REU: Kimberly Ogden  
Terrestrial Invertebrates in Devil’s Postpile National Monument:
Preliminary Study on Community Structure Adjacent to Trails and a Comparison of Sampling Methods




The National Park Service is interested in finding an indicator group in order to efficiently measure the ecological health of various meadows in the Eastern Sierra Nevada. In addition, the National Park Service in interested in the basic community structure and movement patterns of invertebrates in meadows since very little is known about montane meadows in general. The primary objective of the study was to determine if trails through the heavily visited meadows in Devil's Postpile National Monument, CA are having a measurable effect on the meadow fauna as well as collecting community structure data for invertebrates in the meadow using pitfall traps. No significant affects on meadow invertebrates could be linked specifically to trails through Soda Springs Meadow. However, the results seem to suggest a correlation between vegetation type and community structure. There is a need for further studies in heavily used montane meadows.


In the course of human inhabitation, natural habitat is inevitably artificially fragmented. Fragmentation has proven to be significantly detrimental for many species of flora and fauna (Forman 1995) and is thus an extremely important subject of study. Studies on habitat fragmentation are not only of interest to the ecological scientist, but to those that manage natural areas for the use of human recreational activities as well.

Fragmentation can come in many shapes, sizes and degrees. Natural boundaries can be created, for example, by a river traversing a meadow or a forest fire that burns a large patch of trees. In contrast to areas of human created fragmentation, natural fragments are rarely permanent and actively maintained. A common source of fragmentation caused by humans is the creation roads and railways for transportation (Forman 1995). Disturbances from roads and railways that cause fragmentation are usually in place for an extended period of time (Forman 1995). In any case of fragmentation, a number of ecological effects could be occurring.

Besides the obvious shrinkage of natural habitat for native flora and fauna to simply exist in as a result of fragmentation, other ecological effects have been shown to occur. Previous studies have shown that landscape boundaries can have an effect on ecological flows (Wiens 1992). Nutrient flow, genetic flow between populations, and individual movement, among others, can be altered depending on the nature of the fragmentation boundary (Wiens 1992). Genetic flow between populations and simple movement of individuals can be a serious issue when considering metapopulation dynamics and the recolonization of habitat patches (Hansson, Fahrig, and Merriam 1995). In a landscape mosaic considering metapopulation movement, the actual distances between patches are not straight-line measurements, but rather a complex function of boundary permeabilities to moving organisms (Wiens 1997).

When considering the management of significantly reduced natural habitat, a category so many of the national and state reserves in the United States find themselves in, most question whether it is possible to maintain a high population survival probability of organisms (Hansson et al. 1995). In most cases, habitat fragmentation cannot be qualified using one measure; it involves the simultaneous loss of habitat, reduction of patch size and increase in patch isolation (Hansson et al. 1995). As a result of difficulties in measuring the affect of habitat fragmentation on biotic communities, there is no simple strategy for landscape managers trying to protect reserves. In considering the plight of landscape managers, it is then particularly important to do fine scale ecological studies in areas under active wildlife management in order to develop a strategy to reduce the threat of habitat fragmentation on the survival of organisms.

Many studies on habitat fragmentation have shown that paved roads can be detrimental to the movement patterns of fauna (Trombulak and Frissell 2000). Interestingly enough, very few studies have focused on smaller disturbances that segment habitats like trails and footpaths. Although a dirt or gravel path may not seem as obviously adverse to faunal movements as a paved highway with a high traffic volume, research has shown that even grassy field-tracks can pose difficulties for some animals (Mader, Schell, and Kornacker 1990). In fragile ecosystems, such as those of mountain meadows, the effects of fragmentation might be magnified. In effect, a grassy path could possibly act just as though a paved highway were placed through the meadow.

Devil's Postpile National Monument, part of the Sierra Nevada Network parks, is home to many ecologically important features, including montane meadows. The park is also a popular tourist destination with more than 100,000 recreational visitors per year (National Park Service Survey 2003). Hence, the trails through the park meadows are in frequent use. The meadows in the monument provide habitats for a variety of plants and animals in addition to capturing sediment and buffering stream flows (Hagberg 1995). Furthermore, the meadows are highly valued for scenic vistas and colorful flora, causing a disproportionate amount of public use (Hagberg 1995).

The National Park Service is interested in finding an indicator group in order to efficiently measure the ecological health of various meadows in the Eastern Sierra Nevada. In addition, the National Park Service in interested in the basic community structure and movement patterns of invertebrates in meadows. Insects have the potential to be excellent indicators for the status of montane meadows. They are diverse, easy to sample, and sensitive to certain microclimates (Rosenberg, Danks, and Lehmkuhl 1986). Insects constitute a significant portion of terrestrial biomass and play a significant role in ecosystem functioning (McGeoch 1998). Terrestrial invertebrates have the potential to be of excellent use to the National Park Service and are thus the subject of this study.

