Impact of Cogongrass (Imperata cylindrica) Presence and Management Strategies on 
Arthropod Natural Enemy Populations in Longleaf Pine Stands 
 
by 
 
Sallie Martin 
 
 
 
 
A thesis submitted to the Graduate Faculty of 
Auburn University 
in partial fulfillment of the 
requirements for the Degree of 
Master of Science 
 
Auburn, Alabama 
May 7, 2012 
 
 
 
 
Keywords: cogongrass, arthropod diversity, natural enemies,  
invasion, management strategies 
 
 
Copyright 2012 by Sallie Martin 
 
 
Approved by 
 
Dr. David Held, Chair, Associate Professor of Entomology 
Dr. Stephen Enloe, Associate Professor of Agronomy 
Dr. Lori Eckhardt, Associate Research Professor of Forestry 
Dr. Robert Boyd, Professor of Biological Sciences 
 
 
 
 
 ii 
 
 
 
 
Abstract 
 
 
Cogongrass (Imperata cylindrica Beav.) is an aggressive, invasive weed that is 
threatening the integrity of longleaf pine (Pinus palustris Mill.) ecosystems in southeast 
Alabama. Various management strategies are implemented to suppress cogongrass. In 
Objective 1, populations of natural enemies and bark beetles (Hylastes spp.) were 
monitored over 2 years in response to four different management strategies; herbicide, 
seeding with native species, seeding plus herbicide, and control (no treatment) in burned 
and unburned longleaf pine plots.  In Objective 2, the impacts of cogongrass on the 
abundance of arthropods and soil dwelling microfauna in longleaf pine were investigated.  
Insects were collected via pitfall traps and sweep samples, then sorted and counted in the 
lab.  For Objective 1, sentinel species (mainly generalist predators) were monitored.  For 
Objective 2, all arthropods were sorted to family level and counted.  In spring and fall 
2011, earthworms were sampled using a disclosing solution in longleaf pine with or 
without cogongrass.  Soil cores were taken in May 2011 to determine the relative 
abundance and diversity of nematodes.  Cogongrass management strategies had minimal 
impacts on natural enemies and no impact on Hylastes populations.  The seeding plus 
herbicide treatment had significantly fewer spiders than other treatments but this was the 
only affected group of natural enemies. Thirty of the >100 families collected were 
significantly impacted by cogongrass but not all impacts were negative. For example, 
Curculionidae, Tettigoniidae, Acrididae, and Cercopidae were more abundant in longleaf 
pine infested with cogongrass.  Similarly, earthworm abundance and weight were 
 
 iii 
significantly greater in longleaf pine with cogongrass present. Plant parasitic nematodes 
were significantly greater in longleaf pine with cogongrass present, the opposite being 
true for free-living nematodes. These results suggest that cogongrass management does 
not significantly affect natural enemy and bark beetle populations, but its presence does.  
The families that were impacted, particularly Scarabaeidae (and their relatives) or 
Gryllidae, can be used in future evaluations of ecosystem impacts of cogongrass invasion.  
Also, families of herbivores that increased in the longleaf pine stands where cogongrass 
was present may provide insight for future studies seeking herbivores that may be able to 
utilize this invasive grass.   
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 iv 
 
 
 
 
 
Dedication 
 
 
 I would like to dedicate my thesis and its work to my ever encouraging parents, 
Herb and Marynm, and my brothers, Sam and Wade. Also, to my fianc?e Josh, who stood 
by my side through every bit of this process and never let me quit. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 v 
 
 
 
 
 
Acknowledgments 
 
 
There is no way I can begin to thank everyone who played a part in the 
completion of this project. First, if it weren?t for Dr. Held convincing me not to apply for 
an internship at Disney World, but to stay here instead, I would not have gone on this 
incredible journey. He pushed me to work harder than I knew I could and accomplish 
more than I thought I ever would. He let me learn from my own mistakes and guided me 
through endless challenges. I owe a large part of my success to him. 
 Without help in the field and in the lab, I would still be drowning in antifreeze 
soaked arthropods. Alyse Hobgood, David Cummings, Amanda Loy, Shane Parker, 
Kristyn Carroll, Josh Sells, Matt Oliver, Chase Brown, Anthony Bowden, Scott Clem and 
my fellow lab mates-I could have never done this without you! Your help on days of 
endless heat, long drives, smelly bugs, filtering, and sorting is appreciated more than 
you?ll ever know. 
 A special thanks to Drs. Enloe and Loewenstein who spent countless hours 
communicating with landowners, scoping out sites and setting up plots for this research. 
Thanks to the owners of Beebe Farms, Mr. Rainwater, the Delaney brothers, and Keville 
Larson for the use of their land. They were endlessly flexible with our use of their land 
for my research. Also to my committee for their continual guidance throughout this 
journey. 
 
 
 vi 
 
 
 
 
Table of Contents 
 
 
Abstract ......................................................................................................................................... ii 
Dedication .................................................................................................................................... iv 
Acknowledgments ......................................................................................................................... v 
List of Tables ............................................................................................................................. viii 
List of Figures .............................................................................................................................. ix 
Chapter 1 ...................................................................................................................................... 1 
 Longleaf Pine Ecosystems ................................................................................................ 1 
 Insects in the Pine Ecosystems ........................................................................................ 1 
 Earthworms and Nematodes ............................................................................................. 5 
 Cogongrass Morphology, Biology, and Distribution ........................................................ 6 
 Ecosystem Impacts of Cogongrass ................................................................................... 7 
 Cogongrass Management .................................................................................................. 9 
 Interactions ...................................................................................................................... 11 
Chapter 2   ................................................................................................................................... 15 
 Introduction ..................................................................................................................... 16 
 Materials and Methods .................................................................................................... 18 
 Site description .................................................................................................... 18 
 Experimental Design and Treatments ................................................................. 18 
 Arthropod Sampling ............................................................................................ 21 
 Analyses .............................................................................................................. 23 
 
 vii 
 Results  ............................................................................................................................ 23 
 Discussion ....................................................................................................................... 25 
Chapter 3   ................................................................................................................................... 40 
 Introduction ..................................................................................................................... 41 
 Materials and Methods .................................................................................................... 44 
 Site description .................................................................................................... 44 
 Experimental Design ........................................................................................... 44 
 Arthropod Sampling ............................................................................................ 44 
 Earthworm Sampling .......................................................................................... 45 
 Nematode Sampling ............................................................................................ 46 
 Analyses .............................................................................................................. 46 
 Results  ............................................................................................................................ 48 
 Discussion ....................................................................................................................... 49 
References   ................................................................................................................................. 62 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 viii 
 
 
 
 
 
List of Tables 
 
 
Table 2.1 Seed Mix for Seeding and Seeding + Herbicide in 2010 ............................................ 31 
Table 2.2 Seed Mix for Seeding and Seeding + Herbicide in 2011 ............................................ 31 
Table 2.3 Average soil characteristics by site ............................................................................. 32 
Table 3.1 Average soil characteristics for cogon(C) and no cogon (NC) plots for objective 
2.
a
 ................................................................................................................................................ 54 
 
Table 3.2 Comparison of arthropods collected over entire season in longleaf with and 
without cogongrass present by Wilcoxon In cogongrass present plots, 18,968 arthropods 
were collected and counted. In cogongrass absent plots, 20,182 arthropods were collected 
and counted. Both were collected over an 8 month period in 2011. ........................................... 55 
 
Table 3.3 Comparison of populations of selected arthropods using all observations from 
each sample collected in pitfalls and sweep samples in longleaf pine with cogongrass 
present and cogongrass absent, Bay Minette and Grand Bay, AL, 2011, using repeated 
measures analyses ....................................................................................................................... 58 
 
Table 3.4 Mean abundance of different nematode taxa. Total nematodes, plant parasitic 
(PP), and free-living (FL) are represented. If P <0.05, difference is noted between 
cogongrass present (C) and cogongrass absent (N). Repeated measures test used. ................... 59 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ix 
 
 
 
 
 
List of Figures 
 
 
Figure 1.1 Longleaf pine stand with shrub and forb understory ................................................. 14 
Figure 1.2 Map of cogongrass distribution in Alabama in 2009 ................................................ 14 
Figure 2.1. Map of plot locations in Mobile County (Rainwater) and Baldwin County 
(Beebe) ........................................................................................................................................ 33 
 
Figure 2.2 Pitfall traps were arranged in an X pattern 3 meters apart on each 10 x 10 m 
plot .............................................................................................................................................. 3 
 
Figure 2.3 Composition of the natural enemy complex in the cogongrass treatment plots 
from pitfall traps and sweep samples. Samples contained mainly these three groups with 
parasitic hymenoptera and syrphids comprising the ?other? natural enemy group. 
Percentages reflect total number of individuals collected in 31 sampling periods, 2010 
and 2011. ..................................................................................................................................... 34 
 
Figure 2.4 Natural enemies collected in each treatment per month over 31 collection trips 
in 2010 and 2011. Biweekly samples pitfall traps and sweep samples were pooled into 
monthly total. ANOVA revealed: year (ANOVA, F= 6.72, df=1,63, P=0.0119) and 
treatment (F=2.06, df=3,59, P=0.1152). Significant interaction was: year by site 
(ANOVA, F=43.41, df=1,63, p <0.0001). .................................................................................. 34 
 
Figure 2.5 Abundance of total natural enemies sampled from May 2010-Oct 2011 from 
pitfall traps and sweep samples. Main effect of site (ANOVA, F=10.36, df=1,59, 
P=0.0021), burn treatment (ANOVA, F=14.69, df=1,59, P=0.0003), and site by burn 
treatment (ANOVA, F=28.67, df=1,59, P=<0.0001) were significant. ...................................... 35 
 
Figure 2.6 Abundance of predatory beetles sampled from May 2010-Oct 2011 from 
pitfall traps and sweep samples. Main effect of site, burn (ANOVA, F =29.46, df=1,59, 
P= <0.0001), and site by burn interaction (ANOVA, F =26.20, df=1,59, P =<0.0001) 
were significant ........................................................................................................................... 35 
 
Figure 2.7 Abundance of ants sampled from May 2010-Oct 2011 from pitfall traps and 
sweep samples. Main effect of site (ANOVA, F=0.06, df=1, 59, P=0.8093), burn 
treatment (ANOVA, F =10.89, df=1,59, P = 0.0016), and site by burn treatment 
(ANOVA, F =10.51, df=1,59, P = 0.0020) ................................................................................. 36 
 
 
 x 
Figure 2.8 Abundance of spiders sampled from May 2010-Oct 2011 from pitfall traps and 
sweep samples. Main effect of site (ANOVA, F =11.45, df=1,59, P =0.0013), burn 
treatment (ANOVA, F=0.11, df=1, 59, P=0.7398), and site by burn treatment (ANOVA, 
F =28.84, df=1,59, P =<0.0001). ................................................................................................ 36 
 
Figure 2.9 Predatory beetles collected in each treatment over 31 collection trips in 2010 
and 2011. ANOVA revealed: year (ANOVA, F =37.96, df=1,63, P =<0.0001), and 
treatment (ANOVA, F=1.75, df=3,59, P=0.1670). Significant interaction was: site by 
treatment (ANOVA, F value=4.25, df=3,59, P = 0.0087) .......................................................... 37 
 
Figure 2.10 Ants collected in each treatment over 31 collection trips in 2010 and 2011. 
ANOVA revealed: year (ANOVA, F =6.39, df=1,63, P= 0.0140), and treatment 
(ANOVA, F=1.94, df=3,59, P=0.1334). Significant interaction was: year by site 
(ANOVA, F=24.52, df=1,63, P = <0.0001 ................................................................................. 37 
 
Figure 2.11.  Spiders collected in each treatment over 31 collection trips in 2010 and 
2011. ANOVA revealed: year (ANOVA, F=1.21, df=1,63, P=0.2758), and treatment 
(ANOVA; F =4.11; df=1,63; P =0.0102). Significant interaction was: year by site 
(ANOVA, F =46.01, df=1,63, P =<0.0001) ............................................................................... 38 
 
Figure 2.12. Total Hylastes spp. collected in each treatment over 31 collection trips in 
2010 and 2011. Biweekly samples from baited pitfall traps were pooled into monthly 
totals. ANOVA revealed: year (ANOVA, F=6.38, df=1,63, P=0.0141), site (ANOVA, 
F=0.81, df=1,59, P=3730), burn treatment (ANOVA, F=0.14, df=1,59, P=0.7072), and 
treatment (ANOVA, F =0.43, df=3,59, P =0.7317). Significant interaction was seen in 
year by site (ANOVA, F =4.18, df=1,63, P =0.0452). ............................................................... 38 
 
Figure 2.13. Mean monthly temperatures in 2010 and 2011 for Baldwin County (Beebe) 
and Mobile County (Rainwater). (AWIS 2008) ......................................................................... 39 
 
Figure 2.14. Mean monthly precipitation in 2010 and 2011 for Baldwin County (Beebe) 
and Mobile County (Rainwater). (AWIS 2008) ......................................................................... 39 
 
Figure 3.1 Average soil characteristics for cogon(C) and no cogon (NC) plots for 
objective 2.
a
 ................................................................................................................................. 60 
 
Figure 3.2 Comparison of arthropods collected over entire season in longleaf with and 
without cogongrass present by Wilcoxon In cogongrass present plots, 18,968 arthropods 
were collected and counted. In cogongrass absent plots, 20,182 arthropods were collected 
and counted. Both were collected over an 8 month period in 2011 ............................................ 60 
 
Figure 3.3 Earthworm abundance in 2011 on cogongrass present and cogongrass absent 
plots by abundance. Year was not significant, therefore samples were combined. 
ANOVA: F= 8.00, df=1, 37, P=0.0075 ...................................................................................... 61 
 
 
 xi 
Figure 3.4 Earthworms collected in 2011 on cogongrass present and cogongrass absent 
plots by weight (mg). Year was not significant, therefore samples were combined.  
ANOVA F=7.57, df=1, 37, P=0.0091. ....................................................................................... 61 
 
 1 
Chapter 1 
 
Introduction 
Longleaf Pine Ecosystems 
 In the southeastern United States, longleaf pine (Pinus palustris Mil.) savanahs are 
unique ecologicaly (Bartholomew et al. 206) because they are composed of many threatened 
and endangered endemic plant and animal species, and host an extremely species rich plant 
comunity (Brewer and Crale 203). They are characterized by an open canopy, lack of mid-
story tres or shrubs, and wel-developed, highly diverse herbaceous understory which is 
dominated by bunch grases, graminoids, shrubs, vines, and legumes and forbs (Figure 1.1) 
(Boyer 190, Daneshgar and Jose 209). Without frequent, low intensity fires, this ecosystem 
quickly suceds to a closed canopy hardwod mixed pine or evergren shrub habitat (Boyer 
190). Historicaly, the range of longleaf pine stretched from southeastern Virginia to central 
Florida and west into Texas (Haywod 209). Within Alabama, the majority of remaining 
longleaf sites are located in the southern portion of the state. Litle of the original vast longleaf 
forests remain, and conservation and restoration of longleaf ecosystems is a high priority in the 
Southeast (Outcault and Shefield 196). 
  
