Citation
Riparian Zone Management in Coastal Plain Streams

Material Information

Title:
Riparian Zone Management in Coastal Plain Streams Multi-Scale Effects of Habitat Fragmentation
Creator:
Griswold, Marcus
Place of Publication:
[Gainesville, Fla.]
Florida
Publisher:
University of Florida
Publication Date:
Language:
english
Physical Description:
1 online resource (206 p.)

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Environmental Engineering Sciences
Committee Chair:
Crisman, Thomas L.
Committee Members:
Holt, Robert D.
Wise, William R.
Bolker, Benjamin M.
Graduation Date:
8/9/2008

Subjects

Subjects / Keywords:
Drought ( jstor )
Ecology ( jstor )
Fresh water ( jstor )
Invertebrates ( jstor )
Landscapes ( jstor )
Macroinvertebrates ( jstor )
Macrophytes ( jstor )
Species ( jstor )
Streams ( jstor )
Watersheds ( jstor )
Environmental Engineering Sciences -- Dissertations, Academic -- UF
bmp, drought, forestry, logging, macroinvertebrates, quality, riparian, water
City of Tallahassee ( local )
Genre:
bibliography ( marcgt )
theses ( marcgt )
government publication (state, provincial, terriorial, dependent) ( marcgt )
born-digital ( sobekcm )
Electronic Thesis or Dissertation
Environmental Engineering Sciences thesis, Ph.D.

Notes

Abstract:
Riparian zones act as filters for nutrients and sediment, and provide food and habitat for terrestrial and aquatic organisms. Preserving riparian structure in headwater streams is critical to protecting local and downstream aquatic biota. Forestry practices along streams are capable of degrading riparian zone function, leading to increased sediment and nutrient inputs, limiting organic matter availability, and altering light and temperature levels in streams. The effects of forestry practices on aquatic invertebrate communities were evaluated in coastal plain streams by experimentally manipulating two harvest regimes in headwater streams based on Georgia?s best management practices. Though the primary goal of the study was to relate anthropogenic disturbances to water quality, a drought occurring prior to the study created degraded streams with low invertebrate abundance and diversity. The drought resulted in streambeds with large amounts of stored organic matter and nutrients, that became available with re-wetting. A core set of species appeared immediately following drought in the streams, reflecting a shared species pool. These species shared resilient traits, including short life cycles and resistance to desiccation, which allowed for rapid recovery from disturbance. However, temporal shifts in biological traits reflected a more stable hydrologic regime over time. As communities recovered, a shift occurred from individuals that were small, sclerotized, and abundant in drift, to those that were larger, soft-bodied, and rare in drift, indicating that the habitat was more favorable. Thus, such shifts in trait structure and the role of natural disturbances need to be accounted for when bioassessment programs are implemented. To evaluate the effects of logging on streams, macroinvertebrates were sampled in reference and harvest streams before and after an experimental harvest. In response to harvest, communities shifted from detritivores to herbivores, following a shift in the food source from organic matter to algae and macrophytes. This change was most apparent in the thinned SMZ, where chlorophyll a was 50-100% higher than in the intact SMZ and reference streams. In general, changes in community structure were most apparent the first year following the harvest and began to follow a trajectory of recovery over the next four years. Interestingly, multimetric indices of water quality based on macroinvertebrates suggested more favorable conditions in the most disturbed treatment (Thinned SMZ). This relates to increases in food quality, due to an increase in algae and macrophytes, and a decrease in C:N ratios in terrestrially derived leaves. However, invertebrates in the thinned SMZ were represented by species preferring to live in sand, highlighting the increased isolation of patches apparent in these reaches. Observational and experimental field work was used to determine the effects of altered habitat amount and type on macroinvertebrate colonization and movement patterns. Macrophyte patches were more complex, stable, and trapped higher quantities of organic matter; attracting more diverse invertebrate communities than leaf packs. Patch size was a determinant of community structure for habitat specialists, where shredders were more common in large leaf packs and scrapers more common in large macrophytes. This reflected the higher biomass of chlorophyll a in macrophytes and bacteria in leaf packs. However, in general, invertebrate abundance and taxon richness decreased with increasing patch size. This indicates unfavorable conditions in larger patches, not evidenced by changes in dissolved oxygen or canopy cover. Behavioral experiments utilizing a habitat specialist (Trichoptera: Anisocentropus) and generalist (Pleuroceridae: Elimia) invertebrate indicated that the availability of both macrophytes and leaf packs is preferred by both groups and decreases emigration rates from landscapes. Thus, the increased diversity of habitats created by harvest potentially balanced the effects of habitat fragmentation and isolation. This was achieved by the addition of stable macrophyte patches that provided habitat islands, decreasing isolation and storing organic matter. Deforestation potentially has detrimental consequences for aquatic an terrestrial organisms, fragmenting in-stream habitat and the terrestrial landscape. Evidence from this study indicates that properly managed riparian zones effectively maintain water quality in small coastal plain streams. However, community composition shifted in the thinned SMZ and did not completely recover within five years. This suggests that harvest within the established riparian zone may eliminate habitat specialists, especially those consuming leaf litter, while replacing them with opportunistic species. Although water quality was preserved, managers should consider the consequences of reducing habitat specialists and its potential effects on food-web structure. ( en )
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis:
Thesis (Ph.D.)--University of Florida, 2008.
Local:
Adviser: Crisman, Thomas L.
Statement of Responsibility:
by Marcus Griswold.

Record Information

Source Institution:
UFRGP
Rights Management:
Copyright Griswold, Marcus. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Classification:
LD1780 2008 ( lcc )

Downloads

This item has the following downloads:


Full Text





and mass (F2,372 = 5.9, P<0.01). More CPOM was trapped in the 4 g patches than the 1 and 2 g

patches (Fig. 4-11). The amount of CPOM trapped in patches ranged from 0.2 to 1.7 grams. The

most CPOM was trapped in Ludwigia and the least in Pinus (Fig. 4-12). Additionally, more

CPOM was trapped in patches that were collected after fifteen days and were not disturbed (Fig.

4-13). FPOM trapped within patches was related to leaf species (F3,406 = 40.9, P<0.0001),

disturbance (F2,406 = 5.7, P<0.01), and mass (F2,406 = 11.1, P<0.0001). More FPOM was trapped

over time and with increasing patch size (Fig. 4-14).The amount of FPOM ranged from 0.05 to

0.3 g and was greatest in Ludwigia and least in Pinus and Quercus (Fig. 4-15). Oxygen within

patches was not different between any treatment.

Invertebrate abundance changed significantly between leaf species (F3,402 = 5.3, P<0.01)

and initial leaf mass (F2,402 = 12.6, P<0.001). Abundance was lowest in Ludwigia with an

average of 15 individuals and highest in Pinus and Liriodendron with an average of 30

individuals (Fig. 4-16). Invertebrate abundance increased with increasing patch size, from 20 to

44 individuals (Fig. 4-17). However, there was no apparent evidence for a relationship between

patch size and expected species richness whe sample size was accounted for. Taxon richness

changed significantly between leaf species (F3,402 = 6.6, P<0.001) and initial leaf mass (F2,402

19.5, P<0.001). In general, the number oftaxa did not differ greatly, averaging between 3 and 5,

with Quercus patches having the least number of taxa (Fig. 4-18).

The proportion of predators did not differ between treatments. The proportion of scrapers

was dependent on initial leaf mass (F2,402 = 4.3, P<0.02) and was higher in the 4 g than 1 g

patches (Fig. 4-19). The proportion of shredders changed in response to mass (F2,402 = 3.1,

P<0.05), leaf type (F3,402 = 3.4, P<0.02), and disturbance (F2,402 = 5.6, P<0.01), however, the

effect of leaf type depended on disturbance treatment (F6,402 = 2.5, P<0.03). In general,









1.0 Harv



So A2
SA DO ** o3
S o *4
Flow TN
SNH4 Turb
A sc
0.0 A
A A U
A / pH

Co A A LF p
CL LF





-1.0- U










-2.0
-1.0 0.0 1.0
Axis 1



Figure 3-8. NMDS of taxonomic composition in watersheds A and B in pre-harvest (1) and in
post-harvest reference (2), thinned SMZs (3), and intact SMZ treatments (4).









Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

RIPARIAN ZONE MANAGEMENT IN COASTAL PLAIN STREAMS: MULTI-SCALE
EFFECTS OF HABITAT FRAGMENTATION

By

Marcus Wayne Griswold

August 2008

Chair: Thomas Crisman
Major: Environmental Engineering Sciences

Riparian zones filter nutrients, sediment, and provide food and habitat for terrestrial and

aquatic organisms. Georgia's forestry practices were evaluated in coastal plain streams by

manipulating harvest regimes in headwater streams. Macroinvertebrate and their food sources

were sampled before and after harvest.

A drought occurring prior to the study degraded streams, depressing invertebrate

abundance and diversity. A core set of species appeared immediately following drought,

displaying short life cycles and resistance to desiccation, allowing for rapid recovery from

disturbance. Communities shifted from small, sclerotized individuals abundant in drift, to those

that were larger, soft-bodied, and rare in drift, indicating more favorable habitat.

In response to harvest, communities shifted from detritivores to herbivores, following

shifts in food availability from organic matter to algae and macrophytes. This was most apparent

immediately following harvest and followed a trajectory of recovery over the next four years.

Interestingly, multimetric indices of water quality based on macroinvertebrates suggested more

favorable conditions in the most disturbed treatment. This relates to increases in food quality,

due to an increase in algae and macrophytes, and a decrease in C:N ratios in terrestrially derived


































2008 Marcus Wayne Griswold









between streams. For instance, many water quality indices were initially derived to

understand downstream effects of points sources such as sewage (Kolkwitz and Marsson,

1909). However, the challenge to create indices that respond to non-point sources as well

as multiple stressors has brought this approach to the extent of its limits. In these indices,

lower values reflect poor water quality (e.g., pollution tolerant organisms), while high

values indicate good water quality (e.g., pollution sensitive species). However, the

Florida SCI was not responsive to the harvest treatments, suggesting the harvest streams

had better water quality than the reference streams two years after the harvest. The SCI

was highly responsive to natural disturbances, and values increased from poor water

quality to excellent water quality as streams responded to restoration of flow and

precipitation following the 1998-2002 drought..

A strong relationship existed between SCI scores and both flow and dissolved

oxygen, the primary factors responsible for recovery of invertebrate communities

following drought (Chapter 2). Harvest created a diverse range of microhabitats (e.g.,

light and temperature patches), likely providing more niches for other species.

Additionally, discharge was greater in the selective harvest treatment, which may have

buffered these streams from any drying over the course of the study. Lastly, as

periphyton levels increased in the selective harvest treatment, increased Ephemeroptera

abundance drove the SCI scores higher since this group tends to be a good indicator of

water quality.

The SCI has not been able to differentiate between reference and disturbed

streams in other cases. Vowell (2001) did not find evidence that the SCI was able to

discriminate between reference and logged sites in Florida. In a survey of 167 headwater










Doledec, S., B. Statzner, and M. Bournard. 1999. Species traits for future biomonitoring across
ecoregions: patterns along a human-impacted river. Freshwater Biology 42:737-758.

Doledec, S. and B. Statzner. 2008. Invertebrate traits for the biomonitoring of large
European rivers: an assessment of specific types of human impact. Freshwater Biology
53:617-634.

Douglas, M. and P. S. Lake. 1994. Species Richness of Stream Stones: An Investigation of the
Mechanisms Generating the Species-Area Relationship. Oikos 69:387-396.

Downes,B.J., P. S. Lake, E. S. G. Schreiber, and A. Glaister. 1998. Habitat structure and
regulation of local species diversity in a stony, upland stream. Ecological Monographs 68:237-
257.

Downes,B.J., P. S. Lake, E. S. G. Schreiber, and A. Glaister. 2000. Habitat structure, resources
and diversity: the separate effects of surface roughness and macroalgae on stream invertebrates.
Oecologia 123:569-581.

Drew, C.A. and D.B. Eggleston. 2006. Currents, landscape structure, and recruitment success
along a passive-active dispersal gradient. Landscape Ecology 21:917-931.

Dudley, T.L. 1988. The role of plant complexity and epiphyton in colonization of macrophytes
by stream insects. Verhandlungen der Internationalen Vereinigung
fur Theoretische und Angewandte Limnologie 23:1153-1158.

Dufrene, M. and P. Legendre. 1997. Species assemblages and indicator species: the need for a
flexible asymmetrical approach. Ecological Monographs 67:345-366.

Dunning, J. B., B. J. Danielson, and H. R. Pulliam. 1992. Ecological processes that affect
populations in complex landscapes. Oikos 65:169-175.

Eckman, J. E. 1990. A model of passive settlement by planktonic larvae onto bottoms of
differing roughness. Limnology and Oceanography 35:887-901.

Elliot, J.M. 1971. Upstream movements of benthic invertebrates in a Lake District stream,
Journal of Animal Ecology 40:235-252.

Elliott, J.M. 2002. Time spent in the drift by downstream-dispersing invertebrates in a Lake
District stream. Freshwater Biology 47:97-106.

Elliott, J.M. 2003. A comparative study of the dispersal of 10 species of stream invertebrates.
Freshwater Biology 48:1652-1668.









Invertebrates in the reference streams shared traits indicative of an undisturbed

forested headwater coastal plain stream. In general, species were soft-bodied and bluff,

indicating lower flow and less scouring. When left undisturbed, streams in this region

have riparian zones that limit high peak flows during storm events in these streams. For

example, soft-bodied tipulid larvae have limited ability to resist scouring and are easily

washed downstream. Thus, species with this trait are adapted to low-gradient,

undisturbed streams. Additionally, species in the reference streams were more likely to

prefer living in plant material derived from the riparian zone and thus are closely linked

to the riparian zone. Thus, biological traits were accurate predictors of functional

changes occurring in watersheds following a logging disturbance.

The utility of using biological traits and fuzzy coding for linking trophic habits

to disturbance lies in the catholic food preferences of many invertebrates. Species

historically thought to be shredders supplement their diet with algae, especially when this

food source becomes dominant (Zah et al., 2001). This flexibility in food choice may

limit the ability of bioassessment protocols to detect disturbance. However, many species

rely heavily on a primary food source, and little is known of their reproductive capacity

when faced with a less preferred food choice. Additionally, traits are stable over

interannual periods, allowing for more flexible sampling protocols (Snook and Milner,

2002). Thus, analysis of trophic habitat, combined with fuzzy coding, which allows

species to be assigned to multiple groups, will ultimately enhance the robustness of

sampling programs.

Anthropogenic disturbance in the face of natural disturbances

Water quality indices derived from ecological and taxonomic information on

aquatic invertebrates should be responsive to a gradient of disturbances within and































S 0 o00
OA
: .-


0*


Axis 1


h6 r3
tr. mh4
.v3
htl arlp slmh2 tr2 tr3
h .sh2 d
mh3* *ec2
mlh6
mh "6
d1
ti
**


hi


v2

r2
dfl mh7

ar3


ecl
sl mh5

tr4
rs2 *.shl
ar2
ht2
ht3 ds2 *
ds3 *s3 rs
s2 *vl


Axis 1


Figure 3-11. NMDS of biological traits in watersheds C and D in pre-harvest (1) and in post-

harvest reference (2), thinned SMZs (3), and intact SMZ treatments (4).















103


Harv

"1
A2
03
*4


10.











10-
.2 0 0







-1 0







-20


*

A A
A
U


Harv

.1
A2
*3
*4


r4

3


I I I I I I I









decrease in shredders and scrapers with increased drought duration, suggesting a

predictable cyclic change in feeding habits with drought.

After a major disturbance, body size is hypothesized to increase as communities

stabilize. Accordingly, medium-sized species became more common as the community

stabilized, while smaller species were more common during the less-stable time periods.

Small body size is often related to shorter-life cycles, and might therefore serve as a

resilience trait. However, small body size may also allow for exploitation of refugia such

as the hyporheic zone during droughts or floods (Townsend, 1989), serving as a potential

source of colonizers (Dole-Olivier, Marmonier and Beffy, 1997). Thus, small body size

may also be useful for resisting impacts of disturbance (e.g., Townsend, Doledec and

Scarsbrook, 1997).

Organisms with longer life cycles require more stable habitat and water

chemistry. Adaptations to high variability were less common by the third year of

sampling, when species commonly had uni- or semivoltine life cycles and lacked

adaptations for resisting desiccation. Additionally, hardening of the exoskeleton reduces

mortality during periods of drying and floods. Sclerotized and hard-shelled organisms

were more common during the first two years after drought, while traits favoring soft-

bodied organisms became more common with time. Although there was a general trend

toward species lacking resistance to desiccation, many species in the streams appeared to

be adapted to some level of drying through a desiccation-resistant or diapause stage.

Although many traits responded predictably to the drought, streamlined

individuals were more common immediately following drought, reflecting the abundance

of adult dytiscid beetles and amphipods as early colonizers. Dytiscid beetles colonize













2500
0 Ludwigia
MMixed
S2000 Liriodendron
2000



O 1500
.U)

0 1000



500 -
z

0
10 20 30
Percent Cover


Figure 5-5. Average net squared displacement (+ SE) of Anisocentropus in microlandscapes.









cylinder with water to determine the volume of the patch by water displaced. Each portion

(CPOM, FPOM, and macrophytes) was then dried at 600C for at least 48 hours and weighed to

compare dry weights to volume for a given area. FPOM samples were ashed at 550C for five

hours to determine organic content.

Since many invertebrates consume bacteria and periphyton, chlorophyll a, ash-free dry

weight (AFDW), and bacteria counts were used as indicators of patch quality. Chlorophyll a and

AFDW were analyzed as in Chapter 3. Leaves for chlorophyll and AFDW measurement were

vigorously shaken in 50 mL of deionized water for 30s, after which leaves were removed to

measure surface area. Water samples were filtered through 45 [im GFF filters, and bacteria on

the filters were stained with SYBR Green and counted under an epiflourescent microscope.

Leaves were photographed, and Scion Image (Scion Corp., Frederick, MD, U.S.A.) was used to

calculate total surface area.

Bacteria enumeration followed the protocol outlined in Buesing (2005). A 0.2 [im,

25mm aluminum oxide membrane filter (Whatman Anodisc) was placed on top of a wetted 0.45

[lm, 25 mm cellulose nitrate filter on a glass filter manifold. Leaves for bacteria counts were

thawed and sonicated for one minute at 80 W while on ice. Then, the sample was vortexed, and

a 100 pl aliquot was removed after 10s. The sample and one ml nanopure water was added to

the filter manifold to ensure mixing and a homogeneous slide mount and pressure applied using a

vacuum.

The Anodisc filter was removed and gently dried using a Kimwipe. Filters were placed

face-up on a 100ul drop of SYBR Green II fluorescent stain diluted 400 fold (Molecular Probes,

Eugene, Oregon, USA) on labelled petri dishes. Filters were stained in the dark for 15 minutes,

then dried and placed face-up on a glass slide. A 30-uL drop of antifade mounting solution (50%









Goodwin, B.J. and L. Fahrig. 2002. How does landscape structure influence landscape
connectivity? Oikos 99:552-570.

Goodwin, B.J. 2003. Is landscape connectivity a dependent or independent variable? Landscape
Ecology 18:687-699.

Grace, J. M., III, R. W. Skaggs, H. R. Malcom, G. M. Chescheir, and D. K. Cassel. 2003.
Increased water yields following harvesting operations on a drained coastal watershed. ASAE
Paper No. 032038. St. Joseph, Mich.: ASAE.

Grassle, J. P., C. A. Butman, and S. W. Mills. 1992. Active habitat selection by Capitella sp. I
larvae. II. Multiple-choice experiments in still water and flume flows. J. Mar. Res. 50:717-743.

Griswold, M.W., R.T. Winn, T. L. Crisman, and S. W. Golladay. Impacts of climatic stability on
structural and functional aspects of macroinvertebrate communities following drought
(In Revision: Freshwater Biology)

Growns, I. O., and J. A. Davis. 1994. Effects of forestry activities (clearfelling) on stream
macroinvertebrate fauna in south-western Australia. Australian Journal of Marine and
Freshwater Research 45:963-975.

Gurtz, M. E. and J. B. Wallace. 1984. Substrate mediated response of stream invertebrates to
disturbance. Ecology 65:1556-1569.

Haefner, J. D., and J. B. Wallace. 1981. Shifts in aquatic insect populations in a first-order
southern Appalachian stream following a decade of old field succession. Canadian Journal of
Fisheries and Aquatic Sciences 38:353-359.

Hanski, I. 1994. Patch-Occupancy Dynamics in Fragmented Landscapes. Trends in Ecology and
Evolution 9:131-135.

Hanski, I. 1994. A Practical Model of Metapopulation Dynamics. Journal of Animal Ecology
63:151-162.

Hanski, I. 1994. Spatial Scale, Patchiness and Population-Dynamics on Land. Philosophical
Transactions of the Royal Society of London Series B-Biological Sciences 343:19-25.

Hanski, I. 1998. Metapopulation Dynamics. Nature 396: 41-49.

Hanski, I. 1999. Habitat connectivity, habitat continuity, and metapopulations in dynamic
landscapes. Oikos 87:209-219.

Harding, J. S., 2003. Historic deforestation and the fate of endemic invertebrate species in
streams. New Zealand Journal of Marine and Freshwater Research 37:333-345.









CHAPTER 3
TESTING BMP EFFECTIVENESS FOR SMALL COASTAL PLAIN STREAMS
USING MACROINVERTEBRATES AS BIOINDICATORS

Introduction

Headwater streams are tightly coupled with the surrounding riparian landscape.

Thus, changes in the structure of the riparian zone can affect water quality of larger

streams and rivers, since they are heavily influenced by headwater streams that feed them

(Meyer and Wallace, 2001). Headwater streams make up a majority of channel length

within stream networks and serve important ecological and biological functions by

delivering water, sediment, organic material, prey items, and nutrients to downstream

reaches (Sidle et al., 2000; Gomi et al., 2001; Meyer and Wallace, 2001; Wipfli and

Gregovich, 2002). As the importance of ecosystem services, such as water quality and

biodiversity, becomes more widely recognized, the need to protect aquatic resources

increases. Thus, proper management of terrestrial landscapes must take into account

needs of aquatic organisms and communities.

Numerous studies have found significant impacts of logging on physical and

chemical aspects of streams, including reduced large woody debris in streams (Golladay,

Webster and Benfield, 1987), increased sediment input (Beschta, 1978), discharge

(Hartman and Scrivener, 1990), nutrient inputs (Likens et al., 1969; McClurkin et al.,

1985), and decreased shading resulting in higher water temperature (Swift and Messer,

1971; Webster and Waide, 1982). Elevated light, temperature and nutrient concentration

can increase algal biomass within the stream, shifting the base of the food web from

allochthonous to autochthonous sources (Likens et al., 1970; Wallace and Gurtz, 1986;

Bilby and Bisson, 1992). The extent and impact of these effects are influenced by














5- 1

u) 0.8


o. 0.6


0 0.4
a-
0
0.2


0 -- -----
1 2 4
Mass

Figure 4-11. CPOM trapped in patches (+ SE) in relation to patch size.









category per trait was multiplied by the invertebrate abundances. This resulted in a trait-

by-site array that contained the density of individuals for each trait category for each site;

density was transformed (ln(x+l)) to approximate a normal distribution for the statistical

analyses.

Statistical Analysis

Environmental variables

Environmental variables were analyzed over time with repeated measures

ANOVA (SAS Institute, 2002). When differences were significant, post-hoc analysis was

conducted using Tukey's test and Bonferroni corrections. Additionally, environmental

stability was assessed by calculating Bray-Curtis distances (Bray and Curtis, 1957)

between adjacent years. Bray-Curtis distances are a measure of dissimilarity with values

ranging from 0 to 1. Zero denotes identical samples; thus, higher values denote lower

compositional stability. This measure is computed as:


DRh Z= a-, I


where DRh is the distance between samples i and h.

Stability

Compositional stability of invertebrate communities was examined separately for

the two streams between pairs of successive years. Stability was measured by calculating

Bray-Curtis distances between adjacent years based on abundance data and biological

traits. ANOVA was used to examine between year differences in compositional and traits

stability scores for the streams. The relationship between Bray-Curtis values and flow

and SPI values were regressed to assess the impact of hydrologic scale on community and

trait stability.










Table 3-7. Indicator values for watersheds C and D based on biological traits. Groups are
defined as pre-harvest all sites (1), post-harvest reference (2), post-harvest thinned
SMZ (3), and post-harvest intact SMZs (4).

Trait Group Indicator Value p-value
v2 1 27.1 0.012
ar2 1 29.9 0.013
h3 1 32.7 0.023
tr4 1 33.1 0.001
mh7 1 30.9 0.002
ds2 1 30.3 0.001
ht2 1 29.5 0.007
h5 2 44.7 0.003
mh5 2 32 0.003
rs3 2 33.3 0.009
tr3 3 29.9 0.010
mhl 3 28.9 0.023
mh6 4 30.3 0.022
ec2 4 26.3 0.047









et al., 1989), but this process is additionally linked to abiotic conditions. Less BOM was

stored in the sediment with increases in flow, ammonia, conductivity, and dissolved

oxygen. However, these changes were related to factors unique to harvest treatments. In

reference streams, a negative relationships with BOM and conductivity, dissolved

oxygen, and turbidity suggest an interaction between abiotic and biotic factors affecting

loss. As dissolved oxygen levels increased, higher decomposition rates may have been

responsible for decreases in BOM. This was likely related to an increase in microbial

biofilm and invertebrate abundance and diversity typically found at higher dissolved

oxygen levels (Allan, 1995). Decreases in conductivity were linked to flushing of the

streams as flow was restored following the drought. Additionally, drying of the

streambed releases S042- as reduced sulfur is oxidized, leading to an increase in

conductivity (Bayley et al., 1986; Devito, 1999). This increase reduces the solubility of

carbon, thus decreasing decomposition rates (Clark et al., 1005) and potentially reducing

invertebrate abundance.

In the harvested watersheds, the strong negative relationship found between BOM

and flow, suggests the loss of BOM is controlled primarily by physical factors.

Movement of organic matter and sediment occurs during most storm events in sandy-

bottomed, coastal plain streams and is more pronounced in the clearcut streams due to

increased runoff and peak flow (Golladay et al., 1987). Ultimately, this leads to trapping

of litter in discrete, spatially variable habitats, such as debris dams (Palmer et al., 1996).

Although flow was an important predictor of BOM storage, ammonia and turbidity also

played a role. In the streams with an intact SMZ, BOM increased as the water became

more turbid. Reaches draining the intact SMZ retained silt from upstream sections,










Table 3-6. Indicator values for watersheds A and B based on biological traits. Groups are
defined as pre-harvest all sites (1), post-harvest reference (2), post-harvest thinned
SMZ (3), and post-harvest intact SMZs (4).

Trait Group Indicator Value p-value
df3 1 31.3 0.003
ar2 1 31.9 0.014
h3 1 38 0.000
trl 1 29.9 0.007
mh2 1 28.5 0.015
ds2 1 28.3 0.036
ht2 1 32.1 0.003
arl 2 26.6 0.005
tr5 2 29.4 0.007
sh2 2 26.2 0.008
mh4 2 29.7 0.002
dsl 2 29 0.004
s2 3 32.1 0.034
h4 3 28.8 0.030
mhl 3 30 0.040
mh3 3 32.2 0.014
v1 4 36.5 0.026
dfl 4 27 0.042
tr2 4 33.8 0.007









geology, soils and vegetation of the catchment, the extent to which the riparian buffer

strip remains after logging, stream discharge, and channel gradient and morphology.

Changes in abiotic characteristics of a stream following logging can affect the

structure and function of the stream community, including periphyton (Lowe, Golladay

and Webster, 1986), fish (Garman and Moring, 1993) and macroinvertebrates. Logging

activities can disrupt stream invertebrate communities, but the magnitude and trajectory

of effects vary. Increased light penetration and warmer temperatures from canopy

removal, and nutrient enrichment in runoff from ground disturbance, increase aquatic

invertebrate density and/or biomass in streams (Newbold et al., 1980; Murphy et al.,

1981; Hawkins et al., 1982; Wallace and Gurtz, 1986; Campbell and Doeg, 1989; Brown

et al., 1997). Fine sediment loading, particularly in watersheds with steep slopes, can

reduce invertebrate populations following logging (Growns and Davis 1994, Waters

1995, Wood and Armitage 1997), clogging tracheal gills, and burying food sources. In

many cases, invertebrate communities shift from shredders to grazers (algae consumer) or

detritivores (collector-gatherer) (Haefner and Wallace 1981; Gurtz and Wallace, 1984;

Webster et al., 1992).

Although Stone and Wallace (1998) found shifts in dominance of taxa, there was

no loss oftaxa in logged versus unlogged sites. They posited that measures of taxon

richness may be useful for indicating presence of pollutants, but not for more discrete

disturbances. Long-term impacts of clear cutting were documented decades later,

stemming from recovery of riparian vegetation and canopy cover (Growns and Davis,

1991; Stout et al., 1993; Stone and Wallace 1998).









streams and is closely related to water quality. The relationship of this variable with long-

term SPI values indicates the advantage of incorporating regional climatic data into

bioassessment protocols.

Temporal Variation in Traits

Traits were more related to local abiotic variables than to flow or long-term

precipitation indices. Initially, traits may be filtered by large-scale factors, including

climate and geology (Poff, 1997); thus at the smaller scale of two adjacent watersheds,

traits may vary locally. Across small, physiographically homogeneous regions (e.g.,

watersheds, ecoregions), sites are likely to be located within a single regional species

pool (Zobel, 1997). Thus, as predicted by the habitat templet model (Southwood, 1977;

1988; Townsend and Hildrew 1994), local characteristics at the reach scale directly

influence biological traits.

Traits responded predictably to changes in local environmental conditions. As pH

increased and nutrients decreased, species less likely to drift became more common in the

stream reaches. Initial availability of nitrogen and phosphorous allowed for early

colonization of scrapers (e.g., Boulton, 1991), as algal sources within the stream

accumulated on available substrate. However, this effect was not apparent in WF,

reflecting the lower productivity of grazers typically associated with colored, acidic

streams (Rosemond et al., 1992). Additionally, shredders became more common in the

second and third year, as previously exposed patches of organic matter became

submerged. In addition, species typically classified as detritivores (e.g., nemourid

stoneflies) may consume algal biomass, assuming the role of scrapers, especially in

streams with lower pH (Ledger and Hildrew, 2005). In a comparison of rivers affected by

drought in Italy, Fenoglio et al.(2007) documented an increase in collectors and a














O Wet Season
* Dry Season


0.1



0.08

E
E 0.06



0.04



0.02



0




Figure 3-2.


Average chlorophyll a biomass (+SE) during the wet (May-September) and dry
season (October-April) from 2004-2008 in reference, thinned SMZs, and intact SMZ
streams after harvest.


Reference Thinned Intact









by the data since abundance and taxon richness of invertebrates were higher in macrophyte

patches.

In the observational study, filterers were best predicted by chlorophyll a in macrophytes

and bacteria in leaf packs. This supports recent findings that many invertebrates exhibit

plasticity when selecting resources (Friberg and Jacobsen, 1994). In addition to organic matter

sloughing from epiphyton, the structure provided by algae will aid in development of a biofilm

matrix. Additionally, filterers were positively related to dissolved oxygen within the patch,

which was higher in macrophytes since they extend into the water column and release oxygen

during photosynthesis. Switching feeding behavior has also been observed in shredders, mixing

algal and detritus based carbon sources (Friberg and Jacobsen, 1994).

Patch quality is also linked to refractory compounds in leaves that may alter biofilm

structure, decomposition rates, and nutrient availability for colonizing species (Ostrofsky, 1993,

1997). This may be especially true for shredding invertebrates that depend on biofilm as well as

leaf properties (e.g., Lignin content). Habitat selection by shredders was apparent in the short-

term experiment in relation to leaf palatability. After seven days, shredders were more common

in Pinus and Liriodendron than in the less palatable Quercus and Ludwigia. However, shredders

became similar among all treatments after fifteen days and were similar between macrophytes

and leaf packs in the observational study. This indicates that although shredders initially select

more suitable habitat, accumulation of organic matter in other patches creates adequate habitat

for this group. Bastian (2007) found that shredders were distributed across a broad range of leaf

species in a stream, with no leaf species being preferentially colonized by shredders. However,

most studies find that shredder species exhibit clear leaf preferences (Anderson and Sedell, 1979;

Mackay and Kalff, 1973; Nolen and Pearson, 1993), and selectively feed on food resources of









(Plecoptera) of Florida. State of Florida Department of Environmental Protection
Division of Water Resource Management. Tallahassee, Florida. 94 pp.

Peters, R. H. and K. Wassenberg. 1983. The Effect of Body Size on Animal Abundance.
Oecologia 60:89-96.

Petersen, R.C. and K.W. Cummins. 1974. Leaf processing in a woodland stream.
Freshwater Biology 4:343-368.

Petersen, Jr., R.C., K. W. Cummins, and G. M.Ward. 1989. Microbial and animal processing of
detritus in a woodland stream. Ecological Monographs 59:21-39.

Pickett, S. T. A. and J. N. Thompson. 1978. Patch dynamics and design of nature reserves.
Biological Conservation 13:27-37.

Pither, J. and P.D. Taylor. 1998. An experimental assessment of landscape connectivity. Oikos
83:166-174.

Poff, N. L. and J. V. Ward. 1992. Heterogeneous currents and algal resources mediate in situ
foraging activity of a mobile stream grazer. Oikos 65:465-478.

Poff, N.L., Palmer, M.A., P.L. Angermeier, R.L. Vadas, Jr., C.C. Hakenkamp, A. Bely, P.
Arensburger, and A.P. Martin. 1993. Size structure of the metazoan community in a Piedmont
stream. Oecologia 95:202-209.

Poff, N. L. and J. V. Ward. 1995. Herbivory under different flow regimes a field experiment
and test of a model with a benthic stream insect. Oikos 72:179-188.

Poff, N.L. 1997. Landscape filters and species traits: toward mechanistic understanding and
prediction in stream ecology. Journal of the North American Benthological Society 16:391-409.

Poff, N.L., J.D. Olden, N.K.M. Vieira, D.S. Finn, M.P. Simmons, and B.S. Kondratieff 2006.
Functional trait niches of North American lotic insects: traits-based ecological applications in
light of phylogenetic relationships. Journal of the North American Benthological Society 25:
730-755.

Polyakov, V., A. Fares, and M. H. Ryder. 2005. Precision riparian buffers for the control of
nonpoint source pollutant loading into surface water: A review. Environmental Reviews 13:129-
144.

Pusey, B.J. and A.H. Arthington. 2003. Importance of the riparian zone to the conservation and
management of freshwater fish: a review. Marine and Freshwater Research 54:1-16.

Rahel, F.J. 1990 The hierarchical nature of community persistence: A problem of scale.
American Naturalist 136:328-344.









treatment-factor combinations were tested using post hoc Tukey's honest significant difference

(HSD) tests (c = 0.05).

Colonization

Habitat selection based on patch amount and type was examined in a short term

colonization study. Microlandscapes were created along an -75 m stretch of stream, separated

by at least 3 m. Landscapes were the same as those used in the short-term behavioral

experiment, but were half the size (45.7 cm W X 50.8 cm L), and were replicated three times in a

randomized block using each replicate as a block. Prior to creation of the landscape, the

streambed was raked to 0.5 m to remove any apparent organic matter or habitat and allowed to

settle for four hours. Drift nets were placed at the end of the landscape to trap emigrating

invertebrates. Macrophyte and leaf patches were anchored to the sediment in the appropriate

configuration (Fig. 1).

Invertebrates for the experiment were collected from the streambed and, leaf packs and

lengths of individuals were measured. Due to low abundance of Anisocentropus, only one

individual was used for each replicate, however, six individuals of Elimia were used. The shell

or case of the individual was blotted dry and marked with a drop of paint and the number of

landscape (from 1 to 27). Individuals were released at the center of the landscape after the paint

dried (- 5 minutes). After 24 hours, all patches were collected and placed in individually

labelled bags. Velocity was measured at the upstream and downstream end of the landscape with

a Flomate 2000 (Marsh McBirney). In addition, drift nets were collected and any marked

individuals in the matrix (sand) were collected. A surber sample was also taken from the

landscape to determine recolonization by other invertebrates.









movement distances. A drift net was placed at the end of the landscape to trap emigrating

individuals.

Experimental organisms were collected from the stream each morning from the

streambed and naturally occurring leaf packs. Individuals were placed in separate flow-through

trays and allowed to acclimate to the stream reach for at least an hour. A video camera was set

up on a tripod to record movement of individuals. During each trial, individuals were placed at

the center of the landscape, facing downstream. Behavior was recorded for 30 minutes, with the

observer leaving the reach while trials were occurring. After 30 minutes, the length and width of

each individual was measured, removed from the landscape, and released downstream. On

average, Anisocentropus individuals were 4.2 cm long ( 0.2 SE) and 2 cm wide ( 0.1 SE),

while Elimia individuals were 4.3 cm long ( 0.1 SE) and 1.7 cm wide ( 0.1 SE). No individual

was used more than once, and trials were repeated for at least four individuals (more for Elimia

due to availability). After each trial, the streambed was gently scoured to remove any traces of

the individual path. Trials were run between 7 AM and 4 PM daily for a period of seven days

beginning 7 March, 2007. The grid was left in place each night, and pvc pipes were inserted into

the streambed as placeholders for the tripod.

Videos were digitized and manually analyzed on a computer screen with coordinates

(x,y) recorded every 10 s to determine movement parameters. For each path, total path length,

correlations between turning angles, mean cosine of turning angle, mean path length, and net

squared displacement were calculated to assess distance covered (Turchin et al. 1991). The

above parameters were used for correlated random walk models, which are useful for making

inter-specific comparisons (Kareiva and Shigesada 1983; Cain 1985; Crist et al.1992). Each path









Merriam, G.. 1984. Connectivity: a fundamental ecological characteristics of landscape pattern. -
In: Brandt, J. and Agger, P. (eds), Methodology in landscape ecological research and planning.
Roskilde Universitetsforlag, GeuRuc, Roskilde, Denmark, pp. 5- 15.

Merritt, R.W. and K.W. Cummins. 1996. An Introduction to the Aquatic Insects of
North America, 3rd edition. Kendall/Hunt Publishing Company, Dubuque, Iowa.

Meyer, J.L. and J.B. Wallace. 2001. Lost linkages and lotic ecology: rediscovering small
streams. In:Ecology: Achievement and Challenge (eds M.C. Press, N. Huntly and S. Levin), pp.
295-317. Blackwell Science, Oxford, UK.

Meyer, J.L., D.L. Strayer, J.B. Wallace, S.L. Eggert, and G.S. Helfman, 2007. The Contribution
of Headwater Streams to Biodiversity in River Networks. Journal of the American Water
Resources Association 43:1752-1688.

Miltner, R.J. and E.T. Rankin. 1998. Primary nutrients and the biotic integrity of rivers and
streams. Freshwater Biology 40:145-158.

Minshall, G.W. 1984. Aquatic insect-substratum relationships. In: The Ecology of Aquatic
Insects (Eds V.H. Resh and D.M. Rosenberg), pp. 358-400. Praeger Publishers,
New York.

Miyake,Y., T. Hiura, N.Kuhara, and S.Nakano. 2003. Succession in a stream invertebrate
community: A transition in species dominance through colonization Ecological Research 18:
493-501.

Moilanen, A. and I. Hanski. 1998. Metapopulation dynamics : effects of habitat quality and
landscape structure. Ecology 79:2503-2515.

Moloney, K.A. and S.A. Levin. 1996. The effects of disturbance architecture on landscape-level
population dynamics. Ecology 77:375-394.

Moore, R.D. and J.S. Richardson. 2003. Progress towards understanding the structure, function
and ecological significance of small stream channels and their riparian zones. Canadian Journal
of Forest Research. 33:1349-1351.

Moore, J.C., E. L. Berlow, D. C. Coleman, P. C. de Ruiter, Q. Dong, A. Hastings, N. C. Johnson,
K. S. McCann, K. Melville, P. J. Morin, K. Nadelhoffer, A. D. Rosemond,
D. M. Post, J. L. Sabo, K. M. Scow, M. J. Vanni and Diana H. Wall. 2004. Detritus, trophic
dynamics and biodiversity. Ecology Letters 7:584-600.

Morgan, M. D. 1985. Photosynthetically elevated pH in acid waters with high nutrient content
and its significance for the zooplankton community. Hydrobiologia 128:239-247.









effects of local filters on invertebrate communities. Thus, the quality of the riparian and

availability of instream habitat create a filter to limit presence of certain species.

Results from this study support the idea that instream habitat availability controls small

scale community composition. The habitat specialist, Anisocentropus, left landscapes without

preferred leaf litter habitat. In streams with little organic matter storage, this species may be

driven locally extirpated. Although it may be supported where riparian zones are left

undisturbed, this species prefers small streams and may be driven out of entire headwater

streams if they are logged along their length.

Logging limits the amount and quality of habitat available for aquatic invertebrates in

streams. This is accomplished by reducing leaf fall, as well as through an increase in peak flow

with increasing surface runoff (Beasley and Granillo, 1982; Williams et al., 1999; McBroom et

al., 2002; Grace et al., 2003). Thus, any leaf fall that does reach the stream is easily washed

downstream during storm events.

The results of this study suggest that increased habitat cover decreases emigration rates,

regardless of habitat configuration. Very few Anisocentropus individuals were able to colonize

patches successfully, but when successful, they remained there for the duration of the trial. This

suggests that, although small, patches were able to be used as refugia from flow and exposure.

Both Elimia and Anisocentropus were likely to remain in the microlandscapes with 30% cover.

Changes in landscape structure, such as reduction of the proportion of one or more patch

types or increased patch isolation, will alter the ability of organisms to disperse (Merriam 1984;

Fahrig and Merriam 1985). Species that can not disperse effectively as a result of a change in

structure will suffer reductions in regional population sizes (Fahrig and Merriam 1994). As a

result, relative abundances of Anisocentropus decreased in treatment watersheds following









Environmental variables

Environmental variables were analyzed over time with repeated measures

ANOVA (SAS Institute, 2002). Since macroinvertebrate samples were taken during the

winter/spring period, only environmental data from this period were analyzed. When

differences were significant, post-hoc analysis was conducted using Tukey's test and

Bonferroni corrections. Additionally, environmental stability was assessed by calculating

Bray-Curtis distances (Bray and Curtis, 1957) between adjacent years. These measure

dissimilarity with values ranging from 0 to 1. Zero denotes identical samples; thus, higher

values denote lower stability and unity implies complete turnover.

Macroinvertebrates

The Florida Stream Condition Index (SCI) combines metrics that respond to

changes in human induced disturbance to yield a score reflecting water quality (Florida

Department of Environmetal Protection, 2004). Higher values indicate better water

quality. SCI values were analyzed by time and treatment effects using repeated measures

ANOVA (SAS Institute, 2002).

Stability of invertebrate communities

Compositional stability of invertebrate communities was examined for streams

between pairs of successive years (e.g., 1 vs 2, 2 vs 3, etc.). Stability was measured by

calculating Bray-Curtis distances between adjacent years based on abundance data and

biological traits. ANOVA then was used to examine between year differences in

compositional and biological trait stability scores for the streams.

Ordination: species composition and traits

Nonmetric multidimensional scaling (NMDS; Kruskal, 1964) was used to explore

temporal patterns in species composition and biological traits as in Chapter 2. Since the





















A U

A A
A A



AA
A
.0


h5












dsl rl
tr
rs3 htl r *s1 ih2
mhl dl h6* '2*es*tri
S d2*sh2*.mh4ec2
tr2 mh3 h m.7.dfl 3 h2 f3 .h3
s2 es2 r3
t3 h ds3 shf2 ht2
e mh
r4 rs2
ecl
tr4


Axis 1


Tss


LF
DO
NH4
Flow
TN
Turb
SC


Axis 1


Figure 3-10. NMDS of biological traits in watersheds A and B in pre-harvest (1) and in post-

harvest reference (2), thinned SMZs (3), and intact SMZ treatments (4).





102


I I I I I










current study focused primarily on natural variation in climatic conditions and indicated

changes in trait composition as a result of this variation. Thus, additional effort should be

devoted to distinguishing the effects of natural and anthropogenic disturbances when

devising biological monitoring programs.


















Table 2-1. Definition and codes for biological traits and modalities.

Trait Code Modality Trait Code Modality
Voltinism v1 Semivoltine Habit hi Clingers


Drying Resistance

Drift


Armoring


Maximum Size


Rheophily


Univoltine
Multivoltine
Absent
Present
Rare
Common
Abundant
Soft
Sclerotized
Case/Shell
Small (<9mm)
Medium (9-16mm)
Large (>16mm)
Standing
Slow
Fast


h2
h3
h4
h5
h6
Trophic trl
tr2
tr3
tr4
tr5
Shape shl
sh2


Burrowers
Swimmer
Sprawler
Skater
Climber
Gatherer
Filterer
Scraper/Herbivore
Shredder
Predator
Streamlined
Not Streamlined
(Bluff, Tubular)










Table 4-1. Multiple regressions for leaf packs averaged over all time periods for the
observational study.
Dependent Variable Parameter Estimate SE t P
Invertebrate Abundance (F4,59=15.0, P < 0.0001, R2 = 0.58)
Size 0.28 0.23 0.05 0.96
Chlorophyll a -17.7 50.4 -0.35 0.72
Bacteria Abundance 0.08 0.14 0.57 0.57
Bacteria Biomass 0.33 0.1 3.2 0.003
FPOM 0.38 0.13 2.9 0.005
Taxon Richness (F4,59=7.7, P < 0.0001, R2 = 0.42)
Size 0.13 0.14 0.9 0.37
Chlorophyll a -26.8 30.4 -0.88 0.38
Bacteria Abundance 0.05 0.08 0.56 0.58
Bacteria Biomass 0.12 0.06 1.9 0.06
FPOM 0.17 0.08 2.2 0.03
Shredders (F4,59=2.0, P = 0.08, R2 = 0.16)
Size 0.86 0.35 2.5 0.02
Chlorophyll a -134.6 76.2 -1.77 0.08
Bacteria Abundance -0.52 0.21 -2.51 0.01
Bacteria Biomass -0.13 0.16 -0.8 0.13
FPOM -0.17 0.2 -0.87 0.39
Filterers (F4,59=8.9, P < 0.0001, R2 = 0.45)
Size -0.29 0.27 -1.1 0.28
Chlorophyll a 72.2 58.8 1.23 0.22
Bacteria Abundance 0.38 0.16 2.3 0.02
Bacteria Biomass 0.24 0.12 2 0.04
FPOM 0.46 0.15 3 0.005









CHAPTER 6
CONCLUSIONS

Best management practices for forestry in the U.S. clearly depend on the geographic

region under review. For example, coastal plain streams in the southern U.S. are

characteristically low-gradient, sandy-bottomed systems with dynamically changing instream

habitat. In contrast, those managed for forestry in the western U.S. are typically high-gradient

montane streams with high habitat and substrate diversity and are susceptible to mass-wasting as

vegetation removal reduces bank stability. Although forestry practices in the Northwest can lead

to drastic reductions in water quality, evidence from the coastal plain indicates limited changes

in water quality and biotic diversity in streams impacted by logging, as long as stream

management zones are left intact.

Although there were few changes in biotic community structure following logging, this

does not discount use of aquatic invertebrates as indicators of water quality. Many biotic indices

weigh heavily upon the use of EPTs (Ephemeroptera, Plecoptera, and Trichoptera) in their

formation (e.g., Lenat, 1993). However, logging ultimately increases primary productivity in

streams, leading to higher densities of Baetid/Leptophlebiid ephemeropterans (Chapter 3; Stone

and Wallace, 1998). As a result, the FLSCI biotic index is inflated, suggesting an increase in

water quality with logging. One short-coming of local management organizations is the long-

term fascination with EPTs, sometimes leading to redundant use of this group by utilizing

metrics on the number of EPT taxa, % EPT, number of Trichoptera, and number of

Ephemeroptera, to name a few (e.g., Maxted et al., 2000).

As an alternative, use of biological traits has recently been advocated as a potential tool

for assessing aquatic ecosystems by academia and the federal government (Poff et al., 2006).

Biological traits are more informative indicators of ecosystem function than are changes in









3-5 Stream condition index (SCI) scores (SE) for reference, thinned SMZs, and intact
SMZ streams. Samples below the red line indicate poor water quality, those above
the red line, fair water quality, and those above the blue line, good water quality. ..........97

3-6 Taxonomic stability (+SE) for reference, thinned SMZs, and intact SMZ streams. .........98

3-7 Trait stability (SE) for reference, thinned SMZs, and intact SMZ streams ...................99

3-8 NMDS of taxonomic composition in watersheds A and B in pre-harvest (1) and in
post-harvest reference (2), thinned SMZs (3), and intact SMZ treatments (4)................100

3-9 NMDS of taxonomic composition in watersheds C and D in pre-harvest (1) and in
post-harvest reference (2), thinned SMZs (3), and intact SMZ treatments (4)................101

3-10 NMDS of biological traits in watersheds A and B in pre-harvest (1) and in post-
harvest reference (2), thinned SMZs (3), and intact SMZ treatments (4)......................102

3-11 NMDS of biological traits in watersheds C and D in pre-harvest (1) and in post-
harvest reference (2), thinned SMZs (3), and intact SMZ treatments (4)......................103

4-1 Total biomass of chlorophyll a (mg) ( SE) in each patch type ..................................... 127

4-2 Total number of bacterial cells (1 X 106) ( SE) in each patch type.............................128

4-3 Bacterial biomass (pg C/cm3) ( SE) in each patch type........................... .............129

4-4 Number of bacterial cells per cm3 (1 X 106) ( SE) in each patch type ........................130

4-5 Chlorophyll a biomass (mg/cm3) ( SE) in each patch type............. .................131

4-6 Volume-weighted taxon richness (Taxa/cm3) (+ SE) in each patch type ......................132

4-7 Volume weighted invertebrate density (Individuals/cm3) ( SE) in each patch type......133

4-8 Proportion of filtering invertebrates ( SE) in each patch type..................................134

4-9 Proportion of leaf mass decomposed ( SE) in relation to patch type and disturbance. .135

4-10 Amount of leaf mass decomposed (g) ( SE) in relation to initial patch mass. ..............136

4-11 CPOM trapped in patches ( SE) in relation to patch size. ...........................................137

4-12 Average amount of coarse particulate organic matter (g) ( SE) trapped in each
patch type. ................................................................................ 138

4-13 Average amount of coarse particulate organic matter (g) ( SE) trapped in patches by
disturbance type. ........................................................................ 139











4
-A- Leaf Packs
3.5 --- Ludwigia

3
E

02.5
U)
2

r1.5
0
x
1


0.5

0
November 2005 January 2006 April 2006 June 2006


Figure 4-6. Volume-weighted taxon richness (Taxa/cm3) ( SE) in each patch type.

































Day 15 Disturbed


Day 15 Undisturbed


Figure 4-13. Average amount of coarse particulate organic matter (g) (+ SE) trapped in patches
by disturbance type.


1
-


u) 0.8
()
C--


C

S 0.4


0.2


0


Day 7


T i









Wallace J.B. and M.E. Gurtz. 1986. Response ofBaetis mayflies (Ephemeroptera) to catchment
logging. American Midland Naturalist 115:25-41.

Wallace, J. B., M. R.Whiles, S. Eggert, T. F. Cuffney, G. J. Lugthart, and K. Chung. 1995. Long-
term dynamics of coarse particulate organic matter in three small Appalachian Mountain
streams. Journal of the North American Benthological Society 14:217-232.

