The Response of Beetles to Group Selection Harvesting in A

Home , Agonum decorum, Amerinus, Calathus giganteus, Cicindela sexguttata, Climax (beetle), Clivinini

THE RESPONSE OF BEETLES TO GROUP SELECTION HARVESTING IN A

SOUTHEASTERN BOTTOMLAND HARDWOOD FOREST

MICHAEL DARRAGH ULYSHEN

(Under the Direction of James L. Hanula)

ABSTRACT

The environmental protection and sustainable management of our remaining forests are

increasingly important concerns. Group selection harvesting is an uneven-aged forest

management practice that removes patches of desirable trees to create small openings mimicking

natural disturbances. To determine the effects of this technique on beetles, malaise and pitfall traps were placed at the center, edge, and in the forest surrounding artificially created gaps of different size (0.13, 0.26, and 0.50 ha) and age (1 and 7 years) in a South Carolina bottomland hardwood forest. Beetles were generally more abundant and species rich in the centers of younger gaps than in the centers of older gaps or in the forest surrounding them. There were relatively few differences in the abundance and richness of beetles between old gaps and the

surrounding forest but species composition differed considerably. These differences may be

explained by the uneven distribution of various resources.

INDEX WORDS: Coleoptera, logging, coarse woody debris, carabidae, herbivores, silphidae, canopy gaps, swamps, flooding THE RESPONSE OF BEETLES TO GROUP SELECTION HARVESTING IN A

SOUTHEASTERN BOTTOMLAND HARDWOOD FOREST

MICHAEL DARRAGH ULYSHEN

B.S., Miami University, 2002

A Thesis Submitted to the Graduate Faculty of The University of Georgia in Partial Fulfillment

of the Requirements for the Degree

MASTER OF SCIENCE

ATHENS, GEORGIA

2005

Michael Darragh Ulyshen

THE RESPONSE OF BEETLES TO GROUP SELECTION HARVESTING IN A

SOUTHEASTERN BOTTOMLAND HARDWOOD FOREST

MICHAEL DARRAGH ULYSHEN

Major Professor: James L. Hanula

Committee: Joseph McHugh Wayne Berisford

Electronic Version Approved:

Maureen Grasso Dean of the Graduate School The University of Georgia May 2005 iv

ACKNOWLEDGEMENTS

I would like to thank Dr. James Hanula for advice and guidance during this study, as well

as for permitting me to travel while pursuing this degree. I am also grateful for the help of Dr.

Joseph McHugh, Dr. Wayne Berisford, Scott Horn, Dr. John Kilgo, Dr. Christopher Moorman,

Dr. Cecil Smith, Harry Lee Jr., Danny Dyer, Walter Sikora, Lee Reynolds, Nicole Hamilton,

Stephanie Cahill, Ryan Malloy, Josh Campbell, and Francis Brookshire. Finally, I thank and am in particular debt to the Ulyshens and Keisters. v

TABLE OF CONTENTS

Page

ACKNOWLEDGEMENTS...... iv

CHAPTER

1 INTRODUCTION AND LITERATURE REVIEW ...... 1

2 THE RESPONSE OF BEETLES TO CANOPY GAP CREATION IN A

SOUTHEASTERN BOTTOMLAND HARDWOOD FOREST...... 4

3 SPATIAL AND TEMPORAL PATTERNS OF BEETLES ASSOCIATED WITH

COARSE WOODY DEBRIS IN MANAGED BOTTOMLAND HARDWOOD

FORESTS...... 38

4 HERBIVOROUS INSECT RESPONSE TO GROUP SELECTION CUTTING IN A

SOUTHEASTERN BOTTOMLAND HARDWOOD FOREST...... 77

5 THE RESPONSE OF GROUND BEETLES (COLEOPTERA: CARABIDAE) TO

SELECTION CUTTING IN A SOUTH CAROLINA BOTTOMLAND

HARDWOOD FOREST ...... 105

6 USING MALAISE TRAPS TO SAMPLE GROUND BEETLES (COLEOPTERA:

CARABIDAE) ...... 137

7 RESPONSE OF CARRION BEETLES (COLEOPTERA: SILPHIDAE) TO

FLOODING AND GAP CREATION IN A SOUTHEASTERN BOTTOMLAND

HARDWOOD FOREST ...... 154

8 CONCLUSIONS...... 175 1

CHAPTER 1

INTRODUCTION AND LITERATURE REVIEW

Southeastern bottomland hardwood forests are important for water quality and control, nutrient cycling, and wildlife habitat, and support among the most diverse plant and animal communities in North America (Kellison and Young 1997). The amount of remaining bottomland forest continues to decrease, and that which remains has been largely mismanaged.

Virtually all forests in the eastern United States have been logged at least once (Heavrin 1981), and selective removal of only the most desirable trees, has resulted in the comparatively low- value stands existing today (Kellison and Young 1997). Restoring and protecting environmental health has become the priority among progressive foresters and modern

‘ecosystem management’ attempts to produce commodities “not for the capital value they represent, but rather as a byproduct of ecologically based interventions in stands and landscapes”

(Guldin 1996).

How to best manage forests in order to minimize the negative impacts of timber removal will depend, in many cases, on the forest type. This is due to the fact that different forests have different patterns of natural disturbance (Hunter 1990, Guldin 1996). Even-aged management

(e.g. clearcutting, shelterwood cutting, etc.) may be best used in forests, such as conifer forests, that experience infrequent but widespread disturbances like fire (Hunter 1990, Bonnicksen

1994). In contrast, uneven-aged management (e.g. selection cutting) that mimics frequent but 2

localized disturbances may be best in forests, such as deciduous forests, that generally

experience disturbance in the form of individual tree deaths (Hunter 1990).

Group-selection harvesting is an uneven-aged forest management practice that removes

patches of merchantable trees leaving small (< 0.55 ha) openings similar to those created by

insect infestations, severe wind damage or other localized disturbances (Hunter 1990, Meadows

and Stanturf 1997, Guldin 1996). Because it emulates disturbance patterns natural to bottomland

hardwood forests, group-selection harvesting may be a good compromise between protecting

forest health and harvesting valuable timber. This research investigates the response of beetles

to the creation of group-selection gaps of different size (0.13, 0.26, and 0.50 ha) and age (1 or 7

years) in a bottomland hardwood forest in South Carolina, USA.

This thesis consists of six sections. The first examines the overall response of beetles to

gap creation as well as the individual responses of the 40 most common families. The second

looks specifically at the bark and wood-boring families and their dependency upon coarse woody

debris. The third considers the effects of plant succession on leaf beetles and other insect herbivores. The response of ground beetles (Carabidae) and the importance of flight to carabids

in flooded habitats are considered in the fourth and fifth sections, respectively. Finally, the last

section deals with the response of Silphidae to both gap creation and flooding.

References Cited

Bonnicksen, T.M., 1994. Nature’s clearcuts: Lesson from the past. P. 22-28 in Closer

look: An on-the ground investigation of the Sierra Club book, Clearcut. Am. For.

And Pap. Assoc. Washington D.C. 28 p.

Guldin, J. M., 1996. The role of uneven-aged silviculture in the context of ecosystem

management. Western Journal of Applied Forestry 11: 4-13.

Heavrin, C. A., 1981. Boxes, Baskets and Boards: A History of Anderson-Tully Co.

Memphis State Univ. Press, Memphis, TN, 178 pp

Hunter, M. L. 1990. Wildlife, forests, and forestry. Prentice-Hall, Englewood Cliffs,

New Jersey, USA.

Kellison, R. C., Young, M. J., 1997. The bottomland hardwood forest of the southern

United States. Forest Ecology and Management 90: 101-115.

Meadows, J. S., Stanturf, J. A., 1997. Silvicultural systems for southern bottomland

hardwood forests. Forest Ecology and Management 90: 127-140. 4

CHAPTER 2

THE RESPONSE OF BEETLES TO CANOPY GAP CREATION IN A

SOUTHEASTERN BOTTOMLAND HARDWOOD FOREST¹

¹Ulyshen, M. D., Hanula, J. L., Horn, S., Kilgo, J. C., Moorman, C. E. To be submitted to the

Journal of Entomological Science. 5

Introduction

Due to ever-increasing demands for timber products and an increasingly fragmented

landscape, identifying the least disruptive and most sustainable harvesting practices has become

a central concern. This research investigates the response of beetles to the creation of canopy

gaps of different size (0.13, 0.26, and 0.50 ha) and age (1 or 7 years) in a 75-100 year-old

bottomland hardwood forest in South Carolina. The gaps were created by group selection

harvests, an uneven-aged forest management practice that removes patches of merchantable trees leaving small (< 0.55 ha) openings similar to those created by treefalls and wind damage

(Hunter 1990, Meadows and Stanturf 1997, Guldin 1996).

Materials and Methods

This study was conducted on the Savannah River Site (SRS), an 80,269-ha nuclear production facility near Aiken, South Carolina. The SRS is owned and operated by the United

States Department of Energy (DOE), and is managed as a National Environmental Research

Park. The study area used was a 75-100-year-old bottomland hardwood forest approximately

120 ha in size. The forest canopy consisted of bald cypress (Taxodium distichum (L.)), laurel oak (Quercus laurifolia Michaux), willow oak (Q. phellos L.), overcup oak (Q. lyrata Walter), cherrybark oak (Q. falcata var. pagodaefolia Elliott), swamp chestnut oak (Q. michauxii

Nuttall), sweetgum (Liquidambar styraciflua L.), red maple (Acer rubrum L.), and loblolly pine

(Pinus taeda L.). The midstory consisted predominantly of red mulberry (Morus rubra L.), ironwood (Carpinus caroliniana Walter) and American holly (Ilex opaca Aiton). The understory 6

was dominated by dwarf palmetto (Sabal minor (Jacquin) Persoon) and switchcane (Arundinaria

gigantean (Walter) Muhlenberg).

Of the 24 gaps used in this study, 12 were created in December of 1994 (“old gaps”) and

12 in August of 2000 (“young gaps”). Each size of old and young gaps was replicated 4 times.

We defined gap size by the area within the boles of the perimeter trees. At the time of sampling,

the vegetation in the old gaps ranged from 1-8 m in height and consisted of pioneer species such

as sweetgum, sycamore (Platanus occidentalis L.), green ash (Fraxinus pennsylvanica Marshall),

black willow (Salix nigra Marshall), tulip poplar (Liriodendron tulipifera L.), oaks, switchcane,

and dwarf palmetto. Young gaps also contained seedlings and stump sprouts of the tree species

above plus an abundance of fireweed (Erechtites hieracifolia (L.) Rafinesque), blackberries

(Rubus spp.), plumegrass (Erianthus giganteus (Walter) Muhlenberg), and various grass and sedge (Cyperus spp.) species.

Insects were sampled four times in 2001 (17-23 May, 10-16 July, 7-13 September, and 3-

9 November) at three locations (gap center, gap edge, and in the forest 50 m from gap edge) in and around each gap. Overall, we sampled for 4 weeks in each of the 72 locations. Each sample location had a malaise trap (Canopy Trap, Sante Traps, Lexington, KY), suspended from a 3-m pole to capture flying insects, and two pitfall traps to capture ground-dwelling insects (Ulyshen et al. in press). The pitfall traps were placed 5 m apart at each sample location, and were identical to those used by Hanula et al. (2002). The collecting jars for both pitfall and malaise traps were filled with a NaCl-2% formaldehyde solution to preserve specimens and a drop of detergent was added to reduce surface tension (New and Hanula 1998). Once collected, the insects were brought back to the lab and immediately stored in 70 % alcohol. Specimens were sorted and later identified to morphospecies. 7

A 3-way analysis of variance with abundance and richness as response variables and gap age, trap location, and gap size as the main effects showed a significant interaction between gap age and location so we analyzed the data for each age separately. The square-root transformation was used to satisfy assumptions of normality and homogeneity of variance where appropriate.

Data were analyzed using the General Linear Model procedure of SAS (SAS institute, 1985), and the Ryan-Einot-Gabriel-Welsch Multiple Range Test (α<0.05) was used to determine differences in relative abundances of insects between trap locations or gap sizes for each gap age. We used

Raabe’s percent similarity to compare herbivore community similarity among trap locations and gap ages.

Results

We collected a total of 33,423 beetles representing 1,105 species and 74 families (Table

2.1). The five most abundant families accounted for 56.2% of the total and were: Carabidae,

Staphylinidae, Scirtidae, Curculionidae, and Elateridae (Figure 2.2). The five most diverse families accounted for 42.1% of the total number of species and were: Staphylinidae,

Curculionidae, Carabidae, Chrysomelidae, and Elateridae (Figure 2.2).

Malaise and pitfall traps collected 80.2% and 19.8% of total beetles, respectively. Of the

74 families collected, 38 were collected only in malaise traps and 2 were collected only in pitfall

traps (Table 2.1). The remaining 34 families were represented in both traps, but were, in most

cases, much more commonly collected in malaise traps than in pitfall traps.

Overall beetle richness declined from the center of young gaps to the forest, but the

differences among locations were not significant (Figure 2.3). Richness was lower at the center 8 of old gaps than at the edge or in the forest surrounding old gaps, but this was true only for malaise trap samples. No differences were observed among the locations in pitfall samples.

Overall richness was higher in new gap locations than in corresponding locations in old gaps, but forest locations were not different in the malaise trap samples, and there were no differences between gap ages in the pitfall samples (Figure 2.2).

The overall abundance (Figure 2.3) and total biomass (Figure 2.4) of beetles was greater in the center of young gaps than in the edge or in the forest surrounding young gaps, but this was only true for malaise traps. Pitfall traps showed no differences in abundance among locations.

There were no differences observed among old gap locations for either trap type. Overall abundance was higher at all young gap locations than at corresponding old gap locations. For pitfall traps, however, only in the forest locations were more beetles collected in young than in old gaps (Figure 2.3). Total biomass was higher at the centers of young gaps than at the centers of old gaps (Figure 2.4). This was true for both malaise and pitfall trap samples. Additionally, the biomass collected at the edge of young gaps was greater than that collected at the edge of old gaps (Figure 2.4).

No differences in species richness, abundance, or biomass were found among the three gap sizes (Figure 2.5). There is a gradual increase in richness with increasing gap size for young gaps, but the differences were not significant (Figure 2.5A). The centers of young gaps had greater species richness and greater biomass than did the centers of old gaps regardless of size

(Figure 2.5A,C). Abundance was greater at the centers of 0.50 ha young gaps than in old gaps of the same size (Figure 2.5B).

Overall, the greatest percent similarity in beetle communities was found in the forests surrounding young and old gaps (61.1%), while the lowest similarity was found between the 9

center of young gaps and the forest surrounding young gaps (41.3%) (Table 2.2). In contrast, the

centers of old gaps and the forest surrounding old gaps were relatively similar (56.1%). The

centers of young and old gaps were the second least similar (43.5%) (Table 2.2). The beetle

communities collected by pitfalls in the different locations were much more similar (69.6-80.8%)

than were those collected by malaise traps (38.1-53.7%) (Table 2.2).

Individual families were variable in their response to gap creation. Of the 40 most abundant families (>50 individuals), 13 were more abundant in young gaps than in the

surrounding forest (Table 2.3). Only 3 families (Aderidae, Cryptophagidae, and Nitidulidae)

were less abundant in young gaps than in the forest. In contrast, just two families (Anthicidae

and Carabidae) were more abundant in old gaps than in the forest, while eight others were less

abundant (Table 2.3). One family (Throscidae) was less abundant in both young and old gaps

compared to the forest, but no families were more abundant in both young and old gaps (Table

2.3).

Discussion

Although many studies have investigated the response of insects to disturbances, few

have dealt with canopy gaps or group selection, and even fewer have compared the richness and

abundance of insects in canopy gaps to that of the surrounding forest. Although Carabidae

(Koivula and Niemelae 2003), Homoptera (Gorham et al. 2002), and Lepidoptera (Hill et al.

2001) were found to be more abundant in canopy gaps than in closed forest, Shelly (1988) found

the opposite to be true for Coleoptera, Formicidae, and Psocoptera in Panama. Shelly’s results

may be misleading, however, since his traps were operated during the daytime only, and

managed to collect only very small insects. 10

That insect abundance tends to be higher in gaps than in the forest is supported by studies

involving insect predators. For example, group selection harvests increased bat foraging

behavior in a bottomland hardwood forest (Menzel et al. 2002). Similar increases in predator

presence were noted for treefrogs (Cromer et al. 2002, Horn et al. 2004), and birds (Kilgo et al.

1999). Foraging dragonflies were also much more abundant in canopy gaps than in the surrounding forest (pers. obs.).

Increased insect abundance suggests that young canopy gaps contain greater amounts of important resources than does the surrounding forest. The increased availability of light, water, and nutrients in gaps increases plant diversity and net primary productivity, and encourages the growth and regeneration of less shade tolerant species (Bormann and Likens 1979, Boring et al.

1981, Brokaw 1982, Phillips and Shure 1990, Shure and Phillips 1991, Wilder et al. 1999).

Other factors encouraging herbivory in gaps include increased nutrient levels (i.e. soluble nitrogen) in plant tissues, fewer plant defenses in short-lived pioneer species, and increased consumption and growth rates of herbivorous insects that feed on plants receiving direct sunlight

(Boardman 1977, White 1978, Scriber and Slansky 1981, Coley 1983, Harrison 1987).

Another resource that is likely to increase above intact-forest levels following gap creation is coarse woody debris (CWD). CWD provides food and habitat for many animal species and can be a long-term source of nutrients for both plants and animals (Harmon et al.,

1986; Hagan and Grove, 1999). Elton (1966) considered it to be one of the three greatest resources for animal species in natural forests and Huston (1996) claimed that CWD has a greater impact on biodiversity than does any other manageable forest property. The large quantity of CWD created during timber removal should encourage the colonization of many wood-inhabiting organisms. 11

There are also noticeable changes in the resource base available to insects in gaps of

different age. The fact that the abundance and richness of insects was higher in young gaps than

in old gaps suggests that there are fewer usable resources in old gaps. This was definitely the case for CWD, which was virtually nonexistent in old gaps compared to young gaps. The taxonomic and structural diversity of plants in the two gap ages also differed considerably.

While young gaps contained an abundance of young herbaceous growth, old gaps were largely dominated by dense stands of young trees, consisting of just a few species. Given these differences in resource availability, it is not surprising that the insect communities differed in species composition (being only 43.5% similar) as well as in abundance and richness. While the species composition of insects in new gaps was only 41.3% similar to that in the nearby forest, old gaps and their corresponding forest locations had 56.1% similar insect communities. This suggests that after 7 years the insect community has become more forest-like in composition, indicating a gradual recovery.

The size of canopy gaps should influence both the presence and abundance of certain resources as well as the consumers of those resources. Larger canopy gaps, for example, tend to receive greater amounts of sunlight, and consequently have greater amounts of vegetation than

do smaller gaps (Shure and Phillips 1991). However, there is some evidence that the vegetation

in smaller gaps is more palatable (less defensive compounds, and softer leaves) than that in

larger gaps (Shure and Wilson 1993). This is supported by a study in which herbivorous insect

load (mg of herbivores/g of foliage) was found to be higher in smaller gaps (0.016 ha) than in

larger gaps (2.0 and 10.0 ha) or in the surrounding forest (Shure and Phillips 1991). In addition,

small gaps appear to provide increased protection from both invertebrate and vertebrate

predators- both of which are more abundant in increasingly large gaps and in the surrounding 12 forest (Forman et al. 1976, Shure and Phillips 1991). Shure and Phillips (1991) found that the abundance and richness of insects was lower in mid-sized gaps (0.08 and 0.4 ha) than in smaller or larger gaps. The authors attribute this to patchy vegetation and higher soil and air temperatures.

Because gaps of different size differ in the quantity and quality of vegetation, the distribution of plants, and the soil temperature, it becomes difficult to predict differences in insect abundance among different sized gaps. The fact that we found no differences in beetle richness or abundance between gaps of different size may be due to the relatively small size range under consideration. Another complicating factor is the potential bias of passive collecting techniques (Malaise and Pitfall trapping). These traps may only sample the immediate vicinity without accurately reflecting differences in total richness or abundance among different sized gaps.

The edge separating forest plant communities from those of canopy gaps is very distinct, and contains a mixture of vegetation from both habitats. Although it is widely believed that edges support great diversities of wildlife, there is relatively little experimental evidence in support of a distinct edge effect (Yahner 1988). While there may be greater concentrations of insects near the edges between habitats, it may be relatively difficult to detect. Some research suggests that edges do not delineate the ranges of many insects, but act instead as “broad transition zones” (Dangerfield et al. 2003). We placed traps 50 meters into the forest from the edge of each gap, but our results suggest that at this distance the beetle community was still substantially influenced by the gaps. For example, we recorded greater beetle richness and abundance in the forests surrounding new gaps than in the forests surrounding old gaps (Figures

2.3 and 2.4). Given this, it seems reasonable to suppose that our “forest” sites may have actually 13 been within a broad transition zone, as described by Dangerfield et al. (2003), and should be considered part of the edge separating interior forest from gap.

The response of beetles collected in pitfall traps to gap creation was not as strong as that of those collected in malaise traps. This may be due partly to the fact that malaise traps collected many times more beetles than did the pitfall traps. Another possibility is that flying insects may be attracted to the increased light levels and higher temperatures of gaps. If so, one would expect malaise traps, but not pitfall traps, to collect greater numbers of insects in gaps, regardless of differences in resource availability. While this may explain differences observed among locations for some families, differences in the availability of certain resources (i.e. CWD, vegetation) among locations must surely affect the abundance of specific families (i.e. woodboring groups, herbivores). Regardless of the precise cause, it is clearly the case that young canopy gaps contain greater densities of beetles than do old gaps or the nearby forest.

Group selection appears to be a relatively low-impact harvesting technique. Overall, young gaps support a greater richness and abundance of beetle species than does the surrounding forest, while the richness and abundance of species in older gaps is comparable to forested sites.

Individually, most families were either more or equally abundant in young gaps than in the surrounding forest. Although several families were less abundant in old gaps than in the forest

(especially those that depend on CWD), the majority were as abundant in old gaps as in the forest. Only one family (Throscidae) was less abundant in both young and old gaps. Based on our insect data, group selection seems to minimize the impact harvesting has on the forest and its wildlife, and warrants further consideration for use in bottomland hardwood forests.

References Cited

Boardman, N. K., 1977. Comparative photosynthesis of sun and shade plants. Annual

Review of Plant Physiology 28: 355-377.

Bonnicksen, T.M., 1994. Nature’s clearcuts: Lesson from the past. P. 22-28 in Closer

look: An on-the ground investigation of the Sierra Club book, Clearcut. Am. For.

And Pap. Assoc. Washington D.C. 28 p.

Boring, L. R., Monk, C. D., Swank, W. T., 1981. Early regeneration of a clear-cut

southern Appalachian forest. Ecology 62: 1244-1253.

Bormann, F. H., Likens, G. E., 1979. Pattern and process in a forested ecosystem.

Springer-Verlag, New York, New York, USA.

Brokaw, N. V. L. 1982. Treefalls: frequency, timing and consequences. In: Leigh EG,

Rand AS, Windsor DM (eds) The ecology of a tropical forest. Seasonal rhythms and

long-term changes. Washington, DC: Smithsonian Institute.

Buckner, C. A., Shure, D.J., 1985. The Response of Peromyscus to forest opening size in

the southern Appalachian Mountains. Journal of Mammalogy. 66: 299-307.

Coley, P. D. 1983. Herbivory and defensive characteristics of tree species in a lowland

tropical forest. Ecological Monographs 53: 209-233.

Cromer R. B., Lanham, J. D., Halin H. H., 2002. Herpetofaunal Response to gap and

skidder-rut wetland creation in a southern bottomland hardwood forest. Forest Science.

48:407-413.

Dangerfield, J. M., Pik, A. J., Britton, D., Holmes, A., Gillings, M., Oliver, I., Briscoe, 15

D., Beattie, A. J., 2003. Patterns of invertebrate biodiversity across a natural edge.

Austral Ecology. 28: 227-236.

Elton, C. S., 1966. Dying and dead wood. In: “The Pattern of Animal Communities.”

