Relative Susceptibility Among Arborvitae Cultivars to Arborvitae Leafminer, Argyresthia thuiella (Lepidoptera: Argyresthiidae)

Host Plant Characteristics

The results from this study suggest that canopy density appears to play a role in ovipositional preference by ALM. Short, globe-shaped plants with very dense or dense canopies, such as ‘Hetz Midget’ and ‘Hoopesii’, had very low infestation levels (<7% IBTs) compared to significantly higher infestation levels (10%–42% IBTs) for pyramidal- and columnar-shaped species and cultivars. In addition, a weak but significant inverse relationship exists between canopy density and IBTs.

While these two cultivars had lower levels of IBTs, it is not clear what factors may influence insect distribution within the tree canopy, including the insect’s mobility, interspecific and intraspecific competition, natural enemies, micro-climate, nutritional quality of foliage, and leaf production and leaf phenology of the host trees (Brown et al. 1997). The purpose of this study was to document the relative susceptibility of arborvitae species/cultivars to ALM. Identifying a specific factor or factors responsible for differences in feeding damage was determined to be beyond the scope of this study. However, the study authors would be remiss for not offering possible suggestions for observed differences in ALM feeding damage between arborvitae species/cultivars.

Numerous studies have revealed a variety of factors responsible for differences in ovipositional and feeding preference among insect herbivores, such as leafminers, including across natural light gradients. However, these studies have primarily focused on shaded plants or plants growing in full sun, not canopy density (Moore et al. 1988, and references therein; Bentz 2003; Connor 2006). To the best of the authors’ knowledge, no definitive studies have been conducted on the effect of canopy density on ovipositional preference, particularly on leafminers feeding on evergreen trees. Numerous studies of the effects of light-gradient on oviposition and host susceptibility are conflicting at best. Some herbivores are more abundant or damaging on plants in sunny areas, while others are more abundant or cause more damage in shaded environments. Bultman and Faeth (1988) showed that densities of two leafminer species feeding on Emory oak (Quercus emoryi) were negatively correlated with solar radiation, and they further concluded that even subtle differences in shading influences leafminer density and mortality. Further still, in some studies, no difference in feeding or abundance was found between sun and shade environments (Collinge and Louda 1988; Collinge and Louda 1989; Potter 1992; Connor 2006). It is possible, though not confirmed in this study, that light levels may impact ovipositional preference and susceptibility when comparing plants with very dense foliage compared to sparsely foliated plants. Dirr (2009) stated that short, globe-shaped arborvitae cultivars, such as ‘Hetz Midget’ and ‘Hoopesii’, tend to have shorter internodes and denser canopies compared to taller trees. Informal field measurements indicate short cultivars with very dense, thick foliage (e.g., ‘Hoopesii’) had approximately 50% less light penetration compared to tall cultivars with sparse foliage (e.g., ‘Smaragd’). In addition to diminished light penetration, short, very dense cultivars had thick foliage with branching extending approximately 8 cm into the canopy interior, while sparse cultivars had much thinner foliage and less branching, extending approximately 3 cm from the spray tip into the interior tree canopy.

Female leafminer ovipositional behavior may also be a factor. Connor (2006) indicated that daytime fliers discriminate between sun and shade foliage, while nighttime fliers do not. While not confirmed in this study, it might be inferred that female ALMs, being daytime fliers, may discriminate by preferring to oviposit on foliage receiving more sunlight. This might explain why there was significantly higher tip damage in the upper canopies of shorter cultivars, and middle and upper portions of taller cultivars.

Leaf thickness, toughness, water content, secondary compounds, and nutritional compounds have all been implicated at one time or another as possible factors responsible for various feeding patterns observed in insect herbivores (Connor 2006, and references therein). While none of these factors were directly measured in this study, field and laboratory visual observations at the time of IBT assessments indicated no obvious and/or discernable differences in leaf thickness for the arborvitae cultivars studied. In virtually all of the IBT assessments, only one larva was found per leaf tip, and only in very rare cases were two larvae discovered. This suggests there was adequate food (i.e., parenchyma tissue) for larval development for all species and cultivars in this study, regardless of canopy density. However, more formal laboratory and field studies would be required to support these findings.