Wallace, J. B., S. L. Eggert, J. L. Meyer, and J. R. Webster. 1999 Effects of resource
limitation on a detrital-based ecosystem. Ecological Monographs 69:409-442.

Ward, J. V., G. Bretschko, M. Brunke, D. Danielopol, J. Gibert, T. Gonser, and A.G.
Hildrew. 1998. The boundaries of river systems: the metazoan perspective. Freshwater Biology
40:531-569

Waters, T. F. 1972. Drift of stream insects. Annual Review of Entomology 17:253-278.

Waters, T. F. 1995. Sediment in streams. Sources, biological effects, and control. Monograph 7.
American Fisheries Society, Bethesda, Maryland.

Webster, J. R. and J. B. Waide. 1982. Effects of forest clearcutting on leaf breakdown in a
southern Appalachian stream. Freshwater Biology 12:331-344.

Webster, J. R. and E. F. Benfield. 1986. Vascular plant breakdown in freshwater ecosystems.
Annual Review of Ecology and Systematics 17:567-594.

Webster, J.R., S. W. Golladay, E. F. Benfield, D. J. D'Angelo, and G. T. Peters.1990. Effects of
forest disturbance on particulate organic matter budgets of small streams. Journal of the North
American Benthological Society 9:120-140.

Webster, J. R., S. W. Golladay, E. F. Benfield, J. L. Meyer, W. T. Swank, AND J. B. Wallace.
1992. Catchment disturbance and stream responses: an overview of stream research at Coweeta
Hydrologic Laboratory. Pages 231-253 In P. J. Boon, P. Calow, and G. E. Petts (editors). River
conservation and management. John Wiley and Sons Ltd., Chichester, UK.

West, B. 2002. Water quality in the South. In Southern Forest Resource Assessment, 455-477.
General Tech. Report SRS-53. D. N. Wear and J. Greis, eds. Asheville, N.C.: USDA Forest
Service, Southern Research Station.

Wharton, C.H. 1978. The Natural Environments of Georgia. Georgia Department of Natural
Resources.

Wiegand, T., K.A. Moloney, and N. J, Knauer. 1999. Finding the missing link between
landscape structure and population dynamics: a spatially explicit perspective. American
Naturalist 154:605-627.

Wetherald R.T. and S. Manabe. 2002. Simulation of hydrologic changes associated



























Table 3-1. Biological trait definitions and modalities.


Trait Code Modality IlTrait Code Modality


Life History
Voltinism


Drying Resistance

Eggs cemented to substrate

Development Time


Egg Hatch Time


Mobility
Drift


Morphology
Armoring


Maximum Size


Shape


Respiration


Semivoltine
Univoltine
Multivoltine
Absent
Present
Yes
No
< 6 weeks
< 1 year
> 1 year
< 1 week
< 1 month
> 1 month

Rare
Common
Abundant

Soft
Sclerotized
Case/Shell
Small (<9mm)
Medium (9-16mm)
Large (>16mm)
Streamlined
Not Streamlined
(Bluff, Tubular)
Cutaneous
Tracheal Gills
Spirales/Plastron


Ecology
Habit






Trophic





Rheophily



Microhabitat


h1
h2
h3
h4
h5
h6
trl
tr2
tr3
tr4
tr5
rl
r2
r3
r4
mhl
mh2
mh3
mh4
mh5
mh6
mh7


Clingers
Burrowers
Swimmer
Sprawler
Skater
Climber
Gatherer
Filterer
Scraper/Herbivore
Shredder
Predator
Standing
Slow
Fast Laminar
Fast Turbulent
Sand
Rock
Gravel
Macrophyte/Algae
Detritus
Woody debris
Silt









drought depends on specific life history traits, including resistance to desiccation and an

ability to colonize habitats rapidly through drift and aerial migration or oviposition

(Williams, 1987, 1996; Boulton, 1989). Further colonization reflects subsequent changes in

water chemistry, habitat availability, and resource base following flow resumption.

Biological traits are more informative indicators of ecosystem function than are

changes in abundance of individual species, and they are expected to change across a

gradient of anthropogenic and natural disturbances (Charvet et al., 2000; Dole'dec et al.,

1999; Statzner, Hildrew and Resh, 2001). However, species loss decreases the ability of

ecosystems to resist disturbances and leads to lowered stability (Hooper et al., 2005).

Therefore, an integrative approach should utilize both species composition and biological

traits to predict community responses to disturbances (Richards et al., 1997) Biological

traits are regulated at a hierarchy of scales, with environmental filters (e.g., climate and

geology) creating a template for traits that are present in a specific region (Townsend and

Hildrew, 1994; Poff, 1997). Thus, a subset of traits is expected to respond to disturbances

within a certain region. For example, species that are resilient to disturbance display a

series of traits, including small size and multiple generations per year, that allow them to

expand their population densities rapidly (Townsend and Hildrew, 1994). As functional

redundancy is common among stream invertebrates, biological traits can be compared

across large regions to understand the large-scale impacts of anthropogenic change

(Statzner et al., 2004).

This study utilized a six-year (2001-2007) dataset of macroinvertebrates from

headwater streams after an intense drought in the southeastern U.S. (1998 to 2002) to

characterize inter-year successional patterns following flow restoration relative to water









BIOGRAPHICAL SKETCH

Marcus Griswold was born in Baltimore, Maryland, on September 30, 1978. He pursued a

B.S. in biology at the University of Maryland at College Park. His master's work took him to

the University of Florida to work on predator-prey dynamics of larval mosquitoes under the

direction of Phil Lounibos. His interest in aquatic ecology and background in Entomology led

him to Thomas Crisman to pursue a PhD in environmental engineering sciences, with a focus on

riparian zone management in aquatic ecosystems. During this time, his work was funded by the

U.S. EPA, Sigma Xi, and the Friends of the Osa. He has worked in a variety of stream systems

in the southeastern U.S. and Costa Rica, from primary tropical forests to degraded urban streams.

His goal is to utilize his knowledge of aquatic stressors to properly manage aquatic ecosystems,

balancing human needs and maintenance of ecosystem function and biodiversity.









the southeastern part of the country. Furthermore, the effects of partial harvest within SMZs on

water quality are not well documented. More research is necessary to fill in gaps that currently

exist regarding BMP effectiveness in the coastal plain and effects of partial harvesting within

SMZs.

Habitat Fragmentation and Forestry Practices

Forestry practices potentially have adverse effects on communities by limiting dispersal

between watersheds, eliminating suitable environmental conditions, and altering predator-prey

dynamics. Even with current regulations for stream water quality, clear cutting of a watershed

down to the buffer zone commonly occurs. Although this can maintain local biodiversity,

dispersal across this newly created, potentially hostile landscape may be difficult for small

organisms such as invertebrates and amphibians (Hughes et al., 1996; Fagan, 2002; Briers et al.,

2004).

Although distance between watersheds can serve as a template for determining

population structure and species composition (e.g., Harding, 2003), locally influenced

microhabitats may be the strongest drivers of community structure at the reach and microhabitat

scales. Indirect effects of logging or riparian zone modification lead to changes in microhabitat

structure in streams. This was clearly demonstrated in afforested agricultural streams that

displayed an 87 % reduction in the leaf litter storage compared to forested streams (Benstead and

Pringle, 2004). Similarly, Noel et al.(1986) found that 50% of logged streams were covered by

macrophytes, while unlogged reference streams had only 10% macrophyte cover. Thus, a

gradient of tree removal from the riparian zone should change the physical and biotic structure of

the stream in a predictable manner.

In logged streams, leaf pack formation is often slow, resulting in increased patch isolation

and fragmentation. Rooted macrophytes, however, become more abundant in logged streams









Mulholland, P.J., A. V. Palumbo, J.W. Elwood, and A. D. Rosemond. 1987. Effects of
acidification on leaf decomposition in streams. Journal of the North American Benthological
Society 6:147-158.

Miller, K. The colonization cycle of freshwater insects. Oecologia 53:202-207.

Murphy, M. L. C., P. Hawkins, and N. H. Anderson. 1981. Effects of canopy modification and
accumulated sediment on stream communities. Transactions of the American Fisheries Society
110:469-478.

Murphy, M.L., J. Heifetz, S.W. Johnson, K.V. Koski, and J.F. Thedinga. 1986. Effects of clear-
cut logging with and without buffer strips on juvenile salmonids in Alaskan streams. Canadian
Journal of Fisheries and Aquatic Sciences 43:1521-1533.

Murphy, F., P. S. Giller, and M. A Horan. 1998. Spatial scale and the aggregation of stream
macroinvertebrates associated with leaf packs Freshwater Biology 39:325-337.

Nakano, S., H. Miyasaka, and N. Kuhara. 1999. Terrestrial-aquatic linkages: riparian arthropod
inputs alter trophic cascades in a stream food web. Ecology 80:2435-2441.

Nams, V. 0. and M. Bourgeois. 2004. Fractal dimension measures habitat use at different spatial
scales: an example with marten. Canadian Journal of Zoology 82:1738-1747.

Newman, R.M. 1991. Herbivory and detritivory on freshwater macrophytes by invertebrates: a
review. Journal of the North American Benthological Society 10:89-114.

Niederlehner, BR and J. Cairns Jr. 1990. Effects of ammonia on periphytic communities.
Environmental Pollution 66:207-21.

Newbold, J. D., D. C. Erman, and K. B. Roby. 1980. Effects of logging on macroinvertebrates in
streams with and without buffer strips. Canadian Journal of Fisheries and Aquatic Sciences
37:1076-1085.

Noel, D. S., C. W. Martin, and C. A. Federer. 1986. Effects of forest clearcutting in New-
England on stream macroinvertebrates and periphyton. Environmental Management 10:661-670.

Nolen, J.A. and R.G. Pearson. 1993. Factors affecting litter processing by Anisocentropus
kirramus (Trichoptera: Calamoceratidae) from an Australian tropical rainforest stream.
Freshwater Biology 29:469-479.

O'Hare, M.T. and K. J. Murphy. 1999. Invertebrate hydraulic microhabitat and community
structure in Callitriche stagnalis Scop. Patches. Hydrobiologia 415:169-176.

O'hop, J., J. B. Wallace, and J.D. Haefner. 1984. Production of a stream shredder, Peltoperla
maria (Plecoptera: Peltoperlidae) in disturbed and undisturbed hardwood catchments Freshwater
Biology 14:13-21.












Table 2-2. Mean annual values for environmental variables for the wetland-fed (WF) and seep-fed (SF) streams


Year Flow TSS NH4 o- NO2/NO3 Total Total pH SC DO Turbidity Temperature Leaffall
(L/s) (g/L) (pg/L) phosphate (pg/L) Phosphorous Nitrogen (uS/cm) (mg/L) (NTU) (C) (g/m2)
(pg/L) (pg/L) (pg/L)


WF 2001- 0.99 0.015 2.48 2.73 1.08 8.22 238.00 5.53 42.28 4.47 1.78 16.13 34.29
2002
2002- 2.52 0.001 6.28 1.75 0.00 3.69 278.28 4.73 30.60 7.34 0.19 12.15 12.29
2003
2003- 1.04 0.017 11.97 1.85 0.00 13.26 345.87 5.03 35.95 4.51 1.18 16.08 19.26
2004
2004- 1.76 0.003 2.88 2.49 2.34 4.90 233.30 4.87 24.90 7.62 1.23 12.00 21.26
2005
2005- 1.47 0.013 6.20 1.78 0.00 16.94 343.30 5.35 26.85 6.87 1.10 13.63 30.19
2006
2006- 2.58 0.008 3.93 3.01 0.00 9.61 291.87 5.11 33.68 8.99 0.63 13.05 24.55
2007


SF 2001- 0.02 0.004
2002
2002- 2.74 0.004
2003
2003- 3.07 0.008
2004
2004- 3.39 0.003
2005
2005- 5.36 0.016
2006
2006- 3.90 0.001
2007


45.46

27.24

23.74

18.86

12.43

25.69


77.00

51.25

51.75

39.00

30.25

27.36


212.46

285.97

237.73

232.05

218.54

203.76


84.03

94.85

70.90

74.98

60.93

82.80


15.68

12.73

15.90

11.73

13.05

12.05


38.25

16.23

22.18

26.06

29.76

25.84









CHAPTER 2
IMPACTS OF CLIMATIC STABILITY ON THE STRUCTURAL AND
FUNCTIONAL ASPECTS OF MACROINVERTEBRATE COMMUNITIES AFTER
SEVERE DROUGHT

Introduction

Natural disturbances regulate community structure and ecosystem function, and

thus play a crucial role in shaping aquatic and terrestrial communities (Sousa, 1984; Resh

et al., 1988). Aquatic ecosystems are especially vulnerable to extreme climatic changes,

such as drought, because these disturbances alter flow regimes, water chemistry, and

ultimately, the biotic community (Wood and Petts, 1999). The long-term effects of

drought on the economy, wildlife habitat, and recreation occur as ramp disturbances over

periods of years (sensu Lake, 2003), as opposed to the effects of flooding events, which

subside after weeks or months. The frequency and predictability of droughts are generally

low. However, when drought does occur, it can potentially act as a destabilizing agent for

aquatic communities. The forecast for climate change suggests increased frequency of

extreme events, particularly drought, over the next century (Wetherald and Manabe,

2002; Kundzewicz et al., 2007). Increased intensity and frequency of natural disturbances

will ultimately affect ecosystem stability and influence organisms' resistance and

resilience to change.

During extreme drought, streams typically form a series of disconnected pools

and lose evidence of surficial flow over time, a response that can potentially reset the

aquatic community. Furthermore, toxic accumulation of nutrients and waste (Towns

1985, 1991; Closs and Lake 1995; Dahm et al., 2003), coupled with increased

temperature (Matthews, 1998) and lowered dissolved oxygen (Stanley, 1997; Golladay

and Battle, 2002), add stress to the remaining species pools. Species survival after









shredders became more common over time, more so in the undisturbed treatments, while more

abundant in the largest patches, they were not abundant overall and only ranged from 0-6 percent

of the community (Figs. 4-20,4-21). The proportion of filterers changed in response to mass

(F2,402 = 5.9, P < 0.01 and leaf type (F3,402 = 10.4, P<0.0001) and were twice as common in

Ludwigia than any other patch type and were more abundant in larger patches (Fig. 4-22,4-23).

Collector-gatherers were the dominant feeding group in all patches, ranging from 40-60 percent

of the community. Collector-gatherers differed between leaf species (F3,402 = 6.02, P<0.001) and

were least abundant in Ludwigia (Fig. 4-24).

Regressions

Invertebrate abundance was positively related to CPOM, velocity, and canopy cover and

negatively related to FPOM. Taxon richness had a positive relationship with velocity and

CPOM. The proportion of scrapers was negatively related to increased canopy cover and

positively related to velocity. Filterers were positively related to FPOM and oxygen within the

patch. Shredders were not predicted by any environmental variable. Collector-gatherers were

negatively related to FPOM and positively related to canopy cover (Table 4-3).

Discussion

Stream invertebrate communities are structured by a mosaic of habitats ranging from

macrophytes and substrate diversity to small-scale changes in flow patterns. Community

composition is related to the quantity, quality, and distribution of detritus on the streambed in

headwater streams (Arsuffi and Suberkropp, 1985; Murphy et al., 1998), and plays a significant

role in the distribution, species composition, and total biomass of benthic invertebrates

(Hearnden and Pearson, 1991; Reice 1974). Thus, patch size and quality are two key factors

affecting colonization patterns of patches. In the current study, invertebrate density and



















- 01-02
A 02-03
03-04
04-05
m 05-06
S06O-07


no **
-SPI12


SPC \ 3 PI44
DO


+ +


-0 .5 .


























D.s .
os


















*OS.








-02 -


Sr


Figure 2-8. NMDS ordinations of logl0-abundance in site-year space and taxon-space for WF.

Time periods are indicated by different symbols. Ordination plots of taxa are based on

weighted-averaging.


1 .5 -


0n


C~ r,rr'

J4 llr
,' +4* f 4
ch.
+ +Hr +
fi:". tA~ih;. ".M! Afima.


C^*^ fe ^ "yi ,..
,,^ -E- r "'"l^ "-g
rk+.ll'i


""* rn,- 'I **'p
V F Xi i -h
*- -A, 4e:, .,
(, ~ a tC*1 he* (la .
*NI the
1: + u ng.
4ff, Lie I.M '.l
+ + V


1.4,:
Ki^~* rr l









Regressions. Invertebrate abundance had a positive relationship with both bacterial

biomass and FPOM for leaf packs, but was only related to FPOM for macrophytes. Taxon

richness was positively related to bacterial biomass and FPOM for leaf packs and FPOM for

macrophytes. The proportion of shredders was positively related to patch size and negatively to

bacterial abundance for leaf packs and did not relate to any parameter for macrophytes. Scrapers

were positively related to patch size, chlorophyll a, and bacterial abundance for macrophytes.

Filterers were positively related to bacterial abundance, biomass, and FPOM for leaf packs and

to FPOM and chlorophyll a for macrophytes. Collector-gatherers were related to chlorophyll a

and FPOM for macrophytes (Tables 4-1, 4-2).

Field Experiment

C:N ratios varied among leaf species with Pinus (40.29) having the highest and Ludwigia

(14.7) have the lowest ratio. Quercus and Liriodendron were similar with ratios of 30.5 and

26.6, respectively. Leaf mass decomposition over time was dependent on leaf species (F2,273

148.4, P<0.001), disturbance (F2,273 = 9.9, P<0.001), and mass (F2,273 = 205.5, P<0.001). Pinus

and Liriodendron lost two to three times more mass than Quercus patches (Fig. 4-9). The

percent of leaf mass loss increased with time, but did not differ between disturbance treatments.

Larger leaf packs lost more mass over time than smaller leaf packs, with 4-gram packs losing

five times more than 1-gram packs (Fig. 4-10). However, when corrected for percent loss over

time, mass was not significant. On average, leaves lost twenty five percent of their mass,

regardless of initial mass.

Changes in velocity due to disturbance depended on mass (F4,408 = 5.8, P<0.001) and leaf

species (F6,408 = 2.2, P<0.04). Average velocity ranged from 0.04 to 0.06 cm/s. In general

velocities were lower in Liriodendron and higher in larger leaf packs. CPOM trapped within

patches was related to leaf species (F3,372 = 50.8, P <0.001), disturbance (F2,372 = 4.7, P<0.01),











1.6

-0- Intact SMZ
1.4 -A- Thinned SMZ
S-- Reference

1.2


5 1


S0.8


S0.6


0.4


0.2


0 -
2001-2002 2002-2003 2003-2004 2004-2005 2005-2006 2006-2007


Figure 3-6. Taxonomic stability (+SE) for reference, thinned SMZs, and intact SMZ streams.









Species dominating the overstory in riparian areas were: Nyssa biflora, Liriodendron

tulipifera, Pinus glabra, Magnolia virginiana, Fagus grandifolia, Liquidambar

styraciflua, Quercus nigra, and Quercus michauxii. Magnolia grandiflora was most

common in watersheds C and D (International Paper unpublished data). The upland of

each watershed was dominated by Pinus taeda, which was established at varying times

by hand planting. The midstory of all watersheds was generally composed of Carpinus

caroliniana, Ostrya virginiana, Acer rubrum, Acer barbatum, and Oxydendrum

arboretum. Magnolia pyramidata occurred in riparian areas and midslopes of watersheds

C and D.

Climate

The climate of the region is characterized by warm, humid summers and mild

winters with average annual precipitation of 1412 mm (SERCC, 2007). Temperatures

range from an average maximum of 33.5C to a minimum of 2.8 C. June has the highest

mean rainfall (152.1 mm) and October the lowest (77.5 mm) (SERCC, 2007). Summer

rains are usually short, high intensity events giving way to low intensity frontal events

from late fall to early spring. Due to proximity to the Gulf of Mexico, spin-off from

hurricanes and tropical storms in late summer is not unusual. Drought conditions

occurred during 1998-2002 and resulted in an accumulated rainfall deficit of 711-1270

mm in parts of southwestern Georgia (see Chapter 2).

Hydrology

Surface water flow in the Apalachicola-Chattahoochee-Flint river basin is lowest

from September to November and peaks during January to April due to higher rainfall

and decreased evapotranspiration (Couch et al., 1996). Streams and rivers in the Coastal

Plain receive substantial groundwater because they are typically deeply incised into












0.03



0.025


Q,
V 0.02



0.015
O

0
0. 0.01

aC

0.005



0 --
1 2 4
Mass


Figure 4-20. Proportion of shredders (+ SE) in each patch based on initial patch mass.









RIPARIAN ZONE MANAGEMENT IN COASTAL PLAIN STREAMS: MULTI-SCALE
EFFECTS OF HABITAT FRAGMENTATION





















By

MARCUS WAYNE GRISWOLD


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

2008























LF




Oph_ NN

M


* 01-02

A 02-03

o03-04

* 04-05

*05-06
o 06-07


p *.


SPL4





0] "--_


-21J -1nj O. 1.0


Figure 2-9. NMDS ordinations oflogl0-abundance in site-year space and taxon-space for SF.
Time periods are indicated by different symbols. Ordination plots of taxa are based on
weighted-averaging.


-25 -15 -05 5 15


as







CN


L *.J
CI:f.


'OL~
:ci,
~ ~.p
* ~n









Alotanypus (r = -0.6), Nippotipula (r = 0.5), Crangonyx (r = -0.8), Habrophlebiodes (r =

0.6), Helichus (r = 0.6), Stenelmis (r = 0.7), and Sphaerium (r = 0.6) were most strongly

correlated with Axis 1. Axis 2 was most related to dissolved oxygen (r = 0.5),

Nanocladius (r = 0.6), Parametriocnemus (r = 0.6), Bezzia (r = 0.6), and Erioptera (r =

0.5). Axis 3 was correlated with dissolved oxygen (r = 0.6), flow (r = 0.5), leaf fall (r = -

0.5), Cryptochironomous (r = 0.7), Polypedilum (r = 0.5), Conchepelopia (r = 0.6),

Alluaudomyia (r = 0.6), Simulium (r = 0.6), Sphaerium (r = 0.8), and Tanytarsus (r = 0.7).

For watersheds A and B, drought impacts separated along axis 3 of the NMDS, with

positive values leading to a recovery from disturbance. Harvest effects separated along

Axis 1, with positive values indicating the harvest induced disturbance.

Only one chironomid species, Parachaetocladius, was a significant species for

pre-harvest samples. By contrast, seven species were significant indicators for reference

streams after harvest. These were primarily predators or those consuming organic matter

and included Alotanypus, Caecidiota, Corethrella, Crangonyx, Ptilostomis, Sciomyzidae,

and Stenochironomus. Thirteen species were indicators for the thinned SMZ treatment.

They occupied a range of trophic habits and included Ablabesmyia, Calopteryx,

Cheumatopsyche, Cryptochironomous, Habrophlebiodes, Hemerodromia, Orthocladius,

Paralauterborniella, Peltodytes, Sphaerium, Stenelmis, Tanytarsus, and Theinemaniella.

Indicator species reflective of the intact SMZ treatment were primarily predators and

shredders including, Anisocentropus, Procladius, and Hexatoma (Table 3-4).

Watersheds C (Harvested) and D (Reference). NMDS ordination (stress = 12.9,

P = 0.001) explained 88 % of variance in the dataset, with 39 %, 31 % and 18 %

explained by Axes 1, 2, and 3, respectively. Overall, the ordination indicated separation









in: Proceedings of the 12th Biennial Southern Silviculture Research Conference, pp. 161-165.
Biloxi, MS. 24-28 February 2003. Southern Research Station: USDA Forest
Service, Asheville, NC.

Jonsen, I.D. and L. Fahrig. 1997. Response of generalist and specialist insect herbivores to
landscape spatial structure. Landscape Ecology 12:185-197.

Jonsen, I.D. and P.D. Taylor. 2000. Fine-scale movement behaviors of Calopterygid damselflies
are influenced by landscape structure: an experimental manipulation. Oikos 88:553-562.

Justic,D., N.N. Rabalais, R.E. Turner, and W.J. Wiseman. 1993. Seasonal coupling between
riverborne nutrients, net productivity and hypoxia, Marine Pollution Bulletin 26:184-189.

Kareiva, P.M. and N. Shigesada. 1983. Analyzing insect movement as a correlated random walk.
Oecologia 56:234-238.

Kareiva, P. 1990. Population dynamics in spatially complex environments: theory and data.
Philosophical Transactions of the Royal Society of London : B. 330:175-190.

Kareiva, P., and U. Wennergren. 1995. Connecting landscape patterns to ecosystem and
population processes. Nature 373:299-302.

Kedzierski, W.M. and L.A. Smock. 2001. Effects of logging on macroinvertebrate production in
a sand-bottomed, low-gradient stream. Freshwater Biology 46:821-833.

Kemp, M.J. and W.K. Dodds. 2001. Centimeter-scale patterns in dissolved oxygen and
nitrification rates in a prairie stream. Journal of the North American Benthological Society
20:347-357.

Kiffney, P.M., J.S. Richardson, and J.P. Bull. 2003. Responses of periphyton and insects to
experimental manipulation of riparian buffer width along forest streams. Journal of Applied
Ecology 40:1060-1076.

Kirchman, D.L. 1993. Statistical analysis of direct counts of microbial abundance. In: P.F.
Kemp, B.F. Sherr, E.B. Sherr, and J.J. Cole (eds.). Handbook of Methods in Aquatic Microbial
Ecology, pp. 117-119. Lewis Publishers. Boca Raton.

Knoepp, J. D. and W.T. Swank. 1993. Site preparation burning to improve southern Appalachian
pine-hardwood stands: Nitrogen responses in soil, soil water and streams. Canadian Journal of
Forest Research. 23:2263-2270.

Kolkwitz, R. and M. Marsson. 1909. 0 kologie der tierischen Saprobien. International Revue
der gesamten Hydrobiologie und Hydrographie 2:126-152.














O Ludwigia
0.2
Mixed
0.1 Liriodendron



S-0.1-

2 -0.2
-0.3



-0.5

-0.6

-0.7

-0.8
10 20 30
Percent Cover


Figure 5-2. Average deviation from a correlated random walk (+ SE) (CRW) (Rdiff) for
Anisocentropus.









bodies and less than one generation per year (vl, r = -0.8) were negatively related to Axis

1. Axis 2 was correlated with N03/NO2 (r = 0.4). Small individuals (sl, r = 0.5) lacking

resistance to desiccation (d2, r = 0.8) with more than one generation per year (v3, r =

0.5), are common in drift (df2, r = 0.7), prefer fast flowing water (r3, r = 0.7) and cling to

substrates (hl, r = 0.8) were positively related to Axis 2. Medium-sized (s2, r = -0.7)

individuals rare in drift (dfl, r = -0.6) resistant to desiccation (dl, r = -0.8) with one

generation per year (v2, r = -0.5) that prefer slow flowing water (r2, r = -0.5), are

predators (tr5, r = -0.6), and burrow into the substrate (h2, r = -0.8) were negatively

related to Axis 2.

Indicator traits in the first two years included scrapers/herbivores with hard shells

or cases that climb on substrate. Those in the third year included genera with less than

one generation per year and not resistant to desiccation. The last two years following the

drought were represented by predators and skaters preferring fast flowing water.

Discussion

Few studies have attempted to dissect the functional and structural responses of

aquatic communities to a severe, unpredictable drought event (Boulton and Lake, 1992;

Wood and Petts, 1999; Wright et al., 2001; Churchel and Batzer, 2007). Studies on the

impacts of short-term wet and dry season cycles have provided insight into predictable

climatic variation, primarily in Mediterranean and arid climates (Gasith and Resh, 1999;

Acuna et al., 2005; Beche, McElravy and Resh, 2006; Bonada et al., 2006). In this study,

regional precipitation indices (SPI) were good predictors of temporal changes in both

taxonomic composition and biological trait structure in perennial streams. Communities

became more stable over time and were significantly more stable in wet, rather than dry,

years. Temporal changes in community composition and trait structure resulted in a









Elwood, J.W., J. D. Newbold, A. F. Trimble, and R. W. Stark. 1981. The limiting role of
phosphorus in a woodland stream ecosystem: Effects of P enrichment on leaf decomposition and
primary producers. Ecology 62:146-158.

Epler, J.H. 1995. Identification manual for the larval Chironomidae (Diptera) of Florida. FL
Department of Environmental Protection, Tallahassee, FL.

Epler, J.H. 1996. Identification Manual for the Water Beetles of Florida (Coleoptera:
Dryopidae, Dytiscidae, Elmidae, Gyrinidae, Haliplidae, Hydraenidae,
Hydrophilidae, Noteridae, Psephenidae, Ptilodactylidae, Scirtidae). State of Florida
Department of Environmental Protection Division of Water Facilities. Tallahassee,
Florida. 228 pp.

Erman, N. A. 1986. Movements of self-marked caddisfly larvae, Chyranda centralis
(Trichoptera, Limnephilidae) in a Sierran spring stream, California, U.S.A. Freshwater Biology
16:455-464.

Escudero, A. and J.M. del Arco. 1987. Ecological significance of the phenology of leaf
abscission. Oikos 49:11-14.

Essafi, K., H. Chergui, E. Pattee, and J. Mathieu. 1994. The breakdown of dead leaves buried in
the sediment of a permanent stream in Morocco. Archiv Fur Hydrobiologie 130:105-112.

Fagan, W. E., R. S. Cantrell, and C. Cosner. 1999. How habitat edges change species
interactions. American Naturalist 153:165-182.

Fagan, W.E. 2002. Connectivity, fragmentation, and extinction risk in dendritic metapopulations.
Ecology 83:3243-3249.

Fagan, W.F., M-J. Fortin, and C. Soykan. 2003. Integrating edge detection and dynamic
modeling in quantitative analyses of ecological boundaries Bioscience 53:730-738.

Fahrig, L. and G. Merriam. 1985. Habitat patch connectivity and population survival. Ecology
66:1762-1768.

Fahrig L. and G. Merriam. 1994. Conservation of fragmented populations. Conservation Biology
8:50-59.

Fahrig, L., and J. Paloheimo. 1988. Determinants of local population size in patchy habitats.
Theoretical Population Biology 34:194-213.

Fenoglio, S., P. Agosta, T. Bo, and M. Cucco. 2002. Field experiments on colonization and
movements of stream invertebrates in an Apennine river (Visone, NW Italy). Hydrobiologia
474:125-130.














*01-02
SA02-03
03-04
.*04-05
A" 05-06
A-* 306-07


I S P 4.I










0.15









D+
Sa


a:
D.15





1.1+ +,













.-0.2 D D2
,,Ik





Axis I








Figure 2-11. NMDS ordinations of biological traits in site-year space and trait-space for SF.
Time periods are indicated by different symbols. Ordination plots of taxa are based on
weighted-averaging.
-0.15










weighted-averaging.









Tolonen, K.T., H. Hamalainen, I.J. Holopainen, K. Mikkonen, and J. Karjalainen. 2003. Body
size and substrate association of littoral insects in relation to vegetation structure. Hydrobiologia
499:179-190.

Towns, D.R. 1985. Limnological characteristics of a South Australian intermittent stream,
Brown Hill Creek. Australian Journal of Marine and Freshwater Research 36:821-837.

Towns, D.R. 1991. Ecology of leptocerid caddisfly larvae in an intermittent South Australian
stream receiving Eucalyptus litter. Freshwater Biology 25:117-129.

Townsend, C.R. 1989. The patch dynamics concept of stream community ecology. Journal of the
North American Benthological Society 8:36-50.

Townsend, C.R. and A.G. Hildrew A.G. 1994. Species traits in relation to a habitat templet for
river systems. Freshwater Biology 31:265-275.

Townsend, C.R., S. Doledec, and M. Scarsbrook. 1997. Species traits in relation to temporal and
spatial heterogeneity in streams: a test of the habitat templet theory. Freshwater Biology 37:367-
387.

United States Department of Agriculture (USDA). 1939. Soil Survey of Decatur County.
United States Department of Agriculture, Soil Conservation Service, Albany, GA.

United States Environmental Protection Agency. 1999. Update of ambient water quality criteria
for ammonia. EPA 822-R-99-014. Office of Water, US Environmental Protection Agency,
Washington, DC.

United States Environmental Protection Agency. 2000. National water quality inventory: 1998
report to Congress. EPA 841-R-00-001. Washington, D.C.: U.S. Environmental Protection
Agency.

United States Environmental Protection Agency. 2003. Nonpoint-source pollution: The nation's
largest water quality problem. Washington, D.C.: U.S. Environmental Protection Agency, Office
of Water, Nonpoint Source Control Branch. www.epa.gov/owow/nps/facts/pointl.htm.

Vannote, R. L., G. W. Minshall, K. W. Cummins, J. R. Sedell, and C. E. Cushing. 1980. River
Continuum Concept. Canadian Journal of Fisheries and Aquatic Sciences 37:130-137.

Viera, N.K.M, N.L. Poff, D.M. Carlisle, S.R. Moulton, M.L. Koski, and B.C. Kondratieff 2006.
A Database of Lotic Invertebrate Traits for North America. U.S. Geological Survey Data Series
187. US Geological Survey, US Department of the Interior, Reston, Virginia. (Available from:
http://pubs.usgs.gov/ds/ds187/)

Vowell, J.L. Using stream bioassessment to monitor best management practice effectiveness.
Forest Ecology and Management 143:237-44.









Temporal Variation and Successional Patterns in Taxonomic Abundance

Broad-scale measures of macroinvertabrate communities were responsive to

temporal changes in environmental conditions. Overall abundance was more responsive

to short-term environmental variation than were taxa richness or stability. Taxa richness

was lowest immediately following drought, peaking in the 2004-2005 sampling period.

Changes in taxa richness have been linked to disturbance history in relation to short-term

and long-term droughts (Beche, 2006). As flow regimes recover, more favorable

conditions exist including increased habitat availability and heterogeneity, higher

dissolved oxygen levels, and dilution of nutrients.

A set of nine core taxa existed in both streams immediately following drought

forming a regional species pool adapted to extreme conditions. Most either have multiple

generations per year or a desiccation resistant stage. Crayfish typically respond to

drought by creating deeper burrows, which may also provide other species a refuge from

receding water levels (Boulton, 1989). The chironomid genus, Polypedilum makes

cocoons to resist periods of drying (Hinton, 1960). The presence of coleopteran adults

and hemipterans in WF immediately following drought reflect their ability to survive

outside of the stream and rapidly colonize via aerial dispersal (Ortega et al., 1991;

Wissinger, 1997).

Changes in taxonomic composition were linked to both local- and large-scale

environmental variables. Dissolved oxygen and pH were most linked to changes in

taxonomic conditions for local variables, while long-term 12- and 48-month precipitation

most influenced overall changes in taxonomic composition in WF. Low pH values in WF

excluded entire taxonomic groups, including Ephemeroptera and Plecoptera. Dissolved

oxygen is often posited to be a controlling variable for invertebrate communities in










Table 3-5. Indicator values for watersheds C and D based on taxonomic composition. Groups
are defined as pre-harvest all sites (1), post-harvest reference (2), post-harvest thinned
SMZ (3), and post-harvest intact SMZs (4).


Parachaetocladius
Hexatoma
Leptophlebia
Corynoneura
Thienemaniella
Nippotipula
Pseudolimnophila
Bezzia
Alluaudomyia
Dixella
Psychoda
Sciomyzidae
Ophiogomphus
Cordulegaster
Amphinemura
Perlesta
Allocapnia
Anisocentropus
Helichus
Neoporus
Microvelia
Paracladopelma
Brillia
Cambaridae
Elimia
Cryptochironom us
Polypedilum
Stenochironomus
Tanytarsus
Tribelos
Hexagenia
Molanna
Triaenodes
Laevapex


Group
1
1
1
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
3
3
3
3
4
4
4
4
4
4
4
4
4


Indicator Value
31.2
31.6
35.7
34.2
44.3
39.5
34
36.4
35.4
48.9
50
37.5
54.9
42.6
36.7
42.5
38
37.1
41.8
36.1
46.7
41.1
34.8
45.6
38.4
33.5
30.4
42.1
33.6
36.2
34.6
24.2
27.6
39.8


p-value
0.023
0.034
0.019
0.049
0.003
0.002
0.002
0.001
0.038
0.001
0.001
0.010
0.000
0.005
0.027
0.006
0.018
0.028
0.009
0.035
0.002
0.004
0.006
0.001
0.005
0.022
0.039
0.008
0.026
0.019
0.033
0.031
0.043
0.005










Holt, R.D. 1996. Temporal and spatial aspects of food web structure and dynamics. In: Food
Webs: Contemporary Perspectives, G. Polis and K. Winemiller, eds. Chapman and Hall. pp. 255-
257.

Hooper, D.U., F.S. Chapin, J.J. Ewel, A. Hector, P. Inchausti, S. Lavorel, J.H. Lawton, D.M.
Lodge, M. Loreau, S. Naeem, B. Schmid, H. Seta, A.J. Symstad, J. Vandermeer, and D.A.
Wardle. 2005. Effects of biodiversity on ecosystem functioning: a consensus of current
knowledge. Ecological Monographs 75:3-35.

Howe, M. J. and K. Suberkropp. 1994. Effects of isopod (Lirceus sp.) feeding on aquatic
hyphomycetes colonizing leaves in a stream. Arch. Hydrobiol. 130:93-103.

Hughes, J. M., S. E. Bunn, D. A. Hurwood, S. Choy, and R. G. Pearson. 1996. Genetic
differentiation among populations of Caridina zebra (Decapoda: Atyidae) in tropical rainforest
streams, northern Australia. Freshwater Biology 36:289-296.

Huryn, A.D. and M.W. Denny. 1997. A biomechanical hypothesis explaining upstream
movements by the freshwater snail Elimia. Functional Ecology 11:472 483.

Hynes, H. B. N. 1960, The biology of polluted waters: Liverpool University Press, Liverpool,
UK.

Hynes, H.B.N. 1975. The stream and its valley. Verh. Internat. Verein Limnol. 19:1-15.

Ims, R. A. 1995. Movement patterns related to spatial structures. pp. 85-109 In Hansson, L.,
Fahrig, L. and Merriam, G. (eds.). Mosaic landscapes and ecological processes. Chapman and
Hall, London.

J. K. Jackson, E. P. Mcelravy, and V. H. Resh. 1999. Long-term movements of self-marked
caddisfly larvae (Trichoptera: Sericostomatidae) in a California coastal mountain stream.
Freshwater Biology 42:525-536.

Jackson, C. R., C. A. Sturm, and J. M. Ward. 2001. Timber harvest impacts on small headwater
stream channels in the coast ranges of Washington. Journal of the American Water Resources
Association 37:1533-1549.

Johnson, A.R., B.T. Milne, and J.A. Wiens. 1992. Diffusion in fractal landscapes: simulations
and experimental studies of Tenebrionid beetle movements. Ecology 73:1968-1983.

Johst, K., R. Brandl, and S. Eber. 2002. Metapopulation persistence in dynamic landscapes: the
role of dispersal distance. Oikos 98:263-270.

Jones D.G., W.B. Summer, M. Miwa, and C.R. Jackson. 2003. Baseline characterization
of forested headwater stream hydrology and water chemistry in southwest Georgia.









leaves. However, invertebrates in the thinned SMZ were represented by species preferring to

live in sand, highlighting the increased isolation of patches apparent in these reaches.

At the microhabitat scale, macrophyte patches were more complex, stable, and trapped

higher quantities of organic matter; attracting more diverse invertebrate communities than leaf

packs. Shredders were more common in large leaf packs and scrapers more common in large

macrophytes. This reflected the higher biomass of chlorophyll a in macrophytes and bacteria in

leaf packs. This was supported during a behavioral study utilizing a habitat specialist and

generalist where the availability of both macrophytes and leaf packs was preferred by both

groups and decreased emigration rates from landscapes. Increased diversity of habitats created

by harvest potentially balanced the effects of habitat fragmentation and isolation.

Evidence from this study indicates that properly managed riparian zones effectively

maintain water quality in small coastal plain streams. However, managers should consider the

consequences of reducing habitat specialists and its potential effects on food-web structure.




























To those who helped me balance my life. To my wife Ann, for her adventurous spirit and her
attempts to reduce my carbon footprint. To Leif for the changes he will inspire. To my mother
for the gift of learning, teaching, compassion, and independence. To my family for their support
and sense of home that will never fade.










Gambi, M. C., A. R. M. Nowell, and P. A. Jumars. 1990. Flume Observations on Flow Dynamics
in Zostera-Marina (Eelgrass) Beds. Marine Ecology-Progress Series 61:159-169.

Garman, G. C. and J. R. Moring. 1993. Diet and annual production of two boreal river fishes
following clearcut logging. Environmental Biology of Fishes 36:301-311.

Gasith, A. and V.H. Resh. 1999. Streams in mediterranean climate regions: abiotic influences
and biotic responses to predictable seasonal events. Annual Review of Ecology and Systematics
30:51-81.

Gates, J.E., amd L.W. Gysel. 1978. Avian nest predation and fledgling success in field-forest
ecotones. Ecology 59:871-883.

Gelhaus, J.K. 2002. Manual for the identification of aquatic crane fly larvae for
southeastern United States. Prepared for the carolina area benthological workshop. Durham,
North Carolina. 57 pp.

Georgia Forestry Commission. 1999. Georgia's best management practices for forestry. 71 pp.

Gilliam, J. F., and D. F. Fraser. 2001. Movement in corridors: Enhancement by predation threat,
disturbance, and habitat structure. Ecology 82:258-273.

Golladay S.W., J.R. Webster, and E.F. Benfield E.F. 1987. Changes in stream morphology and
storm transport of seston following watershed disturbance. Journal of the North American
Benthological Society 6:1-11.

Golladay, S. W., and C. L. Hax. 1995. Effects of an engineered flow disturbance on meiofauna in
a north Texas prairie stream. Journal of the North American Benthological Society 14:404-413.

Golladay, S.W. and J.B. Battle. 2002. Effects of flooding and drought on water quality in
Gulf Coastal Plain streams in Georgia. Journal of Environmental Quality 31:1266-1272.

Gomi, T., R. C. Sidle, and J. S. Richardson. 2002. Understanding processes and downstream
linkages of headwater systems. BioScience 52:905-916.

Gregg, W.W. and F.L. Rose. 1982. The effects of aquatic macrophytes on the stream
microenvironment. Aquatic Botany 14:309-324.

Gurtz, M.E. and J.B. Wallace. 1984. Substrate-mediated response of stream invertebrates to
disturbance. Ecology 65:1556-1569.

Guttman, N.B. 1999. Accepting the standardized precipitation index: a calculation algorithm.
Journal of the American Water Resources Association 35:311-322.









with the number of zero-flow days reaching 20-50 year recurrence levels and the Flint

River displaying record low daily flows (Barber and Stamey, 2000).

The current study (late 2001 to 2007) occurred during a period of average

precipitation, with slightly above-average SPI values for months 3 and 12 (i.e., 0.13, SD

= 1.02 and 0.14, SD = 1.00, respectively) and a slightly below-average SPI value for

month 48 (i.e., -0.33, SD = 1.22). Additionally, hydrographs recorded flow throughout

most of the sampling period (Fig. 2), and the number of zero-flow days progressively

decreased over time in both streams, indicating a period of stream recovery. However,

SPI values in 2006-2007 indicate a return to a drought period, an observation supported

by occurrence of a substantial drought in Georgia in 2007-2008.

Environmental Variables

Although highly variable, environmental stability was relatively high throughout

the study, with Bray-Curtis values ranging from 0.03 to 0.15 (Fig. 3). Most

environmental parameters fluctuated over time regardless of changes in precipitation or

discharge (Table 2), however, some parameters changed significantly with time.

Ammonia remained low throughout most of the study, but doubled in the third year in

both streams (F5,41 = 2.3, P = 0.05). Values for pH were variable, but were highest

immediately following drought, decreasing thereafter (F5,41 = 4.7, P < 0.01).

Additionally, WF remained more acidic than SF throughout the study. Orthophosphate

decreased over time (F5,41 = 5.4, P = 0.02), but increased again in the 2006-2007

sampling period. In general, conductivity decreased following flow resumption (F5,41 =

2.3, P = 0.05) but increased again during the 2006-2007 sampling period. Temperature

decreased by four degrees over the study period (F5,41 = 5.1, P < 0.001), ranging from









quality and climatic parameters, biological traits and taxonomic composition, and

community stability. Biological traits were expected to respond similarly in the two

streams because they are adjacent headwater streams in the same basin and have access to

the same species pool. Additionally, traits were anticipated to respond primarily to local

environmental variation (e.g., water quality parameters) as a reflection of large-scale

environmental filters. However, changes in regional climatic data are expected to

structure the overall successional pattern of the community.

Materials and Methods

Site Description

The two study streams were located in southwestern Georgia (30049'N /

84037'W), approximately 16 km south of Bainbridge in the Coastal Plain physiographic

province. They lie within the Dry Creek watershed, which discharges to the Flint River

approximately 22 km upstream of the Jim Woodruff Dam of Lake Seminole. Surface

water flow in this basin is lowest from September to November and peaks during January

to April due to higher rainfall and decreased evapotranspiration (Couch et al., 1996).

Streams and rivers in the Coastal Plain receive substantial amounts of groundwater

because they are typically deeply incised into underlying aquifers (Couch et al., 1996).

These streams were first order (width 1.25m), perennial, groundwater-influenced, low

to medium gradient, with sand-dominated substrate (DsoWF = 0.54mm, DsoSF =

0.71mm). The wetland-fed stream (WF) has a broader, flatter valley floor with several

lateral wetlands and drained a catchment of 26.2 ha with a gradient of 1.96%. The seep-

fed stream (SF) was more incised with a steeper, v-shaped valley, a 43.9 ha drainage

basin and a 2.11% gradient (Summer et al., 2003). Both watersheds are forested with WF

dominated by Nyssa biflora, Liriodendron tulipifera, Pinus taeda, and Quercus alba, and











14
-A- Leaf Packs
u 12 -- Ludwigia
U,

S10


-. 8

u,
S6


S4-
.0
,Q

,- 2


0
November 2005 January 2006 April 2006 June 2006


Figure 4-7. Volume weighted invertebrate density (Individuals/cm3) (+ SE) in each patch type.













li---- Fi* .L
A. F W,

SHanw Cia "
1 I Ihi- H- ..Y-t H ?I -
.
....--

II .



-,. '





.. --. ,









Heck, K.L. and G.S. Westone. 1977. Habitat complexity and invertebrate species richness and
abundance in tropical seagrass meadows. Journal of Biogeography 4:135-142.

Heino J., H. Mykra, J. Kotanen, and T. Muotka. 2007. Ecological filters and variability in stream
macroinvertebrate communities: do taxonomic and functional structure follow the same path?
Ecography 30:217-230.

Hepinstall, J.A. and R.L. Fuller. 1994. Periphyton reactions to different light and nutrient levels
and the response of bacteria to these manipulations Archiv fur Hydrobiologie. 131:161-173.

Herlihy, A.T., W. J. Gerth, J. Li, and J. L. Banks. 2005. Macroinvertebrate community response
to natural and forest harvest gradients in western Oregon headwater streams
Freshwater Biology 50:905-919.

Hershey, A. E. and S. I. Dodson. 1985. Selective predation by a sculpin and a stonefly on 2
Chironomids in laboratory feeding trials. Hydrobiologia 124:269-273.

A.E. Hershey, J. Pastor, B.J. Peterson, and G.W. Kling. 1993. Stable isotopes resolve the drift
paradox for Baetis mayflies in an arctic river, Ecology 74:2315-2325.

Hildrew, A. G. And P. S. Giller.1994. Patchiness, species interactions and disturbance in stream
benthos. Pages 21-61 in PS. Giller, A.G. Hildrew, and D.G. Rafaelli (editors). Aquatic ecology:
scale, pattern and process. Blackwell Scientific, London.

Hill, W.R. and A. W. Knight 1988. Nutrient and light limitation of algae in two northern
California streams. Journal of Phycology 24:125-132.

Hill, W.R., M.G. Ryon, and E.M. Schilling. 1995. Light limitation in a stream ecosystem:
responses by primary producers and consumers. Ecology 76:1297-1309.

Hillebrand, H. 2002. Top-down versus bottom-up control of autotrophic biomass: a meta-
analysis on experiments with periphyton. Journal of the North American Benthological
Society 21:349-369.

Hinton H.E. 1960. Cryptobiosis in the larva of Polypedilum vanderplanki Hint. (Chironomidae).
Journal of Insect Physiology 5:286-300.

Hoeinghaus, D. J., K.O. Winemiller, and J.S. Birnbaum. 2007. Local and regional determinants
of stream fish assemblage structure: Inferences based on taxonomic vs. functional groups.
Journal of Biogeography 34:324-338.

Hogg, I.D. and D. D.Williams 1996. Response of stream invertebrates to a global-warming
thermal regime: An ecosystem-level manipulation Ecology 77:395-407.

Holt, R.D. 1977. Predation, apparent competition and structure of prey communities. Theoretical
Population Biology 12:197-229.









debris dams, and the addition of this habitat was preferred to landscapes with only leaf packs by

a specialist detritivore (Anisocentropuspyraloides). This situation may be unique to coastal plain

streams, where fine substrate is often entrained in storm events, creating a dynamically changing

landscape. This is in stark contrast to mountain streams with higher substrate diversity and

stability in the form of boulders and cobble.

Testing of any best management practice ultimately requires an understanding of

mechanisms behind changes in stream communities, as well as long-term monitoring data.

Results from this study provide information on the mechanisms leading to apparent improved

water quality in streams impacted by logging. Thus, additional effort should be placed on

developing assessments specific to coastal plain streams, since most are based upon expected

habitat diversity and channel structure found in Piedmont streams.










Table 4-2. Multiple regressions for Ludwigia averaged over all time periods for the observational
study.

Dependent Variable Parameter Estimate SE t P
Invertebrate Abundance (F4,61=3.5, P = 0.008, R2 = 0.24)
Size 0.15 0.17 0.9 0.37
Chlorophyll a 20.6 13.9 1.6 0.12
Bacteria Abundance 0.12 0.15 0.17 0.44
Bacteria Biomass -0.04 0.13 -0.28 0.78
FPOM 0.28 0.1 2.8 0.007
Taxon Richness (F4,63=2.7, P = 0.03, R2 = 0.19)
Size 0.11 0.08 1.5 0.14
Chlorophyll a 9.7 5.8 1.7 0.1
Bacteria Abundance -0.02 0.07 -0.28 0.78
Bacteria Biomass -0.04 0.06 -0.67 0.51
FPOM 0.08 0.04 1.9 0.06
Scrapers (F4,63=2.6, P = 0.04, R2 = 0.20)
Size 0.62 0.25 2.5 0.01
Chlorophyll a -42.7 19.3 -2.2 0.03
Bacteria Abundance -0.56 0.22 -2.5 0.01
Bacteria Biomass 0.08 0.2 0.39 0.7
FPOM -0.21 0.15 -1.4 0.16
Filterers (F4,63=2.9, P = 0.02, R2 = 0.20)
Size -0.11 0.21 -0.52 0.61
Chlorophyll a 31.9 16.6 1.9 0.06
Bacteria Abundance 0.35 0.19 1.8 0.07
Bacteria Biomass -0.17 0.17 -0.99 0.33
FPOM 0.29 0.12 2.3 0.02
Collector-gatherers (F4,63=4.5, P = 0.002, R2 = 0.28)
Size 0.25 0.21 1.22 0.23
Chlorophyll a -46 15.9 -2.9 0.005
Bacteria Abundance -0.29 0.18 -1.57 0.11
Bacteria Biomass 0.25 0.16 1.55 0.13
FPOM -0.29 0.12 -2.4 0.02









CHAPTER 1
INTRODUCTION

Overview of forestry practices in southeastern U.S.