279-305. Wiley, New York.

Forman, R. T. T., Galli, A. E., Leck, C. F., 1976. Forest size and avian diversity in New

Jersey woodlots with some land use implications. Oecologia 26: 1-8.

Gorham, L. E., King, S. L., Keeland, B. D., Mopper, S., 2002. Effects of Canopy Gaps

and Flooding on Homopterans in a Bottomland Hardwood Forest. Wetlands 22: 541-549.

Guldin, J. M., 1996. The role of uneven-aged silviculture in the context of ecosystem

management. Western Journal of Applied Forestry 11: 4-13.

Hagan, J.M., Grove, S.L., 1999. Coarse woody debris. Journal of Forestry. January: 6-11.

Hanula, J. L., Meeker, J. R., Miller, D. R., Barnard, E. L., 2002. Association of wildfire

on tree health and numbers of pine bark beetles, reproduction weevils, and their

associates in Florida. Forest Ecology and Management 170:233-247.

Harmon, M.E., Franklin, J.F., Swanson, F.J., Sollins, P., Gregory, S.V., Lattin, J.D.,

Anderson, N.H., Cline, S.P., Aumen, N.G., Sedell, J.R., Lienkaemper, G.W., Cromack,

J.R., Cummins, K.W., 1986. Ecology of coarse woody debris in temperate ecosystems.

Advancess in Ecological Research. 15: 133-302.

Heavrin, C. A., 1981. Boxes, Baskets and Boards: A History of Anderson-Tully Co.

Memphis State Univ. Press, Memphis, TN, 178 pp

Hill, J., Hamer, K., Tangah, J., Dawood, M. 2001. Ecology of tropical butterflies in

rainforest gaps. Oecologia 128: 294-302.

Horn, S., Ulyshen, M. D., Hanula, J. L., (2004 in review) Abundance of Green Tree 16

Frogs and Insects in Artificial Canopy Gaps in a Bottomland Hardwood Forest. The

American Midland Naturalist.

Hunter, M. L. 1990. Wildlife, forests, and forestry. Prentice-Hall, Englewood Cliffs,

New Jersey, USA.

Huston, M.A., 1996. Modeling and management implications of coarse woody debris

impacts on biodiversity. In: McMinn, J., Crossley, D.A., eds. 1996. Biodiversity and

coarse woody debris in southern forests. Athens Ga: USDA For. Ser. Gen. Tech. Rep.

SE-94.

Kellison, R. C., Young, M. J., 1997. The bottomland hardwood forest of the southern

United States. Forest Ecology and Management 90: 101-115.

Kellison, R. C., Martin, J. P., Hansen, G. D., Lea, R., 1988. Regenerating and managing

natural stands of bottomland hardwoods. APA88-A-6 Amer. Pulpwood Assoc., 26 pp.

Kilgo, J. C., Miller, K. V., Smith, W. P., 1999. Effects of group-selection timber harvest

in bottomland hardwoods on fall migrant birds. Journal of Field Ornithology. 70:404-413.

Koivula, M., Niemelae, J., 2003. Gap felling as a forest harvesting method in boreal

forests: responses of carabid beetles (Coleoptera, Carabidae). Ecography. 26: 179-187.

Maser, C. 1994. Sustainable Forestry. St. Lucie Press, Delray Beach, FL, 373 pp.

Meadows, J. S., Stanturf, J. A., 1997. Silvicultural systems for southern bottomland

hardwood forests. Forest Ecology and Management 90: 127-140.

Menzel, M. A., Carter, T. C., Menzel, J. M., Ford, M. F., Chapman, B. R., 2002. Effects

of group selection silviculture in bottomland hardwoods on the spatial activity

pattern of bats. Forest Ecology and Management. 162:209-218.

New, K.C., Hanula, J.L., 1998. Effect of time elapsed after prescribed burning in 17

longleaf pine stands on potential prey of the Red-Cockaded Woodpecker. SJAF 22(3):

175-183.

Phillips, D. L., Shure, D. J., 1990. Patch-size effects on early succession in

southern Appalachian forests. Ecology 71: 204-212.

SAS INSTITUTE, 1985. SAS guide for personal computers. Version 6 ed., SAS

Institute, Cary, NC. 378 pp.

Scriber, J. M., Slansky Jr., F., 1981. The nutritional ecology of immature insects.

Annual Review of Entomology 26: 183-211.

Shelly, T. E., 1988. Relative Abundance of Day-Flying insects in Treefall Gaps vs.

Shaded Understory in a Neotropical Forest. Biotropica 20(2): 114-119.

Shure, D. J., Wilson, L. A., 1993. Patch-size effects on plant phenolics in successional

openings of the southern Appalachians. Ecology. 74: 55-67.

Shure, D. J., Phillips, D. L., 1991. Patch size of forest openings and arthropod

populations. Oecologia. 86: 325-334.

White, T. C. R., 1978. The importance of a relative shortage of food in animal ecology.

Oecologia (Berlin) 33: 71-86.

Wilder, C. M., Holtzclaw Jr., F. W., Clebsch, E. E. C., 1999. Succession, Sapling

density and growth in canopy gaps along a topographic gradient in second growth east

Tennessee forest. American Midland Naturalist 142: 201-212.

Yahner, R. H., 1988. Changes in Wildlife Communities Near Edges. Conservation

Biology. 2: 333-339.

Table 2.1. The abundance and richness of beetle families collected in malaise and pitfall traps at the Savannah River Site, South Carolina, in 2001.

Family Number of Individuals Total Number of Number of

Malaise/Pitfall Individuals Species

Aderidae 100/1 101 13

Anobiidae 196/0 196 16

Anthicidae 79/115 194 9

Anthribidae 86/1 87 15

Attelabidae 7/0 7 2

Biphyllidae 7/0 7 1

Bostrichidae 9/0 9 3

Bothrideridae 3/0 3 2

Brentidae 58/0 58 1

Bruchidae 2/0 2 1

Buprestidae 601/2 603 17

Cantharidae 143/0 143 19

Carabidae 1430/4068 5498 88

Cebrionidae 1/0 1 1

Cerambycidae 1136/11 1147 51

Ceratocanthidae 45/0 45 2

Cerylonidae 4/0 4 2

Chrysomelidae 1446/35 1481 71 19

Ciidae 19/4 23 7

Clambidae 1/0 1 1

Cleridae 376/0 376 16

Coccinellidae 184/0 184 17

Colydiidae 26/1 27 13

Corylophidae 50/0 50 7

Cryptophagidae 589/4 593 6

Cupedidae 3/0 3 1

Curculionidae 2603/261 2864 127

Dermestidae 6/1 7 4

Dryopidae 0/2 2 1

Dytiscidae 197/38 235 15

Elateridae 1622/12 1634 55

Elmidae 1/0 1 1

Endomychidae 32/20 52 10

Erotylidae 76/21 97 14

Eucinetidae 1/1 2 2

Eucnemidae 284/1 285 13

Heteroceridae 218/3 221 2

Histeridae 275/15 290 19

Hydraenidae 146/0 146 2

Hydrophilidae 129/35 164 18

Laemophloeidae 25/0 25 6 20

Lampyridae 127/7 134 9

Languridae 3/0 3 1

Latridiidae 63/0 63 8

Leiodidae 16/6 22 7

Limnichidae 1/0 1 1

Lycidae 296/0 296 8

Melandryidae 123/2 125 16

Meloidae 10/0 10 2

Melyridae 56/0 56 8

Monommatidae 35/0 35 2

Monotomidae 22/0 22 2

Mordellidae 708/5 713 23

Mycetophagidae 21/0 21 11

Nitidulidae 55/223 278 22

Oedemeridae 32/0 32 4

Passalidae 0/13 13 1

Passandridae 5/0 5 2

Phalacridae 139/3 142 14

Phengodidae 2/0 2 1

Psephenidae 4/0 4 1

Ptiliidae 9/0 9 4

Ptilodactylidae 1270/11 1281 1

Pyrochroidae 7/0 7 1 21

Scarabaeidae 270/675 945 50

Scirtidae 3688/2 3690 9

Scydmaenidae 1053/57 1110 14

Silphidae 1/0 1 1

Silvanidae 50/1 51 9

Sphindidae 13/0 13 4

Staphylinidae 4186/896 5082 144

Tenebrionidae 1515/52 1567 37

Throscidae 793/0 793 9

Trogossitidae 11/0 11 3

Unidentified 12/6 18 5

Totals 26812/6611 33423 1105

Table 2.2. Raabe’s percent similarity of beetles collected in young (~1 yr.) or old (~7 yrs) canopy gaps by location (center, edge, or 50 m into surrounding forest) in a South Carolina bottomland hardwood forest, 2001.

Comparison Trap type

Malaise Pitfall Total

New Forest vs. Old Forest 53.7±2.8 76.1±2.8 61.1±1.9

Old Gap vs. Old Edge 45.6±1.9 79.3±2.7 56.8±2.2

Old Gap vs. Old Forest 40.4±2.1 69.6±3.9 56.1±2.3

New Edge vs. Old Edge 42.8±2.8 80.8±1.7 50.4±2.8

New Gap vs. New Edge 46.0±3.8 78.4±0.6 48.7±3.0

New Gap vs. Old Gap 45.0±.0.7 79.2±1.3 43.5±1.5

New Gap vs. New Forest 38.1±1.7 80.3±2.4 41.3±1.6 23

Table 2.3. The mean (n=12) abundance (above) and richness (below) of common (> 50 individuals) beetle families collected at the

center, edge, and in the surrounding forest of artificial canopy gaps in a bottomland hardwood forest, South Carolina, 2001. Richness

is not included for those families that were represented by less than 10 species. The young and old gaps were created in 2000 and

1994, respectively. For each gap age, values with the same letter are not significantly different (Ryan-Einot-Gabriel-Welsch Multiple

Range Test, p<0.05). Asterisks denote significant differences (p<0.05) between the same trap locations (e.g. center vs. center) in

young and old gaps. A prime (‘) symbol indicates that the square root transformation was used.

Young Gaps Old Gaps

Gap Edge Forest Gap Edge Forest

0.67±0.19 B 1.17±0.27 B 2.58±0.67 A 0.58±0.29 a 1.58±0.81 a 1.83±0.82 a Aderidae 1.08±0.08 A 1.16±0.11 A 1.58±0.23 A 1.00±0.00 a 1.33±0.19 a 1.25±0.18 a

2.17±0.81 A* 3.50±1.51 A 4.67±1.71 A 0.75±0.46 a* 4.17±1.78 a 1.08±0.36 a Anobiidae 1.33±0.14 B 1.17±0.11 B 2.00±0.33 A* 1.33±0.26 a 1.33±0.22 a 1.17±0.11 a*

Anthicidae 5.42±4.07 A 0.42±0.15 A 0.67±0.22 A 7.83±3.66 a’ 1.00±0.37 b’ 0.83±0.44 b’

Anthribidae 3.42±0.63 A* 1.00±0.21 B 1.08±0.29 B 0.33±0.19 a* 0.58±0.26 a 0.83±0.27 a

2.67±0.28 A* 1.25±0.13 B 1.42±0.19 B 1.00±0.00 a* 1.08±0.08 a 1.17±0.17 a

Brentidae 3.50±0.90 A* 0.92±0.36 B* 0.17±0.11 B 0 a* 0.08±0.08 a* 0.17±0.11 a 24

Buprestidae 35.17±4.60 A* 5.92±1.36 B 3.75±0.90 B 2.08±0.57 a* 1.42±0.73 a 1.91±0.78 a

3.67±0.43 A* 2.33±0.33 B* 2.25±0.41 B 1.50±0.26 a* 1.25±0.13 a* 1.33±0.19 a

Cantharidae 3.33±0.54 A 1.58±0.43 A 2.50±0.68 A 1.33±0.47 a 1.00±0.33 a 2.17±0.91 a

2.08±0.23 A* 1.42±0.19 A 1.67±0.19 A 1.17±0.11 a* 1.08±0.08 a 1.58±0.23 a

Carabidae 111.08±16.07 A 88.17±12.86 A 70.25±10.49 A* 84.17±15.44 a 62.00±11.67 ab 42.50±5.65 b*

22.67±1.30 A* 20.25±1.07 AB* 17.00±1.55 B 13.33±1.34 a* 14.92±1.52 a* 14.08±0.96 a

Cerambycidae 45.5±10.14 A* 22.75±4.13 B 15.83±6.07 B 1.58±0.31 b*’ 4.92±1.61 a’ 5.00±0.73 a’

8.83±1.28 A* 7.75±0.98 A* 6.00±0.66 A* 1.67±0.22 b* 2.75±0.55 ab* 3.58±0.54 a*

Chrysomelidae 30.25±6.14 A 23.33±6.95 A 14.33±2.54 A 18.58±2.69 a 19.33±4.54 a 17.58±2.67 a

9.92±0.65 A* 6.50±0.72 B 5.67±0.50 B 5.92±0.51 a* 5.42±0.47 a 5.92±0.50 a

Cleridae 17.08±2.93 A 5.08±0.82 B 3.67±0.57 B 0.58±0.19 b 1.75±0.41 ab 3.17±0.80 a

3.33±0.26 A* 3.00±0.41 AB* 2.17±0.27 B 1.00±0.00 b* 1.58±0.19 b* 2.33±0.33 a

Coccinellidae 10.50±1.88 A* 1.17±0.27 B* 1.75±0.45 B 0.25±0.13 b* 0.42±0.19 b* 1.25±0.33 a

1.58±0.23 A* 1.17±0.11 A 1.50±0.23 A 1.00±0.00 a* 1.08±0.08 a 1.25±0.13 a

Corylophidae 1.50±0.97 A 2.00±1.91 A 0.42±0.23 A 0 a 0.17±0.11 a 0.08±0.08 a

Cryptophagidae 4.58±1.55 B* 7.92±1.29 B 13.33±1.88 A* 4.58±1.28 b 11.08±2.43 a 7.92±1.73 ba*

Curculionidae 47.58±9.50 B* 88.50±20.86 A* 40.83±3.27 B* 10.83±1.35 b* 22.92±2.52 a* 28.00±3.65 a*

9.42±0.80 B* 12.83±0.97 A 12.83±0.92 A 6.33±0.76 b* 10.83±0.98 a 12.17±0.86 a 25

Dytiscidae 4.58±0.74 A 2.42±0.60 A 2.50±0.82 A 3.42±1.20 a 3.17±0.64 a 3.50±0.70 a

2.25±0.33 A 1.75±0.28 A 1.50±0.26 A 1.83±0.27 a 1.75±0.22 a 2.00±0.25 a

Elateridae 50.00±11.09 A* 18.08±2.18 B 29.50±3.35 B 3.50±0.48 c* 12.50±1.68 b 22.58±3.84 a

7.33±0.84 A* 6.83±0.56 A 7.08±0.60 A 2.75±0.30 b* 5.42±0.62 a 6.42±0.65 a

Endomychidae 0.67±0.28 A 0.83±0.51 A 1.08±0.31 A 0.17±0.17 a 0.58±0.19 a 1.00±0.48 a

1.00±0.00 A 1.08±0.08 A 1.25±0.13 A 1.08±0.08 a 1.08±0.08 a 1.00±0.00 a

Erotylidae 1.33±0.33 A 1.33±0.36 A 1.42±0.45 A 1.00±0.43 a 1.42±0.36 a 1.58±0.53 a

1.17±0.17 A 1.25±0.13 A 1.58±0.26 A 1.33±0.22 a 1.50±0.19 a 1.58±0.34 a

Eucnemidae 4.17±2.65 A 3.33±1.25 A 7.42±1.47 A 1.25±0.84 b 1.58±0.65 b 6.00±1.17 a

1.33±0.14 B 1.58±0.15 AB 2.17±0.32 A 1.25±0.18 b 1.33±0.19 a 2.08±0.23 a

Heteroceridae 2.75±1.09 AB 6.75±2.48 A 1.08±0.36 B 4.33±2.14 b 2.92±1.29 b 0.58±0.29 a

Histeridae 2.92±1.31 A 1.75±0.41 A 4.08±1.39 A 1.50±0.85 a 9.67±5.37 a 4.25±1.57 a

1.25±0.13 A 1.25±0.18 A 1.58±0.23 A 1.50±0.29 a 1.75±0.28 a 1.42±0.23 a

Hydraenidae 0.42±0.26 A 0.42±0.26 A 7.25±5.11 A 0.58±0.42 a 2.50±1.26 a 1.00±0.83 a

Hydrophilidae 2.17±0.41 A 3.00±0.85 A 3.75±2.06 A 1.83±0.72 a 1.58±0.34 a 1.33±0.26 a

1.75±0.22 A 2.00±0.35 A 1.33±0.19 A 1.42±0.26 a 1.42±0.19 a 1.25±0.13 a

Lampyridae 3.83±0.86 A 1.25±0.41 B 1.67±0.53 AB 2.33±0.78 a 1.33±0.43 a 0.75±0.28 a

Lathridiidae 1.17±0.55 A* 0.92±0.31 A 1.33±0.54 A 0 a* 0.75±0.46 a 1.08±0.48 a 26

Lycidae 7.83±1.00 A 5.33±1.28 AB 3.00±0.66 B 3.33±0.58 a 2.50±0.42 a 2.67±0.58 a

Melandryidae 2.75±0.76 A 2.25±0.58 A 1.33±0.47 A 2.33±1.23 a 1.08±0.34 a 0.67±0.22 a

2.08±0.34 A* 1.92±0.36 A 1.25±0.13 A 1.25±0.13 a* 1.42±0.19 a 1.17±0.11 a

Melyridae 0.25±0.13 A 0.75±0.33 A 2.08±0.92 A* 0.33±0.14 a 0.92±0.36 a 0.33±0.19 a*

13.17±1.85 A* 8.58±1.45 A 13.42±1.76 A 4.25±1.07 b* 11.33±1.62 a 8.67±1.47 a Mordellidae 3.67±0.58 A 3.25±0.48 A 3.50±0.38 A 2.42±0.48 a 3.08±0.36 a 3.00±0.33 a

Nitidulidae 1.00±0.41 B 2.00±0.30B 5.33±0.96 A 5.58±1.76 a 3.00±0.91 a 6.25±1.67 a

1.17±0.17 A 1.67±0.19 A 1.83±0.24 A 1.67±0.22 a 1.50±0.26 a 1.75±0.22 a

Phalacridae 10.25±3.63 A* 0.67±0.35 B 0.17±0.11 B 0.42±0.19 a* 0.17±0.17 a 0.17±0.11 a

1.75±0.39 A 1.08±0.08 A 1.00±0.00 A 1.08±0.08 a 1.00±0 a 1.00±0 a

Ptilodactylidae 35.92±8.97 A* 13.00±3.09 B* 5.08±1.44 B 17.42±3.50 a* 22.00±3.02 a* 13.33±5.56 a

Scarabaeidae 11.17±1.40 A 10.67±1.23 A 14.25±3.64 A 8.83±1.00 a 23.75±9.27 a 10.08±2.82 a

4.58±0.40 A 5.58±0.54 A 4.33±0.53 A 4.83±0.46 ab 6.17±0.90 a 3.08±0.48 b

Scirtidae 196.83±77.24 A 30.50±10.96 B 18.67±3.66 B 31.50±8.31 a 12.67±3.66 a 17.33±3.92 a

Scydmaenidae 10.75±4.14 A 38.92±16.04 A 10.58±3.48 A 5.83±1.85 ab’ 23.08±10.38 a’ 3.33±1.08 b’

1.58±0.26 A 1.75±0.30 A 1.92±0.31 A 1.92±0.36 a 2.08±0.38 a 1.42±0.15 a

Silvanidae 1.25±0.52 AB*’ 1.83±0.74 A’ 0.17±0.11 B’ 0.08±0.08 a* 0.42±0.19 a 0.50±0.29 a

Staphylinidae 115.42±30.62 A* 80.92±23.14 A 70.75±14.00 A 54.58±7.70 a* 58.00±13.03 a 43.83±9.54 a 27

17.58±2.39 A 16.17±1.77 A 17.83±1.62 A 16.67±1.63 a 17.25±1.78 a 14.50±1.12 a

Tenebrionidae 43.00±5.11 A* 23.00±7.64 B 24.92±3.40 B 7.67±1.50 a* 14.25±3.09 a 17.75±5.23 a

3.75±0.46 A 3.25±0.41 A 3.00±0.37 A 3.08±0.47 a 2.92±0.42 a 2.42±0.19 a

Throscidae 6.42±3.03 B 7.75±6.22 B 34.42±11.92 A* 2.25±0.76 b 5.17±1.87 ab 10.08±2.34 a*

Fig. 2.1. The five most species rich and abundant families collected in a bottomland hardwood forest, South Carolina, in 2001. 29

C a r Richness a Chr b Curculionidae i d ys a om e E la e te li rid dae ae Staphylinidae

Other

Staphylinidae Abundance S ci rt id a e Carabidae

Curculionidae

dae Elateri Other 30

Fig. 2.2. Mean±SE (n=12) richness of beetles collected in malaise and pitfall traps in 2001 at different locations in bottomland hardwood forest gaps created in 1994 and 2000. A: both malaise and pitfall trap samples, B: malaise trap samples, C: pitfall trap samples. Within graphs

(for each treatment), bars with the same letter above them are not significantly different (Ryan-

Einot-Gabriel-Welsch Multiple Range Test, p < 0.05). Asterisks denote significant differences (p

< 0.05) between the same trap locations (e.g. center vs. center) in old (1994) and new (2000) gaps.

160 A A* Malaise and Pitfall 140 A* A* 120 a* a* 100 b* 80

0 160 B Malaise 140 A* 120 A* 100 A a* a 80 b* Mean Richness (n=12) 60 40 20 0 C Pitfall 40 A A A 30 a a a 20

0 Center Edge Forest Center Edge Forest 2000 Treatment 1994 Treatment Location

Fig. 2.3. Mean±SE (n=12) abundance of beetles collected in malaise and pitfall traps in 2001 at

different locations in bottomland hardwood forest gaps created in 1994 and 2000. A: both

malaise and pitfall trap samples, B: malaise trap samples, C: pitfall trap samples. Within graphs

(for each treatment), bars with the same letter above them are not significantly different (Ryan-

Einot-Gabriel-Welsch Multiple Range Test, p < 0.05). Asterisks denote significant differences (p

< 0.05) between the same trap locations (e.g. center vs. center) in old (1994) and new (2000) gaps.

1200 A Malaise and Pitfall A* 1000

800

B* 600 B*

a* 400 a* a*

200

0 A* B Malaise 800

600 B*

Mean Abundance (n=12) Abundance Mean B* 400 a* a* a* 200

0 160 C Pitfall 140 a 120 A* A A a 100 80 a* 60 40 20 0 Center Edge Forest Center Edge Forest 2000 Treatment 1994 Treatment Location

Fig. 2.4. Mean±SE (n=12) biomass of beetles collected in malaise and pitfall traps in 2001 at different locations in bottomland hardwood forest gaps created in 1994 and 2000. A: both malaise and pitfall trap samples, B: malaise trap samples, C: pitfall trap samples. Within graphs

(for each treatment), bars with the same letter above them are not significantly different (Ryan-

Einot-Gabriel-Welsch Multiple Range Test, p < 0.05). Asterisks denote significant differences (p

< 0.05) between the same trap locations (e.g. center vs. center) in old (1994) and new (2000) gaps. Prime symbols (‘) indicate that the log transformation was used.

8 A A*' Malaise and Pitfall

a' a' a*' 2

0 B A*' Malaise 5 B' 4 3

Mean Biomass (n=12) Biomass Mean 2 B*' a' 1 a*' a*' 0 C Pitfall 3 A* A A a 2 a a*

0 Center Edge Forest Center Edge Forest 2000 Treatment 1994 Treatment Location

Fig. 2.5. Mean±SE (n=4) Richness (A), Abundance (B), and Weight (C) of beetles collected in malaise and pitfall traps in 2001 at the centers of different sized (0.13, 0.26, and 0.50 ha) bottomland hardwood forest gaps created in 1994 and 2000. Within graphs (for each treatment), bars with the same letter above them are not significantly different (Ryan-Einot-Gabriel-Welsch

Multiple Range Test, p < 0.05). Asterisks denote significant differences (p < 0.05) between the same trap locations (e.g. center vs. center) in old (1994) and new (2000) gaps.