The potential impact of very dense foliage versus sparse foliage on natural enemy habitat was not included in this study. Predation and parasitism are inconsistent at best. Some studies have shown higher rates of arthropod predation and parasitism in shaded areas and by canopy aspect (Trumbule and Denno 1995; Pottinger and LeRoux 1971). In contrast, a study by Bultman and Faeth (1988) on four leafminer species on Emory oak discovered that one of the species had higher survivorship on shaded trees due to higher levels of parasitism on sun trees. Further, Barrett (1994) found higher levels of spotted tentiform leafminers (Phyllonorycter blancardella) and apple blotch leafminers (P. crataegella) in the lower to middle canopy of apple trees while parasitoid activity was more evenly distributed throughout the canopy with a trend for higher parasitism in the middle to upper canopy regions. While no infested foliage was held for parasitoid emergence, examination of infested tips and ALM larvae during field and laboratory IBT assessments did not reveal any obvious and/or discernable evidence of parasitism, suggesting natural enemies were not a significant factor in contributing to differences in IBTs across cultivars.

The remaining 16 cultivars exhibited a gradation of damage from moderate (10%–20% branch tip damage) to severe (>20% branch tip damage). Observations in the current study are somewhat consistent with at least one other report noting pyramidal, globose, and golden arborvitae seem to be preferred, in that order (NCSU 2015). However, in this study, it is unclear why ‘Globosa’ and ‘Umbraculifera’—both of which are short, globose, and have dense canopies—had high levels of damage (22% and 31%, respectively). Apparently, other factors are contributing to ovipositional preference with these two cultivars.

In addition to plant height, height above ground also appears to influence ovipositional preference. Non-random distribution of leafminers is common among plants and among leaves within a given plant (Cornelissen and Stiling 2008). Findings in the current study suggest that adult female ALMs prefer to oviposit in a narrow range within the tree canopy, judging from the higher-level feeding damage observed at three to five meters above ground, which corresponds to the lower and middle canopy zones of tall trees and the upper canopy zone of short trees. Shorter trees tended to be noticeably less infested, with the highest levels of injury near the top of the canopy, and taller trees had generally higher levels of infestation particularly in the lower and middle sections of the canopy.

These findings are consistent with a study by Miller and Ware (2014) in which percent leaf area consumed by elm leafminer larvae (Fenusa ulmi) was greater in the lower exterior zones of medium-sized (5–7 m) elm trees. Barrett (1994) found the mean number of spotted tentiform leafminers and apple blotch mines per leaf were generally greater in the lower-middle canopy regions of apple trees compared to upper canopy regions. Auerbach et al. (1995) observed ovipositing P. tremuloidella female leafminers prefer small quaking aspen trees for oviposition.

A possible explanation for higher ALM tip damage in the middle and upper zones of arborvitae trees may be partially explained by Brown et al.’s (1997) hypothesis on lack of movement and egg depletion. In studying a leaf-mining moth (Cameraria hamadryadella) on oak (Quercus spp.), Brown et al. (1997) suggested higher leafminer densities on foliage in the lower oak tree crown may be due to the lack of other potential causal factors and due to the fact that the insect overwinters in the leaf litter. Females begin ovipositing on foliage closest to overwintering, emergence, and resting sites. While density of C. hamadryadella, in the lower crown of oaks conflicts with the current study’s findings, Brown et al. (1997) further stated this same hypothesis could also account for vertical stratification for insects more abundant in the upper tree crown. This hypothesis is supported by studies by Stark (1952; 1958), of a leaf-mining weevil (Recurvaria starki), and Harris (1958), studying Ocnerostoma piniariella, a needle miner of white pine, in which adults emerge from needles high in the tree crown, and these same needle miners are more dense in the upper tree crown. Field observations by the study authors and others have revealed adult arborvitae leafminers tend to fly briefly in a cloud no more than two to four meters off the ground, and then settle within a minute or so on the host tree or nearby vegetation (Busck 1907; Maier et al. 2011). It is possible that adult female ALMs, being weak fliers, may tend to oviposit at plant canopy levels as high or higher, from which they emerge, congregate, and rest. In addition, ALM adults emerging from mines higher on the plant may find sunny portions and congregation of adults more attractive for mating and oviposition. Additional studies are needed to support and confirm Brown et al.’s (1997) hypothesis on lack of movement and egg depletion.