Managed forests practices comprise a significant land area within the U.S., thus their

proper management has broad scale consequences for biodiversity and ecosystem functions.

Previous disregard for these ecosystems resulted in loss of nearly 120 million hectares of

forested land in the U.S. from 1630-2005, of which 40 million was lost in the southeastern U.S.

(Alvarez, 2007). Currently, approximately 59 % of land in the southeastern U. S. is forested,

with 98% managed for timber (Alvarez, 2007), representing more than 10% of timberland in the

U.S. In Georgia alone, there are 9.5 million hectares of commercial forest land, comprising an

area covering nearly 67% of the state (Georgia Forestry Commission, 1999). Additionally, the

Coastal Plain is extremely productive, with the fastest pine growth rates in the country, thus

attracting forestry operations. (Demmon, 1951).

Historically, logging has occurred along rivers and streams, in part to facilitate

downstream transport of timber, with little regard for preserving stream habitat or biota.

However, following enactment of the Clean Water Act in 1972, land managers recognized the

importance of protecting water quality. In recent years, nonpoint-source (NPS) pollution has

become one of the greatest threats to U.S. water quality as point sources were eliminated or

controlled (USEPA, 2003). Silviculture accounts for 5,900 km of impaired rivers and streams in

the U.S. and is ranked 9th of the 10 leading sources of nonpoint pollution of rivers and streams in

the South (West, 2002). Currently, two percent of all assessed stream kilometers (7% of all

impaired kilometers) are considered degraded through forestry activities (US EPA, 2000). In

addition, 53 % of the freshwater supply, originates on forestlands (e.g., headwaters) (Alvarez,











0.2

0.18

0.16

S0.14

0 0.12

0.1 -

S0.08
0
O
a-
L 0.06

0.04

0.02

0 --- ------
Day 7 Day 15 Disturbed Day 15 Undisturbed


Figure 4-14. Average amount of fine particulate organic matter (g) ( SE) trapped in each patch
based on disturbance.









streams in Oregon, Herlihy et al.(2005) also found that environmental variation was a

stronger driver of changes in taxonomic composition than logging history. Further

support exists for the short-tem impact of harvest on streams. Kreutzweiser et al.(2005)

only found an initial peak in scrapers and filterers immediately following harvest in

watersheds with selective harvest. They also found that taxonomic structure differed

among headwater streams with similar characteristics within the same basin, providing

further support for the use of biological traits in bioassessment. Given the predicted

increase in natural disturbances, the value of these indices becomes questionable for

detecting anthropogenic disturbances. However, they have been used successfully for

detecting large disturbances such as urbanization and agricultural practices.

Describing and understanding variability in stream systems is difficult because

processes and patterns vary at different spatial and temporal scales (Wiens et al., 1986;

Roth et al., 1996; Allan and Lammert, 1999). Assemblages can vary at small spatial

scales, yet appear stable, or at least resilient, at larger scales (Rahel, 1990). This

phenomenon has been referred to as the shifting mosaic, steady-state model (Clark, 1991;

Moloney and Levin, 1996). The study streams were exposed to two press disturbances

and at least one pulse disturbance over a decade. The former included a drought lasting

from 1998-2002, logging in 2003, and a hurricane in 2004. The enhanced discharge

resulting from the storms did not influence taxonomic composition or trait structure.

However, the impacts of harvest and drought, discussed here and in Chapter 2, indicate

that natural variability needs to be taken into account when attempting to link changes in

land use to changes in structural and functional aspects of aquatic ecosystems.









Upstream movement has been proposed as one part of the solution to the drift paradox,

whereby species need to recolonize upstream habitats to account for downstream drift.

Displacement along the longitudinal axis is of particular interest to stream ecologists, partly

because of its relevance to concepts such as Muller's colonisation cycle and the paradox of

upstream- downstream movement (e.g., Muller, 1982; Hershey et al., 1993; Anholt, 1995). In

essence, there needs to be a balance between downstream movement (both passive and active)

and upstream migration by larvae and adults to maintain position in suitable stream habitats (e.g.,

Elliott, 1971b; Soderstrom, 1987). In this study, most Elimia individuals (90 %) moved

upstream, regardless of landscape type, suggesting that this is a compensation mechanism for

potential disturbances such as floods. However, Anisocentropus did not exhibit a significant

displacement direction. Although this species spent much of its time attempting to move

upstream, the shape of its case made it susceptible to downstream drift.

Field studies of up- versus downstream movement of individually marked invertebrates

(i.e. at larger spatial and temporal scales than this study) provide contrasting results with regard

to directional movement. Among cased caddisflies, Jackson et al.(1999) recorded no directional

bias in net displacement at low discharge, but there was a downstream bias at higher discharges;

Erman (1986) reported some seasonal dependence but, generally, a net downstream

displacement. However, Hart and Resh (1980) reported no bias in net displacement direction, but

did not report displacement distance along the longitudinal axis. As in this study, upstream

displacement occurs commonly in snails (Schneider and Lyons, 1993; Huryn and Denny, 1997).

Although species capable of upstream flight, such as stoneflies, tend to move downstream

(Freilich, 1999).









more surface area for bacteria and fungi, thus providing more food for invertebrates.

Additionally, since FPOM is easily flushed from habitats during storm events, higher FPOM

may indicate greater stability of the patch, providing more reliable habitat for invertebrates.

Ludwigia patches trapped the most CPOM and FPOM in the short-term experimental study.

This ultimately increased diversity of niches available to invertebrates and improved suitability

for colonization. Since macrophytes are anchored in sediment, they may act like debris dams,

trapping and holding organic matter during storm events. Thus, macrophytes have the potential

to take over some of the function of woody debris typically absent in logged streams. Although

macrophytes became abundant following logging, inputs of pine needles will likely increase over

the next decade since the watershed was planted with a monoculture of pine. As expected, pine

patches created the least heterogeneity and trapped little if any organic matter. Many timber

operations in the southern U.S. utilize pine plantations, which could have a negative impact on

invertebrates by decreasing structural complexity and overall storage of organic matter.

Patch Stability

In addition to structural complexity, habitat stability plays a large role in determining the

composition of patch inhabitants. Although the southern coastal plain does not typically receive

high-energy flows such as those present in snow-melt, relatively large events may occur during

hurricanes and smaller events with storm events common during summer. Thus, more stable

habitats are likley to be more attractive to invertebrates. Stability provided by Ludwigia

enhanced colonization by filtering invertebrates. Additionally, sandy-bottomed streams in the

coastal plain do not provide relatively immobile substrates such as cobble and boulders present

in the piedmont. Thus, invertebrates depend on availability of organic substrate introduced from

the riparian zone or growing within the stream including woody debris, rootwads, macrophytes,

and leaf packs. However, leaf packs are ephemeral, rapidly decomposing, and are subject to










Robinson, C.A., T. Thom, and M. Lucas. 2000. Ranging behaviour of a large freshwater
invertebrate, the white-clawed crayfish Austropotamobiuspallipes. Freshwater Biology 44:509
-521 .

Roitberg, B. D. and M. Mangel. 1997. Individuals on the landscape: behavior can mitigate
landscape differences among habitats. Oikos 80:234-240.

Rooke, J.B. 1984. The invertebrate fauna of four macrophytes in a lotic system. Freshwater
Biology 14:507-513.

Rooke, J.B. 1986. Macroinvertebrates associated with macrophytes and plastic imitations in the
Erasoma River, Ontario, Canada. Archiv fur Hydrobiologie 106:307-325.

Rosemond, A.D., S.R. Reice, J.W. Elwood, and P.J. Mulholland. 1992. The effects of stream
acidity on benthic invertebrate communities in the south-eastern United States Freshwater
Biology 27:193-209.

Rossi, L. 1985. Interactions between invertebrates and microfungi in freshwater ecosystems.
Oikos 44:175-184.

Roth, N.E., J.D.Allan, and D. L. Erickson. 1996. Landscape influences on stream biotic integrity
assessed at multiple spatial scales Landscape Ecology 11:141-156.

Rounick, J.S. and M. J. Winterbourn. 1983. The formation, structure and utilization of stone
surface organic layers in two New Zealand streams Freshwater Biology 13:57-72.

Russell, R. E., R. K. Swihart, and Z. Feng. 2003. Population consequences of foraging decisions
in a patchy matrix. Oikos 103:142-152.

Sagova-Mareckova, M. and J. Kvet. 2002. Performance of Sparganium emersum Rehm. shoots
in response to sediment quality. Hydrobiologia 479:131-141.

Sand-Jensen, K. 1998. Influence of submerged macrophytes on sediment composition and near-
bed flow in lowland streams. Freshwater Biology. 39:663-679.

Sand-Jensen, K. and O. Pedersen. 1999. Velocity gradients and turbulence around macrophyte
stands in streams. Freshwater Biology 42:315-328.

Sartory, D.P. and J.E. Grobbelaar. 1984. Extraction of chlorophyll a from freshwater
phytoplankton for spectrophotometric analysis. Hydrobiol. 114:177-187.

SAS Institute Inc. 2002. SAS OnlineDoc 9.1.3, Cary, NC: SAS Institute Inc.

Schneider, D.W. and J. Lyons. 1993. Dynamics of upstream migration in two species of tropical
freshwater snails Journal of the North American Benthological Society 12:3-16. .









Field Experiment

The goal of the field experiment was to control for leaf species, patch size, and patch age

to examine initial macroinvertebrate colonization patterns. Leaf packs consisted of dominant tree

species shared among watersheds B and C; Liriodendron tulipifera, Quercus nigra, and Pinus

spp. Leaves were collected in August 2006 prior to abcission and air dried for seven days.

Macrophytes were collected from seeps along the stream, washed thoroughly in distilled water,

and examined for invertebrates and biofilm before use. The macrophytes and leaf species were

used to create patches of 1, 2, or 4 g. A separate set often macrophyte samples were dried at

60C to determine a wet to dry mass regression and create an equivalent to the leaf packs prior to

the beginning of the experiment.

The three size classes were crossed with two levels of stability and four species in a

randomized block design. Blocks were created in a 10-20 m stretch of stream and replicated

three times along at 70 m length of each reach in the intact and thinned SMZ treatments in

watersheds B and C. Leaf packs were created by loosely tying leaves together using nylon line.

Macrophyte patches were anchored in the sediment using mesh produce bags (10" Vexar bags,

Avis Bag Co.). Leaf packs and macrophytes were tethered to pvc pipe driven into the streambed.

Stable patches were left undisturbed for 15 days, while unstable patches were disturbed once on

day 7 by rinsing the patch through the water column for one minute. Patches were then collected

after 7 and 15 days to determine colonization patterns.

Velocity, oxygen, and canopy cover were measured at each patch as potential

determinants of patch quality. Canopy cover was measured by taking four measurements using a

densitometer. Velocity was measured using a Marsh McBirney Flowmate 2000 (Frederick,MD).

Oxygen samples were taken by first removing a 10ml water sample from the patch with a 10 ml









Bronmark, C. 1985. Freshwater snail diversity: effects of pond area, habitat heterogeneity, and
isolation. Oecologia 67: 127- 131.

Brosofske, K.D., J. Chen, R.J. Naiman, and J.F. Franklin. 1997. Effects of harvesting on
microclimatic gradients from streams to uplands in western Washington, USA. Ecological
Applications 7: 1188-1200.

Brown, A. V., Y. Aguila, K. B. Brown, and W. P. Fowler. 1997. Responses of benthic
macroinvertebrates in small intermittent streams to silvicultural practices. Hydrobiologia
347:119-125.

Buesing, N. 2005. Bacterial counts and biomass determination by epifluorescence microscopy in
M.A.S. Graca, F. Barlocher and M.O. Gessner (eds.), Methods to Study Litter Decomposition: A
Practical Guide, pp. 203 208. Springer, Netherlands.

Butcher, R.W. 1933. Studies on the ecology of rivers I. On the distribution of macrophytic
vegetation in the rivers of Britain. Journal of Ecology 21: 58-91.

Cadenasso, M.L., M. M. Traynor, S.T.A. Pickett. 1997. Functional location of forest edges:
Gradients of multiple physical factors. Canadian Journal of Forest Research 27: 774-782.

Cain, M. L. 1985. Random search by herbivorous insects: a simulation model. Ecology 66: 876-
888.

Campbell, I.C. and T.J. Doeg. 1989. Impact of timber harvesting and production on streams: a
review. Australian Journal of Marine and Freshwater Research 40: 519-539.

Campbell, I.C. and L. Fuchshuber. 1995. Polyphenols, condensed tannins, and processing rates
of tropical and temperate leaves in an Australian stream. Journal of the North American
Benthological Society 14: 174-182.

Caraco, N.F. and J.J. Cole. 2002. Contrasting impacts of a native and alien macrophyte on
dissolved oxygen in a large river. Ecological Applications 12: 1496-1509.

Caruso, B.S. 2002. Temporal and spatial patterns of extreme low flows and effects on stream
ecosystems in Otago. New Zealand. Journal of Hydrology 257: 115-133.

Charvet S., B. Statzner, P. Usseglio-Polatera, and B. Dumont. 2000. Traits of benthic
macroinvertebrates in semi-natural French streams: an initial application to biomonitoring in
Europe. Freshwater Biology 43: 277-296.

Chevenet F., S. Doledec, and D. Chessel. 1994. A fuzzy coding approach for the analysis of
long-term ecological data. Freshwater Biology 31: 295-309.

Churchel M.A. and D.P. Batzer. 2006. Recovery of aquatic macroinvertebrate communities from
drought in Georgia piedmont headwater streams. American Midland Naturalist 156: 259-272.











0.7


0.6
C-
E

0 0.5
0)

S0.4


.2 0.3
CD



i \

0.1


0
November 2005 January 2006 April 2006


Figure 4-3. Bacterial biomass (pg C/cm3) ( SE) in each patch type.


June 2006












0.35


0.3


S 0.25


0.2


0.15

a-
LL 0.1


0.05


0 -
Pinus Liriodendron Ludwigia Quercus




Figure 4-15. Average amount of fine particulate organic matter (g) (+ SE) trapped in each patch
type.













45

40 -

35
C
0 30
C
25 25

0 20

S15

S10

5

0
1 2 4
Mass


Figure 4-17. Average number of invertebrate individuals (+ SE) in each patch based on initial
patch mass.
































143











80
-A- Leaf Packs
70 --- Ludwigia
Co
0
60
X

S50


40 -
40







0
"u20


0
I -


November 2005 January 2006 April 2006 June 2006


Figure 4-2. Total number of bacterial cells (1 X 106) ( SE) in each patch type.









Physical and Biological Measurements

Eight 50m sample reaches, two per watershed, were established 30.8 m

upstream of hydrology flumes. Three transects were established perpendicular to the

thalweg within each reach at 15, 30, and 45m to serve as in-stream data collection points

for physical measurements including channel cross-sections, canopy cover, and percent

cover of in-stream habitat. A survey of habitat unit and channel characteristics was

conducted longitudinally within established macroinvertebrate sample reaches once

before harvest (December 2001) and once after (October 2004). A 50 m fiberglass tape

was placed in the thalweg of the stream, along which boundaries between habitat unit

types (riffle, run, glide, pool, backwater pool, step, and undercut bank) were determined

and physical characteristics recorded. A backwater pool was defined as being slower and

deeper than a glide but lacking characteristics of a pool, such as scouring, deposition, and

presence of a deep section followed by a shallow tail downstream (i.e. measurable

residual pool depth). For each unit type, length, wetted width, and maximum water depth

were recorded. For steps and pools, a step height and residual pool depth were taken. The

length and diameter of channel obstructions (e.g., wood, roots) were recorded when the

object was primarily responsible for pool formation. The number of functional (e.g.,

ability to change stream morphology) and non-functional wood pieces greater than 10 cm

in diameter was recorded, and texture of the streambed (e.g., sand, silty-sand) was

visually assessed.

Habitat data were converted into percent cover, to define major habitat types to be

sampled for macroinvertebrates. Canopy photos were taken at each transect once before

and once after harvest with a digital camera fitted with a 180 hemispherical fisheye lens

to calculate % canopy cover.















-2- Reference
-A-Thinned
-e- Intact SMZ


50.00



40.00



30.00

Z

20.00



10.00



0.00



Figure 3-4


SAverage ammonia (NH4) concentrations (+SE) in reference, thinned SMZs, and
intact SMZ streams. Harvest treatments were applied prior to the third sampling
period.


60.00


-


2001-2002 2002-2003 2003-2004 2004-2005 2005-2006 2006-2007









across large areas via aerial dispersal, while amphipods may persist in refugia by

aestivating in moist sediments or disperse through underground routes (Wiggins et al.,

1980; Harris and Roosa, 2002). In small coastal plain streams, discharge is lower than in

many montane headwater streams, thus streamlined bodies may not be necessary to resist

flow forces. However, within four years, species preferring fast flow, such as clingers,

became more abundant, suggesting that behavioral rather morphological adaptations to

flow may be a distinguishing trait of coastal plain streams.

Drought Prediction

A major hurdle that is often encountered when assessing impacts of extreme,

unpredictable events on aquatic ecosystems is the availability of data prior to the

disturbance (Lake, 2000; Lake, 2003). Although long-term datasets (>10 years) are

increasingly common in ecology, many geographical regions lack such data. Long-term

datasets allow for the prediction of drought via precipitation indices, changes in stability

and trait composition. The SPI utilized in this study indicated a trend toward a major

drought prior to the occurrence of such an event (Fig. 1). However, the 24-month SPI

may bemost relevant since most invertebrate life cycles require months to years. Values

fell toward 2001 levels in 2006-2007, and a drying period was evident after the last

sampling period. This observation is further supported by the occurrence of a severe

drought in the same region during 2007-2008 (U.S. Drought Monitor,

http://www.drought.unl.edu/dm/archive.html). Thus, use of long-term regional

precipitation data, which is more widely available than discharge and ecological data,

may provide a unique opportunity to study pre- and post-recovery aspects of extreme

events.










Table 3-4. Indicator values for watersheds A and B based on taxonomic composition. Groups
are defined as pre-harvest all sites (1), post-harvest reference (2), post-harvest thinned
SMZ (3), and post-harvest intact SMZs (4).


Parachaetocladius
Alotanypus
Caecidiota
Corethrella
Crangonyx
Ptilostomis
Sciomyzidae
Stenochironomus
Ablabesmyia
Calopteryx
Cheumatopsyche
Cryptochironom us
Habrophlebiodes
Hemerodromia
Orthocladius
Paralauterborniella
Peltodytes
Sphaerium
Stenelmis
Tanytarsus
Thienemaniella
Anisocentropus
Hexatoma
Procladius


Group
1
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
3
3
3
4
4
4


Indicator Value
37.5
44.3
39.5
27.8
38.6
43.7
44.7
39.8
44.2
36.6
26.4
41.8
35.5
25
30.3
42.3
30
44.8
49.6
34.5
30.3
32.4
34.2
37.1


p-value
0.007
0.000
0.012
0.040
0.000
0.001
0.005
0.008
0.004
0.012
0.042
0.008
0.020
0.045
0.039
0.005
0.016
0.004
0.001
0.025
0.040
0.020
0.016
0.019


- -- -- -









respectively. For example the three-month index for November 2002 is the average of

August, September, and October 2002.

Water temperature was measured from October 2001 through February 2007 with

an Onset HOBO temperature logger (Pocasset, MA), programmed to record

temperature every 15 minutes. Water chemistry and meteorological measurements have

been collected by other investigators as part of the Dry Creek Study, and these data were

available for use in this study. Monthly in-situ measurements for dissolved oxygen,

specific conductance, temperature, pH, and turbidity were made at eight sites (two per

stream) with portable meters. Grab samples were taken from a midstream location and

analyzed for inorganic nitrogen, inorganic phosphorus, and ammonium. Specific details

of data collection and sample analysis are in Jones et al.(2003). Values were In (X+1)

transformed prior to analysis to normalize data.

Invertebrate Sampling

Benthic macroinvertebrates were collected from four sample reaches (two per

stream, separated by 50 meters) with a 500-tm-mesh D-frame net (0.3 m wide) in

December and February for six consecutive years beginning December 2001, which

marked return of flowing water in both streams. Twenty samples (- 0.5 m) were taken

from each reach for a total of 3.1 m2 area sampled from all available habitats and were

combined into a single sample. Samples were preserved in 95% ethanol and identified to

genus using regional and national keys(Pescador et al., 1995; Epler, 1995;1996; Merritt

and Cummins, 1996; Pescador et al., 2000; Gelhaus, 2002; Richardson, 2003).

Chironomid larvae were quantitatively subsampled, mounted and identified following

Epler (1995) and Merritt and Cummins (1996).
















0.14


0.12


0.1 -




0.04






0.02


0
Figure 2-3.
0 -






Figure 2-3.


-m- WF
-e-SF


01-02 02-03 03-04 04-05 05-06 06-07


Temporal variability of Bray-Curtis stability values for environmental variables in
WF and SF ( SE).









Ranney, J.W., Bruner, M.C., and Levenson, J.B. 1981. The importance of edge in the structure
and dynamics of forest islands. In: R.L. Burgess and D.M. Sharpe, eds. Forest island dynamics in
man-dominated landscapes. Springer-Verlag, New York, pp. 67-95.

Rawer-Jost, J., J. Bohmer, J. Blank, and H. Rahmann. 2000. Macroinvertebrate functional
feeding groups in biological assessment. Hydrobiologia 422/423:225-232.

Rempel, R.S. and J.C.H. Carter. 1986. Experimental study on the effect of elevated temperature
on the heterotrophic and autotrophic food resources of aquatic insects in a forested stream.
Canadian Journal of Zoology 64:2457-2466.

Resh V.H., A.V. Brown, A.P. Covich, M.E. Gurtz, H.W. Li, G.W. Minshall, S.R. Reice, A.L.
Sheldon, J.B. Wallace, and R.C. Wissmar. 1988. The role of disturbance in stream communities.
Journal of the North American Benthological Society 7:433-455.

Reice, S. R. 1974. Environmental patchiness and the break-down of leaf litter in a woodland
stream. Ecology 55:1271-1282.

Reice, S. R. 1991. Effects of detritus loading and fish predation on leafpack breakdown and
benthic macroinvertebrates in a woodland stream. Journal of the North American Benthological
Society 10:42-56.

Remer, L. C. and S. B. Heard. 1998. Local movement and edge effects on competition and
coexistence in ephemeral-patch models. American Naturalist 152:896-904.

Richards, C. and G. W. Minshall. 1988. The influence of periphyton abundance on Baetis
bicaudatus distribution and colonization in a small stream. Journal of the North American
Benthological Society 7:77-86.

Richards, C., G. E. Host, and J. W. Arthur. 1993. Identification of Predominant Environmental-
Factors Structuring Stream Macroinvertebrate Communities within a Large Agricultural
Catchment. Freshwater Biology 29:285-294.

Richards C., R.J. Haro, L.B. Johnson, G.E. Host. 1997. Catchment and reach-scale properties as
indicators of macroinvertebrate species traits. Freshwater Biology 37:219-230.

Richardson, J.S. 2003. Identification Manual for the Dragonfly Larvae (Anisoptera) of
Florida. Tallahassee, Florida. 114 pp.

Rier, S.T.and R.J. Stevenson. 2002. Effects of light, dissolved organic carbon, and inorganic
nutrients on the relationship between algae and heterotrophic bacteria in stream periphyton.
Hydrobiologia 489:179-194.

Roberts, C.R. 2002. Riparian tree associations and storage, transport, and processing of
particulate organic matter in a subtropical stream. PhD Dissertation, University of Florida,
Gainesville,FL. 98 pp.










0.012
-A- Leaf Packs
-- Ludwigia
0.01


0.008


0.006

0
0.004


0.002


0
November 2005 January 2006 April 2006 June 2006


Figure 4-1. Total biomass of chlorophyll a (mg) (+ SE) in each patch type.









species using local and regional keys (Pescador et al., 1995; Epler, 1995;1996; Merritt

and Cummins, 1996; Pescador et al., 2000; Gelhaus, 2002; Richardson, 2003).

Biological Traits

Fourteen biological traits were selected to characterize body morphology (size,

body shape, body armoring, respiration), life history (voltinism, resistance to desiccation,

eggs cemented to substrate, and development and hatch times), mobility (occurrence in

drift), and ecology (rheophily, behavior, feeding preferences, microhabitat preference)

(Table 3-1) to delineate responses to changes in disturbance regime (Poff et al., 2006).

Some desired traits were omitted due to the lack of available information (e.g.,

fecundity), particularly for chironomid genera. The fourteen biological traits were divided

into 49 modalities ranging from two to seven levels per trait. Trait information was

collected from literature (Viera et al., 2006), as well as through communication with

taxonomic experts in the United States. Traits were coded and analyzed as in Chapter 2.

Data Analysis

Energy sources

Changes in leaf fall C:N ratios, periphyton, and BOM with time and treatment

were analyzed with repeated measures ANOVA (SAS Institute, 2002). Seasonal impacts

of harvest on periphyton biomass were assessed by grouping time periods into by wet or

dry seasons. The wet season was defined as May September, while the dry season was

October April. Multiple regressions were utilized to relate changes in BOM and

chlorophyll a in relation to environmental variables for each treatment. To reduce

impacts of multicollinearity on the regression model, Pearson's correlations were

calculated for each pair of environmental variable, and values greater than 0.6 were

removed. This resulted in pH and orthophosphate being removed from the dataset.









Droughts act on local stream variables by concentrating nutrients and organic

matter, and potentially increasing temperature (Closs and Lake, 1995; Stanley, Fisher and

Grimm, 1997; Matthews, 1998; Golladay and Battle, 2002; Dahm, 2003). A decrease in

o-phosphate following drought reflected flushing of stored nutrients during increased

flow periods (Dahm, 2003). Ammonia peaked in the third year in both streams, reflecting

increased microbial activity and organic matter. Massive amounts of organic matter are

typically stored in the stream channel and floodplain during drought. Initial flushes from

early flow events may not have been enough to carry organic matter from the floodplain

into the stream, however, a large flow event in the spring prior to the third sampling

period likely made a large amount of organic matter available. Baldwin (2005) also

suggested that peaks in ammonia following drying events might have originated from

dead bacterial cells. As in other studies, a coupled decrease in water temperature and

increase in discharge led to higher overall dissolved oxygen values (Stanley, Fisher and

Grimm, 1997; Matthews, 1998; Golladay and Battle, 2002). Conductivity also decreased

with time, reflecting dilution of concentrated ions typically found during drought

(Stanley, Fisher and Grimm, 1997; Caruso, 2002; Line et al., 2006) and may have been

linked to greater contribution of groundwater versus surface flow found during dry

periods (Rider and Belish, 1999; Caruso, 2002).

Limited precipitation and water availability in the riparian zone altered local

landscape dynamics. Leaf fall within the riparian zone decreased following the drought,

indicating a recovery period from the drought, as trees often drop their leaves during

periods of moisture deficit (Escudero and del Arco, 1987). Although this may provide

more resources for invertebrates, leaf quality may be lower and thus limit decomposition.









with global warming. Journal of Geophysical Research 107:4379-4394.


Wiens, J. A. 1976. Population responses to patchy environments. Annual Review of Ecology and
Systematics 7:81-120.

Wiens, J. A., J. T. Rotenberry, and B. Vanhorne. 1985. Territory size variations in shrubsteppe
birds. Auk 102:500-505.

Wiens, J., J. Addicott, T. J. Case, and J. Diamond. 1986. Overview: The importance of spatial
and temporal scale in ecological investigations. Pages 145-153 In J. Diamond and T. J. Case
(editors).

Wiens, J. A., N. C. Stenseth, B. Vanhorne, and R. A. Ims. 1993. Ecological mechanisms and
landscape ecology. Oikos 66:369-380.

Wiens, J. A., R. L. Schooley, and R. D. Weeks. 1997. Patchy landscapes and animal movements:
Do beetles percolate? Oikos 78:257-264.

Wiggins, G.B. 1996. Larvae of the north american caddisfly genera. 457 pp. University of
Toronto Press.

Wiggins B., R.J. Mackay, and I.M. Smith. 1980. Evolutionary and ecological strategies of
animals in annual temporary pools. Archiv fur Hydrobiologie Supplement 58:97-206.

Wilcove, D. S. 1985. Nest predation in forest tracts and the decline of migratory songbirds.
Ecology 66:1211-1214.

Williams, D.D. 1987. The Ecology of Temporary Waters. Timber Press, Portland, Oregon.

Williams, D.D. 1996. Environmental constraints in temporary fresh waters and their
consequences for the insect fauna. Journal of the North American Benthological Society 15:634-
650.

Williams, T. M., D. D. Hook, D. J. Lipscomb, X. Zeng, and J. W. Albiston. 1999. Effectiveness
of best management practices to protect water quality in the South Carolina Piedmont. In Proc.
Tenth Biennial Southern Silvicultural Research Conference, 357-362. General Tech. Report
SRS-30. T. A. Waldrop, ed. Asheville, N.C.: USDA Forest Service, Southern Research Station.

Wipfli, M.S. and D.P. Gregovich. 2002. Export of invertebrates and detritus from fishless
headwater streams in southeastern Alaska: implications for downstream salmonid production.
Freshwater Biology 47:957-969.

Wissinger, S.A. 1997. Cyclic colonization in predictably ephemeral habitats: a template for
biological control in annual crop systems. Biological Control 10:4-15.









Barlocher, F. and B. Kendrick. 1975. Assimilation efficiency of Gammarus pseudolimnaeus
(Amphipoda) feeding on fungal mycelium or autumn-shed leaves. Oikos 26: 55-59.

Bastian, M., L. Boyero, B. R. Jackes, and R. G. Pearson. 2007. Leaf litter diversity and shredder
preferences in an Australian tropical rain-forest stream Journal of Tropical Ecology 23: 219-229

Bayley, S.E., R.S. Behr, and C.A. Kelly. 1986. Retention and release of sulphur from a
freshwater wetland. Water, Air, and Soil Pollution 31:101-114.

Beasley, R. S., and A. B. Granillo. 1982. Sediment losses from forest practices in the Gulf
Coastal Plain of Arkansas. In Proc. Second Biennial Southern Silviculture Research Conference,
461-467. E. P. Jones Jr., ed. General Tech. Report SE-24. Asheville, N.C.: USDA Forest Service,
Southeastern Forest Experiment Station.

B&che, L. A., E.P. McElravy, and V.H. Resh. 2006. Long-term seasonal variation in the
biological traits of benthic-macroinvertebrates in two Mediterranean-climate streams in
California, U.S.A. Freshwater Biology 51: 56-75.

Bender, D. J., L. Tischendorf, and L. Fahrig. 2003. Using patch isolation metrics to predict
animal movement in binary landscapes. Landscape Ecology 18: 17-39.

Benson, B.J. and J.J. Magnusson. 1992. Spatial heterogeneity of littoral fish assemblages in
lakes: relation to species diversity and habitat structure. Canadian Journal of Fisheries and
Aquatic Sciences 49: 1493-1500.

J.P. Benstead and C.M. Pringle. 2004. Deforestation alters the resource base and biomass
ofendemic stream insects in eastern Madagascar. Freshwater Biology 49: 490-501.

Beschta, R.L. 1978. Long-term patterns of sediment production following road construction and
logging in the Oregon Coast Range. Water Resources Research 14: 1011-1016.

Best, L. B., T. M. Bergin, and K. E. Freemark. 2001. Influence of landscape composition on bird
use of rowcrop fields. Journal of Wildlife Management 65:442-449.

Bider, J. R. 1968. Animal activity in uncontrolled terrestrial communities as determined by a
sand transect technique. Ecological Monographs 38:269.

Bilby, R.E. and P.A. Bisson, 1992. Relative contribution of allochthonous and autochthonous
organic matter to the trophic support of fish populations in clear-cut and old-growth forested
headwater streams. Canadian Journal of Fisheries and Aquatic Sciences 49:540-551.

Bilton, D. T., A. Foggo, and S. D. Rundle. 2001. Size, permanence and the proportion of
predators in ponds. Archiv Fur Hydrobiologie 151:451-458.

Blindow, I. 1987. The composition and density of epiphyton on several species of submerged
sacrophytes the neutral substrate hypothesis tested. Aquatic Botany 29:157-168.









null hypothesis of no difference between two or more apriori defined groups. The test

statistic A describes the degree of within-group homogeneity compared with that

expected by chance. MRPP was based on In (x+1) transformed abundance data and the

Bray-Curtis coefficient. Indicator species analysis (IndVal; Dufrene and Legendre 1997)

was used to identify significant indicator species discriminating among the time periods

for the species composition and biological trait data. IndVal is based on a comparison of

relative abundance and relative frequencies of taxa in different apriori groups. Good

indicator taxa are those occurring at all sites in a given group and never in any other

groups (Dufrene and Legendre, 1997). The indicator value ranges from zero to 100 and is

maximized when all individuals occur within a single group of sites. The significance of

the indicator values for each taxon was tested by Monte Carlo tests with 1000

permutations. All ordinations, MRPP, and indicator species analyses were performed in

PC-Ord ver. 5 (McCune and Mefford, 1999).

Results

Hydrologic and Climatic Patterns

SPI values ranged from -2.54 to 4.29 during the 50 year period from 1956 to

2006 in southwest Georgia (Fig. 1). Values greater than 2 are classified as extremely wet

and values below -2 as extremely dry (Guttman, 1999). Mean values for the 3-, 12-, and

48-month SPI during the 1998-2002 drought were -0.25 (SD = 1.06), -0.29 (SD =

1.42), and 0.55 (SD = 0.86) respectively. The drought prior to the study period (1998-

2002) was the worst of the past 50 years and the third worst of the past 100 years,

exceeded only by droughts from 1930 to 1935 and 1938 to 1944 (Barber and Stamey,

2000). The 1998-2002 drought had serious impacts on streams and rivers in the region,









related increased nitrogen may accelerate leaf litter decomposition, altering organic matter

dynamics and potentially limiting resources available for detritivore populations (Bormann et al.,

1974; Likens et al., 1978; Martin et al., 2000; Swank et al., 2001).

However, these changes tend to be short-lived, with water chemical parameters

recovering within one to two years (Corbett et al., 1978; Martin and Pierce, 1980; Arthur et al.,

1998). Vowell (2001) did not find any change in water chemistry in Florida when Best

Management Practices (BMPs) were utilized nor did Adams (1995) in South Carolina.

However, neither study connected long-term pre or post harvest data, nor did they selectively

harvest within the buffer zone, an acceptable practice in Florida and Georgia (Georgia Forestry

Commission, 1999).

Current Status of Riparian Zone Management in the Southeastern U.S.

Regulations for Stream Management Zone (SMZ) width vary among states, however, most

rely on watershed slope as a predictor of sediment inputs following harvest. Although Georgia

recommends a buffer width for a perennial stream beginning at 12.2 meters (40 feet), with

increases as slope of the adjacent watershed increases (Georgia Forestry Commission, 1999),

current regulations allow for limited harvest within the SMZ. Such harvest, known as thinning or

partial harvesting, may be conducted until either there is a minimum of 11.5 square meters of

basal area per hectare (50 square feet of basal area per acre) or 50% canopy cover remaining.

Aust and Blinn (2004) examined published research on the effects of forest practices on water

quality in the southeastern U.S. for the previous 20 years. They concluded that forestry BMPs

were effective for minimizing potentially negative effects of forest practices on water quality, but

needed to be refined to reflect site specific conditions in the southeast. Impacts of logging on

stream biota have been well studied in high gradient streams in the northwest and the eastern

Appalachians of the U.S., but little emphasis has been placed on small, low gradient streams in









Stout, B. M., E. F. Benfield, And J. B. Webster. 1993. Effects of a forest disturbance on shredder
production in southern Appalachian headwater streams. Freshwater Biology 29:59-69.

Strommer, J.L. and L.A. Smock. 1989. Vertical distribution and abundances of invertebrates
within the sandy substrate of a low-gradient headwater stream. Freshwater Biology 22:263-274.

Suberkropp, K. 1992. Interactions with invertebrates. In F. Barlocher, editor. The ecology of
aquatic Hyphomycetes. Pp.113-134. Springer-Verlag, New York, New York,
USA.

Suberkropp, K. 1998. Effect of dissolved nutrients on two aquatic hyphomycetes growing on leaf
litter Mycological Research 102:998-1002.

Sugihara, G. and R. May. 1990. Nonlinear forecasting as a way of distinguishing chaos from
measurement error in a data series. Nature 344:734-741.

Summer W.B., C.R. Jackson, D.G. Jones, and M. Miwa. 2003. Characterization of hydrologic
and sediment transport behavior of forested headwater streams in southwest Georgia. In:
Proceedings of the 2003 Georgia Water Resources Conference, pp 157-160. Athens, GA. 23-24
April 2003. The Institute of Ecology: The University of Georgia, Athens, GA.

Swank, W.T., J.M. Vose, and K.J. Elliot. 2001. Long-term hydrologic and water quality
responses following commercial clearcutting of mixed hardwoods on a southern Appalachian
catchment. Forest Ecology and Management 143:163-178.

Sweeney, B.W. and R. L. Vannote. 1986. Growth and production of stream stonefly: Influences
of diet and temperature. Ecology 67:1396-1410.

Swift, L.W. Jr and J.B. Messer. 1971. Forest cuttings raise temperatures of small streams in the
southern Appalachians. Journal of Soil and Water Conservation 26:111-116.

Taylor, B.R., D. Parkinson, and W. F. J. Parsons. 1989. Nitrogen and lignin content as predictors
of litter decay rates: A microcosm test. Ecology 70:97-104.

Tett, P., C. Gallegos, M. G. Kelly, G. M. Hornberger, and B. J. Cosby. 1978. Relationships
among Substrate, Flow, and Benthic Microalgal Pigment Density in Mechums River, Virginia.
Limnology and Oceanography 23:785-797.

Tischendorf, L. and L. Fahrig. 2000. How should we measure landscape connectivity?
Landscape Ecology 15: 633-641.

Tokeshi,M. and L.C.V. Pinder. 1985. Microhabitats of stream invertebrates on two submerged
macrophytes with contrasting leaf morphology. Holarctic Ecology 8: 313-319.









Biological Traits

Nine biological traits were selected to characterize body morphology (i.e., size,

body shape, body armoring), life history (i.e., voltinism, resistance to desiccation),

mobility (i.e., occurrence in drift), and ecology (i.e., rheophily, habits, feeding

preferences) (Table 1). These were anticipated to vary in response to changes in

precipitation and display low statistical and phylogenetic dependence (Poff et al., 2006).

Some desired traits were omitted due to the lack of available information (e.g.,

fecundity), particularly within the chironomid genera. The nine biological traits were

divided into 30 modalities ranging from two to six levels per trait. Trait information was

collected from the literature (e.g., Viera et al., 2006), as well as through communication

with taxonomic experts. Trait information was coded at the generic level, except for some

Diptera and non-insect taxa, which were coded at the family or order level, respectively.

Where information on a particular trait could not be obtained for a taxon (in <5 % of

cases), zero scores were entered for each category so it did not influence overall results

(Chevenet, Doledec and Chessel, 1994). Individual taxa were then scored for the extent to

which they displayed the categories of these traits using a 'fuzzy coding' procedure

(Chevenet et al., 1994). Fuzzy coding allows taxa to exhibit trait categories to different

degrees (Chevenet et al., 1994) to take account of intraspecific variations in trait

expression (Charvet et al., 2000). The scoring range of 0 to 3 was adopted, with 0 being

no affinity to a trait category and 3 being high affinity. Traits were rescaled as

proportions (sum = 1), such that for each trait modality, values ranged from 0 (no affinity

among individuals for the modality) to 1 (all individuals had exclusive affinity for the

modality) and modalities summed to 1 for each trait. To describe the functional

composition of communities in terms of density of individuals, the proportion of each




Full Text

PAGE 1

1 RIPARIAN ZONE MANAGEMENT IN COASTAL PLAIN STREAMS: MULTI-SCALE EFFECT S OF HABITAT FRAGMENTATION By MARCUS WAYNE GRISWOLD A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE RE QUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2008

PAGE 2

2 2008 Marcus Wayne Griswold

PAGE 3

3 To those who helped me balance my life. To m y wife Ann, for her adventurous spirit and her attempts to reduce my carbon footprint. To Leif fo r the changes he will inspire. To my mother for the gift of learning, teaching, compassion, and independence. To my family for their support and sense of home that will never fade.

PAGE 4

4 ACKNOWLEDGMENTS I would like to thank my advisor, T.L. Crisman, for his guidance and insight into my research as well as giving me numerous opportunities to expand my knowledge base. I benefited greatly from discussions with my committee B. Bolker, R. Holt, and W. Wise and their experience in a large breadth of disciplines. I am grateful to those who helped me find myself and my mentoring skills throughout this journe y. I thank those who let me pry my way into their research, just to discover something new and to those who reminded me that everyone has something to contribute. Included in this, I thank those who assisted with the fieldwork, and torturous days and nights of sorting: R. Sandidge, M. Dornberg, O. Stern, K. Alvarez, C. Cruz, L. Burhans, M. Diedrick, and M. Bell. I would like to thank Scott Terrell for his willingness to share data and Rebecca Winn for the initiation of the preharvest work.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........8 LIST OF FIGURES.........................................................................................................................9 ABSTRACT...................................................................................................................................12 CHAPTER 1 INTRODUCTION..................................................................................................................14 Overview of forestry practic es in southeastern U.S. ..............................................................14 Buffer Zones and Aquatic Ecosystems................................................................................... 15 Buffer Zones and Water Quality..................................................................................... 15 Current Status of Riparian Zone Management in the Southeastern U.S......................... 17 Habitat Fragmentation and Forestry Practices........................................................................18 2 IMPACTS OF CLIMATIC STABILITY ON THE STRUCTURAL AND FUNC TIONAL ASPECTS OF MACROIN VERTEBRATE COMMUNITIES AFTER SEVERE DROUGHT............................................................................................................. 20 Introduction................................................................................................................... ..........20 Materials and Methods...........................................................................................................22 Site Description...............................................................................................................22 Hydrologic and Environmental Variables....................................................................... 23 Invertebrate Sampling..................................................................................................... 24 Biological Traits.............................................................................................................. 25 Statistical Analysis.......................................................................................................... 26 Environmental variables........................................................................................... 26 Ordination: species co mposition and traits .............................................................. 27 Results.....................................................................................................................................28 Hydrologic and Climatic Patterns................................................................................... 28 Environmental Variables................................................................................................. 29 Benthic Macroinvertebrates............................................................................................. 30 Community succession.............................................................................................30 Community stability.................................................................................................31 Taxonomic Composition................................................................................................. 31 Wetland-Fed stream (WF)............................................................................................... 31 Seep-Fed stream (SF)...............................................................................................32 Biological Traits.............................................................................................................. 33 Wetland-Fed stream................................................................................................. 33 Seep-Fed stream.......................................................................................................34 Discussion...............................................................................................................................35

PAGE 6

6 Environmental Variation................................................................................................. 36 Temporal Variation and Successional Patterns in Taxonom ic Abundance..................... 38 Temporal Variation in Traits........................................................................................... 39 Drought Prediction..........................................................................................................41 3 TESTING BMP EFFECTIVENESS FOR SMALL COASTAL PLAIN STREAMS USING MACROINVER TEBRATES AS BIOINDICATORS .............................................56 Introduction................................................................................................................... ..........56 Materials and Methods...........................................................................................................59 Site Description...............................................................................................................59 Geology....................................................................................................................59 Vegetation................................................................................................................59 Climate..................................................................................................................... 60 Hydrology.................................................................................................................60 Experimental Harvest............................................................................................... 61 Physical and Biological Measurements........................................................................... 62 Physical measurements............................................................................................ 63 Energy sources.........................................................................................................63 Macroinvertebrates...................................................................................................64 Biological Traits....................................................................................................... 65 Data Analysis...................................................................................................................65 Energy sources.........................................................................................................65 Environmental variables........................................................................................... 66 Macroinvertebrates...................................................................................................66 Results.....................................................................................................................................67 Energy Source.................................................................................................................67 Environmental Variables................................................................................................. 68 Macroinvertebrates..........................................................................................................69 Stability.................................................................................................................... 69 Taxonomic composition........................................................................................... 69 Biological traits........................................................................................................ 72 Discussion...............................................................................................................................74 Energy Sources................................................................................................................74 Environmental Variables................................................................................................. 78 Macroinvertebrates..........................................................................................................80 Anthropogenic disturbance in the face of natural disturbances....................................... 82 4 EFFECTS OF PATCH TYPE, QUALI TY, AND SIZE ON MACROINVERTEBRATE COMMUNI TY STRUCTURE ............................................................................................104 Introduction................................................................................................................... ........104 Materials and Methods.........................................................................................................106 Field Sampling of Patches.............................................................................................106 Field Experiment...........................................................................................................109 Data Analysis.................................................................................................................110 Field obervations....................................................................................................110

PAGE 7

7 Experimental manipulation of patches................................................................... 110 Results...................................................................................................................................110 Field Observations......................................................................................................... 110 Field Experiment...........................................................................................................113 Regressions............................................................................................................. 115 Discussion.............................................................................................................................115 Patch Complexity.......................................................................................................... 116 Patch Stability................................................................................................................ 117 Patch Quality.................................................................................................................118 Patch Size..................................................................................................................... .121 5 HABITAT SELECTION IN FRAGMENTED LANDSCAPES: COMPARING GENE RALISTS TO SPECIALISTS ...................................................................................151 Introduction................................................................................................................... ........151 Materials and Methods.........................................................................................................153 Study Organisms...........................................................................................................153 Behavioral Observations............................................................................................... 154 Colonization..................................................................................................................157 Results...................................................................................................................................158 Movement......................................................................................................................158 Anisocentropus.......................................................................................................158 Elimia..................................................................................................................... 158 Colonization..................................................................................................................159 Discussion.............................................................................................................................159 6 CONCLUSIONS.................................................................................................................. 172 LIST OF REFERENCES.............................................................................................................175 BIOGRAPHICAL SKETCH.......................................................................................................206

PAGE 8

8 LIST OF TABLES Table page 2-1 Definition and codes for bi ological traits and m odalities................................................. 43 2-2 Mean annual values for environmental va riables for the wetla nd-fed (WF) and seepfed......................................................................................................................................44 3-1 Biological trait definitions and modalities........................................................................ 86 3-2 Results of multiple regressions for ch lorophy ll a biomass and benthic organic matter (BOM). Significance of R2 values is given by ( P < 0.05), ** ( P < 0.01), *** ( P < 0.001).................................................................................................................................87 3-3 Average environmental conditions for wi nter sampling periods in reference (A,D), thinned SMZs (B1,C1), and intact S MZs (B2,C2). Data are for pre-harvest (20012003) and post-harvest (2004-2008).................................................................................. 88 3-4 Indicator values for watersheds A and B based on taxonomic composition. Groups are defined as pre-h arvest all sites (1), post-harvest reference (2), post-harvest thinned SMZ (3), and post-h arvest intact SMZs (4).......................................................... 89 3-5 Indicator values for watersheds C and D based on taxonomic composition. Groups are defined as pre-h arvest all sites (1), post-harvest reference (2), post-harvest thinned SMZ (3), and post-h arvest intact SMZs (4).......................................................... 90 3-6 Indicator values for watersheds A and B based on biological traits. Groups are defined as pre-harvest all sites (1), post-h a rvest reference (2), post-harvest thinned SMZ (3), and post-harvest intact SMZs (4)....................................................................... 91 3-7 Indicator values for watersheds C an d D based on biological traits. Groups are defined as pre-harvest all sites (1), post-h a rvest reference (2), post-harvest thinned SMZ (3), and post-harvest intact SMZs (4)....................................................................... 92 4-1 Multiple regressions for leaf packs averaged over all time periods for the observational study. ......................................................................................................... 124 4-2 Multiple regressions for Ludwigia averaged over all tim e periods for the observational study.......................................................................................................... 125 4-3 Multiple regressions for the field experiment averaged over all treatments for each invertebrate m etric........................................................................................................... 126

PAGE 9

9 LIST OF FIGURES Figure page 2-1 Plot of Standardized Precipitation Inde x (SPI) values for southwestern Georgia, from 1956 to 2007......................................................................................................................45 2-2 Hydrograph based on mean daily discharge (m3/s) for each stream..................................46 2-3 Temporal variability of Bray-Curtis st ability values for environm ental variables in WF and SF ( SE).............................................................................................................. 47 2-4 Temporal changes in taxon ri chness and invertebrate abundance. .................................... 48 2-5 Changes in compositional stability (BrayCurtis distance) in WF and SF ( SE)............. 49 2-6 Linear regression of SPI valu es versus taxonom ic stability.............................................. 50 2-7 Changes in trait stability (Bray-Cu rtis distance) in W F and SF ( SE)............................. 51 2-8 NMDS ordinations of log10-abundance in site-year space an d taxon-space for W F. Time periods are indicated by different symbols. Ordination plots of taxa are based on weighted-averaging.......................................................................................................52 2-9 NMDS ordinations of log10-abundance in site-year space an d taxon-space for SF. Tim e periods are indicated by different symbols. Ordination plots of taxa are based on weighted-averaging.......................................................................................................53 2-10 NMDS ordinations of biological traits in site-y ear space and trait-space for WF. Time periods are indicated by different symbols. Ordination plots of taxa are based on weighted-averaging.......................................................................................................54 2-11 NMDS ordinations of biological traits in site-y ear space and trait-space for SF. Time periods are indicated by diffe rent symbols. Ordination pl ots of taxa are based on weighted-averaging............................................................................................................55 3-1 Topographic map and aerial photo of the four study watersheds (A-D). .......................... 93 3-2 Average chlorophyll a biomass (SE) during the wet (May-S eptember) and dry season (October-April) from 2004-2008 in refe rence, thinned SMZs, and intact SMZ streams afte r harvest.......................................................................................................... 94 3-3 C:N ratios of leaf fall from the riparian zone in reference and harvested watersheds before (2001-2003) and af ter (2004-2007) harvest. ........................................................... 95 3-4 Average ammonia (NH4) concentrations (SE) in reference, thinned SMZs, and intact SMZ streams. Harvest treatments were applied prior to the third sampling period.................................................................................................................................96

PAGE 10

10 3-5 Stream condition index (SCI) scores (SE ) for reference, thinned SMZs, and intact SMZ stream s. Samples below the red line indicate poor water quality, those above the red line, fair water quality, and those above the blue line, good water quality........... 97 3-6 Taxonomic stability (SE) for reference, thinned SMZs, and intact SMZ stream s.......... 98 3-7 Trait stability (SE) for reference, thinned SMZs, and intact SMZ stream s.....................99 3-8 NMDS of taxonomic composition in watersheds A and B in pre-harves t (1) and in post-harvest reference (2), thinned SMZs (3), and intact SMZ treatments (4)................ 100 3-9 NMDS of taxonomic composition in watersheds C and D in pre-harves t (1) and in post-harvest reference (2), thinned SMZs (3), and intact SMZ treatments (4)................ 101 3-10 NMDS of biological traits in watershe ds A and B in pre-harvest (1) and in postharvest reference (2), thinned SMZs (3), and intact SMZ treatm ents (4)........................ 102 3-11 NMDS of biological traits in watershe ds C and D in pre-harvest (1) and in postharvest reference (2), thinned SMZs (3), and intact SMZ treatm ents (4)........................ 103 4-1 Total biomass of chlorophyll a (m g) ( SE) in each patch type...................................... 127 4-2 Total number of bacterial cells (1 X 106) ( SE) in each patch type............................... 128 4-3 Bacterial biomass (pg C/cm3) ( SE) in each patch type.................................................129 4-4 Number of bacterial cells per cm3 (1 X 106) ( SE) in each patch type.......................... 130 4-5 Chlorophyll a biomass (mg/cm3) ( SE) in each patch type............................................ 131 4-6 Volume-weighted taxon richness (Taxa/cm3) ( SE) in each patch type........................ 132 4-7 Volume weighted inverteb rate density (Individuals/cm3) ( SE) in each patch type...... 133 4-8 Proportion of filtering invertebrates ( SE) in each patch type....................................... 134 4-9 Proportion of leaf mass decomposed ( SE) in relation to patch type and disturbance. 135 4-10 Amount of leaf mass decomposed (g) ( SE) in relation to initial patch m ass............... 136 4-11 CPOM trapped in patches ( SE) in relation to patch size. ............................................. 137 4-12 Average amount of coarse particulate organic matter (g ) ( SE) trapped in each patch type.........................................................................................................................138 4-13 Average amount of coarse particulate orga nic matter (g) ( SE) tr apped in patches by disturbance type. .............................................................................................................. 139

PAGE 11

11 4-14 Average amount of fine particulate organi c m atter (g) ( SE) tra pped in each patch based on disturbance........................................................................................................140 4-15 Average amount of fine particulate organi c m atter (g) ( SE) tra pped in each patch type...................................................................................................................................141 4-16 Average number of invertebrate indi viduals ( S E) in each patch type......................... 142 4-17 Average number of invertebrate individua ls ( S E) in each patch based on initial patch mass..................................................................................................................... ...143 4-18 Average number of taxa ( SE) in each patch in relation to initial patch mass. ............. 144 4-19 Proportion of scrapers ( SE) in each patch based on initial p atch mass........................ 145 4-20 Proportion of shredders ( SE) in each patch based on initial patch mass...................... 146 4-21 Proportion of shredders ( SE) in each patch based on patch type and disturbance. ...... 147 4-22 Proportion of filterers ( SE) in each patch bas ed on initial patch mass......................... 148 4-23 Proportion of filterers ( SE) in each patch type............................................................. 149 4-24 Proportion of collector-gatherers ( S E) in each patch type............................................ 150 5-1 Microlandscape designs us ed in the behavioral and colonization experim ents. Liriodendron leaf packs (brown squares) and Ludwigia macrophyte patches at A) 10 B) 20, and C) 30 percent cover...................................................................................... 165 5-2 Average deviation from a correlated random walk ( SE) (CRW) (Rdiff) for Anisocentropus .................................................................................................................166 5-3 Average probability ( SE) of each turn being in the sam e direction for Anisocentropus .................................................................................................................167 5-4 Average correlation ( SE) between turning angles for Anisocentropus. .......................168 5-5 Average net squared displacement ( SE) of Anisocentropus in microlandscapes......... 169 5-6 Mean step length ( SE) in each landscape for Elimia ...................................................170 5-7 Average deviation ( SE) from a correlated random walk (Rdiff) for Elimia ................171

PAGE 12

12 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy RIPARIAN ZONE MANAGEMENT IN COASTAL PLAIN STREAMS: MULTI-SCALE EFFECT S OF HABITAT FRAGMENTATION By Marcus Wayne Griswold August 2008 Chair: Thomas Crisman Major: Environmental Engineering Sciences Riparian zones filter nutrients, sediment, and pro vide food and habitat for terrestrial and aquatic organisms. Georgias forestry practi ces were evaluated in coastal plain streams by manipulating harvest regimes in headwater stream s. Macroinvertebrate and their food sources were sampled before and after harvest. A drought occurring prior to the study degr aded stream s, depressing invertebrate abundance and diversity. A core set of speci es appeared immediately following drought, displaying short life cycles and resistance to desiccation, allowing for rapid recovery from disturbance. Communities shifted from small, sclero tized individuals abundant in drift, to those that were larger, soft-bodied, and rare in drift, indicating more favorable habitat. In response to harvest, communities shifted from detritivores to herbivores, following shifts in food availability from organic matter to algae and macrophytes. This was most apparent immediately following harvest and followed a trajecto ry of recovery over the next four years. Interestingly, multimetric indices of water qua lity based on macroinvertebrates suggested more favorable conditions in the most disturbed treatme nt. This relates to in creases in food quality, due to an increase in algae and macrophytes, and a decrease in C:N ratios in terrestrially derived

PAGE 13

13 leaves. However, invertebrates in the thinned SMZ were represented by species preferring to live in sand, highlighting the increased isolation of patches appare nt in these reaches. At the microhabitat scale, macrophyte patche s were m ore complex, stable, and trapped higher quantities of organic matter; attracting more diverse invertebrate communities than leaf packs. Shredders were more common in large l eaf packs and scrapers more common in large macrophytes. This reflected the higher biomass of chlorophyll a in macrophytes and bacteria in leaf packs. This was supported during a behavi oral study utilizing a ha bitat specialist and generalist where the availability of both m acrophytes and leaf packs was preferred by both groups and decreased emigration rates from landscap es. Increased diversity of habitats created by harvest potentially balanced the effects of habitat fragmentation and isolation. Evidence from this study indicates that pr operly m anaged riparian zones effectively maintain water quality in small coastal plain st reams. However, managers should consider the consequences of reducing habitat specialists a nd its potential effects on food-web structure.