180 A A* 160 A* Richness A* 140

120 a* 100 a* a* 80

0 B A 1200 Abundance A* 1000 A

800

600

a 400 a Mean Value (n=12) Value Mean a* 200

0 C Biomass 10 A*

8 A* A* 6

4 a* a* a* 2

0 0.13 ha 0.26 ha 0.50 ha 0.13 ha 0.26 ha 0.50 ha 2000 Treatment 1994 Treatment Gap Size

CHAPTER 3

SPATIAL AND TEMPORAL PATTERNS OF BEETLES ASSOCIATED WITH COARSE

WOODY DEBRIS IN MANAGED BOTTOMLAND HARDWOOD FORESTS¹

¹Ulyshen, M. D., Hanula, J. L., Horn, S., Kilgo, J. C., Moorman, C. E., 2004. Accepted by

Forest Ecology and Management.

Reprinted here with permission of publisher.

Abstract

Malaise traps were used to sample beetles in artificial canopy gaps of different size (0.13, 0.26,

0.50 ha) and age in a South Carolina bottomland hardwood forest. Traps were placed at the

center, edge, and in the surrounding forest of each gap. Young gaps (~ 1 yr) had large amounts

of coarse woody debris compared to the surrounding forest, while older gaps (~ 6 yrs) had virtually none. The total abundance and diversity of wood-dwelling beetles (Buprestidae,

Cerambycidae, Brentidae, Bostrichidae, and Curculionidae (Scolytinae and Platypodinae)) was

higher in the center of young gaps than in the center of old gaps. The abundance was higher in the center of young gaps than in the surrounding forest, while the forest surrounding old gaps and the edge of old gaps had a higher abundance and diversity of wood-dwelling beetles than did the center of old gaps. There was no difference in wood-dwelling beetle abundance between gaps of different size, but diversity was lower in 0.13 ha old gaps than in 0.26 ha or 0.50 ha old gaps. We suspect that gap size has more of an effect on woodborer abundance than indicated here because malaise traps sample a limited area. The predaceous beetle family Cleridae showed a very similar trend to that of the woodborers. Coarse woody debris is an important resource for many organisms, and our results lend further support to forest management practices that preserve coarse woody debris created during timber removal.

1. Introduction

Coarse woody debris (CWD) provides food and habitat for many animal species and can be a long-term source of nutrients for both plants and animals (Harmon et al., 1986; Hagan and

Grove, 1999). Elton (1966) considered it to be one of the three greatest resources for animal species in natural forests and Huston (1996) claimed that CWD has a greater impact on biodiversity than does any other manageable forest property. Because of its documented value and the fact that it is likely to have other functions in ecosystems yet to be discovered, management of CWD is an increasingly important consideration for industrial and public forests

(Harmon et al., 1986)

That beetles associated with dead and dying wood typically increase in abundance following logging operations has long been known. For example, in his 12 years collecting beetles in the tropics, Wallace (1869) found recent clearings to be the most productive areas for collecting cerambycids and other woodborers. The large amounts of CWD created by timber removal is short-lived, however, and is followed by extended periods without replacement until the regenerating forest enters the stem exclusion stage (Hagan and Grove, 1999). How the relative abundance of wood-dwelling beetles changes with time in these disturbed areas is not well understood.

We consider six families of beetles. Buprestidae, Cerambycidae, Bostrichidae, Brentidae

(genus Arrhenodes), and Curculionidae (subfamilies Scoytinae and Platypodinae) all generally depend on dead or dying wood, and Cleridae (“checkered beetles”) are predaceous on the adults and larvae of many bark and woodboring beetles.

Among the inhabitants of CWD, some of the most conspicuous are woodborers belonging to the families Cerambycidae and Buprestidae. While some feed on living trees, most are found on dead or dying wood (Bellamy and Nelson, 2002; Warriner et al., 2002; Fellin, 1980;

Solomon, 1995). Some species are very host specific while others, largely those that feed on

dead wood, tend to have wide host ranges (Yanega, 1996). Many species show specificity for

such factors as bark thickness, height, and branch diameter which reduces competition between

species thereby increasing diversity on a single resource (Yanega, 1996; Hanula, 1996; Bellamy

and Nelson, 2002). Some species also prefer certain stages of decay. As decomposition

proceeds, wood undergoes a series of chemical changes and experiences a predictable succession

of cerambycid species over time (Linsley, 1961). Logs infested by members of these families

can be completely degraded, resulting in reduced timber value (Barbosa and Wagner, 1989).

Despite these negative aspects, woodboring beetles may play an important role in nutrient

cycling by breaking down dead wood.

Scolytinae and Platypodinae are often the first to arrive at fresh CWD. Scolytines can be

either true bark beetles, feeding on the inner phloem of wood, or ambrosia beetles (as are all

platypodines), feeding on cultivated “ambrosia” fungi that grows in their tunnels (Solomon,

1995). Most scolytines and platypodines prefer dying or freshly fallen trees, but some also

attack healthy trees, and all require sufficiently moist wood for proper larval development

(Solomon, 1995; Anderson, 2002).

Brentids and bostrichids are few in numbers compared to the other woodborers, but they

have similar lifecycles. Bostrichids typically infest dead trees but some can attack living trees

that have been weakened (Fisher, 1950; Solomon, 1995). Because of their ability to attack dry

wood, such as standing dead trees, bostrichids are able to infest wood that other woodborers find

42 unusable (Hanula, 1996). Adult brentids can be found under dead bark and lay eggs in dying or recently fallen hardwoods after which their larvae burrow deep into the heartwood (Anderson and Kissinger, 2002).

Cleridae is a small family of mostly predaceous beetles (Opitz, 2002). Some species are considered important biocontrol agents against certain scolytines and important for forest health

(Reeve, 2000; Cronin et al., 2000; Erbilgin and Raffa, 2002). Many host plant volatiles, such as

alpha-pinene and ethanol, attract cerambycids, scolytines, and platypodines to stressed or dying trees (Anderson, 2002; Schroeder, 1988; Montgomery and Wargo, 1983). Clerids, in turn, use these chemicals and others, such as aggregation pheromones, in their search for prey (Schroeder,

1988; Montgomery and Wargo, 1983). Many of the clerid genera we collected are known to feed on one or several of the woodborer families considered in this study (Opitz, 2002).

This study was part of a larger study examining the effects of tree harvest gap size and age on arthropod and bird communities in managed southeastern bottomland hardwood forests.

We report here the response of bark and woodborers to the creation of canopy gaps of different size and age. The gaps were created by group selection harvesting, an uneven-aged silvicultural practice that removes patches of desirable trees to create small (< 0.55 ha) openings similar to those created by insect infestations, severe wind damage, and other localized disturbances

(Hunter 1990, Meadows and Stanturf 1997, Guldin 1996). Forest management that fails to mimic natural rates and patterns of disturbance may disrupt the dead-wood dynamics of a forest which can result in extinctions of species adapted to the natural abundance and diversity of CWD

(Grove, 2002). For example, years of intensive forest management has already extirpated or threatened numerous species throughout Europe (Grove, 2002; Twinn and Harding, 1999; Shirt,

1987). Group selection harvesting may be well suited for southeastern bottomland hardwood

43 forest management because these forests are naturally adapted to similar small-scale disturbances.

2. Methods.

2.1. Study Site

The Savannah River Site (SRS) near Aiken, South Carolina, is located in the upper

Atlantic Coastal Plain Province. The SRS is an 80,269-ha nuclear production facility owned and operated by the United States Department of Energy (DOE). The SRS is a DOE National

Environmental Research Park.

Our study was conducted on 120 ha of 75-year-old bottomland hardwoods (Kilgo, 1997).

Common overstory species included bald cypress (Taxodium distichum (L.) Richard), laurel oak

(Quercus laurifolia Michaux), willow oak (Q. phellos L.), overcup oak (Q. lyrata Walter), cherrybark oak (Q. falcata var. pagodaefolia Ell.), swamp chestnut oak (Q. michauxii Nuttall), sweetgum (Liquidambar styraciflua L.), red maple (Acer rubrum L.), and loblolly pine (Pinus

taeda L.). The midstory consisted predominantly of red mulberry (Morus rubra L.), ironwood

(Carpinus caroliniana Walter) and American holly (Ilex opaca Aiton). The understory was dominated by dwarf palmetto (Sabal minor (Jacquin) Persoon) and switchcane (Arundinaria

gigantean (Walter) Muhl.). Pre-harvest basal area of the stands was 33m²/ha (Pauley et al.,

1996). The stands typically experienced seasonal flooding leaving some of the lower-lying areas

under water throughout much of the year. The total rainfall for the 2001 collecting season was

103.99 cm, with the wettest month being June (23.37cm), and the driest being December

(1.17cm). Significant portions of the study area were flooded from late January through early to mid-April.

2.2. Canopy Gaps

Twenty-four gaps were used in this study. Twelve were created in December 1994 (“old gaps”) and twelve in August 2000 (“new gaps”). Both old and new gaps were of three different sizes

(0.13, 0.26, and 0.50 ha), each replicated four times. The gaps were entirely cleared of all trees

(primarily oaks (Quercus spp.) and other standing vegetation, and they encompassed the area

defined by the boles of perimeter trees. Most coarse woody debris was logging “slash” (tops,

malformed boles, dead wood) that was left behind (Figure 3.1).

At the time of sampling, the vegetation in the old gaps ranged from 1-8 m in height and

consisted primarily of sweetgum, sycamore (Platanus occidentalis L.), green ash (Fraxinus

pennsylvanica Marshall), black willow (Salix nigra Marshall), tulip poplar (Liriodendron

tulipifera L.), blackberries (Rubus spp.), oaks, switchcane, and dwarf palmetto. Vegetation in

the new gaps was up to 1 m tall and consisted primarily of, in addition to many listed above, fireweed (Erechtites hieracifolia (L.) Raf.), plumegrass (Erianthus giganteus (Walter) Muhl.),

sedges (Cyperus spp.), and many native grass species.

2.3. Beetle Sampling

A malaise trap was placed in the center of each gap (Figure 3.1), at the gap edge, and 50 m into the forest from each gap. Traps were operated for four one week intervals (17-23 May,

10-16 July, 7-13 September, and 3-9 November) during 2001. The traps (Sante Traps,

Lexington, KY) differed from traditional malaises traps in that beetles could enter them from any direction and they had collecting jars at both the top and bottom so species that respond to barriers by dropping instead of flying upward were also captured. Traps were suspended from 3 m tall poles constructed from EMT electrical conduit. One lower corner of the trap was connected to the conduit pole to hold the trap in place during windy conditions. To further ensure the stability of the trap, the hanger was inserted into a larger pipe which had been driven in the ground. Poles were simply removed from the pipe to lower the trap for sample collecting.

Collecting jars contained a NaCl-2% formaldehyde solution to preserve specimens and a drop of detergent to reduce surface tension (New and Hanula, 1998). Samples were immediately stored in 70 % alcohol until they were identified to morphospecies.

2.4. Coarse Woody Debris Measurements

We estimated the amount of CWD (stumps, logs, branches, and twigs of all sizes) left in the new gaps following logging (Figure 3.1). The ground cover (i.e. plants, CWD, bare soil, etc.) was recorded every 15 cm along 50-m transects through the center of each gap, and a percentage of each was recorded.

We did not quantify the amount of CWD in the old gaps but very little remained from when the gaps were created, and the thicket of young saplings growing there produced very little new woody debris. The surrounding forest had natural amounts of CWD in various stages of decay.

2.5. Statistical Analysis

A 3-way analysis of variance with gap age, trap location, and gap size as the main effects

showed a significant interaction between gap age and location so we analyzed the data for each

gap age separately.

The square-root transformation was used to satisfy assumptions of normality and

homogeneity of variance where appropriate. Data were analyzed using the General Linear

Models procedure of SAS (SAS institute, 1985), and the Ryan-Einot-Gabriel-Welsch Multiple

Range Test was used to determine differences in relative abundances of insects between trap

locations or gap sizes for each gap age (Day and Quinn 1989). Statistical significance was

accepted when α ≤ 0.05. All differences mentioned in the text are significant unless otherwise

stated.

3. Results

3.1. Coarse Woody Debris

We determined that nearly one fourth (23.58 %, n = 12) of the ground in the new gaps was covered by coarse woody debris. The amount of CWD was fairly consistent for all gap sizes

(20.75%, 27.5%, and 22.5% for 0.13 ha, 0.26 ha, and 0.50 ha gaps, respectively).

3.2. Overall Beetle Diversity and Abundance

We collected 3,122 specimens representing 126 species of bark and woodboring beetles

(Table 3.1). Scolytinae was by far the most abundant group and only slightly less diverse than

Cerambycidae (Table 3.1). Clerids were also well represented with 16 species and 376 specimens collected. Only 9 specimens of 3 species of bostrichids were collected, making this the least abundant family in our samples (Table 3.1).

3.3. Influence of Gap Age on Bark Beetles and Woodborers

Regardless of gap size or location, the total abundance and richness of all woodborers

was higher in new gaps than in old gaps (Figures 3.2 and 3.3). Similar trends existed for

individual families or subfamilies. There were higher numbers of all groups in the centers of new

gaps versus the centers of old gaps (Figure 3.4). With the exception of Cerambycidae (Figure

3.4B), we collected more specimens of each group at the edge of the new gaps than at the edge

48 of old gaps (Figure 3.4). Only Scolytinae (Figure 3.4D) were more abundant in the forest surrounding the new gaps than around the old gaps.

We analyzed eight of the most common (≥ 10 individuals collected) cerambycid species individually. Except for Neoclytus sp. 3 (Figure 3.5F), all were more numerous in new gap

centers than old (Figure 3.5). All but two species (Figure 3.5D,G) were more numerous at the

edge of new gaps than the edge of old gaps, and there were no significant differences between

the number of beetles captured in the forest around new and old gaps (Figure 3.5).

Three species of Buprestidae were captured in high enough numbers to allow

comparisons among treatments (Figure 3.6). Of those, two (Figure 3.6B,C) were more abundant

in new gap centers than old gap centers. Agrilus sp. 1 was more abundant in the forest around

new gaps than around old, but the abundance of this species was very low at the edges of new

gaps (Figure 3.6).

The scolytine species Hypothenemus sp. 2 and Xyleborinus sp. 2 were captured in higher

numbers in the centers and edges of new gaps than of old (Figure 3.7). Xyleborinus sp. 2 also was more abundant in the forest surrounding the new gaps than the old (Figure 3.7B).

3.4. Influence of Trap Location on Bark Beetles and Woodborers

3.4.1. New Gaps

In total, woodborers were more abundant at the gap center and edge than in the nearby forest (Figure 3.2A). While species richness shows a similar trend, the differences between the locations is not significant. Buprestids, cerambycids, brentids, and clerids were all captured in

49 higher numbers in gap centers than at the edges or in the nearby forest (Figure 3.4A,B,E,F).

Platypodines and scolytines were both more abundant at the edge than in the forest (Figure

3.4C,D), and traps at the gap centers had fewer platypodines than at the edges (Figure 3.4C). Of the 16 genera we studied individually, only four did not differ in abundance among locations in the new gaps (Figures 3.5F,G,H and Figure 3.7B).

3.4.2. Old Gaps

In the old gaps, we captured more individuals and species of woodborers at the edge and in the forest than at gap center (Figure 3.2). Cerambycids and scolytines were more abundant at the edge and in the forest than in the center of old gaps (Figure 3.4B,D). Clerids were more abundant in the forest than at gap center (Figure 3.4F). Among individual genera, only the cerambycid Neoclytus sp. 3 (Figure 3.5F) differed among locations.

3.5. Influence of Gap Size on Bark Beetles and Woodborers

The different gap sizes had less of an effect on woodborers than did trap location. Except

for lower richness in 0.13 ha old gaps than in 0.26 ha or 0.50 ha old gaps, there was no significant difference in total bark-beetle and woodborer diversity or abundance in gaps of different size (Figure 3.3).

3.6. Response of Cleridae

The trend among clerids closely matched that of their prey. Clerids were more abundant in the center of new gaps than at the edge or in the forest (Figure 3.4F). They also were more abundant at the center and edge of new gaps than in old, and were less abundant in the center of old gaps than in the surrounding forest (Figure 3.4F). Neorthopleura thoracica and Chariessa

pilos were the most numerous species and they closely matched the trend for the entire family

(Figure 3.8). Cregya oculata was more abundant in traps located in new gap centers and edges

than in those located at similar positions in old gaps (Figure 3.8). Neorthopleura thoracica

(Figure 3.8C) and Chariessa pilos (Figure 3.8A) differed among locations in old gaps.

3.7. Specialist Species

Of all families considered, 14 species were collected only in the center of new gaps. Of these, only 2 (Chalcophora (Buprestidae) and Ecyrus (Cerambycidae)) were abundant enough

(27 and 15 individuals, respectively) to consider as possible gap specialists. While other species were only collected in certain locations, they were not collected in high enough numbers to determine that they may specialize in such areas.

4. Discussion

The importance of preserving CWD for wood-dwelling beetle communities is well supported by past research. For example, Warriner et al. (2002) saw increased cerambycid diversity and abundance in recently thinned forest stands (with large amounts of woody debris) compared to unthinned stands. Barbalat (1996) also found cerambycids to favor clearings in a managed forest in Switzerland. Another study found that buprestids, cerambycids, and scolytines increased the year following disturbance, but soon declined in number to below undisturbed forest levels (Werner, 2002). Our study shows a similar trend for cerambycids, buprestids, brentids, clerids, scolytines, and platypodines. We found their numbers to be much higher in new canopy gaps with abundant CWD than in the surrounding forest, while the trend for old canopy gaps with very little CWD was the opposite.

Because the creation of new gaps influenced the diversity of woodboring beetles at least

50 m into the surrounding forest, the traps located in the forest surrounding old gaps may provide a better baseline of bark-beetle and woodborer diversity in undisturbed forest. From this standpoint, recently created gaps and their associated woody debris had, on average, roughly twice the species richness and about 6 times the abundance of beetles than did the forest.

With the exception of there being fewer species in 0.13 ha gaps than in the larger gap sizes, our results show little effect of gap size on the abundance or species richness of wood- dwelling beetles. The larger gaps had proportionally more CWD than did the small gaps, and we expected there to be similar differences in the numbers of wood-dwelling beetles present.

Our sampling method (i.e. a malaise trap in the center of each gap) may have missed such differences. The malaise traps may have estimated the beetle density in the immediate vicinity of the traps (which may have been the same regardless of gap size since CWD was evenly

52 distributed across gaps) without accurately portraying differences in the total number of wood-

dwelling beetles in gaps of each size. Because the sampling radius of our malaise traps is not

known we cannot use our data to accurately estimate differences in the abundance of woodborers

among different sized gaps. Not all studies have found an increase in woodborer abundance

following disturbance. In a five-year study, Gutowski (1986) found lower cerambycid diversity

and abundance in forests where shelterwood cutting was applied compared to unmanaged forests in Poland. Gutowski (1986) observed that in many timber harvesting operations, the slash is burned, but he did not mention the presence or absence of CWD in his study site. If it had been removed, as the author indirectly suggests, then this resource would not have been made available to woodborers.

While preserving some CWD in managed forests is preferable to its removal, we do not suggest that disturbing natural forests would improve wood-dwelling insect habitat. While wood-dwelling insects appear to initially benefit from disturbance, disturbed areas become less favorable with time. As the amount of available CWD decreases, so should the abundance and possibly the diversity of wood-dwelling insects. In our study, 6-year-old gaps had lower abundance and richness of wood boring beetles than did young gaps or the surrounding forest.

In disturbed areas, the amount of CWD present, and the associated beetle fauna, depends on how recently the area was disturbed (Grove, 2002; Siitonen, 2001; Speis et al., 1988). Mature, unmanaged forests, in contrast, have a relatively predictable distribution and stable amount of

available CWD (Grove, 2002).

Forests that have been mostly undisturbed by humans maintain the most abundant and

diverse beetle populations. In Japan, both the diversity and abundance of cerambycids were higher in old-growth forests than in 30-70 year old second-growth forests (Maeto et al., 2002).

Furthermore, some species are found only in old forests, underscoring the importance of preserving stands of mature trees (Barbalat, 1996; Gutowski, 1986; Grove, 2002). While we report here several species found only in the clearings, they are surely adapted to areas of natural disturbances

Although unmanaged forests may be preferable from an ecological perspective, the need for timber necessitates its removal. It is important to identify management practices that minimize impact on the local ecosystem. More information is needed to understand the relative value of group selection harvesting compared to other forest management techniques, and how the size of harvest unit effects arthropods. Our results suggest that the smallest gaps (0.13 ha) had no fewer wood-dwelling beetles than did the largest gaps (0.50 ha), and that with but one exception, diversity was not different among gap sizes. We suggest that future studies incorporate a broader range of gap sizes and consider the movement of insects into and out of the gaps. How insects recolonize mature forest after emerging from areas of disturbance remains largely unknown. In this study we found several species to be more common in the forests surrounding new gaps than old gaps, possibly indicating movement from the gap back into the forest.

While further research is needed to understand the long-term implications of forest management, this study clearly demonstrates the importance of preserving CWD created during timber removal. Had the CWD been removed from the new gaps after logging, the diversity and abundance of the woodborers likely would have been well below that of the surrounding forest.

Given the evidence from this and past studies of CWD, its preservation should be of central concern to forest management. The degree to which logging disrupts forest ecosystems may be lessened by recognizing the value of the woody debris to many wood-inhabiting organisms.

Acknowledgments

We thank C. Smith for assisting with the identification of specimens, D. Dyer and J. Campbell for helping with field work, and C. Asaro, J. McHugh, and W. Berisford for editing early drafts of the manuscript. We also thank the U.S. Department of Energy-Savannah River and the US

Forest Service-Savannah River for access and logistical support. This research was funded by the National Research Initiative Competitive Grants Program of the USDA Cooperative State

Research Education and Extension Service (CSREES Grant No. 00-35101-9307).

References Cited

Anderson, R.S., 2002. Curculionidae. In: Arnett, R. H. and M. C. Thomas, eds. American

beetles volume 2, 722-815.

Anderson, R.S., Kissinger, D.G., 2002. Brentidae. In: Arnett, R. H. and M.

C. Thomas, eds. American beetles volume 2, 711-719.

Barbalat, S., 1996. Influence of forest management on three wood-eating families in the

Areuse Gorges (Canton of Neuchatel Switzerland). Rev. Suisse. Zool. 103, 553-564.

Barbosa, P., Wagner, M.R.. 1989. Introduction to forest and shade tree insects.

Academic Press, Inc.

Bellamy, C.L., Nelson, G.H., 2002. Buprestidae. In: Arnett, R. H. and M. C.

Thomas, eds. American beetles volume 2. 98-112.

Cronin, J.T., Reeve, J.D., Wilkens, R., Turchin, P., 2000. The pattern and range of

movement of a checkered beetle predator relative to its bark beetle prey. Oikos 1,

127-138.

Day, R.W., Quinn, G.P., 1989. Comparisons of treatments after an analysis of variance

in ecology. Ecol. Mongr. 59, 433-463.

Elton, C.S., 1966. Dying and dead wood. In: “The Pattern of Animal Communities.”

279-305. Wiley, New York.

Erbilgin, N., Raffa, K.F., 2002. Association of declining red pine stands with reduced

populations of bark beetle predators, seasonal increases in root colonizing insects, and

incidence of root pathogens. Forest Ecology and Management 164, 221-236.

Fellin, D.G., 1980. A review of some interactions between harvesting, residue

management, fire, and forest insects and diseases. In: Environmental consequences of

timber harvesting in Rocky Mountain coniferous forests, symposium proceedings.

General Technical Report INT-90: U.S. Department of Agriculture, Forest Service.

Fisher, W.S., 1950. A revision of the North American species of beetles belonging to the

family Bostrichidae. Misc. Publ. 698. Washington, DC: U.S. Department of Agriculture,

157p.

Grove, S.J., 2002. Saproxylic insect ecology and the sustainable management of forests.

Annu. Rev. Ecol. Syst. 33, 1-23.

