Ornamental Plants

Annual Reports and Research Reviews

The Flowering Sequence of Ornamental Plants as a Tool for Predicting the Phenology of Insect Pests

Daniel A. Herms

Summary

This report presents the phenological sequence of 56 plant and 22 insect species for Wooster, Ohio, in 1997. A phenological sequence developed in Midland, Michigan, based on five years of data was evaluated for its accuracy in predicting insect activity in Wooster in 1997. For some insects, correspondence was quite close; for other insects, reliance on the Michigan sequence would have resulted in mistimed treatments. These data suggest that phenological indicators developed in one region may need to be evaluated on a case-by-case basis before they can be used with confidence in another region.

Introduction

The tremendous diversity of ornamental plants, each with its own complement of insect pests, creates a logistical challenge for planning and implementing a successful pest management program. Insecticide applications must be timed precisely to maximize their effectiveness and minimize the number required. Improperly timed insecticide applications are expensive, environmentally detrimental, and result in customer dissatisfaction. Many insects are difficult to detect and monitor, further complicating the accurate timing of pesticide applications. Consequently, pesticide applications are frequently scheduled on a calendar-day basis. However, because of variation in patterns of degree-day accumulation from place to place and year to year, calendar-based scheduling is frequently inaccurate.

The use of ornamental plants as phenological indicators provides an alternative approach to predicting insect activity. Phenology is the study of recurring biological phenomena and their relationship to weather. Bird migration, hunting and gathering seasons, blooming of wildflowers and trees, and the seasonal appearance of insects are examples of phenological events that have been recorded for centuries (4,8). Because the development of both plants and insects is temperature dependent (6,12), plants may accurately track the environmental factors that affect insect development. Indeed, the use of plant phenology to predict insect activity is an old practice, with recorded observations that date back at least to the 18th century (7).

In recent years, two sets of phenological indicators have been published that demonstrate that phenological indicators can be useful tools for predicting the activity of insect pests. Donald A. Orton and Thomas L. Green published an extensive list of observations made in Illinois in their book Coincide (10). In addition, this author published a phenological sequence based on five years of data collected in central Michigan (5).

A critical but unresolved question is whether phenological indicators developed in one geographic region can be used accurately in another. The objectives of this report are:

• To present a phenological sequence for Wooster, Ohio, in 1997.
• To test the ability of the Michigan phenological sequence to predict the occurrence of insect activity in Wooster, Ohio.

Methods and Materials

During 1997, the phenology of 56 plant species and/or cultivars and 22 species of insects on or near the Ohio Agricultural Research and Development Center's Wooster campus were monitored (Table 1). For clarity, only common names are listed. To achieve standardization in nomenclature, common names of plants follow Dirr (3), and insect names are official common names as approved by the Entomological Society of America.

Plants were chosen to represent a range of blooming times from early March through late July. This time period corresponds with the activity of most of the important insect pests of ornamental plants. Four individuals of each species or cultivar were monitored. All individuals of a species were located either in uniform sun or shade, depending on the environment to which the species is adapted. Plants in microenvironments obviously altered by buildings, parking lots, and related areas were not used. Plants were monitored at least three times each week, and the dates of "first bloom" and "full bloom" were recorded. "First bloom" was defined as the date on which the first flower bud on the plant opened, revealing pistils and/or stamens, and "full bloom" as the date on which 95% of the flower buds have opened (i.e., one bud out of 20 has yet to open). These phenological events can be identified and recorded with precision.

The phenology of 22 insect pests with diverse life histories was also monitored in 1997, including defoliators, scales, gall formers, wood borers, and leafminers. As opposed to methods used to monitor plant phenology, which were designed to minimize variation in order to increase predictive power, sampling protocols for insects were designed to characterize the phenology of the entire population.