Insects in the Pine Ecosystem 
 Bark betles and wevils (Curculionidae: Scolytinae) are the most destructive forest pests 
in the United States (Moeck and Safranyik 1984, Masey 1974). These include Dendroctonus, 
Ips, and Hylastes sp. Rot-feding Hylastes sp. are curently considered non-agresive and 
not a threat. They typicaly atack unhealthy, declining, wounded or even dead pines (Eckhardt et 
al. 204). Stres-altered alelochemical profiles of pine rots atract the scolytids and they 
agresively fed on the rots releasing these chemicals (Eckhardt et al. 2008). Some other 
insects asociated to Scolytid pheromones are: Rhizophagidae, Trogositidae, Staphylinidae, 
 
 2 
Histeridae, Nitidulidae, Tenebrionidae (Coleoptera), Dolichopodidae, Lonchaeidae, Asilidae, 
Palopteridae (Diptera), Anthocoridae (Hemiptera), and Raphididae (Neuroptera). Betles in the 
Carabidae and Silvanidae familes are losely asociated with bark betles (Kenis et al. 204). 
 Certain arthropod taxa (i.e., ants, betles, spiders) are known to be bioindicators in forest 
management. Ants (Formicidae) are efective bioindicators because of their sensitivity to 
ecosystem disruption caused by forest fires, species invasion, and other forms of disturbance 
(Maleque et al. 2009). They are asociated with important ecosystem proceses such as soil 
aeration, soil nitrogen cycling, sed dispersal and sed predation (H?ldobler and Wilson 190). 
In Alabama, there are aproximately 16 species of ants. Of those, twenty-eight species are 
exotic (MacGown and Forster 205). Red imported fire ants (Solenopsis invicta, or RIFA) play a 
role in simplification of ecosystems by out-competing other ants (Reagan 1982). Red imported 
fire ants were introduced near Mobile, AL around 1930 (Lofgren 1986). Within 10 years of 
introduction, S. invicta had spread to 10 states in the southeast. High reproductive capacity, 
propensity for soil invasion, and eficient foraging behaviors undoubtedly contributed to their 
rapid spread (Lofgren 1986). Ecologicaly, S. invicta can be considered a wed species. Weds 
are plant or animal species adapted for oportunistic exploitation of ecologicaly disturbed 
habitats (Tschinkel 1986). Disturbed habitats can ocur from natural forces such as wind and 
storms as wel as human disturbances like agriculture or homestead establishment. Human 
mediated activities are the greatest encourager of S. invicta range expansion (Tschinkel 1986). 
Fire ants flourish on early and secondary sucesion comunities but are absent or rare in late 
sucesion and climax comunities. They are often found mixed with other early sucesion 
wed species, both plants and animals. Overal, the red imported fire ant in Alabama is 
fundamentaly an oportunistic wed species dependent on habitat disturbance for its continued 
suces and existence (Tschinkel 1986). 
 Although S. invicta is notoriously detrimental to human activities, it is beneficial in some 
situations (Reagan 1986). In the crop fields of southern Louisiana, S. invicta was the most 
 
 3 
important predator relative to other comon predator groups (spiders, carabids, earwigs, and 
staphylinids; Charpentier et al. 1967). In forest systems, ants are important predators of egs and 
larvae, especialy of Lepidoptera. Red imported fire ants, specificaly, are known predators of 
forest pests including Nantucket pine tip moth (Rhyacionia frustrana) (Reagan 1986) and spruce 
budworm (Choristoneura hebenstreitela) (Youngs 1983). 
Like ants, predatory betles in the families Carabidae, Staphylinidae, Cicindelidae, 
Histeridae and Cleridae are also generalist predators of forest pests (Kenis et al. 204, Gandhi et 
al. 201). Specificaly, ground betles (Carabidae) are used to gauge local-scale disturbance 
intensity because they are porly dispersing, large-bodied insects that typicaly decrease in 
abundance with interference (Maleque et al. 2009, Rainio and Niemela 202).  Carabids decline 
quickly in response to disturbance, forest fragmentation, fertilization, and grazing and are 
sensitive to temperature, humidity, vegetation, and patch size. The most important factors 
impacting ground betle abundance are soil water content, fragmentation and vegetation. They 
are also among the first to respond to forestry disturbances (Rainio and Niemela 202). They can 
also function as indicators of arthropod groups on which they fed. Carabid species richnes tends 
to be lower in closed canopy situations compared to open habitats (Humphrey et al. 199). 
Larvae and adults ocur in the leaf liter or soil and are omnivorous or predaceous usualy feding 
on dead or dying insects (White 1983). Adult carabids are ground active, sometimes in large 
numbers. Both larvae and adults are active at night and hide during the day. Ground betles in the 
Carabini tribe wil climb tres in their search for prey and the adults consume a much greater 
amount of fod than the larvae (White 1983). Pedunculate ground betles (tribe Scaritini) have 
the largest mandibles in the family and are some of the most rapacious predators. They are most 
comonly found in open fields, pastures and wodlands feding on caterpilars and other larvae. 
False bombardier betles (tribe Galeritini) are important for reducing cankerworm populations 
(White 1983). Carabids in the genus Dromius (tribe Harpilini) are asociated with Scolytids 
(Kenis et al. 204). 
 
 4 
Similar to the ground betles, tiger betles (Cicindelidae) are large betles and both the 
larvae and adults are predaceous, with the adults being particularly voracious. Usualy, adult tiger 
betles are found in open sandy areas or paths and sek out their prey. Larvae create vertical 
burows in compact soil and are sit-and-wait predators (White 1983). 
Another predaceous betle family is the hister betles (Histeridae). The diet of adults and 
larvae consists almost completely of insects and other smal animals. Usualy these betles are 
found in dead or dying tre bark or in galeries under the bark surface (White 1983). Staphylinid, 
or rove betle, adults and larvae are also predominantly predaceous insects. A few of these 
betles are known to be parasitic (Boror and White 1970). Rove betles are one of the largest 
families of betles and about half are found in leaf liter. They comprise one of the most 
ecologicaly important and comon insect constituents of soil fauna (Bohac 199). The majority 
of these betles are non-specific predators that fed on soil arthropods such as nematodes, mites, 
colembolans, and smal insect larvae or imatures. Staphylinids serve as bioindicators of 
environmental status, especialy when measuring human influence on ecosystems (Maleque et al. 
209). These families comprise the predatory betle group that wil be later discused.  
Natural enemies exist aerialy as wel. Species richnes and diversity of hover flies 
(Diptera: Syrphidae) are positively corelated with stand structural complexity and ground layer 
vegetation (Maleque et al. 209). Syrphids are god indicators because they serve varying 
functional roles in an ecosystem and have diverse habitat requirements. Much like carabids, hover 
fly diversity is reduced in areas with a closed canopy (Humphrey et al. 199). Parasitic wasps are 
complex and specialized because of their high trophic position, intricate biology, narow host 
ranges, and habitat requirements. These wasps are found to be more abundant in species rich 
mixed wodland habitats than in coniferous wodlands (Maleque et al. 209). 
Spiders respond to natural and anthropogenic disturbances and are indicators of 
ecosystem changes caused by forest fires, vegetation development and changes in stand 
complexity (Maleque et al. 209). Spiders and other arthropods play a large role in ecological 
 
 5 
features of soil and detritus of forest flors. Spiders are among the most esential entomophagous 
predators in an ecosystem and a reduction in abundance can interfere with ecosystem function 
(Moulder and Reichle 1972). 
 
Earthworms and Nematodes 
Earthworms (Oligochaeta) are wel known as burowers in soil systems and are important 
bioindicators. Their burowing and feding activities ad inorganic residues to the soil, as wel as 
enhance decomposition, humus formation and nutrient cycling. They also play a critical role in 
soil structure development (Werner 190). Distribution of earthworm species is strongly 
influenced by the physiographic region. In Alabama, 43% of earthworm species are exclusively 
found in the coastal plain (Reynolds 194). 
Nematodes are roundworms (Nematoda) and second only to arthropods in number of 
species and individuals present. There are two main groups of nematodes: fre-living and 
parasitic (Wharton 1986). Free-living nematodes live in the soil or water. Nematodes can be plant 
feding, animal feding (parasitic guilds) or bacterial/fungal feders. About 48 % of nematode 
genera are parasites of plants and animals (Wharton 1986). In the pine system, nematodes 
primarily afect sedlings. The stunt nematode, Tylenchorhynchus claytoni, and a stuby-rot 
nematode, Paratrichodorus minor invade southern forest nurseries (Cram and Fraedrich 207). 
Ruehle (1973) investigated the influence of plant-parasitic nematodes on longleaf pine and found 
several nematodes that decreased weight of sedlings. These include Holoplolaimus galeatus, 
Tylenchorhynchus claytoni, Meloidodera floridensis and Pratylenchus brachyurus (Ruehle 1973). 
Overal, nematodes are not considered a problem in nurseries because their populations are 
controled by soil fumigation with methyl bromide (Cram and Fraedrich 207). 
Nematodes are one of the major biotic factors impacting bark betle populations. 
Members of the Contortylenchus genus are probably the most familiar parasitic nematode on bark 
betles. Most nematodes asociated with bark betles are true parasites; only a few il kil the 
 
 6 
host. The most pronounced efect sen on bark betles from internal parasites is brod reduction. 
This is because, in most cases of mortality, it is the female that perishes (Masey 1974). 
 
Cogongras Morphology, Biology, and Distribution 
Cogongras, Japanese blodgras, or alang-alang (Imperata cylindrica (L.) Beauv) is a 
long-lived, invasive C
4
 gras. Morphologicaly, leaves are a lime gren color (0.3?1.5 m long, 
1.2?2.5 cm wide) with a frequently of set midrib and minutely serated leaf margins. Except for 
the flowering stalk, cogongras is virtualy stemles aboveground (Dozier et al. 198, MacDonald 
204). The rot and rhizome system can, and often does, compose 60% or more of the total plant 
biomas and is found in the uper 15?40 cm of soil (MacDonald 204, Daneshgar and Jose 
2009). Rhizomes are long, white, tough and scaly with very short internodes. The branched 
rhizomes create a dense mat that excludes most other vegetation. Apical ends of rhizomes are 
sharp and may grow through rots of other plants, thus enhancing its ability to debilitate other 
plant species (Dozier et al. 198, MacDonald 204). Unlike seds, rhizomes can invade 
undisturbed pine savanas (Lipincot 200). Site to site variation can ocur, but a single ramet 
can have over 350 descendant shots in just 6 weks under optimal conditions. After 1 weks, 
clones can of a ramet cover at least 4 m
2
 (Brewer and Crale 203). Soil contaminated with 
rhizome fragments as smal in size as 0.1 g can be an important mechanism of cogongras spread 
(Dozier et al. 198, Daneshgar et al. 208, Daneshgar and Jose 209). Cogongras rhizomes can 
also penetrate rocky soils and even asphalt pavement (Bryson and Carter 204).  
Plants typicaly flower in the early spring, but flowering may ocur year round in Florida 
(Van Loan et al. 202). Flowers are a terminal, cylindrical panicle, typicaly 10?20 cm long, 
comprised of spikelets (MacDonald 204). As many as 460 white spikelets, aproximately 3-6 
cm in length are found on each panicle (Dozier et al. 198). Panicles are silky and become flufy 
with maturity. Seds are viable for aproximately one year and can germinate imediately after 
harvest (MacDonald 204). There are contrasting reports on how critical sed production is for 
 
 7 
reproduction (Sajise 1972, MacDonald et al. 204). Human mediated activity, water, and animals 
disperse seds, or they may be caried by wind within 15m of the plant or up to 25 km in open 
areas (Bryson and Carter 204, Daneshgar et al. 208). Anthropogenic means of spread are much 
worse than natural factors (Capo-chichi et al. 208) and soil disturbance (i.e., discing, tiling, road 
creation) encourages seding establishment (Brewer and Crale 203). Cogongras spread in 
Alabama can be linked to the northeastern winds from the gulf as wel as sed spread along 
roadways (Dozier et al. 198). 
 Globaly, cogongras is ranked number seven among the ten worst weds (Coile and 
Schiling 193, MacDonald 204, Brewer 208) because of severe impacts on agriculture and 
forestry.  It is a major hindrance to reforestation eforts in Southeast Asia and the number one 
wed in agriculture production in many parts of Africa (Brewer and Crale 203).   Cogongras 
was acidentaly and intentionaly introduced into North America. Populations introduced in the 
early 190?s to Grand Bay, Alabama, have had significant impacts on native habitat in the 
southeastern United States (Jose et al. 202, Dozier et al. 208). At first, its distribution was slow, 
ocupying 1235 ha by 1950 (Dickens 1974) but curent estimates surpas 20,00 ha of land 
primarily in southwest Alabama (Capo-chichi et al. 204, Brewer 208; Figure 1.2). 
 
Ecosystem Impacts of Cogongras 
 Cogongras invasivenes can be atributed to its being adaptable to por soils, having 
high drought tolerance, wind-diseminated sed production, fire adaptability and high genetic 
plasticity (MacDonald 204, Dozier et al. 198). Cogongras also poseses alelopathic 
characters (Bryson and Carter 204, Coile and Shiling 193, Daneshgar et al. 208) that 
characters inhibit growth, germination, and survival of other plants (Daneshgar et al. 208) and 
aid in the creation of a monotypic expanse (Coile and Shiling 193). Cogongras has not become 
a threat to agricultural production in the US because tilage and herbicides relegate it to field 
edges. However, as its frequency increases worldwide, ecologists and conservationists fear 
 
 8 
cogongras wil displace native animal and plant species as wel as alter the fire regime in some 
forest ecosystems (Brewer and Crale 203). 
 Its fire tolerant capabilities provide it with an advantage over non-tolerant, usualy native, 
species. Cogongras fires burn about 15-20% hoter than other groundcover. However, this 
temperature efect is not sen on the surface, but at 1.2 - 1.8 m of the ground (Duever 208). 
Rhizomes persist below the soil surface and produce new shots quickly folowing a burn 
(Bryson and Carter 204). Cogongras? dense rhizome mat creates a high rot to shot ratio 
alowing it to thrive after fire or mowing. Burns in gras-invaded areas tend to be more frequent 
and have a greater spatial expanse (Daneshgar et al. 208). 
Cogongras invasion can cause significant declines in herbaceous plants in longleaf pine 
savanas and reduce the species richnes of native flora. Brewer and Crale (203) found that the 
older the infestation, the lower the richnes of native plant species. Cogongras stands create dep 
shade, which can displace lower growing native plant species. Cogon-tolerant native species 
therefore are mostly taler and wodier plants. For example, yaupon (Ilex vomitoria), wax myrtle 
(Morela cerifera), and persimon (Diospyros virginiana) al poses the potential to displace 
smaler native species, especialy after long periods of supresed fire (Glitzenstein et al. 203). 
Unlike cogongras, these species lack agresive rhizomes and do not quickly recover from fire 
(Brewer 208). In the longleaf pine ecosystems, already in danger, cogongras presence could 
push endemic plant species to threatened and even endangered listings (Brewer 208). Brewer 
(208) concluded that even though most sugest cogongras poses a greater threat because of the 
altered fire regime, reduced plant diversity might be a more presing conservation concern.    
Despite the rate at which cogongras can and wil overtake a comunity, litle is known 
about its afect on invertebrate populations. Native plant and animal species tend to be most 
susceptible to invasion by exotic species because they are usualy specialists and inferior 
competitors or they lack defense mechanisms and high reproductive outputs (Brewer 208). 
 