Guldin, J.M., 1996. The role of uneven-aged silviculture in the context of ecosystem

management. WJAF 11: 4-13

Gutowski, J.M., 1986. Species composition and structure of the communities of

longhorn beetles (Coleoptera: Cerambycidae) in virgin and managed stands of Tilio-

Carpinetum stachyetosum association in the Bialowieza Forest (NE Poland). J. Appl. Ent.

102, 380-391.

Hagan, J.M., Grove, S.L., 1999. Coarse woody debris. Journal of Forestry.

January: 6-11.

Hanula, J.L., 1996. Relationship of wood-feeding insects and coarse woody debris. In:

McMinn, J. W. and D. A. Crossley, Jr. eds. Biodiversity and coarse woody debris in

southern forests, proceedings of the workshop on coarse woody debris in southern

forests: effects on biodiversity; 1993 October 18-20; Athens, GA. Gen. Tech. Rep. SE-94

Asheville, NC: U.S. Department of Agriculture, Forest Service, Southern Research

Station: 55-81.

Harmon, M.E., Franklin, J.F., Swanson, F.J., Sollins, P., Gregory, S.V., Lattin, J.D.,

Anderson, N.H., Cline, S.P., Aumen, N.G., Sedell, J.R., Lienkaemper, G.W., Cromack,

J.R., Cummins, K.W., 1986. Ecology of coarse woody debris in temperate ecosystems.

Advancess in Ecological Research. 15, 133-302.

Hunter, M.L., 1990. Wildlife, forests, and forestry. Prentice-Hall, Englewood Cliffs,

New Jersey, USA.

Huston, M.A., 1996. Modeling and management implications of coarse woody debris

impacts on biodiversity. In: McMinn, J., Crossley, D.A., eds. 1996. Biodiversity and

coarse woody debris in southern forests. Athens Ga: USDA For. Ser. Gen. Tech. Rep.

SE-94.

Kilgo, J.C., 1997. Establishment report: the gap project. U. S. Forest Service Report,

Savannah River Institutue. 42 pp.

Linsley, E.G., 1961. The Cerambycidae of North America. Part I: Introduction. Berkely,

CA: University of California Press. Vol. 18.

Maeto, K., Sato, S., Miyata, H., 2002. Species diversity of longhorn beetles in

humid warm-temperate forests: the impact of forest management practices on old-growth

forest species in southwestern Japan. Biodiversity and Conservation. 11, 1919-1937.

Meadows, J.S., Stanturf, J.A., 1997. Silvicultural systems for southern bottomland

hardwood forests. Forest Ecology and Management 90, 127-140.

Montgomery, M.E., Wargo, P.M., 1983. Ethanol and other host-derived volatiles as

attractants to beetles that bore into hardwoods. Journal of Chemical Ecology 9, 181-190.

New, K.C., Hanula, J.L., 1998. Effect of time elapsed after prescribed burning in

longleaf pine stands on potential prey of the Red-Cockaded Woodpecker. SJAF 22, 175-

183.

Opitz, W., 2002. Cleridae. In: Arnett, R.H., Thomas, M.C., eds. American beetles volume

2, 267-280.

Pauley, E.F., Collins, B.S., Smith, W.P., 1996. Pp. 132-136 in Flynn, K.M. (Ed.). Early

establishment of cherrybark oak in bottomland hardwood gaps: effects of seed predation,

gap size, herbivory, and competition. Proceedings of the southern forested wetlands

ecology and management conference. Clemson, S.C.

Reeve, J.D., 2000. Complex emergence patterns in a bark beetle predator. Agricultural

and Forest Entomology 2, 233-240.

SAS Institute, 1985. SAS guide for personal computers. Version 6 ed., SAS

Institute, Cary, NC. 378 pp.

Schroeder, L.M., 1988. Attraction of the bark beetle Tomicus piniperda and some other

bark and wood-living beetles to the host volatiles alpha-pinene and ethanol.

Entomologia Experimentalis et Applicata 46, 203-210.

Shirt, D.B., 1987. British red data books: 2. Insects. Peterborough, UK: Nat. Conserv.

Counc.

Siitonen, J., 2001. Forest management, coarse woody debris and saproxylic organisms:

Fennoscandian boreal forests as an example. Ecol. Bull. 49, 11-42.

Solomon, J.D., 1995. Guide to Insect Borers in North American Broadleaf Trees and

Shrubs. Agricultural Handbook 706. Washington, D.C.: U.S. Department of Agriculture,

Forest Service. 735 p.

Spies, T.A., Franklin, J.F., Thomas, T.B., 1988. Coarse woody debris in Douglas-fir

forests of western Oregon and Washington. Ecology. 69, 1689-1702.

Twinn, P.F.G, Harding, P.T., 1999. Provisional atlas of the longhorn beetles

(Coleooptera, Cerambycidae) of Britain. Abbots Ripton, UK: Nat. Environ. Res. Counc.

96 pp.

Wallace, A.R. 1869. The Malay Archipelago: the land of the orang-utan, and the bird of

paradise. 515 pp.

Warriner, M.D., Nebeker T.E., Leininger T.D., Meadows J.S., 2002. The effects

of thinning on beetles (Coleoptera: Carabidae, Cerambycidae) in bottomland hardwood

forests. In: Outcalt, KW, ed. Proceedings of the eleventh biennial southern silvicultural

research conference. Gen. Tech. Rep. SRS-48. Asheville, NC: U.S. Department of

Agriculture, Forest Service, Southern Research Station: 569-573.

Werner, R.A., 2002. Effect of ecosystem disturbance on diversity of bark and wood-

boring beetles (Coleoptera: Scolytidae, Buprestidae, Cerambycidae) in White Spruce

(Picea glauca (Moench) Voss) ecosystems of Alaska. U.S. Department of Agriculture,

Forest Service, Pacific Northwest Research Station, Research Paper PNW-RP-546, 1-15.

Yanega, D., 1996. Field guide to northeastern longhorned beetles (Coleoptera:

Cerambycidae). Natural History Survey Manual. 6.

Table 3.1

Total number/richness of six beetle families collected in malaise traps set at gap center, gap edge, and nearby forest of canopy gaps created in 1994 and 2000 in a hardwood bottomland forest at the Savannah River Site, South

Carolina. Samples were collected from May-November, 2001.

2000 Treatment 1994 Treatment Total

Center Edge Forest Center Edge Forest

Buprestidae 421/12 70/12 45/9 25/5 17/6 23/6 601/17

Cerambycidae 546/34 272/29 187/26 19/13 57/19 55/22 1136/51

Bostrichidae 4/3 3/1 1/1 0 1/1 0 9/3

Brentidae (Arrhenodes) 42/1 11/1 2/1 0 1/1 2/1 58/1

Curculionidae (Platypodinae) 11/2 26/3 6/1 2/1 5/2 6/3 56/5

Curculionidae (Scolytinae) 451/18 839/21 227/30 35/12 131/24 120/24 1803/49

Total Woodborer 1475/70 976/67 468/68 81/31 212/53 206/56 3122/126

Cleridae 205/12 61/12 44/11 7/3 21/7 38/10 376/16

Figure 3.1. The malaise trap (Sante Traps, Lexington, KY) used to sample bark and woodboring beetles in a 0.50 ha canopy gap in a South Carolina bottomland hardwood forest. Note the large amounts of coarse woody created during timber removal.

Figure 3.2. Mean (±SE) abundance and richness of woodboring beetles (Cerambycidae,

Buprestidae, Bostrichidae, Brentidae (Arrhenodes), Curculionidae (Platypodinae and

Scolytinae)) captured in malaise traps at different locations in bottomland hardwood forest gaps

created in 1994 and 2000. Within graphs, bars with the same letter above them are not

significantly different (Ryan-Einot-Gabriel-Welsch Multiple Range Test, p < 0.05). Asterisks

denote significant differences (p < 0.05) between the same trap locations (e.g., center vs. center)

in old (1994) and new (2000) gaps.

160

A A* woodborers 140 A* 120

100

60 B* 40

mean no. of individuals (n=12) no. of mean b* b* 20 a*

B woodborers 20 A*

A* 15

10 b* b* mean no. of species (n=12)

5 a*

0 Center Edge Forest Center Edge Forest 1994 Treatment 2000 Treatment

Figure 3.3. Mean (±SE) abundance and richness of woodboring beetles (Cerambycidae,

Buprestidae, Bostrichidae, Brentidae (Arrhenodes), Curculionidae (Platypodinae and

Scolytinae)) captured in malaise traps in bottomland hardwood forest gaps of different sizes

created in 1994 and 2000. The three malaise samples (a trap was placed at the center, edge, and

in nearby forest for each gap) were combined for these results. Within graphs, bars with the same

letter above them are not significantly different (Ryan-Einot-Gabriel-Welsch Multiple Range

Test, p < 0.05). Asterisks denote significant differences (p < 0.05) between the same trap

locations (e.g., center vs. center) in old (1994) and new (2000) gaps.

500

woodborers A* 400

300 A* A*

200

100 mean no. of individuals (n=4) a* a* a*

0 20 A* A* woodborers

A* 15

species (n=4) 10 b* b*

5 a* mean no. of

0 0.13ha 0.26ha 0.50ha 0.13 ha 0.26 ha 0.50 ha 2000 Treatment 1994 Treatment

Figure 3.4. Mean (±SE) abundance of woodboring beetles (at family or genus level) captured in malaise traps at different locations in bottomland hardwood forest gaps created in 1994 and

2000. Within graphs, bars with the same letter above them are not significantly different (Ryan-

Einot-Gabriel-Welsch Multiple Range Test, p > 0.05). Asterisks denote significant differences

(p < 0.05) between the same trap locations (e.g., center vs. center) in old (1994) and new (2000) gaps. The prime symbol indicates the data were transformed using the square root transformation prior to analysis.

50 A* 40 A Buprestidae

10 B*' B a* a*' a 0 50 B A* Cerambycidae 40

30 B B 20

10 b' b' a*' 0 C A* Platypodinae (Platypus sp.) 2

1 B a a* a* ected at each location (n=12) 0 A* 80 D Scolytinae

60 AB* 40 B* 20 b* b* a* 0 A* 4 E Brentidae (Arrhenodes sp.) 3 mean number of individuals coll 2 B* 1 B a a* a* 0 A* 20 F Cleridae

10 B* 5 B b ab* a* 0 Center Edge Forest Center Edge Forest

2000 Treatment 1994 Treatment

Figure 3.5. Mean (±SE) abundance of notable cerambycid species captured in malaise traps at different locations in bottomland hardwood forest gaps created in 1994 and 2000. Within graphs, bars with the same letter above them are not significantly different (Ryan-Einot-Gabriel-

Welsch Multiple Range Test, p > 0.05). Asterisks denote significant differences (p < 0.05)

between the same trap locations (e.g., center vs. center) in old (1994) and new (2000) gaps. The

prime symbol indicates the data were transformed using the square root transformation prior to

analysis.

8 A A* Acanthocinus 6 AB*' 4

2 B C C* C*' 0 20 B A* Elaphidion sp. 1 15 10 B* 5 B C* C* C 0 AB* 4 C A* Elaphidion sp. 2 3 2 B 1 C* C* C 0 2 D A* Leptostylus sp. 1 1 B C B C* C

cted at each location (n=12) 0 6 A* E Neoclytus sp. 2 4

2 B* B C* C* C 0 A* 0.6 F A Neoclytus sp. 3

0.4 A C 0.2

C C* 0.0 mean number individualsmean of colle A*' 3 G Saperda tridentata

2 A 1 A C*' C C 0 A* H A* A Xylotrechus sp. 2 1 C C* C* 0 Center Edge Forest Center Edge Forest 2000 Treatment 1994 Treatment

Figure 3.6. Mean (±SE) abundance of notable buprestid species captured in malaise traps at different locations in bottomland hardwood forest gaps created in 1994 and 2000. Within graphs, bars with the same letter above them are not significantly different (Ryan-Einot-Gabriel-

Welsch Multiple Range Test, p > 0.05). Asterisks denote significant differences (p < 0.05)

between the same trap locations (e.g., center vs. center) in old (1994) and new (2000) gaps. The

prime symbol indicates the data were transformed using the square root transformation prior to

analysis.

A A*' Agrilus sp. 1

2 A'

1 C* C C B' 0

Agrilus sp. 2 4 B A*

2 C C* 1 B B C 0 A* 30 C Chrysobothris

mean number of individualsmean number of collected each(n=12) location at B C* B C C 0 Center Edge Forest Center Edge Forest 2000 Treatment 1994 Treatment

Figure 3.7. Mean (±SE) abundance of notable scolytine species captured in malaise traps at different locations in bottomland hardwood forest gaps created in 1994 and 2000. Within graphs, bars with the same letter above them are not significantly different (Ryan-Einot-Gabriel-

Welsch Multiple Range Test, p > 0.05). Asterisks denote significant differences (p < 0.05)

between the same trap locations (e.g., center vs. center) in old (1994) and new (2000) gaps.

40 A Hypothenemus sp. 2 A* 30

AB* 20

10 B a* a* a 0

B A* Xyleborinus sp. 2 15 A*

10 A*

5 a* a* a* mean no. individuals collected per location (n=12) location per collected individuals no. mean 0 Center Edge Forest Center Edge Forest 2000 Treatment 1994 Treatment

Figure 3.8. Mean (±SE) abundance of notable clerid species captured in malaise traps at different locations in bottomland hardwood forest gaps created in 1994 and 2000. Within graphs, bars with the same letter above them are not significantly different (Ryan-Einot-Gabriel-Welsch

Multiple Range Test, p > 0.05). Asterisks denote significant differences (p < 0.05) between the same trap locations (e.g., center vs. center) in old (1994) and new (2000) gaps.

18 16 A A* Chariessa pilos 14 12 10 8 6 4 B* 2 B a* b* ab 0 Cregya oculata B A* A*

a a* B a* 0 C Neorthopleura thoracica A* mean no. individuals collected per location (n=12) location per collected individuals no. mean 1 b

B B ab a* 0 Center Edge Forest Center Edge Forest 2000 Treatment 1994 Treatment

CHAPTER 4

HERBIVOROUS INSECT RESPONSE TO GROUP SELECTION CUTTING IN A

SOUTHEASTERN BOTTOMLAND HARDWOOD FOREST

¹Ulyshen, M. D., Hanula, J. L., Horn, S., Kilgo, J. C., Moorman, C. E., 2004. Accepted by

Environmental Entomology.

Reprinted here with permission of publisher

Abstract

Malaise and pitfall traps were used to sample herbivorous insects in canopy gaps created by group-selection cutting in a bottomland hardwood forest, South Carolina, USA. The traps were placed at the centers, edges, and in the forest adjacent to gaps of different sizes (0.13, 0.26, 0.50 ha) and ages (one and seven years old) during four sampling periods in 2001. Overall, the abundance and species richness of insect herbivores were greater at the centers of young gaps than at the edge of young gaps or in the forest surrounding young gaps. There were no differences in abundance or species richness among old gap locations (i.e. centers, edges, and forest) and we collected significantly more insects in young gaps than old gaps. The insect communities in old gaps were more similar to the forests surrounding them than young gap communities were to their respective forest locations, but the insect communities in the two forests locations (surrounding young and old gaps) had the highest percent similarity of all.

Although both abundance and richness increased in the centers of young gaps with increasing gap size, these differences were not significant. We attribute the increased numbers of herbivorous insects to the greater abundance of herbaceous plants available in young gaps.

Keywords selection cutting, uneven-aged silviculture, forest openings, forest management

Introduction

While the effects of insect herbivory on plant communities and rates of succession have been well studied (Breedlove and Ehrlich 1968, Brown 1984, Brown 1985, Hendrix et al. 1988,

Brown and Gange 1992, McBrien et al. 1983), relatively little is known about how plant succession affects herbivorous insects (Bach 1990). Given the relative abundance of young herbaceous growth in early stages of succession, one might expect to find increased numbers of herbivorous insects there compared to more mature habitats. Indeed, past work recognizes the

importance of increasing taxonomic and structural diversity of plants to the creation and

maintenance of a diverse insect community during succession (Murdoch et al. 1972, Lawton

1978, Southwood et al. 1979), and several features of the plants themselves may encourage herbivory in recently created habitats. These include increased nutrient levels (i.e. soluble

nitrogen) in plant tissues (Boardman 1977, McNeill and Southwood 1978, Mattson 1980),

reduced plant defenses in many pioneer species (Coley 1983, Lawton and McNeill 1979), and

increased consumption and growth rates of herbivorous insects that feed on plants receiving

direct sunlight (White 1978, Scriber and Slansky 1981). The situation is complicated by a

number of factors that appear to discourage herbivory, however. For example, increased light

levels may be beneficial in terms of insect growth rates, but they have also been shown to

increase the toughness of leaves and in some cases, the concentration of defensive compounds

(Shure and Wilson 1993).

In many forests, canopy gaps created by treefalls, wind damage, and other minor events

serve as important centers of plant growth and succession (Runkle 1981, 1982, White et al. 1985,

Phillips and Shure 1990, Clinton et al. 1993). The increased availability of light, water, and

80 nutrients in gaps increases plant diversity and net primary productivity, and encourages the

growth and regeneration of less shade tolerant species (Bormann and Likens 1979, Boring et al.

1981, Brokaw 1982, Phillips and Shure 1990, Wilder et al. 1999). Predicting the response of

herbivorous insects to such complicated and dynamic environments is difficult and quickly

confounded by factors such as gap size and age. Large gaps, for example, receive more sunlight

than small gaps (Shure and Wilson 1993) creating differences in soil moisture and plant growth

(Shure and Phillips 1991). The plant communities present in gaps of differing age and stage of

succession should be quite different, with unknown implications to the insect community. In this

paper we compare the species richness, abundance, and composition of herbivorous insects in

artificial canopy gaps of different size (0.13 ha, 0.26 ha, and 0.50 ha) and age (one and seven

years old) in a bottomland hardwood forest in the southeastern United States. The gaps were created by group-selection cutting, an uneven-aged forest management practice that removes patches of merchantable trees leaving small (< 0.55 ha) openings similar to those created by insect infestations, severe wind damage or other localized disturbances (Hunter 1990, Meadows and Stanturf 1997, Guldin 1996).

Materials and Methods

This study was conducted on the Savannah River Site, an 80,269-ha nuclear production facility near Aiken, South Carolina. The SRS is owned and operated by the United States

Department of Energy, and is managed as a National Environmental Research Park. The stand

used was a 75-100-year-old bottomland hardwood forest approximately 120 ha in size. The

81 forest canopy consisted of bald cypress (Taxodium distichum (L.)), laurel oak (Quercus laurifolia

Michaux), willow oak (Q. phellos L.), overcup oak (Q. lyrata Walter), cherrybark oak (Q.

falcata var. pagodaefolia Elliott), swamp chestnut oak (Q. michauxii Nuttall), sweetgum

(Liquidambar styraciflua L.), red maple (Acer rubrum L.), and loblolly pine (Pinus taeda L.).

The midstory consisted predominantly of red mulberry (Morus rubra L.), ironwood (Carpinus

caroliniana Walter) and American holly (Ilex opaca Aiton). The understory was dominated by dwarf palmetto (Sabal minor (Jacquin) Persoon) and switchcane (Arundinaria gigantean

(Walter) Muhlenberg).

Of the 24 gaps used in this study, 12 were created in December of 1994 (“old gaps”) and

12 in August of 2000 (“young gaps”). For both young and old gaps there were three different

sizes (0.13 ha, 0.26 ha, and 0.50 ha), each replicated four times. At the time of sampling, the

vegetation in the old gaps ranged from 1-8 m in height and consisted of pioneer species such as sweetgum, sycamore (Platanus occidentalis L.), green ash (Fraxinus pennsylvanica Marshall),

black willow (Salix nigra Marshall), tulip poplar (Liriodendron tulipifera L.), oaks, switchcane,

and dwarf palmetto. Young gaps also contained seedlings and stump sprouts of the tree species

above plus an abundance of fireweed (Erechtites hieracifolia (L.) Rafinesque), blackberries

(Rubus spp.), plumegrass (Erianthus giganteus (Walter) Muhlenberg), and various grass and sedge (Cyperus spp.) species. Young gaps generally had a much more diverse herbaceous layer as well as numerous sprouts arising from tree stumps and roots, while older gaps and the surrounding forest had comparatively little herbaceous growth. Competition with young trees limited the amount of herbaceous vegetation in old gaps, but they still contained more herbaceous growth per unit area than the nearby forest (L. Bowen, personal communication).

Insects were sampled four times in 2001 (17-23 May, 10-16 July, 7-13 September, and 3-

9 November) at three locations (gap center, gap edge, and in the forest 50 m from gap edge) in and around each gap. Overall, we sampled for four weeks in each of the 72 locations. Each sample location had a malaise trap (Canopy Trap, Sante Traps, Lexington, KY), suspended from a three meter-long pole to capture flying insects, and two pitfall traps to capture ground-dwelling insects. The pitfall traps were placed five meters apart at each sample location, and consisted of a 480-ml plastic cup buried to ground level. A small funnel (8.4 cm diameter) was inserted into the mouth of the cup to direct captured insects into a smaller 120-ml specimen cup below. The cup was positioned at the intersection of four 0.5-m long drift fences. The malaise and pitfall trap samples at each location were combined prior to analysis. The collecting jars for both pitfall and malaise traps were filled with a NaCl-2% formaldehyde solution to preserve specimens and a drop of detergent was added to reduce surface tension (New and Hanula 1998). Once collected, the insects were brought back to the lab and immediately stored in 70 % alcohol. Specimens were sorted and later identified to morphospecies. We included the following herbivores (by order and family) in our analyses: Coleoptera: Chrysomelidae; Lepidoptera: larvae of all families; Thysanoptera: all families; Orthoptera: Acrididae and Tettigoniidae; Homoptera:

Achilidae, Aphididae, Cercopidae, Cicadellidae, Cixiidae, Delphacidae, Derbidae, Flatidae,

Issidae, Membracidae, and Psyllidae; Hemiptera: Lygaeidae, Miridae, and Pentatomidae. We examined adult Lepidoptera in separate analyses because they may indicate the presence of larvae but are not themselves actively herbivorous.

A three-way analysis of variance with abundance and richness as response variables and gap age, trap location, and gap size as the main effects showed a significant interaction between gap age and location so we analyzed the data for each age separately. Data were analyzed using

83 the General Linear Model procedure of SAS (SAS institute, 1985), and the Ryan-Einot-Gabriel-

Welsch Multiple Range Test (α<0.05) was used to determine differences in relative abundances of insects between trap locations or gap sizes for each gap age. We also calculated the percent similarity (Southwood 1966) of herbivore communities among trap locations and gap ages.

Results

We collected a total of 18,583 herbivores representing 429 species, excluding adult

Lepidoptera. Cicadellidae (Homoptera) was by far the most abundant family with over 13,000

specimens collected. Over 1,000 specimens were collected from each of the next two most

abundant families, Chrysomelidae (Coleoptera) and Aphididae (Homoptera) (Table 1). In

contrast, several families of Homoptera (Achilidae, Membracidae, Issidae), Hemiptera

(Pentatomidae), and Orthoptera (Tettigoniidae) were infrequently collected and were represented

by fewer than 50 specimens each (Table 1). Cicadellidae had the greatest species richness with

94 morphospecies followed by Chrysomelidae (71 species) and larval Lepidoptera (56 species),

while Tettigoniidae, Achilidae, Psyllidae, Cercopidae, Issidae, and Flatidae had fewer than ten

morphospecies each (Table 1). Adult Lepidoptera were very well represented in the samples

(Table 1), but were not included with herbivore totals because they are not active plant feeders.

The forests surrounding young and old gaps had the most similar herbivorous insect assemblages

while those in the centers of young gaps and the forests surrounding them were the least similar

(Table 2).

Overall abundance and species richness of herbivorous insects was greater at the center

of young gaps than at other young or old gap locations (Fig. 1A,B). No differences in abundance

84 or richness were observed among the three trapping locations for old gaps. Although insect richness and abundance was higher at the center of young gaps than at the center of old gaps, there were no such differences between gap ages for the edge or forest locations (Fig. 1A,B).

