Abstract
An analysis of data collected from 212 residential landscapes in suburban Maryland, U.S., revealed significant positive relationships between the number of insect and mite pests in the landscape and the total number of plants and plant species at the site. The number of pests in a landscape increased very little in relation to the number of plants found in the landscape. However, the number of arthropod pests increased at a much greater rate as more species of plants were added. Two explanations for these results are likely. Relatively few plants harbored arthropod pests throughout the course of the season. Adding more plants of the same species had little effect on altering the number of pest species in a landscape. Arthropod pests tend to be relatively specialized in their host range. When different species of plants are added to a landscape, more opportunities are created for specialized insects and mites to colonize the site and increase the richness of the arthropod fauna. When used in conjunction with previous investigations involving monitoring approaches, these results help IPM and PHC monitors plan and conduct site inspections more efficiently and effectively.
One of the pillars of both Integrated Pest Management (IPM) and Plant Health Care (PHC) programs is the regular inspection of plant material in the managed landscape (Raupp et al. 1985; Ball 1987,1994; Ball and Marsan 1991).This process, known as monitoring, forms the basis for making decisions regarding corrective actions such as the application of pesticides or ameliorative and restorative actions such as pruning, fertilizing, mulching, or irrigating (Lloyd and Miller 1997; Ball et al. 1999; Harris et al. 1999).
During the past two decades, several advances have helped arborists understand the relationships between landscape plants and their pest complexes. Not the least of these has been the recognition that a relatively small number of pests (called key pests) are responsible for the majority of biotic problems found in the landscape (Raupp and Noland 1984; Nielsen et al. 1985; Raupp et al. 1985, 1992; Harris et al. 1999). There are now several well-developed lists of key pests for different regions of the United States. These lists have been generated in two general ways. Some are compilations of large surveys conducted with groups such as municipal foresters on regional and national scales (Nielsen et al. 1985; Wu et al. 1991). Others were compiled from monitoring records gathered by IPM and PHC technicians during their visits to the landscapes of clients (Holmes and Davidson 1984; Raupp and Noland 1984). Both types of data have been extremely valuable in helping arborists and landscapers prepare bids, train employees, and focus their monitoring and intervention activities.
With the evolution of IPM and PHC programs for landscapes came refinements in our understanding of methods to make monitoring more effective and efficient. Authors such as Holmes and Davidson (1984) demonstrated that the amount of time spent monitoring was directly related to the size of the property under management. Next came the recognition that monitoring activities could be made more efficient if actions centered on plants that were exceptional in their value to the landscape and propensity for problems. This led to the emergence of the concept of key plants in landscapes (Nielsen 1983; Raupp et al. 1985; Lloyd and Miller 1997; Harris et al. 1999). In an important paper, Ball (1987) used this concept to demonstrate that landscapes containing greater numbers of key plants required significantly longer amounts of time for technicians to monitor. By enumerating key plants in residential landscapes, arborists could more accurately estimate the amount of time necessary to monitor and treat plants and thereby estimate jobs more effectively (Ball 1987). Recently, Ball et al. (1999) have linked levels of plant injury to intervention responses of arborists using a system called the Appropriate Response Process. That paper expanded the role of monitors by developing five plant health injury classes that demonstrate appropriate responses monitors should take at each level of plant injury.
Our objective was to gain further insights into the relationships between the abundance and types of plant materials found in landscapes and the diversity of associated insect and mite pests. Relatively few studies have attempted to quantify the diversity of pests found in landscapes (Hellman et al. 1982; Raupp and Noland 1984; Nielsen et al. 1985; Raupp et al. 1985; Ball 1987; Raupp and Shrewsbury 2000). In particular, we were interested in determining if relationships existed between the number of arthropod pests and abundance and diversity of the plant material found in the landscape. It was our goal to provide arborists and landscapers with information that could be used to estimate the magnitude of pests likely to be encountered in home landscapes.
MATERIALS AND METHODS
Data were gathered from suburban home sites in central Maryland, U.S., that were part of demonstration landscape IPM programs conducted by the University of Maryland in 1981 and 1982. Beginning in April of both years, trained IPM monitors visited each landscape and recorded the identity and number of each woody plant species found in the landscape. Samples from plants that could not be identified in the field by monitors were brought to the Plant Diagnostic Laboratory at the University of Maryland where they were identified by plant taxonomists. In the case of woody ground covers such as Vinca minor or Euonymus fortunei, single plants could not be readily enumerated, and plants with coalesced canopies were recorded as single plant units as reported by Raupp and Noland (1984). Starting in May and continuing through September at biweekly intervals, all plants in each landscape were examined and all occurrences of insect and mite infestations were recorded. When monitors were unable to identify pests in the field, samples were returned to the University of Maryland where entomologists assisted with the identification.