 
 9 
Cogongras Management 
 A variety of tols have ben used for cogongras management including biological, 
physical, chemical and cultural methods. The key to suces in cogongras management lies in 
the ability of a given treatment to sucesfuly kil it or supres its recovery from the extensive 
rhizome network that cogongras develops (Dozier et al. 208). Repeated tilage, or discing, 
which fragments the rhizome layer, has ben useful and is primarily why cogongras is not a 
greater problem to agriculture production in the southeastern US. Repeated mowing is only 
helpful for sed head supresion because shot recovery is rapid. Both tiling and mowing 
provide temporary supresion of cogongras, but are not efective in providing long-term control 
and are often impractical and unrealistic for large highly infested areas (Dozier et al. 198). 
Burning is useful for removing dense cogon thatch but provides no long-term control when used 
alone. Burning and mowing remove old growth and dead biomas (thatch) and weaken rhizomes 
through alocation of starch resources to make new shots. Also, removal of the dense thatch 
layer makes secondary management strategies more sucesful. 
Glyphosate and imazapyr are the only herbicides that significantly impact cogongras and 
are aproved for use in pine ecosystems. Imazapyr is an acetolactate synthase (ALS) inhibitor. 
The ALS enzyme is a key enzyme in production of thre amino acids (valine, leucine, and 
isoleucine). Blocking ALS activity prevents the conversion of pyruvate to acetolactate (eventualy 
to valine and isoleucine) and conversion to 2-ketobutyrate to 2-acetohydroxybutyrate and 
eventualy isoleucine (Senseman 207). Imazapyr is a non-selective, foliar, and soil active 
herbicide that is generaly aplied post emergence. Symptoms of imazapyr toxicity apear after 
3?10 d and after 17?24 d plant wil show a dep red color and symptoms of chlorosis, leading to 
necrosis and plant death (Senseman 207). However, symptoms may not be expresed on 
cogongras for up to 60 days. Imazapyr has considerable soil residual activity that may interfere 
with sedling establishment and injure or kil other species where rots are in the treated area. 
 
 10 
Glyphosate is a 5-enolpyruvyl shimate 3-phosphate (EPSP) synthase inhibitor. It inhibits 
the activity of EPSP synthase, which converts shikimate-3-phosphate plus phosphoenolypyruvate 
to 5-enolpyruvyl shikimate-3-phosphate; this inhibits production of aromatic amino acids: 
tryptophan, tyrosine and phenylalanine. Like imazapyr, glyphosate is non-selective with 
symptoms of growth cesation begining within 3?10 d. After 17?24 d, plants wil show 
dehydration and desication and eventualy necrosis and plant death (Senseman 207). High 
aplication rates of imazypyr and glyphosate cause rapid foliar kil, but intermediate levels have 
proven to be more efective (Dozier et al. 198). In fact, the most efective method of herbicide 
treatment involves multiple aplications. Most efective multiple treatment management 
strategies work best under the earliest developmental stages. Older developmental stages show a 
susceptible foliar reaction, but rhizomes are slightly resistant to glyphosate. Also, basal area of 
the leaf proved to be the most likely area to absorb glyphosate. Seasonal diferences can afect the 
impact herbicides have on cogongras management strategies (Dozier et al. 198). It is critical to 
use the corect rate during the apropriate time. Herbicide aplications can be expensive, up to 
$40 ha/year, and multiple aplications per year may be required (Capo-chichi et al. 208). 
Integrating burning, mowing, discing, herbicides, and cultural control wil increase the 
prospect of cogongras supresion. First discing in the fal, then folowing with an herbicide 
aplication has the potential of providing over 90% control and supresion. The sedlings? under 
developed tisues are a perfect target and ofer maximized absorption and increased eficacy. 
Folowing fal mowing and herbicide treatments, it is esential that desirable plant species be 
planted. In order to remove cogongras more efectively, it must be replaced or it wil regain 
dominance in the empty niche. Careful consideration must be taken when deciding which species 
to include. If the planted species canot withstand cogongras, they wil be easily taken over and 
a monotypic expanse wil eventualy be recreated. Some sucesful characters of the new species 
include herbicide tolerance, as wel as tolerance of soil type, canopy cover and other factors 
 
 11 
(Dozier et al. 198). Revegetation is very useful to supres cogongras recovery folowing 
control (MacDonald 204) and is an integral component of ecologicaly based wed management.  
 Long term, highly intensive management strategies have not worked so far. The only 
known survey of insects within cogongras in the US focused solely on herbaceous feding 
insects as potential biological control candidates (Mino and Mino 199). From this work, 51 
insect species were observed in cogongras, but no suitable insects for the purpose of biological 
control were found. In Indonesia, Maeto et al. (209) found reduced braconid abundance and 
species richnes in Imperata graslands compared to Acacia plantations and old secondary 
forests.  Few promising candidates have ben found for biological control and none are curently 
available (Van Loan et al. 202).  
 
Interactions 
 Rot-feding bark betles (Hylastes sp.) spread ophiostomatoid fungi that are linked to 
pine decline (Eckhardt et al. 204). Pine decline in the southeastern United States is another 
growing problem that also severely impacts both the environment and the livelihod of the 
Southeast (Eckhardt et al. 208). Lobloly pine decline (LPD) is an example of a complicated 
syndrome involving the host, insect, pathogen and site interactions (Eckhardt et al. 2007). 
Symptoms of LPD are sen in tres on wel-drained sites as sparse crowns, short chlorotic 
nedles, and gradual mortality. Older tres begin to decline slowly and die prematurely. 
Eventualy, LPD spread by rot-feding bark betles to younger tres and the onset of symptoms 
asociated with fungal and insect species are sen (Eckhardt et al. 204, Eckhardt et al. 207). 
Tres over 40 years old are most frequently weakened by LPD. These declined tres are more 
prone to southern pine betle (SPB) atack (Otrosina et al. 197). 
 Curent studies of pine stands in the southeastern United States containing tres 
expresing symptoms of decline show significantly higher numbers of rot-feding bark betles 
and the presence of ophiostomatoid fungi (Eckhardt et al. 207). Species of Leptographium and 
 
 12 
Ophiostoma are most comonly reported as causes of blue stains in pine, spruce and other 
conifers. These stain fungi rarely kil healthy tres, but they do reduce the value of timber and 
may be asociated with mortality in stresed tres (Wingfield et al. 198). Leptographium species 
are comonly asociated with various species of rot-feding bark betles (primarily Hylastes 
sp.) and rot wevils, which atack healthy and stresed tres (Eckhardt et al. 207).  These 
insects serve as vectors of these fungi and as wounding agents creating courts, which permit 
fungal infection. In a study done on lobloly pine in Alabama, Eckhardt et al. (207) found that 
symptomatic sites were asociated with increased numbers of insects. Rot and lower stem 
feding betles were significantly more abundant in symptomatic sites when compared to 
asymptomatic ones. The rot-feding betles considered for this study were: Hylastes salebrosus, 
H. tenuis, Hylobius pales and Pachylobius picivorus. Betles colected were asociated with 
Leptographium procerum, L. terebrantis and Grosmania. serpens and could be serving as 
vectors (Eckhardt et al. 204).  
 Hylastes sp. are found in greater numbers within pine tres that are symptomatic for 
pine decline when compared to asymptomatic pines (Eckhardt et al. 207). Although the betles 
are responsible for vectoring fungal spores, the fungus alone does more damage because it 
interacts with the vascular tisue of the tres. While not completely understod, pine decline has 
ben atributed to a combination of tre stres, native and exotic insects, and native and exotic 
fungi that can severely stres or kil species such as longleaf, shortleaf, and lobloly pine 
(Eckhardt et al. 208). While these arthropods, disease, and weds comonly co-ocur, litle is 
known concerning how these pests interact or if cogongras promotes cascading negative impacts 
within pine ecosystems (Daneshgar et al. 208). 
 Similarly, burning and herbicide strategies used for managing cogongras can also impact 
the abundance of key arthropods in the pine system (Burke et al. 208). Lobloly pine managed 
with burning and imazypyr actualy had greater arthropod diversity relative to control plots, 
however, impacts among taxonomic groups varied widely (Burke et al. 208). Abundances of 
 
 13 
spiders and Orthoptera were impacted by management strategy but there was no treatment efect 
among Coleoptera, Lepidoptera, Diptera, Hemiptera, or Hymenoptera colected in pitfal or 
swep net samples (Burke et al. 208). Glyphosate aplications can negatively afect 
survivorship and behavior of wolf spiders (Evans et al. 2010). These efects can be direct (Evans 
et al. 2010) or indirect impacts related to reductions in vegetation height and increases in dead 
vegetation resulting from treatment (Bel et al. 202). Wed management can also alter the prey 
taken and ant abundance. In wed-fre areas, smaler ant populations were recorded and a greater 
number of nests were abandoned (Reagan 1986). Similar management strategies for cogongras 
in longleaf pine have not ben asesed for their potential impacts on arthropods. Reductions in 
generalist predators and other natural enemies would potentialy reduce predation on rot-feding 
bark betles and posibly increase the incidence of pine decline in cogongras-infested pines. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 14 
 
 
 
 
 
 
 
 
 
Figure 1.1. Longleaf pine stand with shrub and forb understory 
 
 
 
  
 
 
 
 
 
 
 
 
 
 
Figure 1.2. Map of Cogongras distribution in Alabama in 209 
 
 
 15 
 
Chapter 2 
Impact of Cogongrass Management Strategies on Generalist Predators of Hylastes 
spp. in Longleaf Pine 
 
Abstract 
Cogongrass (Imperata cylindrica Beav.), an aggressive, invasive weed, is 
threatening the integrity of longleaf pine (Pinus palustris Mill.) ecosystems. Various 
management strategies are used to suppress cogongrass such as fire, herbicide, and 
seeding with native plants. To understand the effect management strategies have on 
arthropod natural enemies and root-feeding bark beetles a split plot design to test fire 
(whole plot) and 4 sub plot treatments: control (no additional treatment), herbicide, 
seeding, and herbicide plus seeding were used. Arthropods were sampled using pitfall 
traps and sweep samples. After two years of sampling, counts of all natural enemies were 
not significantly affected subplot by treatment but were affected by burn treatment. Upon 
subdividing natural enemies into groups, only spiders were significantly affected by 
subplot treatment, but predatory beetles and ants were significantly affected by burn 
treatment. Root-feeding bark beetle (Hylastes spp.) abundance was not significantly 
affected by treatment or burn treatment. Cogongrass management strategies overall do 
not affect natural enemy populations. Undisturbed cogongrass provides the best habitat 
for natural enemy assemblages. 
 
 
 
 16 
 
Introduction 
Cogongrass (Imperata cylindrica Beav.) is an aggressive, invasive, perennial, C
4
 
grass originating from south East Asia that was introduced to south Alabama around 
1910 and again in the mid 1940s (Dozier et al. 1998). Since its introduction, it has spread 
into all states bordering Alabama, as well as Louisiana, Texas, and South Carolina. 
Cogongrass changes the fire regime and inhibits commercial forestry and reforestation 
efforts (Lippincott 2000). Invasion of cogongrass into longleaf pine (Pinus palustris 
Mill.) stand results in a monotypic understory (Lippincott 2000) and established pines 
may begin to decline (Daneshgar et al. 2008). Longleaf pine ecosystems are typically 
composed a diverse understory of graminoids, shrubs, forbs, vines and legumes 
(Daneshgar and Jose 2009). Burning is an important regulator of longleaf pine stands.  In 
the presence of cogongrass, prescribed or natural fires are higher and hotter, which causes 
more damage to trees and understory plants than a normal, cogongrass-free burn (Bryson 
and Carter 2004).  
Control of cogongrass requires an integrated approach of cultural and chemical 
controls (MacDonald 2004). Natural or prescribed burning, although not recommended 
for control, removes the thatch layer and forces the plant to allocate carbohydrates to 
shoot growth, temporarily weakening the rhizome (Brewer 2008). Mowing is also not a 
recommended method of control but is used for seedhead suppression, especially on 
roadsides. Because 60% of cogongrass biomass is underground in the extensive rhizome 
mat, it can also quickly recover after a fire or mowing. Discing alone does not provide 
control, but can reduce biomass (Brewer 2008). Discing followed by late fall herbicide 
 
 17 
application, however, can yield greater than 90% control (Dozier et al. 1998). Native 
plants generally do not recover as quickly as cogongrass, and plant diversity is reduced 
after disturbances due to displacement (Brewer 2008).  
Herbicides, mainly imazapyr and glyphosate (MacDonald 2004) are effective, 
widely used controls used for congongrass. However, multiple applications for several 
years are required (Capo-chichi et al. 2008). High application rates of herbicides will 
result in rapid foliar death, but to successfully kill cogongrass, rhizomes must receive a 
lethal dose of herbicide (Dozier et al. 1998).  Herbicides alone can be expensive ($400 
ha/yr) (Capo-chichi et al. 2008). For long-term control, establishing desired plant species 
to ultimately replace cogongrass is essential after suppression or removal with herbicides 
(Dozier et al. 1998). These revegetation treatments are done by introducing native plants 
that produce heavy shade and dense rhizomes that can compete with cogongrass and 
prevent recolonization (Yager et al. 2011, Dozier et al. 1998).  
Although not very well studied, all of these aforementioned control tactics for 
cogongrass may have measureable effects on the arthropod community.  Longleaf pine 
invaded by cogongrass would produce hotter fires subjecting trees to increased stress 
(Lippincott 2000). Beetles in the genus Hylastes (Curculionidae, Scotlytinae) are root-
feeding bark beetles and cannot attack and kill healthy trees, unlike the southern pine 
beetles (Dendroctonus frontalis Zimmerman). Hylastes are attracted to stressed, already 
weakened trees. Environmental disturbance like hotter, more frequent fire can weaken 
trees, stressing them causing them to be and more susceptible to beetles and root 
infections.  Changes in the structure and cover of vegetation resulting from burning or 
herbicide application may impact generalist predators such as ants (Castano-Meneses and 
 
 18 
Palazois-Vargas 2003, Matsuda et al. 2011), spiders (Bell et al. 2002, Burke et al. 2008) 
or predatory beetles (Maleque et al. 2009, Holland and Luff 2000).  Populations of 
parasitic Hymenoptera are negatively impacted by cogongrass invasion into forest 
systems, but populations can recover through replanting and increased plant diversity 
(Maeto et al. 2009). 
The objective for this study was to understand how cogongrass management 
strategies used in longleaf pine communities would influence root-feeding bark beetles 
and diversity and abundance of invertebrate natural enemies. Treatments such as burning 
locally increase bark beetles (Bauman 2003) whereas the recovery of plant diversity 
following removal of cogongrass should increase natural enemies. 
 
Materials and Methods 
Site Description. Two sites in southwest Alabama, located within open longleaf pine 
stands approximately 10 years old, were selected for study. Site one was located in 
Mobile County, AL near the Mississippi State line on Airport Road and will hereafter be 
referred to as the ?Rainwater? site. Site two was located in Baldwin County, AL near Bay 
Minette on White House Fork Road and will hereafter be referred to as the ?Beebe? site 
(Figure 2.1). Each site was heavily infested with established swards of cogongrass, which 
completely dominated the understory.  Both sites had similar soil characteristics 
including texture, pH, CEC, and water holding capacity. However, soil organic matter 
was slightly higher at the Rainwater site than the Beebe Farm site  (Table 2.1). Longleaf 
trees at each site ranged from 3.4 to 12.2 m in height and 11.4 to 25.4 cm diameter at 
breast height (DBH)   
 
 19 
Experimental Design and Treatments. At each site, we utilized a split plot 
experimental design to allow for a consolidated prescribed burn. The whole plot 
treatments were prescribed burning or no burning. Fire breaks were established around 
the perimeter of the burned portion of each site to prevent fire escape. The prescribed 
burn was conducted by a private contractor in early March of 2010. Since cogongrass is a 
pyrogenic species that is not well suited for safe prescribed burns, a backing fire was used 
to minimize fire severity and prevent more severe conditions that typically occur when 
cogongrass is burned. 
Following the burning treatment, four cogongrass management treatments were 
imposed on 10 by 10 meter subplots in the burned and unburned main plots: 1) an 
untreated control; 2) disking and seeding with native herbaceous species; 3) repeated 
glyphosate treatment, and 4) an integrated treatment consisting of glyphosate treatment 
followed by disking and seeding with native herbaceous species and subsequent spot 
treatments of glyphosate as necessary. Subplot treatments were randomly assigned and 
replicated five times within each main plot at each site except for the seeding only 
treatment, which was replicated three times within each main plot. Plots were surrounded 
by a 1.5 m buffer that was treated with glyphosate as needed to prevent cogongrass 
spread between plots.  
 