Both herbivore richness and abundance increased with increasing gap size for the centers of young gaps, but these differences were not significant (Fig. 2A,B). Herbivore richness was higher in young gaps than in old gaps regardless of gap size (Fig. 2A) and abundance was higher in the centers of the two smaller gap sizes but not in the 0.50 ha gaps (Fig. 2B).

Most herbivore families or orders (Table 3) followed the overall trend (Fig. 1) for all

herbivores combined, i.e. greater abundance and richness in young gap centers than at the edges

or in the forest interior. Likewise, the abundance and species richness of many groups were

higher in young gap centers than old gap centers. Cixiidae and Derbidae were the only groups

captured in higher numbers in the forests near old gaps than in the gap centers while no group

was captured in higher numbers in forests near young gaps than in the centers of young gaps.

Discussion

Relatively little research has dealt with group selection cutting and its effects on forest

ecosystems, but several studies have found insect abundance to be greater in canopy gaps than in

closed forest (Koivula and Niemelae 2003, Gorham et al. 2002, Hill et al. 2001). Some studies

suggest otherwise (Shelly 1988), but others involving insect predators (bats, Menzel et al. 2002;

treefrogs, Cromer et al. 2002, Horn et al. 2005; and birds, Kilgo et al. 1999) lend support to the conclusion that gaps generally do contain greater abundances of insects. In this and related

85 papers (Ulyshen et al. 2004, 2006) we report increases in the abundance and species richness of insects in young gaps, but older gaps and the forests surrounding them contained comparable numbers of insects. Only four families of herbivores showed any differences in morphospecies richness or abundance among old gap locations. The abundance and species richness of Miridae and the abundance of Flatidae were higher at the centers of old gaps than at the edges or in the

forests surrounding old gaps, while Cixiidae and Derbidae exhibited the opposite response. In

contrast, the abundance and species richness of all herbivore groups were either greater in young

gaps compared to the surrounding forest or were approximately equal. In no case did we catch

more in the forest.

Because they differed so greatly in the structure and composition of their respective plant

communities, it is not surprising that young and old gaps contained quite different insect

communities (48.91%). Community similarity between old gaps and the forest surrounding old

gaps was considerably greater (63.25%) than the similarity between young gaps and the forest

surrounding them (47.25%), but both old and young gaps were considerably less similar to their

respective forest locations than the two forest locations were to one other (72.85%). These data

suggest that seven years is insufficient time for herbivorous insect communities to reach pre-

disturbance conditions following canopy gap creation but that they are gradually becoming more like the surrounding forest herbivore community.

While the effects of gap age on insect communities were considerable, gap size had surprisingly little effect on their abundance or species richness. The percent area covered by vegetation in young gaps was about the same regardless of gap size (T. Champlin, personal communication), so the amount of vegetation, at least in young gaps, increased with increasing

gap size. From this we would have expected a similar increase in the abundance and possibly

86 species richness of herbivorous insects. Although the trend was there, no significant differences in herbivore species richness or abundance among the gap sizes were detected. Because all traps have a limited sampling radius, malaise and pitfall traps may have somewhat hindered our ability to detect differences in insect abundance among different sized gaps. Also malaise and pitfall traps are somewhat biased toward the most active species, and generally overlook many species that are confined to their host plant. Systematically collecting insects throughout the gaps with vacuums or nets may have eliminated some limitations of passive trapping, but this may or may not have affected our results. Shure and Phillips (1991) used vacuum sampling to effectively measure differences in insect abundance between gaps of different size (0.016 to 10 ha), but they also found relatively little difference in herbivore (Homoptera and Hemiptera) abundance within the range of gap sizes considered in this study.

Our trapping procedure may also explain the difference between adult and larval

Lepidoptera response to gap creation. The abundance and richness of caterpillars between old and young gaps were similar while adults occurred in higher numbers in both old and young gap centers compared to the surrounding forest. Caterpillars were only collected in pitfall traps, so only larvae crawling across the ground were sampled. Because they are generally confined to their food plants and spend little time on the ground, caterpillars were probably underrepresented

in our samples. In contrast, our malaise traps were well suited to sample adult Lepidoptera since

they move readily in search of nectar, mates, and oviposition sites, and the trend for these (Table

3) more closely matched that for herbivores in general (Fig. 1).

The comparatively high abundance and richness of herbivores observed in young gaps

may be attributed in part to the abundance and palatability of the young plants growing there.

Lawton and McNeill (1979) predicted that young foliage would support a higher abundance of

87 herbivores than older foliage of the same species, and that higher abundances would also be supported at earlier rather than later stages of succession. These predictions were supported in a study by Godfray (1985) in which the abundances of leaf-miners were compared among different stages of succession. The apparent preference for plants in early rather than later stages of succession may be explained by increases in plant defenses with time (Lawton and McNeill

1979), softer leaves (Shure and Wilson 1993), fewer secondary plant compounds in the faster- growing pioneer species (Coley 1983, Denslow et al. 1990), and a greater availability of water

(Scriber and Slansky 1981) and nutrients (McNeill and Southwood 1978, Mattson 1980) in young tissues.

Although disparities in resource availability are likely to have affected herbivore numbers, other factors may have been important as well. For instance, some insects may have been drawn to gaps due to the relatively high light levels and temperatures there, as compared to the forest understory. Many may use the increased temperatures in gaps and other open areas to warm up on cold days. A number of herbivores were surely in the gaps for reasons other than herbivory, but their contribution to our results cannot be determined.

Despite these complications, the conditions in young gaps appear to encourage the colonization of a more abundant and species rich assemblage of insect herbivores. Of the nineteen herbivorous insect groups considered in this study, six had significantly greater morphospecies richness in young gaps compared to the forest and 11 were captured in higher numbers in young gaps. Six taxonomic groups had greater species richness in young gaps compared to old gaps and five were captured in higher numbers in young gaps. In most cases, the other groups exhibited similar trends. After seven years of succession, the abundance and species richness of insect herbivores was comparable to that of the surrounding forest, but the

88 communities still differed considerably. These results indicate a substantial change in insect communities with time following gap creation. The relationships between insect and plant communities at other stages of succession, and the time required for the herbivore community to return to pre-harvest levels remains largely unknown and warrants further study.

Acknowledgments

We thank S. Cahill, R. Malloy, W. Sikora, L. Reynolds for sorting insects, D. Dyer and J.

Campbell for helping with field work, and T. Champlin and L. Bowen for sharing vegetation data. Support was provided by the Department of Energy-Savannah River Operations Office through the U.S. Forest Service Savannah River under Interagency Agreement DE-AI09-

00SR22188. Funding was provided by the National Research Initiative Competitive Grants

Program of the USDA Cooperative State Research Education and Extension Service (CSREES

Grant No. 00-35101-9307).

References Cited

Bach, C. E. 1990. Plant successional stage and insect herbivory: flea beetles on sand-dune

willow. Ecology 71: 598-609.

Boardman, N. K. 1977. Comparative photosynthesis of sun and shade plants. Annual Review

of Plant Physiology 28: 355-377.

Boring, L. R., C. D. Monk, and W. T. Swank. 1981. Early regeneration of a clear-cut southern

Appalachian forest. Ecology 62: 1244-1253.

Bormann, F. H., and G. E. Likens. 1979. Pattern and process in a forested ecosystem.

Springer-Verlag, New York, New York, USA.

Breedlove, D. E., and P. R. Ehrlich. 1968. Plant-herbivore evolution: lupines and lycaenids.

Science 162: 671-672.

Brokaw, N. V. L. 1982. Treefalls: frequency, timing and consequences. In: Leigh EG, Rand

AS, Windsor DM (eds) The ecology of a tropical forest. Seasonal rhythms and long-term

changes. Washington, DC: Smithsonian Institute.

Brown, V. K. 1984. Secondary succession: insect-plant relationships. Bioscience 34: 710-716.

Brown, V. K. 1985. Insect herbivores and plant succession. Oikos 44: 17-22.

Brown, V. K., and A. C. Gange. 1992. Secondary plant succession: how is it modified by insect

herbivory? Vegetatio 101: 3-13.

Clinton, B. D., L. R. Boring, and W. T. Swank. 1993. Canopy gap characteristics and drought

influences in oak forests of the Coweeta Basin. Ecology 74: 1551-1558.

Coley, P. D. 1983. Herbivory and defensive characteristics of tree species in a lowland tropical

forest. Ecological Monographs 53: 209-233.

Cromer, R. B., J. D. Lanham, and H. H. Halin. 2002. Herpetofaunal Response to gap and

skidder-rut wetland creation in a southern bottomland hardwood forest. Forest Science.

48: 407-413.

Denslow, J. S., J. C. Schultz, P. M. Vitousek, and B. R. Strain. 1990. Growth responses of

tropical shrubs to treefall gap environments. Ecology 71: 165-179.

Godfray, H. C. J. 1985. The absolute abundance of leaf miners on plants of different

successional stages. Oikos. 45: 17-25.

Gorham, L. E., S. L. King, B. D. Keeland, and S. Mopper. 2002. Effects of Canopy Gaps

and Flooding on Homopterans in a Bottomland Hardwood Forest. Wetlands 22: 541-549.

Guldin, J. M. 1996. The role of uneven-aged silviculture in the context of ecosystem

management. Western Journal of Applied Forestry 11: 4-13.

Hendrix, S. D., V. K. Brown, and A. C. Gange. 1988. Effects of insect herbivory on early plant

succession: Comparison of an English site and an American site. Biological Journal of

the Linnean Society 35: 205-216.

Hill, J., K. Hamer, J. Tangah, and M. Dawood. 2001. Ecology of tropical butterflies in

rainforest gaps. Oecologia 128: 294-302.

Horn, S., M. D. Ulyshen, and J. L. Hanula. 2005. Abundance of Green Tree

Frogs and Insects in Artificial Canopy Gaps in a Bottomland Hardwood Forest. The

American Midland Naturalist. (in press).

Hunter, M. L. 1990. Wildlife, forests, and forestry. Prentice-Hall, Englewood Cliffs, New

Jersey, USA.

Kilgo, J. C., K. V. Miller, and W. P. Smith. 1999. Effects of group-selection timber harvest

in bottomland hardwoods on fall migrant birds. Journal of Field Ornithology. 70: 404-

413.

Koivula, M., and J. Niemelae. 2003. Gap felling as a forest harvesting method in boreal

forests: responses of carabid beetles (Coleoptera, Carabidae). Ecography. 26: 179-187.

Lawton, J. H. 1978. Host-plant influences on insect diversity: the effects of time and space. In

L. A. Mound and N. Waloff (eds), Diversity of Insect Faunas. Symposium of the Royal

entomological Society of London, 9: 105-125. Lawton, J. H., and S. McNeill. 1979. Between the devil and the deep blue sea: on the problems

of being a herbivore. In: Anderson, R. M., B. D. Turner, and L. R. Taylor (eds),

Population dynamics. Blackwell, Oxford, pp. 223-244.

Mattson, W. J. 1980. Herbivory in relation to plant nitrogen content. Annals of the

Entomological Society of America. 11: 119-162.

Meadows, J. S., and J. A. Stanturf. 1997. Silvicultural systems for southern bottomland

hardwood forests. Forest Ecology and Management 90: 127-140.

McBrien, H., R. Harmsen, and A. Crowder. 1983. A case of insect grazing affecting plant

succession. Ecology 64: 1035-1039.

McNeill, S., and T. R. E. Southwood. 1978. The role of nitrogen in the development of

insect/plant relationships. In: Harborn, J. B. (ed), Biochemical aspects of plant and

animal coevolution. Academic Press, London.

Menzel, M. A., T. C. Carter, J. M. Menzel, M. F. Ford, and B. R. Chapman. 2002. Effects

of group selection silviculture in bottomland hardwoods on the spatial activity

pattern of bats. Forest Ecology and Management. 162: 209-218.

Murdoch, W. W., F. C. Evans, and C. H. Peterson. 1972. Diversity and pattern in plants and

insects. Ecology 53: 819-829.

New, K. C., and J. L. Hanula. 1998. Effect of time elapsed after prescribed burning in

longleaf pine stands on potential prey of the Red-Cockaded Woodpecker. Southern Journal

of Applied Forestry 22: 175-183.

Phillips, D. L., and D. J. Shure. 1990. Patch-size effects on early succession in southern

Appalachian forests. Ecology 71: 204-212.

Runkle, J. R. 1981. Gap regeneration in some old-growth forests of the eastern United States.

Ecology 62: 1041-1051.

Runkle, J. R. 1982. Patterns of disturbance in some old-growth mesic forests of eastern

North America. Ecology 63: 1533-1546.

SAS Institute. 1985. SAS guide for personal computers. Version 6 ed., SAS Institute, Cary, NC.

Scriber, J. M., and F. Slansky Jr. 1981. The nutritional ecology of immature insects. Annual

Review of Entomology 26: 183-211.

Shelly, T. E. 1988. Relative Abundance of Day-Flying insects in Treefall Gaps vs.

Shaded Understory in a Neotropical Forest. Biotropica. 20: 114-119.

Shure, D. J., and D. L. Phillips. 1991. Patch size of forest openings and arthropod populations.

Oecologia. 86: 325-334.

Shure, D. J., and L. A. Wilson. 1993. Patch-size effects on plant phenolics in successional

openings of the southern Appalachians. Ecology. 74: 55-67.

Southwood, T. R. E. 1966. Ecological methods with particular reference to the study of insect

populations. Butler and Tanner Ltd., Frome and London.

Southwood, T. R. E., V. K. Brown, and P. M. Reader. 1979. The relationships of plant and

insect diversities in succession. Biological Journal of the Linnean Society. 12: 327-348.

Ulyshen, M. D., J. L. Hanula, S. Horn, J. C. Kilgo, and C. E. Moorman. 2004. Spatial

and temporal patterns of beetles associated with coarse woody debris in managed

bottomland hardwood forests. Forest Ecology and Management. 199: 259-272.

Ulyshen, M. D., J. L. Hanula, S. Horn, J. C. Kilgo, and C. E. Moorman. 2006.

The response of ground beetles (Coleoptera: Carabidae) to selection cutting in a South

Carolina bottomland hardwood forest. Biodiversity and Conservation. (in press).

White, P. S., M. D. MacKenzie, and R. T. Busing. 1985. Natural disturbance and gap phase

dynamics in southern Appalachian spruce-fir forests. Canadian Journal of Forest

Research 15: 233-240.

White, T. C. R. 1978. The importance of a relative shortage of food in animal ecology.

Oecologia (Berlin) 33: 71-86.

Wilder, C. M., F. W. Holtzclaw Jr., and E. E. C. Clebsch. 1999. Succession, Sapling

density and growth in canopy gaps along a topographic gradient in second growth east

Tennessee forest. American Midland Naturalist 142: 201-212.

Table 4.1. List, in decreasing abundance, of herbivorous insect groups collected in artificial

canopy gaps in a South Carolina bottomland hardwood forest, during 2001.

Family (Order) No. Species No. Individuals

Cicadellidae (Homoptera) 94 13186

Chrysomelidae (Coleoptera) 71 1481

Aphididae (Homoptera) 23 1093

Cixiidae (Homoptera) 26 863

Derbidae (Homoptera) 15 404

Flatidae (Homoptera) 5 323

Lygaeidae (Hemiptera) 22 234

Miridae (Hemiptera) 32 222

Immature Lepidoptera 56 123

Adult Lepidoptera 303 15292

Delphacidae (Homoptera) 13 118

Thysanoptera¹ 1 102

Cercopidae (Homoptera) 6 85

Acrididae (Orthoptera) 12 85

Psyllidae (Homoptera) 6 77

Pentatomidae (Hemiptera) 16 45

Achilidae (Homoptera) 7 45

Membracidae (Homoptera) 10 39

Issidae (Homoptera) 6 38

Tettigoniidae (Orthoptera) 8 20

Total Herbivores² 429 18583 96

Table 4.2. The percent similarity of herbivorous insects collected in young (~1 yr.) or old

(~7 yrs) canopy gaps by location (center, edge, or 50 m into surrounding forest) in a South

Carolina bottomland hardwood forest, 2001.

Comparison Percent Similarity

New Forest vs. Old Forest 72.85

New Edge vs. Old Edge 63.60

Old Gap vs. Old Forest 63.25

New Gap vs. Old Gap 48.91

New Gap vs. New Forest 47.25

Table 4.3. The mean (n=12) abundance (above) and richness (below) of herbivorous insects collected at the center, edge, and in

the surrounding forest of artificial canopy gaps in a bottomland hardwood forest, South Carolina, 2001. The young and old

gaps were created in 2000 and 1994, respectively. For each gap age, values with the same letter are not significantly different

(Ryan-Einot-Gabriel-Welsch Multiple Range Test, p<0.05). Asterisks denote significant differences (p<0.05) between the same

trap locations (e.g. center vs. center) in young and old gaps.

Young Gaps Old Gaps

Gap Edge Forest Gap Edge Forest

Coleoptera Chrysomelidae 30.25±6.14 A 23.33±6.95 A 14.33±2.54 A 18.58±2.69 a 19.33±4.54 a 17.58±2.67 a

9.92±0.65 A* 6.50±0.72 B 5.67±0.50 B 5.92±0.51 a* 5.42±0.47 a 5.92±0.50 a

Hemiptera Lygaeidae 10.75±1.92 A 2.17±0.58 B 0.50±0.19 B 2.92±0.99 a 1.42±0.31 a 1.75±0.81 a

4.08±0.51 A 1.75±0.28 B 1.08±0.08 B 1.83±0.21 a 1.42±0.19 a 1.58±0.26 a

Miridae 8.33±1.56 A 3.00±1.58 B 0.75±0.35 B 4.25±1.11 a 0.92±0.31 b 1.25±0.37 b

4.50±0.60 A* 1.50±0.19 B 1.25±0.18 B 2.75±0.41 a* 1.25±0.25 b 1.50±0.29 b

Pentatomidaea 1.25±0.22 0.58±0.19 0.42±0.29 0.67±0.33 0.42±0.26 0.42±0.26

1.25±0.13 A 1.08±0.08 A 1.00±0.00 A 1.17±0.17 a 1.08±0.08 a 1.08±0.08 a

Homoptera Achilidaec 2.33±1.24 0.67±0.28 0±0 0.25±0.25 0.33±0.19 0.17±0.11

1.33±0.19 1.00±0 1.00±0 1.00±0 1.00±0 1.00±0

Aphididae 37.00±11.24 A* 19.00±5.42 AB 6.25±1.01 B* 6.75±1.21 a* 9.75±2.65 a 12.33±2.19 a*

4.17±0.70 A 4.25±0.64 A 2.67±0.22 A 2.83±0.41 a 3.75±0.79 a 2.75±0.43 a

Cercopidaeb 3.17±1.29 A 0.83±0.24 B 0.50±0.19 B 1.42±0.50 a 0.58±0.29 a 0.58±0.15 a

1.33±0.14 1.17±0.17 1.08±0.08 1.17±0.11 1.00±0 1.00±0

Cicadellidae 468.00±70.41 A* 143.67±26.45 B 120.50±20.03 B 124.33±17.58 a* 133.50±20.14 a 108.83±21.25 a

23.83±1.77 A* 13.42±1.42 B 12.33±1.10 B 13.25±0.98 a* 12.08±0.80 a 11.42±0.76 a

Cixiidae 25.67±3.79 A* 11.25±3.25 B 12.83±2.05 B 2.00±0.44 a* 9.17±2.07 b 11.00±1.48 b

4.92±0.58 A* 3.58±0.45 A 3.92±0.51 A 1.50±0.19 a* 3.58±0.45 b 3.42±0.47 b

Delphacidae 4.50±1.51 A 1.33±0.38 B* 0.33±0.19 B 0.75±0.41 a 0.25±0.13 a* 2.67±2.11 a

2.42±0.40 A* 1.08±0.08 B 1.08±0.08 B 1.00±0 a* 1.00±0 a 1.00±0 a

Derbidae 11.42±2.13 A* 2.92±0.60 B 5.33±1.52 B 2.17±0.67 a* 5.33±1.55 ab 6.50±1.33 b

2.25±0.30 A 1.83±0.30 A 1.92±0.19 A 1.67±0.28 a 2.42±0.43 a 1.83±0.24 a

Flatidaeb 11.08±1.62 A 2.08±0.83 B 0.75±0.28 B 10.00±2.25 a 2.00±0.41 b 1.00±0.39 b

1.83±0.11 1.33±0.19 1.08±0.08 1.75±0.18 1.25±0.18 1.08±0.08

Issidaec 0.83±0.24 0.58±0.26 0.50±0.29 0.58±0.34 0.33±0.14 0.33±0.14

1.00±0 1.00±0 1.00±0 1.25±0.18 1.00±0 1.00±0

Membracidaea 0.25±0.13 0.42±0.19 0.75±0.30 0.08±0.08 1.33±1.07 0.42±0.29

1.00±0.00 A 1.00±0.00 A 1.17±0.11 A 1.00±0.00 a 1.00±0.00 a 1.25±0.18 a

Psyllidaeb 1.58±1.07 A 2.00±0.90 A 0.67±0.33 A 0.42±0.23 a 1.67±0.96 a 0.08±0.08 a

1.08±0.08 1.08±0.08 1.08±0.08 1.00±0 1.08±0.08 1.00±0

Lepidoptera

caterpillars 1.33±0.50 A 2.67±0.31 A 1.83±0.61 A 1.83±0.34 a 1.67±0.53 a 0.92±0.19 a

1.42±0.23 A 2.33±0.26 A 2.08±0.53 A 1.92±0.26 a 1.75±0.30 a 1.17±0.11 a

adults 350.33±31.11 A 173.67±15.19 B 130.17±12.06 B 223.17±34.40 a 207.17±25.21 a 189.83±17.76 a

61.08±2.44 A* 41.83±2.13 B 28.83±1.53 C 43.58±3.80 a* 39.58±3.86 a 28.50±1.16 b

Orthoptera Acrididae 3.25±0.62 A* 1.25±0.52 B 0.58±0.29 B 0.25±0.13 a* 1.42±0.63 a 0.33±0.19 a

2.17±0.30 A* 1.42±0.19 B 1.00±0.00 B 1.00±0.00 a* 1.25±0.18 a 1.00±0.00 a

Tettigoniidaec 0.42±0.15 0.25±0.18 0.08±0.08 0.33±0.14 0.42±0.15 0.17±0.17

1.00±0.00 1.00±0.00 1.00±0.00 1.00±0.00 1.00±0.00 1.08±0.08

Thysanopterad

5.00±1.89 A* 0.42±0.26 B 1.33±0.43 AB 0.50±0.19 a* 0.25±0.18 a 1.00±0.33 a

100

a The abundance among locations is not compared statistically due to insufficient numbers (<50 specimens) b The species richness among locations is not compared statistically due to insufficient numbers (<10 species) c both abundance and species richness are excluded from analysis due to insufficient numbers d Thrips were only identified to order so we have no measure of richness

101

Fig. 4.1. Mean±SE (n=12) richness (A) and abundance (B) of herbivorous insects collected in malaise and pitfall traps in 2001 at different locations in a bottomland hardwood forest in 1994 or 2000. Traps were placed at the “Center”, “Edge”, and in the nearby “Forest” of each gap.

Within graphs (for each treatment), bars with the same letter above them are not significantly different (Ryan-Einot-Gabriel-Welsch Multiple Range Test, p < 0.05). Asterisks denote significant differences (p < 0.05) between the same trap locations (e.g. center vs. center) in old

(1994) and new (2000) gaps.

102

120 A

100

80 A*

60 B 40 B a* a mean richness a

0 B A*

600

400

B a a* mean abundance B a 200

0 Center Edge Forest Center Edge Forest 2000 Treatment 1994 Treatment Location

103

Fig. 4.2. Mean±SE (n=4) richness (A) and abundance (B) of herbivorous insects collected in malaise and pitfall traps in 2001 at the centers of different sized canopy gaps (0.13, 0.26, and

0.50 ha) created in a bottomland hardwood forest in 1994 or 2000. Within graphs (for each treatment), bars with the same letter above them are not significantly different (Ryan-Einot-

Gabriel-Welsch Multiple Range Test, p < 0.05). Asterisks denote significant differences (p <

0.05) between the same trap locations (e.g. center vs. center) in old (1994) and new (2000) gaps.