The primary objective of the study was to determine if trails through the meadows in Devil's Postpile National Monument are having a measurable effect on the meadow fauna. In order to accomplish this task, a matrix of pitfall traps was employed in collecting terrestrial invertebrates adjacent to two trails that traverse the meadow. In addition to the primary objective, a comparison of terrestrial invertebrate sampling methods was accomplished using the community structure data gained from this study as well as of that from Holmquist (unpublished data 2004).

Materials and Methods

Portions of this study have been modeled after prior studies on the movement patterns of Coleopterans, namely, the 1990 study published in Biological Conservation, "Linear Barriers to Arthropod Movements in a Landscape" by Mader, Schell and Kornacker, and the 1984 study published in Biological Conservation, "Animal Habitat Isolation by Roads and Agricultural Fields" by Mader. Modifications have been made, specifically in the spacing of the pitfall traps, to fit the scale of this meadow study.

As an initial step to determine scale for this project, two days and one night of pitfall traps were set in Soda Springs Meadow in Devil’s Postpile National Monument. Initially, a matrix of 3 traps by 4 traps was placed in a North-South orientation approximately 1.0m apart. The traps were set for 6 hours in an area of relatively low vegetative cover. The number and order of invertebrates captured was recorded for each trap. An average of 4 common black ground-dwelling beetles per trap was found. Relatively few other invertebrates were found including spiders (Araneae) and mites (Acari). The traps were then emptied and replaced for the night trial. No invertebrates were found in the traps after the 10-hour night trial. After the night trial, the traps were taken out and the soil cores replaced. The trap matrix was then moved to another portion of the meadow further west with more vegetative cover for comparison to the sparsely vegetated previous site. During this second day trial, the traps were placed in a matrix of 3 traps by 4 traps in an East-West orientation 2m apart. The traps were set for 5 hours and an average of 2 of the indicator beetles per trap was found. Almost no other invertebrates were found including ants (Hymenoptera), spiders (Araneae) and mites (Acari). After the fauna were classified to order, they were released and the soil cores replaced.

In light of the results from the pilot study, the trap matrix at 1.0m apart was feasible for a mark-release-recapture study in this particular meadow. In addition, two ground-dwelling beetles seemed abundant throughout the meadow so as to be easily measured as indicators of ecological health.
During the months of July and August 2004, the commonly employed pitfall trap method (Mader et al. 1990, Mader 1984) was used to survey the community structure of terrestrial invertebrates in relation to two trails through the meadows of Devil’s Postpile National Monument, CA.

The study site was adjacent to the Ranger Station in Devil’s Postpile National Monument, located in Mammoth Lakes, CA. The two trails chosen for this study, a 2.6m wide gravel path delineated with large stones and a 1.0m wide packed dirt trail with no hard edges, originate from the Ranger Station area and traverse the meadow in a general North direction. The site was chosen because the trails through the meadow receive a large amount of foot traffic each year from visitors on their way to the Devil’s Postpile geological formation, Rainbow Falls, or other popular landmarks. In addition, very little is known about the invertebrate population in the Devil’s Postpile National Monument meadows and the park service is eager for any type of census survey to be conducted.

Two common terrestrial beetles (Family Carabidae) were used as an indicator of invertebrate movements in relation to the two trails that traverse the meadow. A mark-release-recapture study using the pitfall traps placed alongside the trails to capture the beetles was conducted. Pitfall traps are useful, and often utilized, for mark-recapture studies involving small vertebrates and invertebrates (Holland and Smith 1999, Mader et al. 1990, Mader 1984, Perner and Schueler 2004). Pitfall traps for this study consisted of clear plastic cups, approximately 7.6cm in diameter and 7.6cm deep. The cups were placed in corresponding holes of the same dimensions, cut out of the meadow with a hand trowel, so that the lip of the plastic container was level with the ground. The cores of soil and plant matter from the holes were placed in a protected area, out of the sun and wind. At the conclusion of the study the soil cores were replaced in the exact location they were taken from so as to minimize any impacts on the meadow.

At each of the two trails, grids of pitfall traps were oriented so that three rows of traps extend into the meadow from each side of the trail (see Figure 1). The first row of pitfall traps was located 1.0m from the edge of the trails and each successive row was located 1.0m from the previous row. Each trail held two replicates; each replicate consisted of 30 traps, 15 on each side of the trail (see Figure 1). Thus, 120 pitfall traps were in use during the study period.