PAGE 14

14 CHAPTER 1 INTRODUCTION Overview of forestry practices in southeastern U.S. Managed forests practices comprise a signifi cant land area within the U.S., thus their proper m anagement has broad scale consequenc es for biodiversity and ecosystem functions. Previous disregard for these ecosystems resulte d in loss of nearly 120 million hectares of forested land in the U.S. from 1630-2005, of which 40 million was lost in the southeastern U.S. (Alvarez, 2007). Currently, approximately 59 % of land in the southeastern U. S. is forested, with 98% managed for timber (Alvarez, 2007), repr esenting more than 10% of timberland in the U.S. In Georgia alone, there are 9.5 million hect ares of commercial forest land, comprising an area covering nearly 67% of the state (Georg ia Forestry Commission, 1999). Additionally, the Coastal Plain is extremely productiv e, with the fastest pine growth rates in the country, thus attracting forestry operations. (Demmon, 1951). Historically, logging has occurred along rive rs and stream s, in part to facilitate downstream transport of timber, with little regard for preserving stream habitat or biota. However, following enactment of the Clean Wate r Act in 1972, land managers recognized the importance of protecting water quality. In recent years, nonpoint-source (NPS) pollution has become one of the greatest threat s to U.S. water quality as point sources were eliminated or controlled (USEPA, 2003). Silvicul ture accounts for 5,900 km of impa ired rivers and streams in the U.S. and is ranked 9th of the 10 leading sour ces of nonpoint pollution of rivers and streams in the South (West, 2002). Currently, two percent of all assessed stream kilometers (7% of all impaired kilometers) are considered degraded through forestry activities (US EPA, 2000). In addition, 53 % of the freshwater supply, origin ates on forestlands (e.g., headwaters) (Alvarez,

PAGE 15

15 2007), and proper management strategies are necessary to protect loca l and downstream water quality. Buffer Zones and Aquatic Ecosystems Buffer Zones and Water Quality Riparian buffer zones (streamside management zones) are forested areas along stream s meant to protect biotic integrity and water quality. Riparian zone s act as important ecotones for aquatic systems, providing food for aquatic (e.g. organic matter and terrestrial insects) and terrestrial organisms (e.g., emerging aquatic adul ts) (Nakano et al., 1999), shading, temperature regulation, and woody debris; provid ing the basis for invertebrate community structure (Kiffney et al., 2003). Small headwater streams are closely linked to the riparian zone since they are relatively narrow and shaded by forest canopy (Cummins, 1974; Hynes, 1975; Vannote, 1980; Moore and Richardson, 2003). They account for 70-80 % of total watershed area in the U.S. and export organic matter (OM), sediment, prey items, and nutrients downstream (Meyer and Wallace, 2001; Kiffney et al., 2003). Logging and thinning of vegetation in the ripa rian zone reduce detrital input to stream s over time. The extent of this reduction is influenced by the remaining canopy cover in the riparian buffer zone. Decreased canopy cover le ads to increased light and temperature (e.g., Swift and Messer, 1971) in str eam channels, and may increase primary productivity, shifting production from heterotrophic to autotrophic pro cesses (Hartman and Scrivener,1990; Fuchs et al., 2003). In faster high gradient streams this process leads to dominance by algal communities, while in low-gradient, coastal plain streams it results in a mix of ma crophyte and algal growth (Noel et al., 1986; Kedzierski a nd Smock, 2001). This change typically results in increased density, biomass and diversity of macroinvertebrates and can sh ift macroinvertebrate dominance from shredders to grazers (Jackson et al., 2001; Kedzierski and Smock, 2001; Fuchs et al., 2003).

PAGE 16

16 Such a shift in foodweb structure potentially al ters ecosystem function (e.g., decomposition) and higher trophic levels, limiting food availability for detritivores. Watershed-level disturbances alter runoff regi mes and evapotranspiration rates. Logging potentially alters the hyd rologic regime, such th at increased surface runo ff contributes sediment and nutrients to the affected streams. The primary hydrological influence of harvesting and thinning is increased water yield due to decreased evapotranspiration that typically in harvest treatments is 69 to 210 mm/year (Beasley and Granillo, 1982; Williams et al., 1999; McBroom et al., 2002; Grace et al., 2003). This change in hydrology may ultimately homogenize microhabitats and exclude inverteb rates that prefer slow flow. Clearcut watersheds typically have large sedi m ent yields, potentially clogging fish gills and smothering invertebrate habitat. Gurtz and Wallace (1984) found that abundance of many invertebrate taxa in habitats susceptible to sediment deposition (i.e., pools and sandy reaches) declined in a stream draining a recently clear -cut watershed, whereas those taxa in less susceptible habitats (i.e. steep-gradient, boulde r outcrops) increased. Th is emphasizes the need for proper management of riparian zones in co astal plain streams as they are primarily low gradient systems dominated by extensive sandy reaches, with few outcroppings. In addition, creation of buffer zones decreases potential fo r sediment movement by promoting sheet flow rather than channelized flow across the landscape. Harvest related changes in nut rient export affect the abundan ce and diversity of aquatic invertebrates. Macroinvertebrat e abundance m ay initially increase as nutrients fuel algal growth, providing food to a typically resource-limited grazer population. However, Miltner and Rankin (1998) found a negative relationshi p between water quality indices based on macroinvertebrates and increased nutrient concentrations, especially in low order streams. Additionally, harvest-

PAGE 17

17 related increased nitrogen may accelerate leaf litter decompos ition, altering organic matter dynamics and potentially limiting resources available for detritivore populations (Bormann et al., 1974; Likens et al., 1978; Martin et al., 2000; Swank et al., 2001). However, these changes tend to be shortlived, with water che mical parameters recovering within one to two year s (Corbett et al., 1978; Martin and Pierce, 1980; Arthur et al., 1998). Vowell (2001) did not find any change in water chemistry in Florida when Best Management Practices (BMPs) were utilized nor did Adams (1995) in South Carolina. However, neither study connected long-term pre or post harvest data, nor did they selectively harvest within the buffer zone, an acceptable practice in Florida and Georgia (Georgia Forestry Commission, 1999). Current Status of Riparian Zone Man agement in the So utheastern U.S. Regulations for Stream Management Zone (S MZ) width vary am ong states, however, most rely on watershed slope as a predictor of sediment inputs foll owing harvest. Although Georgia recommends a buffer width for a perennial stream beginning at 12.2 meters (40 feet), with increases as slope of the adjacent watershed increases (Georgia Forestry Commission, 1999), current regulations allow for limited harvest within the SMZ. Such harvest, known as thinning or partial harvesting, may be conducted until either th ere is a minimum of 11.5 square meters of basal area per hectare (50 square feet of basal area per acre) or 50% canopy cover remaining. Aust and Blinn (2004) examined published research on the effects of forest practices on water quality in the southeastern U.S. for the previous 20 years. They concluded that forestry BMPs were effective for m inimizing potentially negative effects of forest practi ces on water quality, but needed to be refined to reflect site specific c onditions in the southeas t. Impacts of logging on stream biota have been well stud ied in high gradient streams in the northwest and the eastern Appalachians of the U.S., but little emphasis has been placed on small, low gradient streams in

PAGE 18

18 the southeastern part of the count ry. Furthermore, the effects of partial harvest within SMZs on water quality are not well documented. More research is ncessary to fill in gaps that currently exist regarding BMP effectiveness in the coastal plain and effects of partial harvesting within SMZs. Habitat Fragmentation and Forestry Practices Forestry practices potentially have advers e effects on communities by lim iting dispersal between watersheds, eliminating suitable envi ronmental conditions, and altering predator-prey dynamics. Even with current regulations for str eam water quality, clear cutting of a watershed down to the buffer zone commonly occurs. A lthough this can maintain local biodiversity, dispersal across this newly created, potentially hostile landscape may be difficult for small organisms such as invertebrates and amphibian s (Hughes et al., 1996; Fagan, 2002; Briers et al., 2004). Although distance between watersheds can serve as a template for determining population s tructure and species composition (e.g., Harding, 2003), locally influenced microhabitats may be the strongest drivers of co mmunity structure at th e reach and microhabitat scales. Indirect effects of loggi ng or riparian zone modification lead to changes in microhabitat structure in streams. This wa s clearly demonstrated in affore sted agricultural streams that displayed an 87 % reduction in the leaf litter storage compared to forested streams (Benstead and Pringle, 2004). Similarly, Noel et al.(1986) f ound that 50% of logged st reams were covered by macrophytes, while unlogged reference stream s had only 10% macrophyte cover. Thus, a gradient of tree removal from the riparian zone should change the physical and biotic structure of the stream in a predictable manner. In logged streams, leaf pack formation is ofte n slow, resulting in increased patch isolation and fragm entation. Rooted macrophytes, however become more abundant in logged streams

PAGE 19

19 and are more stable, contributing to a less dynamic streambed landscape. Thus, macrophytes may support more permanent coexisting species, wh ile leaf packs may support more transient, inferior competitors. Col onization of streambeds by macr ophytes, coupled with decreased allochthonous input to logged streams, may alter the av ailability patches for stream biota. Key gaps in the current literature lie primarily within their temporal and spatial scales. Most studies have lim ited data on pre-harvest co nditions in the watershed, especially true of studies in the southern coas tal plain (Smock et al., 2001). Water chemistry and biotic communities may vary significantly on a temporal scale, knowledge of which is required to determine whether changes following logging are re lated to natural or anthropogenically related disturbance. Many studies focus on changes in taxonomic structure of the biotic community. However, changes m ay be linked more to biological traits that are sensitive to changes in habitat structure (e.g., habitat templet sensu Southwood, 1978; Townsend and Hildrew, 1994). Studies also are limited to the reach scale, whereas or ganisms disturbed in the riparian zone may be affected at the microhabitat scale. Thus the objectives of this study were to: 1) Determine the impact of two logging regime s considered acceptable in the Georgia BMP m anual on stream communities through cha nges in water quality, taxonomic and trait composition, and the role of natural varia tion (e.g., drought) on th e recovery process (Chapters 2 and 3). 2) Link changes in habitat-scale, community com position to changes in pa tch availability at the m icrohabitat scale (Chapter 4). 3) Relate small scale patterns of dispersal for ins tream habitat fragme ntation in a habitat specialist and generalist invertebra te species (Chapter 5)

PAGE 20

20 CHAPTER 2 IMPACTS OF CLIMATIC STABILITY ON THE STRUCTURAL AND FUNC TIONAL ASPECTS OF MACROIN VERTEBRATE COMMUNITIES AFTER SEVERE DROUGHT Introduction Natural disturbances regulate community structure and ecosystem function, and thus play a crucial role in shaping aquatic and terrestrial commun ities (Sousa, 1984; Resh et al., 1988). Aquatic ecosystems are especially vulnerable to extreme climatic changes, such as drought, because these disturbances alter flow regimes, water chemistry, and ultimately, the biotic community (Wood and Petts, 1999). The long-term effects of drought on the economy, wildlife habitat, and re creation occur as ramp disturbances over periods of years ( sensu Lake, 2003), as opposed to the e ffects of flooding events, which subside after weeks or months. The frequency and predictability of droughts are generally low. However, when drought does occur, it can potentially act as a destabilizing agent for aquatic communities. The forecast for climat e change suggests increased frequency of extreme events, particularly drought, over the next century (Wetherald and Manabe, 2002; Kundzewicz et al., 2007). Increased intens ity and frequency of natural disturbances will ultimately affect ecosystem stability and influence organisms resistance and resilience to change. During extreme drought, streams typically form a series of disconnected pools and lose evidence of surficial flow over ti me, a response that can potentially reset the aquatic community. Furthermore, toxic accu mulation of nutrients and waste (Towns 1985, 1991; Closs and Lake 1995; Dahm et al., 2003), coupled with increased temperature (Matthews, 1998) and lowered dissolved oxygen (Stanley, 1997; Golladay and Battle, 2002), add stress to the remaining species pools. Species survival after

PAGE 21

21 drought depends on specific life history traits including resistance to desiccation and an ability to colonize habitats rapidly through drift and aeria l migration or oviposition (Williams, 1987, 1996; Boulton, 1989). Further colonization reflect s subsequent changes in water chemistry, habitat availability, and resource base following flow resumption. Biological traits are more informative i ndicators of ecosystem function than are changes in abundance of individual species, and they are expected to change across a gradient of anthropogenic and natural disturba nces (Charvet et al ., 2000; Doledec et al., 1999; Statzner, Hildrew and Resh, 2001). Howeve r, species loss decreas es the ability of ecosystems to resist disturba nces and leads to lowered stability (Hooper et al., 2005). Therefore, an integrative approach should ut ilize both species com position and biological traits to predict community responses to disturbances (Richards et al., 1997) Biological traits are regulated at a hierarchy of scales, with environmental filters (e.g., climate and geology) creating a template for traits that are present in a specific region (Townsend and Hildrew, 1994; Poff, 1997). Thus, a subset of tr aits is expected to respond to disturbances within a certain region. For example, species that are resilient to disturbance display a series of traits, including small size and multip le generations per year, that allow them to expand their population densitie s rapidly (Townsend and Hi ldrew, 1994). As functional redundancy is common among stream invertebrate s, biological traits can be compared across large regions to unde rstand the large-scale impact s of anthropogenic change (Statzner et al., 2004). This study utilized a six-y ear (2001-2007) dataset of m acroinvertebrates from headwater streams after an intense drought in the southeastern U.S. (1998 to 2002) to characterize inter-year succession al patterns following flow restoration relative to water

PAGE 22

22 quality and climatic parameters, biologi cal traits and taxonomic composition, and community stability. Biological traits were expected to respond similarly in the two streams because they are adjacent headwater st reams in the same basin and have access to the same species pool. Additionally, traits we re anticipated to respond primarily to local environmental variation (e.g., water quality pa rameters) as a reflec tion of large-scale environmental filters. However, changes in regional climatic data are expected to structure the overall successiona l pattern of the community. Materials and Methods Site Description The two study streams were located in southwestern Georgia (30' N / 84'W), approximately 16 km south of Bai nbridge in the Coastal Plain physiographic province. They lie within the Dry Creek wate rshed, which discharges to the Flint River approximately 22 km upstream of the Jim Woodruff Dam of Lake Seminole. Surface water flow in this basin is lowest from Se ptember to November and peaks during January to April due to higher rainfall and decreased evapotra nspiration (Couch et al ., 1996). Streams and rivers in the Coastal Plain receive substantial amounts of groundwater because they are typically deeply in cised into underlying aquifers (Couch et al ., 1996). These streams were first order (width ~ 1.25m), perennial, groundwater-influenced, low to medium gradient, with sand-dominated substrate (D50WF = 0.54mm, D50SF = 0.71mm). The wetland-fed stream (WF) has a broader, flatter valley floor with several lateral wetlands and drained a catchment of 26.2 ha with a gradient of 1.96%. The seepfed stream (SF) was more incised with a steeper, v-shaped valley, a 43.9 ha drainage basin and a 2.11% gradient (Summer et al., 2003). Both watersheds are forested with WF dominated by Nyssa biflora Liriodendron tulipifera Pinus taeda and Quercus alba, and

PAGE 23

23 SF dominated by Pinus glabra Fagus grandifolia Liriodendron tulipifera and Quercus nigra Hydrologic and Environmental Variables The climate of the region is character ized by warm, hum id summers, and mild winters. Average temperatures in January, the coldest month of the year, range from 2.8C to 16.3C. July is the hottest month, with average temperatures ranging from 21.5C to 33.5C (SERCC, 2004). Mean annua l precipitation is 1412 mm, with June having the highest mean rainfall (152.1 mm) and October the lowest (77.5 mm) (SERCC, 2004). Summer rains are usually short, with high intensity events giving way to low intensity frontal events from late fall to ear ly spring. Due to close proximity to the Gulf of Mexico, heavy rainfall associated with hurricanes and tropical storms is not unusual in late summer. Drought characteristics were based on regi onal precipitation data and flow data from both study streams. Flow data were obt ained from in-stream parshal flumes and ISCO (Teledyne Isco, Lincoln, NE, USA) sa mplers (Summer et al., 2003) beginning in 2001. A standardized precipitation index (SP I) (McKee et al., 1995) was calculated to assess the frequency and duration of droughts in the region based on monthly precipitation averages for southwest Georgia (National Climatic Data Center, 2007). This index is preferable over the Palmer drought severi ty index because it is easier to interpret, more realistic over the long-term, and does not depend on a normal distribution of precipitation (Guttman, 1999). SPI values less th an one indicate a water deficit, and those above one an excess. SPI values were calculated based on 3-, 12, and 48-month running averages to determine the presence of shor t-term, intermediate, and long-term droughts,

PAGE 24

24 respectively. For example th e three-month index for Nove mber 2002 is the average of August, September, and October 2002. Water temperature was measured from October 2001 through February 2007 with an Onset HOBO temperatu re logger (Pocasset, MA), programmed to record temperature every 15 minutes. Water chemistry and meteorological measurements have been collected by other investig ators as part of the Dry Creek Study, and these data were available for use in this study. Monthly in-situ measurements for dissolved oxygen, specific conductance, temperature, pH, and tu rbidity were made at eight sites (two per stream) with portable meters. Grab samples were taken from a midstream location and analyzed for inorganic nitr ogen, inorganic phosphorus, and ammonium. Specific details of data collection and sample analysis are in Jones et al.(2003). Va lues were ln (X+1) transformed prior to anal ysis to normalize data. Invertebrate Sampling Benthic macroinvertebrates were collected from four sample reaches (two per stream, separated by ~ 50 meters) with a 500m-mesh D-frame net (0.3 m wide) in December and February for six consec utive years beginning December 2001, which marked return of flowing water in both streams. Twenty samples (~ 0.5 m) were taken from each reach for a total of ~ 3.1 m2 area sampled from all available habitats and were combined into a single sample. Samples were preserved in 95% ethanol and identified to genus using regional and national keys(Pescador et al., 1995; Ep ler, 1995;1996; Merritt and Cummins, 1996; Pescador et al., 2000; Gelhaus, 2002; Richardson, 2003). Chironomid larvae were quantitatively subs ampled, mounted and identified following Epler (1995) and Merritt and Cummins (1996).

PAGE 25

25 Biological Traits Nine biological traits were selected to characterize body m orphology (i.e., size, body shape, body armoring), life history (i.e., voltinism, resistance to desiccation), mobility (i.e., occurrence in drift), and ecology (i.e., rheophily, habits, feeding preferences) (Table 1). These were anticip ated to vary in re sponse to changes in precipitation and display low statistical and phylogenetic dependence (Poff et al., 2006). Some desired traits were omitted due to the lack of available information (e.g., fecundity), particularly within the chironomi d genera. The nine biol ogical traits were divided into 30 modalities ranging from two to six levels per trait. Trait information was collected from the literature (e.g., Viera et al., 2006), as well as through communication with taxonomic experts. Trait information was coded at the generic le vel, except for some Diptera and non-insect taxa, whic h were coded at the family or order level, respectively. Where information on a particular trait coul d not be obtained for a taxon (in <5 % of cases), zero scores were entered for each cat egory so it did not influence overall results (Chevenet, Doledec and Chessel, 1994). Individual taxa were th en scored for the extent to which they displayed the categories of thes e traits using a fu zzy coding procedure (Chevenet et al., 1994). Fuzzy coding allows taxa to exhibit trait categories to different degrees (Chevenet et al., 1994) to take acc ount of intraspecific variations in trait expression (Charvet et al., 2000). The scoring range of 0 to 3 was adopted, with 0 being no affinity to a trait category and 3 being high affinity. Traits were rescaled as proportions (sum = 1), such that for each trai t modality, values ranged from 0 (no affinity among individuals for the modality) to 1 (all individuals had exclusive affinity for the modality) and modalities summed to 1 for each trait. To describe the functional composition of communities in terms of dens ity of individuals, the proportion of each

PAGE 26

26 category per trait was multiplied by the invertebrate abundances. This resulted in a traitby-site array that contained the density of i ndividuals for each trait category for each site; density was transformed (ln(x+1)) to approxima te a normal distribution for the statistical analyses. Statistical Analysis Environmental variab les Environmental variables were analyzed over tim e with repeated measures ANOVA (SAS Institute, 2002). When differences we re significant, post-hoc analysis was conducted using Tukeys test and Bonferroni corrections. Additionally, environmental stability was assessed by cal culating Bray-Curtis distan ces (Bray and Curtis, 1957) between adjacent years. Bray-Cur tis distances are a measure of dissimilarity with values ranging from 0 to 1. Zero denotes identical samples; thus, higher values denote lower compositional stability. This measure is computed as: hj ij hjij ihaa aa D || whereihD is the distance between samples i and h. Stability Compositional stability of invertebrate co mm unities was examined separately for the two streams between pairs of successive years. Stabil ity was measured by calculating Bray-Curtis distances between adjacent ye ars based on abundance data and biological traits. ANOVA was used to examine between year differences in compositional and traits stability scores for the streams. The relati onship between Bray-Curtis values and flow and SPI values were regressed to assess the impact of hydrologic scale on community and trait stability.

PAGE 27

27 Ordination: species composition and traits. Nonmetric multidimensional scaling (N MDS; Kruskal, 1964) was used to explore temporal patte rns in species composition and bi ological traits. NMDS is an ordination method based on ranked distances between samples, and it is highly suitable for ecological data that typically contain numerous zero values. First, a distance matrix was constructed using Sorensen's metric s. To reduce the chance of local optima (Legendre and Legendre, 1998; McCune, Grace and Urban, 2002), an initial ordination with 1000 runs was conducted, and the ordinatio n with the lowest stress value was used as the starting configuration for NMDS. Stress is the square root of the ratio of the squared differences between a monotoni c transformation of the calculated dissimilarities/distances and the plotted dist ances and the sum of the plotted distances squared. The number of dimensions retained was evaluated after inspecting the stress (goodness of fit) of solutions with dimensions 1 through 6, with values close to 0 being a good fit of the data. Significance was assess ed by conducting Monte Carlo tests using 999 runs of randomized data in the final ordination. A P-value was calculated as a function of the number randomized runs that re sulted in a stress less th an or equal to the observed stress (McCune and Mefford, 1999). Ordi nations were performed separately for each stream because preliminary analysis indi cated that differences between sites masked any temporal effects. Ordinations were performed on species abundances and abundanceweighted biological tr aits individually. A multi-response permutation procedure (MRPP; McCune and Grace 2002) was used to test f or significant differences in taxonomic composition and biological trait structure over time at each stream. MRPP is a nonparametric method that examines the

PAGE 28

28 null hypothesis of no difference between two or more a priori defined groups. The test statistic A describes the degree of within -group homogeneity compared with that expected by chance. MRPP was based on ln (x+1) transformed abundance data and the Bray-Curtis coefficient. Indi cator species analysis (IndVal; Dufrene and Legendre 1997) was used to identify significant indicato r species discriminating among the time periods for the species composition and biological trait data. IndVal is ba sed on a comparison of relative abundance and relative frequencies of taxa in different a priori groups. Good indicator taxa are those occu ring at all sites in a given group and never in any other groups (Dufrene and Legendre, 1997). The indi cator value ranges from zero to 100 and is maximized when all individuals occur within a single group of sites. The significance of the indicator values for each taxon was tested by Monte Carlo tests with 1000 permutations. All ordinations, MRPP, and indi cator species analyses were performed in PC-Ord ver. 5 (McCune and Mefford, 1999). Results Hydrologic and Climatic Patterns SPI values ranged from .54 to 4.29 duri ng the 50 year period from 1956 to 2006 in southwest Georgia (Fig. 1). Values grea ter than 2 are classified as extremely wet and values below as extremely dry (Guttman, 1999). Mean values for the 3-, 12-, and 48-month SPI during the 1998 drought were .25 (SD = .06), .29 (SD = 1.42), and 0.55 (SD = .86) respectively. Th e drought prior to the study period (1998 2002) was the worst of the past 50 years and the third worst of the past 100 years, exceeded only by droughts from 1930 to 1935 and 1938 to 1944 (Barber and Stamey, 2000). The 1998-2002 drought had serious impacts on streams and rivers in the region,

PAGE 29

29 with the number of zero-flow days reachi ng 2050 year recurrence levels and the Flint River displaying record low daily flows (Barber and Stamey, 2000). The current study (late 2001 to 2007) occu rred during a peri od of average precipitation, with slightly above-average SPI values for m onths 3 and 12 (i.e., 0.13, SD = .02 and 0.14, SD = 1.00, respec tively) and a slightly be low-average SPI value for month 48 (i.e., .33, SD = .22). Additionally, hydrographs recorded flow throughout most of the sampling period (Fig. 2), and th e number of zero-flow days progressively decreased over time in both streams, indica ting a period of stream recovery. However, SPI values in 2006 indicate a return to a drought period, an observation supported by occurrence of a substantial drought in Georgia in 2007. Environmental Variables Although highly variable, environmental stability was rela tively high throughout the study, w ith Bray-Curtis values ra nging from 0.03 to 0.15 (Fig. 3). Most environmental parameters fluctuated over time regardless of changes in precipitation or discharge (Table 2), however, some parame ters changed significantly with time. Ammonia remained low throughout most of th e study, but doubled in the third year in both streams (F5,41 = 2.3, P = 0.05). Values for pH were variable, but were highest immediately following drought, decreasing thereafter (F5,41 = 4.7, P < 0.01). Additionally, WF remained more acidic than SF throughout the study. Orthophosphate decreased over time (F5,41 = 5.4, P = 0.02), but increased again in the 2006 sampling period. In general, conductivity decreased following flow resumption (F5,41 = 2.3, P = 0.05) but increased again during th e 2006 sampling period. Temperature decreased by four degrees over the study period (F5,41 = 5.1, P < 0.001), ranging from

PAGE 30

30 12C to 16 C. Leaf fall peaked in the fi rst year following the drought, but was reduced by 50% the following year (F5,41 = 2.6, P = 0.04). Benthic Macroinvertebrates Community succession Although the two streams differed extensively in term s of successional patterns following drought, a number of species responded similarly at both s ites. A core set of taxa were present throughout the six-y ear sampling period at both sites including Ceratopogonidae (Bezzia ), Chironomidae (Parametriocnemus, Polypedilum Tanytarsus Tribelos, Zavrelimyia ), Decapoda (Cambaridae), Tabanidae ( Chrysops ), and Tipulidae ( Pilaria ). Similarities in the second year included Chironomidae ( Cantopelopia Orthocladiinae) and Ptychopteridae ( Bittacomorpha ), while those in the fourth and fifth year of sampling included Trichoptera ( Lepidostoma ), Hemiptera ( Microvelia ), and Odonata ( Boyeria ). The sixth year was the first year that no additional taxa were found (Appendices 1 and 2). Taxon richness increased significantly over time (F4,30 = 122.73, P < 0.001), primarily during the initial three years of th e study (Fig. 2-4A), but was consistently lower in WF (F1,30 = 56.65, P < 0.001). However, taxon richness saturated with the same number of taxa occurring from the fourth to the sixth year. Te mporal progression of abundance was more humped shaped, decreasing after the fourth year. Invertebrate abundance increased similarly in both streams through time (F4,30 = 8.68, P < 0.001)(Fig. 4B), but WF consistently had signifi cantly fewer individuals than SF (F1,30 = 19.94, P < 0.001).

PAGE 31

31 Community stability Bray-Curtis values for taxonomic com position decreased progressiv ely from 2001 to 2006, with increased stability over time at both sites. Ho wever, Bray-Curtis values decreased during the 2006 to 2007 peri od, indicating a change in community structure to an earlier period (Fig. 2-5). Communities were more stable in wetter than drier periods, as indicated by the negative re lationship between Bray-Curtis values and SPI values (Fig. 2-6). For SF, stability was si gnificantly related to both local and regional hydrologic and climatic indicators, however, stronger relationships existed with flow (R2 = 0.33, P <0.001) and the 48-month SPI (R2 = 0.5, P < 0.001). Only long-term 48-month SPI values were related to stability in WF with a similar negative relationship between SPI values and stability. Biological traits were relatively stable over tim e, and low Bray-Curtis values suggested that traits were more stable than taxonomic composition (Fig. 2-7), while those for WF were only significantly correlated with the intermediate 12-month SPI (R2 = 0.2, P = 0.02). Biological traits for SF were not significantly correlated with local hydrologic or regional climatic variables. Taxonomic Composition Wetland-Fed stream (WF) NMDS ordination (stress = 10.8, P = 0.001) explained 88.4% of the variance in the dataset, with 22%, 36%, and 31% explai ned by Axis 1, 2, and 3 respectively. Overall, the ordination indicated tem poral separati on of species composition (Fig. 2-8) and was supported by significant diffe rences between all time periods (MRPP, A = 0.5, P < 0.001). Axis 1 was primarily represented by local variables including pH (r = .5), dissolved oxygen (r = .5), and turbidity (r = .4). The genera Calopteryx (r = .6), Chironomus

PAGE 32

32 (r = 0.6), Chrysops (r = 0.6), Eukiefferiella (r = 0.7), Lepidostoma (r = 0.6), and Pycnopsyche (r = 0.7) were most strongly correlated with Axis 1. Axis 2 was most related toboth local and large-scale va riables including pH (r = .4) dissolved oxygen (r = 0.4), and the 12-month (r = 0.4) and 48-month SPI (r = 0.7). The genera Agabus (r = 0.7), Boyeria (r = 0.66), Conchepelopia (r = 0.6), Erioptera (r = 0.8), Microtendipes (r = 0.7), and Orthocladius/Cricotopus (r = 0.7) were most strongly related to Axis 2. Axis 3 was correlated with dissolved oxygen (r = .5), and 48-month SPI (r = .5). The genera Caecidiota (r = .6), Microvelia (r = .8), and Smitia (r = .7) were strongly correlated with Axis 3. No significant indicator sp ecies were found in the W F stream for the first three years following drought. Taxonomic indicators of temporal change (P < 0.05) included genera indicative of the fourth year such as Corethrella Stenochironomus Polypedilum Pseudolimnophila Cordulegaster Lepidostoma and Scirtidae. Those having a maximum indicator value for the fifth year we re primarily predators and included Cryptochironomus Alotanypus, Clinotanypus, Bezzia Alluadomyia and Laevapex Those in the last year of the study included Larsia and Erioptera Seep-Fed stream (SF) NMDS ordination (stress = 12.9, P = 0.001) explained 90.2% of the variance in the SF dataset, with 77% and 13% explained by Axes 1 and 2, respectively. Overall, the ordination indicated separation of species com position with time (Fig.9) and was supported by significant diffe rences between all time periods (MRPP, A = 0.5, P< 0.0001). Axis 1 was primarily repres ented by o-phosphate (r = .6), NO2/NO3 (r = .4), dissolved oxygen (r = 0.8), flow (r = 0.6), 3month (r = 0.4) and 48-month (r = 0.7) SPI.

PAGE 33

33 The genera Alluaudomyia (r = 0.6), Bezzia (r = 0.7), Corynoneura (r = 0.7), Stempellinella (r = 0.6), Thienemaniella (r = 0.8), and Zavriella (r = .8) were most strongly correlated with Axis 1. Axis 2 was most related to pH (r = 0.7), dissolved oxygen (r = .5), and leaffall (r = 0.5). The genera Neoporus (r = .7), Parachaetocladius (r = .6), Sphaerium (r = .6), and Pycnopsyche (r = .5) were most strongly related to Axis 2. Significant indicator species for the second year included Allocapnia Helichus Parachaetocladius, and Stenelmis. The third year was m ostly represented by shredders and scrapers including Anisocentropus Cordulegaster Eurylophella Habrophlebiodes Ophiogomphus Pseudolimnophila and Stempellinella The indicators in the fourth year included Baetidae, Psychoda Scirtidae, Tribelos and Zavrelimyia The fifth year was represented by Caenis Calopteryx Diplectrona, Microvelia Nippotipula Rheotanytarsus, and Stenonema The last year of the study was represented by Corynoneura. Biological Traits Wetland-Fed stream NMDS ordination (stress = 11.5, P = 0.001) explained 87.2% of the variance in the dataset, with 53%, 10%, and 25% explained by Axes 1, 2, and 3 respectively. Tem poral changes were supported by overall si gnificant differences between time periods (MRPP, A = 0.3, P < 0.0001) (Fig. 10). However, pairwi se comparisons indicated weak changes in traits over time. Axis 1 was prim arily represented by specific conductance (r = .4) and 12-month SPI (r = 0.5) and thus related to intermediate temporal changes. Soft bodied (ar1, r=0.9), fast flow preferring (r3, r = 0.6), sprawlers (h4, r = 0.6) with bluff and tubular shapes (sh2, r = 0.8) were positiv ely correlated with Axis 1. Sclerotized (ar2,

PAGE 34

34 r = .9), slow flow preferring (r2, r = .6), streamlined traits (sh1, r = .8) were negatively correlated with Axis 1. Axis 2 was negatively correlated with pH (r = .4). Burrowers (h2, r = 0.8) were strongly correlated with Axis 2. Axis 3 was most related to local, short-term variables including t ss (r = .6), dissolved oxygen (r = 0.5), temperature (r = .5), and 3-month SPI (r = 0.4) Collector-gatherers (tr1, r = 0.8) with several generations per year (v3, r = 0.6) were positively related to Axis 3. Those with one generation a year (v2, r = .7), hard shells or cases (a r3, r = .7), and a predatory lifestyle (tr5, r = .8) were ne gatively related to Axis 3 Analysis of indicator species showed that early colonizers had sclerotized, tubular or bluff bodies and are abundant in drift. Late r years were characterized by species that cling to the substra te, rarely drift, and are hard shelled or made a case. Seep-Fed stream NMDS ordination (stress = 7.9, P = 0.001) explained 97% of the variance in the dataset, with 80% and 17% explained by Axes 1 a nd 2, respectively. Overall, the ordination indicated separation of species com positi on with time (Fig. 11) and was supported by significant effects of time on tr ait composition (MRPP, A = 0.4, P < 0.0001). Axis 1 was most related to leaf fall (r = 0.5) and the 48-month SPI (r = 0.4). Small (s1, r = 0.8) softbodied (ar1, r = 0.9) bluff or tubular (sh2, r = 0.8) individual s that gather food (tr1, r = 0.8), with more than one generation per year (v1, r = 0.8), abundant in drift (df1, r = 0.8), and sprawled (h4, r = 0.6) or climbed (h6, r = 0.6) over the substrate were positively related to Axis 1. Shredders (tr4, r = .7) and scrapers/herbivores (tr3, r = -0.7) uncommon in drift (df1, .8) with medium (s2, r = .6) to large (s3, r = .5) streamlined (sh1, r = .8) and with sclerotized (ar2, r = .8 ) or shelled (ar3, r = .8)

PAGE 35

35 bodies and less than one generation per year ( v1, r = .8) were negativ ely related to Axis 1. Axis 2 was correlated with NO3/NO2 (r = 0.4). Small individuals (s1, r = 0.5) lacking resistance to desiccation ( d2, r = 0.8) with more than one generation per year (v3, r = 0.5), are common in drift (df2, r = 0.7), prefer fast flowing wa ter (r3, r = 0.7) and cling to substrates (h1, r = 0.8) were positively related to Axis 2. Medium-sized (s2, r = .7) individuals rare in drift (df1, r = .6) resi stant to desiccation (d1, r = .8) with one generation per year (v2, r = .5) that pref er slow flowing water (r2, r = .5), are predators (tr5, r = .6), and burrow into th e substrate (h2, r = .8) were negatively related to Axis 2. Indicator traits in the first two years incl uded scrapers/herbivores with hard shells or cases that clim b on substrate. Those in th e third year included genera with less than one generation per year and not resistant to desiccation. Th e last two years following the drought were represented by predators and skat ers preferring fast flowing water. Discussion Few studies have attempted to dissect the functional and structural responses of aquatic communities to a severe, un predicta ble drought event (Boulton and Lake, 1992; Wood and Petts, 1999; Wright et al., 2001; Churchel and Batzer, 2007). Studies on the impacts of short-term wet and dry season cycl es have provided insi ght into predictable climatic variation, primarily in Mediterranean and arid climates (Gasith and Resh, 1999; Acuna et al., 2005; Beche, McElravy and Res h, 2006; Bonada et al., 2006). In this study, regional precipitation indices (SPI) were good predictors of temporal changes in both taxonomic composition and biological trait stru cture in perennial st reams. Communities became more stable over time and were significan tly more stable in wet, rather than dry, years. Temporal changes in community co mposition and trait structure resulted in a

PAGE 36

36 rapidly stabilizing community w ithin the first three to four years after drought, producing highly stable and persistent communities in the fourth and fifth years. Stability was maintained throughout the o ccurrence of a large discharge event 4 months before collection of the fifth-year sam ples (Fig. 2). Recolonization after flood events is rapid, limiting effects to shortterm changes in abundance and community composition (Townsend, Doledec and Scarsb rook, 1997). Beche, McElravy and Resh (2006) also found that invertebrate communities and trait characteristics were more stable in wet, rather than dry, season communities and postulated that droughts have more severe long-term consequences than flooding for invertebrate communities. Traits changed less with time than taxonom ic composition and were more stable. This may be a product of the high functional redundancy existing among aquatic invertebrates (Poff, 1997; Lamouroux, Doledec and Gayraud, 2004). For example, although climatic variation can change species presence, multiple species share similar traits, allowing taxa to survive during changi ng conditions. The role of local and regional abiotic filters are discussed below in relation to temporal changes in structural (e.g., taxonomic) and functional (e.g., traits) aspects of invertebrate commun ities as a result of a disturbance imposed by a long-term drought. Environmental Variation Environmental stability was relatively high but highly variable throughout the study, with Bray-Curtis values ranging from 0.03 to 0.15 (Fig. 3). However, local water chemistry variables did not follow changes in precipitation or flow. This may be linked to high connectivity to the floodplain and oxyge nation of the hyporheic zone; these interactions are typically lost during severe drying even ts (Boulton, 2003; Lake, 2003)

PAGE 37

37 Droughts act on local stream variables by concentrating nutrients and organic m atter, and potentially increasing temperatur e (Closs and Lake, 1995; Stanley, Fisher and Grimm, 1997; Matthews, 1998; Golladay a nd Battle, 2002; Dahm, 2003). A decrease in o-phosphate following drought reflected flushing of stored nutrients during increased flow periods (Dahm, 2003). Ammonia peaked in the third year in both streams, reflecting increased microbial activity and organic ma tter. Massive amounts of organic matter are typically stored in the stream channel a nd floodplain during drought. Initial flushes from early flow events may not have been enough to carry organic matter from the floodplain into the stream, however, a large flow event in the spring prior to the third sampling period likely made a large amount of organi c matter available. Baldwin (2005) also suggested that peaks in ammonia following dr ying events might have originated from dead bacterial cells. As in other studies, a coupled decrease in water temperature and increase in discharge led to higher overall dissolved oxygen values (Stanley, Fisher and Grimm, 1997; Matthews, 1998; Golladay a nd Battle, 2002). Conductivity also decreased with time, reflecting diluti on of concentrated ions typically found during drought (Stanley, Fisher and Grimm, 1997; Caruso, 2002; Line et al., 2006) and may have been linked to greater contribution of groundw ater versus surface flow found during dry periods (Rider and Beli sh, 1999; Caruso, 2002). Limited precipitation and water availability in th e riparian zone altered local landscape dynamics. Leaf fall within the ripa rian zone decreased following the drought, indicating a recovery period from the drought as trees often drop their leaves during periods of moisture deficit (Escudero a nd del Arco, 1987). Although this may provide more resources for invertebrate s, leaf quality may be lower and thus limit decomposition.

PAGE 38

38 Temporal Variation and Successional Patterns in Taxonomic Abundance Broad-scale measures of macroinvertabrate communities were responsive to tem poral changes in environmental conditi ons. Overall abundance was more responsive to short-term environmental variation than we re taxa richness or st ability. Taxa richness was lowest immediately following drought, peaking in the 2004 sampling period. Changes in taxa richness have been linked to disturbance history in relation to short-term and long-term droughts (Beche, 2006). As fl ow regimes recover, more favorable conditions exist including in creased habitat availability and heterogeneity, higher dissolved oxygen levels, a nd dilution of nutrients. A set of nine core taxa existed in both stream s immediately following drought forming a regional species pool adapted to extreme conditions. Most either have multiple generations per year or a de siccation resistant stage. Cr ayfish typically respond to drought by creating deeper burrows, which may al so provide other species a refuge from receding water levels (Boulton, 1989). The chironomid genus, Polypedilum makes cocoons to resist periods of drying (Hinton, 1960). The presence of coleopteran adults and hemipterans in WF immediately followi ng drought reflect their ability to survive outside of the stream and ra pidly colonize via aerial disp ersal (Ortega et al., 1991; Wissinger, 1997). Changes in taxonomic composition were link ed to both localand large-scale environmental variables. Dissolved oxygen and pH were most linked to changes in taxonomic conditions for local variables, while long-term 12and 48-month precipitation most influenced overall changes in taxonomic composition in WF. Low pH values in WF excluded entire taxonomic groups, including Ep hemeroptera and Plecoptera. Dissolved oxygen is often posited to be a controlling variable for invertebrate communities in

PAGE 39

39 streams and is closely related to water quality. The relationshi p of this variable with longterm SPI values indicates the advantage of incorporating regional climatic data into bioassessment protocols. Temporal Variation in Traits Traits were more related to local abiotic variables than to flow or long-term precipitation indices. Initially, traits may be filtered by large-scale factors, including climate and geology (Poff, 1997); thus at the smaller scale of two adjacent watersheds, traits may vary locally. Across small, physiographically homoge neous regions (e.g., watersheds, ecoregions), sites are likely to be located within a single regional species pool (Zobel, 1997). Thus, as predicted by the habitat templet model (Southwood, 1977; 1988; Townsend and Hildrew 1994), local charac teristics at the r each scale directly influence biological traits. Traits responded predictably to changes in local environm ental conditions. As pH increased and nutrients decreased, species le ss likely to drift became more common in the stream reaches. Initial availability of nitrogen and phosphorous allowed for early colonization of scrapers (e.g., Boulton, 1991) as algal sources within the stream accumulated on available substrate. However, this effect was not apparent in WF, reflecting the lower productivity of grazers typically associated with colored, acidic streams (Rosemond et al., 1992). Additionally, shredders became more common in the second and third year, as previously e xposed patches of organic matter became submerged. In addition, species typically cl assified as detritivores (e.g., nemourid stoneflies) may consume algal biomass, assuming the role of scrapers, especially in streams with lower pH (Ledger and Hildrew, 2005). In a comparison of rivers affected by drought in Italy, Fenoglio et al.(2007) docum ented an increase in collectors and a

PAGE 40

40 decrease in shredders and scrapers w ith increased drought duration, suggesting a predictable cyclic change in feeding habits with drought. After a major disturbance, body size is hypot hesized to increase as communities stabilize. According ly, medium-sized species became more common as the community stabilized, while smaller species were more common during the less-stable time periods. Small body size is often related to shorter-life cycles, and might therefore serve as a resilience trait. However, small body size may al so allow for exploitation of refugia such as the hyporheic zone during droughts or fl oods (Townsend, 1989), serving as a potential source of colonizers (DoleOlivier, Marmonier and Beff y, 1997). Thus, small body size may also be useful for resisting impact s of disturbance (e.g., Townsend, Doledec and Scarsbrook, 1997). Organisms with longer life cycles require m ore stable habitat and water chemistry. Adaptations to high variability were less common by the third year of sampling, when species commonly had unior semivoltine life cycles and lacked adaptations for resisting desiccation. Additi onally, hardening of the exoskeleton reduces mortality during periods of drying and floods. Sclerotized and hard-shelled organisms were more common during the first two years after drought, while traits favoring softbodied organisms became more common with time. Although there was a general trend toward species lacking resistan ce to desiccation, many species in the streams appeared to be adapted to some level of drying through a desiccation-resistant or diapause stage. Although many traits responded predic tably to the drought, stream lined individuals were more comm on immediately following drought, reflecting the abundance of adult dytiscid beetles and amphipods as early colonizers. Dytis cid beetles colonize

PAGE 41

41 across large areas via aerial dispersal, while amphipods may persist in refugia by aestivating in moist sediments or disperse through underground rout es (Wiggins et al., 1980; Harris and Roosa, 2002). In small coastal pl ain streams, discharge is lower than in many montane headwater streams, thus streamlined bodies may not be necessary to resist flow forces. However, within four years, sp ecies preferring fast flow, such as clingers, became more abundant, suggesting that behavi oral rather morphologi cal adaptations to flow may be a distinguishing tra it of coastal plain streams. Drought Prediction A major hurdle that is often encountere d when as sessing impacts of extreme, unpredictable events on aquatic ecosystems is the availability of data prior to the disturbance (Lake, 2000; Lake, 2003). Alt hough long-term datasets (>10 years) are increasingly common in ecology, many geographi cal regions lack such data. Long-term datasets allow for the prediction of drought vi a precipitation indices, changes in stability and trait composition. The SPI ut ilized in this study indica ted a trend toward a major drought prior to the o ccurrence of such an event (Fig. 1). However, the 24-month SPI may bemost relevant since most invertebrate li fe cycles require months to years. Values fell toward 2001 levels in 2006, and a dryi ng period was evident after the last sampling period. This observation is further supported by the occu rrence of a severe drought in the same region during 2007 (U.S. Drought Monitor, http://www.drought.unl.edu/dm/archive.html). Thus, use of long-term regional precipitation data, which is more widely av ailable than discharge and ecological data, may provide a unique opportunity to study pr eand post-recovery aspects of extreme events.