The average size of the home landscapes ranged from slightly less than one-quarter acre to slightly more than one-half acre and was typical for homes found in and around the suburbs of Washington, D.C.The composition of landscape plants found in these landscapes has been discussed in greater detail elsewhere (Hellman et al. 1982; Raupp and Noland 1984; Raupp et al. 1985, 1999; Raupp and Shrewsbury 2000).
To examine the relationships between the number of arthropod pests encountered and various attributes of plant diversity, regression analyses were performed as in Shrewsbury and Raupp (2000). Linear regression analyses were used to investigate the relationship between the number of arthropod species found in a landscape and the number of plants, plant species, species diversity (Hʹ), and relative diversity or evenness (Jʹ). Plant species diversity was estimated using a Shannon-Wiener Diversity Index (Zar 1999). Plant species diversity was calculated as follows:
where n = total number of plants; fi = number of plants of each species; and log = natural log. Plant species evenness or relative diversity was calculated as follows:
where Hʹ max = log K, and K = the total number of plant species and was the maximum possible diversity.
To determine which components of diversity best predicted the number of arthropod pests in the landscape, a regression analysis was performed on the number of pests and the number of plants, number of plant species, plant diversity (Hʹ), and plant evenness (Jʹ) (SAS 1988).
RESULTS AND DISCUSSION
The number of home landscapes examined in this study was 100 and 112 in 1981 and 1982, respectively. The total number of plants and plant units sampled was 13,395. The best predictor of arthropod pest diversity was the number of plants found in the landscape (Table 1, Figure 1).This variable explained about 18% of the variation in the number of insect and mite species observed during the course of the growing season (Table 1). Almost as predictive as the total number of plants was the number of plant species in the landscape (Table 1, Figure 2).This variable independently predicted about 17% of the variation observed in the number of arthropod pests found (Table 1). Although there was a significant relationship between plant diversity, Hʹ, and the number of pests observed, this relationship was weak and explained less than 3% of the variation in the number of pests found (Table 1). There was no relationship between plant evenness, Jʹ (that is, the observed diversity relative to the maximum diversity), and the number of pests observed (Table 1).
Relationship between the number of plants in a landscape and the number of arthropod pest species associated with them. The linear regression was significant at P = 0.001, and the regression equation was y = 0.047x + 4.14, where y = number of pest species and x = total number of plants.
Relationship between the number of plant species in a landscape and the number of arthropod pest species associated with them. The linear regression was significant at P = 0.001, and the regression equation was y = 0.227x + 3.08, where y = number of pest species and x = number of plant species.
Summary of regression analyses of number of arthropod pests and the number of plants, plant species, diversity (Hʹ), and evenness (Jʹ) encountered in 212 suburban Maryland landscapes monitored in 1982 and 1983.
A correlation analysis of all independent variables revealed a high and significant relationship between the number of plant species and the total number of plants found in the landscape (r = 0.728, p < 0.001). This is not altogether surprising, for often in many of the landscapes only a single specimen of a plant species was found at the site. Therefore, increasing the number of plant species in a site concomitantly increased the total number of plants at the site.
The relationships between the number of arthropod pests and the number plants and plant species in a landscape are both interesting and important. The slopes of regression equations for total number of plants and number of plant species on number of pests both differ significantly from one (numbers of plants versus number of pest species, Z = 6.554, p < 0.0001; number of plant species versus number of pest species, Z = 6.198, p < 0.0001). Therefore, the rate at which the number of pests accrues in a landscape is less than the rate at which numbers of plants are added (Figure 1).This relationship confirms the observation that there were numerous plants in every landscape that lacked any noticeable insect or mite pests for the entire duration of the study.
Adding additional plants of the same species to a landscape is unlikely to increase the number of resident arthropod pests for at least two reasons. First, most plants never become infested with insects or mites or achieve pest status during the course of the growing season (Raupp and Noland 1984). Adding an additional plant of the same species is unlikely to change this situation; hence, the slope of the regression of number of pest species on number of plants is far less than one. Second, once a pest has successfully colonized a landscape and achieved pest status, adding a plant of the same species will not alter the species richness of pests in the landscape unless a second pest species accompanies or colonizes the second plant. This phenomenon suppresses the rate at which pests accrue in landscapes as plants are added.