104

100 A A* 80 A* A*

a* a* a* 40 mean richness

0 B A 1000

800 A*

600 A*

400 mean abundance a a* 200 a*

0 0.13 ha 0.26 ha 0.50 ha 0.13 ha 0.26 ha 0.50 ha 2000 Treatment 1994 Treatment Gap Size

105

CHAPTER 5

THE RESPONSE OF GROUND BEETLES (COLEOPTERA: CARABIDAE) TO SELECTION

CUTTING IN A SOUTH CAROLINA BOTTOMLAND HARDWOOD FOREST

¹Ulyshen, M. D., Hanula, J. L., Horn, S., Kilgo, J. C., Moorman, C. E., 2004. Accepted by

Biodiversity and Conservation.

Reprinted here with permission of publisher.

106

Abstract

We compared the response of ground beetles (Coleoptera: Carabidae) to the creation of canopy gaps of different size (0.13 ha, 0.26 ha, and 0.50 ha) and age (1 and 7 yrs) in a bottomland hardwood forest (South Carolina, USA). Samples were collected 4 times in 2001 by malaise and pitfall traps placed at the center and edge of each gap, and 50 m into the surrounding forest.

Species richness was higher at the center of young gaps than in old gaps or in the forest, but there

was no statistical difference in species richness between old gaps and the forests surrounding

them. Carabid abundance followed the same trend, but only with the exclusion of

Semiardistomis viridis (Say), a very abundant species that differed in its response to gap age

compared to most other species. The carabid assemblage at the gap edge was very similar to that

of the forest, and there appeared to be no distinct edge community. Species known to occur in

open or disturbed habitats were more abundant at the center of young gaps than at any other location. Generalist species were relatively unaffected by the disturbance, but one species

(Dicaelus dilatatus Say) was significantly less abundant at the centers of young gaps. Forest

inhabiting species were less abundant at the centers of old gaps than in the forest, but not in the

centers of young gaps. Comparison of community similarity at various trapping locations

showed that communities at the centers of old and young gaps had the lowest similarity (46.5%).

The community similarity between young gap centers and nearby forest (49.1%) and old gap

centers and nearby forest (50.0%) was similarly low. These results show that while the

abundance and richness of carabids in old gaps was similar to that of the surrounding forest, the

species composition between the two sites differed greatly.

107

Key words: Carabidae, ground beetles, disturbance, group selection harvesting, edge effect, canopy gaps.

108

Introduction

Southeastern bottomland hardwood forests are important for water quality and control, nutrient cycling, wildlife habitat, and they support among the most diverse plant and animal communities in North America (Kellison and Young 1997). To protect this unique ecosystem, and to satisfy increasing demand for forest products, the remaining stands must be maintained and managed properly. According to Guldin (1996), proper forest management attempts to imitate natural rates of succession and disturbance in order to minimize the environmental impacts of timber removal. One promising method for use in bottomland hardwood forests is group selection cutting, an uneven age forest management practice that emulates small-scale natural disturbances (i.e. tree falls, insect outbreaks, wind damage, etc.) to create small openings throughout the forest (Hunter 1990, Meadows and Stanturf 1997, Guldin 1996).

Ground beetles (Carabidae) are taxonomically well known, easily and inexpensively surveyed, and respond quickly to environmental change (Rainio and Niemelä 2003). These attributes have made them useful bioindicators in numerous studies involving disturbance

(Rainio and Niemelä 2003, Allegro and Sciaky 2003).

While the response of ground beetles to clearcuts in the conifer forests of Europe and northeastern North America has been well studied (Niemelä et al. 1993, Altegrim et al. 1997,

Beaudry et al. 1997, Niemelä 1997, Duchesne et al. 1999, Heliola et al. 2001, Koivula 2002a,

Koivula et al. 2002, Magura et al. 2003, Pearce et al. 2003), little work has been done on alternative harvesting methods (Altegrim et al. 1997, Werner and Raffa 2000, Vance and Nol

2003, Koivula 2002b, Koivula and Niemelä 2003, Moore et al. 2004), or in hardwood forests

(Lenski 1982a, Warriner et al. 2002, Vance and Nol 2003, Moore et al. 2004).

109

Here we report the results of the first study to examine the response of carabids to group selection cutting in a bottomland hardwood forest in the southeastern United States. We compare the abundance and species richness of carabids in canopy gaps of different size (0.13,

0.26, and 0.50 ha) and age (1 or 7 years) to those at gap edge and in the surrounding forest.

Materials and Methods

Study Site

This study was conducted from May to November 2001 on the Savannah River Site

(SRS), an 80,269-ha nuclear production facility near Aiken, South Carolina. The SRS is owned

and operated by the United States Department of Energy (DOE) as a National Environmental

Research Park. Our study site was an approximately 120-ha stand of 75-100 year-old

bottomland hardwoods. Common forest trees included numerous oak species (Quercus spp.),

bald cypress (Taxodium distichum (L.) Richard), sweetgum (Liquidambar styraciflua L.), red

maple (Acer rubrum L.), and loblolly pine (Pinus taeda L.). The midstory consisted

predominantly of red mulberry (Morus rubra L.), ironwood (Carpinus caroliniana Walter) and

American holly (Ilex opaca Aiton). The understory was dominated by dwarf palmetto (Sabal

minor (Jacquin) Persoon) and switchcane (Arundinaria gigantean (Walter) Muhl.). Pre-harvest

basal area of the stands was 33m²/ha (Pauley et al. 1996). The study site often experiences

seasonal flooding (January-April) with some low-lying areas remaining under water much of the

year. Total rainfall in 2001 was 104 cm with the wettest month being June (23.4 cm) and the

driest being December (1.2 cm).

110

Gaps

Of the 24 gaps used in this study, 12 were created in December 1994 (“old gaps”) and 12 in August 2000 (“young gaps”). There were 4 replicates of three different sizes (0.13, 0.26, and

0.50 ha) for each gap age. The gap area was defined as the area surrounded by the boles of the peripheral dominant forest trees. The gaps were located throughout the 120 ha bottomland hardwood forest, and were spaced at least 200 m apart. Vegetation in old gaps was 1-8 m in height and consisted of pioneer species such as sweetgum, sycamore (Platanus occidentalis L.), green ash (Fraxinus pennsylvanica Marshall), black willow (Salix nigra Marshall), tulip poplar

(Liriodendron tulipifera L.), oaks, switchcane, and dwarf palmetto. Young gaps contained small stump sprouts or seedling of these species as well as fireweed (Erechtites hieracifolia (L.) Raf.), blackberries (Rubus spp.), and plumegrass (Erianthus giganteus (Walter) Muhl.), other native grasses, and various sedge species (Cyperus spp.).

Beetle Sampling and Identification

Ground beetles were sampled at the center and edge of each gap and in the surrounding forest 50 m from gap edges during four 7-day trapping periods (17-23 May, 10-16 July, 7-13

September, and 3-9 November). Each sample location had a malaise and two pitfall traps to capture flying and crawling beetles, respectively. Malaise traps (“Canopy Traps”, Sante Traps,

Lexington, KY) differed from the traditional design in that they contained collecting jars at the top and bottom so insects that fall when encountering a barrier were also collected. The traps were suspended from 3 m tall metal hangers.

111

Pitfall traps consisted of a 480 ml plastic cup buried to ground level. A small funnel (8.4 cm diameter) inserted into the cup directed captured beetles into a smaller 120 ml specimen cup below. The pitfall was positioned at the intersection of four 0.5 m long drift fences. Two pitfall traps were placed 5 m apart at each sample station, and the samples from each were combined for each location (center, edge, and forest). The collecting jars for both pitfall and malaise traps were filled with NaCl-2% formaldehyde solution to preserve specimens and a drop of detergent to reduce surface tension (New and Hanula 1998). Once collected, beetles were brought back to the lab and immediately stored in 70% alcohol. Specimens were sorted to morphospecies and later identified using a reference collection and a key to South Carolina Carabidae (Ciegler

2000). In the interest of accuracy, we were unable to assign species-level names to several morphospecies.

We assigned the most abundant species to categories (open-habitat species (fields, meadows, and disturbed areas), generalist species (open or forested areas), and forest species

(forested areas)) based on known habitat data (Larochelle and Lariviere 2003). Not all species were classified into these categories due to inadequate information or to the genus-level identification of several morphospecies (Table 5.1).

Statistical Analysis

We combined malaise and pitfall trap captures at each location before analyzing the

results. A 3-way analysis of variance with gap age, trap location, and gap size as the main

effects showed a significant interaction between gap age and trap location so we analyzed the

data for each gap age separately. The General Linear Model procedure of SAS (SAS institute

112

1985) was used for all analyses and the Ryan-Einot-Gabriel-Welsch Multiple Range Test was

used to determine differences (α < 0.05 unless otherwise stated) in relative abundance of insects

between trap locations or gap sizes for each gap age (Day and Quinn 1989). We used Raabe’s

percent similarity (Southwood 1966) to compare similarity among trap locations and trap types.

Results

In total, 5,498 ground beetles were collected representing 26 tribes, 60 genera, and 87 species. Species richness was higher at the center of young gaps than in old gaps or in the forest, but there was no statistical difference in species richness between old gaps and the forests surrounding them (Figure 5.1). Carabid abundance followed the same trend (Figure 5.2), but only with the exclusion of Semiardistomis viridis (Say), a very abundant species (23% of the total number) that differed in its response to gap age compared to most other species (Table 5.1).

There was no statistical difference in abundance or species richness among gaps of differing size

(Figure 5.3).

We were able to classify 19 of the 31 most abundant (>25 individuals) species as open- habitat species, generalists, or forest dwellers (Table 5.1). In general, carabids associated with open-habitat responded positively to canopy gap creation, and were more abundant at the centers of young gaps than at other young or old gap locations (center, edge, or forest) (Figure 5.4). The number of open habitat species at the centers of old gaps was comparable to that of the surrounding forest. Likewise, the abundance of generalist carabids was similar among both young and old gap locations (Figure 5.4). Carabids that prefer forest habitats were less abundant

113 at the centers of young and old gaps than in their respective forest locations, but this was only significant (P < 0.1) for old gaps (Figure 5.4).

Of the 31 most abundant species, ten species exhibited a significant difference among young gap locations and five differed significantly among old gap locations (Table 5.1). Eight of the ten species that differed among young gap locations were more abundant in the centers of young gaps than in the surrounding forest. Conversely, only two (Acupalpus sp. 2 and S. viridis

) were more abundant at the center of old gaps than in the surrounding forest (Table 5.1). While

Acupalpus sp. 2 was more abundant at the centers of both young and old gaps, S. viridis was much more numerous in old gap centers than at the edge or in the forest. It was also significantly more abundant in old gap centers than in young gap centers. Several abundant species appeared to respond positively to recent disturbance, but only Notiobia terminata (Say) was found exclusively in young gaps (Table 5.1).

The most similar carabid assemblages were those at the edges of gaps and the forests surrounding them (Table 5.2). The edges of old and young gaps also had a high degree of similarity (72%). The least similar carabid communities were those at the centers of young and old gaps (Table 5.2), but carabid assemblages in gap centers and surrounding forests also had relatively low similarity.

Discussion

Many studies have shown an overall increase in the species richness and/or abundance of carabids following disturbance (Niemelä et al. 1993, 1994, Warriner et al. 2002, Beaudry et al.

1997, Heliola et al. 2001, Koivula et al. 2002, Thompson and Allen 1993, Eryschov and

Trophimova 1984). While some studies have found no overall change in carabid abundance or

114 species richness, they have identified significant effects at the species level (Atlegrim et al. 1997) as well as differences in species composition between disturbed and undisturbed sites

(Greenburg and Thomas 1995, Butterfield 1997, Werner and Raffa 2000). As might be expected, habitat specificity appears to determine the response of many carabids. The abundance of open habitat species, for example, has been shown to increase in disturbed areas, while the numbers of forest-dwelling species often decreases or disappears following disturbance (Niemela et al. 1993).

Our results are generally consistent with these trends, but there appear to be substantial differences in the abundance, species richness, and community composition of carabids with time after disturbance. For example, the carabid abundance, richness, and species composition differed greatly between the centers of young and old gaps. Furthermore, while species composition differed greatly between both young and old gaps centers and their respective forest locations, differences between the abundance and species richness of carabids at the centers of gaps and the forests surrounding them was significant only for young gaps. Open-habitat species were more abundant at the centers of young gaps than in the surrounding forest, but there was no difference in abundance between the centers of old gaps and the forests surrounding them.

Conversely, forest species were less abundant at the centers of gaps than in the forest, but only for old gaps was this difference significant. Thus, the carabid communities present at the centers of old gaps differed greatly from those found at the centers of young gaps as well as from those in the forests surrounding old gaps.

Past studies have also noted changes in carabid communities with time after disturbance.

For example, in a study involving single-tree selection cutting, Vance and Nol (2003) found reduced activity densities in recently (0.5-3 yrs) cut stands compared to reference stands, while

115 the activity densities for certain species was higher in older (15-20 yrs) cut stands. The authors attribute these differences to significant reductions in leaf litter in the recently cut stands, and to differences in the vegetation in older stands. The importance of factors such as vegetation structure, temperature, humidity, light intensity, and soil moisture to ground beetles is well supported by past research (Antvogel and Bonn 2001, Magura et al. 1997, Cardenas and Bach

1989, Lenski 1982a, Thompson and Allen 1993, Warriner et al. 2002).

Reduced competitive exclusion may have played a role in the higher abundance and species richness of carabids observed in young gaps (Lenski 1982a, 1982b, Allen and Thompson

1977), but we suspect that it had relatively little effect in this study. The increase in habitat

heterogeneity following disturbance was probably much more important. For example, timber

removal created large amounts of coarse woody debris and greatly increased the complexity of

the gap floor. While young gaps contained an abundance of CWD, little remained in the old

gaps. Differences in vegetation between young and old gaps were similarly dramatic. In

contrast to young gaps in which there were scattered clumps of grasses, tree sprouts, and

herbaceous growth, old gaps were covered in a dense growth of young trees competing for

sunlight. Because young and old gaps were so different in habitat structure, it is not surprising

that carabid abundance, species richness, and composition differed greatly between the two

locations.

Because the carabid communities at the edges of young and old gaps were so similar to

those in the surrounding forest, we have little indication of a distinct edge community. Although

researchers in Hungary reported unique edge communities as well as several species unique to

edge habitats (Magura and Tothmeresz 1997, Magura et al. 2001, Magura 2002), the results from

116 other studies are similar to our own (Heliola et al. 2001, Spence et al. 1996, Kotze and Samways

2001).

The carabid community in seven-year old gaps is far from recovered, despite comparable

abundance and species richness between old gaps and the surrounding forest. This is indicated

by the low degree of similarity between the two sites. In fact, carabids at the centers of old gaps

are only slightly more similar to those in the forest than are the carabids at young gap centers

(50.0 % and 49.1 % similar, respectively). These results emphasize the fact that abundance

should not be used alone (Moore et al. 2004) to determine the recovery time of carabid

assemblages.

Although Vance and Nol (2003) found an increase in both open habitat species and forest

generalists 0.5-3 and 15-20 years after single-tree selection harvests, we could identify no

common trend among carabids between young and old canopy gaps. The response of carabids to

young and old gaps differed greatly, even among species with similar habitat preferences. For

example, of the 6 common open-habitat species in this study, 3 were significantly more abundant at young gap centers than at the edges or in the forest surrounding young gaps, but there were no differences among old gap locations. Furthermore, of the 31 most abundant morphospecies collected, 13 exhibited a significant difference among either young or old gap locations. Of these, only 5 differed significantly among old gap locations and just two responded similarly to young and old gaps.

These differences in abundance between young and old gaps are probably due to the specific habitat requirements of each species. Given this, it is interesting to note that just two species (S. viridis and an Acupalpus species) were more abundant at the centers of old gaps than in the nearby forest. While the Acupalpus species was more abundant at the center of young

117 gaps than old gaps, S. viridis was more abundant at the centers of old gaps than at any other young or old gap location. This result further emphasizes the importance of time considerations when studying the effects of disturbance on ground beetles, as well as the species-specific response of carabids to disturbance.

While many species tend to be more abundant in disturbed habitats, several have been shown to exist there exclusively (Warriner et al. 2002, Beaudry et al. 1997, Niemelä et al. 1993,

Thompson and Allen 1993). For example, in this study, N. terminata was collected only in the center or at the edge of young gaps. Similarly, a number of forest species were found in much greater numbers in the forest than elsewhere. While we found no substantial evidence for the presence of strict forest specialists, such species may have been collected in low numbers (and could not be analyzed statistically) or not at all. Since total carabid abundance in the forest near young gaps was different from that near old gaps, gap creation had a definite effect on the carabid community at least 50 m into the surrounding forest. Because of this, obligate forest species, if present, may have found the forests surrounding the gaps to be unsuitable. Although many carabids will eventually recolonize an area after disturbance (Koivula et al. 2002) some forest specialists are unable to reestablish populations in regenerating clear-cut stands (Beaudry et al. 1997) and old growth species with poor dispersal ability may face local extinction if stands of mature forest are not preserved (Koivula et al. 2002, Spence et al. 1996, Heliola et al. 2001,

Halme and Niemelä 1993, Beaudry et al. 1997). Because group selection harvesting disturbs smaller patches of bottomland hardwood forest at any one time and is more similar to natural levels of disturbance, it may lessen the detrimental effects of disturbance on these sensitive forest species.

118

How group selection cutting compares to other forestry practices, remains unclear.

Recent work in Finland has found small (0.16 ha) openings to be less disruptive of community structure than larger clear-cut stands (Koivula 2002b, Koivula and Niemelä 2003) but much more comparative work is needed to ascertain the advantages of various harvesting techniques

with respect to environmental health. Different forests have different natural rates of disturbance

(Hunter 1990, Guldin 1996) so the effects of a particular management technique may depend upon the forest type under consideration. Because few carabids were negatively affected by gap

creation, and none seemed to be completely eliminated by the disturbance, we feel that group

selection cutting may be particularly well suited to bottomland hardwood forests and deserves further consideration.

119

Acknowledgments

We thank H. Lee, Jr. and C. Smith for assisting with the identification of specimens, D. Dyer and

J. Campbell for helping with field work, and J. McHugh for editing an early draft of the manuscript. We also thank the U.S. Department of Energy-Savannah River and the USDA Forest

Service-Savannah River for access and logistical support. This research was funded by the

National Research Initiative Competitive Grants Program of the USDA Cooperative State

Research Education and Extension Service (CSREES Grant No. 00-35101-9307).

120

References

Allegro G and Sciaky R (2003) Assessing the potential role of ground beetles

(Coleoptera, Carabidae) as bioindicators in poplar stands, with a newly proposed

ecological index (FAI). Forest Ecology and Management 175: 275-284

Allen RT and Thompson RG (1977) Faunal composition and seasonal activity of

Carabidae (Insecta: Coleoptera) in three different woodland communities in

Arkansas. Annals of the Entomological Society of America 70: 31-34

Altegrim O, Sjoeberg K and Ball JP (1997) Forestry effects on a boreal ground beetle

community in spring: Selective logging and clear-cutting compared. Entomologica

Fennica 8: 19-26

Antvogel H and Bonn A (2001) Environmental parameters and microspatial distribution

of insects: a case study of carabids in an alluvial forest. Ecography 24: 470-482

Beaudry S, Duchesne LC and Cote B (1997) Short-term effects of three forestry practices

on carabid assemblages in a jack pine forest. Canadian Journal of Forest Research 27:

2065-2071

Butterfield J (1997) Carabid community succession during the forestry cycle in conifer

plantations. Ecography 20: 614-625

Cardenas AM and Bach C (1989) The effect of some abiotic factors on the distribution

and selection of habitat by the carabid beetles in the central Sierra Morena Mountains

(SW Cordoba, Spain). Vie et milieu 39: 93-103

Ciegler JC (2000) Wheeler AG, ed. Ground beetles and wrinkled bark beetles of south

121

Carolina (Coleoptera: Geadephaga: Carabidae and Rhysodidae). Clemson University

149pp

Day RW and Quinn GP (1989) Comparisons of treatments after an analysis of variance in

ecology. Ecological Mongraphs 59: 433-463

Duchesne LC, Lautenschlager RA, Bell FW (1999) Effects of clear-cutting and plant

competition control methods on carabid (Coleoptera: Carabidae) assemblages in

northwestern Ontario. Environmental Monitoring and Assessment 56: 87-96

Eryschov VI and Trophimova OL (1984) Changes in the population of ground beetles

(Coleoptera, Carabidae) in a clearing in the mountain taiga of the Kuznetsky Alatau

foothills. Zoologicheskij zhurnal. Moscow 63: 848-852

Greenburg CG and Thomas MC (1995) Effects of forest management on coleopteran

assemblages in sand pine scrub. Florida Entomologist 78: 271-285

Guldin JM (1996) The role of uneven-aged silviculture in the context of ecosytem

management. Western Journal of Applied Forestry 11: 4-13

Halme E and Niemelä J (1993). Carabids beetles in fragments of coniferous forest.

Annales Zoologici Fennici 30:17-30

Heliola J, Koivula M and Niemelä J (2001) Distribution of carabid beetles (Coleoptera,

Carabidae) across a boreal forest-clearcut ecotone. Conservation Biology 15:

370-377

Hunter ML (1990) Wildlife, forests, and forestry. Prentice-Hall, Englewood Cliffs, New

Jersey, USA

Kellison RC and Young MJ (1997) The bottomland hardwood forest of the southern

United States. Forest Ecology and Management 90: 101-115

122

Koivula M (2002a) Boreal carabid-beetle (Coleoptera: Carabidae) assemblages in thinned

uneven-aged and clear-cut spruce stands. Annales Zoologici Fennici 39: 131-

149

Koivula M (2002b) Alternative harvesting methods and boreal carabid beetles

(Coleoptera: Carabidae). Forest Ecology and Management 167: 103-121

Koivula M, Kukkonen J and Niemelä J (2002) Boreal carabid-beetle (Coleoptera:

Carabidae) assemblages along the clear-cut originated succession gradient. Biodiversity

and Conservation 11: 1269-1288

Koivula M and Niemelä J (2003) Gap felling as a forest harvesting method in boreal

forests: responses of carabid beetles (Coleoptera: Carabidae). Ecography 26: 179-187

Kotze DJ and Samways MJ (2001) No general edge effects for invertebrates at

Afromontane forest/grassland ecotones. Biodiversity and Conservation 10: 443-

466

Larochelle A and Lariviere M-C (2003) A natural history of the ground beetles

(Coleoptera:Carabidae) of America north of Mexico. Pensoft Series Faunistica No 7.

Pensoft Publishers. 583 pp.

Lenski RE (1982a) The impact of forest cutting on the diversity of ground beetles

(Coleoptera: Carabidae) in the southern Appalachians. Ecological Entomology 7: 385-

390

Lenski RE (1982b) Effects of forest cutting on two Carabus species: evidence for

competition for food. Ecology 63: 1211-1217

Magura T and Tothmeresz B (1997) Testing edge effect on carabid assemblages in an

oak-hornbeam forest. Acta Zoologica Academiae Scientarum Hungaricae 43:

123

303-312

Magura T, Thothmeresz B and Bordan Z (1997) Comparison of the carabid communities

of a zonal oak-hornbeam forest and pine plantations. Acta Zoologica Academiae

Scientiarum Hungaricae 43: 173-182

Magura T, Tothmeresz B and Molnar T (2001) Forest edge and diversity: carabids along

forest-grassland transects. Biodiversity and Conservation 10: 287-300

Magura T (2002) Carabids and forest edge: spatial pattern and edge effect. Forest

Ecology and Management 157: 23-37

Magura T, Tothmeresz B, Elek Z (2003) Diversity and composition of carabids during a

forestry cycle. Biodiversity and Conservation 12: 73-85

Meadows JS and Stanturf JA (1997) Silvicultural systems for southern bottomland

hardwood forests. Forest Ecology and Management 90: 127-140

Moore J-D, Ouimet R, Houle D, Camire C (2004) Effects of two silvicultural practices on

ground beetles (Coleoptera: Carabidae) in a northern hardwood forest, Quebec, Canada.