PAGE 42

42 Stability of species composition d ecreased in both stream s in 2006, reflecting the oncoming drought. This change was more striking in WF than SF, likely because SF is buffered by groundwater inputs, while WF is dependent on surface water inputs from a wetland headwater. Traits were relatively stable ove r the study period, but decreased in the wetland-fed stream fro m 2005. Additionally, predators and species common in drift became more abundant in 20062007. Predators are linked to disturbance and become more common in distur bed or degraded streams as resources for other guilds are patchy (Rawer-J ost et al., 2000). This suggests return to a disturbed condition where organisms have the ability to emigrate if envi ronmental conditions become suboptimal. Biological traits have been advocated as good indicators of disturbance in aquatic ecosystem s (i.e. Doledec, Statzner and Bour nard, 1999). One associated issue is whether they predict changes in disturbance regime more effectively than taxonomic structure or they are a complementary aspect that shoul d be examined in disturbance ecology and biomonitoring programs (Doledec and Statzner, 2008). The results are unclear, as studies across biomes suggest they are more, less or equally as informative than taxonomic structure (Stevens et al., 2003; Finn and Poff, 200 5; Heino, 2007, Hoeinghaus, Winemiller and Birnbaum, 2007). Although less apparent than changes in taxonomic structure, stability of biologi cal traits in this study indi cated a disturbance during the drought period and the potential for another disturbance during the 2006 sampling period. Thus, traits may be useful indicators of disturbance that can be compared across watersheds. One caveat to this is that func tional traits are thought to be relatively insensitive to natural variation (Charvet et al., 2000; Statzner et al., 2001; 2005). The

PAGE 43

43 current study focused primarily on natural vari ation in climatic conditions and indicated changes in trait composition as a result of this variation. T hus, additional effort should be devoted to distinguishing the effects of na tural and anthropogenic disturbances when devising biological monitoring programs. Table 2-1. Definition and codes for biological traits and modalities. Trait Code Modality Trait Code Modality Voltinism v1 Semivoltine Habit h1 Clingers v2 Univoltine h2 Burrowers v3 Multivoltine h3 Swimmer Drying Resistance d1 Absent h4 Sprawler d2 Present h5 Skater Drift df1 Rare h6 Climber df2 Common Trophic tr1 Gatherer df3 Abundant tr2 Filterer Armoring ar1 Soft tr3 Scraper/Herbivore ar2 Sclerotized tr4 Shredder ar3 Case/Shell tr5 Predator Maximum Size s1 Small (<9mm) Shape sh1 Streamlined s2 Medium (9-16mm) sh2 Not Streamlined s3 Large (>16mm) (Bluff, Tubular) Rheophily r1 Standing r2 Slow r3 Fast

PAGE 44

44Table 2-2. Mean annual values for environmental variable s for the wetland-fed (WF) and seep-fed (SF) streams Year Flow (L/s) TSS (g/L) NH4 (g/L) ophosphate (g/L) NO2/NO3 (g/L) Total Phosphorous (g/L) Total Nitrogen (g/L) pH SC (uS/cm) DO (mg/L) Turbidity (NTU) Temperature ( C) Leaffall (g/m2) WF 20012002 0.99 0.015 2.48 2.73 1.08 8.22 238.00 5.53 42.28 4.47 1.78 16.13 34.29 20022003 2.52 0.001 6.28 1.75 0.00 3.69 278.28 4.73 30.60 7.34 0.19 12.15 12.29 20032004 1.04 0.017 11.97 1.85 0.00 13.26 345.87 5.03 35.95 4.51 1.18 16.08 19.26 20042005 1.76 0.003 2.88 2.49 2.34 4.90 233.30 4.87 24.90 7.62 1.23 12.00 21.26 20052006 1.47 0.013 6.20 1.78 0.00 16.94 343.30 5.35 26.85 6.87 1.10 13.63 30.19 20062007 2.58 0.008 3.93 3.01 0.00 9.61 291.87 5.11 33.68 8.99 0.63 13.05 24.55 SF 20012002 0.02 0.004 4.54 45.46 9.42 77.00 212.46 7.23 84.03 5.25 4.03 15.68 38.25 20022003 2.74 0.004 0.00 27.24 7.85 51.25 285.97 5.88 94.85 7.85 2.95 12.73 16.23 20032004 3.07 0.008 7.65 23.74 4.41 51.75 237.73 6.85 70.90 6.82 4.03 15.90 22.18 20042005 3.39 0.003 1.68 18.86 3.00 39.00 232.05 6.61 74.98 9.39 2.98 11.73 26.06 20052006 5.36 0.016 6.59 12.43 9.40 30.25 218.54 7.12 60.93 8.79 4.51 13.05 29.76 20062007 3.90 0.001 4.85 25.69 2.38 27.36 203.76 7.09 82.80 9.88 2.61 12.05 25.84

PAGE 45

45 -3 -2 -1 0 1 2 3 4 51956 1958 1960 1963 1965 1968 1970 1972 1975 1977 1980 1982 1985 1987 1989 1992 1994 1997 1999 2001 2004 2006SPI(a) -3 -2 -1 0 1 2 31956 1958 1960 1963 1965 1968 1970 1972 1975 1977 1980 1982 1985 1987 1989 1992 1994 1997 1999 2001 2004 2006SPI(b) -3 -2 -1 0 1 2 31956 1958 1960 1963 1965 1968 1970 1972 1975 1977 1980 1982 1985 1987 1989 1992 1994 1997 1999 2001 2004 2006SPI(c) -3 -2 -1 0 1 2 3 4 51956 1958 1960 1963 1965 1968 1970 1972 1975 1977 1980 1982 1985 1987 1989 1992 1994 1997 1999 2001 2004 2006SPI(a) -3 -2 -1 0 1 2 31956 1958 1960 1963 1965 1968 1970 1972 1975 1977 1980 1982 1985 1987 1989 1992 1994 1997 1999 2001 2004 2006SPI(b) -3 -2 -1 0 1 2 31956 1958 1960 1963 1965 1968 1970 1972 1975 1977 1980 1982 1985 1987 1989 1992 1994 1997 1999 2001 2004 2006SPI(c) C B A -3 -2 -1 0 1 2 3 4 51956 1958 1960 1963 1965 1968 1970 1972 1975 1977 1980 1982 1985 1987 1989 1992 1994 1997 1999 2001 2004 2006SPI(a) -3 -2 -1 0 1 2 31956 1958 1960 1963 1965 1968 1970 1972 1975 1977 1980 1982 1985 1987 1989 1992 1994 1997 1999 2001 2004 2006SPI(b) -3 -2 -1 0 1 2 31956 1958 1960 1963 1965 1968 1970 1972 1975 1977 1980 1982 1985 1987 1989 1992 1994 1997 1999 2001 2004 2006SPI(c) -3 -2 -1 0 1 2 3 4 51956 1958 1960 1963 1965 1968 1970 1972 1975 1977 1980 1982 1985 1987 1989 1992 1994 1997 1999 2001 2004 2006SPI(a) -3 -2 -1 0 1 2 31956 1958 1960 1963 1965 1968 1970 1972 1975 1977 1980 1982 1985 1987 1989 1992 1994 1997 1999 2001 2004 2006SPI(b) -3 -2 -1 0 1 2 31956 1958 1960 1963 1965 1968 1970 1972 1975 1977 1980 1982 1985 1987 1989 1992 1994 1997 1999 2001 2004 2006SPI(c) C B A Figure 2-1. Plot of Standardized Precipitation Index (SPI) values for southwestern Georgia, from 1956 to 2007. A) 3-month, B) 12-mont h, and C) 48-month. Values above 2 indicate an extremely wet year, while valu es below indicate an extremely dry year.

PAGE 46

46 (a) A (a) A (b) B (b) B Figure 2-2. Hydrograph based on m ean daily discharge (m3/s) for each stream. A) stream with wetland headwater (WF) and B) stream with seep headwater (SF).

PAGE 47

47 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.1601-0202-0303-0404-0505-0606-07Bray-Curtis Distance WF SF Figure 2-3. Temporal variability of Bray-Curtis stability values f or environmental variables in WF and SF ( SE).

PAGE 48

48 0 10 20 30 40 50 60 200120022003200420052006Total Taxa WF SFA 0 200 400 600 800 1000 1200 1400 1600 1800 200120022003200420052006Abundance WF SFB Figure 2-4. Temporal changes in taxon richness and invertebrate abundance. A) taxon richness and B) abundance values (per ~ 3.1m2) (SE) for individual years.

PAGE 49

49 0 0.1 0.2 0.3 0.4 0.5 0.6 0.701-0202-0303-0404-0505-0606-07Bray-Curtis Distance WF SF Figure 2-5. Changes in compositional stability (B ray-Curtis distan ce) in WF and SF ( SE).

PAGE 50

50 R2 = 0.11 R2 = 0.14 R2 = 0.490 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 -2.5-2-1.5-1-0.500.511.52SPIBray-Curtis Values 48-month SPI 12-month SPI 3-month SPI Linear (3-month SPI) Linear (12-month SPI) Linear (48-month SPI) R2 = 0.18 R2 = 0.21 R2 = 0.510 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 -2.5-2-1.5-1-0.500.511.52SPIBray-Curtis Values 48-month SPI 12-month SPI 3-month SPI Linear (3-month SPI) Linear (12-month SPI) Linear (48-month SPI)A B R2 = 0.11 R2 = 0.14 R2 = 0.490 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 -2.5-2-1.5-1-0.500.511.52SPIBray-Curtis Values 48-month SPI 12-month SPI 3-month SPI Linear (3-month SPI) Linear (12-month SPI) Linear (48-month SPI) R2 = 0.18 R2 = 0.21 R2 = 0.510 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 -2.5-2-1.5-1-0.500.511.52SPIBray-Curtis Values 48-month SPI 12-month SPI 3-month SPI Linear (3-month SPI) Linear (12-month SPI) Linear (48-month SPI)A B Figure 2-6. Linear regression of SPI values versus taxonom ic stability. A) Wetland-Fed (WF) and B) Seep-Fed streams.

PAGE 51

51 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.1601-0202-0303-0404-0505-0606-07Bray-Curtis Distance WF SF Figure 2-7. Changes in trait stab ility (Bray-Curtis distan ce) in WF and SF ( SE).

PAGE 52

52 Figure 2-8. NMDS ordinations of log10-abundance in site-year spa ce and taxon-space for W F. Time periods are indicated by different symbol s. Ordination plots of taxa are based on weighted-averaging.

PAGE 53

53 Figure 2-9. NMDS ordinations of log10-abundance in site-year spa ce and taxon-space for SF. Tim e periods are indicated by different symbol s. Ordination plots of taxa are based on weighted-averaging.

PAGE 54

54 Figure 2-10. NMDS ordinations of biological traits in s ite-y ear space and trait-space for WF. Time periods are indicated by different symbol s. Ordination plots of taxa are based on weighted-averaging.

PAGE 55

55 Figure 2-11. NMDS ordinations of biological traits in s ite-y ear space and trait-space for SF. Time periods are indicated by different symbol s. Ordination plots of taxa are based on weighted-averaging.

PAGE 56

56 CHAPTER 3 TESTING BMP EFFECTIVENESS FOR SMALL COASTAL PLAIN STREAMS USING MACROINVER TEBRATES AS BIOINDICATORS Introduction Headwater streams are tightly coupled w ith the surrounding riparian landscape. Thus, changes in the structure of the riparian zon e can a ffect water quality of larger streams and rivers, since they are heavily influenced by headwater streams that feed them (Meyer and Wallace, 2001). Headwater stream s make up a majority of channel length within stream networks and serve importa nt ecological and bi ological functions by delivering water, sediment, organic material prey items, and nutrients to downstream reaches (Sidle et al., 2000; Gomi et al., 2001; Meyer and Wallace, 2001; Wipfli and Gregovich, 2002). As the importance of ecosyst em services, such as water quality and biodiversity, becomes more widely recognize d, the need to protect aquatic resources increases. Thus, proper management of te rrestrial landscapes must take into account needs of aquatic organisms and communities. Numerous studies have found significant im pacts of logging on physical and chemical aspects of streams, including redu ced large woody debris in streams (Golladay, Webster and Benfield, 1987), increased se diment input (Beschta, 1978), discharge (Hartman and Scrivener, 1990), nutrient inpu ts (Likens et al., 1969; McClurkin et al., 1985), and decreased shading re sulting in higher water temperature (Swift and Messer, 1971; Webster and Waide, 1982). Elevated light temperature and nutrient concentration can increase algal biomass within the stream shifting the base of the food web from allochthonous to autochthonous sources (L ikens et al., 1970; Wallace and Gurtz, 1986; Bilby and Bisson, 1992). The extent and impact of these effects are influenced by

PAGE 57

57 geology, soils and vegetation of the catchment, the extent to which the riparian buffer strip remains after logging, stream discharg e, and channel gradient and morphology. Changes in abiotic characteristics of a stream following logging can affect the structure and function of th e stream community, includi ng periphyton (Lowe, Golladay and Webster, 1986), fish (Garman and Moring, 1993) and macroinvertebrates. Logging activities can disrupt stream invertebrate communities, but the magnitude and trajectory of effects vary. Increased light penetrat ion and warmer temperatures from canopy removal, and nutrient enrichment in runo ff from ground disturbance, increase aquatic invertebrate density and/or biomass in streams (Newbold et al., 1980; Murphy et al., 1981; Hawkins et al., 1982; Wallace and Gurt z, 1986; Campbell and Doeg, 1989; Brown et al., 1997). Fine sediment loading, particul arly in watersheds w ith steep slopes, can reduce invertebrate populations following logging (Growns and Davis 1994, Waters 1995, Wood and Armitage 1997), clogging trach eal gills, and burying food sources. In many cases, invertebrate communities shift from shredders to grazers (algae consumer) or detritivores (collector-gatherer) (Haefner and Wallace 1981; Gurtz and Wallace, 1984; Webster et al., 1992). Although Stone and Wallace (1998) found shifts in dom inance of taxa, there was no loss of taxa in logged versus unlogged si tes. They posited that measures of taxon richness may be useful for indicating presence of pollutants, but not for more discrete disturbances. Long-term impacts of clea r cutting were documented decades later, stemming from recovery of riparian vege tation and canopy cover (Growns and Davis, 1991; Stout et al., 1993; St one and Wallace 1998).

PAGE 58

58 Although there are exceptions (e.g., Wallace and Gurtz, 1986), most studies exam ining effects of riparian zone manageme nt on streams last only a few years, with only a year of pre-harvest data, limiting th e predictive power of post-harvest data. Perhaps the assumption is that management activities are only important if they supercede natural variati on in environmental conditi ons. However, catastrophic disturbances such as hurricanes and drought s potentially alter impacts of human induced disturbances. Thus, it is important to incor porate natural disturban ces into studies of anthropogenic disturbances in aquatic ecosystem s to generate more ge neral predictions of disturbance (Ward, 1998). The impacts of forest management activ ities on aquatic ecosy stem s have been well studied in the Northwest and Mid-Atlan tic U.S., where there are steep slopes and high-gradient streams, but little effort has been placed on st reams in the Southeast coastal plain (Stone and Wallace 1998; Kedzierski and Smock, 2001). These low-gradient streams have shallow slopes and finer sediment (sand and silt) than montane streams and thus are likely to respond differently to logging. Even fewer studies have investigated the impact of logging within buffer zones (e.g, Kr eutzweiser et al., 2005) even though this is an acceptable practice throughout the South east (e.g., Georgia Forestry Commission, 1999). Additionally, it appears that no study has utilized biologica l traits, other than feeding guilds, to understand impacts of forest ry practices. The goal of this study was to test the effectiveness of Georgias best mana gement practices for forestry along streams. This was assessed through multiple years of pre and post-harvest sampling of macroinvertebrates, water quality parameters, and invertebrate food sources.

PAGE 59

59 Materials and Methods Site Description Four study watersheds were located with in International Papers Southlands experim ental forest in sout hwestern Georgia within th e Dry Creek watershed, which discharges to the Flint River. The first order headwater st reams were labeled from A to D (Fig. 3-1) since they did not have official nomenclature. Watersheds A and B were shallow, floodplain influenced, wetland-fed streams, while C and D were incised, seepfed streams. Geology The watersheds were located on the Pelham escarpment between the Tifton upland and Dougherty plain. Soils of the study sites were dominated by Ultisols. Riparian soils were com prised of Chiefland and Esto series, which feature well drained fine sands over clay loams. The lower slopes comprised Eu stis series soils, which were loamy sands over sandy loams and classified as somewhat excessively well drained. Upland soils were Wagram, Norfolk, Lakeland, Orangeburg, and Lucy series, which are generally well drained loamy sands over sandy clay loams, with the exception of the Lakeland Unit, which has a sandy texture throughout and is ch aracterized as exce ssively well drained (USDA, 1939). This transitional area has bluffs and deep ravines that create cool microclimates supporting rare plant species with northern affi nities (Wharton, 1978). Vegetation Species composition of the vegetation was similar among watersheds. The upland consisted of m ature, planted pine, and the riparian areas were mixed pine and hardwoods.

PAGE 60

60 Species dominating the overstory in riparian areas were: Nyssa biflora Liriodendron tulipifera Pinus glabra, Magnolia virginiana Fagus grandifolia Liquidambar styraciflua Quercus nigra and Quercus michauxii. Magnolia grandiflora was most common in watersheds C and D (Internati onal Paper unpublished data). The upland of each watershed was dominated by Pinus taeda which was established at varying times by hand planting. The midstory of all wa tersheds was generally composed of Carpinus caroliniana Ostrya virginiana, Acer rubrum Acer barbatum and Oxydendrum arboretum Magnolia pyramidata occurred in riparian areas and midslopes of watersheds C and D. Climate The climate of the region is character ized by warm, hum id summers and mild winters with average annual precipita tion of 1412 mm (SERCC, 2007). Temperatures range from an average maximum of 33.5C to a minimum of 2.8 C. June has the highest mean rainfall (152.1 mm) and October the lowest (77.5 mm) (SERCC, 2007). Summer rains are usually short, high intensity events giving way to low intensity frontal events from late fall to early spring. Due to proxi mity to the Gulf of Mexico, spin-off from hurricanes and tropical storms in late summer is not unusual. Drought conditions occurred during 1998-2002 and resulted in an accumulated rain fall deficit of 711-1270 mm in parts of southwestern Georgia (see Chapter 2). Hydrology Surface water flow in the Apalachicola-Chat tah oochee-Flint river basin is lowest from September to November and peaks duri ng January to April due to higher rainfall and decreased evapotranspiration (Couch et al ., 1996). Streams and rivers in the Coastal Plain receive substantial groundwater because they are typically deeply incised into

PAGE 61

61 underlying aquifers (Couch et al., 1996). The study streams are first order, groundwaterinfluenced, low to medium gradient, and have sand-dominated substrate. In-stream habitat includes coarse woody debris, undercut banks, leaf packs, and fine roots. Fiberglass parshall flumes (Tracom Inc., A tlanta, GA) were placed at the downstream end of each reach. The four study watersheds average 39 ha, with average annual discharge of 1.5 L/s prior to harvest (Summer et al., 2003; Summer unpublished data). Experimental Harvest The statistical design was BACI (Before After Control Im pact) using a paired watershed design, with two treatment and two re ference first-order watersheds varying in area from 24 to 44 ha. Watershed pairs were determined based on landscape morphology and vegetative community, and treatment waters heds were randomly selected within each pair. Watersheds A and B formed the first pair with Watershed B selected for treatment. Watersheds C and D formed the second pair, with Watershed C selected for treatment. Each watershed was divided into an upstr eam and downstream reach, separated by at least 50 m. The reference watersheds did not receive silviculture treatm e nts during the study period. The remaining two watersheds were clea rcut after 27 months of baseline data collection (June 2001 to September 2003). Post harvest data collec tion continued until February 2007. In treatment watersheds, the SMZ in the upstream reach was left intact (intact SMZ), while 50% basal area was re moved in the downstream portion (thinned SMZ). SMZ widths were determined acco rding to minimum recommendations in Georgia BMP manual(Georgia Forestry Commission, 1999). Slopes less than 20% received a 40 foot (12m) buffer and while those greater than 20% received a 70 foot (21m) buffer.

PAGE 62

62 Physical and Biological Measurements Eight 50m sample reaches, two per watershed, were established 30.8 m upstream of hydrology flumes. Three transects w ere established perpendicular to the thalweg within each reach at 15, 30, and 45m to serve as in-stream data collection points for physical measurements including channel cross-sections, canopy cover, and percent cover of in-stream habitat. A survey of habitat unit and channel characteristics was conducted longitudinally within established macroinverteb rate sample reaches once before harvest (December 2001) and once after (October 2004). A 50 m fiberglass tape was placed in the thalweg of the stream along which boundaries between habitat unit types (riffle, run, glide, pool, backwater pool step, and undercut bank) were determined and physical characteristics recorded. A backwater pool was defined as being slower and deeper than a glide but lacking characteristics of a pool, such as scouring, deposition, and presence of a deep section followed by a shallow tail downstreeam (i.e. measurable residual pool depth). For each unit type, le ngth, wetted width, and maximum water depth were recorded. For steps and pools, a step height and residual pool depth were taken. The length and diameter of channel obstructions (e.g., wood, roots) were recorded when the object was primarily responsible for pool formation. The number of functional (e.g., ability to change stream morphology) and nonfunctional wood pieces greater than 10 cm in diameter was recorded, and texture of the streambed (e.g., sand, silty-sand) was visually assessed. Habitat data were converted into percent cover, to define m ajor habitat types to be sampled for macroinvertebrates. Canopy photos were taken at each transect once before and once after harvest with a digital camera fitted with a 180 hemispherical fisheye lens to calculate % canopy cover.

PAGE 63

63 Physical measurements Water temperature was measured from October 2001 through February 2007 with an Onset HOBO temperatu re logger (Pocasset, MA), programmed to record temperature at 15 minute interv als. Stream flow, water chem istry, and meteorological measurements were collected by other investigators and technicians as part of the D ry Creek Study, and these data were available for use in this study. Stream stage and discharge were recorded every 15 minutes by Isco Model 4320 Bubbler Flow Meters at six sites: one in the stream at the ou tlet of watersheds A, B, C, D, and one in the upstream portion of watersheds B and C (Sum mer, 2003). Monthly in-situ measurements for dissolved oxygen, specific conductance, temper ature, pH, and turbidity were made at the downstream portion of each reach with porta ble m eters. Grab samples of water were also taken and analyzed for inorganic ni trogen, inorganic phosphorus, and ammonium (Jones et al., 2003). Energy sources Sixteen leaf litter tr ap s (surface area 0.26 m2 each) were positioned within the riparian area: streambank (6), 10 m from the st ream (6), and 20 m from the stream (4). Following harvest, leaf litter traps 20 m fr om the stream were beyond the SMZ, while the 10 m samples were within but near the edge of the SMZ. Leaf litter was collected monthly and dried at 60C for 48 hours a nd sorted into pine, hardwood, small woody debris, and mast. A subsample of leaf litter from riparian zone samples was used for nutrient analysis. Three samples were comb ined from each reach on a quarterly basis (May, September, December, and February) and ground to a fine powder. A 3-5 mg subsample was then analyzed for C:N ratios using a Carlo-Erba CNS analyzer.

PAGE 64

64 Within each stream reach, ten randomly se lected location s were sampled monthly following harvest for periphyton, benthic orga nic matter (BOM), and macrophytes from 2003-2007. At each sampling point, a 0.25 m2 sampling quadrat was randomly tossed onto the streambed. Periphyton and BOM samp les were collected by inserting petri dishes (17.34 cm2) into the streambed (Tett et al., 1978). Chlorophyll a was analyzed spectrophotometrically (Sartory and Grobbek aar, 1984) to estimate periphyton biomass. The contents of addistional petri dishes were dried at 60C for at least 48 hours, weighed, burned at 550C for five hours, and reweighed for as h-free dry weight determination after cooling. Macrophytes were sampled by removing all vegetation above the sediment surface that existed within a 0.25 m2 quadrat. These were then ri nsed and dried at 60C. BOM was square root transformed, and chlo rophyll a was log-transformed prior to statistical analysis. Macroinvertebrates Benthic macroinvertebrates were collected within established sample reaches with a 500-m-mesh D-frame net (0.3 m wide) using a multi-habitat sampling procedure (Barbour et al., 1999) during December and Fe bruary from 2001 to 2007. Within each reach, 20 sampling sweeps (~3.1 m2) were made through all ha bitat types including sand, woody debris, fine roots, and leaf packs. Samp les were placed in 1 L bottles, preserved in 95 % Ethanol and returned to the lab for pr ocessing. Macrophytes were included after 2003 because they became a significant habitat type in the treatment watersheds. All samples were processed by washing organic de bris (leaves and woody debris) with water onto a 500-m-mesh sieve. Larval Chironomid ae were subsampled (randomly selected 100 individuals) and mounted in CMC moun ting media for both voucher specimens and identification to genus. Macroi nvertebrates were enumerated and identified to genus or

PAGE 65

65 species using local and regional keys (Pes cador et al., 1995; Ep ler, 1995;1996; Merritt and Cummins, 1996; Pescador et al., 2000; Gelhaus, 2002; Richardson, 2003). Biological Traits Fourteen biological traits were select ed to characterize body morphology (size, body shape, body arm oring, respiration), life hist ory (voltinism, resistance to desiccation, eggs cemented to substrate, and development and hatch times), mobility (occurrence in drift), and ecology (rheophily, behavior, feeding preference s, microhabitat preference) (Table 3-1) to delineate responses to change s in disturbance regime (Poff et al., 2006). Some desired traits were omitted due to the lack of available information (e.g., fecundity), particularly for chironomid genera. The fourteen biological traits were divided into 49 modalities ranging from two to seven levels per trait. Trait information was collected from literature (Viera et al., 2006), as well as through communication with taxonomic experts in the United States. Traits were coded and analyzed as in Chapter 2. Data Analysis Energy sources Changes in leaf fall C:N ratios, perip hyton, and BOM with tim e and treatment were analyzed with repeated measures ANOVA (SAS Institute, 2002). Seasonal impacts of harvest on periphyton biomass were a ssessed by grouping time periods into by wet or dry seasons. The wet season was defined as May September, while the dry season was October April. Multiple regressions were utilized to relate changes in BOM and chlorophyll a in relation to environmental variables for each treatment. To reduce impacts of multicollinearity on the regressi on model, Pearsons correlations were calculated for each pair of environmental va riable, and values greater than 0.6 were removed. This resulted in pH and orthophosphate being removed from the dataset.

PAGE 66

66 Environmental variables Environmental variables were analyzed over tim e with repeated measures ANOVA (SAS Institute, 2002). Since macroinvert ebrate samples were taken during the winter/spring period, only environmental data from this period were analyzed. When differences were significant, post-hoc anal ysis was conducted usi ng Tukeys test and Bonferroni corrections. Additi onally, environmental stability was assessed by calculating Bray-Curtis distances (Bray and Curtis, 1957) between adjacent years. These measure dissimilarity with values ranging from 0 to 1. Zero denotes identical samples; thus, higher values denote lower stability and unity implies complete turnover. Macroinvertebrates The Florida Stream Condition Index (SC I) com bines metrics that respond to changes in human induced disturbance to yiel d a score reflecting wa ter quality (Florida Depatment of Environmetal Protection, 2004). Higher values indicate better water quality. SCI values were analyzed by time a nd treatment effects using repeated measures ANOVA (SAS Institute, 2002). Stability of invertebrate communities Compositional stability of invertebrate communities was examined for streams between pairs of success ive years (e.g., 1 vs 2, 2 vs 3, etc.). Stability was measured by calculating Bray-Curtis distances between ad jacent years based on abundance data and biological traits. ANOVA then was used to examine between year differences in compositional and biological trait stability scores for the streams. Ordination: species co mposition and traits Nonmetric multidimensional scaling (NMDS; Kruskal, 1964) was used to explore temporal patterns in species composition and biological traits as in Chapter 2. Since the

PAGE 67

67 experiment was designed as a paired waters hed study (A paired with B and C with D), ordinations were performed on the pairs se parately. For comparing treatments, preharvest samples were grouped together and co mpared with post-harvest reference and treatment reaches. Ordinations were pe rformed on species a bundances and abundance weighted biological tr aits individually. A multi-response permutation procedure (MRPP; McCune and Grace, 20 02) was used to test for significant differences in taxonomic composition and biological trait structure over time at each stream. Indicator species analys is (IndVal; Dufrene and Legendre 1997) was used to identify signifi cant indicator species discriminating among the time periods for the species composition and biological trait da ta. All ordinations, MRPP, and indicator species analyses were performed in PC-Ord ver. 5 (McCune and Mefford, 1999). Results Energy Source Benthic algal biomass estimated from ch lorophyll a differed significantly between trea tments (F2, 2834 = 102.4, P < 0.001) and seasons (F1,2834 = 40.8, P < 0.001) was twice as high in the selective harves t treatment as in the reference and intact treatments during the dry season (0.04 vs. 0.08 mg/m2). In the wet season, ch lorophyll a was 50% greater in the selective treatment compared to re ference and intact treatments (Fig. 3-2). Chlorophyll a was best predicted by phosphorous in the reference streams, conductivity in the thinned SMZs, and was not predicted by any variable in the intact SMZ streams. (Table 3-2). BOM differed significantly between treatments (F1, 1992 = 66.5, P < 0.001), increasing along a gradient fr om selective (12.5 0.5 g/m2) to intact (17.4 0.7 g/m2) to

PAGE 68

68 reference (22.8 0.6 g/m2) treatments. Significant re gressions were found for all treatments, but were explained by different predictors. BOM was best predicted by conductivity, oxygen, and turbidity in reference streams, by flow and ammonia in thinned SMZs, and by flow and turbidity in intact SMZs (Table 3-2). C:N ratios in leaf fall were generally high, ranging from 40 to 70. Values were similar in the reference streams between pr e and post harvest samples, but decreased significantly after harvest in the intact (F1, 36 = 12.8, P < 0.01) and thinned (F1, 36 = 4.2, P < 0.05) SMZs (Fig. 3-3). Environmental Variables Ammonia (F2,82 = 31.8, P < 0.001), total nitrogen (F2,82 = 55.8, P < 0.001), and total phosphorous (F2,82 = 3.2, P < 0.05) varied significantly with harvest, but not over time. Ammonia peaked three years following ha rvest, with levels 9 times higher than the reference streams in the intact SMZ treatme nt and 6 times higher in the thinned SMZ treatment (Fig. 3-4). Total nitrogen also peaked in the third year following harvest, with values increasing by twenty percent in th e harvested watersheds. Total phosphorous peaked in the third year in the harvested wa tersheds, but was always lower than in the reference watersheds (Table 3-3). Dissolved oxygen levels increased over time (F4,82 = 13.9, P < 0.001) and ranged from values of 5 9 mg/L, but were not affected by harvest. Temperature changed significantly over time (F4,82 = 11.2, P < 0.001), ranging from 13-18 C. Although changes relative to harvest were not significant, winter temperatures were 1-2 C higher in treatment streams following harvest. Flow increased over the course of the study (F4,82 = 5.7, P < 0.001) as precipitation increased, more so in harveste d watersheds than reference watersheds (F2,82

PAGE 69

69 = 15.5, P < 0.001) after harvest. Flow ranged from 0.5 to 3.5 L/s in reference streams but reached levels of 7.5 and 10 L/s in intact SMZ and thinned SMZ treatments, respectively. Turbidity varied significantly over the course of the study (F4,82 = 2.6, P < 0.05), but did not have any discernible tem poral pattern. However, values increased significantly following harvest (F2,82 = 22.1, P < 0.001), doubling in intact SMZs and tripling in thinned SMZs compared with the reference in th e first year following harvest (Table 3-3). Macroinvertebrates SCI values became more positive over time and with more extensive harvest, indicating b etter water quality. These increases were most notable in selective harvest, followed by intact SMZs and reference sites (Fig. 3-5). Stability Taxonomic stability increased significan tly over tim e in all treatments (F5,88 = 6.7, P < 0.001) as Bray-Curtis values decreased (Fi g. 3-6). Trends in tr ait stability did not change significantly with site or treatment over the course of the study. However, higher Bray-Curtis values in the inta ct SMZ indicate higher spec ies turnover (Fig. 3-7). Taxonomic composition Watersheds A (Reference) and B (Harvested). NMDS ordination (stress = 11.8, P = 0.001) explained 89 % of the variance in the dataset, with 38 %, 11 %, and 40 % explained by Axes 1, 2, and 3 respectively. Ov er all, the ordination indicated a distinct separation of community composition based on harvest levels (Fig. 3-8) and was supported by significant differe nces between reference and harvest samples (MRPP, A = 0.3, P< 0.001). However, post-harvest samples from thinned and intact SMZs were not different. Axis 1 was primarily represented by total nitrogen (r = 0.8), ammonia (r = 0.5), conductivity (r = 0.8), pH (r = 0.5) flow (r = 0.6), and turbidity (r = 0.8). The genera

PAGE 70

70 Alotanypus (r = .6), Nippotipula (r = 0.5), Crangonyx (r = -0.8), Habrophlebiodes (r = 0.6), Helichus (r = 0.6), Stenelmis (r = 0.7), and Sphaerium (r = 0.6) were most strongly correlated with Axis 1. Axis 2 was most related to dissolved oxygen (r = 0.5), Nanocladius (r = 0.6), Parametriocnemus (r = 0.6), Bezzia (r = 0.6), and Erioptera (r = 0.5). Axis 3 was correlated with dissolved oxyge n (r = 0.6), flow (r = 0.5), leaf fall (r = 0.5), Cryptochironomous (r = 0.7), Polypedilum (r = 0.5), Conchepelopia (r = 0.6), Alluaudomyia (r = 0.6), Simulium (r = 0.6), Sphaerium (r = 0.8), and Tanytarsus (r = 0.7). For watersheds A and B, drought impacts se parated along axis 3 of the NMDS, with positive values leading to a recovery from disturbance. Harvest effects separated along Axis 1, with positive values indicating the harvest induced disturbance. Only one chironomid species, Parachaetocladius was a significant species for pre-harvest sam ples. By contrast, seven spec ies were significant indicators for reference streams after harvest. These were primarily predators or those consuming organic matter and included Alotanypus, Caecidiota Corethrella Crangonyx Ptilostomis, Sciomyzidae and Stenochironomus Thirteen species were indicators for the thinned SMZ treatment. They occupied a range of trophic habits and included Ablabesmyia Calopteryx Cheumatopsyche, Cryptochironomous, Habrophlebiodes, Hemerodromia Orthocladius Paralauterborniella Peltodytes, Sphaerium Stenelmis Tanytarsus and Theinemaniella Indicator species reflective of the intact SMZ treatment were primarily predators and shredders including, Anisocentropus, Procladius, and Hexatoma (Table 3-4). Watersheds C (Harvested) and D (Reference). NMDS ordination (stress = 12.9, P = 0.001) explained 88 % of variance in the dataset, with 39 %, 31 % and 18 % explained by Axes 1, 2, and 3, respectively. Ov erall, the ordination indicated separation

PAGE 71

71 of the invertebrate communities by harvest regime (Fig. 3-9) and was supported by significant differences between reference and harvest sites, but not between thinned and intacts SMZs. Axis 1 was primarily represented by ammonia (r = 0.5), total phosphorous (r = .5), dissolved oxygen (r = 0.6), flow (r = 0.7), and leaf fall (r = -0.4). Polypedilum (r = 0.5), Tanytarsus (r = 0.5), Tribelos (r = 0.8), Simulium (r = 0.5), Habrophlebiodes (r = 0.6), and Diplectrona (r = 0.5) were most strongly corr elated with Axis 1. Axis 2 was most related to ammonia (r = -0.5), total n itrogen (r = .5), and tu rbidity (r = -0.6). Nippotipula (r = 0.6), Pseudolimnophila (r = 0.7), Bezzia (r = 0.8), Chrysops (r = 0.6), Stenelmis (r = 0.6), and Helichus (r = 0.7) were most strongly related to Axis 2. Axis 3 was most related to to tal nitrogen (r = 0.6). Stempellinella (r = -0.6), Tanytarsus (r = 0.6), Corynoneura (r = -0.6), Thienemaniella (r = -0.7), Stenelmiss (r = 0.6), and Helichus (r = 0.7) were most strongly related to Axis 3. For watersheds C and D, Axis 1 of the NMDS appears related to both disturbances, with drought samples having lower values than reference, which had lower valu es than harvest samples. Significant indicator taxa for the pre-ha rves t samples were primarily predators and collector-gathers, including Parachaetocladius, Hexatoma, and Leptophlebia Eighteen taxa were significant indicators for the reference streams after the harvest occurred, primarily consisting of predators a nd shredders. Four taxa were indicators for the thinned SMZ treatment. All were prim arily scrapers and collector gatherers and included Decapoda (Cambaridae), Brillia Elimia and Paracladopelma The indicator taxa reflective of the intact SMZ treatment were predators, scrapers, filterers, gatherers, and shredders, including, Cryptochironomous, Polypedilum Stenochironomous Tanytarsus Tribelos, Molanna, Triaenodes Laevapex and Hexagenia (Table 3-5).

PAGE 72

72 Biological traits Watersheds A (Reference) and B (Harvested). NMDS ordination (stress = 14.3, P = 0.001) explained 91.1% of the variance in the dataset, with 66 % and 25 % explained by Axis 1 and 2 respectively. Overall, the or dination indicated sepa ration of comm unity composition with harvest regime (Fig.3-10) and was supported by significant differences between reference and harvest sites, but not between thinned and intact harvest treatments (MRPP, A = 0.2, P< 0.01). Axis 1 was primarily represented by ammonia (r = -0.4), total nitrogen (r = -0.5), dissolved oxygen (r = -0.6 ), and flow (r = -0.4). Traits positively associated with Axis 1 included bu rrowers (h2, r = 0.7) and collector-gatherers (tr1, r = 0.8), with sclerotized bodies (ar2, 0.7) and slow-hatch ing eggs (ht2, r = 0.8), that are abundant in drift (df3, r = 0.6) and live in gravel (mh3, r = -0.7) and woody debris (mh6, r = 0.7). Those negatively associated w ith Axis 1 included medium-sized (s2, r = 0.7) sprawlers (h4, r = -0.6), filte rers (tr2, r = -0.7), and herb ivores (tr3, r = -0.6), with less than one generation per year (v1, r = 0.6) and fast-hatching (h t1, r = -0.7) cemented eggs (ec1, r = -0.6) living in sand (mh1, r = -0.7) or rocks (mh2, r = -0.7). Axis 2 was most related to total nitrogen (r = -0.4) and turbidity (-0.5). Traits positively associated with Axis 2 included small (s 1, r = 0.7), soft-bodied (ar 1, r = 0.7) individuals with cutaneous respiration (rs1, r = 0.7), and rapid development rates (ds1, r = 0.7). Those negatively associated with axis 2 included la rge (s3, r = -0.6) shredders (tr4, r = -0.6) with tracheal gills (rs2, r = -0.8), slow development (ds2, r = -0.6) less than one generation per year (v1, r = -0.6), an d cemented eggs (ec1, r = 0.6). Traits indicative of pre-harvest samples included sclerotized (ar2) collectorgatherers (tr1) and swimm ers (h3) common in drift (df3), with long hatching (ht2) and development (ds2) times. Those indicative of the reference streams in the post-harvest

PAGE 73

73 period included bluff (sh2), soft-bodied (ar1 ) predators (tr5) with short development times (ds1) living in plant matter (mh4). Species in the thinned SMZ treatment were medium-sized (s2) sprawlers (h4), living in sand (mh1) and gravel (mh3) substrate. Species in the intact SMZ treatment included fi lterers (tr2) with semi-voltine life cycles (v1) that are rare in drif t (df1) (Table 3-6). Watersheds C (Harvested) and D (Reference). NMDS ordination (stress = 11.1, P = 0.001) explained 95.1% of variance in the dataset, with 81 % and 15 % explained by Axes 1 and 2, respectively. Overall, ordination did not indicate separation of comm unity composition by harvest (Fig.3-11) although MRPP did indicate signifi cant differences between reference and harvest samples. Axis 1 did not strongly relate to any environmental variables. Shredders (tr4, r = 0.7) and swimmers (h3, r = 0.7) with tracheal gills (rs2, r = 0.8), cemented eggs (ec1, r = 0.8), long development (ds2, r = 0.8) and hatch times (ht2, r = 0.8) in fast turbulent water (r4, r = 0.7) with streamlined (sh1, r = 0.8), sclerotized (ar2, r = 0.9) bodies univoltine life cycles (v2, r = 0.5), living in detritus (mh5, r = 0.7) were positively related to Axis 1. Collector-gatherers (tr1, r = -0.6) and sprawlers (h4, r = -0.6) with cutaneous resp iration (rs1, r = -0.9) bluff bodies (sh2, r = 0.8) multivoltine life cycles (v3, r = -0.6), short development (ds1, r = -0.9) and hatch times (ht1, r = -0.9), abundant in drift (df3, r = -0.7), small (s1, r = -0.7), soft-bodies (ar1, r = -0.8) were negatively related to Axis 1. Axis 2 did not relate to any environmental variable. Large-bodied (s3, r = 0.8) individuals were positivel y related to axis 2. Small bodied (s1, r = -0.7), burrowers (h2, r = -0.7) not common in drift (d f1, r = -0.8) were negatively associated with axis 2.

PAGE 74

74 Traits indicative of pre-harvest sample s included individuals with m id-length hatching (ht2) and development times (ds2), semivoltinism (v2), sclerotized bodies (ar2), swimmers (h3) and shredders (t r4), residing in silt substrate (mh7). Species in reference streams during the post-harvest period included predators (tr5) respir ing via spiracles or plastrons living in detritus (mh5). Species in the thinned SMZ treatment included herbivores (tr3) preferring sandy substrate (mh1). Species in the intact SMZ treatment included individuals without cemented eggs living in woody debris (Table 3-7). Discussion The need for properly managed watersheds has becom e clear as estuaries and deltas become inundated with sediment, nutrien ts, and chemical pollutants (Justic et al., 1993; Long et al., 1994). Proper management of small streams will contribute significantly to reductions in the downstream tr ansport of these materials since headwater streams account for ~ 80 % of all stream mile s (Gomi et al., 2002). Historically, logging has been the most prominent land use in h eadwater streams, highlighting the importance of protecting these systems during this practi ce. In this study, the impacts of logging in Georgias coastal plain had small, but si gnificant impacts on aquatic communities and their food sources. Although strong bottom-up eff ects occurred in the disturbed streams; in general best management practices eff ectively protected the streams during clearcut harvest. Energy Sources Terrestrially derived organic matter is the prim ary resource in many headwater streams (e.g., Vannote et al., 1980). In the sout heastern U.S., this food base is available throughout the year due to the long grow ing season and may be less limiting in undisturbed streams than in temperate zones (Roberts, 2002). In this study, leaf fall

PAGE 75

75 quantity and quality was altered by harvest as well as natural disturbances. A severe drought prior to the study led to a sharp decline in riparian le af fall in all streams, likely due to physiological responses of vegetati on to changes in precipitation regimes. Following harvest, a decline in leaf fall oc curred in the harves ted watersheds, while reference sites continued to accumulate litt er. Thus, the decrease in habitat and food availability was accentuated in response to both the pulse and press disturbances. Additionally, lower C:N ratios in rapidly grow ing herbaceous litter in the thinned SMZs may have given invertebrates access to higher quality food. The decrease in leaf fall was directly related to less storage of BOM in harvested watersheds. As expected, a decrease in canopy in the logged sites decreased organic m atter inputs and availability, and increased algal and macrophyt e biomass. Studies have found logged sites to have significantly lower leaf biomass than reference streams when no buffer strip was established (Golladay et al ., 1989; Stout et al., 1993). However, this study shows a clear loss in orga nic matter storage even with the retention of a protected buffer zone. Although leaf fall from riparian vegetation is closely related to BOM, the contributing area may depend on watershed characteristics. Lateral inputs into the stream (Fisher and Likens, 1973) emphasize the impor tance of maintaining a wide buffer zone. Additionally, leaves are carried into the stream with surface runoff. Thus, the extent of the buffer zone may influence organic matter, altering food and habitat availability for aquatic organisms. Although less BOM storage is linked to changes in canopy cover, factors affecting decom position rates may play a role in BOM loss. In many streams microorganisms and invertebrates are primar ily responsible for decomposition (Petersen

PAGE 76

76 et al., 1989), but this proce ss is additionally linked to ab iotic conditions. Less BOM was stored in the sediment with increases in flow, ammonia, conductivity, and dissolved oxygen. However, these changes were related to factors unique to harvest treatments. In reference streams, a negative relationsh ips with BOM and conductivity, dissolved oxygen, and turbidity suggest an interaction between abiotic a nd biotic factors affecting loss. As dissolved oxygen levels increase d, higher decomposition rates may have been responsible for decreases in BOM. This was likely related to an increase in microbial biofilm and invertebrate abundance and dive rsity typically found at higher dissolved oxygen levels (Allan, 1995). D ecreases in conductivity were linked to flushing of the streams as flow was restored following the drought. Additionally, drying of the streambed releases SO4 2as reduced sulfur is oxidize d, leading to an increase in conductivity (Bayley et al., 1986; Devito, 1999). This increase reduces the solubility of carbon, thus decreasing decomposition rates (Cla rk et al., 1005) and potentially reducing invertebrate abundance. In the harvested watersheds, the stro ng negative relationshi p found between BOM and flow, suggests the loss of BOM is c ontrolled prim arily by physical factors. Movement of organic matter and sediment occurs during most storm events in sandybottomed, coastal plain streams and is more pronounced in the clearcut streams due to increased runoff and peak flow (Golladay et al., 1987). Ultimately, this leads to trapping of litter in discrete, spatially variable habitats such as debris dams (Palmer et al., 1996). Although flow was an important predictor of BOM storage, ammonia and turbidity also played a role. In the streams with an in tact SMZ, BOM increased as the water became more turbid. Reaches draining the intact SMZ retained silt from upstream sections,

PAGE 77

77 leading to habitat smothering, clogging biofilm and gills of macroinvertebrates (Allan, 1995). In the thinned SMZ, stored BOM decr eased with increasing levels of ammonia, reflecting the contribution of b acteria to leaf litter decomposition. Thus, discrete changes in the physical structure of th e stream due to harvest potentially limit ecosystem function and food resources. Although forested headwater streams obtain most of their energy from allochthonous sources, periphyt on is expected to become the dominant food and habitat source as canopy cover is eliminated. In our study, streams with intact buffer zones did not differ from reference streams, however, streams in thinned reaches had periphyton biomass nearly double that of reference str eams. Murphy et al.(1986) reported that clearcut streams averaged 130% greater peri phyton biomass than buffered and old growth streams. Additionally, Brosofske et al.(1997) show ed that logging practices that affect the width of riparian reserves al ong streams also alter the amount of light reaching the stream surface. However, other factors may interact w ith photosynthetically active radiation to determ ine standing stock of periphyton. In a study examining the impacts of selective harvest in Canada, periphyton biomass in creased both as light levels and water temperature increased and buffer width na rrowed (Kiffney et al.2003). Additionally, foodweb structure (Wootton and Power, 1993; Hillet al., 1995) and nutrients (Hillebrand, 2002) can also be important in controlling al gal accrual. In the reference streams, chlorophyll a increased with in creasing total phosphorous. In general, small headwater streams are nutrient limited (phosphorous a nd nitrogen) (Elwood et al., 1981) and thus periphyton responds rapidly to any increase in the water column. Additionally, Hart and

PAGE 78

78 Robinson (1990) found strong bot tom-up effect of phosphorous addition in streams linking increased scraper abundance with hi gher periphyton biomass, emphasizing the importance of nutrients in changing community structure. However, none of the measured variables had a strong relationship with periphyton in th e harvested streams, indicating increases in light may be the primary factor contributing to periphyton biomass in harvested streams. Although few variables explained changes in periphyton due to harvest, discharge levels in these stream s were linked to the the type of primary production established in these streams. Low flow in low-gradient compared to montane streams may allow for growth of macrophytes as well as macroa lga. Colonization by the macrophyte, Ludwigia repens in the harvested streams increased attach ment surfaces available for algal cells. Kedzierski and Smock (2001) found an increase in the macrophyte Sparganium and the algal species Chara in response to logging in coastal plain streams in Virginia, which they linked to increased macroinvertebrate a bundance and biomass. They found that this macrophyte served as an ideal attachment site for filterers (e.g., Simulidae and Rheotanytarsus), thus increasing microhabitat dive rsity in logged reaches. Long-term availability of periphyton and macrophytes may influence invertebrate community structure years after logging, since overstor y canopy cover will take years to decades to limit light penetration (Fuchs et al., 2003). Environmental Variables Water temperature frequently increases in logg ed areas and has complex effects on life cycles of stream bi ota (Hogg andWilliams, 1996). For example, it influences the rate at which eggs develop and juvenile fi sh and invertebrates grow, which, in turn, determines voltinism, rates of growth, and productivity (Allan, 1995; O'Hop et al., 1984;

PAGE 79

79 Wallace and Gurtz 1986). Both winter and summ er temperatures were 1-2 C higher in treatment than reference streams following harv est, indicating long-term effects on biota. While temperatures peaked at 18 C duri ng winter sampling periods, temperatures exceeding 26C were common in the treatment wa tersheds in the late summer, potentially excluding cool-water adapted invertebrates a nd fish. However, this did not result in lower dissolved oxygen concentrations in the treatment watersheds, in part due to the increased growth of macrophytes present th roughout the water column. Additionally, since a 1-2C is the predicted increase in temperature resulting from climate change (Kundzewicz et al., 2007), the longterm effects of this temperature change may be useful for predicting changes in temp erature on aquatic biota. Buffer zones along streams are expected to retain nutrients (Polyakov et al., 2005), allowing them to be taken up by vegeta tion and assimilated in to the terrestrial ecosystem. However, ammonia levels tripled or quadrupled following harvest, increasing from 10-15 g/L to 30-50 g/L. The U.S. EPA standard for ammonia is 27 g/L (USEPA, 1999), the threshold fo r potential toxicity for aqua tic biota. Surprisingly, concentrations were higher in the intact SMZ than in the thinned SMZ. In the intact SMZ, runoff may still reach the streams, but is more likely retained with fine particles and organic matter, contributing to bacterial produc tion. Additionally, retention of silt in these streams likely reduced benthic oxygen, leading to more ammonia. Ammonia adsorbs to silt-clay fractions in stre ams (Silva and Williams, 2001) and may increase retention in these reaches. Additionally, a prescribed burn in the watersheds may have also contributed ammonia to the streams (Knoepp and Swank 1993). However, in the thinned

PAGE 80

80 SMZs there appeared to be a balance between substrate flushing a nd inputs from runoff so that benthic silt levels were reduced. Macroinvertebrates Both taxonomic and trait composition are expe cted to change in respons e to large scale disturbances. The taxonomic similarity within streams increased over time regardless of treatment. This was linked to recovery of the invertebrate community following a long term drought (Griswold et al., 2008). Biological traits were stable over the course of the study; however, there were trends related to harvest. Statzner et al.(2004) also found that traits were relatively stable ove r large temporal and spatial scales in Europe. This stability may be linke d to the finite number of traits that are available within a region based on climat e and geologic features. However, strong changes in environmental conditions are likely to alter stability of trait composition. Winter flow values continued to escalate in the harvested watersheds over the final two years, nearly doubling the flow rate compared to the reference watersheds. A decrease in trait stability during this period in the treatment watersheds suggests that disturbed sites may be less resistant to change. However, the extension of the study over a longer period would be necessary to determine if this is the case. Harvest led to a shift in dominant species associated with chan ges in the food base and environm ental conditions. The species re sponding to harvest in watersheds B and C were different taxonomically, but shared simila r ecological roles. For example, species responding to harvest consumed benthic peri phyton or organic matter present within the water column. In watershed B this included sphaeriid clams, elmid beetles, and mayflies, while in watershed C blackflies ( Simulium ), Tanytarsus midges, and mayflies ( Habrophlebiodes ) increased in abundance. This shif t to from detritus to algae and fine

PAGE 81

81 matter is common in streams impacted by fo rest activities (e.g., Noel et al., 1986). Differences in species composition between the harvested watersheds are likely linked to habitat stability and flow regime. The greatest structural and functional changes occurred in the thinned SMZs, resulting in the lowest B OM and canopy cove r, and the highest periphyton biomass. Species in the thinned SMZs preferred to live in sand and were larger than in the other stream reaches. This preference for sand reflec ts the regular scouring and lack of organic matter in the thinned treatments. Additio nally, larger body sizes and abundance of herbivores are likely related to greater availa bility of periphyton in these treatments. Most studies link herbivore abundance to increased periphyton biomass as a result of logging (Gurtz and Wallace 1984). The intact SMZ was dominated by filterers (e.g., Simulium ), like ly linked to enhanced transport of organic matter from the clearcut section of this watershed. Blackflies require faster flow and a stable source of attachment, conditions provided by larger substrate particle size and abundant macrophytes present in watershed C. The sphaeriid clams present in watershed B prefer slow flow and burrow in the fine sediment. Fine particulates may enter the stream thr ough bank erosion, or lateral inputs from runoff and resuspension providing additional food (Anderson and Sedell, 1979). Harvested watersheds typically export significantly more particulate organic matter than undisturbed reference watersheds (Webster et al., 1990). Additionally, species living in woody debris were an important component in this treatment. Windthrow resulting from the 2004 hurricanes provided organic matter to this system that may have been flushed out in the thinned treatment.