 20 
In 2010, plots assigned to the glyphosate herbicide treatment received the 
herbicide glyphosate (Accord XRT 2; 646 g ai per liter; Dow AgroSciences, Indianapolis, 
IL) at 2.5% and 4% v/v solution in the spring and fall respectively. The spring treatments 
were broadcast applied to each plot on April 14, 2010 and fall treatments on October 14, 
2010. In 2011, glyphosate was applied as a spot treatment only to cogongrass within plots 
on November 11. For all applications, glyphosate treatments were applied at 514 l/ha (55 
gpa) with a single nozzle Solo CO
2
 pressurized backpack sprayer. 
Plots receiving the disking and seeding treatment also received glyphosate 
treatments as previously described. A mixture of commercially available, competitive, 
native species (Table 2.2) was selected to create a desirable herbaceous understory. This 
species mix is a functionally diverse mix of species that are drought tolerant, create heavy 
shade, grow densely, are N-fixers, and are rhizomatous or deeply rooted. The species in 
the mix also provide numerous benefits including habitat and food sources for insects and 
wildlife. The plots were seeded at a total seeding rate of 40 kg per hectare (36 lb/a) with 
varying percentages of each species in the mix (Table 2.2). After year one, the species 
mix was altered due to an extremely low germination and establishment rate due to 
drought. Four species were removed, and six were added (Table 2.3). The seed mix was 
sown by hand in April of 2010 and February of 2011. Each plot was divided into 1.5 m 
strips and an equal portion of seed was then hand broadcasted across each strip. Before 
 
 21 
seeding, plots were lightly disked (5-8 cm deep) to break up cogongrass thatch and 
provide bare soil for seed contact. 
Arthropod Sampling.  Arthropods were sampled using two methods: pitfall traps for 
ground active arthropods, and sweep samples primarily targeting arthropods associated 
with vegetation.  Five pitfall traps (15 cm dia funnel type traps, Multi-pher, Bio-Controle, 
Quebec City, Quebec), were installed in each plot, arranged in an X pattern 3 m apart 
along each diagonal with one in the center (Figure 2.2).  Inside each trap, a 453 ml plastic 
cup (16K, Dart, Mason, MI) was filled with 50 ml of a 1:1 (v/v) mixture of water and 
propylene glycol (Peak, Old World Industries, Northbook, IL) to serve as a preserving 
solution.  Every 2 wk, contents from all five pitfall traps were pooled into a labeled 
plastic container.  Empty traps were then recharged with preserving solution.  Pitfall traps 
were active from May 1 through Oct 28, 2010. Pitfall traps remained empty in the field 
overwinter.  In 2011, traps were activated on Feb 15 and serviced biweekly through Oct 
25. Soil displaced from pitfall trap installation was collected and brought back to the AU 
soil dynamics lab for analysis. 
 Sweep samples were taken biweekly, concurrent with servicing pitfall traps.  One 
sample was made per plot, following each diagonal making approximately 30?40 sweeps 
per plot.  Insects collected in the sweep net were put into a 3.8 liter plastic zipper bags 
and placed into a cooler for transport back to the laboratory.  
Inside the laboratory, arthropods from pitfall traps contents were filtered to 
remove dirt and other debris.  The remaining arthropods were stored in 50 ml plastic vials 
with 90% ethanol in a freezer until they were sorted.  Arthropods were sorted into 
taxonomic groups and stored in 22 mL glass vials with 90% ethanol. The taxonomic 
 
 22 
groups of interest were: predatory beetles (Staphylindiae, Carabidae, Cleridae, Histeridae, 
Cicindelidae, Coccinellidae, and predatory larvae), ants (Formicidae), spiders (Aranae), 
all of which are bioindicators in forest systems (Maeto et al. 2009, Maleque et al. 2009) 
and important in ecosystem services (Isaacs et al. 2009).  
In the lab, sweep samples were placed in the freezer for ?1 d. Arthropods were 
then removed from the bags and placed into 15 mL vials by plot in 90% ethanol. From 
these samples, Formicidae, Aranae, hover flies (Syrphidae), and parasitic Hymenoptera 
were identified and counted.  
 In addition to monitoring beneficial arthropods, root-feeding bark beetles 
(Scolytinae: Hylastes spp.) were also collected in baited pitfall traps commonly used for 
root-feeding or butt feeding beetles in the pine system (adapted from Eckhardt et al. 
2007).  Pitfall traps were located 1 m from the center in six plots per treatment on both 
sites.  Each trap contained an insert with two freshly cut pine stems and two glass vials, 
one with 8 ml of 95% ethanol, and the other with 8 ml of turpentine. The interior of each 
trap was coated with liquid Teflon? (Northern Products, Woonsockets, RI) to discourage 
insect escape. Eight evenly spaced entrance holes were drilled around the pipe, as well as 
two smaller holes in the bottom for drainage. Traps were inserted so that entrance holes 
were at ground level. Traps were baited and serviced biweekly from May-Oct in 2010 
and Mar?Oct in 2011.  The number and species of beetles collected were sorted to 
species and counted.   
 Weather data were obtained for the 2010 and 2011 sample seasons from Alabama 
Mesonet Weather Data (AWIS Weather Services Inc) stations in Grand Bay and 
 
 23 
Fairhope. Mean monthly temperatures and average monthly rainfall were summarized for 
each site (AWIS 2008). 
Analyses. Counts of insects for each taxonomic group of natural enemies, total natural 
enemies (all groups collectively), and lower bole and root-feeding bark beetles and 
weevils were analyzed by fitting each insect count using a Poisson generalized linear 
mixed model (GLMM) with log-link function (PROC GLIMMIX, SAS Institute 2002).  
GLMM allows for a Poisson regression (insect counts) with the inclusion of random 
effects in the model.  A Poisson GLMM with repeated measures was fitted to the count 
data.  The random effects were sampling error and plot error.  Compound symmetry (CS) 
was used for the within-subject covariance structure.  If significant, pairwise mean 
comparisons among main effects and interactions were made using the least-squares 
means (LSMEANS) method. 
 
Results 
 
Total arthropod natural enemies (>58,000 individuals) collected over the 2 years 
of sampling were dominated by ants, spiders, and predatory beetles, with parasitic 
Hymenoptera and hover flies accounting for <1% of the total (Figure 2.3).  Cogongrass 
management treatments, including the revegetation treatment, did not significantly affect 
total counts of natural enemies, however, there were significantly (P < 0.0003) fewer 
natural enemies in burned plots (Figure 2.4) (Eckhardt 2003). Although not all groups 
revealed a significant burn response, total natural enemies, predatory beetles, ants, and 
spiders all significantly differed in site by burn interactions (Figure 2.5, 2.6, 2.7, 2.8). The 
most total natural enemies and predatory beetles were found in not burned Beebe plots. 
Ants and spiders were more abundant on not burned Rainwater plots. There was no 
 
 24 
significant treatment effect on predatory beetles or ants among cogongrass management 
treatments (Figures 2.9 and 2.10), but there were significantly more spiders in the 
untreated control plots (Figure 2.11).  Differences were also apparent at the whole plot.  
For all taxa except spiders, 2011 had significantly more arthropods collected than 2010. 
 Predatory beetles were represented by: Staphylindiae, Carabidae, Histeridae, 
Cicindelidae, and Coccinellidae. Staphylinidae and Carabidae were the most 
abundant(about 58% and 39% respectively) of the predatory beetles, followed in 
abundance by Histeridae, Cicindelidae, and then Coccinellidae, which accounted for 
about <2%. No Clerids were collected. Predatory beetle larvae (Elateridae, Cantharidae) 
were also included but accounted for <1% of the total predatory beetles. Predatory beetle 
populations had peak collections in June?July 2010, March 2011, and October 2011 
(Figure 2.9). 
 Ant captures were greatest in summer (June?Sept) of each year (Figure 2.6) with 
Solenopsis invicta Buren accounting for > 98% of ants collected. Spiders had very 
distinctive peaks in Sept 2010 and June 2011. Almost consistently through the 2011 
sample period, fewer spiders were found in the seeding plus herbicide, or eco-based, plots 
(Figure 2.11). Spiders were mostly represented by Lycosidae, Gnaphosidae, Liocranidae, 
and Ctenidae families with lycosids and gnaphosids accounting for the largest portion. 
In 31 samples across 2 years, only 164 parasitic Hymenopterans and 21 syrphid flies 
were collected: too few collected to run statistical tests for differences by treatment. Over 
two years, 381 Hylastes spp. were collected and their abundance showed no significant 
effect by treatment (Figure 2.12) or by burn treatment. 
 
 25 
Beebe site (Baldwin County) also had significantly more insects for each 
taxonomic group than the Rainwater site (Mobile County). Of all of the natural enemy 
groups, predatory beetles differed the most between the two sites; more beetles were 
found on Beebe. Beebe was also host to a larger carabid morphotype (Pasimachus sp) not 
collected on the Rainwater sites. Mean temperatures were relatively similar throughout 
both collecting seasons (Figure 2.13). Rainwater had greater levels of precipitation in 
2010 than 2011 (Figure 2.14).   
 
Discussion 
In this 2-year study, we determined how management strategies for cogongrass 
impacted all arthropod natural enemies and impacts on ants, spiders, and predatory 
beetles, which were the dominant predators in our samples. This is the first study 
investigating the effects of management strategies for cogongrass on arthropods. Overall, 
all arthropod natural enemies were not significantly affected by treatment (subplot), and 
only spiders were significantly greater in the control treatment. When compared to other 
treatments, seeding + herbicide had significantly fewer spiders. The greatest number of 
spiders were collected on control plots, which were the only ones left undisturbed. 
Seeding + herbicide plots, disturbed twice, yielded the fewest spiders and the greatest 
differences between other treatments. Complete recolonization was required on the 
seeding + herbicide plots, but on the control plots spiders present remained undisturbed.  
These findings reiterate spiders? sensitivity to disturbance (Maleque et al. 2009). Given 
that most of the spiders collected were lycosids and gnaphosids, ground dwelling instead 
of plant residing spiders, these differences may not be due directly to vegetation changes. 
 
 26 
More than half of the total collected natural enemies were fire ants. Fire ants have 
the ability to establish rapidly and take advantage of disturbed areas (Lofgren 1986). 
Since fire ants are more aggressive and have the ability to take over an ecosystem, this 
could explain the lack of significant differences between treatments. More fire ant 
mounds were observed on plots with a reduced presence of cogongrass. Since fire ants 
forage up to 37 m from their mound (Martin et al. 1998), it is plausible that ants would 
venture into nearby plots and return to the mound. If this is in fact the case, fire ant 
foraging distance would explain the lack of significance by treatment. 
The lack of change in natural enemies, predatory beetles and ants contradicts 
literature that implicates changes in arthropod communities are seen due to various 
management strategies (Evans et al. 2010, Maleque et al. 2009, Burke et al. 2008, 
Guiseppe et al. 2006, Ulyshen et al. 2004, Buckelew et al. 2000). Changes in arthropod 
communities can be linked to factors such as habitat fragmentation, vegetation change, 
herbicides and mowing (Maleque et al. 2009, Guiseppe et al. 2006).  Cogongrass is 
reported to drastically change ecosystems (Brewer 2008) and impact animal life 
(Simberloff et al. 1997), but arthropods seem to follow a different trend. Direct impacts 
of cogongrass on arthropods and soil invertebrates are being further investigated (Chapter 
Three). 
Few parasitic Hymenoptera, even fewer syrphids, and no checker beetles were 
collected during this two-year study. These groups are found at the top of the food chain 
and are easily disrupted if their prey become less abundant or absent (Maleque et al. 
2009).  
 
 27 
Whether ant populations increase or decrease after fire is dependent on species. 
(Hanula and Wade 2003, Castano-Meneses and Palacois-Vargas 2003, Jackson and Fox 
2006).  Solenopsis spp. usually survive high intensity burns and species density increases 
after burning in disturbed areas (Castano-Meneses and Palacois-Vargas 2003). Although 
spiders were not impacted by the burn treatment, abundances of predatory beetles and 
ants were greater in the unburned plots. Previous studies on predatory beetles suggest the 
impact may be family-specific.  For example, fire regimes that do not affect carabid 
populations and diversity may negatively impact staphylinids. Furthermore, these impacts 
on staphylinids may last more than 30 years post-fire (Gandi et al. 2001). Staphylinidae 
compose about 60% of the predatory beetles collected and could explain the significant 
difference among predatory beetles from the burning treatment. Further analysis is being 
conducted to determine if there is a significant difference by treatment between the two 
families. Ground and rove beetles (Carabidae and Staphylinidae) live predominantly in 
the litter layer of the soil. Cogongrass creates a dense but open thatch allowing 
arthropods and vertebrates to easily navigate. Since cogongrass quickly recovers after a 
burn, ground cover is only absent for a short period of time (Daneshgar et al. 2008).  
Site by burn interactions were inconsistent across natural enemy groups. Sites 
differences, such as more evergreen vegetation on Rainwater plots, could explain the 
inconsistencies. Burned rainwater plots consistently had a lower amount of natural 
enemies across all groups; presence of evergreen woody plants persisting after the burn 
may have influenced the natural enemy populations (S. Enloe and N. Loewenstein, 
unpublished data). Since longleaf communities are fire regulated (Brockway et al. 2005), 
the suppression of fire could allow less fire tolerant species to inhabit the community. 
 
 28 
After invasion of cogongrass and an altered fire regime, arthropod communities may 
respond differently. 
Hylastes spp., or root-feeding bark beetles, in this study were trapped at very low 
numbers. However, they were greater in burned plots but not significantly. Although not 
significant, the greatest numbers of beetles were captured on the herbicide only plots. 
Campbell et al. (2008) found that Hylastes were most abundant on herbicide plus burn 
plots in longleaf pine. Longleaf pines are typically less susceptible to decline and 
environmental as well as forest pest related stressors (Eckhardt and Menard 2009). 
Longleaf are actually planted as a substitute for loblolly pine (Pinus taeda L.) and are 
better adapted to anthropogenic disturbances (Eckhardt and Menard 2009). The low 
numbers of Hylastes spp. collected could be a reflection of longleaf pine?s ability to 
withstand attack and tolerate LPD. 
Total natural enemies and all taxonomic groups except spiders were more 
abundant in 2011.  Yearly differences can be linked to temperature, precipitation or the 
change in vegetation between 2010 and 2011. Temperatures varied more in 2011 than in 
2010 (Figure 2.6).  Rainwater (Site 1) had a greater mean precipitation in 2010 compared 
to Beebe (Site 2) (Figure 2.7). More total natural enemies were collected at Beebe.  
We found the management strategy most effective for removing cogongrass to be 
the seeding + herbicide treatment. Herbicide only plots had thinned cogongrass density, 
but cogongrass was still present. Seeded only plots revealed quick cogongrass recovery 
with limited native plant growth (S. Enloe and N. Loewenstein, unpublished data). 
Between year one and year two, the species mix was changed to the US Forest Service?s 
recommended seed mix for longleaf pine forests. After the cogongrass recovered from 
 
 29 
the burn, it overwintered and created a thatch of the prior year?s grass. At the start of the 
2011 sampling season, control plots looked the same (S. Enloe and N. Loewenstein, 
unpublished data). 
 Most cogongrass research focuses on its aggressive behavior (MacDonald 2004, 
Jose et al. 2002, Dozier et al. 1998), and its impact on forest systems (Daneshgar et al. 
2008, Ramsey et al. 2003, Lippincott 2000), and fire regimes (King and Grace 2000, 
Lippincott 2000). After two years of sampling natural enemies under four different 
treatments, only one arthropod group, spiders, exhibited a significant effect. Burning had 
a greater affect on arthropods, with total natural enemies, predatory beetles and ants all 
found in greater numbers in unburned plots. Natural enemies were not only chosen to 
serve as indicators of the community, but to monitor the root-feeding bark beetle 
populations as well. Root-feeding bark beetles, namely Hylastes, do not have known 
specialist predators but a variety of associated generalist predators (Kenis et al. 2004, 
Moeck and Safranyik 1984). Pine decline is spread by beetles being attracted to stressed 
and weakened trees (Eckhardt 2004), but we did not find a significant increase in root-
feeding bark beetles due to treatments or burn treatments. Natural enemy numbers were 
high and root-feeding bark beetles numbers were low consistently through all treatments. 
Cogongrass management strategies did not significantly impact natural enemies. Since 
Mobile and Baldwin Counties were in a state of drought the past two years, the impact of 
management strategies could be completely different in a year with average amounts of 
rainfall. As cogongrass was removed, there was a lack of understory replacement by 
native species. Eventually, cogongrass with overtake the niche and once again create a 
monotypic expanse. Given more time and better weather conditions, natural enemies may 
 
 30 
begin to resemble longleaf pine without cogongrass present on plots where it was 
removed.
 
