the branch. Characteristics of the stand (climate, exposure) and of the sampled trees (i.e., type of crown: horizontal vs. pyramidal) and tree position in the stand were also noted at the time of sampling

In the laboratory, the 1-year-old cones and half of the 2 and 3- year-old cones were dissected to look for internal insect damage and for the presence of larvae. For each dissected cone, the seeds were extracted, and each seed lot was individually irradiated with X-rays using a Faxitron-43855® apparatus (20 Kv, 3 mA, 4 min) and recorded using X-ray sensitive film (Kodak® “Industrex M”). The seed quality (i.e., insect-infected seeds, larval development) was assessed from the radiographic images.

Emergence period and sex-ratio of M. wachtli

The remainder of the 2 and 3-year-old cones were placed into rearing boxes (50 × 35) and stored in an outdoor insectary at Baïnem, Algiers. Adult insects were killed at emergence and identified to species. The emergence pattern of both sexes of the Megastigmus species was monitored daily during the 1998 autumnal period from 140, 2-year-old cones seed of C. sempervirens collected in natural stands before any emergence of the adults to investigate possible differences in the emergence period among sexes of the infected seeds. The sex ratio of the M. wachtli adults that emerged in September 1998 from seeds collected from C. sempervirens collected in natural stands during the summer of the same year was measured.

M. wachtli development

To check if the female M. wachtli can develop parthenogenetically, 48 unfertilized females that emerged in September and October 1995 from individual infested seeds held in the laboratory, were fed honey water, and then placed into tulle bags that were attached to a branch containing cones in 4 trees, 2 bags per tree. Six females were put in each bag to observe if they produced progeny, and their gender. For comparison, 36 mating pairs of M. wachtli were introduced into 12 bags placed in 6 trees (2 bags per tree). Three pairs were put per trap to observe the emergence of progeny the following year.

Longevity of adult M. wachtli was determined under laboratory conditions and corroborated with periodic field observations. In the field, longevity of individuals was studied by following 904 adults emerged from seed of 2-year-old C. sempervirens cones in September 1995 and placed on the day of emergence in groups of identical emergence date into rearing boxes stored in an outdoor insectary at Baïnem and fed honey water. In a laboratory experiment, longevity of 279 adults of M. wachtli was studied during April and May 1996. The adults emerged from seed from 2-year-old C. sempervirens cones.

Egg development in M. wachtli and P. tessulatana

The development of eggs was examined by dissection of ovaries of thirty M. wachtli females that emerged in September 1998 from seeds of 2-year-old cones of C. sempervirens. These females were placed on the first day of emergence in groups of identical emergence date into rearing boxes stored in an outdoor insectary at Baïnem and fed honey water. The chalcids were dissected by removing the complete ovary. The ovarioles were transferred to a slide, placed in Bouins fixative, and the eggs were counted. Microscopic observations were made on the oocytes of thirty females 1 to 15 days old. Development of the oocytes was recorded as pre-vitellogenic and vitellogenic oocytes, or fully developed eggs.

Egg development in P. tessulatana was determined as described above for M. wachtli. Microscopic observations of were made on the ovaries of twenty females 1 to 16 days old.

Relationship between M. wachtli and P. tessulatana

The relationship between colonization by a cypress seed chalcid and the tortricid was characterized by their presence/absence ratio in 1000 colonized 2-year-old cones of C. sempervirens. 10 cones were collected from each of 100 trees. Cones attacked by M. wachtli and P. tessulatana were identified on the basis of their external appearance, as they dried and opened precociously when compared to cones containing healthy seed. For each attacked cone, the presence or the absence of attack of the chalcid and the tortricid was noted to see whether M. wachtli prefers, rejects or is indifferent to the previous attack of the cones by P. tessulatana.