PAGE 82

82 Invertebrates in the refere nce stream s shared traits indicative of an undisturbed forested headwater coastal plain stream. In general, species were soft-bodied and bluff, indicating lower flow and less scouring. When left undisturbed, streams in this region have riparian zones that limit high peak flows during storm events in these streams. For example, soft-bodied tipulid larvae have limite d ability to resist sc ouring and are easily washed downstream. Thus, species with th is trait are adapted to low-gradient, undisturbed streams. Additionally, species in the reference streams were more likely to prefer living in plant material derived from the riparian zone and thus are closely linked to the riparian zone. Thus, biological tra its were accurate predictors of functional changes occurring in watersheds fo llowing a logging disturbance. The utility of using biologi cal traits and fuzzy coding f or linking trophic habits to disturbance lies in the catholic food pr eferences of many inve rtebrates. Species historically thought to be shredde rs supplement their diet with algae, especially when this food source becomes dominant (Zah et al., 2001 ). This flexibility in food choice may limit the ability of bioassessment protocols to detect disturbance. However, many species rely heavily on a primary food source, and li ttle is known of their reproductive capacity when faced with a less preferred food choi ce. Additionally, traits are stable over interannual periods, allowing for more flex ible sampling protocols (Snook and Milner, 2002). Thus, analysis of trophic habitat, combined with fuzzy coding, which allows species to be assigned to multiple groups, will ultimately enhance the robustness of sampling programs. Anthropogenic disturbance in the face of natural disturbances Water quality indices derived from ec ological and taxonomic information on aquatic invertebrates should be responsive to a gradient of disturbances within and

PAGE 83

83 between streams. For instance, many water quality indices were initially derived to understand downstream effects of points source s such as sewage (Kolkwitz and Marsson, 1909). However, the challenge to create indi ces that respond to nonpoint sources as well as multiple stressors has brought this approach to the extent of its limits. In these indices, lower values reflect poor wa ter quality (e.g., pollution tolera nt organisms), while high values indicate good water quality (e.g., po llution sensitive species). However, the Florida SCI was not responsive to the harvest treatments, suggesting the harvest streams had better water quality than the reference streams two years after the harvest. The SCI was highly responsive to natural disturban ces, and values increased from poor water quality to excellent water quality as streams responded to restoration of flow and precipitation following the 1998-2002 drought.. A strong relationship existed between SCI scores and both flow and dissolved oxygen, the prim ary factors responsible for recovery of invertebrate communities following drought (Chapter 2). Harvest created a diverse range of microhabitats (e.g., light and temperature patche s), likely providing more niches for other species. Additionally, discharge was greater in the selective harvest treatment, which may have buffered these streams from any drying ove r the course of the study. Lastly, as periphyton levels increased in the selective harvest treatm ent, increased Ephemeroptera abundance drove the SCI scores higher since this group tends to be a good indicator of water quality. The SCI has not been able to differen tiate between reference and disturbed stream s in other cases. Vowell (2001) did not find evidence that the SCI was able to discriminate between reference and logged sites in Florida. In a survey of 167 headwater

PAGE 84

84 streams in Oregon, Herlihy et al.(2005) also found that environmental variation was a stronger driver of changes in taxonomic composition than logging history. Further support exists for the short-tem impact of harvest on streams. Kreutzweiser et al.(2005) only found an initial peak in scrapers and filterers immediately following harvest in watersheds with selective ha rvest. They also found that taxonomic structure differed among headwater streams with si milar characteristics within the same basin, providing further support for the use of biological trai ts in bioassessment. Given the predicted increase in natural disturbances, the valu e of these indices becomes questionable for detecting anthropogenic disturba nces. However, they have been used successfully for detecting large disturbances such as urba nization and agricultural practices. Describing and understanding va riability in stream system s is difficult because processes and patterns vary at different spat ial and temporal scales (Wiens et al., 1986; Roth et al., 1996; Allan and Lammert, 1999). Assemblages can vary at small spatial scales, yet appear stable, or at least resilient, at larger scales (Rahel, 1990). This phenomenon has been referred to as the shifting mosaic, steady-stat e model (Clark, 1991; Moloney and Levin, 1996). The study streams were exposed to two press disturbances and at least one pulse disturbance over a de cade. The former included a drought lasting from 1998-2002, logging in 2003, and a hurricane in 2004. The enhanced discharge resulting from the storms did not influence taxonomic composition or trait structure. However, the impacts of harvest and drought, discussed here and in Chapter 2, indicate that natural variab ility needs to be taken into account when attempting to link changes in land use to changes in structural and f unctional aspects of aquatic ecosystems.

PAGE 85

85 Changes in forestry management practic es over the past couple decades have driven the need to understand im pacts of logging along streams on water quality and biodiversity. The assumption cannot be made that simply leaving a few trees along the stream will protect it from land use within the watershed. Thus, more state management programs have incorporated watershed slope into the equation for determining buffer width (e.g., Georgia Forestry Commission, 1999). This study found evidence for longterm impacts of properly managed SMZs on a quatic biodiversity and basal resources. However, these effects were most pronounced in the first year following harvest. Thus, models examining the impacts of SMZ manageme nt must incorporate direct and indirect effects of forestry activities.

PAGE 86

86 Table 3-1. Biological trait definitions and modalities. Trait CodeModality Trait CodeModality Life History Ecology Voltinism v1SemivoltineHabith1Clingers v2Univoltine h2Burrowers v3Multivoltine h3Swimmer Drying Resistance d1Absent h4Sprawler d2Present h5Skater Eggs cemented to substrateec1Yes h6Climber ec2No Trophictr1Gatherer Development Time ds1< 6 weeks tr2Filterer ds2< 1 year tr3Scraper/Herbivore ds3> 1 year tr4Shredder Egg Hatch Time ht1< 1 week tr5Predator ht2< 1 monthRheophilyr1Standing ht3> 1 month r2Slow Mobility r3Fast Laminar Drift df1Rare r4Fast Turbulent df2Common Microhabitatmh1Sand df3Abundant mh2Rock Morphology mh3Gravel Armoring ar1Soft mh4Macrophyte/Algae ar2Sclerotized mh5Detritus ar3Case/Shell mh6Woody debris Maximum Size s1Small (<9mm) mh7Silt s2Medium (9-16mm) s3Large (>16mm) Shape sh1Streamlined sh2Not Streamlined (Bluff, Tubular) Respiration rs1Cutaneous rs2Tracheal Gills rs3Spirales/Plastron

PAGE 87

87 Table 3-2. Results of multiple regressions for chlorophyll a biomass and benthic organic m atter (BOM). Significance of R2 values is given by ( P < 0.05), ** ( P < 0.01), *** ( P < 0.001). Response Variable TreatmentRegression Equation R2ChlaReference streams .02 + 0.005 ( TP ) 0.27*** Intact SMZNSNS Thinned SMZ -0.14 + 0.07 ( SC ) .36* BOMReference streams 2.6 0.47( SC ) 0.24 ( DO ) 0.18(Turbidit y ) 0.67*** Intact SMZ 0.55 0.39( Flow ) + 0.64 ( Turbidit y ) 0.31* Thinned SMZ -0.21 .38 ( Flow ) 0.2 ( A mmonia ) 0.50***

PAGE 88

88 Table 3-3. Average environmental conditions for winter sampling pe riods in referen ce (A,D), thin ned SMZs (B1,C1), and intact SMZs (B2,C2). Data are for pre-harvest (2001-2003) and post-ha rvest (2004-2008). YearFlow (L/s)TSS (g/L) NH4 (g/L) o-phosphate (g/L) NO2/NO3 (g/L) Total Phosphorous (g/L) Total Nitrogen (g/L) pH SC (uS/cm) DO (mg/L) Turbidity (NTU) Temperature ( C) Leaffall (g/m2) A2001-20020.9980.0152.482.731.088.22238.005.5342.284.471.7816.1334.29 2002-20032.5200.0016.281.750.003.69278.284.7330.607.340.1912.1512.29 2003-20041.0410.01711.971.850.0013.26345.875.0335.954.511.1816.0819.26 2004-20051.7590.0032.882.492.344.90233.304.8724.907.621.2312.0021.26 2005-20061.4720.0136.201.780.0016.94343.305.3526.856.871.1013.6330.19 2006-20072.5790.0083.933.010.009.61291.875.1133.688.990.6313.0524.55 D2001-20020.0190.0044.5445.469.4277.00212.467.2384.035.254.0315.6838.25 2002-20032.7350.0040.0027.247.8551.25285.975.8894.857.852.9512.7316.23 2003-20043.0680.0087.6523.744.4151.75237.736.8570.906.824.0315.9022.18 2004-20053.3860.0031.6818.863.0039.00232.056.6174.989.392.9811.7326.06 2005-20065.3610.0166.5912.439.4030.25218.547.1260.938.794.5113.0529.76 2006-20073.9000.0014.8525.692.3827.36203.767.0982.809.882.6112.0525.84 B12001-20021.6510.00312.0902.880366.82011.810621.0906.65061.6505.1203.95016.40038.770 2002-20033.6500.0069.482.24412.5310.5078.194.8596.156.965.6512.4511.18 2003-20044.0500.00419.102.12655.404.83970.006.2489.506.326.5517.107.53 2004-20055.9700.00723.412.41891.942.571165.216.1572.008.655.1814.158.30 2005-20065.4100.00531.801.901041.0030.681256.396.6171.408.064.5514.559.02 2006-200710.3700.00217.882.27346.906.43535.036.2687.958.404.1913.058.75 B22001-20021.9900.00210.782.39835.908.421058.006.5080.804.713.4016.7051.13 2002-20032.6500.0031.892.56824.674.691182.005.1582.657.423.3012.6021.10 2003-20042.9300.00727.801.801161.006.841230.006.1977.006.493.8517.5511.18 2004-20054.4300.00429.652.461480.004.781655.005.9766.758.224.2114.2512.75 2005-20064.1500.00543.361.781523.005.051813.006.4868.407.584.2414.6012.35 2006-20077.4500.00331.272.52656.206.53866.906.2271.358.593.1513.0011.06 C12001-20020.0880.0016.895.611099.0010.111344.007.80101.558.134.1513.7039.97 2002-20033.7300.0010.004.281189.008.521580.005.80106.309.173.1512.3014.15 2003-20045.2800.00513.054.97900.8015.771075.006.7988.707.639.2017.1513.73 2004-200510.1000.00314.704.45816.9010.80969.306.5575.209.215.4412.759.42 2005-20069.9000.00830.554.74998.9016.201290.006.9078.908.198.8314.2515.57 2006-20079.8300.00116.586.20818.207.941004.006.74103.909.833.3012.7511.59 C22001-20021.6300.00314.183.891386.0012.371541.007.4584.906.875.1515.9024.20 2002-20033.2900.00214.942.561541.005.932062.005.9589.958.523.0513.4016.10 2003-20046.1900.00516.924.261204.0013.051413.006.5775.657.227.4017.2510.55 2004-20057.4500.00626.482.821094.0011.351267.006.3268.108.975.9313.1521.17 2005-20066.7600.01348.333.421338.0021.101653.006.7870.907.708.0814.9519.70 2006-20077.6100.00216.752.401277.007.861410.006.4691.508.714.6213.2017.30

PAGE 89

89 Table 3-4. Indicator values for watersheds A an d B based on taxonomic composition. Groups are defined as pre-harvest all sites (1), pos t-harvest reference (2), post-harvest thinned SMZ (3), and post-harvest intact SMZs (4). GroupIndicator Valuep-value Parachaetocladius 137.50.007 Alotanypus 244.30.000 Caecidiota 239.50.012 Corethrella 227.80.040 Crangonyx 238.60.000 Ptilostomis 2 43.70.001 Sciomyzidae 2 44.70.005 Stenochironomus 2 39.80.008 Ablabesmyia 3 44.20.004 Calopteryx 3 36.60.012 Cheumatopsyche 3 26.40.042 Cryptochironomus 3 41.80.008 Habrophlebiodes 3 35.50.020 Hemerodromia 3 250.045 Orthocladius 3 30.30.039 Paralauterborniella 3 42.30.005 Peltodytes 3 300.016 Sphaerium 3 44.80.004 Stenelmis 3 49.60.001 Tanytarsus 3 34.50.025 Thienemaniella 3 30.30.040 Anisocentropus 4 32.40.020 Hexatoma 4 34.20.016 Procladius 4 37.10.019

PAGE 90

90 Table 3-5. Indicator values for watersheds C an d D based on taxonomic composition. Groups are defined as pre-harvest all sites (1), pos t-harvest reference (2), post-harvest thinned SMZ (3), and post-harvest intact SMZs (4). GroupIndicator Valuep-value Parachaetocladius 131.20.023 Hexatoma 131.60.034 Leptophlebia 135.70.019 Corynoneura 234.20.049 Thienemaniella 244.30.003 Nippotipula 2 39.50.002 Pseudolimnophila 2 340.002 Bezzia 2 36.40.001 Alluaudomyia 2 35.40.038 Dixella 2 48.90.001 Psychoda 2 500.001 Sciomyzidae 2 37.50.010 Ophiogomphus 2 54.90.000 Cordulegaster 2 42.60.005 Amphinemura 2 36.70.027 Perlesta 2 42.50.006 Allocapnia 2 380.018 Anisocentropus 2 37.10.028 Helichus 2 41.80.009 Neoporus 2 36.10.035 Microvelia 2 46.70.002 Paracladopelma 3 41.10.004 Brillia 3 34.80.006 Cambaridae3 45.60.001 Elimia 3 38.40.005 Cryptochironomus 4 33.50.022 Polypedilum 4 30.40.039 Stenochironomus 4 42.10.008 Tanytarsus 4 33.60.026 Tribelos 4 36.20.019 Hexagenia 4 34.60.033 Molanna 4 24.20.031 Triaenodes 4 27.60.043 Laevapex 4 39.80.005

PAGE 91

91 Table 3-6. Indicator values for watersheds A and B based on biological traits. Groups are defined as pre-harvest all sites (1), post-h a rvest reference (2), post-harvest thinned SMZ (3), and post-harvest intact SMZs (4). TraitGroupIndicator Valuep-value df3131.30.003 ar2131.90.014 h31380.000 tr1129.90.007 mh2128.50.015 ds2128.30.036 ht2132.10.003 ar1226.60.005 tr5229.40.007 sh2226.20.008 mh4229.70.002 ds12290.004 s2332.10.034 h4328.80.030 mh13300.040 mh3332.20.014 v1436.50.026 df14270.042 tr2433.80.007

PAGE 92

92 Table 3-7. Indicator values for watersheds C and D based on biological traits. Groups are defined as pre-harvest all sites (1), post-h a rvest reference (2), post-harvest thinned SMZ (3), and post-harvest intact SMZs (4). TraitGroupIndicator Valuep-value v2127.10.012 ar2129.90.013 h3 1 32.70.023 tr4 1 33.10.001 mh7 1 30.90.002 ds2 1 30.30.001 ht2 1 29.50.007 h5 2 44.70.003 mh5 2 320.003 rs3 2 33.30.009 tr3 3 29.90.010 mh1 3 28.90.023 mh6 4 30.30.022 ec2 4 26.30.047

PAGE 93

93 Figure 3-1. Topographic map a nd aerial photo of the four study watersheds (A-D).

PAGE 94

94 0 0.02 0.04 0.06 0.08 0.1 0.12 ReferenceThinnedIntactChla (mg/m2) Wet Season Dry Season Figure 3-2. Average chlorophyll a biomass ( SE) during the wet (May-S eptember) and dry season (October-April) from 2004-2008 in refe rence, thinned SMZs, and intact SMZ streams after harvest.

PAGE 95

95 0 10 20 30 40 50 60 70 80Reference Thinned SMZ Intact SMZC:N Ratio Pre-harvest Post-harvest Figure 3-3. C:N ratios of leaf fall from the ripari an zone in reference and harvested watersheds before (2001-2003) and af ter (2004-2007) harvest.

PAGE 96

96 0.00 10.00 20.00 30.00 40.00 50.00 60.002001-20022002-20032003-20042004-20052005-20062006-2007NH4 (ug/L) Reference Thinned Intact SMZ Figure 3-4. Average ammonia (NH4) concentrations (SE) in reference, thinned SMZs, and intact SMZ streams. Harvest treatments were applied prior to the third sampling period.

PAGE 97

97 0 10 20 30 40 50 60 70 80Stream Condition Index (SCI) Reference Thinned SMZ Intact SMZ 2002 2003 2004 2005 2006 2007 Figure 3-5. Stream condition inde x (SCI) scores (SE) for re fere nce, thinned SMZs, and intact SMZ streams. Samples below the red line i ndicate poor water quality, those above the red line, fair water quality, and those a bove the blue line, good water quality.

PAGE 98

98 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 2001-20022002-20032003-20042004-20052005-20062006-2007Bray-Curtis Values Intact SMZ Thinned SMZ Reference Figure 3-6. Taxonomic stability (SE) for refere nce, thinned SMZs, and intact SMZ stream s.

PAGE 99

99 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 2001-20022002-20032003-20042004-20052005-20062006-2007Bray-Curtis Values Intact Thinned SMZ Reference Figure 3-7. Trait stability (SE) for referen ce, thinned SMZs, and intact SMZ stream s.

PAGE 100

100 NH4 TN pH SC DO Turb Flow LF -1.0 -2.0 0.01.0 -1.0 0.0 1.0Axis 1Axis 3 Harv 1 2 3 4 Figure 3-8. NMDS of taxonomic composition in watersheds A and B in pre-harves t (1) and in post-harvest reference (2), thinned SMZs (3), and intact SMZ treatments (4).

PAGE 101

101 NH4 Phosph TN DO Turb Temp Flow LF -2.0 -1.5 -1.00.01.0 -0.5 0.5 1.5Axis 1Axis 2 Harv 1 2 3 4 Figure 3-9. NMDS of taxonomic composition in watersheds C and D in pre-harves t (1) and in post-harvest reference (2), thinned SMZs (3), and intact SMZ treatments (4).

PAGE 102

102 -2 -2.0 024 -1.0 0.0 1.0Axis 1Axis 2 Harv 1 2 3 4 v1 v2 v3 d1 d2 df1 df2 df3 ar1 ar2 ar3 s1 s2 s3 r1 r2 r3 r4 h1 h2 h3 h4 h5 h6 tr1 tr2 tr3 tr4 tr5 sh1 sh2 mh1 mh2 mh3 mh4 mh5 mh6 mh7 rs1 rs2 rs3 es1 es2 ec1 ec2 ds1 ds2 ds3 ht1 ht2 ht3 -0.3 -0.3 -0.10.10.3 -0.1 0.1 0.3Axis 1Axis 2 Harv 1 2 3 4 Tss NH4 TN SC DO Turb Flow LF -2 -2.0 024 -1.0 0.0 1.0Axis 1Axis 2 Harv 1 2 3 4 Figure 3-10. NMDS of biological tr aits in watersheds A and B in pre-harvest (1) and in postharvest reference (2), thinned SMZs (3), and intact SMZ treatments (4).

PAGE 103

103 -3 -2.0 -11 -1.0 0.0 1.0 2.0Axis 1Axis 2 Harv 1 2 3 4. v1 v2 v3 d1 d2 df1 df2 df3 ar1 ar2 ar3 s1 s2 s3 r1 r2 r3 r4 h1 h2 h3 h4 h5 h6 tr1 tr2 tr3 tr4 tr5 sh1 sh2 mh1 mh2 mh3 mh4 mh5 mh6 mh7 rs1 rs2 rs3 es1 es2 ec1 ec2 ds1 ds2 ds3 ht1 ht2 ht3 -0.3 -0.15 -0.10.10.3 -0.05 0.05 0.15Axis 1Axis 2 Harv 1 2 3 4 Figure 3-11. NMDS of biological tr aits in watersheds C and D in pre-harvest (1) and in postharvest reference (2), thinned SMZs (3), and intact SMZ treatments (4).

PAGE 104

104 CHAPTER 4 EFFECTS OF PATCH TYPE, QUALITY, AND SIZE ON MACROINVERTEBRATE COMMUNI TY STRUCTURE Introduction Spatial heterogeneity in landscapes strongl y affects comm unity structure (Bond et. al 2000) and population dynamics (Kar eiva 1990). In the simplest case, increasing either habitat types or patches provides more potential niches, allowing in creased species diversity and abundance. However, certain habitat types may be better suited for a species or group of species, resulting in preferential habitat selections. Physical characteristics, environmental conditions (e.g., oxygen, temperature), food resources and predation risk are important factors determining patch suitability, and spa tial aspects of food webs are key to understanding community structure and dynamics (Holt 1977, 1996). Low gradient, headwater streams in the southeastern coastal plain are typically dominated by sandy substrate, yet they also have patches of leaf packs, woody debris and root m ats. Two important habitat patch types in logged streams are leaf packs and macrophyte beds, both of which vary in quality and quantity over space an d time. Thus, streambed heterogeneity results from seasonal inputs of organic matter and the rearrangement of thes e patches in shifting mosaics (Stout et al., 1985; Hildrew and Giller, 1994; Wallace et al., 1995).Leaf packs can peak in autumn after leaf senescence, while macrophytes peak in the late spring and summer during the growing season. However, both are present at some level throughout the year. Thus, patch size and location will be highly variable and depend on changes in canopy cover and allocthonous leaf input. Coloni zation of streambeds by macrophyt es, coupled with decreased allochthonous input in logged streams, can alter the number of patches available for stream biota. Temporal and spatial changes in stream landscap es lead to changes in size, isolation and structure of habitat patches. Leaf packs and m acrophytes potentially differ in their temporal and

PAGE 105

105 spatial dynamics. Depending on stream bed stability and flashiness of flow, the structure of leaf packs can change greatly with time (Palmer 1996, Velasquez 2003). In logged streams, the rate of leaf pack formation is often slow, resulting in increased patch isolation and fragmentation. Leaf pack formation occurs primarily in autumn as leaves senesce and fall into the stream. They then become smaller and are rearranged within the stream as invertebrates process leaves and increased flow scours the channel bottom. Un like leaf packs, rooted macrophytes are more stable in streams and contribute to a less dyna mic streambed landscape. Thus, macrophytes may support superior competitors, while leaf packs may support more transient, inferior competitors. Differences in patch quality and size drive hab ita t selection by stream invertebrates. Leaf packs vary in suitability as ha bitat and food, with fast decomp osing leaves acting more as a resource, and slow decomposing leaves acting more as habitat (Dangles et al.2001). Essafi (1994) found that invertebrate biomass did not decrease in leaf p acks after leaves lost most of their nutritional value, suggesting that leaf packs act as habitat and potential refugia in addition to being a consumable resource. Leaf packs may also enter the hyporheic zone and provide resources for subsurface biota (Strommer and Smock 1989). Although few invertebrates consume macrophytes, high quality resources fo r invertebrates exist in macrophyte beds, including decaying plants, root exudates, root associated bacteria, epilithic periphyton and detritus (Sagova 2002, Tolonen 2003). Previous studies of static stream landscap es suggest that sm all patches support higher densities than large aggregated patches (P almer 2000, Silver 2000). However, these studies focused primarily on leaf packs and a small gr oup of organisms (chironomids and copepods). Silver et al.(2004b) observed that chironomid density was greater in both fragmented landscapes and less stable habitats.

PAGE 106

106 Given the current state of knowledge of patch dynam ics in streams, the goal of the this study was to understand the role of patch type, size, and quality in structuring invertebrate communities and colonization dynamics. It wa s hypothesized that larger macrophyte patches provide both more cover a nd a greater source of food (e piphyton/biofilm) and increase invertebrate abundance and diversity. Invertebrate abundance and diversity should be lower in large leaf packs because their interior will offer reduced water velocity and oxygen. This was assessed using a combination of field observation s and experimental manipulations of patches. Materials and Methods Field Sampling of Patches Macrophytes and leaf packs were mapped three tim es over a year at 20 m intervals along the length of the study reach (~ 200 m) by establishing transect s perpendicular to flow and determining percent cover of leaf packs and macrophytes. Samples were obtained from randomly selected patches of macrophytes ( Ludwigia repens) and leaf packs four times between September 2005 and June 2006. A net (250 m mesh) was positioned downstream of the patch, and three leaves were taken from each patch a nd placed in individual vi als containing 100 ml of deionized water for chlorophyll a analysis. Three additional l eaves were preserved in phosphobuffered formalin (1%) for bacteria counts. Leaf samples were kept on ice until returning to the lab where they were stored at -20C until analyz ed. The remainder of the patch was collected by removing only its above-sediment portion and allowing it to drift into the ne t. The contents of the net were placed in a ziplock bag, placed on ice, and returned to the laboratory for processing. In the laboratory, patch samples were gently rinsed through nested sieves of 1mm (CPOM) and 0.250 m (FPOM). Macroinvertebr ates were sorted from the samples and preserved in 70% ethanol. Each sample was th en separated into terre strially derived CPOM, FPOM, and macrophytes. Patch size was determin ed by placing each sample in a 500ml glass

PAGE 107

107 cylinder with water to determine the volume of the patch by water displaced. Each portion (CPOM, FPOM, and macrophytes) was then dried at 60C for at least 48 hours and weighed to compare dry weights to volume for a given area. FPOM samples were ashed at 550C for five hours to determine organic content. Since many invertebrates consume bacteria and periphyton, chlorophyll a, ash-free dry weight (AFDW ), and bacteria counts were used as indicators of patch quality. Chlorophyll a and AFDW were analyzed as in Chapter 3. Leaves for chlorophyll and AFDW measurement were vigorously shaken in 50 mL of deionized water for 30s, after which leaves were removed to measure surface area. Water samples were filt ered through 45 m GFF filters, and bacteria on the filters were stained with SYBR Green and counted under an epiflourescent microscope. Leaves were photographed, and Scion Image (Scion Corp., Frederick, MD, U.S.A.) was used to calculate total surface area. Bacteria enumeration followed the protocol outlined in Buesing (2005). A 0.2 m, 25mm aluminum oxide membrane filter (Whatm an Anodisc) was placed on top of a wetted 0.45 m, 25 mm cellulose nitrate filter on a glass filter manifold. Leav es for bacteria counts were thawed and sonicated for one minute at 80 W while on ice. Then, the sample was vortexed, and a 100 l aliquot was removed after 10s. The sample and one ml nanopure water was added to the filter manifold to ensure mixing and a homog eneous slide mount and pressure applied using a vacuum. The Anodisc filter was removed and gently dr ied using a Kimwipe. Filters were placed face-up on a 100ul drop of SYBR Green II fluorescent stain diluted 400 fold (Molecular Probes, Eugene, Oregon, USA) on labelled petri dishes. Filters were stained in the dark for 15 minutes, then dried and placed face-up on a glass slide. A 30-uL drop of antifade mounting solution (50%

PAGE 108

108 glycerol, 0.1% p-phenylenediamine, 50% PBS: 120 mM NaCl, 10 mM NaH2PO4, pH 7.5) was added, and a cover slip was placed on top. Slides we re then counted or stored frozen at -20C for up to six weeks. An Epifluorescence microscope equipped with a high-pressure mercury lamp (HPO 100 W), with a Chroma filter set ( no. 41001; excitation filter 480 nm, beam splitter 505 nm, emission filter 530 nm) was used to count bacteria. Cell numbers were counted from at least 10 fields until a total of 400 bacterial cel ls was reached (Kirchman 1993). Preliminary counts from ~ 25 slides were used to determine a size class distri bution with a calibrated micrometer and placed in the following classes: cocci (< 0.5 m, > 0.5 m diameter), vibrio, filamentous, and rod (< 0.35 m, > 0.35 m). In subsequent slides, at least 15 cells from each size class were measured. Volumes (V) of individual ce lls wer e calculated under the assumption that cells are cylindrical with hemispheric e nds (Fry, 1988), which works for both rods and cocci. The total biovolume (BV) of bacterial cells per g of leaf material was calculated as: lcf fsi iDMAS AVb DM BV )( where bi is the biovolume of an individual bacteria cell, Vs the sample volume, Af the total filtration area, Sf the volume of the subsample passed over the filter, Ac the filtration area, in which bacteria were counted, and DMl the litter dry mass. Bacterial dry mass or carbon was calculated from bacteria BV based on empirically determ ined conversion factors. For pelagic freshwater bacter ia, LofererKrbacher et al.(1998) established the following relationship: bv dmb435 0.86 where dmb is the dry mass and bv the biovolume of a bacteria cell.

PAGE 109

109 Field Experiment The goal of the field experiment was to contro l for leaf species, patch size, and patch age to exam ine initial macroinvertebrate colonization patterns. Leaf packs cons isted of dominant tree species shared among watersheds B and C; Liriodendron tulipifera Quercus nigra and Pinus spp. Leaves were collected in August 2006 prior to abcission and air dried for seven days. Macrophytes were collected from seeps along the st ream, washed thoroughly in distilled water, and examined for invertebrates and biofilm before use. The macrophytes and leaf species were used to create patches of 1, 2, or 4 g. A separa te set of ten macrophyte samples were dried at 60C to determine a wet to dry mass regression and cr eate an equivalent to the leaf packs prior to the beginning of the experiment. The three size classes were crossed with two lev els of stab ility and four species in a randomized block design. Blocks were created in a 10-20 m stretch of stream and replicated three times along at 70 m length of each reach in the intact and thinned SMZ treatements in watersheds B and C. Leaf packs were created by loosely tying leaves toge ther using nylon line. Macrophyte patches were anchored in the sedime nt using mesh produce bags (10 Vexar bags, Avis Bag Co.). Leaf packs and macrophytes were teth ered to pvc pipe driven into the streambed. Stable patches were left undisturbed for 15 days, while unstable patches were disturbed once on day 7 by rinsing the patch through the water column for one minute. Patches were then collected after 7 and 15 days to determine colonization patterns. Velocity, oxygen, and canopy cover were m easured at each patch as potential determ inants of patch quality. Canopy cover was measured by taking four measurements using a densitometer. Velocity was measured using a Marsh McBirney Flowmate 2000 (Frederick,MD). Oxygen samples were taken by firs t removing a 10ml water sample from the patch with a 10 ml

PAGE 110

110 syringe. Then, dissolved oxygen was measured using the micro winkler technique (Peck and Uglow, 1990). Samples were fixed within two h ours and returned to the lab for processing. Leaf packs were rinsed through a 250 m mesh sieve, and inve rtebrates were sorted from the sample and identified. CPOM and FPOM trappe d in the patch were se parated, dried at 60C for 48 hours, and weighed. Additionally, subsamples were taken and ashed at 550C to correct for inorganic accumulation on leaf litter. Data Analysis Field obervations Independent and dependent variables were transformed to meet assumptions of normality and independence. Analysis of covarian ce (ANC OVA) was utilized to determine the relationship between invertebrates and patch type using size as a covariate. Multiple regression was used to relate invertebrate metrics to pa tch size, epiphyton biomass, and bacteria abundance and biomass. Experimental manipulation of patches A three-way ANOVA was utilized to relate changes in macroinvertebrate metrics to initial leaf mass, species, a nd disturbance. Rarefaction wa s used to com pare taxon richness across samples after standardizing for patch size. Linear regression wa s used to relate the expected number of species to the patch size. Multiple regressions were used to examine the influence of canopy cover, dissolved oxygen, trapped FPOM and CPOM, and velocity on macroinvertebrates. Results Field Observations The total biomass of epiphyton (chlorophyll a) was not related to patch type or size. Values were lower during autumn/w inter than in spring/summer (F3,122 = 4.2, P<0.01) (Fig. 4-1).

PAGE 111

111 Differences in total bacterial abundance relate d to patch type were dependent on patch size (F1,115 = 4.6, P<0.05). Overall, abundance increased w ith linearly with patch size, but the slope was greater for leaf packs than macrophytes. Temporally the number of b acteria cells decreased from fall to winter and incr eased from spring to summer (F3,119 = 5.6, P<0.01), but did not depend on patch size (Fig. 4-2). Total bacterial biomass was higher in leaf packs than on macrophytes (F1,118 = 11.8, P<0.001), but was not influenced by patch size. Total biomass per patch changed with date, but was related to the size of the patch (F3,114 = 4.0, P<0.01). In general, biomass increased with patch size, bu t an outlier led to high biomass in a small macrophyte patch. FPOM trapped within patches increased in both patch types from fall to summer (F3,122 = 15.0, P<0.0001). FPOM changed with patc h type, but was dependent on patch size (F1,1118 = 25.8, P<0.0001). FPOM in leaf packs incr eased with patch size, but did not change with patch size of macrophytes. Further, the volume of FPOM trapped in leaf packs was significantly higher than that in macrophytes (29.4 vs. 21.2 cm3). Patch quality parameters were weighted for patch size. Bacterial biom ass/cm3 changed with date (F3,116 = 10.2, P<0.0001), but depended on patch type (F3,116 = 3.7, P<0.02). Biomass was greatest in November for both leaf packs and macrophytes, but was higher in leaf packs (Fig. 4-3). Bacterial abundance/cm3 changed with date (F3,116 = 12.4, P<0.0001), but depended on patch type (F3,116 = 10.0, P<0.0001). Abundance was lowest in April for both patch types (Fig. 4-4). Chla/cm3 changed with date (F3,122 = 4.0, P<0.001) and patch type (F1,122 = 6.4, P<0.02) and was higher for macrophytes on a ll dates except November (Fig. 4-5). Changes in taxon richness with date (F3,118 = 6.0, P<0.01) and patch type (F1,118 = 6.7, P<0.001) were dependent on patch size. Taxon richne ss increased with patch size for leaf packs, but did not have any relationship to patch size for macrophytes. After ad justing taxon richness

PAGE 112

112 based on taxa/cm3, there was a significant e ffect of patch type (F1,116 = 53.8, P<0.001) date (F3,116 = 12.6, P<0.001) and their interaction (F3,116 = 19.2, P<0.001) on taxon richness. Taxon richness was higher in Ludwigia and greatest in the winter samp ling period (Fig. 4-6). Changes in invertebrate abundance with date (F3,116 = 7.1, P<0.001) and patch type (F1,116 = 17.5, P<0.001) were dependent on patch size. There was a positive relationship between patch size and leaf packs, but no relationship between patch size and macrophytes. After adjusting abundance based on individuals/cm3, patch type (F1,116 = 43.8, P<0.001) date (F3,116 = 10.8, P<0.001) and their interaction (F3,116 = 13.3, P<0.001) significantly affected abundance. In general, invertebrates were more abundant in Ludwigia peaking during winter (Fig. 4-7). Changes in the proportion of shredders did not depend on patch size, but were different between patch types (F1,122 = 8.3, P<0.01) and over time (F3,122 = 3.5, P<0.02). Shredders were more common in leaf packs moreso in the winter than in the summer. Differences in filterers with patch type depended on patch size (F1,118 = 12.9, P<0.001). Filterers were positively related to patch size in leaf packs, but did not display any relationship to patch size in macrophytes, and they were more abundant in summer than fall and winter (F3,122 = 5.9, P<0.01) (Fig. 4-8). Predators did not change significantly over time but differences betw een patches depended on patch size (F1,118 = 10.1, P<0.001). Predators increased w ith patch size in leaf packs, but decreased with size in macrophytes. However, predators were more abundant in macrophytes than in leaf packs ( 25 vs. 20 percent of comm unity composition). Collector-gatherers did not change over time, but the effects of patch type differed by patch size (F3,118 = 2.7, P<0.05). They increased with patch size in macrophytes, but de creased with size in l eaf packs. Overall, collector-gatherers comprised a larger proportion in leaf packs than in macrophytes (34 vs 23 percent).

PAGE 113

113 Regressions. Invertebrate abundance had a positive relationship with both bacterial biom ass and FPOM for leaf packs, but was only related to FPOM for macrophytes. Taxon richness was positively related to bacterial biomass and FPOM for leaf packs and FPOM for macrophytes. The proportion of shredders was positiv ely related to patch size and negatively to bacterial abundance for leaf packs and did not relate to any para meter for macrophytes. Scrapers were positively related to patch size, chlor ophyll a and bacterial a bundance for macrophytes. Filterers were positively related to bacterial a bundance, biomass, and FPOM for leaf packs and to FPOM and chlorophyll a for macrophytes. Colle ctor-gatherers were related to chlorophyll a and FPOM for macrophytes (Tables 4-1, 4-2). Field Experiment C:N ratios varied among leaf species with Pinus (40.29) having the highest and Ludwigia (14.7) have the lowest ratio. Quercus and Liriodendron were sim ilar with ratios of 30.5 and 26.6, respectively. Leaf mass decomposition ov er time was dependent on leaf species (F2,273 = 148.4, P<0.001), disturbance (F2,273 = 9.9, P<0.001), and mass (F2,273 = 205.5, P<0.001). Pinus and Liriodendron lost two to three times more mass than Quercus patches (Fig. 4-9). The percent of leaf mass loss increased with time, but did not differ between disturbance treatments. Larger leaf packs lost more mass over time than smaller leaf packs, with 4-gram packs losing five times more than 1-gram packs (Fig. 4-10) However, when corrected for percent loss over time, mass was not significant. On average, leav es lost twenty five percent of their mass, regardless of initial mass. Changes in velocity due to disturbance depended on mass (F4,408 = 5.8, P<0.001) and leaf species (F6,408 = 2.2, P<0.04). Average velocity ranged from 0.04 to 0.06 cm/s. In general velocities were lower in Liriodendron and higher in larger leaf pa cks. CPOM trapped within patches was related to leaf species (F3,372 = 50.8, P <0.001), disturbance (F2,372 = 4.7, P<0.01),

PAGE 114

114 and mass (F2,372 = 5.9, P<0.01). More CPOM was trapped in the 4 g patches than the 1 and 2 g patches (Fig. 4-11). The amount of CPOM tra pped in patches ranged from 0.2 to 1.7 grams. The most CPOM was trapped in Ludwigia and the least in Pinus (Fig. 4-12). Additionally, more CPOM was trapped in patches that were collected after fifteen da ys and were not disturbed (Fig. 4-13). FPOM trapped within patche s was related to leaf species (F3,406 = 40.9, P<0.0001), disturbance (F2,406 = 5.7, P<0.01), and mass (F2,406 = 11.1, P<0.0001). More FPOM was trapped over time and with increasing patch size (Fi g. 4-14).The amount of FPOM ranged from 0.05 to 0.3 g and was greatest in Ludwigia and least in Pinus and Quercus (Fig. 4-15). Oxygen within patches was not different be tween any treatment. Invertebrate abundance changed sign ificantly between leaf species (F3,402 = 5.3, P<0.01) and initial leaf mass (F2,402 = 12.6, P<0.001). Abundance was lowest in Ludwigia with an average of 15 individuals and highest in Pinus and Liriodendron with an average of 30 individuals (Fig. 4-16). Inverteb rate abundance increased with increasing patch size, from 20 to 44 individuals (Fig. 4-17). However, there was no apparent evidence for a relationship between patch size and expected species richness whe sample size was accounted for. Taxon richness changed significantly between leaf species (F3,402 = 6.6, P<0.001) and in itial leaf mass (F2,402 = 19.5, P<0.001). In general, the number of taxa di d not differ greatly, averaging between 3 and 5, with Quercus patches having the least numbe r of taxa (Fig. 4-18). The proportion of predators did not differ betw een treatm ents. The proportion of scrapers was dependent on initial leaf mass (F2,402 = 4.3, P<0.02) and was higher in the 4 g than 1 g patches (Fig. 4-19). The proportion of shredders changed in response to mass (F2,402 = 3.1, P<0.05), leaf type (F3,402 = 3.4, P<0.02), and disturbance (F2,402 = 5.6, P<0.01), however, the effect of leaf type depend ed on disturbance treatment (F6,402 = 2.5, P<0.03). In general,

PAGE 115

115 shredders became more common over time, more so in the undisturbed treatments, while more abundant in the largest patches, they were not abundant overall and only ranged from 0-6 percent of the community (Figs. 4-20,4-21). The proportion of filterers changed in response to mass (F2,402 = 5.9, P < 0.01 and leaf type (F3,402 = 10.4, P<0.0001) and were twice as common in Ludwigia than any other patch type and were more abundant in larger patc hes (Fig. 4-22,4-23). Collector-gatherers were the dominant feeding group in all patches, ranging from 40-60 percent of the community. Collector-gathere rs differed between leaf species (F3,402 = 6.02, P<0.001) and were least abundant in Ludwigia (Fig. 4-24). Regressions Invertebrate abundance was positively relate d to CPOM, velocity, and can opy cover and negatively related to FPOM. Taxon richness ha d a positive relationship with velocity and CPOM. The proportion of scrapers was negativ ely related to increased canopy cover and positively related to velocity. Filterers were positiv ely related to FPOM and oxygen within the patch. Shredders were not predicted by any environmental variable. Collector-gatherers were negatively related to FPOM and positively related to canopy cover (Table 4-3). Discussion Stream invertebrate communities are structur ed b y a mosaic of habitats ranging from macrophytes and substrate diversity to small-s cale changes in flow patterns. Community composition is related to the quantity, quality, and dist ribution of detritus on the streambed in headwater streams (Arsuffi and Suberkropp, 1985; Murphy et al., 1998), and plays a significant role in the distribution, species composition, and total biomass of benthic invertebrates (Hearnden and Pearson, 1991; Reice 1974). Thus, patch size and quality are two key factors affecting colonization patterns of patches. In the current study, invertebrate density and

PAGE 116

116 community structure were determined by complex interactions among patch size, type, quality, and abiotic variables. Patch Complexity Patches with greater structural complexity generally support more species and higher abundances as potential niches increase (Dean and Connell, 1987; Douglas and Lake, 1994; Downes et al., 1998, Downes et al., 2000). In the study stream s, Ludwigia typically fills the entire water column, providing habitat for benthic species, swimmers, and clingers, while leaf packs rest on the surface of the st reambed. Submerged macrophytes increase the physical complexity of aquatic environments, pr oviding habitat for colonisation by invertebrates (Heck and Westone, 1977; Crowder and Cooper, 1982; Gregg and Rose, 1982; Tokeshi and Pinder, 1985; Lodge, 1991; Newman, 1991). A dditionally, macrophyte architecture has a influences food supply through detritus trappi ng (Rooke, 1984) and growth of epiphytic algae (Dudley, 1988), leading in some cases to di stinct invertebrate communities on different macrophytes (Minshall, 1984; Rooke, 1986). As a result, macrophytes in the current study supported higher densities and taxon richness on a per volume basis than did leaf packs. However, in the short-term experimental study, they supported the lowest invertebrate density. Macrophyte leaves are not consumed by invertebrate s, but the epilithon and biofilm matrix is in most cases (Newman, 1991). Additionally, sinc e macrophytes are growin g within the stream, they may exude less nutrients than decomposing allocthonous leaf litter. Thus, macrophytes may need more time than terrestrially derived leav es both to attract invertebrates and be to conditioned with suitable biofilm, as seen in the current study. Structural complexity may also enhance resour ces available within habitat patches (Diehl and Kornijow, 1998). FPOM trapped within patche s provided the basis for higher invertebrate abundance and taxon richness in the observationa l study. Higher am ounts of FPOM provide

PAGE 117

117 more surface area for bacteria and fungi, thus providing more food for invertebrates. Additionally, since FPOM is easily flushed from habitats during stor m events, higher FPOM may indicate greater stability of the patch, providing more reliabl e habitat for invertebrates. Ludwigia patches trapped the most CPOM and FPOM in the short-term experimental study. This ultimately increased diversity of niches av ailable to invertebrates and improved suitability for colonization. Since macrophytes are anchored in sediment, they may act like debris dams, trapping and holding organic matter during storm events. Thus, m acrophytes have the potential to take over some of the func tion of woody debris typically abse nt in logged streams. Although macrophytes became abundant following logging, inputs of pine needles will likely increase over the next decade since the waters hed was planted with a monoculture of pine. As expected, pine patches created the least heterogeneity and trap ped little if any organi c matter. Many timber operations in the southern U.S. utilize pine pl antations, which could have a negative impact on invertebrates by decreasing structural complexity and overall storage of organic matter. Patch Stability In addition to structural complexity, habitat st ability p lays a large role in determining the composition of patch inhabitants. Although the sout hern coastal plain does not typically receive high-energy flows such as those present in snow-m elt, relatively large events may occur during hurricanes and smaller events with storm even ts common during summer. Thus, more stable habitats are likley to be more attractive to invertebrates. Stability provided by Ludwigia enhanced colonization by filtering invertebrate s. Additionally, sandy-bottomed streams in the coastal plain do not provide relatively immobile substrates such as cobble and boulders present in the piedmont. Thus, invertebrates depend on ava ilability of organic subs trate introduced from the riparian zone or growing within the st ream including woody debris, rootwads, macrophytes, and leaf packs. However, leaf packs are ephe meral, rapidly decomposing, and are subject to

PAGE 118

118 being scoured from the streambed during storms. In addition, Ludwigia patches provide multiple food sources for filterers, such as Simuliidae, by allowing them access to the water column and by trapping large amounts of organic matter. Thus, Ludwigia can sustain filterers even at low flows, when only small amounts of FPOM and bact eria are present in the water column. Hydrologic disturbance can create a mosaic of stable and unstable patches within stream s. Olsen et al.(2007) found that invertebra te densities were greate st in stable patches following an experimental simulation of flooding in streams. However, this difference only existed for 14 days following the disturbance. In the current study, a small scale disturbance had little impact on colonization patt erns of invertebrates. The ex pectation was that disturbed samples would be more similar to the sevenday samples than the undisturbed fifteen-day samples. The disturbed samples appeared to rese mble the fifteen-day samples in most cases and were even higher than the undisturbed in some cases. This may be linked to the size of the disturbance and presence of source populations nearby. Coloni zation is a rapid process in streams, and most areas recover in 10-30 days following localized disturbances (Mackay, 1992). Melo and Froelich (2001) found that invertebrates recolonized ove rturned stones within 4 days, and densities became higher than those on control stones within a month. Thus, 7 days between sampling may have been too long to see any differences. Although not significant, total abundance and mass-weighted abundance were hi gher in disturbed than in seven day or undisturbed samples. This may be linked to higher amounts of FPOM trapped in disturbed samples, and an increase in collector-gatherers. Additionally, small scale di sturbances that leave biofilm intact are less likely to have long la sting impacts on habitat occupancy (Miyake, 2003). Patch Quality The quality of leaves as food affects the perform ance (i.e. gr owth rates and densities) of benthic macroinvertebrates (Cummins and Klug, 1979; Sweeney and Vannote, 1986) and is

PAGE 119

119 determined by the leaf composition and attached biofilms (Lock et al., 1984; Hax and Golladay, 1993), which consist of autotrophic and heterotr ophic components. In this study, patch quality based on biofilm composition was in fluenced by temporal changes in environmental parameters. Bacterial biomass was highest in autumn, while chlorophyll a was highest in spring. This reflects changes in canopy cover typical in head water streams since light is a primary factor limiting primary production (Hill and Knight, 1988; Hepinstall and Fuller, 1994; Hill et al., 1995) and consequently influences the develo pment and biomass of biofilms (Ledger and Hildrew, 1998). Higher bacterial biomass is likel y linked to the greater surface area provided by decaying organic matter derived from the riparian zone, as well as decaying algae and macrophytes present in the spring and summer samp les. Additionally, higher bacterial biomass in autumn may fuel algal growth in spring. Severa l studies have indicated the existence of a link between algae and bacteria (Rounick and Wi nterbourn, 1983; Hepinstall and Fuller, 1994; Ledger and Hildrew, 1998) whereby bacteria benefit from algal exudates for an energy source, or as a substratum for colonisation (Rier and Stevenson, 2002). Historically the primary energy source in headw ater streams was thought to be terrestrially-dervived leaf litter a nd the bacteria and fungi associ ated with it (e.g., Vannote et al., 1980). However, more current research found suffic ient algal growth even in streams with high canopy cover (Mayer and Likens, 1987). Though epiphyton is typically associated with macrophytes, the present study found similar ch lorophyll a for leaf packs and macrophytes, except during periods of maximum irradiance (e .g., spring). Thus, both habitats have the potential to support diverse macroinvertebrate communities. However, bacterial biomass was higher on leaf packs, suggesting these are a high er quality food source. This was not supported

PAGE 120

120 by the data since abundance and taxon richness of invertebrates were higher in macrophyte patches. In the observational study, filterers were be st p redicted by chlorophyll a in macrophytes and bacteria in leaf packs. This supports recent findings that many invertebrates exhibit plasticity when selecting resources (Friberg a nd Jacobsen, 1994). In addition to organic matter sloughing from epiphyton, the struct ure provided by algae will aid in development of a biofilm matrix. Additionally, filterers were positively related to dissolved oxygen within the patch, which was higher in macrophytes since they ex tend into the water co lumn and release oxygen during photosynthesis. Switching feed ing behavior has also been observed in shredders, mixing algal and detritus based carbon sour ces (Friberg and Jacobsen, 1994). Patch quality is also linked to refractory com pounds in leaves that may alter biofilm structure, decomposition rates, and nutrient availability for colonizing species (Ostrofsky, 1993, 1997). This may be especially true for shredding invertebrates that depend on biofilm as well as leaf properties (e.g., Lignin conten t). Habitat selection by shredde rs was apparent in the shortterm experiment in relation to leaf palatabilit y. After seven days, shredders were more common in Pinus and Liriodendron than in the less palatable Quercus and Ludwigia However, shredders became similar among all treatments after fifteen days and were similar between macrophytes and leaf packs in the observational study. This in dicates that although shre dders initially select more suitable habitat, accumulation of organic matter in other patches creates adequate habitat for this group. Bastian (2007) f ound that shredders were distributed across a broad range of leaf species in a stream, with no leaf species being preferentially colonized by shredders. However, most studies find that shredder species exhibit clear leaf preferences (Anderson and Sedell, 1979; Mackay and Kalff, 1973; Nolen and Pearson, 1993), and selectively feed on food resources of

PAGE 121

121 different palatability or quality (Arsuffi a nd Suberkropp, 1985, Campbell and Fuchshuber, 1995). Although shredders are implicated in breakdown of organic matter in streams, they usually colonize leaf packs later than other feeding groups Shredders usually select leaves at advanced stages of conditioning (Arsuffi and Suberkr opp 1984, 1985; Mackay and Kalff, 1973, Petersen and Cummins, 1974) due to increased microbial biomass and fungal de gradative enzymes and, thus, increased leaf palatabi lity (Suberkropp, 1998). Thus, Quercus leaves may not be colonized as fast due to their refractory properties, but st ill may provide more than adequate habitat. In addition, a case-making caddisfly, Anisocentropus, was commonly found in cases made from Quercus This is likely due to its resistance to br eakdown, to provide long term protection. Patch Size Increased patch size potentially creates more niches, providing a diversity of resources and refugia from predators. Although the amount of mass lost from patches increased with patch size, breakdown rates were similar when compari ng initial masses. Contrary to my hypothesis, this indicates that conditions inside larger leaf packs do not necessarily become less suitable for decomposition and provide equal opportunity for biofilm formation. In the observational study, bacterial biomass increased with patch size for Ludwigia but not for leaf packs. However, invertebrates did not respond positivel y to this increased re source base and niche availability. In the observational study, the expected species richness was not re lated to patch size. This suggests a lack of differences in resources with larg er patches. Patch size was a determinant of feeding guild st ructure. Scrapers did not select habitat based on leaf type and were m ost abundant in larg er patches. Most scra pers are classified as clingers and thus need habitat that will support their mass and provide a substantial food. Larger patches should support more mass and protect th is group from moderate flow events. In addition, many grazing invertebrates quickly deplete their resour ces (McAuliffe, 1984), thus

PAGE 122

122 emphasizing the need for great surface areas to support a significant community. In the observational study, scrapers increased with patc h size, bacterial abundan ce, and chlorophyll a, but only in Ludwigi a. Higher bacterial abundance may provide additional resources for scrapers, since many invertebrates have flexible feeding ha bits. However, in the observational study, there was no link between patch size and sc raper abundance in leaf packs. Many of the larger leaf patches in the observational study were multi-ti ered, and much of the surface area was not exposed to light, thus lim iting primary productivity. In the observational study, multiple conf ounding factors lim ited interpretation of relationships between patch size and invertebrate communities. Leaf packs were diverse, multi species assemblages in varying stages of decompos ition. To control for this, the field experiment used freshly abcissed leaves and only created single species patches. Liriodendron and Pinus patches decomposed more rapidly than Quercus Liriodendron leaves are soft and pliable with lower C:N ratios than Quercus. However, pine needles ha d much higher C:N ratios, but provided a larger exposed surface area for bact eria colonization. Nitrogen content, C:N ratio, total phenolics, percentage lignin and lignin:N ra tio explain much of the variability in leaf processing rates (Taylor, Parkins on and Parsons, 1989; Ostrofsky, 1997). Patch occupancy may be related to interactio ns between biotic a nd abiotic factors. Collecto r gatherers are bottom-feeders in streams and tend to consume any type of small organic particle. They are also the most abundant group in sandy-bottomed streams (Smock et. al, 1985). However, this group was more abundant in leaf packs than in macrophytes, increasing with patch size in macrophytes, but decreasing in leaf p acks. An opposite relationship was found for predatory invertebrates, sugges ting predation on this group. This is supported by the higher abundance of predators in macrophytes even thou gh trapped organic matter was similar between

PAGE 123

123 the patch types for the observational study. Although streams are thought to be primarily structured by abiotic factors, biotic factors are likely to influence community structure effectively at smaller scales (e.g., Peckarsky, 1983). The proportion of shredders was negatively relate d to bacterial abundan ce for leaf packs. This potentially suggests that com petitive interactions may exist between these groups, since both consume leaf organic carbon. This may ex plain the lack of a relationship between shredders and bacteria in macrophytes. Inte ractions between shredders, organic matter decomposition and microbes (bacteria and fungi) are complex. For example, fungi and bacteria convert a portion of detrital organic matter into microbial bi omass, transforming the detrital substrate into a more nutritious food source fo r detritus feeders (Barlocher and Kendrick, 1975; Suberkropp, 1992). At the same time, shredder fr agmentation of the detrital matrix promotes microbial activity, increasing av ailable detrital surface for colo nisation (Hargrave, 1970; Howe and Suberkropp, 1994) and spreading microfungal spores (Rossi, 1985). The results of this study s upport the expected response of invertebrates to changes in habitat type and quality as l ogging reduces leaf packs and in creases primary productivity. Typically, headwater stream s would lose shredders and gain more scrapers. However, in warm temperate coastal plain systems, collector gath erers may be the dominant consumer of organic matter. This may explain the decrease in collec tor-gatherers with the subsequent increase in scrapers. In addition, scrapers were negatively related to canopy cover, while collector-gatherers were positively related. This indi cates that collectors may be the more natural feeding group in occuring in undisturbed, sandy-botto med streams. Since much of the substrate is highly mobile, scouring of leaf packs may act as the initial d ecomposition mechanism, creating smaller particles available for collectors, work usually done by shredders.