 31 
    Table 2.1. Average soil characteristics
a
 by site 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
  
H
2
O 
availability 
  
Site Burned pH Soil type CEC % OM 
Rainwater 1 
 
Burned 5.12 Sandy Loam 0.13 0.36 3.83 
Rainwater 2 
 
Unburned 5.08 Sandy Loam 0.1 0.3 4.17 
Bebe 1 
 
Unburned 5.38 Sandy Loam 0.09 0.35 1.94 
Bebe 2 
 
Burned 5.26 Sandy Loam 0.09 0.29 1.68 
a
Soil tests conducted by AU Soil Testing Laboratory.   
Table 2.2. Sed Mix for seding and seding + herbicide plots in 2010 
Comon Name Scientific Name Family Life History Growth Form Height (ft) 
Min rot 
depth (in) 
% of 
mix 
Comon sunflower Helianthus anus Asteraceae anual Single stem 1.5-8' 8" 35 
Switchgras Panicum virgatum Poaceae perenial 
Bunch gras, 
rhizomatous 4-6' 12" 32 
Partidge Pea 
Chamaecrista 
fasciculata Fabaceae anual Bunch-type 3' 14" 15 
Purpletop Tridens Tridens flavus Poaceae perenial 
Bunch gras, 
rhizomatous 2-5' 10" 10 
Dixie ticktrefoil Desmodium tortuosum Fabaceae perenial Single crown 2-9' 6" 2 
Coreopsis Coreopsis grandiflora Asteraceae perenial Single crown 1-2' - 2 
Greyheaded 
coneflower Ratibida pinata Asteraceae perenial Single crown 2-4' 14" 2 
Black-eyed susan Rubeckia hirta Asteraceae anual/perenial Single stem 1-3' 10" 2 
 
 32 
 
 
Table 2.3. Sed Mix for seding and seding + herbicide plots 201 
Comon Name Scientific Name Family Life History Growth Form 
Height 
(ft) 
Min rot 
depth (in) 
% of 
mix 
Partridge pea Chamaecrista fasciculata Fabaceae Anual Bunch-type 2.4' 14" 34 
Indian gras Sorghastrum nutans Poaceae Perenial Graminoid 6' 24" 17 
Virginia wild rye Elymus virginicus Poaceae Perenial Graminoid 2.5' 16" 17 
Purple top tridens Tridens flavus Poaceae Perenial 
Bunch gras, 
rhizomatous 2.5' 10" 12 
Bromsedge Andropogon virginicus Poaceae Perenial 
Bunch gras, 
rhizomatous 3-5' 14" 19 
White wingstem Verbesina virginica Asteraceae Perenial Forb/herb 4-6' - 2 
Grey headed coneflower Ratibida pinata Asteraceae Perenial Single crown 4' 18" 2 
Black-eyed susan Rudbeckia hirta Asteraceae Anual, Perenial Single stem 1' 10" 2 
Perplexed tick trefoil Desmodium perplexum Fabaceae Perenial Forb/herb 4' 12" 2 
Lance-leaved coreopsis Coreopsis lanceolata Asteraceae Perenial Single crown 1.5' - 2 
Spured buterfly pea Centrosema virginianum Fabaceae Perenial Vine, Forb/herb 6-12' - 1 
 
 33 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Figure 2.1. Map of plot locations in Mobile County (Rainwater) and Baldwin County 
(Beebe). 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Figure 2.2. Pitfall traps were arranged in an X pattern 3 meters apart on each 10 x 10 m 
plot 
 
 
 34 
M a y J u n J u l A u g S e p t O c t M a r A p r M a y J u n J u l A u g S e p t O c t
2 0 1 0 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 2 0 1 1
0
100
200
300
400
500
M
e
a
n
?
A
m
o
u
n
t
C o n t r o l
H e r b
S e e d
S e e d ? + ? H e r b
 
 
 
 
 
 
 
 
 
 
 
 
 
 
     
 
Figure 2.3. Composition of the natural enemy complex in the cogongrass treatment plots 
from pitfall traps and sweep samples. Samples contained mainly these three groups with 
parasitic hymenoptera and syrphids comprising the ?other? natural enemy group. 
Percentages reflect total number of individuals collected in 31 sampling periods, 2010 
and 2011. 
           
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Figure 2.4. Natural enemies collected in each treatment per month over 31 collection trips 
in 2010 and 2011. Biweekly samples pitfall traps and sweep samples were pooled into 
monthly total. ANOVA revealed: year (ANOVA, F= 6.72, df=1,63, P=0.0119) and 
treatment (F=2.06, df=3,59, P=0.1152). Significant interaction was: year by site 
(ANOVA, F=43.41, df=1,63, p <0.0001). 
Other?
1%?
Beetles?
12%?
Ants?
59%?
Spiders?
28%?
N?=?58,274?
 
 35 
 
Figure 2.5. Abundance of total natural enemies sampled from May 2010-Oct 2011 from 
pitfall traps and sweep samples. Main effect of site (ANOVA, F=10.36, df=1,59, 
P=0.0021), burn treatment (ANOVA, F=14.69, df=1,59, P=0.0003), and site by burn 
treatment (ANOVA, F=28.67, df=1,59, P=<0.0001) were significant. 
 
 
 
 
 
Figure 2.6. Abundance of predatory beetles sampled from May 2010-Oct 2011 from 
pitfall traps and sweep samples. Main effect of site, burn (ANOVA, F =29.46, df=1,59, 
P= <0.0001), and site by burn interaction (ANOVA, F =26.20, df=1,59, P =<0.0001) 
were significant.  
 
 36 
 
Figure 2.7. Abundance of ants sampled from May 2010-Oct 2011 from pitfall traps and 
sweep samples. Main effect of site (ANOVA, F=0.06, df=1, 59, P=0.8093), burn 
treatment (ANOVA, F =10.89, df=1,59, P = 0.0016), and site by burn treatment 
(ANOVA, F =10.51, df=1,59, P = 0.0020).  
 
 
 
Figure 2.8. Abundance of spiders sampled from May 2010-Oct 2011 from pitfall traps 
and sweep samples. Main effect of site (ANOVA, F =11.45, df=1,59, P =0.0013), burn 
treatment (ANOVA, F=0.11, df=1, 59, P=0.7398), and site by burn treatment (ANOVA, 
F =28.84, df=1,59, P =<0.0001). 
 
 
 
 37 
M a y J u n e J u ly A u g S e p t O c t M a r A p r M a y J u n e J u ly A u g S e p t O c t
2 0 1 0 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 2 0 1 1
0
10
20
30
M
e
a
n
?
?
?
S
E
C o n t r o l
H e r b ic id e
S e e d in g
S e e d + H e r b
M a y J u n e J u ly A u g S e p t O c t M a r A p r M a y J u n e J u ly A u g S e p t O c t
2 0 1 0 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 2 0 1 1
0
10
20
30
40
50
60
70
M
e
a
n
?
?
?
S
E
C o n t r o l
H e r b ic id e
S e e d in g
S e e d + H e r b
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Figure 2.9. Predatory beetles collected in each treatment over 31 collection trips in 2010 
and 2011. ANOVA revealed: year (ANOVA, F =37.96, df=1,63, P =<0.0001), and 
treatment (ANOVA, F=1.75, df=3,59, P=0.1670). Significant interaction was: site by 
treatment (ANOVA, F value=4.25, df=3,59, P = 0.0087).   
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Figure 2.10. Ants collected in each treatment over 31 collection trips in 2010 and 2011. 
ANOVA revealed: year (ANOVA, F =6.39, df=1,63, P= 0.0140), and treatment 
(ANOVA, F=1.94, df=3,59, P=0.1334). Significant interaction was: year by site 
(ANOVA, F=24.52, df=1,63, P = <0.0001). 
 
 38 
M a y J u n J u l A u g S e p t O c t M a r A p r M a y J u n J u l A u g S e p t O c t
2 0 1 0 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 2 0 1 1
0
10
20
30
40
50
60
T
o
t
a
l
?
H
y
l
a
s
t
e
s
?
s
p
p
.
C o n t r o l
H e r b ic d e
S e e d in g
H e r b + S e e d
M a y J u n e J u ly A u g S e p t O c t M a r A p r M a y J u n e J u ly A u g S e p t O c t
2 0 1 0 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 2 0 1 1
0
10
20
30
40
M
e
a
n
?
?
?
S
E
C o n t r o l
H e r b ic id e
S e e d in g
S e e d + H e r b
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Figure 2.11.  Spiders collected in each treatment over 31 collection trips in 2010 and 
2011. ANOVA revealed: year (ANOVA, F=1.21, df=1,63, P=0.2758), and treatment 
(ANOVA; F =4.11; df=1,63; P =0.0102). Significant interaction was: year by site 
(ANOVA, F =46.01, df=1,63, P =<0.0001). 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Figure 2.12. Total Hylastes spp. collected in each treatment over 31 collection trips in 
2010 and 2011. Biweekly samples from baited pitfall traps were pooled into monthly 
totals. ANOVA revealed: year (ANOVA, F=6.38, df=1,63, P=0.0141), site (ANOVA, 
F=0.81, df=1,59, P=3730), burn treatment (ANOVA, F=0.14, df=1,59, P=0.7072), and 
treatment (ANOVA, F =0.43, df=3,59, P =0.7317). Significant interaction was seen in 
year by site (ANOVA, F =4.18, df=1,63, P =0.0452). 
 
 
 
 39 
 
 
Figure 2.13. Mean monthly temperatures in 2010 and 2011 for Baldwin County (Beebe) 
and Mobile County (Rainwater). (AWIS 2008) 
 
 
 
 
Figure 2.14. Mean monthly precipitation in 2010 and 2011 for Baldwin County (Beebe) 
and Mobile County (Rainwater). (AWIS 2008) 
 
 
 
0?
5?
10?
15?
20?
25?
30?
35?
May?June? July? Aug? Sept? Oct? Mar? Apr? May?June? July? Aug? Sept?
Mean?Temperature?(?C)?
Beebe?
Rainwater?
0?
5?
10?
15?
20?
25?
30?
35?
May?June? July? Aug? Sept? Oct? Mar? Apr? May?June? July? Aug? Sept?
Mean?Precipitation?(cm)?
Beebe?
Rainwater?
 
 40 
Chapter 3 
Impact of cogongrass on diversity and abundance on arthropods and soil dwelling 
microfauna in longleaf pine 
 
Abstract 
 Cogongrass (Imperata cylindrica Beav.) is an invasive C
4
 grass currently 
invading the southeastern coastal plains of Alabama and its surrounding states. To assess 
the impact cogongrass has on arthropod and soil dwelling invertebrates, we used pitfall 
traps, sweep samples, soil purging and soil samples. Although cogongrass does impact 
arthropod diversity, it is necessary to evaluate each individual taxa to understand the 
impact it has on arthropod communities. Seasonally, Gryllidae, Curculionidae, 
Scarabaeidae and Araneae were all significantly impacted by cogongrass presence. Most 
herbivore groups were significantly more abundant in longleaf pine with cogongrass 
present. Dung beetles and their relatives arose as excellent bioindicators of cogongrass 
presence in longleaf pine. Earthworms yielded significantly greater numbers in 
cogongrass present plots by both abundance and weight. Nematodes reacted differently 
by functional group: plant parasitic nematodes were significantly more abundant in 
cogongrass present plots while free-living nematodes revealed the opposite. 
 
 
 
 
 
 41 
Introduction 
The introduction of aggressive, non-indigenous species poses one of the greatest 
threats to ecosystems worldwide (Capo-chichi et al. 2008, Mack et al. 2000. Cogongrass 
(Imperata cylindrica), an aggressive, rhizomatous grass indigenous to Southeast Asia, 
was introduced to South Alabama around 1910 (MacDonald 2004, Dozier et al. 1998). 
Cogongrass? ability to hinder nutrient uptake and retention, interfere with respiration, and 
create monotypic expanses makes it especially threatening to ecosystems (MacDonald 
2004). It is ranked as one of the ten worst weeds worldwide (Holm et al. 1977), 
fundamentally because of its impacts on agriculture, forestry, and fire regimes 
(Lippincott 2000, Dozier et al. 1998). It is a threat to forest systems because even after 
burning, cogongrass remains dominant, prevents forest succession, and displaces native 
plants (Lippincott 2000). It possesses several strategies that promote its survival such as 
drought tolerance, fire adaptability, adaption to poor soils, a widespread rhizome 
network, and prolific seed production (MacDonald 2004, Dozier et al. 1998). 
Longleaf pine savannas are particularly vulnerable to cogongrass invasion. These 
fire dependent savannas previously stretched from Virginia to Texas but due to 
anthropogenic activity are now isolated to a few areas (Brockway et al. 2005). They are 
home to a variety of rare and endangered plant and animal species sensitive to 
disturbance (Brewer 2008). Game animals such as white-tailed deer, turkey and northern 
bobwhite as well as rare, non-game animals including red-cockaded woodpecker, eastern 
indigo snake, gopher tortoise, and fox squirrel all take advantage of longleaf pines for a 
stable habitat. Longleaf ecosystems are among the most species-rich plant communities 
 
 42 
outside the tropics due to the rich diversity of understory plants found there (Brockway et 
al. 2005). 
 Measuring biodiversity in changing ecosystems is critical for determining health 
conservation needs. Forests support the greatest global biodiversity in terrestrial 
ecosystems, therefore conserving forest biodiversity should be an important objective for 
forest management (Maleque et al. 2009). Entire inventories of forest biodiversity, even 
if possible, are very time-consuming and expensive (Maleque et al. 2009). As an 
alternative to complete inventories, bioindicators are used to monitor ecosystems and 
biodiversity because they are typically functionally important to the ecosystem and are 
determined by disturbance regimes (Maleque et al. 2009, Langor and Spence 2006). 
Arthropods are frequently used as bioindicators because they have short generation times, 
have high species richness, and are sensitive to environmental changes and habitat 
disturbance. They also play critical roles in an ecosystem as herbivores, parasites, 
predators, decomposers, pollinators and seed dispersers (Maleque et al. 2009, Maleque et 
al. 2006).  
Gerber et al. (2007) proposed a link between replacement of native plant species 
by exotic species and the overall reduction of abundance and richness of arthropods. 
Invaded areas supported lower numbers of native plants as well as a lowered abundance 
and richness of invertebrates when compared to uninvaded areas (Gerber et al. 2007). In 
the grasslands of Indonesia, braconid (Hymenoptera: Braconidae) diversity declined in 
cogongrass dominated areas. As grasslands returned to Acacia forests, parasitoid 
abundance and species richness increased (Maeto et al. 2009). 
 