The Michigan phenological sequence developed at Dow Gardens in Midland consisted of 55 plant and 24 insect species, the phenology of which was monitored for five years (1985-1989) using the exact protocols described previously (5). Since the Michigan and Wooster phenological sequences have in common 40 plant and 14 insect species, they can be directly compared in order to test the accuracy of a sequence developed in one region (Michigan) for predicting insect activity in another (Wooster).

Results and Discussion

The Phenological Sequence

The phenological sequence observed in Wooster, Ohio, in 1997 is presented in Table 1. The sequence of occurrence is much more important than the dates, which are only included to provide perspective, and which will vary dramatically with year-to-year variation in degree-day accumulation. During the cool spring of 1997, plant and insect phenology was substantially delayed relative to "normal" years.

It is important to emphasize that any conclusions drawn from this sequence are preliminary, as they are based on only one year of data. Because consistency in the sequence is essential if the sequence is to be useful for predicting insect activity, this study will continue for several more years to test this critical assumption. However, in the five-year Michigan study, there was substantial year-to-year variation in degree-day accumulation, yet little variation in the order in which plants bloomed and insects appeared from one year to the next.

When a phenological sequence of plants can be shown to correspond with the appearance of insects, pest managers can use the easily monitored plant sequence as a "biological clock" to anticipate the order and the time when pests reach vulnerable stages. This can greatly facilitate the logistics of serving many clients with a variety of problems. For example, one can see from Table 1 that when common lilac was in full bloom:

• Crawler emergence of pine needle scale was well underway, and sprays were appropriate.
• It was too late to worry about first generation birch leafminer.
• Lilac borer pheromone traps should have been deployed promptly.
• There was still time to plan for bronze birch borer.

Again, it must be emphasized that this first year of data should be considered preliminary. It remains to be seen if this sequence proves valid as the study continues.

A useful attribute of phenological sequences is that they can be readily expanded and customized. Once the basic sequence is in place, any new plants or pests can be added as the need occurs. For example, if a pest manager has made a particularly successful treatment, any plants in bloom at the time can be noted, added to the sequence, and used to duplicate the timing in future years.

Are Phenological Indicators Accurate Across Regions?

This critical question must be answered before published phenological sequences can be used with confidence outside the regions in which they were developed. To the degree that phenological relationships do not hold across regions, it will be necessary to develop localized sequences. Table 2 can be used to evaluate the accuracy of the Michigan plant sequence for predicting insect phenology in Ohio. The order in which phenological events occurred in Ohio in 1997 is compared with the average order in which they occurred in Michigan from 1985-1989. The magnitude of any disparity in the order in which a particular event occurred in the two sequences is also shown. A positive value indicates that the event occurred earlier in the sequence in Michigan that it did in Ohio, while a negative value indicates the opposite.

The correspondence in the flowering sequences of the plants common to both studies was quite close. Flowering of silver maple occurred first in both locations, followed by flowering of corneliancherry dogwood. Red maple flowered just before border forsythia in Ohio, and just after it in Michigan; first bloom of Norway maple was the ninth event in both sequences. In fact, in no case did the order of flowering by any species vary by more than the five places between the two sequences. As more plant species are included in the sequence, the chance of being led astray by one species that departs from the pattern will diminish.

The sequence of insect phenological events did not correspond as closely. Egg hatch of eastern tent caterpillar occurred before first bloom of red maple, border forsythia, and Korean rhododendron in Michigan, but after their first bloom in Ohio. Some of the discrepancies were substantial enough that any pesticide applications on which they had been based would have been mistimed and probably inefficacious. For example, egg hatch of oystershell scale closely followed first bloom of Vanhoutte spirea in Michigan, but did not occur until more than two weeks after first bloom in Ohio. There was even less correspondence in the phenology of euonymus scale egg hatch, which occurred just after first bloom of black cherry in Ohio, but not until after first bloom of common ninebark in Michigan. The much closer correspondence in other species including birch leafminer, pine needle scale, bronze birch borer, and juniper scale suggests that phenological sequences may have regional generality, at least in some cases.