PAGE 124

124 Table 4-1. Multiple regressions for leaf packs averaged over all time periods for the observational study. Dependent VariableParameterEstimate SEt P Size0.280.230.050.96 Chlorophyll a-17.750.4-0.350.72 Bacteria Abundance0.080.140.570.57 Bacteria Biomass0.330.13.2 0.003 FPOM0.380.132.9 0.005 Size0.130.140.90.37 Chlorophyll a-26.830.4-0.880.38 Bacteria Abundance0.050.080.560.58 Bacteria Biomass0.120.061.90.06 FPOM0.170.082.2 0.03 Size0.860.352.5 0.02 Chlorophyll a-134.676.2-1.770.08 Bacteria Abundance-0.520.21-2.51 0.01 Bacteria Biomass-0.130.16-0.80.13 FPOM-0.170.2-0.870.39 Size-0.290.27-1.10.28 Chlorophyll a72.258.81.230.22 Bacteria Abundance0.380.162.3 0.02 Bacteria Biomass0.240.122 0.04 FPOM0.460.153 0.005 Invertebrate Abundance (F4,59=15.0, P < 0.0001, R2 = 0.58) Taxon Richness (F4,59=7.7, P < 0.0001, R2 = 0.42) Shredders (F4,59=2.0, P = 0.08, R2 = 0.16) Filterers (F4,59=8.9, P < 0.0001, R2 = 0.45)

PAGE 125

125 Table 4-2. Multiple regressions for Ludwigia av eraged over all time peri ods for the observational study. Dependent VariableParameter Estimate SEt P Size 0.150.170.90.37 Chlorophyll a 20.613.91.60.12 Bacteria Abundance0.120.150.170.44 Bacteria Biomass-0.040.13-0.280.78 FPOM 0.280.12.8 0.007 Size 0.110.081.50.14 Chlorophyll a 9.75.81.70.1 Bacteria Abundance-0.020.07-0.280.78 Bacteria Biomass-0.040.06-0.670.51 FPOM 0.080.041.90.06 Size 0.620.252.5 0.01 Chlorophyll a -42.719.3-2.2 0.03 Bacteria Abundance-0.560.22-2.5 0.01 Bacteria Biomass0.080.20.390.7 FPOM -0.210.15-1.40.16 Size -0.110.21-0.520.61 Chlorophyll a 31.916.61.90.06 Bacteria Abundance0.350.191.80.07 Bacteria Biomass-0.170.17-0.990.33 FPOM 0.290.122.3 0.02 Size 0.250.211.220.23 Chlorophyll a -4615.9-2.9 0.005 Bacteria Abundance-0.290.18-1.570.11 Bacteria Biomass0.250.161.550.13 FPOM -0.290.12-2.4 0.02 Filterers (F4,63=2.9, P = 0.02, R2 = 0.20) Collector-gatherers (F4,63=4.5, P = 0.002, R2 = 0.28) Invertebrate Abundance (F4,61=3.5, P = 0.008, R2 = 0.24) Taxon Richness (F4,63=2.7, P = 0.03, R2 = 0.19) Scrapers (F4,63=2.6, P = 0.04, R2 = 0.20)

PAGE 126

126 Table 4-3. Multiple regressions for the field expe rim ent averaged over all treatments for each invertebrate metric. Dependent VariableParameterEstimate SEt P CPOM0.450.153.6 0.0004 FPOM-0.990.5-2 0.04 Velocity5.91.523.9 0.0001 Canopy Cover0.670.116.1 <0.0001 Oxygen-0.520.35-1.50.14 CPOM0.480.192.5 0.01 FPOM-0.280.63-0.450.65 Velocity8.61.924.45 <0.0001 Canopy Cover0.120.140.840.4 Oxygen-0.070.44-0.150.88 CPOM-0.030.24-0.140.89 FPOM0.290.780.370.71 Velocity5.022.412.08 0.04 Canopy Cover-0.870.17-5.05 <0.0001 Oxygen-0.250.55-0.450.65 CPOM0.320.191.650.1 FPOM2.220.623.57 0.0004 Velocity-0.391.91-0.210.84 Canopy Cover-0.090.14-0.680.5 Oxygen10.432.3 0.02 CPOM0.010.10.120.9 FPOM-1.180.33-3.58 0.0004 Velocity-0.551.02-0.550.59 Canopy Cover0.460.076.3 <0.0001 Oxygen-0.260.23-1.120.26 Filterers (F5,255=6.9, P < 0.0001, R2 = 0.12) Collector-gatherers (F5,255=11.0, P < 0.0001, R2 = 0.18) Invertebrate Abundance (F5,255=15.0, P < 0.001, R2 = 0.23) Taxon Richness (F5,255=6.2, P < 0.0001, R2 = 0.11) Scrapers (F5,255=6.1, P < 0.0001, R2 = 0.11)

PAGE 127

127 0 0.002 0.004 0.006 0.008 0.01 0.012 November 2005January 2006April 2006June 2006Chla (mg) Leaf Packs Ludwigia Figure 4-1. Total biomass of chlorophyll a (m g) ( SE) in each patch type.

PAGE 128

128 0 10 20 30 40 50 60 70 80 November 2005January 2006April 2006June 2006Total Bacteria Cells (1 X 106) Leaf Packs Ludwigia Figure 4-2. Total number of bacterial cells (1 X 106) ( SE) in each patch type.

PAGE 129

129 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 November 2005January 2006April 2006June 2006Bacterial Biomass (pg C/cm3) Leaf Packs Ludwigia Figure 4-3. Bacterial biomass (pg C/cm3) ( SE) in each patch type.

PAGE 130

130 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 November 2005January 2006April 2006June 2006Number Bacterial Cells/cm3 (1 X 106) Leaf Packs Ludwigia Figure 4-4. Number of bacte rial cells per cm3 (1 X 106) ( SE) in each patch type.

PAGE 131

131 0 0.00005 0.0001 0.00015 0.0002 0.00025 0.0003 0.00035 0.0004 November 2005 January 2006April 2006June 2006Chl a (mg/cm3) Leaf Packs Ludwigia Figure 4-5. Chlorophyll a biomass (mg/cm3) ( SE) in each patch type.

PAGE 132

132 0 0.5 1 1.5 2 2.5 3 3.5 4 November 2005January 2006April 2006June 2006Taxon Richness/cm3 Leaf Packs Ludwigia Figure 4-6. Volume-weighted taxon richness (Taxa/cm3) ( SE) in each patch type.

PAGE 133

133 0 2 4 6 8 10 12 14 November 2005January 2006April 2006June 2006Invertebrate Density (Individuals/cm3) Leaf Packs Ludwigia Figure 4-7. Volume weighted inve rtebrate density (Individuals/cm3) ( SE) in each patch type.

PAGE 134

134 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 November 2005January 2006April 2006June 2006Proportion Filterers Leaf Packs Ludwigia Figure 4-8. Proportion of filtering invert eb rates ( SE) in each patch type.

PAGE 135

135 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45Pinus Liriodendron QuercusProportion Leaf Mass Loss Day 7 Day 15 Disturbed Day 15 Undisturbed Figure 4-9. Proportion of leaf mass decomposed ( SE) in relation to patch type and disturbance.

PAGE 136

136 0 0.2 0.4 0.6 0.8 1 1.2124MassMass Lost (g) Figure 4-10. Amount of leaf mass decomposed (g ) ( SE) in relation to initial patch mass.

PAGE 137

137 0 0.2 0.4 0.6 0.8 1 1.2124MassCPOM in patches (g) Figure 4-11. CPOM trapped in patches ( SE) in relation to patch size.

PAGE 138

138 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8PinusLiriodendronLudwigiaQuercusCPOM in patches (g) Figure 4-12. Average amount of co arse particulate organic m atter (g) ( SE) trapped in each patch type.

PAGE 139

139 0 0.2 0.4 0.6 0.8 1 1.2Day 7 Day 15 DisturbedDay 15 UndisturbedCPOM in patches (g) Figure 4-13. Average amount of co arse particulate organic m atter (g) ( SE) trapped in patches by disturbance type.

PAGE 140

140 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2Day 7 Day 15 DisturbedDay 15 UndisturbedFPOM in patch (g) Figure 4-14. Average amount of fine particulate or ganic m atter (g) ( SE) trapped in each patch based on disturbance.

PAGE 141

141 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35PinusLiriodendronLudwigiaQuercusFPOM in patch (g) Figure 4-15. Average amount of fine particulate or ganic m atter (g) ( SE) trapped in each patch type.

PAGE 142

142 0 5 10 15 20 25 30 35 40 45PinusLiriodendronLudwigiaQuercusInvertebrate Abundance Figure 4-16. Average number of invertebrate in dividuals ( SE) in each patch type.

PAGE 143

143 0 5 10 15 20 25 30 35 40 45124MassInvertebrate Abundance Figure 4-17. Average number of invertebrate in dividuals ( SE) in each patch bas ed on initial patch mass.

PAGE 144

144 0 1 2 3 4 5 6124MassTaxon Richness Figure 4-18. Average number of taxa ( SE) in each patch in relation to initial pa tch mass.

PAGE 145

145 0 0.05 0.1 0.15 0.2 0.25 0.3124MassProportion Scrapers Figure 4-19. Proportion of scra pers ( SE) in each patch based on initial patch m ass.

PAGE 146

146 0 0.005 0.01 0.015 0.02 0.025 0.03124MassProportion Shredders Figure 4-20. Proportion of shredders ( SE) in each patch based on initial patch m ass.

PAGE 147

147 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07PinusLiriodendronLudwigiaQuercusProportion Shredders Day 7 Day 15 Disturbed Day 15 Undisturbed Figure 4-21. Proportion of shredders ( SE) in each patch based on patch type and disturbance.

PAGE 148

148 0 0.02 0.04 0.06 0.08 0.1 0.12124MassProportion Filterers Figue 4-22. Proportion of filterers ( SE) in each patch bas ed on initial patch mass.

PAGE 149

149 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18PinusLiriodendronLudwigiaQuercusProportion Filterers Figure 4-23. Proportion of filterers ( SE) in each patch type.

PAGE 150

150 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7PinusLiriodendronLudwigiaQuercusMassProportion Collector-Gatherers Figure 4-24. Proportion of collector-gat herers ( SE) in each patch typ e.

PAGE 151

151 CHAPTER 5 HABITAT SELECTION IN FRAGMENTED LANDSCAPES: COMPARING GENERALISTS TO SPECIALISTS Introduction Dispersal of individuals rela tive to patch variability has im portant implications for ecological processes, including population disp ersal and redistribution, local population and metapopulation dynamics, and intensity of species interactions (Hanks i, 1998; Gilliam and Fraser 2001). Habitat patch arrangement, amount and perimeter:area ra tio (e.g., edge) strongly influence both community structure and interpatch dispersal both in terrestrial and aquatic systems (Wiens, 1997; Pither and Taylor, 1998; Hanski, 1999; McIntyre and Wiens, 1999; Jonsen and Taylor, 2000; Palmer, 2000). The impor tance of patches, and especially isolation between patches, depends on dispersal ability of focal organism(s) (Kareiva andWennergren, 1995). Since invertebrates respond to patches at a sm aller scale than other groups, they are an ideal m odel group for testing hypotheses related to habitat fragmentati on (Bowne and Bowers, 2004). For instance, insects residing in patchy hab itats display higher dens ity when resources are dispersed among many small patches as opposed to large, aggregated patches (Hanski, 1994; Remer, 1998; Roitberg, 1997; Heard, 1998; Silver, 2004a). In this way, habitat connectivity increases with increased habitat fragme ntation (Tischendorf and Fahrig, 2000). Both natural and anthropogenic forces leadi ng to habitat loss and fragm entation have been considered in debates over the relative import ance of habitat amount versus arrangement in determining community response to habitat frag mentation (Sih, 2000). Flather (2002) argued that habitat amount is a more plausible explanation for population size, but arrangement becomes important when total habitat cover declines to ~ 30-50 %, emphasizing the need to study processes over a range of total cover. Additionally, patches may be highly dynamic, changing in

PAGE 152

152 shape and size over time (Pickett and Thomps on, 1978) in response to disturbance and succession. Movement is a primary factor determini ng the effect of spatial heterogeneity on ecological processes (Diffendorfer et al., 2000). Movem ent and dispersal provide escape from competitors, predators and parasites. Risks associ ated with dispersal include increased mortality due to predation and an inability to find suitable sites (Bilton et al., 2001). Interactions between dispersal and landscape structure determine the ab ility of an organism to move through the landscape (Merriam, 1984). Thus, the colonizat ion rate of new patches by individuals is influenced by emigration rate, mean dispersal distance relative to patch distance, mortality incurred during dispersal, and mean number of po tential dispersers (Johnst, 2002). The idea that animal movement and dispersal occur as a result of behavioral choices made in response to environmental heterogeneity across spatial and temporal scales emphasizes the importance of linking behavioral and landscape ecology (Lima and Zollner, 1996). Studies of movement across heterogenous lands capes are necessary to determ ine impacts of habitat loss for management and conservation (O lden et al., 2004). Movement patterns in the landscape are central to connect ivity, patch and boundary dynamics, spread of disturbances, source-sink and metapopulation dynamics (Ims, 1995) Dispersal is cons idered active when attributed to behavioral decisions and passive wh en due to displacement. Biased flow in streams emphasizes the importance of both passive and ac tive dispersal. Movement of an individual through a heterogeneous landscape is influenced by a number of abiotic (flow, temperature, light levels) and biotic cues (food, predation risk). Movement between patches also depends on the proportion of different habitat type s, as well as spatial configur ation of the landscape (Moilanen and Hanski, 1998). Connectivity within a landsc ape depends on the spatial configuration of

PAGE 153

153 patches and movement patterns of the organism. Thus, connec tivity based on movement links behavior and landscape structure (Goodwin and Fahrig, 2002). Movement of invertebrates in streams is in fluenced by available ha bitat amount and type (e.g., Palm er, 2000). In many cases, leaf packs and macrophytes share similar macroinvertebrates, including Ephemeroptera, Chironomidae, Trichoptera, and Simuliidae (Velasquez, 2003), making them useful for studying movement across different patch types. However, they differ in the composition of other invertebrate fauna, sugge sting that these patch types are somewhat unique. Interpatch movement and ability to find patches of suitable quality are key factors influencing species persistence. The ability of dispersing invertebrates to find and settle in patches may be influenced by patch quality (Palmer, 1996) and physical arrangement of patches on the stream bed (Silver, 2000). The goal of this study was to de termine implications of changing habitat availability and type on invertebrate movement, focusing specifically on how patch type and amount affect habitat selection. I hypothesized that reduction of available patc hes and increased isolation will negatively affect the ability of ha bitat specialists to locate patches. Additionally habitat specialists should be more efficient at finding th e next patch and will follow a relatively straight path to it, while habitat generalis ts will choose a random, tortuous path. Materials and Methods Study Organisms The habitat specialist, A nisocentropus pyraloides (Trichoptera: Calamoceratidae) is a slow-moving detritivore that inhabits small st reams flowing through deciduous forest throughout the eastern U.S. (Wiggins, 1996). Larvae constr uct notched, oblong cases made of two leaves sealed together with silk. A dditionally, its diet consists prim arily of organic matter and thus depends on leaf packs for food and its case. This species is semivoltine and emerges in the spring

PAGE 154

154 (Wiggins, 1996). Although dorso-ventrally flat tened, the case likely creates drag while the organism is crawling over the streambed. The habitat generalist used in experim ents was the snail, Elimia sp.(Gastropoda: Pleuroceridae), which commonly o ccurs in all landscape units in cluding macrophytes, leaf packs, and the sandy matrix (persona l observation, Chapter 4). This genus occurs throughout the southern U.S. and is abundant in the study streams, with dens ities as high as 25 individuals/m2. Elimia produces more than one generation per year and is parthenogenic, wh ich contributes to its abundance in the streams (Viera et al., 2006). Elimia is conical, limiting its drag as it moves across the streambed. Behavioral Observations The effects of habitat type and amount on m ovement were examined using short-term behavioral exeriments within the stream. During the experiment, conditions were consistent with average conditions thoughout the stream. Over the seven day pe riod, average water temperature was 18.5 C, velocity was 0.13 cm/s, and canopy cover was 79 %. Leaf packs ( Liriodendron tulipifera ) and macrophytes (Ludwigia repens ) were collected as described in Chapter 4. Habitat mosaics were created in a 5 m2 section of the channel in watershed C by adding leaf packs:macrophytes at 1:0, 1:1, or 0:1 ratios, with total percent cover of 10, 30 and 50 percent of the entire landscape (Fig. 1). Macrophytes and leaf packs were arranged randomly since effects of patch configuration were not being addressed. Each leaf pack or macrophyte patch had a surface area of 16 cm2. The sediment in the selected r each was raked to a depth of 0.75 m to remove organic matter and invertebrates, then was smoothed to create a landscape completely dominated by sand. A 3 X 3 grid composed of co lored nylon was attached to pvc pipes at the perimeter of the landscape and was placed 15 cm above the water as a reference point for

PAGE 155

155 movement distances. A drift net was placed at the end of the landscape to trap emigrating individuals. Experimental organisms were collected from the stream each morning from the streambed and naturally occuring leaf packs. Individuals were placed in separate flow-through trays and allowed to acclimate to the stream reach for at least an hour. A video camera was set up on a tripod to record movement of individuals. During each trial, individuals were placed at the center of the landscape, facing downstream. Behavior was recorded for 30 minutes, with the observer leaving the reach while trials were occu ring. After 30 minutes, th e length and width of each individual was measured, removed from the landscape, and released downstream. On average, Anisocentropus individuals were 4.2 cm long ( 0.2 SE) and 2 cm wide ( 0.1 SE), while Elimia individuals were 4.3 cm long ( 0.1 SE) a nd 1.7 cm wide ( 0.1 SE). No individual was used more than once, and trials were repe ated for at least four individuals (more for Elimia due to availability). After each trial, the streambed was gently scoured to remove any traces of the individual path. Trials were run between 7 AM and 4 PM daily for a period of seven days beginning 7 March, 2007. The grid was left in place each night, and pvc pipes were inserted into the streambed as placeholders for the tripod. Videos were digitized and manually analyzed on a com puter screen with coordinates (x,y) recorded every 10 s to determine movement parameters. For each path, total path length, correlations between turning angles, mean cosine of turning angle, mean path length, and net squared displacement were calculated to assess distance covered (Turchin et al.1991). The above parameters were used for correlated ra ndom walk models, which are useful for making inter-specific comparisons (Kareiva and Shigesad a 1983; Cain 1985; Crist et al.1992). Each path

PAGE 156

156 was compared to the correlated random wa lk model of Nams and Bourgeois (2004) by calculating Rdiff: k n n n n diffRE RER k R1 2 2 2)( )( 1 where E(R2 n) is the expected net squared displa cement (Kareiva and Shigesada 1983), n is the number of moves, and R2 n is the mean net squared displacement. Positive values of Rdiff indicate that the path is l onger than predicted by correlated random walk models and negative values indicate shorter paths. An individuals overall rate of movement across a landscape is contingent upon its tendency to move (or rem ain sedentary), moveme nt velocity, and path tortuosity (Russell et al.2003). Tortuosity of movement was assessed by calculating the fractal dimension (D) of each movement path, whereby estimates near 1 indica te highly linear movement and near 2 suggest approximate Brownian (plane-filling) moveme nt (Hastings and Sugihara 1993). Fractal dimensions were estimated with Fractal 4.0 software (http://www.nsac.ns.ca/envsci/staff/vnams/Fractal.htm). The fractal mean method was used, which is based on the traditional dividers method (Mandelbrot 1967, Sugihara and May 1990), but corrects for estimation errors created when th e last divider step does not fall exactly on the end of the path (Nams and Bourgeois 2004). Frac tal dimensions were estimated based on the entire recorded movement path of each individual. Paths of four moves or less were not used in the analyses because estimates of their fracta l dimension sometimes fell below the theoretical limit of 1. To test whether the above movement be haviors differed am o ng habitats, ANOVA was used after normalizing the data. Where signif icant effects were observed, differences among

PAGE 157

157 treatmentfactor combinations were tested us ing post hoc Tukeys honest significant difference (HSD) tests ( = 0.05). Colonization Habitat selection based on patch amount a nd type was exam ined in a short term colonization study. Microlandscapes were created along an ~75 m stretch of stream, separated by at least 3 m. Landscapes were the same as those used in the short-term behavioral experiment, but were half the size (45.7 cm W X 50.8 cm L), and were replicated three times in a randomized block using each replicate as a bloc k. Prior to creation of the landscape, the streambed was raked to 0.5 m to remove any apparent organic matter or habitat and allowed to settle for four hours. Drift nets were placed at the end of the landscape to trap emigrating invertebrates. Macrophyte and l eaf patches were anchored to th e sediment in the appropriate configuration (Fig. 1). Invertebrates for the experiment were collect ed f rom the streambed and, leaf packs and lengths of individuals were meas ured. Due to low abundance of Anisocentropus, only one individual was used for each rep licate, however, six individuals of Elimia were used. The shell or case of the individual was bl otted dry and marked with a dr op of paint and the number of landscape (from 1 to 27). Individu als were released at the center of the landscape after the paint dried (~ 5 minutes). After 24 hours, all patche s were collected and placed in individually labelled bags. Velocity was measured at the upstream and downstream end of the landscape with a Flomate 2000 (Marsh McBirney). In addition, drift nets were collected and any marked individuals in the matrix (sa nd) were collected. A surber sa mple was also taken from the landscape to determine recolonizat ion by other inve rtebrates.

PAGE 158

158 Results Movement Anisocentropus Mean step length did not diffe r for any of the m icrolandscapes, averaging 0.7 cm ( 0.3 SE) per step. Deviation from the correlated ra ndom walk between the leaf species depended on amount of patch cover (F4,37 = 3.1, P = 0.03) (Fig. 2). Paths became more random (closer to the CRW) with increased cover for single species landscapes, but were shorter than a CRW for mixed landscapes. The probability of turning in the same direction differed by leaf type, but depended on total amount of cover (F4,37 = 4.6, P = 0.004). This parameter increased with increasing cover in Ludwigia dominated landscapes, but decreased in mixed landscapes (Fig. 3). Correlation between adjacent angles differed by leaf type, but depended on total amount of cover (F4,37 = 2.6, P = 0.04). In general, correlations betw een angles were negative, but became more negative with increasing cover in mixed landscape s (Fig. 4). Net square d displacement differed by leaf type, but depended on total amount of cover (F4,37 = 4.5, P = 0.004). Displacement increased with increasing cover in Liriodendron dominated landscapes, but decreased in mixed landscapes (Fig. 5). Mean D did not differ between leaf species or percent of habitat cover, averaging 1.15 ( 0.02 SE). Elimia Changes in mean step length for leaf species depended on percent c over in the landscape (F4,39 = 3.7, P = 0.01). Step length incr eased with increasing cover in Liriodendron landscapes from 0.1 cm to 0.7 cm per step. It was higher in Ludwigia landscapes with 30 % cover, increasing from 0.5 to 0.8 cm (Fig. 6). Deviati on from CRW differed between leaf types (F2,38 = 4.4, P = 0.02) and total amount of cover (F2,38 = 4.8, P = 0.01). In general, paths were greater than expected by CRW at 20 % cove r, with the lowest values in Liriodendron (Fig. 7). The

PAGE 159

159 mean cosine differed between the percent cove r treatments, but depended on leaf type (F4,39 = 3.4, P = 0.01). Probability of turning in the same direction did not differ between treatments and ranged from 0.25 to 0.45. Correlation between adjacen t angles did not differ between treatments, and ranged between 0.5 and -0.5. Net squared di splacement did not differ between any of the landscapes and ranged from 50 to 780 cm. Mean D did not differ between leaf species or percent of habitat cover and averaged 1.03 ( 0.004 SE). Colonization Neither effects of patch type or amount si gnificantly influenced the probability of Elimia or Anisocentropus stay ing in the microlandscape. Howe ver, general trends existed, indicating that the amount of cover affects the likelihood of these species remaining in the landscape. Both invertebrate species were less li kely to leave the landscape as the proportion of cover increased in mixed habitats. The proportion of Elimia leaving the landscape decreased from 90 % to 40 % with increasing cover. The proportion of Anisocentropus leaving the landscap e decreased with increasing cover in all patch types, and no indivi duals left the landscape in the mixed species treatment with 30 % cover. Discussion Habitat fragmentation is typically viewed at a sc ale of kilometers; however, the scale at which fragmentation alters loca l population dynamics likely lies at a much smaller scale, particularly for invertebrates. This is one of a few studies to examine individual movement patterns of aquatic invertebrate s in response to patch struct ure (Olden 2004, Lancaster 2006; Drew and Eggleston, 2006). In logged streams, the amount of habitat may be more important than spatial configurat ion since reduced canopy cover limits overall leaf inputs. In the study streams, invertebrate communities differed greatly between four adjacent streams, suggesting

PAGE 160

160 effects of local filters on invertebrate commun ities. Thus, the quality of the riparian and availability of instream habitat create a filte r to limit presence of certain species. Results from this study support the idea that in stream habitat availability controls small scale community composition. The habitat specialist, Anisocentropus, left landscapes without preferred leaf litter habitat. In streams with lit tle organic matter storage, this species may be driven locally extirpated. Although it may be supported where ripa rian zones are left undisturbed, this species prefers small streams and may be driven out of entire headwater streams if they are logged along their length. Logging limits the amount and quality of hab itat available for aquatic invertebrates in stream s. This is accomplished by reducing leaf fall, as well as through an increase in peak flow with increasing surface runoff (Beasley and Gr anillo, 1982; Williams et al., 1999; McBroom et al., 2002; Grace et al., 2003). Thus, any leaf fall that does reach the stream is easily washed downstream during storm events. The results of this study sugge st that increased habitat cove r decreases em igration rates, regardless of habitat configuration. Very few Anisocentropus individuals were able to colonize patches successfully, but when successful, they remained there for the duration of the trial. This suggests that, although small, patches were able to be used as refugia from flow and exposure. Both Elimia and Anisocentropus were likely to remain in the microlandscapes with 30% cover. Changes in landscape structure, such as reduc tion of the proportion of one or m ore patch types or increased patch isolati on, will alter the ability of organisms to disperse (Merriam 1984; Fahrig and Merriam 1985). Species that can not disperse effectivel y as a result of a change in structure will suffer reductions in regional population sizes (Fah rig and Merriam 1994). As a result, relative abundances of Anisocentropus decreased in treatment watersheds following

PAGE 161

161 harvest. In addition to decrea sed food availability with increa sing isolation, availability of refugia decreases. Landscapes in this study were in a particularly inhos pitable matrix of sand, with little heteroge neity. In addition, Anisocentropus individuals were commonly displaced from the streambed in landscapes with low or no habitat available. Habitat loss may also lead to more time bei ng expended searching for suitable habitats, potentially contributing to lowe red survival rates and decrease d fecundity. Correlation between turning angles was always negative for Anisocentropus leading to a wobbly path. It was a clum sy crawler and appeared to have limite d capacity for crawling over a sand dominated streambed with little structure to cling to. Thus, it is likely this species is washed downstream easily during storm events. However, this incr eased with increasing cover in mixed landscapes, suggesting a search strategy. Elimia took larger steps with increasing am ount of cover in the microlandscape. This may be related to the perceptual range of the or ganism, as it may not perceive patches as habitat when they are farther apart as in the 10 percent c over treatment. This sugg ests that the scale of the study may be larger than the scale of perceive d habitat, but potentially defines this scale as lying between the isolation found in the 10 and 30 percent cover treatments. Perceptual range differs greatly among species (Zollner, 2000), regarding the ability of the spec ies to visualize a three dimensional landscape, and it ultimately determines the individuals movement behavi or, search strategy, and respons e to fragmentation (Lima and Zollner, 1996; With and Crist, 1996). However, perceptual range may be dynamic even for individuals of the same species and it changes with environm ental conditions, such as flow variability in streams. For example, downstream flow bias may increase perceptual distance to an upstream patch, increasing isolati on (Olden, 2004). However, in general, increased habitat

PAGE 162

162 amount lengthens the time spent within the la ndscape and may provide protection from scouring and predators. Although habitat amount was a good predictor of em igration rates and movement rates, habitat heterogeneity also played a role. Anisocentropus remained in landscapes longer when both macrophytes and leaf packs were available, s uggesting that habitat di versity leads to higher abundances. Bronmark (1985) found that freshwater snails were more diverse in ponds with more macrophyte species, reflecting the presence of different niches and refugia from predators. In larger, agricultural landscap es, Jonsen and Fahrig (1997) found that more species and individuals colonized landscapes with higher diversity. This suggests a scale-independent relationship between the probability of colonization and diversity of patches in a landscape. The latter increases the number of pot ential refuges and resources avai lable in a landscape. However, I did not expect this to act at a scale independent of resources. Anisocentropus was more likely to remain within the microlandscape in the presence of Ludwigia This suggests a preference for this habitat, possibly due to in creased three dimensional area a nd protection from flow provided from Ludwigia In streams, even species considered specia lists may display flexibility in feeding preferences. Thus, both species considered to be generalists and specia lists may be able to supplement their diet with alternative food sources, enhancing the actual connectivity of the landscape in contrast to the perceived c onnectivity (Dunning et al., 1992). Additionally, Ludwigia traps organic matter faster than newly form ed leaf packs (Chapter 4), creating a higher quality resource. Thus, macrophytes may provide adequate resoures for dispersing detritivores living in patchy landscapes.

PAGE 163

163 Upstream movement has been proposed as one part of the solution to the drift paradox, whereby species need to recolon ize upstream habitats to account for downstream drift. Displacement along the longitudinal ax is is of particular interest to stream ecologists, partly because of its relevance to concepts such as Mullers colonisation cycle and the paradox of upstream downstream movement (e.g., Muller, 1982; Hershey et al., 1993; Anholt, 1995). In essence, there needs to be a balance between downstream movement ( both passive and active) and upstream migration by larvae a nd adults to maintain position in suitable stream habitats (e.g., Elliott, 1971b; Soderstrom, 1987). In this study, most Elimia individuals (90 %) moved upstream, regardless of landscape type, suggesti ng that this is a compensation mechanism for potential disturbances su ch as floods. However, Anisocentropus did not exhibit a significant displacement direction. Although this species spent much of its time attempting to move upstream, the shape of its case made it susceptible to downstream drift. Field studies of upversus downstream moveme nt of individually m arked invertebrates (i.e. at larger spatial and temporal scales than this study) provide contras ting results with regard to directional movement. Among cased caddisflies Jackson et al.(1999) recorded no directional bias in net displacement at low discharge, but th ere was a downstream bias at higher discharges; Erman (1986) reported some seasonal depe ndence but, generally, a net downstream displacement. However, Hart and Resh (1980) re ported no bias in net displacement direction, but did not report displacement distance along the longitudinal ax is. As in this study, upstream displacement occurs commonly in snails (Schneider and Ly ons, 1993; Huryn and Denny, 1997). Although species capable of upstr eam flight, such as stoneflie s, tend to move downstream (Freilich, 1999).

PAGE 164

164 Clearly, multiple factors can influence displ acement at the s tream scale (e.g., body shape, temperature, discharge, life hist ory stage and food availability), a nd generalizations are difficult. Although discharge remained fairly unifrom during this study, there is strong evidence from other studies suggesting th at discharge is a primary determinan t of movement rates and direction (Olden, 2004; Lancaster, 2006). Thus, respons e to landscape structur e may differ between logged and unlogged streams. Studies attempting to understa nd the role of patch structur e an d arrangement in streams lag far behind those in terrestrial systems. Ho wever, it has become clear that both spatial arrangement and habitat amount are determinan ts of community structure stemming from changes in emigration and immigration rate s (Palmer, 2000; Olden, 2004; Lancaster, 2006; Olden, 2007). Additional studies are needed to de termine if generalities exist in streams, including long-term mark-recaptur e studies across life stages. A quatic invertebrates are unique in that they spend their larval period in the wate r and their adult stages on land. Thus, dispersal studies will need to account for small scale move ments in streams as well as the response of adults to spatial structure in the terrestrial landscape. Clearcut watersheds may limit this dispersal, creating streams that act as isolated islands. This is particularly important in headwater streams since some species are depend ent on specicific environmental conditions only available in small, forested headwater sy stems (Lowe, 2002; Meyer et al., 2007).

PAGE 165

165 A B C A B C Figure 5-1. Microlandscape desi gns used in the behavioral an d colonization experim ents. Liriodendron leaf packs (brown squares) and Ludwigia macrophyte patches at A) 10 B) 20, and C) 30 percent cover. The same c onfiguration was used for landscapes with a single patch type.

PAGE 166

166 -0.8 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 102030Percent CoverDeviation from CRW Ludwigia Mixed Liriodendron Figure 5-2. Average deviation from a correl ated random walk ( SE) (CRW) (Rdiff) for Anisocentropus

PAGE 167

167 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 102030Percent CoverProbability of Turning in Same Direction Ludwigia Mixed Liriodendron Figure 5-3. Average probability ( SE) of each turn being in the sam e direction for Anisocentropus

PAGE 168

168 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 102030Percent CoverCorrelation between Turning Angles Ludwigia Mixed Liriodendron Figure 5-4. Average correlation ( S E) between turning angles for Anisocentropus.

PAGE 169

169 0 500 1000 1500 2000 2500 102030Percent CoverNet Squared Displacement (cm) Ludwigia Mixed Liriodendron Figure 5-5. Average net squa red disp lacement ( SE) of Anisocentropus in microlandscapes.

PAGE 170

170 0 0.2 0.4 0.6 0.8 1 1.2 102030Percent CoverMean Step Length (cm) Ludwigia Mixed Liriodendron Figure 5-6. Mean step length ( SE) in each land scape for Elimia

PAGE 171

171 0 0.5 1 1.5 2 2.5 3 3.5 102030Percent CoverDeviation from CRW Ludwigia Mixed Liriodendron Figure 5-7. Average deviation ( SE) fr om a correlated random walk (Rdiff) for Elimia

PAGE 172

172 CHAPTER 6 CONCLUSIONS Best management practices for forestry in the U. S. clearly de pend on the geographic region under review. For example, coastal plain streams in the southern U.S. are characteristically low-gradient, sandy-bottome d systems with dynamically changing instream habitat. In contrast, those managed for forestry in the western U.S. ar e typically high-gradient montane streams with high habitat and substrate diversity and are susceptible to mass-wasting as vegetation removal reduces bank stability. Although forestry practices in the Northwest can lead to drastic reductions in water quality, evidence fr om the coastal plain indicates limited changes in water quality and biotic diversity in st reams impacted by logging, as long as stream management zones are left intact. Although there were few changes in biotic community structure following logging, this does not discount use of aquatic invertebrates as indicators of water qualit y. Many biotic indices weigh heavily upon the use of EPTs (Ephem eropt era, Plecoptera, and Trichoptera) in their formation (e.g., Lenat, 1993). However, loggi ng ultimately increases pr imary productivity in streams, leading to higher densities of Baetid/Leptophlebiid ephemeropter ans (Chapter 3; Stone and Wallace, 1998). As a result, the FLSCI biotic index is inflated, suggesting an increase in water quality with logging. One short-coming of local management organizations is the longterm fascination with EPTs, sometimes leadi ng to redundant use of this group by utilizing metrics on the number of EPT taxa, % EPT, number of Trichoptera, and number of Ephemeroptera, to name a few (e.g., Maxted et al., 2000). As an alternative, use of biol ogical traits h as recently been advocated as a potential tool for assessing aquatic ecosystems by academia and the federal government (Poff et al., 2006). Biological traits are more informative indicato rs of ecosystem functi on than are changes in

PAGE 173

173 abundance of individual species, and they are expected to change across a gradient of anthropogenic and natural disturbances (Charvet et al., 2000; Doledec et al., 1999; Statzner et al., 2001). Additionally, biological traits are re gulated at a hierarchy of scales, with environmental filters (e.g., clim ate and geology) creating a templa te for traits present in a specific region (Townsend and Hildrew, 1994; Poff, 1997) Thus, a subset of traits is expected to respond to disturbances within a certain region. In the logged streams, traits were consistent with changes in the stream and were represen ted by species preferring algae and organic matter in the water column, as well as those preferring to live in sandy habitat, reflecting reduction in other habitat types (e.g., leaf litter). The ability of the Florida Stream Condition Index to indicate the im pacts of drought effectively, but not forestry impacts, emphasizes th e need to incorporate natural disturbances into bioassessment programs. Most programs determin e the condition of streams based on a single sample. Even when multiple sampling time periods are included, they typically are 3-4 years apart and occur in different locations than the first sample, since many large scale surveys are probability based (Stoddard et al., 2005). Thus, predictive models need to be developed based on current environmental conditions in the region as compared to historical conditions (e.g., amount of precipitation). The standardized pr ecipitation index was a good indicator of changes in water quality and thus could be incorporated into such a model. Another confounding feature for predicting impacts of logging was related to habitat availability and quality. Logged stre am s were colonized by the macrophyte, Ludwigia repens, which was able to support higher densities and a more diverse invertebrate community. This was accomplished through the stability provided by this ha bitat and its role in trapping organic matter as compared to less stable leaf packs. Trapping of organic matter creates patches similar to

PAGE 174

174 debris dams, and the addition of this habitat was preferred to la ndscapes with only leaf packs by a specialist detritivore ( Anisocentropus pyraloides ). This situation may be unique to coastal plain streams, where fine substrate is often entrained in storm events, creating a dynamically changing landscape. This is in stark c ontrast to mountain streams with higher substrate diversity and stability in the form of boulders and cobble. Testing of any best management practice ultim ately requires an understanding of mechanisms behind changes in stream communiti es, as well as long-term monitoring data. Results from this study provide information on the mechanisms leading to apparent improved water quality in streams impacted by logging. Thus, additional effo rt should be placed on developing assessments specific to coastal plain streams, since most are based upon expected habitat diversity and cha nnel structure found in Piedmont streams.

PAGE 175

175 LIST OF REFERENCES Adams, T.O., D.D. Hook, and M.A. Floyd. 1995. Eff ectiveness m onitoring of silvicultural best management practices in South Carolina. S outh Journal of Applie d Forestry 19: 170-176. Acua, V., I. Muoz, A. Giorgi, M. Omella, F. Sabater, and S. Sabater. 2005. Drought and postdrought recovery cy cles in an intermittent Mediterranean stream: structural and functional aspects. Journal of the North Ameri can Benthological Society 24: 919-933. Allan, J.D. 1995. Stream Ecology: Structure and F unction of Running W aters. Dordrecht, Neth.: Kluwer. 388pp. Allan, J. D. and M. Lammert. 1999. Assessing biotic integrity of stream s: Effects of scale in measuring the influence of land use/cover and ha bitat structure on fish and macroinvertebrates Environmental Management 23: 257-270. Anderson, N.H. and J. R. Sedell. 1979. Detritus processing by m acroinvertebrates in stream ecosystems. Annual Review of Entomology 24: 351-377. Anholt, B.R. 1995. Density dependence resolv es the stream drift paradox. Ecology 76: 2235 2239. Arthur, M. A., G.B. Coltharp, and D.L. Brown. 1998. Effects of best m anagement practices on forest streamwater quality in eastern Kentuc ky. Journal of the American Water Resources Association 34: 481. Alverez, M. 2007. The State of Americas Forests. Society of Am erican Foresters, Bethesda, MD. 76 pp. Arsuffi, T.L. and K. Suberkropp. 1985. Selectiv e feeding by stream ca ddisfly (Trichoptera) detritivores on leaves with fungalcolonized patches. Oikos 45: 50-58. Aust, W.R. and C.R. Blinn. 2004. Forestry best m anagement practices for timber harvesting and site preparation in the eastern United States : An overview of water quality and productivity research during the past 20 years (1982). Wa ter, air, and soil po llution: Focus 4: 5-36. Baldwin D.S., G.N. Rees, A.M. Mitchell, and G. W atson. 2005. Spatial and temporal variability of nitrogen dynamics in an upland stream before and after a drought. Marine and Freshwater Research 56: 457-464. Barber N.L. and T.C. Stamey. 2000. Droughts in Georgia. USGS Open File Report no. 380. Barbour, M.T., J. Gerritsen, B.D. Snyder, and J.B. Stribling. 1999. Rapid bioassessm ent protocols for use in streams a nd wade able rivers: Periphyton, benthic macroinvertebrates and fish, 2nd ed. U.S. EPA, Office of Wa ter. Washington, D.C. EPA 841-B-99-002.

PAGE 176

176 Barlocher, F. and B. Kendrick. 1975. Assimila tion efficiency of Ga mmarus pseudolimnaeus (Amphipoda) feeding on fungal mycelium or autumn-shed leaves. Oikos 26: 55. Bastian, M., L. Boyero, B. R. Jackes, and R. G. Pearson. 20 07. Leaf litter diversity and shredder preferences in an Australian tropical rain-forest stream Journal of Tropical Ecology 23: 219-229 Bayley, S.E., R.S. Behr, and C.A. Kelly. 1986. R etention and release of sulphur from a freshwater wetland. Water, Air, and Soil Pollution 31:101. Beasley, R. S., and A. B. Granillo. 1982. Sedime nt losses from forest practices in the Gulf Coastal Plain of Arkansas. In Proc. Second Biennial Southern Silviculture Research Conference, 461-467. E. P. Jones Jr., ed. General Tech. Report SE-24. Asheville, N.C.: USDA For est Service, Southeastern Forest Experiment Station. Bche, L. A., E.P. McElravy, and V.H. Re sh. 2006. Long-term seasonal variation in the biological traits of benthicm acroinvertebrates in two Mediterranean-climate streams in California, U.S.A. Fr eshwater Biology 51: 56. Bender, D. J., L. Tischendorf, and L. Fahri g. 2003. Using patch isola tion m etrics to predict animal movement in binary la ndscapes. Landscape Ecology 18: 17. Benson, B.J. and J.J. Magnusson. 1992. Spatial heterogeneity of littoral fish assemblages in lakes: relation to species divers ity and habitat structure. Cana dian Journal of Fisheries and Aquatic Sciences 49: 1493-1500. J.P. Benstead and C.M. Pringle. 2004. Defore station alters the resource base and biom ass ofendemic stream insects in eastern Madagascar. Freshwater Biology 49: 490. Beschta, R.L. 1978. Long-term patterns of sedi m ent production following road construction and logging in the Oregon Coast Range. Wa ter Resources Research 14: 1011-1016. Best, L. B., T. M. Bergin, and K. E. Freemark. 2001. Influence of landscape composition on bird use of rowcrop fields. Journal of W ildlife Management 65:442-449. Bider, J. R. 1968. Animal activity in uncontrolle d terrestrial communitie s as determ ined by a sand transect technique. Ecological Monographs 38:269. Bilby, R.E. and P.A. Bisson, 1992. Relative co ntribution of allochthonous and autochthonous organic m atter to the trophic support of fish popu lations in clear-cut and old-growth forested headwater streams. Canadian Journal of Fisheries and Aquatic Sciences 49:540-551. Bilton, D. T., A. Foggo, and S. D. Rundle. 2001. Size, perm anence and the proportion of predators in ponds. Archiv Fur Hydrobiologie 151:451-458. Blindow, I. 1987. The composition and density of epiphyton on several species of submerged sacrophytes the neutral substrate hypot hesis tested. Aquatic Botany 29:157-168.

PAGE 177

177 Blindow, I. 1992. Long-term and short-rerm dynami cs of subm erged macrophytes in 2 shallow eutrophic lakes. Freshwater Biology 28:15-27. Bonada N., M. Rieradevall, N. Prat, a nd V.H. Re sh. 2006. Benthic macroinvertebrate assemblages and macrohabitat connectivity in Mediterranean-climate streams of northern California. Journal of the North Amer ican Benthological Society 25: 32. Bond, N.R., G.L.W. Perry, and B.J. Downes. 2000. Dispersal of organism s in a patchy stream environment under different settlement scenar ios. Journal of Animal Ecology 69: 608-619. Bormann, F. H., G. E. Likens, T. G. Siccama, R. S. Pierce, an d J. S. Eaton. 1974. The export of nutrients and recovery of stable conditions fo llowing deforestation at Hubbard Brook. Ecological Monographs 44:255-277. Boulton A.J. 1989. Over-summering refuges of a quatic m acroinvertebrates in two intermittent streams in central Victoria. Transactions of the Royal Society of South Australia 113: 23. Boulton A.J. and P.S. Lake. 1992. The ecology of two interm ittent streams in Victoria, Australia. II. Comparisons of faunal composition between ha bitats, rivers, and years. Freshwater Biology 27: 99. Boulton, A.J. and J.G. Foster. 1998. Effects of buried leaf litter a nd vertical hydrologic exchange on hyporheic water chem istry and fauna in a gravel-bed river in northern New South Wales, Australia. Freshwater Biology. 40: 229-243. Boulton, A.J. 2003. Parallels and contrast s in the effects of drought on stream macroinvertebrate assemblages. Freshwater Biology 48: 1173-1185. Bowersox, M.A.and D.G. Brown. 2001. Measuri ng the abruptness of patchy ecotone a sim ulation-based comparison of landscape pattern statistics. Plant Ecology 156: 89. Bowne, D.A. and M.A. Bowers. 2004. Interpatch m ove ments in spatially structured populations: a literature review. Landscape Ecology 19: 1. Boyero, L., R. G. Pearson, and M. Bastian. 2007. How biological diversity influences ecosystem function: a test with a tropi cal stream detritivore guild Ecological Research 22: 551-558 Bray J.R. and J.T. Curtis. 1957 An ordination of the upland forest comm unities of southern Wisconsin. Ecological Monographs 27: 325-349. Briers, A., H.R. John, H.M. Gee, and R.G. Cariss. 2004. Inter-population dispersal by adult stoneflies detected by stable isotope en richm ent Freshwater Biology 49: 425. Brittain, J. E., and T. J. Eikeland. 1988. Invertebrate drift a review Hydrobiologia 166:77-93.

PAGE 178

178 Brnmark, C. 1985. Freshwater snail diversity: effects of pond area, habitat heterogeneity, and isolation. Oecologia 67: 127 131. Brosofske, K.D., J. Chen, R.J. Naiman, and J.F. Franklin. 1997. Effects of harvesting on m icroclimatic gradients from streams to upl ands in western Wash ington, USA. Ecological Applications 7: 1188-1200. Brown A. V., Y. Aguila, K. B. Brown, a nd W P. Fowler. 1997. Responses of benthic macroinvertebrates in small intermittent streams to silvicultural practices. Hydrobiologia 347:119. Buesing, N. 2005. Bacterial counts and biomass determ ination by epifluorescence microscopy in M.A.S. Graa, F. Brlocher and M.O. Gessner (eds.), Methods to Study Litter Decomposition: A Practical Guide, pp. 203 208. Springer, Netherlands. Butcher, R.W. 1933. Studies on the ecology of ri vers I. On the dist ribution of m acrophytic vegetation in the rivers of Br itain. Journal of Ecology 21: 58. Cadenasso, M.L., M. M. Traynor, S.T.A. Picke tt. 1997. Functional locatio n of forest edges: Gradients of multiple physical factors. Cana dian Journal of Forest Research 27: 774782. Cain, M. L. 1985. Random search by herbivorous insects: a simulation m odel. Ecology 66: 876888. Campbell, I.C. and T.J. Doeg. 1989. Impact of timber harvesting and production on stream s: a review. Australian Journal of Mari ne and Freshwater Research 40: 519. Campbell, I.C. and L. Fuchshuber. 1995. Polyphe nols, condensed tannins, and processing rates of tropical and tem perate leaves in an Austra lian stream. Journal of the North American Benthological Society 14: 174-182. Caraco, N.F. and J.J. Cole. 2002. Contrasti ng im pacts of a native and alien macrophyte on dissolved oxygen in a large river. Ecological Applications 12: 1496. Caruso, B.S. 2002. Temporal and spatial patterns of extrem e low flows and effects on stream ecosystems in Otago. New Zealand. Journal of Hydrology 257: 115. Charvet S., B. Statzner, P. Usseglio-Polat era, and B. Dum ont. 2000. Traits of benthic macroinvertebrates in semi-natural French stre ams: an initial application to biomonitoring in Europe. Freshwater Biology 43: 277-296. Chevenet F., S. Doledec, and D. Chessel. 1994. A fuzzy coding approach for the analysis of long-term ecological data. Freshwater Biology 31: 295. Churchel M.A. and D.P. Batzer. 2006. Recovery of aquatic m acroinvertebrate communities from drought in Georgia piedmont headwater str eams. American Midla nd Naturalist 156: 259.