Adding additional species of plants to a landscape has a somewhat different effect on the numbers of arthropod pests found there. The rate at which the number of pests increased in relation to the number of plant species added was less than one (Figure 2). However, adding plant species quintuples the rate at which pest species increase relative to adding more individuals of the same species. This pattern likely reflects at least two important relationships between arthropod pests and their host plants. Several earlier studies established the fact that certain plants—key plants—were likely to be attacked by pests, while others were less likely to be (Raupp and Noland 1984; Holmes and Davidson 1984; Raupp et al. 1985; Ball 1987; Raupp et al. 1999; Raupp and Shrewsbury 2000). The relative absence of arthropod pests on many common woody plants likely causes the relationship between number of plant species and number of pest species to be less than one to one. Species of plants can be added to a landscape without simultaneously increasing the number of arthropod pests because most herbivorous arthropods are relatively specialized in their diets and eat relatively few taxa of plants (Bernays and Chapman 1994). There is strong evidence for this trend in landscapes where many of the key pests such as lace bugs, armored scales, aphids, defoliating caterpillars and sawflies, and borers tend to have relatively restricted host ranges (Holmes and Davidson 1984; Raupp and Noland 1984; Ball 1987; Johnson and Lyon 1988).Therefore, adding a new species of plant to a landscape will provide new opportunities for additional pest species to colonize and become resident. This will cause pest species richness to increase at a greater rate relative to the rate observed when additional individuals of the same species are added. Hence, numbers of arthropod pests increase at a greater rate when species of plants rather than individual plant specimens are added to a landscape.
Increasing the taxonomic diversity of plants in a landscape increases the number of pests that will be found in a site. However, the rate at which pests accumulate is far less than the rates at which plants or plant species are added. While diversifying landscape plantings may increase the absolute number of pests, recent studies have shown that diversification may actually reduce the severity of pest outbreaks in landscape systems (Hanks and Denno 1993; Trumble and Denno 1995; Shrewsbury 1996; Shrewsbury and Raupp 2000).
The findings of previous researchers and those presented here can be incorporated into IPM and PHC schemes in the following way. Arborists gain valuable insight into how long it will take to monitor a property simply by knowing the absolute size of the landscape. Holmes and Davidson (1984) clearly demonstrated a direct relationship between the size (acreage) and amount of time required to monitor a property. As we have shown, landscapes containing many plant species tend to have more pests than ones with fewer plants or species of plants. We caution that while we demonstrated relationships between numbers of pests and plants and species, we did not attempt to quantify the levels of damage pests were causing. In most cases where pest species were observed, they were not at levels that warranted intervention (Holmes and Davidson 1984; Raupp and Noland 1984). On the contrary, there is mounting evidence that the severity of pest damage will be reduced as plant diversity in landscapes is increased (Hanks and Denno 1993; Trumble and Denno 1995; Shrewsbury 1996; Shrewsbury and Raupp 2000).
Arboricultural and landscape management firms will need well-qualified Plant Health Care monitors to inspect plants at sites composed of highly diversified plant materials as a greater diversity of pest species is found at these sites. Landscapes with many individuals of the same type of plant will have fewer species of arthropod pests than landscapes with many types of plants. The diagnostic skills of sales representatives and Plant Health Care technicians may require consideration when work assignments are made. Technicians monitoring diverse landscapes need to recognize a wider variety of plants and their associated arthropod pests than those monitoring simple landscapes.
Finally, as Ball (1987) clearly demonstrated, the amount of time spent monitoring and controlling pests will be strongly influenced by the number of key plants found in a landscape.The propensity of key plants to harbor damaging levels of pests and intensities of pest damage may override other factors such as the size of the property or diversity of the landscape planting (Ball 1987; Ball et al. 1999).
Using information including the size of the property, number of plants and plant species, number of key plants, and intensity of plant care, arborists and landscapers gain considerable insight into the number of pest problems and amount of time that will be required to monitor and intervene in landscapes (Holmes and Davidson 1984; Raupp et al. 1985; Ball 1987;Ball et al. 1999).This information will be useful in making work estimates more accurate and monitoring activities efficient, effective, and reflective of the intensity of intervention.
Acknowledgments
We are grateful to the many homeowners in the suburban Maryland communities who allowed the use of their landscapes as study sites. We thank two reviewers whose editorial comments improved the manuscript. This work was funded in part by the Chesapeake Bay Program of the Environmental Protection Agency.
- © 2001, International Society of Arboriculture. All rights reserved.
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.