Figure 1 – Pictorial representation of pitfall trap grid
around a trail in Devil’s Postpile National Monument.

The traps were in use for 6 daylight hours during weekdays. When not in use, the traps were closed by removing the plastic container from its hole. After each day of trapping, the number and order of all invertebrates in each individual trap was noted. Considering the results of the pilot study, an average of three beetles per trap was expected. Thus, an average trap day would yield 360 beetles. Beetles captured in the traps were individually marked on their elytrae using enamel paint applied with a blunted hypodermic needle while being held still with a light vacuum. Using a set of nine enamel paints (red, orange, yellow, green, blue, silver, gold, brown and white), captured beetles were given a unique color code that did not change throughout the study period. For example, the first beetle caught was marked with a singular red dot. Single dots continued until all colors had been exhausted; at this point, two-color combinations were utilized. An example of the two-color system would be a beetle with a red and an orange dot (see attached Table 2 for color combinations).

Once all invertebrates had been counted and the paint was dry on the beetles, the invertebrates were released 50.0cm from their place of capture. The position of release with respect to the pitfall trap of capture was determined using a random number table (see attached Table 3 for random numbers). In this study, the area within a 50.0cm radius of the trap was divided into twelve sections, similar to a clock face, with “noon” at magnetic North. The random numbers range from 1 to 12. For releases, 12 corresponded to “noon” or magnetic North, and all other numbers corresponded to their respective “clock face” direction. The pitfall trap location was noted for any recaptures and the recaptured beetles were released using the method as described above to the most recent place of capture.

At the conclusion of the study, an analysis was conducted on the community structure of invertebrates adjacent to the trails that included the difference in general abundance of terrestrial invertebrates between the gravel trail and the dirt trail, the abundance of different orders of invertebrates in relation to distance from a trail, and regression analyses. Community structure data was analyzed using JMP 4.0 software (2001). Unfortunately, it was impossible to conduct an analysis on the movement patterns of the indicator beetle both between traps on the same side of the trail and across the trail, for not enough data was collected.


As mentioned previously, the initial premise for this study was to focus on two particular species of beetle. Unfortunately, only one recapture occurred out of 69 marked beetles. The beetle was initially captured in the matrix surrounding the hard-packed dirt trail. Two days after the initial release, the beetle was found in a trap 3.0m South from the original trap. The beetle did not cross the trail.

Fortunately, however, the method for capturing the two beetle species, outlined as the major experiment, allowed for extensive terrestrial invertebrate community structure data to be collected at both trails. A two-tailed t-test analysis (n = 4, p = .05) of the average number of each order of invertebrate found at each site showed that there was significantly fewer (p = .005) invertebrates found that fell into the “other” category including true bugs (Hemiptera), flies (Diptera), caterpillars (Lepidoptera), leafhoppers (Homoptera), centipedes (Scolopendrida), and grasshoppers (Orthoptera). There was no statistical difference found between the rest of the orders. However, there were generally more spiders (Araneae) found at the hard-packed dirt trail and generally more ants (Hymenoptera) and mites (Acari) at the gravel trail.

In addition to a difference in the community structure between trail types, a difference was found in abundance among most invertebrate orders in the overall study area. Although the abundance of mites and the abundance of ants were found not to be significantly different from one another (n = 2, p = .52), they were both generally more abundant than all other orders of invertebrates. Ants constituted 50% of all invertebrates collected. Mites (33%), spiders (11%), beetles (6%), and other invertebrates (1%) comprised the rest of the invertebrates captured (Figure 2). The abundances of each order varied markedly throughout the study period. In late July 2004, all orders showed a general decline in abundance at the study site (Figure 3).

The invertebrate community data also showed some general trends in order abundance in relation to distance from the trails. The hard-packed dirt trail showed an increase in the number of beetles captured with an increase in distance from the trail (Figure 4d). Similarly, an increase in distance from the gravel trail showed a decrease in the number of ants captured in the pitfall traps (Figure 4a). Not much difference was found in the other orders of invertebrates (Figure 4b,c,e). Although there were general trends in the abundance of invertebrates at varying distances from the trails, no statistical difference (n = 4, p = .05) was found (Table 1). The decrease in invertebrate abundance with an increase in distance from the object of disturbance is markedly different from the findings of J. Holmquist (unpublished data, 2004) using the vacuum sampling technique. Holmquist found that an increase in distance from trails showed an increase in the number of invertebrates captured.