Canadian Journal of Forest Research 34: 959-968

New KC and JL Hanula (1998) Effect of time elapsed after prescribed burning in

longleaf pine stands on potential prey of the Red-Cockaded Woodpecker. Southern

Journal of Applied Forestry 22(3): 175-183

Niemelä J (1997) Invertebrates and boreal forest management. Conservation Biology 11:

601-610

Niemelä J, Langor D and Spence JR (1993) Effects of clear-cut harvesting on boreal

ground-beetle assemblages (Coleoptera:Carabidae) in western Canada. Conservation

Biology 7: 551-561

124

Niemelä J, Spence JR, Langor D, Haila Y, Tukia H (1994) Logging and boreal ground-

beetle assemblages on two continents: implications for conservation. In: Gaston K,

Samways M, New T (Eds) Persepctives in Insect Conservation, Intercept, Andover: 29-

Pauley EF, Collins BS, Smith WP (1996) In: Flynn KM (Ed.). Early

establishment of cherrybark oak in bottomland hardwood gaps: effects of seed predation,

gap size, herbivory, and competition. Proceedings of the southern forested wetlands

ecology and management conference. Clemson, S.C. Pp. 132-136

Pearce JL, Venier LA, McKee J, Pedlar J, McKenney D (2003) Influence of habitat and

microhabitat on carabid (Coleoptera: Carabidae) assemblages in four stand types. The

Canadian Entomologist 135: 337-357

Rainio J and Niemelä J (2003) Ground beetles (Coleoptera: Carabidae) as bioindicators.

Biodiversity and Conservation 12: 487-506

SAS Institude (1985) SAS Guide for Personal Computers, 6th ed. SAS Institue, Cary, NC,

378 pp

Southwood TRE (1966) Ecological methods with particular references to the study of

insect populations. Methuen & Co. Ltd. London 391 pp.

Spence JR, Langor DW, Niemelä J, Carcamo HA and Currie CR (1996) Northern

forestry and carabids: The case for concern about old-growth species. Annales Zoologici

Fennici 33: 173-184

Thompson LC and Allen RT (1993) Site preparation affects ground beetles in a clearcut

bottomland hardwood forest in southeastern Arkansas. In: Brissette JC, ed. Proceedings

of the seventh biennial southern silvicultural research conference; 1992 November 17-19,

125

Mobile, AL. Gen. Tech. Rep. SO-93. New Orleans, LA: U.S. Department of Agriculture,

Forests Service, Southern Forest Experiment Station: 57-64

Vance CC, Nol E (2003) Temporal effects of selection logging on ground beetle

communities in northern hardwood forests of eastern Canada. Ecoscience 10: 49-56

Warriner MD, Nebeker TE, Leininger TD and Meadows JS (2002) The effects of

thinning on beetles (Coleoptera: Carabidae, Cerambycidae) in bottomland hardwood

forests. In: Outcalt, KW, ed. Proceedings of the eleventh biennial southern silvicultural

research conference. Gen. Tech. Rep. SRS-48. Asheville, NC: U.S. Department of

Agriculture, Forest Service, Southern Research Station: 569-573

Werner SM and Raffa KF (2000) Effects of forest management practices on the diversity

of ground-occurring beetles in mixed northern hardwood forests of the great lakes region.

Forest Ecology and Management 139: 135-155

126

Table 5.1. Mean (±SE) number of the most abundant (>25 specimens collected) carabid species collected in malaise and pitfall traps in

2001 at different locations in bottomland hardwood forest gaps (n=12) created in 1994 (old) and 2000 (new). For each species and gap age,

values with the same letter are not significantly different (Ryan-Einot-Gabriel-Welsch Multiple Range Test, p < 0.05). Asterisks denote

significant differences (p < 0.05) between the same trap locations (e.g. center vs. center) in young and old gaps.

Young Gaps Old Gaps Species Center Edge Forest Center Edge Forest Open-Habitat Species Acupalpus testaceous Dejean A 1.50±1.04 A 1.17±.63 A .42±.23 a 1.17±.56 a 2.25±1.92 a 0 Clivina bipustulata (Fabricius) A* 5.50±1.64 A 2.33±.67 A 3.67±.96 a* .92±.36 a 2.25±.72 a 2.00±.28 Harpalus pennsylvanicus (De Geer) A *2.00±.77 B .25±.12 B .08±.08 a *.17±.11 a .08±.08 a .17±.17 Notiobia terminata (Say) A* 1.92±.74 B* .42±.19 B 0 a* 0 a* 0 a 0 Poecilus chalcites (Say) A .33±.19 A .33±.19 A .83±.39 a .17±.11 a .42±.23 a .50±.29 Scarites spp. Fabricius A* 3±.70 B* 1.17±.37 B .92±.29 a*.17±.11 a*.17±.17 a .58±.26 Generalist Species Brachinus alternans Dejean A *21.83±4.93 A *16.25±5.05 A 9.50±3.48 a *5.33±2.25 a *4.08±1.74 a *3.67±1.97 Carabus sylvosus Say A 1±.66 A 1.5±.87 A .58±.23 a .25±.13 a .17±.11 a .42±.19 Dicaelus dilatatus Say A .08±.08 AB .58±.19 B 1.08±.36 a .17±.11 a .83±.34 a .92±.36 Dicaelus elongatus Bonelli A*.08±.08 A .75±.33 A 1.17±.56 a* 1.00±.35 a .5±.26 a .33±.14 Galerita spp. Fabricius A .58±.34 A .75±.33 A .42±.26 a 1.08±.36 a .42±.26 a .67±.43 Stenolophus ochropezus (Say) A 1.75±.62 A* 1.08±.42 A .33±.26 a 1.42±.74 a .08±.08 a 0

127

Forest Species Chlaenius aestivus Say A 5.08±1.39 A 7.58±2.92 A 5.67±2.60 a 5.42±1.78 a 3.92±1.49 a 3.50±1.98 Chlaenius erythropus Germar A 1.42±.71 A 1.25±.65 A .83±.21 a .58±.29 a 1.58±.99 a .42±.19 Cyclotrachelus brevoorti (LeConte) A .25±.13 A .75±.28 A 1.08±.80 a .08±.08 a 1.00±.83 a .5±.34 Diplocheila assimilis (LeConte) A .83±.58 A .50±.42 A .75±.35 a .25±.13 ab 1.33±.61 b 2.25±.72 Lophoglossus gravis LeConte A 2.50±.77 A 6.58±2.28 A 5.17±1.36 a 2.42±.82 a 4.92±1.32 a 4.25±1.49 Olisthopus spp. Dejean A .33±.19 A .92±.67 B 4.83±1.10 a 0.58±.19 b 1.67±.64 b 4.08±.88 Piesmus submarginatus (Say) A .75±.75 A 1.83±.76 A .67±.36 a .08±.08 a 1.5±.60 a 3.67±1.77 Unknown Habitat Requirements Acupalpus sp. 2 A* 8.67±2.81 B* 3.00±1.04 B 1.33±0.80 a* 1.17±0.32 b* 0.50±0.15 b 0.25±0.13 Agonum aeruginosum Dejean A .67±.28 A .25±.18 A .83±.21 a .08±.08 a .08±.08 a .33±.19 Agonum decorum (Say) A* 3.17±.89 AB* 2.33±.69 B .42±.26 a *.08±.08 a *.58±.36 a .33±.19 Chlaenius sp. 3 A .92±.42 A 4.17±1.73 A 3.75±1.41 a .75±.45 a 1.42±.68 a 1.58±.63 Clivina rubicunda LeConte A* 2.0±.62 B .42±.15 B 0 a* 0 a .17±.11 a .08±.08 Cymindis spp. A .33±.14 A 0.75±.33 A .92±.36 a .33±.19 a 1.17±.34 a 1.5±.65 Loxandrus sp. LeConte A* 4.83±1.09 A 2.58±0.50 A 2.42±0.70 a* 0.67±0.22 a 2.08±0.82 a 1.75±0.73 Micratopus aenescens (LeConte) A* 14.83±5.96 A 7.17±1.67 A* 2.67±.53 b* 1.08±.34 a 5.33±1.80 ba* 4.25±.84 Oodes amaroides Dejean A* 3.42±.92 B* 1.50±.54 B .25±.18 a* .42±.19 a* .25±.13 a .17±.11 Oodes sp. 2 A .83±.59 A .83±.59 A .83±.44 a .25±.18 a .17±.11 a .17±.11 Paratachys spp. Casey A* 4.33±1.51 B .58±.19 AB 2.08±1.02 a* .92±.51 a 1.67±.90 a .83±.30 Semiardistomis viridis (Say) A* 5.42±1.90 A 13.25±5.54 A* 14.08±3.40 a* 54.33±12.29 b 16.67±6.27 b* 1.58±.43

128

Table 5.2. Raabe’s percent similarity of carabids in new (1 yr) vs. old (7 yrs) canopy gaps by location (center, edge, or 50 m into surrounding forest) in a South Carolina bottomland hardwood forest, 2001.

Comparison Percent Similarity

New Edge vs. New Forest 76.21

Old Edge vs. Old Forest 75.30

New Edge vs. Old Edge 72.33

New Forest vs. Old Forest 67.66

New Center vs. New Edge 60.64

Old Center vs. Old Edge 58.47

Old Center vs. Old Forest 50.01

New Center vs. New Forest 49.11

New Center vs. Old Center 46.49

129

Fig. 5.1. Mean (±SE) richness of carabids collected in malaise and pitfall traps in a bottomland hardwood forest, South Carolina, USA in 2001. The traps were placed at the center, edge, and in the forest surrounding “young” (created in 2000) and “old” (created in 1994) canopy gaps.

Within graphs (for each gap age), bars with the same letter above them are not significantly different (Ryan-Einot-Gabriel-Welsch Multiple Range Test, p < 0.05). Asterisks denote significant differences (p < 0.05) between the same trap locations (e.g. center vs. center) in old and young gaps.

130

25 A* AB* 20 B a* a* a 15

Mean No. Species (n=12) 0 Center Edge Forest Center Edge Forest Young Gaps Old Gaps

131

Fig. 5.2. Mean (±SE) number of carabids collected in malaise and pitfall traps in a bottomland hardwood forest, South Carolina, USA in 2001. The traps were placed at the center, edge, and in the forest surrounding “young” (created in 2000) and “old” (created in 1994) canopy gaps.

Figure 2(ii) depicts total beetle abundances excluding Semiardistomis viridis. Within graphs (for

each gap age), bars with the same letter above them are not significantly different (Ryan-Einot-

Gabriel-Welsch Multiple Range Test, p < 0.05). Asterisks denote significant differences (p <

0.05) between the same trap locations (e.g. center vs. center) in old and young gaps.

132

140 (i) A Including 120 Semiardistomis viridis A a 100 A* 80 ab 60 b* 40 20 0 140 (ii) A* Excluding 120 Semiardistomis viridis 100 AB* 80 B 60 a* a Mean No. Individuals (n=12) Individuals Mean No. 40 a* 20 0 Center Edge Forest Center Edge Forest Young Gaps Old Gaps

133

Fig. 5.3. Mean (±SE) abundance (i) and richness (ii) of carabids collected in malaise and pitfall traps in 2001 in bottomland hardwood forest gaps of different size (0.13 ha, 0.26 ha, and 0.50 ha) created in 1994 and 2000 in South Carolina, USA. Within graphs (for each gap age), bars with the same letter above them are not significantly different (Ryan-Einot-Gabriel-Welsch Multiple

Range Test, p < 0.05). Asterisks denote significant differences (p < 0.05) between the same trap locations (e.g. center vs. center) in old and young gaps.

134

200 (i) A Abundance a 150 A A a 100 a 50

0 Richness 30 (ii) A A* A* Mean No. (n=12) No. Mean a 20 a* a*

0 0.13 ha 0.26 ha 0.50 ha 0.13 ha 0.26 ha 0.50 ha Young Gaps Old Gaps

135

Fig. 5.4. Mean (±SE) number of carabids collected in malaise and pitfall traps in a bottomland hardwood forest, South Carolina, USA in 2001. The traps were placed at the center, edge, and in the forest surrounding “young” (created in 2000) and “old” (created in 1994) canopy gaps. The species (Table 2) were categorized as preferring open-habitat (i), being generalists (ii), or preferring intact forests (iii) based on information in Larochelle and Lariviere (2003). Within graphs (for each gap age), bars with the same letter above them are not significantly different

(Ryan-Einot-Gabriel-Welsch Multiple Range Test, p < 0.05). Asterisks denote significant differences (p < 0.05) between the same trap locations (e.g. center vs. center) in old and young gaps.

136

25 (i) Open-Habitat Species 20 A* 15

10 B B a 5 a* a

0 Generalist Species (ii) A* 30 A

20 A a* 10 a a

0 Mean Number Collected (iii) Forest Species 25 A for α= 0.1 A b 20 ab 15 A a 10 5 0 Center Edge Forest Center Edge Forest

Young Gaps Old Gaps

137

CHAPTER 6

USING MALAISE TRAPS TO SAMPLE GROUND BEETLES (COLEOPTERA:

CARABIDAE)¹

¹Ulyshen, M. D., Hanula, J. L., Horn S., 2004. Accepted by Canadian Entomologist.

Reprinted here with permission of publisher.

138

Pitfall traps provide an easy and inexpensive way to sample ground-dwelling arthropods

(Spence and Niemela 1994, Spence et al. 1997, Abildsnes and Tommeras 2000) and have been used exclusively in many studies on the abundance and diversity of ground beetles (Carabidae).

Despite the popularity of this trapping technique, pitfall traps have many disadvantages. For example, they often fail to collect both small (Spence and Niemela 1994) and “trap shy” species

(Benest 1989), eventually deplete the local carabid population (Digweed et al. 1995), require a species to be ground-dwelling in order to be captured (Liebherr and Mahar 1979), and produce different results depending on trap diameter and material, type of preservative used, and trap placement (Work et al. 2002, Greenslade 1964, Luff 1975). Further complications arise from seasonal patterns of movement among the beetles themselves (Maelfait and Desender 1990) as well as numerous climatic factors, differences in plant cover, and variable surface conditions

(Adis 1979).

Because of these limitations, pitfall trap data give an incomplete picture of the carabid community, and should be interpreted carefully. Additional methods, such as Berlese funnels and litter washing (Spence and Niemela 1994), collecting from lights (Usis and MacLean 1998), and deployment of flight intercept devices (Paarmann and Stork 1987, Liebherr and Mahar 1979) should be incorporated in surveys in order to better ascertain the species composition and relative numbers of ground beetles. Flight intercept devices, like pitfalls, have the advantage of being easy to use and replicate, but their value to carabid surveys is largely unknown. Here we

demonstrate the effectiveness of Malaise traps for sampling ground beetles in a bottomland

hardwood forest.

This is part of a larger study investigating the response of insects to the creation of

canopy gaps in a bottomland hardwood forest in the southeastern United States. The gaps were

139 created within a 120 ha stand of 75 year-old bottomland hardwoods at the Savannah River Site

(near Aiken, South Carolina), an 80,269 ha nuclear production facility and Environmental

Research Park owned and operated by the United States Department of Energy (DOE). For a detailed description of the study site, including the dominant plant species present, consult

Ulyshen et al. (2004).

We established 72 trapping locations in and around canopy gaps of varying size (0.13,

0.26, and 0.50 ha) and age (1 or 7 yrs). The gaps were located throughout the forest, and were

separated by at least 200 m. We placed one Malaise trap and two pitfall traps (all three spaced

~5 m apart) at the center and edge of each gap as well as 50 m into the surrounding forest. We

sampled at the following intervals during 2001: 17-23 May, 10-16 July, 7-13 September, and 3-9

November.

The Malaise traps used in this study (canopy trap, Sante Traps, Lexington, KY) have

collecting jars at the bottom of each trap in addition to one at the top. They were suspended from

3 m tall hangers constructed from metal tubing. The pitfall traps consisted of 480 ml plastic cups

with 8.4 cm diameter funnels. The funnels directed beetles into 120 ml specimen cups

containing preservative. Each trap was positioned at the intersection of four 0.5 m long metal

drift fences to increase trap catch. The two pitfall traps were combined at each location prior to

analysis. The preservative used in both Malaise and pitfall traps was a 2% formaldehyde and

saturated NaCl solution with detergent added to reduce surface tension (New and Hanula 1998).

Samples were stored in 70% ethanol, sorted to morphospecies, and identified using a key to the carabids of South Carolina (Ciegler 2000). This reference was also used to assign our species to size classes (< 5 mm, 5-10 mm, 10-15 mm, and > 15 mm).

140

We collected a total of 5,498 individuals representing 87 carabid species (including

Amerinus linearis LeConte, a new state record) (Table 6.1). Although the average pair of pitfall traps collected more species and individuals than did the average Malaise trap (Figure 6.1),

Malaise traps collected more species overall (Table 6.1). Furthermore, 33 of the species

captured in Malaise traps were not collected in pitfalls (Table 6.1). Pitfall traps also collected

many unique species (29). Of these, 10 were brachypterous, and incapable of flight (Table 6.1).

While smaller carabid species were better represented in Malaise than in pitfall trap

samples, pitfalls collected a greater proportion of the larger species (Figure 6.2). Relatively few

carabids above 10 mm in length were collected in Malaise traps, but large numbers of these were

collected in pitfalls (Figure 6.2). Similarly, while pitfalls captured few species under 5 mm,

many were captured in Malaise traps (Figure 6.2).

Many of the carabids (11) captured exclusively in Malaise traps live primarily on

vegetation. For example, we collected 12 species of Lebiini (the “colorful foliage ground

beetles”), a group of primarily plant-dwelling species. Nine of these were captured only in

Malaise traps (Table 6.1).

Malaise traps greatly increased the number and diversity of carabids sampled in this

study. If only pitfall traps had been used, the number of individuals and species collected would

have been reduced by 26% and 38%, respectively. These results emphasize the importance of

using more than one trapping method when conducting ground beetle surveys. Despite their

success in this study, the efficacy of Malaise traps in different habitats remains uncertain.

Past research recognizes the importance of flight to the dispersal of carabids and the

prevalence of macropterous species in unstable habitats (Darlington 1943, Boer 1970, Cardenas

and Bach 1992). Cardenas and Bach (1992) found a frequently flooded site to contain

141 predominantly macropterous carabid species, while a nearby stable environment had many apterous and brachypterous forms. Because our forest was flooded seasonally, and many of the low-lying areas were under water throughout the study, flight may be a more important mode of

dispersal here than in other, more stable habitats. Further studies are needed to elucidate the

value of Malaise traps to carabid surveys in different habitats and regions before any general

recommendations on their use can be made.

Trap design is another important consideration. The collecting jar at the base of our traps

was of particular value since beetles often fall upon encountering a barrier during flight. We

recently set out the same Malaise trap design in the Oconee National Forest (Greene Co.,

Georgia, USA) to compare the numbers of beetles captured in the upper and lower collecting jars. We ran 12 traps for a month, and collected 275 carabids. Of these, 223 (81.1%) were collected in the lower chamber (unpublished data).

We have demonstrated the value of one Malaise trap design to carabid surveys in a bottomland hardwood forest. The expense of these traps, and the inability of alternative designs

to capture specimens that fall upon contact may limit the use of Malaise traps by many

researchers. Other less expensive flight intercept devices (such as window-pane traps) are

specifically designed to capture fallen insects, and may prove similarly useful to future carabid

surveys.

142

Acknowledgements

We thank H. Lee, Jr., L. Reynolds, and C. Smith for taxonomic help, D. Dyer, and J. Campbell for assisting with field work, and C. Asaro, W. Berisford, and J. McHugh for editing early drafts of the manuscript. This research was funded by the National Research Initiative Competitive

Grants Program of the USDA Cooperative State Research Education and Extension Service

(CSREES Grant No. 00-35101-9307). Use of product names does not constitute endorsement by

the USDA Forest Service.

143

Literature Cited

Abildsnes J, Tommeras BA. 2000. Impacts of experimental habitat fragmentation on

ground beetles (Coleoptera, Carabidae) in a boreal spruce forest. Annales Zoologici

Fennici 37: 201-212

Adis J. 1979. Problems of interpreting arthropod sampling with pitfall traps.

Zoologischer Anzeiger 202: 177-184

Benest G. 1989. The sampling of a carabid community: I. The behavior of a carabid

when facing the trap. Revue D’Ecologie et de Biologie du Sol 26: 205-211

Boer, PJ den. 1970. On the significance of dispersal power for populations of carabid

beetles (Coleoptera, Carabidae). Oecologia 4: 1-28

Cardenas AM, Bach C. 1992. The influence of environmental changes on wing

development in carabids (Col. Carabidae) in the Guadiato River basin (SW Spain). Vie

Milieu 42: 277-282

Ciegler JC, Wheeler AG (Editor). 2000. Ground beetles and wrinkled bark beetles of

South Carolina (Coleoptera: Geadephaga: Carabidae and Rhysodidae). Clemson

University.

Darlington PJ. 1943. Carabidae of mountains and islands: data on the evolution of

isolated faunas and on atrophy of wings. Ecological Monographs 13: 37-64

Digweed SC, Currie CR, Carcamo HA, Spence JR. 1995. Digging out the

“digging-in effect” of pitfall traps: Influences of depletion and disturbance on catches of

ground beetles (Coleoptera: Carabidae). Pedobiologia 39: 561-576

Greenslade PJM. 1964. Pitfall trapping as a method for studying populations of

144

Carabidae (Coleoptera). J. Anim. Ecol. 33: 301-310

Larochella A, Lariviere M-C. 2001. Natural history of the tiger beetles of north America

north of Mexico. Cicindela 33: 41-162

Larochelle A, Lariviere M-C. 2003. A natural history of the ground beetles

(Coleoptera:Carabidae) of America north of Mexico. Pensoft Series Faunistica No 7.

Pensoft Publishers. 583 pp.

Liebherr J, Mahar J. 1979. The carabid fauna of the upland oak forest in Michigan:

Survey and analysis. The Coleopterists Bulletin 33: 183-197

Luff ML. 1975. Some features influencing the efficiency of pitfall traps. Oecologia 19:

345-357

Maelfait JP, Desender K. 1990. Possibilities of short-term carabid sampling for site

assessment studies. In: The role of ground beetles in ecological and environmental

studies. Stork NE (Editor) 217-225

New KC, Hanula JL. 1998. Effect of time elapsed after prescribed burning in

longleaf pine stands on potential prey of the Red-Cockaded Woodpecker. Southern

Journal of Applied Forestry 22: 175-183

Paarmann W, Stork NE. 1987. Seasonality of ground beetles (Coleoptera:Carabidae) in

the rain forests on N. Sulawesi (Indonesia). Insect Science and Its Application 8: 483-

487

Spence JR, Niemela JK. 1994. Sampling carabid assemblages with pitfall traps:

the madness and the method. The Canadian Entomologist 126: 881-894

Spence JR, Langor DW, Hammond, HEJ, Pohl GR. 1997. Beetle abundance and diversity

145

in a boreal mixed-wood forest. In: Forests and Insects. Watt AD, Stork NE, Hunter MD

(Editors). 287-301

Ulyshen MD, Hanula JL, Horn S, Kilgo JC, Moorman CE. 2004. Spatial and temporal

patterns of beetles associated with coarse woody debris in managed bottomland

hardwood forests. Forest Ecology and Management 199: 259-272

Usis JD, MacLean DB. 1998. The ground beetles (Coleoptera: Carabidae) of Stillfork

Swamp Nature Preserve, Carroll County, Ohio. Ohio Journal of Science 98: 66-68

Work TT, Buddle CM, Korinus LM, Spence JR. 2002. Pitfall trap size and

capture of three taxa of litter dwelling arthropods: implications for biodiversity studies.

Environmental Entomology 31: 438-448

146

Table 6.1. List of ground beetles (Carabidae) collected by Malaise (Mal) and pitfall (Pit) traps in a bottomland hardwood forest (South Carolina, USA). Information on the habits and wing morphology of each species was taken from Larochell and Lariviere (2001, 2003).