 43 
Soil dwelling organisms provide a series of ecosystem services, including 
decomposition of organic matter, nutrient cycling, and bioturbation (Brussaard et al. 
1997, Syers and Springett 1984). Earthworm diversity is determined by soil type, nutrient 
content, organic matter, and disturbance more than by plant diversity (Brussaard et al. 
1997). Their body tissues and cast materials are enriched with nutrients and their ability 
to ingest organic material increases rate of nutrient cycling (Syers and Springett 1984). 
Earthworm abundance can be influenced by fertilizers, physical disruption to soils, and 
changes of soil nutrients (Kladivko 2001). In some forests, earthworm communities have 
high species richness and diversity and a relatively high biomass (Blanchart and Julka 
1997).  Although earthworms can serve as beneficial species, many earthworms are 
invasive in forests. They can have significant impacts on plant communities and soil 
processes in native ecosystems (Hendrix and Bohlen 2002). 
Nematode species composition is effected by soil disturbance, erosion and water 
quality (Brussaard et al. 1997). Nematodes have a very limited range and exist in the top 
few centimeters of soil. Changes in temperature and moisture, as well food supply, pore 
disruption and moisture conditions after forest disturbance all influence nematode 
richness and diversity. After soil disturbance, nematodes must cope with many of these 
changes while the soil profile is reassembled (Ettema 1998). Forest nematodes are not 
frequently studied because of the volume of unknown species and amount of resources 
required (Bloemers et al. 1997). The objective of this study was to better understand the 
impact of cogongrass invasion in longleaf pine on arthropods and soil dwelling 
microfauna. 
 
 44 
 
 
 
Materials and Methods 
Site Description. Site one, Rainwater, is in Mobile County, and site two, Beebe, is in 
Baldwin County, each with 10 plots of well-established cogongrass in longleaf pine per 
site. On original Rainwater and Beebe sites, cogongrass was too prevalent for cogongrass 
absent plots to be established. Additional plots were established nearby on land with 
similar geographic traits. Both sites consisted of sandy loam soils, which is typical for 
this region (Table 3.1). 
Experimental Design. Plots of cogongrass present and cogongrass absent sites were 
established in longleaf pine stands in south Alabama to understand the impacts of 
cogongrass invasion on arthropod communities of Southeastern pine forests. On each of 
the four sites, ten 10 x 10 m plots were established containing 5-8 trees each with a DBH 
ranging between of 11.4 and 54.6 cm, and a height ranging from 6.22 to 25.27 m.  
Arthropod Sampling. Arthropods were sampled using two methods: pitfall traps for 
ground active arthropods, and sweep samples primarily targeting arthropods associated 
with vegetation.  Five pitfall traps (15 cm dia funnel type traps, Multi-pher, Bio-Controle, 
Quebec City, Quebec), were installed in each plot, arranged in an X pattern 3 m apart 
along each diagonal with one in the center (Figure 2.2).  Inside each trap, a 453 ml plastic 
cup (16K, Dart, Mason, MI) was filled with 50 ml of a 1:1 (v/v) mixture of water and 
propylene glycol (Peak, Old World Industries, Northbook, IL) to serve as a preserving 
 
 45 
solution.  Every 2 wk, contents from all five pitfall traps were pooled into a labeled 
plastic container.  Empty traps were then recharged with preserving solution.  Pitfall traps 
were active from May 1 through Oct 28, 2010. Pitfall traps remained empty in the field 
overwinter.  In 2011, traps were activated on Feb 15 and serviced biweekly through Oct 
25. Soil displaced from pitfall trap installation was collected and brought back to the AU 
soil dynamics lab for analysis. 
 Sweep samples were taken biweekly, concurrent with servicing pitfall traps.  One 
sample was made per plot, following each diagonal making approximately 30?40 sweeps 
per plot.  Insects collected in the sweep net were put into a 3.8 liter plastic zipper bags 
and placed into a cooler for transport back to the laboratory.  
Inside the laboratory, arthropods from pitfall traps contents were filtered to 
remove dirt and other debris.  The remaining arthropods were stored in 50 ml plastic vials 
with 90% ethanol in cooler until they were sorted.  Arthropods were sorted into 
taxonomic groups (insects and other arthropods) and stored in 22 ml vials with 90% 
ethanol. From these samples, specimens were sorted to family. Sweep samples were 
placed in the freezer for ?1 d. Insects were then removed from the bags and placed into 
15 ml vials by plot in 90% ethanol. Like the arthropods collected from pitfall traps, all 
specimens that could be identified to family were, and those that could not be were given 
to Dr. Charles Ray, Insect Diagnostic Lab, Auburn University, to be classified. 
Earthworm Sampling. Earthworm samples were collected twice (spring and fall 2011) 
from two, 0.5 m
2
 samples in each plot. Vegetation or leaf litter was manually cleared 
from each area and a plastic frame (40.6 cm wide, 30.5 cm tall) defining the area placed 
 
 46 
and inserted slightly (about 2.5 cm) into the soil surface (Figure 3.1).  An expellant 
solution of mustard seed powder (allyl isothiocyanate) in water (100 mg per liter) was 
prepared (Zaborski 2003), and 2 liters of that solution were applied to each frame in 
successive pourings (Lawrence and Bowers 2002).  All earthworms that emerged within 
10 minutes of pouring were collected into small plastic bottles with moist paper towels 
for transport to the lab.  In the lab, the number and mass of worms from each plot were 
determined (Zaborski 2003, Lawrence and Bowers 2002). 
Nematode Sampling.  Soil samples of nematodes were collected in late spring 2011. 
Each plot was sampled in a systematic pattern using a cone soil probe with a 1.9 cm 
opening.  Five uniform soil samples were taken, composited into one sample per plot and 
stored in separate one-quart plastic ziploc bags. In the AU Nematology Lab, individual 
soil samples were mixed thoroughly, and 100 cm
3
 of soil was extracted from each 
sample.  Nematodes were extracted using gravity screening and centrifugal flotation 
methods. Numbers of nematodes present in each plot were identified to genus level and 
nematodes will be identified to trophic level (plant parasitic or free-living).  All nematode 
samples were quantified under an inverted microscope (Nikon T-100). 
Analyses. Counts of arthropods exceeding 900 within each family were analyzed by 
fitting each insect count using a Poisson  distribution generalized linear mixed model 
(GLMM) with log-link function (PROC GLIMMIX, SAS Institute 2002).  GLMM allows 
for a Poisson regression (insect counts) with the inclusion of random effects in the model.  
A Poisson GLMM with repeated measures was fitted to the count data.  The random 
effects were sampling error and plot error.  Compound symmetry (CS) was used for the 
 
 47 
within-subject covariance structure.  If significant, pairwise mean comparisons among 
main effects and interactions were made using the least-squares means (LSMEANS) 
method.  Non-parametric analysis (Wilcoxon Rank Sum Test) was conducted to compare 
insect families between cogon and no cogon treatments (SAS 2005).   
Earthworm weight was analyzed using a split-plot experimental design, where site 
(RW and BB) and treatment (cogon and no cogon) was sub-plot factor.  Earthworm count 
data were analyzed by fitting each counts using Poisson generalized linear mixed model 
(GLMM) with log-link function.  The random effects were site (RW and BB) and 
sampling error.  Data analysis was conducted with SAS software using generalized linear 
mixed model procedure (PROC GLIMMIX, SAS Institute 2002).  If significant, pairwise 
mean comparisons among main effects and interactions were made using the least-
squares means (LSMEANS) method. 
Diversity of nematode genera and orders in cogon and no cogon fields in different 
sites (RW and BB) were examined by the following metrics: Shannon Diversity, species 
richness, species number, species evenness and ratio of plant parasitic/free-living 
nematodes.  All statistical tests for diversity were performed in R using (R Development 
Core Team 2011).  Counts of nematodes were analyzed by fitting each nematode count 
using Poisson generalized linear mixed model (GLMM) with log-link function (PROC 
GLIMMIX, SAS Institute 2002).  GLMM allows for a Poisson regression (nematode 
counts) with the inclusion of random effects in the model.  A Poisson GLMM with 
repeated measures was fitted to the count data.   
 
 
 48 
Results 
Over 40,000 arthropods in over 100 families were recorded from eight months of 
sampling in longleaf pine. Coenagrionidae (Odonata), Nabidae (Hemiptera), 
Hippoboscidae, Tachinidae, Ulididae (all Diptera), Sphecidae (Hymenoptera) were only 
found in longleaf pine with cogongrass present. Gryllotalpidae (Orthoptera), Achilidae, 
Aradidae, Caliscelidae, Tingidae (all Hemiptera), Attelibidae, Bostrichidae, Buprestidae, 
Cicindellidae (all Coleoptera), Ceratopogonidae, Cloropidae, Sarcophagidae, Simuliidae 
(all Diptera), Scoliidae and Trichogrammatidae (both Hymenoptera) were only recorded 
in longleaf pine with cogongrass absent (Table 3.2). After identifying and counting all 
arthropods to the family level, it was found that only 8 groups revealed differences across 
the season (Table 3.3). Only Gryllidae (crickets), Scarabaeidae (scarabs), and Aranae 
(spiders) were significantly more in abundant in longleaf pine without cogongrass while 
Curculionidae (weevils) were significantly more abundanc in longleaf pine with 
cogongrass.  
The Wilcoxan rank-sum test revealed 30 taxa to be significantly impacted by 
cogongrass presence (Table 3.2). First, parasitic hymenopterans (Ichneumonidae and 
Bracondiae) and chalcids (Chalcoidea) were more abundant in longleaf pine with 
cogongrass present. Herbivores such as katydids (Tettigoniidae), grasshoppers 
(Acrididae), walking sticks (Phasmatodea), seed bugs (Alydidae), spittle bugs 
(Cercopidae), jumping plant lice (Psyllidae), leaf beetles (Chrysomelidae), gall midges 
(Cecidomyiidae), and caterpillars (Lepidoptera) were generally twice as abundant in 
longleaf pine with cogongrass present. A few other herbivores, leaf hoppers 
 
 49 
(Cicadellidae), plant hoppers (Issidae), and tree hoppers (Membracidae) revealed the 
opposite. Three members of the Scarabaeoidae superfamily, scarabs (Scarabaeidae), 
earth-boring scarab beetles (Geotrupidae), and hide beetles (Trogidae), were all found to 
be at least seven times more abundant in longleaf pine with cogongrass absent than those 
with cogongrass present. Lastly, springtails (Collembola) were three times more abundant 
in longleaf pine with cogongrass present. 
 Earthworm weight (Figure 3.3) and abundance (Figure 3.4) were significantly 
greater in longleaf with cogongrass present. Similarly, for total nematodes, total plant 
parasitic and Mesocriconema spp, nematodes were significantly greater in longleaf pine 
with cogongrass present (Table 3.4). In contrast, free-living nematodes were significantly 
more abundant in longleaf pine without cogongrass. Diversity and evenness are fairly low 
in longleaf pine with and without cogongrass. More species were recorded in longleaf 
pine with cogongrass present than absent. Also, longleaf pine with cogongrass present 
had a greater ratio of plant parasitic nematodes than cogongrass absent where the ratio 
was equal.  
 
Discussion    
In this 8 month study, we compared abundance and diversity of arthropod and soil 
microfauna in longleaf pine. The only group of natural enemies to show a significant 
seasonal impact due to cogongrass presence was spiders (Table 3.3). Spiders are 
extremely sensitive to disturbance (Maleque et al. 2009), especially soil disturbance and 
litter type (Pearce and Venier 2006). Other families included Gryllidae, Scarabaeidae and 
 
 50 
Curculionidae. A study conducted on wood crickets (Nemobius spp) suggested open, 
wooded edges were their preferred habitat (Brouwers and Newton 2009). Conservation of 
forests and reducing edge effect is critical for habitat preservation. 
Those beetles in the Scaraboidea super family are important nutrient recyclers, 
soil aerators, seed dispersers and indicators in the forest ecosystem (Maleque et al. 2009, 
Vulinec 2000). They feed on dung, carrion, fungi and decaying leaves (Maleque et al. 
2009). They are especially susceptible to ecosystem change with respect to soil 
characteristics. Degraded soils are not conducive to large scarab populations (Vulinec 
2000) and beetle size and diversity increase in intact forests (Scheffler 2005). Scarab 
distribution between longleaf pine with cogongrass present compared to longleaf pine 
with cogongrass absent reveals the greatest difference in all the families surveyed (Table 
3.2). With almost 3,000 scarabs collected in longleaf pine with cogongrass absent, 
evidence of community change due to cogongrass is overwhelming. Scarabs were 
collected abundantly throughout the season in longleaf pine with cogongrass absent. At 
least eight different morphotypes were observed consistently in longleaf pine with 
cogongrass absent, reducing of possibility of one species population exploding and 
skewing abundance results. Also, there were no carcasses of vertebrates detected in the 
vicinity of pitfall traps, which could have biased captures in adjacent plots.  
Parasitic Hymenoptera were found significantly more abundant in longleaf with 
cogongrass present. These findings contradict research (Maeto et al. 2009) where 
cogongrass presence in forest systems reduced braconid richness and diversity. Braconid 
larvae are parasitoids on larvae of beetles, flies, butterflies and moths, as well as aphids 
 
 51 
and Heteroptera (Matthews 1974). Larvae of click beetles, weevils, leaf beetles, flies, 
Lepidopterans, and a few Heteropteran families were more abundant in longleaf pine 
with cogongrass present (Table 3.3) 
Generalist herbivores may be in greater abundance in longleaf pine with 
cogongrass because of the greater amount of biomass available compared to longleaf pine 
with cogongrass absent. Chewing marks, like those left from grasshoppers or katydids, 
were observed on cogongrass leaf blades. Spittle bugs (Cercopidae), a family of grass-
feeding Hemipterans (Borror and White 1970), were also significantly more abundant in 
longleaf pine with cogongrass. Spittle bug immatures were observed on leaf bases of 
cogongrass on multiple occasions. Jumping plant lice (Psocoptera) were twenty times 
more abundant in longleaf pine with cogongrass present. Since they are monophagous or 
oligophagous plant feeders, it could be suggested that they are feeding on cogongrass, but 
further research would be needed to confirm since other plants were present (Enloe and 
Loewenstein unpublished data). Weevils were the only family significantly greater in 
longleaf pine with cogongrass present over the collecting season and total (Table 3.2 and 
3.3). Most of the weevils collected were billbugs (Sphenophorous spp.). Billbugs are 
obligate grass feeders with adults feeding and larvae developing as a stem borer. It is 
possible that billbugs are using the dense root mass of cogongrass as a nest although this 
behavior was not observed. 
Earthworm abundance and weight were more significant in longleaf pine with 
cogongrass present. Because earthworms are active decomposers of the litter layer, a link 
between cogongrass thatch and earthworm presence is possible. Cogongrass? ability to 
 
 52 
alter the nutrient availability of the soil (Jose et al. 2002) could explain the vast difference 
between the cogongrass presence and absence. If the presence of cogongrass seriously 
alters the earthworm community, the soil and soil dwelling arthropod community could 
experience a cascading effect. Increased abundance of earthworms could lead to a spike 
in small mammals or ground-active birds that prey on earthworms. In fact, multiple small 
mice were accidental victims to pitfall traps, but only in longleaf pine with cogongrass 
present. 
 Nematodes are a major component of all soil food webs and soil disturbance 
affects species composition (Brussaard et al. 1997). Samples revealed seven different 
species of free-living nematodes and 3 orders of free-living nematodes (Table 3.4). Plant 
parasitic nematodes were significantly more abundant in longleaf pine with cogongrass 
present and only Mesocriconema spp. were significantly impacted by cogongrass 
presence. Free-living nematodes had an inverse response, having greater numbers in 
longleaf pine with cogongrass absent. This leads us to conclude free-living nematodes are 
more sensitive to cogongrass invasion than plant parasitic one. The ratio indicates plant 
parasitic nematodes composed 70% of collected species in longleaf pine with cogongrass 
present. In longleaf pine without cogongrass present, the ratio was approximately 1:1, 
leading us to believe the absence of cogongrass provides a more suitable habitat for both 
types of nematodes. Overall, diversity and evenness were fairly low in longleaf pine with 
and without cogongrass even though they are significantly different. Without a pristine 
habitat for comparison, it is hard to compare 
 
 53 
 Cogongrass presence significantly impacted arthropod abundance in 
approximately 30 taxa. Scarabs, the most diverse family collected, present a hard to 
ignore indication of disturbance due to cogongrass. Significant differences in these 
arthropod groups highlight the necessity for more research. In cogongrass systems, 
especially those within longleaf pine, scarabs, geotrupid beetles, trogid beetles, carabids, 
and crickets could all serve as indicators of disturbance by invasion. These taxa should be 
of interest in other systems where cogongrass invades. Herbivores such as grasshoppers, 
katydids, seed bugs, spittle bugs, psyllids, and leaf beetles present the potential for a 
biological control that might have been overlooked in previous studies. Earthworms seem 
to take advantage of the altered soil regime found in longleaf pine with cogongrass 
present. Nematodes, however, are more sensitive to cogongrass based on functional 
group, with plant parasitic nematodes significantly more abundant in longleaf pine with 
cogongrass present. Cogongrass does impact arthropod diversity and for future studies it 
is necessary to investigate each individual taxa to understand the overall impact 
cogongrass has on an entire arthropod community. These taxa present opportunities for 
further investigation. Although families and subfamilies share similar life histories, 
perhaps species within these larger taxonomic groups maybe found as specific indicators 
of disturbance for cogongrass. 
 