The vacuum sampling method employed by Holmquist in his invertebrate study in Devil’s Postpile National Monument illustrated a difference in the numbers and kinds of orders captured from that of the pitfall traps used in this study. Not surprisingly, the vacuum sampling method managed to capture more flying invertebrates like flies and leafhoppers (Figure 5). The pitfall trap method was useful for capturing ants and mites (Figure 5).

Table 1 - Statistical difference between the abundance of invertebrates at a Dirt Trail and a Gravel Trail by distance from trail.


Figure 2 – No statistical difference in overall order abundance was found in community structure near trails in Devil’s Postpile National Monument.

Figure 3 – Ants (a) and mites (b) were similarly more abundant than spiders (c), beetles (d), and all other invertebrates (e) captured over the study period in Devil’s Postpile National Monument. All orders showed a decrease in abundance during the course of the study. Error bars represent one standard error from the mean of 4 replicate areas.


Figure 4 – The hard-packed dirt trail showed an increase in the number of beetles (d) captured with an increase in distance from the trail. The gravel trail showed a decrease in the number of ants (a) captured with an increase in distance from the trail in Devil’s Postpile National Monument. Not much difference was found in the mites (b), spiders (c), or other (e) orders of invertebrates.


Figure 5 – By vacuum net sampling and pitfall trapping terrestrial invertebrates in Devil’s Postpile National Monument two completely different community assemblages are produced. Vacuum sampling captures many more flying insects than pitfall trapping.


The purpose of this study was to determine if trails through meadows in Devil’s Postpile National Monument are having a measurable effect on the local fauna, specifically, terrestrial invertebrates. Unfortunately, the results of this study are quite inconclusive.

The lack of recapture data from the mark-release-recapture study on the ground-active beetles was highly detrimental to the purpose of the study. It is possible that the paint used to mark the elytra of the beetles caused the invertebrates to be subject to increased predation or other sources of mortality. In addition, few beetles overall were captured, possibly caused by a natural decrease in abundance due to the lateness of the season.

The results of the supplemental study in community structure did show a difference between the two trails in invertebrate assemblages. The difference found in the abundance of spiders at the hard-packed dirt trail and the abundance of ants at the gravel trail most likely were dependent on the vegetation type surrounding the trail. The vegetation surrounding the dirt trail was predominated by an abundance of willow (Salix sp.) whereas grasses and sedges dominated the vegetation surrounding the gravel trail. In addition, the availability of moisture may have had an effect on the invertebrate community structure. The dirt trail followed alongside the San Joaquin River while the gravel trail was much drier. Since there are many compounding factors that cause variation in the relationship between invertebrates and vegetation type, it is difficult to make an absolute conclusion (Strong, Lawton, and Southwood 1984).

It is surprising that there was no statistical difference in the abundance of invertebrates in relation to distance from the trail. It is possible, since measurements were only taken 3.0m into the meadow, that a difference would not have been detectible until further out into the meadow. Holmquist and Schmidt-Gengenbach (Sierra Nature Notes, 2003) found detectable differences in invertebrate abundance at distances up to 10m from trails in meadows in Yosemite National Park.

The two different sampling methods, vacuum netting and pitfall trapping, provided very different snapshots of the community structure in the Devil’s Postpile meadows. It is impossible and also unnecessary to determine which sampling method is ultimately more useful than the other. Ideally, using both methods in conjunction would give the most complete representation of invertebrate community structure. Both methods have their drawbacks. The vacuum method cannot be used in wilderness areas due to the mechanized parts, though using pitfall traps involves digging a matrix of holes and thus disturbing wild habitat. Depending on what needs to be sampled, a variety of non-mechanized sampling methods could be used (Jervis and Kidd 1996). For example, a combination of sweep netting and pitfall trapping would be useful in wilderness areas, although both methods are not useful for collecting quantitative data.

Overall, more data is needed to make any conclusions concerning the degree of impact that trails are having on the meadows in Devil’s Postpile National Monument. In addition, further comparisons should be made between vacuum sampling and pitfall trapping data to determine the degree of biases associated with each method.


I would like to thank my mentor Dr. Jeff Holmquist and Jutta Schmidt-Gengenbach for advice and assistance throughout this study. I would also like to thank the National Science Foundation, the University of California at San Diego, and the White Mountain Research Station for funding and the use of facilities for this ecological research.


Table 2 – Color combination chart used for marking captured
beetles in Devil’s Postpile National Monument.


Table 3 – Random numbers for use in releasing captured beetles from pitfall traps in Devil’s Postpile National Monument.

About Us Science Fridays Staff Corie Cann: Bio Kimberly Ogden: References Kimberly Ogden: About Me Kimberly Ogden: Home