In this study, it appears that canopy aspect was not a major factor in plant susceptibility; no significant differences in damage level were associated with cardinal direction. However, the findings are not completely consistent with other leafminer studies and/or other ALM field observations. Miller and Ware (2014) found no significant biological difference in mean percent leaf area consumed by canopy aspect for elm trees by the elm leafminer. In contrast, Davidson and Raupp (2010) advised monitoring for ALM by looking on the south side of the plants, where damage is heavier and damage is first seen in full sun. Exposure and plant age may also be a factor affecting severity of damage. Cloyd (2001) stated that ALM is attracted to Thuja planted in shady locations. He suggested locating plants in sun to help alleviate ALM defoliation problems. Hanson and Walker (2015) mentioned ALM is mostly attracted to older, slow-growing trees exposed to sunlight. Further studies are needed to better understand the relationship between canopy aspect in ALM ovipositional preference and subsequent larval feeding damage.

In the current study, no clear association between foliage color and infestation level was observed. While cultivars ‘Aurea’ and ‘Douglasii Aurea’ had yellow foliage and had high amounts of branch tip damage, similar levels of infestation were also seen in green cultivars (‘Smaragd’) and both species. Previous studies on other insect herbivores have been less than consistent regarding ovipositional preference, leaf color, and larval suitability. ‘Sunburst’ honeylocust, a yellow-leaf cultivar, has been found more susceptible to feeding damage by honeylocust plant bug compared to green-leaf cultivars (Cloyd 2008), but Bastian (1988) found no significant difference in mean larval development times of mimosa webworm (Homadaula anisocentra) feeding on the ‘Sunburst’ and green honeylocust cultivars. In addition, Miller and Hart (1987) found no significant differences in overwintering survival of mimosa webworm pupae when larvae fed on all common varieties of honeylocust, including ‘Sunburst.’ Further study with a larger range of yellow cultivars might help determine whether foliage color affects ALM ovipositional preference and larval suitability.

While not formally examined in this study, additional factors may influence ovipositional preference and larval success. This can affect the level of larval feeding damage, including but not limited to intra- and interspecific larval competition, biological control by natural enemies, host plant physical and chemical characteristics, plant age, plant condition (stress, nutrition, etc.), and exposure. Discussed here are some field and laboratory observations that may help to partially explain the role of these factors in female ALM ovipositional preference and subsequent plant damage.

Some studies have examined intra- and interspecific larval competition with externally and internally feeding insect herbivores, and the results have been inconsistent (Faeth 1991; Barrett 1994; Kagata and Ohgushi 2001; Cornelissen and Stiling 2006). In this study, neither intraspecific nor interspecific larval competition appears to affect larval development, survival, and subsequent feeding damage. Lepidopteran leafminers can detect the presence of conspecific eggs (Simberloff and Stiling 1988; Auerbach and Simberloff 1989), and this behavior could be present with adult female ALMs, thus mitigating intra-specific competition. Informal field and laboratory observations and measurements at the time of IBT assessments revealed mined tips on very dense cultivars were approximately one centimeter apart, while mine tips on less dense and sparse canopy cultivars were two to three centimeters apart. In this study, dissection of mines during IBT damage assessments consistently found only one ALM larva per mine, with more than one larva per mine in very rare cases. While not fully investigated in this study, the difference in branch tip spacing between very dense (1 cm) and sparse (2 to 3 cm) canopy species/cultivars might be partially responsible for lower levels of IBTs on very dense plants due to the ability of the ALM female to detect the presence of conspecific eggs, mitigating intraspecific larval competition.

The evergreen bagworm (Thyridopteryx ephemeraeformis), Fletcher scale (Parthenolecanium fletcheri), and two-spotted spider mite (Tetranychus urticae) are common leaf-feeding and sap-feeding pests of arborvitae. At least in this study, none of these pests were present and were not in competition with ALM larvae for plant resources indicating a lack of interspecific competition. However, this may not necessarily be true in other cases.

Over the course of this study, biological control by natural enemies did not appear to be an important factor in regulating ALM populations. While no infested mined leaf tips were held to check for parasitism, a visual examination of mined leaf tips and larvae during two IBT assessments in autumn–winter 2000–2001, and late-summer 2001, did not reveal evidence of parasitism, and no parasitoids were found associated with ALM larval mines over two successive ALM generations.