Distribution of the emergence holes of M. wachtli and P. tessulatana

To determine the relative distribution of M. wachtli and P. tessulatana damage to the cones of C. sempervirens and C. arizonica, external damage and the position of the adult's emergence holes on the cones was noted and drawn. 244 cones infested by M. wachtli and 62 cones infested by P. tessulatana were collected from 80 C. sempervirens trees. Cones having weak and strong M. wachtli damage were examined separately. 30 cones infested by M. wachtli and 60 by P. tessulatana were collected from 50 C. arizonica trees and examined.

Damage by Caruslaspis minima Targioni (Homoptera, Diaspididae), to seed cones of C. arizonica glabra

In December 1994, at C. arizonica glabra stand, 2 branches bearing 1 and 2-year-old cones were randomly chosen from each of 10 flowering trees selected randomly. A total of 62 sound cones and 20 infected-cones by C. striata were collected. At the laboratory, the 1 and 2-year-old cones were dissected and the seeds were extracted. All seed lots were individually irradiated with X-rays to evaluate the respective number of normal, C. minima-infested, and empty seeds per cone.

Insect damage to cones of Cupressus dupreziana

513 cones were selected from 2-and 3- year-old cones of C. dupreziana that had been collected in April 1995. 496 cones were selected from 2-and 3-year-old cones collected in June 1995 in Tassili. Thirty surviving cones, those in their third year of development, were harvested just before the natural seed dispersal occurred and the seeds were removed. A total of 358 seeds were individually irradiated with X-rays to evaluate the respective number of filled, insects-damaged and empty seeds per cone.

Data treatment

The percentage of cones colonized by Megastigmus were compared among tree, stand, and species using analysis of variance (Anova, Statsoft Statistica/W package). ANOVA was used to compare the percentage of M. wachtli and P. tessulatana emergence holes on the colonized cones, and to test localization of the colonization per cone shape, insect pests and cypress species. The percentages data were transformed by arcsin √ p to achieve homogeneity of the variance before statistical analysis. ANOVA was followed by Tukey's test to look for differences between locations and species. The results of the longevity of M. wachtli were analyzed by the χ² test. The relationship between the colonization by the seed chalcid and by the tortricids was subjected to a Pearson correlation test (R).

Entomofauna and description of insect damage on cypress seed cones

Seven insect species and one diaspine cochineal were observed to attack the cones and seeds of Cupressus in the Baïnem arboretum during 1995 and 1996. According to the feeding habits defined by Turgeon et al. (1994), two species are conophagous (i.e., feed on cone tissues only), one is conospermatophagous (i.e., feed on both cone tissues and seeds), and three species are spermatophagous (i.e., feed on seeds only), the diaspine cochineal is opophagous (i.e., suck the sap contents of the scales) (Table 1). The three cypress species studied, C. arizonica arizonica, C. arizonica glabra and C. sempervirens, were attacked by four main pests: O. maculatus, O. depressus, M. wachtli and P. tessulatana. Two other species, Brachyacma oxycedrella and Nanodiscus transversus were also found on C. sempervirens.

Two of these species, the spermatophagous M. wachtli and the conospermatophagous P. tessulatana, had already been observed on the cone and seeds of C. sempervirens and C. macrocarpa in the Meurdja arboretum and Chrea National Park (Bouaziz 1993; Bouaziz and Chakali 1997). We found during two years of observation that M. wachtli emerged from mature cones and laid eggs in seeds of green fully-grown cones that had just begun the seed maturation phase. The conospermatophagous P. tessulatana presented two generations per year with a third partial generation. The same observation was made by Guido et al. (1995) in Italy.

The conospermatophagous P. tessulatana attacked green cones during the first year of development. Three spermatophagous pests, M. wachtli, O. maculatus and O. depressus attacked maturing and mature seeds inside 2-and 3-year-old cones (Table 1). The phenological relationships between cone development and pest attacks from bud burst to seed dispersal were similar to those reported by Guido et al. 1995.