 
 54 
 
Table 3.1. Average soil characteristics for cogon(C) 
and no cogon (NC) plots for objective 2.
a
 
Site pH Soil Type CEC 
RainwaterC 4.97 Sandy < 4.6 cmol
c
kg
-2
 
RainwaterNC 4.83 Sandy < 4.6 cmol
c
kg
-2
 
BebeC 5.39 Sandy < 4.6 cmol
c
kg
-2
 
BebeNC 5.41 Sandy < 4.6 cmol
c
kg
-2
 
a
Soil tests conducted by AU Soil Testing Laboratory. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 55 
Table 3.2. Comparison of arthropods collected over entire season in longleaf with and 
without cogongrass present by Wilcoxon In cogongrass present plots, 18,968 arthropods 
were collected and counted. In cogongrass absent plots, 20,182 arthropods were collected 
and counted. Both were collected over an 8 month period in 2011. 
 Total  
Arthropod Group Cogon No cogon P 
Odonata 
Coenagrionidae 6 0 0.020 
Orthopteroid 
Rhaphidiophoridae 509 560 0.457 
Grylidae 338 1022 <0.001 
Grylotalpidae 0 2 0.342 
Tetrigidae 4 7 0.624 
Tetigonidae 212 41 <0.001 
Acrididae 123 32 <0.001 
Mantidae 1 4 0.164 
Phasmatidae 15 6 0.026 
Blatodea 111 86 0.108 
Hemiptera 
Achilidae 0 1 0.342 
Acanalonidae 3 6 0.270 
Aphididae 39 57 0.835 
Alydidae 117 69 0.05 
Aradidae 0 6 0.040 
Berytidae 6 12 0.325 
Caliscelidae 0 1 0.342 
Cercopidae 26 16 0.012 
Cicadelidae 115 207 0.04 
Cixidae 15 22 0.24 
Coreidae 5 3 0.706 
Cydnidae 30 29 0.505 
Delphacidae 4 9 0.426 
Dictyopharidae 7 12 0.945 
Flatidae 1 3 0.54 
Geocoridae 1 0 0.342 
Isidae 2 19 0.09 
Lygaeidae 5 7 0.508 
Membracidae 9 35 0.05 
Miridae 2 8 0.21 
Nabidae 3 0 0.163 
Pentatomidae 36 23 0.589 
Psylidae 435 21 <0.001 
Reduvidae 53 32 0.036 
Tingidae 0 1 0.342 
 
 
 56 
Coleoptera 
Atelibidae 0 1 0.342 
Bostrichidae 3 2 0.979 
Buprestidae 0 1 0.342 
Cantheridae 7 5 0.307 
Carabidae 768 1416 0.017 
Cerambycidae 2 6 0.573 
Chrysomelidae 238 68 <0.001 
Cicindelidae 0 1 0.342 
Cocinelidae 2 3 0.654 
Curculionidae 1425 141 <0.001 
Elateridae 835 338 0.01 
Erotylidae 2 0 0.162 
Geotrupidae 18 754 <0.001 
Histeridae 4 6 0.758 
Lampyridae 3 0 0.080 
Mordelidae 8 19 0.091 
Nitidulidae 460 495 0.978 
Scarabaeidae 368 2898 <0.001 
Staphylindidae 982 848 0.409 
Tenebrionidae 723 332 0.14 
Trogidae 5 179 
<0.001 
Neuroptera 
Predatory Larvae
a 
15 12 0.569 
Diptera 
Asilidae 26 8 0.37 
Bibionidae 1 1 1.00 
Caliphoridae 3 5 0.652 
Cecidomyidae 49 9 0.01 
Ceratopogonidae 0 1 0.342 
Chironomidae 223 139 0.015 
Cloropidae 0 2 0.162 
Culicidae 10 10 0.716 
Dolichipodidae 58 16 <0.001 
Drosophilidae 30 3 0.04 
Empididae 4 2 0.615 
Hipoboscidae 1 0 0.342 
Muscidae 228 145 0.069 
Phoridae 5 13 0.130 
Pipunculidae 1 1 1.00 
Sarcophagidae 0 1 0.342 
Sciaridae 451 344 0.15 
Simulidae 0 2 0.162 
Tabanidae 1 3 0.310 
Tachinidae 1 0 0.342 
 
 57 
Tephritidae 41 26 0.131 
Tipulidae 10 8 0.904 
Ulididae 1 0 0.342 
Hymenoptera 
Apidae 1 7 0.040 
Chalcoidea 72 22 <0.001 
Evanidae 3 1 0.54 
Formicidae 5525 5680 0.636 
Halictidae 4 12 0.219 
Mutilidae 84 55 0.924 
Parasitic Hymenoptera 67 44 0.052 
Pompilidae 7 5 0.508 
Scolidae 0 5 0.342 
Sphecidae 3 0 0.163 
Tenthredinidae 16 1 0.54 
Trichogramatidae 0 1 0.342 
Vespidae 3 8 0.141 
Other Orders 
Acari 10 2 0.197 
Aranae 2905 3429 0.10 
Colembola 913 238 <0.001 
isoptera 40 16 0.194 
Lepidoptera 60 31 0.07 
Thysanoptera 10 2 0.030 
a
Insects in the families Ascalaphidae and Myrmeliontidae. 
b
Insects in the families Ichneumonidae and Braconidae. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 58 
 
 
Table 3.3.    Comparison of populations of selected arthropods 
using all observations from each sample collected in pitfalls and 
sweep samples in longleaf pine with cogongrass present and 
cogongrass absent, Bay Minette and Grand Bay, AL, 2011, using 
repeated measures analyses 
Effect Arthropod group F  df P 
Treatment Gryllidae 23.65 1, 38 <0.0001 
 Carabidae 1.72 1, 38 0.1982 
 Curculionidae 10.48 1, 38 0.0025 
 Nitidulidae 0.09 1, 38 0.763 
 Scarabaeidae 18.45 1, 38 <0.0001 
 Staphylinidae 0.88 1, 38 0.3544 
 Formicidae 1.48 1, 38 0.2318 
 Araneae 5.14 1, 38 0.0292 
Time Gryllidae 3.76 17, 646 <0.0001 
 Carabidae 8.74 17, 646 <.00001 
 Curculionidae 2.65 17, 646 0.0003 
 Nitidulidae 12.12 17, 646 <0.0001 
 Scarabaeidae 2.54 17, 646 0.0006 
 Staphylinidae 5.87 17, 646 <0.0001 
 Formicidae 13.73 17, 646 <0.0001 
 Araneae 18.99 17, 646 <0.0001 
Interaction Gryllidae 1.75  17, 646 0.031 
 Carabidae 4.01 17, 646 <0.0001 
 Curculionidae 1.53 17, 646 0.0768 
 Nitidulidae 2.52 17, 646 0.0007 
 Scarabaeidae 1.68 17, 646 0.0413 
 Staphylinidae 6.62 17, 646 <0.0001 
 Formicidae 3.74 17, 646 <0.0001 
 Araneae 2.65 17, 646 0.0003 
 
 
 
 
 
 
 
 
 
 
 
 
 
 59 
 
Table 3.4. Mean abundance of different nematode taxa 
Total nematodes, plant parasitic (PP), and free-living (FL) are represented. 
If P <0.05, difference is noted between cogongrass present (C) and  
cogongrass absent (N). Repeated measures test used. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Taxon 
Cogon 
Present 
Cogon 
Absent 
(P <0.05) 
Difference
 
 
Total 212.4 196.3 C > N 
Total PP 159.7 107.5 C > N 
        Helicotylenchus 5.8 28.3  . 
        Tylenchulus 79.8 54.7  . 
        Paratylenchulus 12.2 1.3  . 
        Meloidogyne 3.9 3.9  . 
        Mesocriconema 46.4 3.9 C > N 
        Hoplolaimus 11.6 0.0  . 
        Tylodorus 0.0 15.5  . 
Total FL 52.8 88.8 C < N 
        Rhabditida 39.3 77.3 . 
        Desmodorida 13.5 3.9 
. 
        Enoplida 0.0 7.7 
. 
Shannon Diversity 0.8 0.5 
. 
Species Richness 1.5 1.1 
. 
Species Number 2.5 1.9 
. 
Species Evenness 0.6 0.5 
. 
Ratio: PP/(PP+FL) 0.7 0.5 
. 
 
 
 60 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Figure 3.1. Plastic frame with mustard solution applied for earthworm extraction. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Figure 3.2. Pitfall traps were arranged in an X pattern 3 meters apart on each 10 x 10 m 
plot 
 
 
 
 61 
C o g o n N o ? C o g o n
0
100
200
300
400
500
M
e
a
n
?
?
?
S
E
C o g o n N o ? C o g o n
0
1
2
3
4
5
6
7
M
e
a
n
?
?
?
S
E
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Figure 3.3 Earthworm abundance in 2011 on cogongrass present and cogongrass absent 
plots by abundance. Year was not significant, therefore samples were combined. 
ANOVA: F= 8.00, df=1, 37, P=0.0075.  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Figure 3.4 Earthworms collected in 2011 on cogongrass present and cogongrass absent 
plots by weight (mg). Year was not significant, therefore samples were combined.  
ANOVA F=7.57, df=1, 37, P=0.0091. 
 
 
 
 62 
 
 
References 
 
Alabama Forestry Commission. 2009.
 http://www.forestry.alabama.gov/forest_facts.aspx 
 
AWIS Weather Service Inc. 2008. http://www.awis.com/mesonet/index.html 
 
Bartholomew, C.S., Prowell, D., and Griswold, T. 2006. An annotated checklist of
 bees (Hymenoptera: Apoidea) in longleaf pine savannas of southern Louisiana
 and Mississippi. Journal of the Kansas Entomological Society. 79: 184-198. 
 
Bauman, Tessa Annette. 2002. ?Interaction of Fire and Insects in the Restoration and
 Management of Longleaf Pines.? MS thesis. Louisiana State University. Print. 
 
Bell, J. R., Haughton, A. J., Boatman, N. D., and Wilcox, A. 2002. Do incremental
 increases of the herbicide glyphosate have indirect consequences for spider
 communities? Journal of Arachnology. 30:288-297. 
 
Blanchart, E. and Jukla, J. M. 1997. Influence of forest disturbance on earthworm
 (Oligochaeta) communities in the western Ghats (south India). Soil Biology and
 Biochemistry. 29:303-30. 
 
 
 63 
Bloemers, G. F., Hodda, M.,  Lambshead, P. J. D., Lawton, J. H., and Wanless, F. R.
 1997. The effects of forest disturbance on diversity of tropical soil nematodes.
 Oecologia. 111:575-582. 
 
Bohac, J. 1999. Staphylinid beetles as bioindicators. Agriculture, ecosystems, and
 Environment 74: 357-372. 
 
Borror and White. 1970. Insects. Boston: Houghton Mifflin. 
 
Boyer, W. D. 1990. In Burns, Russell M., and Barbara H. Honkala, tech coods. Silvics of
 North America: Vol 1. Conifers; Agriculture Handbook 654. U.S. Department of
 Agriculture, Forest Service, Washington, DC. 
 
Brewer, S. 2008. Declines in plant species richness and endemic plant species in longleaf
 pine savannas invaded by Imperata cylindrica. Biological Invasions 10:1256-
1264. 
 
Brewer, J.S., and Cralle, S.P. 2003. Phosphorus addition reduces invasion of a longleaf
 pine savanna (Southeastern USA) by a non-indigenous grass (Imperata
 cylindrica). Plant Ecology. 167: 237-245 
 
Brockway, D.G., Outcalt, K.W., Tomczak, D.J., and Johnson, E.E. Restoration of
 Longleaf Pine Ecosystems. 
 
 64 
 
Brouwers, N. C. and Newton, A. C. 2009. Habitat requirements for the conservation of
 wood cricket (Nemobius sylverstris) (Orthoptera: Gryllidae) on the Isle of Wright,
 UK. Journal of Insect Conservation. 13:529-541. 
 
Brussaard, L., Behan-Pelletier, V., Bignell, D., Brown, V., Didden, W., Folgarait, P.,
 Fragoso, C., Freckman, D., Gupta, V., Hattori, T., Hawksworth, D., Kopatek,
 C., Lavelle, P., Malloch, D., Rusek, J., So?destroo?m, B., Tiedje, J. and
 Virginia, R. 1997. Biodiversity and ecosystem function in soil. Ambio. 26: 563
 570. 
 
Bryson, C.T. and Carter, R. 2004. Biology of Pathways for Invasive Weeds. Weed
 Technology. Volume 18: 1216-1220 
 
Buckelew, L.D., Pedigo, L.P., Mero, H.M., Owen, M.D.K., and Tylka, G.L. 2000.
 Effects of weed amnagement systems on canopy insects in herbicide-resistant
 soybeans. Journal of Entomologic Entomology. 93: 1437-1443. 
 
Burke, J. D., Chamberlain, J., Geaghan, J. P. 2008. Effects of understory vegetation
 management on brood habitat for northern bobwhites. Journal of Wildlife
 Management 72: 1361-1368. 
 
 
 65 
Campbell, J. W., Hanula, J. L., and Outcault, K. W. 2008. Effects of prescribed fire
 and other plant community restoration treatments on tree mortality, bark beetles,
 and other saproxylic Coleoptera of longleaf pine, Pinus palustris Mill., on the
 coastal plain of Alabama. Forest Ecology and Management. 354:134-144. 
 
Capo-chichi, L.J.A., Faircloth, W.H., Williamson, A.G., Patterson, M.G., Miller,
 J.H., and van Santen, E. 2008. Invasion dynamics and genotypic diversity of
 cogongrass (Imperata cylindrica). Invasive Plant Science and Management. 1:
 133-141. 
 
Castano-Meneses G. and Palacois-Vargas J. 2003. Effects of fire and agricultural
 practices on neotropical ant communities. Biodiversity and Conservation. 12:
 1913-1919. 
 
Charpentier, L. J., McCormick, and W. J., Mathes, R. 1967. Beneficial arthropods
 inhabiting sugarcane fields and their effects on borer infestations. Sugar Bulletin.
 45: 276-277. 
 
Coile and Schilling. 1993. Cogongrass, Imperata cylindrica (L.) Beauv.: A good grass
 gone bad! Botany Circular 28:1-3. 
 
 
 66 
Cram, M. M. and Fraedrich, S. W. 2007. Detection and management of stunt and
 stubby-root nematodes in a southern forest nursery. USDA Forest Service
 Proceedings RMRS-P-50. 
 