Van Driesche et al. (1996) noted two species of parasitoids, Pentacnemus bucculatricis and Apanteles dedelliae, associated with ALM in Connecticut, U.S. (Britton and Zappe 1922). In Quebec, Canada, Juillet (1972) recorded 26 species of parasitoids collectively destroying 50% of several species of arborvitae leafminers. In addition, a study by Bazinet and Sears (1979) found seven parasitoids species caused 9%–18% mortality of ALM, but their impact was reduced due to 63% winter-related mortality in 1977.

It is highly probable that other morphological and/or chemical plant characteristics may be responsible for making these very dense plants less favorable or desirable, but their investigation was beyond the scope of this study.

Results from the current study suggest some important guidelines for ALM management useful to arborists, urban foresters, landscape professionals, and nursery growers. First, while ALM can be a serious defoliator of arborvitae, and early detection and management is critical for overall plant health, some varieties appear to be more affected than others, and there may be different tolerances for damage depending upon a given situation (e.g., plants grown in a nursery for sale versus those in established landscapes). A study by Raupp et al. (1988) on bagworm defoliation on American arborvitae to determine aesthetic injury levels (AIL) and economic injury levels found 50% of observers perceived damage at 8% defoliation and discoloration. Further, Koehler and Moore (1983) examined the relationship between density of cypress tip miner leafminer as an indicator of aesthetic injury using an “unsightliness rating” as one of the first attempts at quantifying the relationship between pest abundance and aesthetic quality of a landscape plant. Weston (2008) calculated woody plant pest indexes incorporating detectability, damage potential, and ease of control, with leafminers falling in the moderate range with suggested treatment thresholds of 50% of leaves infested. He based his leafminer threshold on the notion that 25%–50% defoliation can be tolerated by a woody plant before damage is done. However, he did not distinguish between leafminers feeding on deciduous and evergreen plants. The authors of the current study did not conduct a formal analysis of “unsightliness” or conduct a formal damage assessment; the authors did receive verbal communication from The Morton Arboretum’s Manager of Plant Collections conveying his concern over the level of plant damage on the arborvitae study plants (K. Bachtell, personal communication). In the current study, this suggests an AIL had been approached, if not exceeded; two cultivars (‘Hetz Midget’ and ‘Hoopesii’) were below the 8% defoliation–discoloration level identified by Raupp et al. (1988). According to Raupp et al. (1988), evergreen plants (i.e., arborvitae) can tolerate up to 20% defoliation, and entirely re-foliate in one to two years. Using a suggested AIL of 20% defoliation, an additional five species/cultivars would probably not require plant protection treatments. The remaining nine species/cultivars with IBTs >20% would require some type of appropriate control measures.

Second, when selecting arborvitae for the landscape and/or for managing existing plants, selecting and incorporating short, dense, and globe-shaped arborvitae species/cultivars may be less prone to ALM feeding damage compared to taller, columnar, and pyramidal-shaped trees.

Third, when monitoring for ALM, visually assessing and/or sampling from the upper canopy on shorter plants, and the lower and middle canopy areas of larger trees, will provide an accurate assessment of ALM population levels and potential plant damage. Scouting for IBTs would best be performed in spring, prior to adult emergence in preparation for adult foliar sprays, if warranted, or in late-summer to early-autumn, when most leafmining activity has peaked.

Fourth, examining any aspect (N, S, E, W) of the tree canopy should provide an accurate assessment of existing feeding damage.

Fifth, biological control may not always be a reliable management tactic as seen in this study. Finally, because of ALMs univoltine life cycle, and late-summer and early-spring larval feeding, premature leaf abscission will not be a major larval mortality factor as seen with other univoltine or bivoltine deciduous plant feeding leafminers.

In conclusion, it is recognized that this study only looked at susceptibility of arborvitae cultivars for ALM and not ovipositional preference and associated insect-plant relationships. The authors further recognize that observations and conclusions in this study are limited by the small number of available plants per cultivar available for study. Additional laboratory and field studies are needed to better examine the role(s) of plant growth habit, plant nutrition, environment, production of defensive chemicals and plant genetics, natural enemies, and female ovipositional searching behavior in ALM population dynamics. Further information is also needed on the susceptibility of other common plant cultivars, such as ‘Green Giant’, and ovipositional and feeding preference of other arborvitae leafmining species.

Information gathered from these studies will provide even more reliable, accurate, and efficient management practices improving ALM pest management.