Another species, Caruslaspis minima Targioni (Homoptera, Diaspididae), was abundant on the cypress cones studied. These insects were unusual. They were motionless and live in colonies, fixed on the various aerial parts (sheets, fruits, branches and trunk). They show an important sexual dimorphism. The females which were, in general, the most abundant forms and most recognizable in nature, were protected by a shield. This diaspine cochineal caused yellowing of the attacked parts where sap was sucked from the plant, but otherwise did not impact the seeds.

Table 1.

List of pest insects of the cypress cones indexed in the Baïnem arboretum.

The total number of seeds in sound cones of C. arizonica glabra was 119.450 ± 76.762 (Mean ± SD) (n = 62) and the total number of seeds in cones infested by C. minima was 135.452 ± 96.982 (n = 20). Analysis of the x-ray radiographic image of the overall seed yield of C. arizonica glabra cones collected at maturity revealed that an average percentage of 77% of the seeds were filled in sound cones, and 23 % of the seeds were empty and aborted. X-ray analyses of the seed content of C. minima cones showed that 89.4 % of seed were filled, and 10.6% of the seeds were empty and aborted. The percentages of normal, aborted and empty seeds per cone were not significantly different among sound and damaged C. minima cones (ANOVA: F 1.80 = 2.701, P = 0.104).

Symptoms of insect attack were observed on the cones of C. dupreziana. The percentage of infestation of 2-year-old cones varied from 46.77% to 52.24%. However, it was not possible to identify the insects that were responsible for this damage. The cones were attacked by Lepidoptera whose exit holes shapes were irregular and of a size identical to those of P. tessulatana. The X-ray analyses of seeds extracted from 30 attacked cones showed that only 2.5% of the seeds were normal. C. dupreziana did not manage to ensure its regeneration. The attack of the pest seed cones, presumably P. tessulatana, seemed to be one of the causes.

Other factors causing mortality

Once seed cone buds have been formed, their development can be stopped by several factors such as climatic conditions that can cause abortion either of buds before they burst open, or of immature seed cones. In most cases, abortion of seed cones was the result of poor or inadequate pollination.

Tagged cones were attacked by several fungi, the most important species of which were Seiridium (=Coryneum) cardinale (Wag.), also responsible for disease known as cypress canker, and Pestalotia (=Pestalotiopsis) funerea Desm. The fungus attack was either limited to cone scales or it extended to the whole cone. The fungi could easily penetrate the seed coat, on which they produced abundant spores (Bouaziz 2003).

Biology of M. wachtli

The female reproductive system of M. wachtli was composed of two ovaries, each with six ovarioles. At emergence, the ovarioles were not differentiated. The female abdominal content was essentially fatty tissue that regressed after the second day of feeding. The genital apparatus did not differentiate until after the sixth day of feeding when the genital apparatus formed. From the 12th day of feeding, the presence of one egg was observed in vitellogenesis. It thus seemed that this species could start to lay eggs only after the twelfth day (Figure 1). The average fecundity of a female under our rearing conditions was 7.8 ± 6.4 eggs (n = 20) (±SD). This observation differed from that made on other Megastigmus species, such as M. spermotrophus, where the females were able to lay eggs shortly after emergence (Hussey 1955; Roux 1995).

Figure 1.

Percentage of eggs at different stages in the ovaries of female cypress seed chalcids, Megastigmus wachtli, from thirty females 1 to 15 days old, collected in September of 1995.

When virgin females were placed in bags in the absence of males they produced 45 males and 0 females in October 1995. In 1996 the fertilized females of the same origin laid eggs producing 38 females and 37 males (Table 2). These results support the evidence that M. wachtli is capable of arrhenotokous parthenogenesis as are other species of Megastigmus (Roux 1995).

Figure 2.

Emergence of males and females of Megastigmus wachtli from seeds of Cupressus sempervirens collected in autumn of 1998 (number of collected cones = 140) in a natural stand of Baïnem. Drawn according to Camus (1914).