Daneshager, P. and Jose, S. 2009. Imperata cylindrical, an alien invasive grass,
 maintains control over nitrogen availability in an establishing pine forest. Plant
 Soil. 320: 209-218. 
 
Daneshgar, P., Jose, S., Collins, A., and Ramsey, C. 2008. Cogongrass (Imperata
 cylindrica), an alien invasive grass, reduces survival and productivity of an
 establishing pine forest. Forest Science. 54: 579-586. 
 
Dickens, R. 1974. Cogongrass in Alabama after sixty years. Weed Science 22:177-179 
 
Dozier, H., Gaffney, J. H., McDonald, S. K., Johnson, E.R.R.L., and Shilling, D. G.
 1998. Cogongrass in the United States: history, ecology, impacts, and
 management. Weed Technology. 12:737-743. 
 
Duever, L. C. 2008. Observations from the Cogongrass Conference: Implications for
 research, management and control. Wildland Weeds. 5-10. 
 
 
 67 
Eckhardt, L, G. ?2003. Biology and Ecology of Leptographium Species and Their
 Vectors as Components of Loblolly Pine Decline.? Diss. Louisiana State
 University. Print. 
 
Eckhardt, L. G., Goyer, R. A., Klepzig, K. D., and Jones, J. P. 2004. Interactions of
 Hylastes species (Coleoptera: Scolytidae) with Leptographium species associated
 with loblolly pine decline. J. Econ. Entomol. 97:468-474. 
 
Eckhardt, L. G. and Menard, R. D. 2009. Declining loblolly pine stands: symptoms,
 causes, and management options. Alabama?s Treasured Forests. 10-12. 
 
Eckhardt, L. G., Sayer, M. A. S., and Imm, D. W. 2008. What is pine decline in the
 southeastern United States and how widespread is it? Ecological Society of
 America. Draft. 
 
Eckhardt, L. G., Weber, A. M., Menard, R. D., Jones, J. P. 2007. Insect-fungal
 complex associated with loblolly pine decline in central Alabama. Forest Science
 53:84-92. 
 
Ettema, C. 1998. Soil nematode diversity: species coexistence and ecosystem function.
 Journal of Nematology. 30:159-169. 
 
 
 68 
Evans, S. C., Shaw, E. M., and Rypstra, A. L.  2010. Exposure to a glyphosate-based
 herbicide affects agrobiont predatory arthropod behaviour and long-term survival.
 Ecotoxicology 19:1249-1257. 
 
Gandhi, K. J. K., Spence, J. R., Langor, D. W., and Morgantini, L. E. 2001. Fire
 residuals as habitat reserves for epigaeic beetles (Coleoptera: Carabidae and
 Staphylinidae). Biological Conservation 102: 131-141. 
 
Gerber, E., Krebs, C., Murrell, C., Moretti, M., Rocklin, R., and Schaffner, U. 2008.
 Exotic invasive knotweeds (Fallopia spp.) negatively affect native plant and
 invertebrate assemblages in European riparian habitats. Biological Conservation.
 141:646-654. 
 
Glitzenstein, J. S., Platt, W. J., and Streng, D. R. 1995. Effects of fire regime and
 habitat on tree dynamics in north Florida longleaf pine savannas. Ecological
 Monographs. 65:441-476. 
 
Guiseppe, K. F. L., Drummond, F. A., Stubbs, C., and Woods, S. 2006. The use of
 glyphosate herbicide in managed forest ecosystems and their effects on non-target
 organisms with particular reference to ants as bioindicators. Technical Bulletin
 192. Maine Agricultural and Forest Experimental Station, University of Maine. 
 
 
 69 
Hanula, J. L. and Wade, D. D. 2002. Influence of long-term dormant-season burning
 and fire exclusion on ground-dwelling arthropod populations in longleaf pine
 flatwoods ecosystems. Forest and Ecology Management. 175:163-184. 
 
Haywood, J. D. 2009. Eight years of seasonal burning and herbicidal brush control
 influence sapling longleaf pine growth, understory vegetation and the outcome of
 an ensuing wildfire. Forest Ecology and Management. 258:295-305. 
 
Hendrix, P. F. and Bohlen, P. J. 2002. Exotic Earthworm Invasions in North America:
 Ecological and Policy Implications. American Institute of Biological Sciences. 
 801-811. 
 
Holland, J. M. and Luff, M. L. 2000. The effects of agricultural practices on Carabidae
 in temperate agroecosystems. Integrated Pest Management Review. 5:109-129. 
 
H?lldobler, B and Wilson, E. O. 1990. The Ants. Cambridge, MA: Belknap. 
 
Holm, L.G., Pucknett, D.L., Pancho, J.B., and Herberger, J.P. 1977. The Worlds
 worst weeds. Distribution and Biology. Univ Press of Hawaii, Honolulu, HI.
 609p. 
 
 
 70 
Humphrey, J. W., Hawes, C., Peace, A. J., Ferris-Kaan, R., and Jukes, M. R. 1999.
 Relationships between insect diversity and habitat characteristics in plantation
 forests. Forest Ecology and Management. 113:11-21. 
 
Isaacs, R., Tuell, J., Fiedler, A., Gardiner, M., and Landis, D. 2009. Maximizing
 arthropod-mediated ecosystem services in agricultural landscapes: the role of
 native plants. Front Ecological Environment. 7:196-203. 
 
Jackson, G. P. And Fox, B. J. 1996. Comparison of regeneration following burning,
 clearing or mineral sand mining at Tomago, NSW: II. Succession of ant
 assemblages in a coastal forest. Australian Journal of Ecology, 21: 200?216. 
 
Jose, S., Cox, J., Miller, D.L., Shilling, and D.G., Merritt, S. 2002. Alien plant
 invasions: The story of cogongrass in southeastern forests. J. Forestry 100:41-44. 
 
Kenis, M., Wemelinger, B. and Gregoire, J.-C. 2004 Research on parasitoids and
 predators of Scolytidae-a review. In: Lieutier, F., Day, K. R., Battisti A.,
 Gregoire, J.-C., and Evansm H. F. (eds). Bark and wood boring insects in living
 trees in Europe, a synthesis. Kluwer Academic Publishers, Dordrecht. 237-290. 
 
King, S. E. and Grace, J. B. 2000. The effects of gap size and disturbance type on
 invasion of wet pine savanna by cogongrass, Imperata cylindrica. American
 Journal of Botany. 87:1279-1286. 
 
 71 
 
Kladivko, E. J. 2001. Tillage systems and soil ecology. Soil and Tillage Research 61:
 61-72. 
 
Langor, D. W. and Spence, J. R. 2006. Arthropods as ecological indicators of
 sustainability in Canadian forests. The Forestry Chronicle. 82:344- 
 
Lawrence, A.P. and Bowers M.A. 2002. A test of the ?hot? mustard extraction method
 of sampling earthworms. Soil Biochem., 34:549-552. 
 
Lippincott, C.L. 2000. Effects of Imperata cylindrica (L.) Beauv. (Cogongrass) Invasion
 on Fire Regime in Florida Sandhill (USA). Natural Areas Journal. 20: 140-149. 
 
Lofgren, L. 1986. History of imported fire ants in the United States. In C. Lofgren and O
 Vander Meer (Eds.), Fire ants and leaf-cutting ants: Biology and management.
 Boulder: Westview Press. 
 
MacDonald, G.E. 2004. Cogongrass (Imperata cylindrica)-Biology Ecology and
 Management. Critical Reviews in Plant Sciences 23(5):367-380. 
 
MacGown, J. A. and Forster, J.A. 2005. A preliminary list of the ants (Hymenoptera:
 Formicidae) of Alabama, U.S.A. Entomological News 116: 61-74 
 
 
 72 
Mack, R. N., Simberloff, D., Lonsdale, W. M., Evans, H., Clout, M., and Bazzaz, F.
 A. 2000. Biotic invasions: Causes, epidemiology, global consequences and
 control. Ecol. Appl. 10:689-710. 
 
Maeto, K., Noerdijito, W. A., Belokobyskij, S. A., and Fukuyama, K. 2009. Recovery
 of species diversity and composition of braconid parasitic wasps after
 reforestation of degraded grasslands in lowland East Kalimantan. Journal of
 Insect Conservation. 13:245-257. 
 
Maleque, M.A., Maeto, K., and Ishii, H.T. 2009. Arthropods as bioindicators of
 sustainable forest management, with a focus on plantation forests. Appl. Entomol.
 Zool. 44:1-11. 
 
Maleque, M.A., Ishii, H.T., and Maeto, K. 2006. The use of arthropods as indicators of
 ecosystem integrity in forest management. Journal of Forestry. 104:113-117. 
 
Martin, J. B., B. M. Drees, W. E. Grant, E. K. Pedersen, C. L. Barr, and S. B.
 Vinson. 1998. Foraging range of the polygynous form of the red imported fire
 ant, Solenopsis invicta Buren. Southwest. Entomol. 23: 221-228. 
 
Massey, C. L. 1974. Biology and taxonomy of nematode parasites and associates of bark
 beetles in the United States.  Agriculture Handbook. 446. Washington, DC: US
 Department of Agriculture, Forest Service. 
 
 73 
 
Matthews, R. W. Biology of Braconidae. Annual Review of Entomology. 19:15-32. 
 
Miller, W. E. 1979. Fire as an insect management tool. Bulletin of the Entomological
 Socitey of America. 25:137-140. 
 
Minno, M. C. and Minno, M. 1999. Survey of herbivorous arthropods and pathogens of
 cogongrass in Florida. Final Report to D. G. Shilling, University of Florida, Jay,
 Florida, and J. Meeker, Florida Department of Agriculture and Consumer
 Services. Gainesville, Florida, USA. 
 
Moeck H. A., Safranyik L. 1984. Assessment of predator and parasitoid control of bark
 beetles. Environment Canada. Canadian Forest Service, Pacific Forest Research
 Centre, Information Report Bc-X-248. 
 
Moulder, B. C. and Reichle, D. E.  1972. Significance of Spider Predation in the energy
 dynamics of forest-floor arthropod communities. Ecological Monographs. 42:473
 498. 
 
Otrosina, W. J., Hess, N. J., Zarnoch, S. J., Perry, T. J., and Perry, J. P. 1997. Blue
 stain fungi associated with roots of southern pine trees attacked by the southern
 pine beetle, Dendroctonus frontalis. Plant Dis. 81:942-945. 
 
 
 74 
Outcault, K.W. and Sheffield, R.M. 1996. The longleaf pine forest: Trends and current
 conditions. SRS-9, USDA Forest Service Southern Research Station, 10p. 
 
Pearce, J. L. and Venier, L. A. 2006. The use of ground beetles (Coleoptera: Carabidae)
 and spiders (Araneae) as bioindicators of sustainable forest management: A
 review. Ecological Indicators. 6:780-793. 
 
R Development Core Team. 2011. R: A language and environment for statistical
 computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3
 900051-07-0, URL http://www.R-project.org/. 
 
Ramsey, C.L., Jose, S., Miller, D.L., Cox, J., Portier, K.M., Shilling, D.G., and
 Merritt, S. 2003. Cogongrass (Imperata cylindrica) response to herbicides and
 disking on a cutover site and in a mid-rotation pine plantation in southern USA.
 Forest Ecol. Management. 179:195-209. 
 
Rainio, J. and Niemela, J. 2003. Ground beetles (Coleoptera: Carabidae) as
 bioindicators. Biodiversity and Conservation 12:487-506. 
 
Reagan, T. E. 1982. Sugarcane borer pest management in Louisiana: Leading to a more
 permanent system. Pro. II Inter-American Sugarcane Seminar-Insect and Rodent
 Pests-1981, Florida International University. 2:100-110. 
 
 
 75 
Reagan, T. E. 1986. Beneficial aspects of the imported fire ant: a field ecology approach.
 In C. Lofgren and O Vander Meer (Eds.), Fire ants and leaf-cutting ants: Biology
 and management. Boulder: Westview Press. 
 
Reynolds, J. W. 1994. Earthworms of Alabama (Oligochaeta: Acanthodrilidae,
 Eudrilidae, Lumbricidae, Megascolecidae, Ocnerodrilidae, Sparganophilidae).
 Megadrilogica. 
 
Ruehle, J. L. 1973. Influence of plant-parasitic nematodes on longleaf pine seedlings.
 Journal of Nematology. 5:7-9. 
 
SAS 2005. SAS Users Guide. Statistics. Version 9.1 ed Cary, NC: SAS institute 
 
Sajise, P. E. 1972. Evaluation of cogon [Imperata cylindrica (L.) Beav.] as a seral stage
 in Philippine vegetational seccession. I. The cogonal seral stage and plant
 succession. II. Autecological studies on cogon. Ph. D. dissertation, Cornell
 University, Ithaca, New York, USA. 
 
Scheffler, P. Y. 2005. Dung beetle (Coleoptera: Scarabaeidae) diversity and community
 structure across three disturbance regimes in eastern Amazonia. Journal of
 Tropical Ecology. 21:9-19. 
 
 
 76 
Senseman, S. A. 2007. Herbicide handbook 9
th
 Edition. Lawrence, KS Weed Science
 Society of America. 
 
Simberloff, D., Schmitz, D. C., and Brown, T. C. 1997. Strangers in paradise: impact
 and management of nonindigenous species in Florida. Island press. 291. 
 
Syers, J. K. and Springett, J. A. 1984. Earthworms and soil fertility. Plant and Soil.
 76:93-104. 
 
Tschinkel, W.  1986. The ecological nature of the fire ant: some aspects of the colony
 function and some unanswered questions. In C. Lofgren and O Vander Meer
 (Eds.), Fire ants and leaf-cutting ants: Biology and management. Boulder:
 Westview Press. 
 
Ulyshen, M. D.,  Hanula, J. L., Horn, S., Kilgo, J. C., Moorman, C. E. 2004. Spatial
 and temporal patterns of beetles associated with coarse woody debris in managed
 bottomland hardwood forests. Forest and Ecology Management 199:259-272. 
 
Van Loan, A.N., Meeker, J.R., and Minno, M.C. 2002. Cogongrass. Chapter 23 in
 Biological Control of invasive plants in the eastern United States. Tech.
 Coordinators, Van Drieshe, R., Lyon, S., Blossey, B., Hoddle, M., and Reardon,
 R. USDA Forest Service Publication FHTET-2002-04, 413 p. 
 
 
 77 
Vulinec, K. 2000.  Dung beetles (Coleoptera: Scarabaeidae), monkeys, and conservation
 in Amazonia. 
 
Werner, M. R. 1990. Earthworm ecology and sustaining agriculture. University of
 California Sustainable Agriculture Research & Education Program 
 
White, R. E. 1983. A field guide to the beetles of North America. Houghton Miffin,
 Boston, Massachusetts, USA. 
 
Wharton, D. A. 1986. A functional biology of nematodes. Baltimore: John Hopkins
 Press. 
 
Wingfield, M.J., Capretti, P. and MacKenzie, M. 1988. Leptographium spp. as
 pathogens of conifers, an international perspective. In: Leptographium Root
 Diseases on Conifers. (Harrington TC, Codd FW eds) American
 Phytopathological Society, USA:113-128. 
 
Yager, L.Y., Miller, D.L., and Jones, J. 2011. Woody Shrubs as a Barrier to Invasion
 by Cogongrass (Imperata cylindrica). Invasive Plant Science and Management. 4:
 207-211. 
 
Youngs, L. C. 1983. Predaceous ants in biological control of insect pests of North
 American Forests. Bulletin of ESA. 29:47-50. 
 
 
 78 
Zaborski, E.R. 2003. Allyl isothiocyanate: an alternative chemical expellent for
 sampling earthworms. Appl. Soil Ecol. 22:87-95.