Tribe Species Number (Mal/Pit) Habitat Wing Structure

Bembidiini Bembidion affine Say 7/1 ground macropterous

Elaphropus granarius (Dejean) 3/12 ground dimorphic

Micratopus aenescens (LeConte) 424/0 ground macropterous

Mioptachys flavicauda (Say) 5/4 ground,under bark macropterous

Paratachys spp. Casey 100/25 ground macropterous

Polyderis laevis (Say) 24/0 ground macropterous

Tachyta nana inornata (Say) 6/0 ground,under bark macropterous

Brachinini Brachinus alternans Dejean 1/727 ground macropterous

Carabini Carabus sylvosus Say 0/47 ground brachypterous

Chlaenini Chlaenius aestivus Say 0/374 ground dimorphic

Chlaenius erythropus Germar 0/73 ground macropterous

Chlaenius laticollis Say 6/0 ground macropterous

Chlaenius pusillus Say 0/4 ground macropterous

Chlaenius sp. 5 0/151 ground macropterous

Cicindelini Cicindela punctulata Olivier 0/5 ground macropterous

Cicindela sexguttata Fabricius 2/0 ground macropterous

Megacephala sp. Latreille 0/1 ground macropterous

Clivinini Clivina bipustulata (Fabricius) 29/171 ground macropterous

Clivina dentipes Dejean 15/0 ground macropterous

Clivina rubicunda LeConte 32/0 ground dimorphic

Dyschirius sp. Bonelli 0/4 ground macropterous

Semiardistomis viridis (Say) 6/1258 ground macropterous

147

Ctenodactylini Leptotrachelus dorsalis (Fabricius) 1/0 ground/vegetation macropterous

Cychrini Scaphinotus sp. Dejean 0/7 ground brachypterous

Sphaeroderus sp. Dejean 0/3 ground brachypterous

Cyclosomini Tetragonoderus intersectus (Germar) 3/1 ground macropterous

Galeritini Galerita spp. Fabricius 0/47 ground macropterous

Harpalini Acupalpus testaceus Dejean 77/1 ground macropterous

Acupalpus sp. 2 172/7 ground macropterous

Acupalpus sp. 3 0/8 ground macropterous

Amblygnathus iripennis (Say) 1/0 ground macropterous

Amerinus linearis LeConte 6/0 ground dimorphic

Anisodactylus furvus LeConte 0/3 ground macropterous

Anisodactylus rusticus (Say) 1/0 ground macropterous

Harpalus pennsylvanicus (De Geer) 5/28 ground macropterous

Notiobia terminata (Say) 28/0 ground macropterous

Selenophorus ellipticus Dejean 2/2 ground macropterous

Selenophorus opalinus (LeConte) 8/4 ground macropterous

Selenophorus palliatus (Fabricius) 3/0 ground macropterous

Stenolophus ochropezus (Say) 40/16 ground macropterous

Stenolophus spretus Dejean 9/1 ground macropterous

Helluonini Helluomorphoides Ball 2/5 ground macropterous

Lachnophorini Euphoroticus pubescens(DeJean) 0/2 ground macropterous

Lebiini Apenes sinuatus (Say) 2/3 ground macropterous

Calleida decora (Fabricius) 2/0 vegetation macropterous

Calleida virdipennis (Say) 3/0 vegetation macropterous

Coptodera aerata Dejean 4/0 vegetation macropterous

Cymindis sp. 60/0 ground? macropterous

Dromius piceus Dejean 1/0 vegetation macropterous

Lebia lobulata LeConte 16/0 vegetation macropterous

148

Lebia marginicollis Dejean 5/0 vegetation macropterous

Lebia tricolor Say 7/0 vegetation macropterous

Lebia viridis Say 22/0 vegetation macropterous

Lebia vittata (Fabricius) 1/0 vegetation macropterous

Philorhizus atriceps (LeConte) 0/2 ground brachypterous

Licinini Badister maculatus LeConte 8/0 ground macropterous

Badister ocularis Casey 12/0 ground macropterous

Dicaelus dilatatus Say 0/44 ground brachypterous

Dicaelus elongatus Bonelli 0/46 ground brachypterous

Diplocheila assimilis (LeConte) 0/71 ground macropterous

Loxandrini Loxandrus rectus (Say) 5/3 ground macropterous

Loxandrus sp. 1 10/161 ground macropterous

Loxandrus sp. 2 3/0 ground macropterous

Morionini Morion monilicornis (Latreille) 2/0 under bark unknown

Notiophilini Notiophilus Dumeril 0/3 ground dimorphic

Oodini Anatrichus minuta (Dejean) 2/1 ground macropterous

Oodes amaroides Dejean 42/30 ground/vegetation macropterous

Oodes sp. 2 0/37 ground macropterous

Panagaeini Panagaeus fasciatus Say 0/1 ground macropterous

Pentagonicini Pentagonica flavipes (LeConte) 6/0 vegetation macropterous

Platynini Agonum aeruginosum Dejean 27/0 ground macropterous

Agonum decorum Say 0/83 ground/vegetation macropterous

Calathus opaculatus LeConte 11/4 ground macropterous

Olisthopus sp. 1 2/0 ground macropterous

Olisthopus sp. 2 137/12 ground macropterous

Platynus decentis (Say) 6/0 ground/vegetation submacropterous

Pterostichini Cyclotrachelus brevoorti (LeConte) 0/44 ground brachypterous

Cyclotrachelus spoliatus (Newman) 0/3 ground brachypterous

149

Cyclotrachelus sp. 3 0/23 ground brachypterous

Lophoglossus gravis LeConte 0/310 ground macropterous

Piesmus submarginatus (Say) 9/93 ground macropterous

Poecilus chalcites (Say) 0/31 ground macropterous

Pterostichus sp. 1 Bonelli 0/2 ground brachypterous

Scaritini Scarites sp. Fabricius 5/67 ground macropterous

Zabrini Amara sp. Bonelli 0/2 ground macropterous

Zuphiini Thalpius pygmaeus (Dejean) 1/0 ground macropterous

Unknown unidentified sp. 1/0

Total No. Individuals 1430/4068

Total No. Species 58/54

No. Species Unique to Trap 33/29

150

Figure 6.1. Mean number of individuals and species of carabids collected in Malaise and pitfall traps in a bottomland hardwood forest (South Carolina, USA), in 2001.

151

60 Individuals Species 50

Mean Number Mean 20

0 Malaise Pitfall

152

Figure 6.2. Mean number of individuals (A) and species (B) of carabids by size class collected in

Malaise and pitfall traps in a bottomland hardwood forest (South Carolina, USA), in 2001.

153

30 A Malaise Traps Pitfall Traps 25

5 mean number of individuals (n=72) of individuals number mean

0 B

1 mean number of species (n=72) of species number mean

0 < 5 mm 5 - 10 mm 10 - 15 mm > 15 mm Length

154

CHAPTER 7

RESPONSE OF CARRION BEETLES (COLEOPTERA: SILPHIDAE) TO FLOODING AND

GAP CREATION IN A SOUTHEASTERN BOTTOMLAND HARDWOOD FOREST¹

¹Ulyshen, M. D., Hanula, J. L. To be submitted to the Journal of Entomological Science

155

Abstract

Carrion beetles (Silphidae) appear to respond negatively to flooding, but are able to persist, at least in low numbers, in flooded bottomland hardwood forests. The four species collected, in decreasing abundance, were: Nicrophorus orbicollis Say, Necrophila americana

(L.), Oiceoptoma inaequale (Fabricius), and Nicrophorus carolinus (L.). The amount of

standing water varied with location. Young gaps (~4 yrs) had the most water, followed by old gaps (~10 yrs), bottomland forest, and dry forest. Silphid abundance increased predictably with decreasing water levels during periods of flooding. While flooded, there were more silphids

collected in the forest surrounding the gaps than in the gaps themselves. During a dry period,

however, there were no differences in total silphid abundance between the gaps and the

surrounding forest. There were more silphids collected in adjacent dry forest locations than in

the other locations regardless of water levels.

156

Introduction

Silphidae is a small but widespread group of beetles. Of the 175 known species (Newton

1995), about 30 are found in North America and 13 have been found in the Southeast (Peck and

Kaulbars 1987). With the exception of several species that prey on caterpillars or snails, most

feed on dead organic matter and will readily come to traps baited with carrion (Peck 2001).

Because they are easily surveyed and are of ecological and behavioral interest, silphids

have been the focus of innumerable studies. They appear to be a fairly sensitive group to habitat fragmentation (Gibbs and Stanton 2001) and often have fairly particular habitat requirements.

The American burying beetle (Nicrophorus americanus Olivier), for example, has gone nearly extinct due to the destruction of climax forests throughout its former range (Anderson and Peck

1985). Other species have been less negatively affected by human activities, but still exhibit

strong preferences for certain habitats (Anderson 1982, Shubeck 1983, Shubeck 1993).

Despite numerous studies concerning silphid seasonality (Shubeck and Blank 1982,

Shubeck and Schleppnik 1984, Shubeck et al. 1981, Ulyshen and Hanula 2004), there has been

little work done on the southeastern coastal plain. We surveyed carrion beetles in a coastal plain

bottomland hardwood forest and determined the flood tolerance of different silphid species, as

well as their response to canopy gaps.

Materials and Methods

This study was conducted on the Savannah River Site (SRS), an 80,269 ha nuclear

production facility near Aiken, South Carolina. The SRS is owned and operated by the United

157

States Department of Energy (DOE), and is managed as an Environmental Research Park. The stands used were 75-100 year-old bottomland hardwoods approximately 120 ha in size. This forest remains wet throughout much of the year, with some areas being almost permanently flooded. The bottomland hardwood sites are described in detail in Ulyshen et al. (in press).

Of the 12 canopy gaps used in this study, 6 were created in December of 1994 (“old gaps”) and 6 in August of 2000 (“young gaps”). Because they had less vegetation, and were still rutted from timber removal, young gaps were much wetter than older gaps or the surrounding forest. Therefore, traps in young gaps were often suspended above standing water. Older gaps had less standing water than young gaps but noticeably more than the forest. All bottomland hardwood sites had substantial amounts of standing water close to the traps. In total, we had 22 traps: one in each of the 6 young and 6 old gaps, one in the forest surrounding each old gap, and four in an adjacent “dry” forest (2-3 km from bottomland study sites), which did not experience flooding.

We used one dead adult mouse wrapped in cheesecloth as bait during each trapping session, which we attached to the clear plastic panels of each trap (Figure 7.1). The panels were wired to a bucket and to a funnel situated at the bottom and top of the trap, respectively. The diameter of both the bucket and funnel was 15 cm. The insects were collected in the bucket below or in a plastic bottle attached to the funnel at the top of the trap (Figure 7.1). The length of the panels between the bucket and funnel was 15 cm. The traps were suspended about 1 m above the ground from metal poles. We used a NaCl-2% formaldehyde solution to kill and preserve our catch (New and Hanula 1998). The traps were operated for a week at a time, and we sampled 4 times between May and August of 2003 and three times between February and

158

April of 2004. However we did not begin sampling in the dry forest until the second month of the study (June 2003).

Results

Four species of silphids were collected, but only 3 were collected in the bottomland forest

(Table 1). Overall, Nicrophorus orbicollis was by far the most abundant, followed by

Necrophila americana, and Oiceoptoma inaequale (Table 7.1). Nicrophorus carolinus was the

least abundant species, and was only collected in the dry forest. We collected the largest

diversity and number of silphids in the dry forest, followed by the bottomland forest, young gaps,

and old gaps (Table 7.1). Although we did not collect any in our traps, we observed large

numbers of Necrodes surinamensis (Fabricius) on a dead hog near the gaps, so there were at least

5 silphid species present in and around our study site. Another species known to occur in the

area is Oiceoptoma rugulosum Portevin. The fact that we collected none in our traps suggests

that this species may be fairly intolerant of flooding.

The study site was noticeably drier in the spring of 2004 than it was the previous year

(Figure 7.2), and by April even the young gaps were dry. We collected an increasingly large

number of silphids throughout the summer of 2003, but we collected more in April, 2004 than in

all other months combined (Table 7.2). This was true for all locations, including the dry forest

sites (Table 7.3). There were significantly more silphids in the forest locations than in the gaps

between May 2003 and March 2004 (Figure 7.3). In April 2004, however, there was no

difference in total silphid abundance between the bottomland forest and the gaps but the dry

forest had significantly more silphids than did any other location (Figure 7.3). Necrophila

159 americana was more abundant in young gaps than in any other location but only during the dry trapping period while Nicrophorus orbicollis appeared to prefer forested locations (Figure 7.4).

Nicrophorus carolinus also appears to prefer forested sites and was only collected in the dry forest (Table 7.1).

Discussion

Silphids appeared to respond negatively to flooding, but were able to persist, at least in low numbers, in flooded bottomland hardwood forests. In six weeks of sampling during the flooded period between May 2003 and March 2004, we collected only 6 specimens of two species in young gaps. In contrast, in just one week of sampling in the same gaps when they were dry we collected 234 specimens of 3 species. A similar trend was observed for old gaps, 12 specimens of 2 species were captured in six wet weeks compared to 189 specimens of 3 species in one dry week. We also collected substantially more silphids in the forest locations in April than in the previous months.

During the wet period, we collected many more silphids in the forest than in the gaps and twice as many in old gaps than in young gaps. This is probably because the forest was drier than the gaps and old gaps were noticeably drier than young gaps. Another contributing factor may be the fact that N. orbicollis, the most abundant species, is known to prefer forested habitat

(Shubeck 1983). Despite a preference for the forest, N. orbicollis was collected in higher numbers in both young (5) and old (11) gaps during the flooded period than were any other species. Even though we did not sample in the dry forest in May 2003, we collected more silphids at this location in 5 wet months than we did in the bottomland forest location in 6 wet

160 months (269 vs. 114). This is consistent with the idea that drier habitats are more favorable to carrion beetles.

When we sampled in April 2004, the gaps were no longer submerged, and the exposed ground was fairly dry. Therefore, moisture was no longer a complicating factor, and we can use this data to compare silphid abundances among locations. We again collected the most silphids in the dry forest, possibly because flooding reduced the populations present in the bottomland forest. We found there to be no differences in the total abundance of silphids between young gaps, old gaps, or the surrounding forest. Necrophila americana was much more abundant in young gaps (Figure 4) than it was in old gaps, the surrounding forest, or the dry forest. This is consistent with past studies on the habitat preferences of silphids in which N. americana has been shown to prefer clearings (Shubeck 1983). Nicrophorus orbicollis showed an opposite trend (Figure 4), being more abundant in the dry forest than in the bottomland forest, old gap or young gap. One species, Nicrophorus carolinus, was only collected in the dry forest suggesting that it may be particularly intolerant of flooding.

The fact that we collected substantially greater numbers of silphids in the dry forest in

April than we did in previous months suggests that soil moisture may affect silphid abundance in forests even if they are not prone to flooding. The proportionate increase in silphid abundance from wet conditions to dry conditions was greatest for the wettest habitats. For example, in 4 yr. gaps, we collected 39 times more silphids in April than in all previous months combined (Table

3). In contrast, the number of silphids collected in 10 yr. gaps increased 15.75 times, while that in the bottomland forest and dry forest sites increased only 1.83 and 1.15 times, respectively.

Given the habits of silphids, it is not surprising to find them to be sensitive to flooding.

Because carrion is usually on the soil surface and silphids pupate in the soil (Anderson and Peck

161

1985), immature carrion beetles must be particularly sensitive to soil moisture. The habits of the

adults are less well understood, but it seems likely that flooding would threaten their survival as

well. Flooding may periodically reduce their populations, but the remarkable ability of silphids to detect and locate carrion from long distances may allow individuals to quickly re-colonize the area from drier sites nearby.

Acknowledgements

We thank D. Dyer for assisting with field work. Support was provided by the Department of

Energy-Savannah River Operations Office through the U.S. Forest Service Savannah River under Interagency Agreement DE-AI09-00SR22188. Funding was provided by the National

Research Initiative Competitive Grants Program of the USDA Cooperative State Research

Education and Extension Service (CSREES Grant No. 00-35101-9307).

162

References Cited

ANDERSON, R. S. 1982. Resource partitioning in the carrion beetle (Coleoptera: silphidae)

fauna of southern Ontario: ecological and evolutionary considerations. Can. J. Zool.

60:1314-1325.

ANDERSON, R. S. and S. B. PECK. 1985. The Insects and Arachnids of Canada, Part 13.

The Carrion Beetles of Canada and Alaska (Coleoptera: Silphidae and Agyrtidae).

Research Branch, Agriculture Canada, Ottawa, Publication 1778. 121 pp.

GIBBS, J. P. and E. J. STANTON. 2001. Habitat Fragmentation and Arthropod community

change: Carrion Beetles, Phoretic Mites and Flies. Ecol. Appl. 11:79-85.

NEW, K. C. and J. L. HANULA. 1998. Effect of time elapsed after prescribed burning in

longleaf pine stands on potential prey of the Red-Cockaded Woodpecker. SJAF 22(3):

175-183.

NEWTON, A. F. 1995. unpublished world catalogue of silphidae.

PECK, S. B. 2001. Silphidae. In: Arnett Jr., R. H. and Thomas, M. C. (eds). American Beetles.

Volume 1. CRC Press. Pp. 268-271.

PECK, S. B. and M. M. KAULBARS. 1987. A synopsis of the distribution and bionomics of

the carrion beetles (Coleoptera: silphidae) of the conterminous United States. P.

Entomol. Soc. Ont. 18:47-81.

SHUBECK, P. P. 1983. Habitat preferences of carrion beetles in The Great Swamp National

Wildlife Refuge, New Jersey (Coleoptera: Silphidae, Dermestidae, Nitidulidae,

Histeridae, Scarabaeidae). J. New York Entomol. S. 91: 333-341.

SHUBECK, P. P. 1993. An Ecotonal Study of Carrion Beetles (Coleoptera:Silphidae) in the

163

Great Swamp National Wildlife Refuge, New Jersey. Entomol. News 104: 88-92.

SHUBECK, P.P., N. M. DOWNIE, R. L. WENZEL, and S. B. PECK. 1981. Species

composition and seasonal abundance of carrion beetles in an oak-beech forest in the great

swamp national wildlife refuge (N.J.). Entomol. News. 92:7-16.

SHUBECK, P. P. and A. A. SCHLEPPNIK. 1984. Silphids attracted to carrion near St. Louis,

Missouri (Coleoptera: Silphidae). J. Kansas Entomol. Soc. 57:360-362.

SHUBECK, P.P. and D. L. BLANK. 1982. Silphids attracted to mammal carrion at

Cheltenham, Maryland (Coleoptera, Silphidae) P. Entomol. Soc. Wash. 84: 409-410.

ULYSHEN, M. D. and J. L. HANULA. 2004 (in press). The diversity and seasonal activity of

carrion beetles (coleoptera: Silphidae) in Northeastern Georgia. J. Entomol. Sci..

ULYSHEN, M. D., J. L. HANULA, S. HORN, J. C. KILGO and C. E. MOORMAN. 2004 (in

press). Spatial and temporal patterns of beetles associated with coarse woody debris in

managed bottomland hardwood forests. Forest Ecol. and Manag.

164

TABLE 7.1. List and total number of Silphidae collected in 3 locations in a southeastern bottomland hardwood forest (4 yr-old gaps, 10 yr-old gaps, and closed forest) and in a drier forest nearby at the Savannah River Site, South Carolina in 2003.

4 yr. Gap 10 yr. Gap Forest Dry Forest Total

Necrophila americana (Linnaeus) 117 23 25 11 176

Nicrophorus carolinus (Linnaeus) 0 0 0 9 9

Nicrophorus orbicollis Say 113 167 278 537 1095

Oiceoptoma inaequale (Fabricius) 10 11 20 21 62

Total 240 201 323 578 1342

165

TABLE 7.2. Total number of Silphidae collected by month at the Savannah River Site, South

Carolina, 2003.

May 03 Jun. 03 Jul. 03 Aug. 03 Feb. 04 Mar. 04 Apr. 04

Necrophila americana 2 0 0 2 0 0 172

Nicrophorus carolinus 0 0 0 2 0 0 7

Nicrophorus orbicollis 7 68 107 144 0 55 714

Oiceoptoma inaequale 0 0 0 0 0 14 48

Total 9 68 107 148 0 69 941

166

TABLE 7.3. Total number of Silphidae collected by location and month at the Savannah River

Site, South Carolina, 2003.

May 03 Jun. 03 Jul. 03 Aug. 03 Feb. 04 Mar. 04 Apr. 04

4 yr. Gap 4 0 0 1 0 1 234

10 yr. Gap 0 8 2 0 0 2 189

Forest 5 40 22 15 0 32 209

Dry Forest NA 20 83 132 0 34 309

167

FIG. 7.1. The suspended carrion trap used to collect carrion beetles (Coleoptera: Silphidae) at the Savannah River Site, South Carolina, in 2003.

168

169

FIG. 7.2. Rainfall data for the Savannah River Site, South Carolina, in 2003 and 2004.

170

2003 25 2004

15 rainfall (cm) 10

0 July May April June March August October January February November December September

171

FIG. 7.3. Mean number by location (n=6 for 4 yr gaps, 10 yr gaps, and bottomland forest; n=4 for dry forest) of carrion beetles (Coleoptera: Silphidae) collected in suspended traps baited with mice in a southeastern bottomland hardwood forest, South Carolina, 2003. Different letters indicate significance based on the Ryan-Einot-Gabriel-Welsch Multiple Range Test, p<0.05.

Asterisks denote that the square root transformation was used.

172

100 A* May 2003 - March 2004 (Wet) 80

B* 20

C* C* 0 A April 2004 (Dry) 80 Mean number of Silphidae collected of Silphidae number Mean 60 B B B 40

0 4 yr Gap 10 yr Gap Forest Dry Forest

decreasing moisture by location

173

FIG. 7.4. Mean number by location (n=6 for 4 yr gaps, 10 yr gaps, and bottomland forest; n=4 for dry forest) of Nicrophorus orbicollis and Necrophila americana collected in suspended traps baited with mice in a southeastern bottomland hardwood forest, South Carolina, 2003. For each species, there is a graph of the number collected during a “wet” (i.e. flooded) period (May 2003-

March 2004) and a graph the number collected during a “dry” period (April 2004). Different letters indicate significance based on the Ryan-Einot-Gabriel-Welsch Multiple Range Test, p<0.05 (the square root transformation was used).

174

100 Nicrophorus orbicollis A (Wet) (Dry) 80 A

B 40 B B B 20 C C 0 Necrophila americana 25 (Wet) (Dry) A mean number mean 20

10 B B 5 B A A A A 0 t t t t Gap Gap res res Gap Gap res res yr yr Fo Fo yr yr Fo Fo 4 10 Dry 4 10 Dry

175

CHAPTER 8

CONCLUSIONS

While gaps seem to represent important concentrations of resources such as coarse woody debris and young herbaceous vegetation, a number of other factors surely affected the distribution of insects in and around gaps. For example, insects may have been drawn to the increased temperatures and light levels in gaps as compared to the cool and dark forests surrounding them. How these and other abiotic factors affect insect behavior and the degree to which they may have contributed to our results cannot be determined. Regardless of the exact forces involved, this project clearly demonstrates the fact that insects generally respond positively to canopy gap creation, and are generally more abundant and species rich there than in the surrounding forest.

Because only a few insect groups were negatively affected by gap creation, and most were more numerous in the gaps than in the forest, it can be safely said that group selection harvesting had relatively little negative impact on the arthropod community. Other studies involving insect predators (birds, bats, and frogs) support the conclusion that insects are generally more abundant in forest openings and these, in conjunction with our results, underscore the importance of insects to forest ecosystems. Furthermore, gaps provide habitat for a number of plant and animals species rarely found elsewhere. For example, a number of carabid species were found to be significantly more abundant at the centers of young gaps or old gaps (i.e.

176

Semiardistomis viridis), and one species, Notiobia terminata was found exclusively in young gaps. It should come as no surprise that a number of species have become adapted to areas of

disturbance since disturbance is the naturalrocess of renewal in all forests.

Because group selection harvesting mimics natural rates of disturbance in bottomland

hardwood forests while preserving the integrity of the remaining stand, it may be an

important tool in the sustainable management of our dwindling forest resources.