In stands surveyed both years, the most abundant sex appeared to vary among years, with males becoming most abundant in Baïnem arboretum in 1996 and 1998 (Figure 3). The ratio of males to females was 1.6 in 1996 (n = 408) and 1.2 (n = 549) in 1998. The situation was completely different in 1993 and 1995 when the sex ratio was 0.7 in favor of females in 1993 in the Meurdja arboretum (n = 114) (Bouaziz 1993; Bouaziz and Chakali 1998) and 0.9 in 1995 (n = 1531). This sex ratio suggested that M. wachtli was capable of arrhenotokous parthenogenesis as are other species of Megastigmus. The same result was shown by Roques et al., (1998) for M. wachtli in natural and introduced trees in Greece where they obtained a sex ratio in favor of females in 14 sites, 5 sites in favor of the males, and 4 sites had a sex ratio of 1.1. Ben Jamaa and Roques (1997) found an average sex ratio of 0.8 in favor of the females.

Figure 3.

Sex-ratio of the Megastigmus wachtli adults that emerged during the summer 1993 from seeds collected during the spring of that year and the autumn of 1995. Data for 1996 and 1998 were from seeds collected during the summer of these years.

Radiography of the seeds allowed the examination of later development of larvae during cone maturation. Larval development proceeded entirely in a single seed. The larva initially developed at the expense of the cotyledons and it started to consume seed endosperm the following spring. Similar results were observed previously (Roques 1983; Roques et Raimbault 1995; Guido et al. 1995). Larvae of M. wachtli seemed able to develop in non-fertilized ovules as Guido et al. (1995) suggested, but we could not experimentally infest non-fertilized ovules in order to test this assumption (Bouaziz and Roques, unpublished results). Rouault et al. (2004) hypothesize that all species of Megastigmus associated with Pinaceae can oviposit in unfertilized ovules, whereas those exploiting Cupressaceae cannot, and thus oviposit only in already fully developed fertilized seeds. Infested megagametophytes of unpollinated ovules did not degenerate as would have been expected, but continued to develop (Aderkas et al., 2005). Niwa and Overhulser (1992) and Rappaport et al. (1993) showed that the M. spermotrophus larvae were able to lay eggs on non-fertilized ovules of the Pseudotsuga species and that the larva could complete its development there. The chalcid lays its eggs in seeds of Douglas fir (Pseudotsuga menziesii (Mirbel) Franco) before fertilization has taken place in the plant. Oviposition not only prevents the expected degeneration and death of unfertilized ovules, but it induces energy reserve accumulation (Aderkas et al., 2005). Following oviposition by M. spermotrophus in unpollinated megagametophytes, larvae hatched and began consuming the central zone. Eggs were laid in megagametophytes that were differentiating archegonia which would eventually house the plant eggs. (Aderkas et al., 2005). The torymid chalcid wasp was able to induce identical nuclear behavior in infested, unfertilized megagametophytes, as occurred in uninfested, fertilized megagametophytes (Aderkas et al., 2005).

Table 2.

Resulting emergence of males and females of the cypress seed chalcid, Megastigmus wachtli, from eggs laid in seeds of Cupressus sempervirens by unfertilized female M. wachtli.

The emergence of M. wachtli was spread out over 6 weeks from the beginning of September to mid-October (Figure 2). Emergence of males was earlier as 50% of males emerged before the 21^st of September compared to the 23^rd of September for 50% of females. These emergences were later than those observed in the natural sites in Greece and in France but earlier than those of Tunisia (Roques et al. 1998).

Figure 4.

Adults longevity of a cypress seed chalcid, Megastigmus wachtli, emerged from seed yield of 2-year-old cones Cupressus sempervirens in September (a) in field, placed into rearing boxes stored in an outdoor insectary at Baïnem and nourished with honey water; (b) breeding under laboratory conditions.