A number of factors may affect the accuracy of plant phenological indicators as predictors of insect activity. The assumption that a given phenological correlation will occur in different climatic regions requires that all organisms included in the correlation have the same upper and lower temperature thresholds for development, as well as the same developmental responses to changing temperature. These traits are known to vary widely among both plants and insects (6,12).

Origin, or provenance, of indicator plants can affect their accuracy as phenological indicators. Both elevational and latitudinal gradients can affect plant phenology, as different provenances are adapted to different day-length and temperature regimes (12). Furthermore, the same genotype may display different phenological patterns when planted in different geographic regions (9). Substantial geographic variation has also been documented in the phenological patterns of insects (1, 2, 13).

Data from the Michigan study indicate that insects that overwinter in the soil, such as the imported willow leaf beetle and white pine weevil, are less accurately predicted from year-to-year than insects that overwinter on aerial plant parts exposed to the same ambient conditions as flower buds. The phenology of insects overwintering in the soil may be influenced substantially by a number of factors not having as strong an effect on the phenology of flower buds, including vegetation, snow cover, and soil moisture (11).

Phenological indicators will be more accurate in the region in which they are developed because these factors are less variable within a region than among different regions. Furthermore, at any given location the weather is more similar from one year to the next than is the weather from one location to another. The results of this study suggest that while some phenological indicators may be regionally robust, others are not. Hence, phenological indicators developed in one region will need to be evaluated on a case-by-case basis before they can be used with confidence in another region. None the less, the use of plant phenological indicators for timing pest activity holds tremendous potential for improving the effectiveness of integrated pest management programs in the landscape.

Literature Cited

1.Akers, R. C. and D. G. Nielsen. 1984. Predicting Agrilus anxius Gory (Coleoptera: Buprestidae) adult emergence by heat unit accumulation. Journal of Economic Entomology 77:1459-1463.

2. Brakefield, P. M. 1987. Geographical variability in, and temperature effects on, the phenology of Maniola jurtina and Pyronia tithonus (Lepidoptera, Satyrinae) in England and Wales. Ecological Entomology 12:139-148.

3. Dirr, M. A. 1983. Manual of Woody Landscape Plants: Their Identification, Ornamental Characteristics, Culture, Propagation and Uses, 3rd edition. Stipes Publishing Co., Champaign, Illinois.

4. Glendenning, R. 1943. Phenology: the most natural of sciences. Canadian Field-Naturalist 57:75-78.

5. Herms, D. A. 1990. Biological clocks: using plant phenology to predict insect activity. American Nurseryman 172(8):56-63.

6. Higley, L. G., L. P. Pedigo, and K. R. Ostlie. 1986. DEGDAY: A program for calculating degree-days and assumptions behind the degree-day approach. Environmental Entomology 15:999-1016.

7. Huberman, M. A. 1941. Why phenology? Journal of Forestry 39:1007-1013.

8. Levitt, D. 1981. Aboriginal uses of plants. Groote Eylandt. Australian Institute of Aboriginal Studies, Canberra, Australia.

9. McGee, C. E. 1974. Elevation of seed sources and planting sites affects phenology and development of red oak seedlings. Forest Science 20160-164.

10. Orton, D. A. and T. L. Green. 1989. Coincide. Plantsmen's Publications, Flossmoor, Il.

11. Raffa, K. F. and D. W. A. Hunt. 1989. Microsite and interspecific interactions affecting emergence of root-infesting pine weevils (Coleoptera: Cur-culionidae) in Wisconsin. Annals of the Entomological Society of America 82:438-445.

12. Rathcke, B. and E. L. Lacey. 1985. Phenological patterns of terrestrial plants. Annual Review of Ecology and Systematics 16:179-214.

13. Tauber, C. A. and M. J. Tauber. 1981. Insect seasonal cycles: genetics and evolution. Annual Review of Ecology and Systematics 12:281-308.