The females had an average lifespan significantly longer than that of males. 13.35% of males died at the end of 8 days and only 4.87% of them survived 15 days or more. In contrast the corresponding percentages for females were 2.31% died at the end of 8 days and 7.18% were still alive after 15 days, many of which survived up to 27 days. Given that eggs did not start to mature until day 12, these results suggested that the only these older females would lay eggs. In the field, preliminary observations showed that the lifespan could increase up to 32 days (Figure 4). Breeding in the laboratory showed that longevity could reach 50 days (Fig. 4). This suggests that emergence can occur over 6 weeks and the insects can live up to 32 days under natural conditions.

Table 3.

Cone crop abundance and pest damage to cypress seed cones of different ages collected between 1994 and 1996 in the Baïnem arboretum.

Insect damage to cones

Pest damage varied with stand, cypress species, and stages of cone development (Table 3). Damage resulted mostly from feeding by M. wachtli which dominated the pest complex in 5 of the stands, and destroyed up to 30% of 2-year-old cones and 44% of 3-year-old cones of C. sempervirens, 30% of 2-year-old cones of C. arizonica, and 29.5% of 2-year-old cones of C. glabra. In most of the stands, the average percentage of trees that were attacked by M. wachtli varied significantly. Some trees were free of heavy attacks whereas others had more than 80 % of their cones colonized as observed for the C. arizonica stand (F34, 228 = 6.057, P = 0.000), the artificial parcel of C. sempervirens (F22, 12 = 10.05, P = 0.000) and the hedge (F25, 15 = 9.92, P = 0.000). On the other hand, no significant difference in tree attack by chalcids was observed on trees of the C. sempervirens natural stand.

Table 4.

Regression analysis of cones attack by M. wachtli and P. tessulatana

Differences in damage caused by M. wachtli were also noted between sites and species. The C. sempervirens artificial stand (40.7%), and the C. arizonica glabra stand (28%), were significantly more infested than the C. arizonica stand, and the C. sempervirens natural stand and hedge (F 4, 131 = 4.02, P = 0.004). Tukey's test showed that the infestations in the C. arizonica stand (23%) and the C. sempervirens natural stand (2.8%) and hedge (21%) were not significantly different. These differences in infestation corresponded to insect population fluctuations (Da Ros et al. 1993), or to variation in cone production between sites (Roques 1983).

Conophagous pests infested only 3.5 % of the cones of C. sempervirens during the first year of development. These losses were due mainly to the first generation of P. tessulatana. Guido et al. 1997 reported similar results in Italy, where P. tessulatana infested only 10% of the cones. In natural and naturalized stands in Greece, damage resulted mostly from feeding by P. tessulatana larvae which dominated the pest complex in 13 of the stands, and destroyed up 98% of 1-year-old cones in Eleousa (Rhodos) (Roques et al. 1997). Attack by larvae of the first generation of the tortricid stopped the growth of seed cones; the cones dried up prematurely and usually dropped to the ground. Only the slightly attacked or healthy cones then remained on the tree. The limited decrease in the number of 2-year-old cones resulted mainly from feeding by P. tessulatana larvae of the second and third generation. Unlike what happened to 1-year-old cones, most of the damaged cones did not disappear from the branch, and seeds that were not damaged directly were able to reach maturity. However, the apparent limited impact of insect attack during the cone maturation phase is misleading because damage by spermatophagous insects such as Orsillus spp, which can be detected only by X-ray analysis, was not taken into consideration. No significant difference in attack by tortricids was observed between sites and species (F 4, 131 = 0.885, P = 0.475).

Figure 5.

Mean number of eggs at different stages in the reproductive apparatus female of cypress tortricids, Pseudococcyx tessulatana, on twenty females from of 1 to 16 days old, collected in April, 1995.

Reproduction by P. tessulatana

The female reproductive apparatus of P. tessulatana was composed of two ovaries, with four ovarioles in each. Previtellogenesis lasted less than two days. In general, mature eggs were present by the third day after emergence (Figure 5). Zocchi (1963) observed that P. tessulatana females laid eggs after mating. The length of the mature egg varied from 0.9 to 1 mm. These eggs were attached to the surface either isolated or in series. From the sixth day, egg laying seemed to have begun as the dissections showed that the ovarioles contained few eggs. Eggs can be laid for over 16 days (Fig. 6).

Figure 6.

Mean eggs number by female of Pseudococcyx tessulatana dissected several days after emergence.

Figure 7.

Distribution of the emergence holes of Megastigmus wachtli according to the degree of damage of the cones of Cupressus sempervirens. A: 1 to 9 emergence holes by cone. B: more than 10 emergence holes by cone. Drawn according to Camus (1914).

Figure 8.

Distribution of the emergence holes of Pseudococcyx tessulatana of the cones of Cupressus sempervirens.

Relationships between the seed chalcid and tortricid

The Pearson (R) correlation coefficient of sampling, which reflects the degree of linearity between the attacks of the two insects on the same cone, was negative (R = -0.72). The attacks of these two pests seemed to exclude each other (Table 2). In the event of a strong cone production, M. wachtli chose the cones not colonized by P. tessulatana.

Table 5.

Variance analysis of M. wachtli and P. tessulatana emergence holes repartition on the C. sempervirens lengthened shaped cones.

Distribution of the emergence holes of M. wachtli and P. tessulatana

The distribution of the emergence holes of M. wachtli on the colonized cones is illustrated in Figure 7. This distribution reflects the distribution of the insects inside the cone since larvae are unable to move to different parts of a cone.

In the event of low level damage, the majority of the emergence holes were found together on the level of two cone scales, the lowest and the medium ones situated only on one face of the cone number of holes = 2047, F9 2037 = 45.648, P = 0.000). The highest cone scales had a limited number of emergences holes. Thus, the damage was concentrated certain regions.

Table 6.

Variance analysis of M. wachtli and P. Tessulatana emergence holes repartition on the cones of C. arizonica arizonica and C. arizonica glabra.

In the event of heavy damage, M. wachtli emergence holes were observed on both faces of the cone. They were distributed on the two lowest scales and the medium scales, among which one was significantly more colonized (number of holes = 350, F9, 340 = 4.780, P = 0.000). The two sides of the lowest scales and the highest scales were significantly less colonized. This could mean that M. wachtli preferred to lay eggs on seeds that were near one another.

The long-shaped cones were attacked differently than the round cones. In the event of low-level attack, the majority of the attacks on long cones were located on two scales, one front lower medium (30.7%) and one on the side lower medium scales (17%) (number of holes = 226, F 11,216 = 4.625; P = 0.000) (Table 5). The preference areas of exit holes probably corresponded to the laying place also located on only one face of the cone. As discussed above, the attacks were concentrated in space. The lower and higher scales were significantly more attacked. In the event of heavy attack, no significant differences were observed in distribution of the holes of exit of M. wachtli between the scales. The insects thus randomly attacked the various scales of the cone contrary to what was observed in the round cones.

Attack by M. wachtli on C. arizonica cone scales was random (number of holes = 58, F 7, 48 = 1.310, P < 0.266) (Table 6). Attack by P. tessulatana on C. sempervirens was concentrated on the lower and medium scales located on the same face of the cone (number of holes = 620, F 9, 610 = 15.197, P = 0.000) (Figure 8). In comparison, attack on C. arizonica was condensed on three scales: one lower, one medium and one higher located on the same face of the cone (number of holes = 170, F 7, 160 = 5.51, P = 0.000) (Table 6). The majority of the damage to the C. arizonica glabra variety was centralized on two scales, a lower and a medium. The medium cone scale was significantly more colonized than the lower one (number of holes = 40, F 7, 30 = 28.24, P = 0.00). Thus, attack strategy by both the chalcid and tortricid varied according to the cypress species and location of attack on the cone.