PAGE 179

179 Clark, J.S. 1991. Disturbance and Population Structure on the Shifting Mosaic. Landscape Ecology 72: 1119-1137. Clark, M., J. Chapman, J. K. Adamson, and S. N. Lane. 2005. Influence of drought-induced acidification on the mobility of dissolved organi c carbon in peat soils. Global Change Biology 11:791. Closs G.P. and P.S. Lake. 1995. Drought, differentia l m ortality and the coexistence of native and an introduced fish species in a south east Aust ralian intermittent stream. Environmental Biology of Fishes 47: 17. Collier K.J. and M.J. Winterbourn. 1989. Impacts of wetland afforestation on the distribution of benthic invertebrates in acid st ream s of Westland, New Zealand. Ne w Zealand Journal of Marine and Freshwater Research 23: 479-490. Collinge, S.K. and T.M. Palmer. 2002. The influe nces of patch shape and boundary contrast on insect response to fragmentation in Ca lifornia grasslands. L andscape Ecology 17: 647. Corbett, E.S., J.A. Lynch, and W.E. Sopper. 1978. Tim ber harvesting practices and water quality in the eastern United States. Journal Forestry 76: 484-488. Couch, C.A, E.H. Hopkins, and P.S. Hardy. 1996. Influences of environmental settings on aquatic ecosystems in the Appalach icola-Chattahoochee-F lint river basin. U.S. Geological Survey National Water-Quality Assessm en t Program. WaterResources Investigations Report 95-4278. Covich, A.P., M. A. Palmer, and T. A. Cr owl. 1999. The role of benthic invertebrate species in freshwater ecosystems zoobenthic spec ies influence energy flows and nutrient cycling. Bioscience 49: 119-127. Crist, T.O., D.S. Guertin, J.A. Wiens, and B.T. Milne. 1992. Ani mal movement in heterogeneous landscapes: an experiment with Eleodes beetles in shortgrass prairie. Functional Ecology 6:536 544. Cronin, J. T. 2003. Matrix heterogeneity and host-parasitoid interactions in space. Ecology 84:1506-1516. Cronin, J. T. 2003. Patch structure, oviposition behavior, and the distribution of parasitism risk. Ecological Monographs 73:283-300. Crowder, L. B., and W. E. Cooper. 1982. Habita t structural com plexity and the interaction between bluegills and their prey. Ecology 63:1802-1813. Cummins, K. W. 1974. Structure and function of stream ecosystems. Bioscience 24:631-641.

PAGE 180

180 Cummins, K.W. and M. J. Kl ug. 1979. Feeding ecology of stream invertebrates Annual Review of Ecology and Systematics 10: 147-172. Dahm C.N., M.A. Baker, D.I. Moore, and J. R. Thibault. 2003. Coupled biogeochem ical and hydrological responses of streams and rive rs to drought. Freshwater Biology 48: 1219. Dangles, O., F. Guerold, and P. Usseglio-Polater a. 2001. Role of transported particulate organic m atter in the macroinvertebrate colonization of lit ter bags in streams. Freshwater Biology 46: 575-586. Dawson, F.H. 1978. The seasonal effects of aquatic plant growth on the flow of water in a stream Proceedings of the European W eed Research Council Fifth Symposium on Aquatic Weeds, pp. 71-78. European Weed Research Council, Amsterdam. Dean, R. L. and J.H. Connell. 1987. Marine inverteb rates in an algal succession. I: V ariations in abundance and diversity with succession Journal of Experimental Marine Biology and Ecology 109:195-215. Debinski, D.M. and R. Holt. 2000. A survey and ove rview of habitat fragm entation experiments. Conservation Biology 14:343-355. Demmon, E.L. 1951. Forest Research in the Southeas t. 29th Annual Meeting of the Association of State Foresters, Charleston, S.C October 3, 1951. Devito, K.J. and A.R. Hill. 1999. Sulphate mobiliz ation and po re water chemistry in relation to groundwater hydrology and summer drought in two conifer swamps on the Canadian shield. Water, Air, and Soil Pollution 113:97. Diehl, S. 1992. Fish predation and benthic comm unity structure the role of om nivory and habitat complexity. Ecology 73:1646-1661. Diehl, S. and R. Kornijow. 1998. Influence of subm erged macrophytes on trophic interactions among fish and macroinvertebrates. Pp. 24 in E. Jeppesen, M. Sndergaard, M. Sndergaard, and K. Christoffersen, eds. The structuring role of submerged macrophytes in lakes. Springer, New York. Diffendorfer, J.E., M.S. Gaines and R.D. Holt. 1999. Patterns and im pacts of movements at different scales in small mammals. In : Landscape Ecology of Small Mammals, G. Barrett and J. Peles, eds. Springer-Verlag: New York. pp. 63-88. Dodds, W.K. and B. J.F. Biggs. 2002. Water ve locity attenuation by stream periphyton and m acrophytes in relation to growth form and ar chitecture. Journal of the North American Benthological Society 21:2. Dole-Olivier, M.-J., P. Marmonier, and J.L. Beffy. 1997. Response of invertebrates to lotic disturbance: is the hyporheic zone a pa tchy refugium ? Freshwater Biology 37:257.

PAGE 181

181 Doledec, S., B. Statzner, and M. Bournard. 1999. Species traits for future biom onitoring across ecoregions: patterns along a human-impacted river. Freshwater Biology 42:737-758. Doledec, S. and B. Statzner. 2008. Invertebra te traits for the biom onitoring of large European rivers: an assessment of specific types of hum an impact. Freshwater Biology 53:617-634. Douglas, M. and P. S. Lake. 1994. Species Richness of Stream Stones: An Investigation of the Mechanism s Generating the Species-A rea Relationship. Oikos 69:387-396. Downes,B.J., P. S. Lake, E. S. G. Schreibe r, and A. Glaister. 1998. Habitat structure and regulation of local species diversity in a stony, upland stream Ecological Monographs 68:237257. Downes,B.J., P. S. Lake, E. S. G. Schreiber, and A. Glaister. 2000. Habita t structure, resources and divers ity: the separate effects of surface roughness and macroalgae on stream invertebrates. Oecologia 123:569-581. Drew, C.A. and D.B. Eggleston. 2006. Currents, landscap e structure, and recruitment success along a passive-active dispersal gradient. Landscape Ecology 21:917. Dudley, T.L. 1988. The role of plant complexity and epiphyton in colonization of m acrophytes by stream insects. Verhandlungen der Internationalen Vereinigung fur Theoretische und Angewandte Limnologie 23:1153. Dufrene, M. and P. Legendre. 1997. Species asse mblages and indicator species: the need for a flexible asymmetrical approac h. Ecological Monographs 67:345-366. Dunning, J. B., B. J. Danielson, and H. R. Pulliam. 1992. Ecological pro cesses that affect populations in complex landscapes. Oikos 65:169. Eckman, J. E. 1990. A model of passive sett lem ent by planktonic larvae onto bottoms of differing roughness. Limnol ogy and Oceanography 35:887-901. Elliot, J.M. 1971. Upstream movements of benthic inverteb rates in a Lake District stream, Journal of Animal Ecology 40:235. Elliott, J.M. 2002. Time spent in the drift by downstream -dispers ing invertebrates in a Lake District stream. Freshw ater Biology 47:97-106. Elliott, J.M. 2003. A comparative st udy of the dispersal of 10 species of s tream invertebrates. Freshwater Biology 48:1652.

PAGE 182

182 Elwood, J.W., J. D. Newbold, A. F. Trimble, and R. W Stark. 1981. The limiting role of phosphorus in a woodland stream ecosystem: Effect s of P enrichment on leaf decomposition and primary producers. Ecology 62:146-158. Epler, J.H. 1995. Identification manual for the la rval Chironom idae (Diptera) of Florida. FL Department of Environmental Protection, Tallahassee, FL. Epler, J.H. 1996. Identification Manual for the Water Beetles of Florida (Coleoptera: Dryopidae, Dytiscidae, Elmidae, Gy rinidae, Haliplidae, Hydraenidae, Hydrophilidae, Noteridae, Psephenidae, Ptilodactylidae, Scirtidae). State of Florida Department of Environmental Protection Division of W ater Faci lities. Tallahassee, Florida. 228 pp. Erman, N. A. 1986. Movements of self-marked caddisfly larvae, Chyranda centralis (Trichop tera, Limnephilidae) in a Sierran spring stream, California, U.S.A. Freshwater Biology 16:455-464. Escudero, A. and J.M. del Arco. 1987. Ecological significance of the phenology of leaf abscission. Oikos 49:11-14. Essafi, K., H. Chergui, E. Pattee, and J. Math ieu. 1994. The breakdown of dead leaves buried in the sedim ent of a permanent stream in Morocco. Archiv Fur Hydrobiologie 130:105-112. Fagan, W. E., R. S. Cantrell, and C. Co sner. 1999. How habitat edges change species interactions. Am erican Naturalist 153:165-182. Fagan, W.E. 2002. Connectivity, frag m entation, and extinction risk in dendritic metapopulations. Ecology 83:3243-3249. Fagan, W.F., M-J. Fortin, and C. Soykan. 2003. Integrating edge detection and dynam ic modeling in quantitative analyses of ecological boundaries Bioscience 53:730-738. Fahrig, L. and G. Merriam. 1985. Habitat patc h connectivity and populat ion survival. Ecology 66:1762-1768. Fahrig L. and G. Merriam. 1994. Conservation of fragm ented populations. Conservation Biology 8:50. Fahrig, L., and J. Paloheimo. 1988. Determinants of local population size in patchy habitats. Theoretical Population Biology 34:194. Fenoglio, S., P. Agosta, T. Bo, and M. Cucc o. 2002. Field experim e nts on colonization and movements of stream invertebrates in an Ape nnine river (Visone, NW Italy). Hydrobiologia 474:125.

PAGE 183

183 Fenoglio, S., T. Bo, M. Cucco, and G. Malaca rne. 2007. Response of be nthic invertebrate assem blages to varying drought conditions in the Po river (NW Italy). It alian Journal of Zoology 74:191 201. Finn, D. S. and N.L. Poff. 2005. Variability a nd convergence in benth ic communities along the longitudinal gradients of four physically sim ilar Rocky Mountain streams. Freshwater Biology 50:243-261. Fisher, S.G. and G. E. Likens.1973. Energy flow in Bear Brook, New Ha mpshire: An integrative approach to stream ecosystem meta bolism. Ecological Monographs 43:421-439. Fisher, S.G. and S.R. Carpenter. 1976. Ecosystem and macrophyte primary production of the Fort River, Massachusetts Hydrobiologia. 47:175. Flather, C.H. and M. Bevers. 2002. Patchy re action-diffusion and population abundance: The relative im portance of habita t amount and arrangement. American Naturalist 159:40-56. Flecker, A.S. 1992. Fish predation and the evolution of invertebrate drift periodicity: evidence from neotropical str eams. Ecology 73:438-448. Florida Department of Environmental Protecti on. 2004 L. S. Fore, R.Frydenborg, D. Miller,T. Frick, D.W hiting, J. Espy, and L. Wolfe. 2007. Development And Testing of Biomonitoring Tools for Macroinvertebrates in Florida Stream s (Stream Condition Index and Biorecon). Florida Department of Environmental Protection, Tallahassee, FL. 122 pp. Fonseca, D. M. 1999. Fluid-mediated dispersal in stream s: models of settlement from the drift. Oecologia 121:212-223. Forman, R. T. T. and M. Godron. 1986. Lands cape Ecology. John W iley and Sons, New York. Friberg, N. and D. Jacobsen. 1994. Feeding plastic ity of two d etritivore-shredders. Freshwater Biology 32:133. Freilich, J.E. 1991. Movement patterns and ecology of Pteronarcys nym phs (Plecoptera): observations of marked individuals in a Roc ky Mountain stream Freshwater Biology 25:379 394. Fritz, K.M., M. M. Gangloff, and J.W. Femi nella. 2004. Habitat m odification by the stream macrophyte Justicia americana and its e ffects on biota. Oecologia 140:388. Fry, J.C. 1988. Determination of biomass. In: B. Austin (ed.) Methods in Aquatic Bacteriology (pp. 27-72). John W iley and Sons. New York. Fuchs, S.A., S.G. Hinch, and E. Mellina. 2003. E ffects of streamside logging on stream macroinvertebrate communities and habitat in subboreal forests in British Columbia, Canada. Canadian Journal of Forest Research. 33:1408-1415.

PAGE 184

184 Gambi, M. C., A. R. M. Nowell, and P. A. Jumars. 1990. Flume Observations on Flow Dynamics in Zostera-Marina (Eelgrass) Beds. Marine E cology-Progress Series 61:159-169. Garman, G. C. and J. R. Moring. 1993. Diet a nd annual production of tw o boreal river fishes following clearcut logging. Environm ental Biology of Fishes 36:301. Gasith, A. and V.H. Resh. 1999. Streams in medi terranean clim ate regions: abiotic influences and biotic responses to predictable seasonal events. Annual Review of Ecology and Systematics 30:51-81. Gates, J.E., amd L.W. Gysel. 1978. Avian nest pr edation and fledgling success in field-forest ecotones. Ecology 59:871-883. Gelhaus, J.K. 2002. Manual for the identifi cation of aquatic crane fly larvae for southeastern United States. Pr epared for the carolina area benthological w orkshop. Durham, North Carolina. 57 pp. Georgia Forestry Commission. 1999. Georgias best m anagement practices for forestry. 71 pp. Gilliam, J. F., and D. F. Fraser. 2001. Movement in corridors : Enhancement by predation threat, disturbance, and habitat structure. Ecology 82:258-273. Golladay S.W., J.R. Webster, and E.F. Benfie ld E.F. 1987. Changes in stream morphology and storm transport of seston following watershed disturbance. Journal of the North American Benthological Society 6:1-11. Golladay, S. W., and C. L. Hax. 1995. Effects of an engineered flow disturbance on m eiofauna in a north Texas prairie stream. Journal of the North American Benthological Society 14:404-413. Golladay, S.W. and J.B. Battle. 2002. Effect s of flooding and drought on water quality in Gulf Coastal Plain streams in Georgia. Journal of Environm ental Quality 31:1266-1272. Gomi, T., R. C. Sidle, and J. S. Richards on. 2002. Understanding pr ocesses and downstream linkages of headwater sy stems. BioScience 52:905. Gregg, W.W. and F.L. Rose. 1982. The effect s of aquatic macrophytes on the stream microenvironment. Aquatic Botany 14:309. Gurtz, M.E. and J.B. Wallace. 1984. Substrate-me diated response of stre am invertebrates to disturbance. Ecology 65:1556-1569. Guttman, N.B. 1999. Accepting the st andardized precipitation inde x: a calculation algorithm Journal of the American Water Re sources Association 35:311-322.

PAGE 185

185 Goodwin, B.J. and L. Fahrig. 2002. How doe s landscape structure influence landscape connectivity? Oikos 99:552. Goodwin, B.J. 2003. Is landscape co nnectivity a dependent or inde pendent variable? Landscape Ecology 18:687. Grace, J. M., III, R. W. Skaggs, H. R. Malcom, G. M. Chescheir, and D. K. Cassel. 2003. Increased water yields following harvesting oper ations on a drained coastal watershed. ASAE Paper No. 032038. St. Joseph, Mich.: ASAE. Grassle, J. P., C. A. Butman, and S. W. Mills 19 92. Active habitat sele ction by Capitella sp. I larvae. II. Multiple-choice experiments in still water and flume flows. J. Mar. Res. 50:717-743. Griswold, M.W., R.T. Winn, T. L. Crisman, and S. W Golladay. Impacts of climatic stability on structural and functional aspects of macr oinvertebrate commun ities following drought ( In Revision : Freshwater Biology) Growns, I. O., and J. A. Davis. 1994. Effects of forestry activ ities (c learfelling) on stream macroinvertebrate fauna in south-western Australia. Australian Journal of Marine and Freshwater Research 45:963. Gurtz, M. E. and J. B. Wallace. 1984. Substrate mediated respons e of stream invertebrates to disturbance. Ecology 65:1556. Haefner, J. D., and J. B. Wallace. 1981. Shifts in aquatic insect popula tio ns in a first-order southern Appalachian stream following a decade of old field succession. Canadian Journal of Fisheries and Aquatic Sciences 38:353. Hanski, I. 1994. Patch-Occupanc y Dyna mics in Fragmented La ndscapes. Trends in Ecology and Evolution 9:131-135. Hanski, I. 1994. A Practical Model of Metapopul ation Dynam ics. Journal of Animal Ecology 63:151-162. Hanski, I. 1994. Spatial Scale, Patchiness and Population-Dynam ics on Land. Philosophical Transactions of the Royal Society of London Series B-Biological Sciences 343:19-25. Hanski, I. 1998. Metapopulation Dynamics. Nature 396: 41-49. Hanski, I. 1999. Habitat connectivity, habitat continuity, and m etapopulations in dynamic landscapes. Oikos 87:209-219. Harding, J. S., 2003. Historic defo restation and the fate of endem ic invertebrate species in streams. New Zealand Journal of Mari ne and Freshwater Research 37:333.

PAGE 186

186 Hargrave B.T. 1970. The utilization of benthic microflora by Hyalella azteca (Am phipoda). Journal of Animal Ecology. 39:427. Harris, P.M. and B.R. Roosa. 2002. Unde rground dispersal by am phipods (Crangonyx pseudogracilis) between temporary ponds. J ournal of Freshwater Ecology 17:589-600. Harrison, S. and L. Fahrig. 1994. Landscap e structure and population survival. In : Mos aic Landscapes and Ecological Processes, eds. L. Hansson, L. Fahrig and G. Merriam, pp. 293-308. Chapman and Hall, New York, USA. Hart, D. D. and V. H. Resh. 1980. Movement patterns and f oraging ecology of a stream caddisfly larva. Canadian Journal of Zoology-Revue Canadienne De Zoologie 58:1174-1185. Hart, D.D. and C.T. Robinson. 1990. Resource lim itation in a stream community: Phosphorus enrichment effects on periphyton and grazers. Ecology 71:1494-1502. Hart, D.D. and R.A. Merz. 1998. Predator-prey interactions in a benthic stream community: a field test of flow-mediated refuges. Oecologia 114:263-273. Hartman, G.F. and J.C. Scrivener. 1990. Impact s of forestry practices on a coastal stream ecosystem, Carnation Creek, British Columbia. Canadian Bulletin of Fisheries and Aquatic Sciences 223:1-148. Hastings, H. M. and G. Sugihara. 1993. Fractals: a users guide to the natural sciences. Oxford University Press, Oxford. Hax, C.L. and S. W. Golladay. 1993. Macroinver tebrate colonization an d biofilm development on leaves and wood in a boreal river Freshwater Biology 29:79. Haynes K.J. and J.T. Cronin. 2003. Matrix compos ition affects th e spatial ecology of a prairie planthopper. Ecology 84:2856-2866. Haynes, K.J. and J. T. Cronin. 2004. Confounding of patch quality and matrix effects in herbivore movem ent studies. Landscape Ecology 19:119. Hawkins, C. P., M. L. Murphy, and N. H. Ander son. 1982. Effect of canopy, substrate composition, and gradient on the structure of macroinvertebrate communities in cascade range streams of Oregon. Ecology 63:1840. Heard, S. B. 1998. Resource patch density and larv al aggregation in m ushroom-breeding flies. Oikos 81:187-195. Hearnden, M.N. and R. G. Pearson. 1991 Habi tat partitioning am ong the mayfly species (Ephemeroptera) of Yuccabine Creek, a tropi cal Australian stream Oecologia 87:91-101.

PAGE 187

187 Heck, K.L. and G.S. Westone. 1977. Habitat comple xity and invertebrate species richness and abundance in tropical seagrass m eadow s. Journal of Biogeography 4:135. Heino J., H. Mykr, J. Kotanen, and T. Muotka. 2007. Ecological filters and variability in stream macroinvertebrate communities: do taxonomic and functional structure follow the same path? Ecography 30:217-230. Hepinstall, J.A. and R.L. Fuller. 1994. Periphyton reactions to different light and nutrient levels and the response of bacteria to these m anipulations Archiv fur Hydrobiologie. 131:161-173. Herlihy, A.T., W. J. Gerth, J. Li, and J. L. Banks. 2005. Macroinvertebrate comm unity response to natural and forest harv est gradients in western Oregon headwater streams Freshwater Biology 50:905. Hershey, A. E. and S. I. Dodson. 1985. Selective predation by a scul pin and a stonefly on 2 Chironom ids in laboratory feedi ng trials. Hydrobiologia 124:269-273. A.E. Hershey, J. Pastor, B.J. Peterson, and G. W Kling. 1993. Stable isotopes resolve the drift paradox for Baetis mayflies in an arctic river, Ecology 74:2315. Hildrew, A. G. And P. S. Giller.1994. Patchiness, species interaction s and disturbance in stream benthos. Pages 21-61 in PS. Giller, A.G. Hildrew, and D.G. Rafaelli (editors). Aquatic ecology: scale, pattern and process. Blackwell Scientific, London. Hill, W.R. and A. W. Knight 1988. Nutrient and light limitation of algae in two northern California streams. Journal of Phycology 24:125. Hill, W.R., M.G. Ryon, and E.M. Schilling. 1995 Light lim itation in a stream ecosystem: responses by primary producers and consumers. Ecology 76:1297. Hillebrand, H. 2002. Top-down versus bottom-up contro l of autotrophic biomass: a metaanalysis on experiments with periphyton. J ournal of the North American Benthological Society 21:349. Hinton H.E. 1960. Cryptobiosis in the larva of Po lypedilum vanderplanki Hint. (Chironomidae). Journal of Insect Physiology 5:286-300. Hoeinghaus, D. J., K.O. Winemiller, and J.S. Birnbaum 2007. Local and regional determinants of stream fish assemblage structure: Infere nces based on taxonomic vs. functional groups. Journal of Biogeography 34:324-338. Hogg, I.D. and D. D.Williams 1996. Response of stream invertebrates to a global-warming thermal regime: An ecosystem-lev el manipulation Ecology 77:395-407. Holt, R.D. 1977. Predation, apparent competition a nd structure of prey co mmunities. Theoretical Population Biology 12:197-229.

PAGE 188

188 Holt, R.D. 1996. Temporal and spatial aspe cts of food web structure and dynamics. In : Food Webs: Contemporary Perspectives, G. Polis and K. Winemiller, eds. Chapman and Hall. pp. 255257. Hooper, D.U., F.S. Chapin, J.J. Ewel, A. Hector P. Inchausti, S. Lavorel, J.H. Lawton, D.M. Lodge, M. Loreau, S. Naeem B. Schmid, H. Seta, A.J. Symstad, J. Vandermeer, and D.A. Wardle. 2005. Effects of biodiversity on ecosy stem functioning: a concensus of current knowledge. Ecological Monographs 75:3-35. Howe, M. J. and K. Suberkropp. 1994. Effects of isopod (Lirceus sp.) feeding on aquatic hyphom ycetes colonizing leaves in a stream. Arch. Hydrobiol. 130:93. Hughes, J. M., S. E. Bunn, D. A. Hurwood, S. Choy, and R. G. Pearson. 1996. Genetic differentiation am ong populations of Caridina zebra (Decapoda: Atyi dae) in tropical rainforest streams, northern Australia. Freshwater Biology 36:289-296. Huryn, A.D. and M.W. Denny. 1997. A biomech anical hypothesis explaining upstream movements by the freshwater snail Elimia. Functional Ecology 11:472 483. Hynes, H. B. N. 1960, The biology of polluted wa ters: Liverpool University Press, Liverpool, UK. Hynes, H.B.N. 1975. The stream and its valle y. V erh. Internat. Verein Limnol. 19:1-15. Ims, R. A. 1995. Movement patterns re lated to spatial st ructures. pp. 85-109 In Hansson, L., Fahrig, L. and Merriam G. (eds.). Mosaic la ndscapes and ecological processes. Chapman and Hall, London. J. K. Jackson, E. P. Mcelravy, and V. H. Resh. 1999. Long-term movements of self-marked caddisfly larvae (Trichoptera: Sericostomatidae) in a California coastal mountain stream. Freshwater Biology 42:525. Jackson, C. R., C. A. Sturm, and J. M. War d. 2001. Tim ber harvest impacts on small headwater stream channels in the coast ranges of Wash ington. Journal of the American Water Resources Association 37:1533-1549. Johnson, A.R., B.T. Milne, and J.A. Wiens. 1992 Diffusion in fractal la ndscapes: sim ulations and experimental studies of Tenebrionid beetle movements. Ecology 73:1968. Johst, K., R. Brandl, and S. Eber. 2002. Metapopulation persistence in dynam ic landscapes: the role of dispersal distance. Oikos 98:263-270. Jones D.G., W.B. Summer, M. Miwa, and C.R. Jackson. 2003. Base line characterization of forested headwater stream hydrology and water chem istry in southwest Georgia.

PAGE 189

189 in : Proceedings of the 12th Biennial Southern Silviculture Resear ch Conference, pp. 161-165. Biloxi, MS. 24-28 February 2003. Southern Research Station: USDA Forest Service, Asheville, NC. Jonsen, I.D. and L. Fahrig. 1997. Response of gene ralist and specialist in sect herbivores to landscape spatial structur e. Landscape Ecology 12:185. Jonsen, I.D. and P.D. Taylor. 2000. Fine-scale move ment behavior s of Calopterygid damselflies are influenced by landscape structure: an experimental mani pulation. Oikos 88:553. Justic,D., N.N. Rabalais, R.E. Turner, and W .J. Wiseman. 1993. Seasonal coupling between riverborne nutrients, net productivity a nd hypoxia, Marine Pollution Bulletin 26:184. Kareiva, P.M. and N. Shigesada. 1983. Analyzi ng insect m ovement as a correlated random walk. Oecologia 56:234. Kareiva, P. 1990. Population dynamics in spatia lly com plex environments: theory and data. Philosophical Transactions of the Royal Society of London : B. 330:175-190. Kareiva, P., and U. Wennergren. 1995. Conn ectin g landscape pattern s to ecosystem and population processes. Nature 373:299-302. Kedzierski, W.M. and L.A. Smock. 2001. Effects of logging on m acroinv ertebrate production in a sand-bottomed, low-gradient stre am. Freshwater Biology 46:821-833. Kemp, M.J. and W.K. Dodds. 2001. Centimeter -scale patterns in dissolved oxygen and nitr ification rates in a prairi e stream. Journal of the North American Benthological Society 20:347. Kiffney, P.M., J.S. Richardson, and J.P. Bull. 2003. Responses of periphyton and insects to experim ental manipulation of riparian buffer wi dth along forest streams. Journal of Applied Ecology 40:1060. Kirchman, D.L. 1993. Statistical analysis of direct counts of m icrobial abundance. In : P.F. Kemp, B.F. Sherr, E.B. Sherr, and J.J. Cole (eds.). Handbook of Methods in Aquatic Microbial Ecology, pp. 117-119. Lewis Publishers. Boca Raton. Knoepp, J. D. and W.T. Swank. 1993. Site preparat ion burning to im prove southern Appalachian pine-hardwood stands: Nitrogen responses in soil, soil water and streams. Canadian Journal of Forest Research. 23:2263-2270. Kolkwitz, R. and M. Marsson. 1909. O kologie de r tierischen Saprobien. Internationale Revue der gesam ten Hydrobiologie und Hydrographie 2:126.

PAGE 190

190 Kozakiewicz, M. 1995. Resource tracki ng in space and tim e. Chapter 6 in L. Hansson, L. Fahrig, and G. Merriam, eds. Mosaic Landscapes a nd Ecological Processes. Chapman and Hall, London. Kruskal, J.B. 1964. Nonmetric multidimensional scaling: a numerical method. Psychometrika 29:115. Kreutzweiser, D.P., S. S. Capell, and K. P. Good. 2005. Macroinvertebrate community responses to selection logging in riparian and upland areas of headwater cat chm ents in a northern hardwood forest Journal of the North American Benthological Society 24:208. Kundzewicz Z.W., L.J. Mata, N.W. Arnell, P. Dll, P. Kabat, B. Jim nez, K.A. Miller, T. Oki, Z. Sen, and I.A. Shiklomanov. 2007. Freshwat er resources and their management. In : Climate Change 2007: Impacts, Adaptation and Vulnerabili ty. Contribution of Working Group II to the Fourth Assessment Report of the Intergovernment al Panel on Climate Change. (Eds Parry M.L., Canziani O.F., Palutikof J.P., van der Linde n P.J. and Hanson C.E.), pp. 173-210. Cambridge University Press, Cambridge, UK. Lake, P.S. 2000. Disturbance, patchiness, and di versity in stream s. Journal of the North American Benthological Society 19:573. Lake, P.S. 2003. Ecological effects of perturbation by drought in flowing waters. Freshwater Biology 48:1161-1172. Lamouroux, N., S. Doldec, and S. Gayraud. 2004. Bi ological traits of stream macroinvertebrate communities: effects of microhabitat, reach and basin filters. Journal of the North American Benthological Society 23:449. Lancaster, J. 1996. Scaling the e ffects of predation and disturba nce in a patchy environm ent. Oecologia 107:321-331. Lancaster, J., and L. R. Belyea. 1997. Nested hierarchies an d scale-dependence of mechanisms of flow refugium use. Jour nal of the North American Be nthological Society 16:221-238. Lancaster, J. 2000. Geometric scaling of microhab itat patches and their e fficacy as refugia during disturbance. Journal of Anim al Ecology 69:442-457. Lancaster, J., T. Buffin-Bela nger, I.Reid, and S. Rice. 2006. Flowand substratum -mediated movement by a stream insect Freshwater Biology 51:1053. Ledger, M.E. and A.G. Hildrew. 1998. Temporal and spatial variation in the epilithic biofilm of an acid stream. Freshwater Biology 40:655. Ledger, M.E. and A.G. Hildrew. 2005. The ecology of acidification and recovery: changes in herbivore-algal food web linkages across a stream pH gradient. Environm ental Pollution 137: 103-118.

PAGE 191

191 Legendre, P. and L. Legendre. 1998.Numerical E cology, 2nd edition. Elsevier, Amsterdam. Likens,G.E., F.H. Bormann, and H.M. Johnson. 1969. Nitrification: im porta nce to nutrient losses from a cutover forested ecosystem. Science 163:1205-1206. Likens,G.E., F.H. Bormann, R.S. Pierce, and W.A. Reiners. 1978. Recovery of a deforested ecosystem : replacing biomass and nu trients lost in harvesting nor thern hardwoods may take 60 to 80 years. Science 199:492-496. Lima, S. L., and P. A. Zollner. 1996. Towards a behavioral ecology of ecological landscapes. Trends in Ecology and E volution 11:131-135. Lock, M.A., R. R. Wallace, J. W. Costerton, R. M. Ventullo and S. E. Charlton. 1984. River epilithon: Toward a structural-f unctional model. Oikos 42:10-22. Lodge, D.M. 1991. Herbivory on freshwat er m acrophytes. Aquatic Botany 41:195. Loferer-Krbacher, M., J. Klima, and R. Psenner. 1998. Determ ination of bacterial cell dry mass by transmission electron microscopy and de nsiometric image analysis. Applied and Environmental Microbiology 64:688-694. Long, E.R., D.A. Wolfe, R.S. Carr, K.J. Scott, G. B. Thursby, H.L. W indom, R. Lee, F.D. Calder, G.M. Sloane, and T. Seal. 1994. Magnitude and ex tent of sediment toxicity in Tampa Bay, Florida. NOAA Technical Memorandum NO S ORCA 78. Silver Spring, Maryland. Lowe, R.L., S.W. Golladay, and J.R. Webs ter. 1986. Periphyton response to nutrient m anipulation in streams draining clearcut and forested wate rsheds. Journal of the North American Benthological Society 5:221. Lowe, W.H. 2002. Landscape-scale spatial popula tion dynam ics in human-impacted stream systems. Environmental Management 30:225. Mackay, R.J. and J. Kalff. 1973. Ecology of two related species of caddis fly larvae in the organic substrates of a woodland stream Ecology 54:499-511. MacKay, R.J. 1992. Colonization by lotic macroi nvertebrates: A review of processes and patterns. Canadian Journal of Fish eries and Aquatic Sciences 49:617-628. Mandelbrot, B. 1967. How long is the coast of Brita in? Statistical self-sim ilarity and fractional dimension. Science 156:636. Marklund, O., I. Blindow, and A. Hargeby. 200 1. Distribution and diel m igration of macroinvertebrates within dense submerge d vegetation. Freshwater Biology 46:913-924.

PAGE 192

192 Martin, C.W. and R.S. Pierce. 1980. Clearcutting patte rns effect nitrate and calcium in streams of New Hampshire. J. Forestry 78:268272. Martin, L.A. P.J. Mulholland, J.R. Webster, and H.M. Valett. 2001. Denitrification potential in sedim ents of headwater streams in the southern Appalachian Mountains, USA. Journal of the North American Benthol ogical Society 20:505-519. Matter, S. F. 1996. Interpatch movement of the red m ilkweed beetle, Tetraopes tetraophthalmus: individual responses to patch si ze and isolation. Oecologia 105:447. Matthews, W.J. 1998. Patterns in Freshwater Fish Ecology. Chapm an and Hall, New York. Mayer, M.S. and G.E. Likens. 1987. The importance of algae in a shaded headwater stream as food for an abundant caddisfly (Trichoptera). Jo urnal of the North American Benthological Society 6:262-269. McAuliffe, J.R. 1984. Competition for space, disturba nce, and the structure of a benthic stream community. Ecology 65:894-908. McBroom, M., M. Chang, and A. K. Sayok. 2002. Fore st clearcutting and si te preparation on a salin e soil in east Texas: Im pacts on water quality. In Proc. Eleventh Biennial Southern Silviculture Research Conference, 535-542. K. W. Outcalt, ed. General Tech. Report SRS-48. Asheville, N.C.: USDA Forest Serv ice, Southern Research Station. McClurkin, D. C., P. D. Duffy, S.J. Ursic, and N .S. Nelson. 1985. Water quality effects of clearcutting upper coastal plain loblolly pine plantations. Jo urnal of Environmental Quality 14: 329. McCune, B. and M.J. Mefford. 1999. PC-ORD: Multiv ariate analysis of ecological data,Version 5. MjM Software, Gleneden Beach, OR, U.S.A. McCune, B., J.B. Grace, and D.L. Urban. 2002. Analysis of Ecological Communities. MjM Software Design, Glened en Beach, Oregon. McIntyre, N.E. and J.A. Wiens. 1999. How does ha bitat patch size affect anim al movement? An experiment with darklin g beetles. Ecology 80:2261. McKee, T. B, N.J. Doeskin, and J. Kieist. 1995. Drought m onitoring with multiple time scales. In: Proceedings of the 9th Conferen ce on Applied Climatology, pp. 233-236. Boston, Massachusetts, January 15-20, American Mete orological Society, Boston, Massachusetts. Melo, A.S. and C. Froelich. 2001. Evaluation of m ethods for estimating macroinvertebrate species richness using indivi dual stones in tropical stream s. Freshwater Biology 46:711-721.

PAGE 193

193 Merriam, G.. 1984. Connectivity: a f undam ental ecological characteri stics of landscape pattern. In : Brandt, J. and Agger, P. (eds), Methodology in landscape ecological research and planning. Roskilde Universitetsforlag, GeuRuc, Roskilde, Denmark, pp. 5 15. Merritt, R.W. and K.W. Cummins. 1996. An Introduction to the Aquatic Insects of North America, 3rd edition. Kendall/H unt Publishing Com pany, Dubuque, Iowa. Meyer, J.L. and J.B. Wallace. 2001. Lost lin kages and lotic ecology: rediscovering sm all streams. In :Ecology: Achievement and Challenge (eds M. C. Press, N. Huntly and S. Levin), pp. 295. Blackwell Science, Oxford, UK. Meyer, J.L., D.L. Strayer, J.B. Wallace, S.L. Eggert, and G.S. Helfman, 2007. The Contribution of Headwater Streams to Biodi versity in River Networks. J ournal of the American Water Resources Association 43:1752-1688. Miltner, R.J. and E.T. Rankin. 1998. Primary nutrien ts and the biotic integrity of rivers and stream s. Freshwater Biology 40:145-158. Minshall, G.W. 1984. Aquatic inse ct-substratum relationships. In : The Ecology of Aquatic Insects (Eds V.H. Resh and D.M. Ro senberg), pp. 358. Praeger Publishers, New York. Miyake,Y., T. Hiura, N.Kuhara, and S.Naka no. 2003. Succession in a stream invertebrate community: A transition in species dominance through colonization Ecological Research 18: 493. Moilanen, A. and I. Hanski. 1998. Metapopulation dynam ics : effects of habitat quality and landscape structure. Ecology 79:2503-2515. Moloney, K.A. and S.A. Levin. 1996. The effects of disturbance architect ure on landscape-level population dynam ics. Ecology 77:375-394. Moore, R.D. and J.S. Richardson. 2003. Progress towards understanding th e structure, function and ecolog ical significance of small stream channels and their riparian zones. Canadian Journal of Forest Research. 33:1349-1351. Moore, J.C., E. L. Berlow, D. C. Coleman, P. C. de Ruiter, Q Dong, A. Hastings, N. C. Johnson, K. S. McCann, K. Melville, P. J. Mori n, K. Nadelhoffer, A. D. Rosemond, D. M. Post, J. L. Sabo, K. M. Scow, M. J. Vanni and Diana H. W all. 2004. Detritus, trophic dynamics and biodiversity. Ecology Letters 7:584. Morgan, M. D. 1985. Photosynthetica lly elevated pH in acid wate rs with high nutrient content and its significance for the zooplan kton comm unity. Hydrobiologia 128:239.

PAGE 194

194 Mulholland, P.J., A. V. Palumbo, J.W. El wood, and A. D. Rosem ond. 1987. Effects of acidification on leaf decomposition in streams. Journal of the North American Benthological Society 6:147-158. Mller, K. The colonization cycle of fr eshwater insects. Oecologia 53:202. Murphy, M. L. C., P. Hawkins, and N. H. Ander son. 1981. Effects of canopy modification and accumulated sediment on stream communities. Tran sactions of the American Fisheries Society 110:469. Murphy, M.L., J. Heifetz, S.W. Johnson, K.V. Ko ski, and J.F Thedinga. 1986. Effects of clearcut logging with and without buffe r strips on juvenile salmonids in Alaskan streams. Canadian Journal of Fisheries and Aquatic Sciences 43:1521-1533. Murphy, F., P. S. Giller, and M. A Horan. 1998. Sp atial scale and the agg regation of stream macroinvertebrates associated with l eaf packs Freshwater Biology 39:325. Nakano, S., H. Miyasaka, and N. Kuhara. 1999. Te rrestrial-aquatic linkage s: riparian arthropod inputs alter trophic cascades in a stream food web. Ecology 80:2435-2441. Nams, V. O. and M. Bourgeois. 2004. Fractal dime nsion m easures habitat us e at different spatial scales: an example with marten. Ca nadian Journal of Zoology 82:1738-1747. Newman, R.M. 1991. Herbivory and detritivory on freshwater m acrophytes by invertebrates: a review. Journal of the North Amer ican Benthological Society 10:89. Niederlehner, BR and J. Cairns Jr. 1990. Eff ects of a mmonia on periphytic communities. Environmental Pollution 66:207-21. Newbold, J. D., D. C. Erman, and K. B. Roby. 1980. Effects of logging on m acroinvertebrates in streams with and without buffer strips. Canadian Journal of Fisheries and Aquatic Sciences 37:1076. Noel, D. S., C. W. Martin, and C. A. Federe r. 1986. Effects of forest clearcutting in N ewEngland on stream macroinvertebrates and pe riphyton. Environmental Management 10:661-670. Nolen, J.A. and R.G. Pearson. 1993. Factors a ffecting litter processing by Anisocentropus kirram us (Trichoptera: Calamoceratidae) from an Australian tropical rainforest stream. Freshwater Biology 29:469. OHare, M.T. and K. J. Murphy. 1999. Invert ebrate hydraulic m icrohabitat and community structure in Callitriche stagnalis Scop. Patches. Hydrobiologia 415:169. O'hop, J., J. B. Wallace, and J.D. Haefner. 1984. Production of a stream shredder, Peltoperla maria (Plecoptera: Peltoperlid ae) in disturbed and undisturbed hardwood catchments Freshwater Biology 14:13.

PAGE 195

195 Olden, J.D., A. L. Hoffman, J. B. Monroe, a nd N. L. Poff. 2004. Movem ent behaviour and dynamics of an aquatic insect in a stream bent hic landscape. Canadian Journal of Zoology 82: 1135. Olden, J.D. 2007. Critical threshold effects of benthiscape structure on stream herbivore movement. Philosophical Transactions of the Royal Society B: Biol ogical Sciences, 362:461472. Olsen D.A., C.D. Matthaei, and C.R. Townsend. 2007. Patch history, invertebrate patch dynam ics and heterogeneous community composition: perspectives from a manipulative stream experiment. Marine and Fr eshwater Research 58:307. Ortega M., M.L. Suarez, M.R. Vida l-Abarca, R. Gom ez, and D.L. Ram rez. 1991. Aspects of postflood recolonization of macroi nvertebrates in a rambla of south-east Spain (Rambla del Moro: Segura River Basin). Verhandlungen der Internationalen Vereinigung fur theoretische und angewandte Limnologie 24:1994. Ostrofsky, M.L. 1993. Effect of tannins on leaf processing and condition ing rates in aquatic ecosystem s: An empirical approach. Canadian Journal of Fisheries and Aquatic Sciences 50:1176-1180 Ostrofsky, M.L. 1997. Relationship between chemical characteristics of autum n-shed leaves and aquatic processing rates. Journal of the No rth American Benthological Society 16:750-759. Palmer, M. A., P. Arensburger, A. P. Marti n, and D. W Denman. 1996. Disturbance and patch specific responses: The interactiv e effects of woody debris and floods on lotic invertebrates. Oecologia 105:247-257. Palmer, M.A., C. M. Swan, K. Nelson, P. Silver, and R. Alvestad. 2000. Stream bed landscapes: evidence that stream invertebrates respond to th e type and spatial arrangement of patches. Landscape Ecology 15:563. Peck, L.S. and R.F. Uglow. 1990. Two methods for the assessment of the oxygen content of sm all volumes of seawater. Journal of E xperimental Marine Biology and Ecology 141:53-62. Peckarsky, B.L. 1983. Biotic interactions or ab iotic lim itations? A model of lotic community structure. In : Dynamics of Lotic Ecosystems (Eds T.D. Fontaine and S.M. Bartell), pp. 303. Ann Arbor Science Publishers, Ann Arbor, Michigan. Pescador, M.L., A.K. Rasmussen, and S.C. Harris. 1995. Identification Manual for the Caddisfly (Trichoptera) La rvae of Florida. State of Florida Departm ent of Environmental Protection Division of W ater Facilities. Tallahassee, Florida. 132 pp. Pescador, M.L., A.K. Rasmussen, and B.A. Richard. 2000. A Guide to the Stoneflies

PAGE 196

196 (Plecoptera) of Florida. State of Florid a Department of Environm ental Protection Division of Water Resource Managem ent. Tallahassee, Florida. 94 pp. Peters, R. H. and K. Wassenberg. 1983. The Effect of Body Size on Animal Abundance. Oecologia 60:89-96. Petersen, R.C. and K.W. Cummins. 1974. Leaf processing in a woodland stream Freshwater Biology 4:343. Petersen, Jr., R.C., K. W. Cu mm ins, and G. M.Ward. 1989. Microbi al and animal processing of detritus in a woodland stream. Ecological Monographs 59:21-39. Pickett, S. T. A. and J. N. Thompson. 1978. Patch dynam ics and design of nature reserves. Biological Conservation 13:27-37. Pither, J. and P.D. Taylor. 1998. An experiment al assessm ent of landscape connectivity. Oikos 83:166. Poff, N. L. and J. V. Ward. 1992. Heterogeneous currents and algal resources m ediate in situ foraging activity of a mobile stream grazer. Oikos 65:465-478. Poff, N.L., Palmer, M.A., P.L. Angermeier, R. L. Vadas, Jr., C.C. Hakenkam p, A. Bely, P. Arensburger, and A.P. Martin. 1993. Size structur e of the metazoan community in a Piedmont stream. Oecologia 95:202. Poff, N. L. and J. V. Ward. 1995. Herbivory under different flow regim es a field experiment and test of a model with a bent hic stream insect. Oikos 72:179-188. Poff, N.L. 1997. Landscape filters and species traits : toward mechanistic understanding and prediction in stream ecology. J ournal of the North American Benthological Society 16:391. Poff, N.L., J.D. Olden N.K.M. Vieira, D.S. Finn, M.P. Simmons, and B.S. Kondratieff. 2006. Functional trait niches of North Am erican lotic in sects: traits-based ecological applications in light of phylogenetic relationships Journal of the North Ameri can Benthological Society 25: 730. Polyakov, V., A. Fares, and M. H. Ryder. 2005. Precision riparian buffers for the control of nonpoint so urce pollutant loading into surface wa ter: A review. Environmental Reviews 13:129 144. Pusey, B.J. and A.H. Arthington. 2003. Importance of the riparian zone to the conservation and m anagement of freshwater fish: a review Marine and Freshwat er Research 54:1-16. Rahel, F.J. 1990 The hierarchical nature of community persistence: A problem of scale. American Naturalist 136:328-344.

PAGE 197

197 Ranney, J.W., Bruner, M.C., and Levenson, J.B. 1981. The importance of edge in the structure and dynam ics of forest islands. In : R.L. Burgess and D.M. Sharpe, eds. Forest island dynamics in man-dominated landscapes. Spri nger-Verlag, New York, pp. 67-95. Rawer-Jost, J., J. Bohmer, J. Blank, and H. Rahmann. 2000. Macroinv ertebrate functional feeding groups in biological asse ssm ent. Hydrobiologia 422/423:225-232. Rempel, R.S. and J.C.H. Carter. 1986. Experiment al study on the effect of elevated tem perature on the heterotrophic and autotrophic food resources of aquatic insects in a forested stream. Canadian Journal of Zoology 64:2457-2466. Resh V.H., A.V. Brown A.P. Covich, M.E. Gurt z, H.W Li, G.W. Minshall, S.R. Reice, A.L. Sheldon, J.B. Wallace, and R.C. Wissmar. 1988. The role of disturbance in stream communities. Journal of the North American Benthological Society 7:433-455. Reice, S. R. 1974. Environmental patchiness and the break-d own of leaf litter in a woodland stream. Ecology 55:1271-1282. Reice, S. R. 1991. Effects of detritus loading and fish predation on leafpack breakdown and benthic m acroinvertebrates in a woodland stream. Journal of the North American Benthological Society 10:42-56. Remer, L. C. and S. B. Heard. 1998. Local movement and edge effects on competition and coexistence in ephem eral-patch mode ls. American Naturalist 152:896-904. Richards, C. and G. W. Minshall. 1988. The influence of periphyton abundance on Baetis bicaudatus distribution and colonization in a sm all stream Journal of the North American Benthological Society 7:77-86. Richards, C., G. E. Host, and J. W. Arthur. 1993. Identification of Pred ominant Environm entalFactors Structuring Stream Macroinvertebrate Communities within a Large Agricultural Catchment. Freshwater Biology 29:285-294. Richards C., R.J. Haro, L.B. Johnson, G.E. Ho st. 1997. Catchm ent and r each-scale properties as indicators of macroinvertebrate spec ies traits. Freshwater Biology 37:219-230. Richardson, J.S. 2003. Identification Manual for the Dragonfly Larvae (A nisoptera) of Florida. Tallahassee, Florida. 114 pp. Rier, S.T.and R.J. Stevenson. 2002. Effects of light, dissolved organi c carbon, and inorganic nutrients on the relationship be tween algae and heterotrophic b acteria in stream periphyton. Hydrobiologia 489:179-194. Roberts, C.R. 2002. Riparian tree associations and storage, tr ansport, and processing of particulate organic m atter in a subtropical stream. PhD Disserta tion, University of Florida, Gainesville,FL. 98 pp.

PAGE 198

198 Robinson, C.A., T. Thom, and M. Lucas. 2000. Ranging behaviour of a large freshw ater invertebrate, the white-clawed crayfish Austropotamobius pallipes Freshwater Biology 44:509 521 Roitberg, B. D. and M. Mangel. 1997. Individual s on the landscape: behavior can m itigate landscape differences among habitats. Oikos 80:234-240. Rooke, J.B. 1984. The invertebrate fauna of four m acrophytes in a lotic system. Freshwater Biology 14:507. Rooke, J.B. 1986. Macroinvertebrates associated w ith m acrophytes and plastic imitations in the Erasoma River, Ontario, Canada. Archiv fur Hydrobiologie 106:307. Rosemond, A.D., S.R. Reice, J.W. Elwood, and P.J. Mulholland. 1992. The effects of stream acidity on benthic invertebrate communities in the south-eastern United States Freshwater Biology 27:193-209. Rossi, L. 1985. Interactions between invertebrate s and m icrofungi in freshwater ecosystems. Oikos 44:175. Roth, N.E., J.D.Allan, and D. L. Erickson. 1996. Lands cape influences on stream biotic integrity assessed at multiple spatial sc ales Landscape Ecology 11:141-156. Rounick, J.S. and M. J. Winterbourn. 1983. The fo r mation, structure and utilization of stone surface organic layers in two New Zeal and streams Freshwater Biology 13:57. Russell, R. E., R. K. Swihart, and Z. Feng. 2003. Population consequences of foraging decisions in a patchy m atrix. Oikos 103:142. Sagova-Mareckova, M. and J. Kvet. 2002. Performa nce of Sparganium emersum Rehm. shoots in response to sediment quality. Hydrobiologia 479:131-141. Sand-Jensen, K. 1998. Influence of submerged m acrophytes o n sediment composition and nearbed flow in lowland streams. Freshwater Biology. 39:663-679. Sand-Jensen, K. and O. Pedersen. 1999. Velo city gradients and turbulence around m acrophyte stands in streams. Fres hwater Biology 42:315-328. Sartory, D.P. and J.E. Grobbelaar. 1984. Extr action of chlorophyll a from freshwater phytoplankton for spectrophotom etri c analysis. Hydrobiol. 114:177. SAS Institute Inc. 2002. SAS OnlineDoc 9.1.3, Cary, NC: SAS Institute Inc. Schneider, D.W. and J. Lyons. 1993. Dynamics of upstream migration in tw o species of tropical freshwater snails Journal of the North American Benthological Society 12:3-16.

PAGE 199

199 Schultz, G.B. and E.E. Crone. 2001. Edge-mediated dispersal behavior in a prairie butterfly. Ecology 82:1879. Shofner, M.A. 1999. Predation, habitat patchiness and prey exchange: interactions between stream meiofauna and juvenile fish. Ph.D. disse rtation, University of Maryland, College Park, Maryland. Sidle, R.C., Y. Tsuboyama, S. Noguchi, I. Hosoda, M. Fujieda, and T. Shimizu. 2000. Stream flow generation in steep headwaters: A linked hydro-geomorphic paradigm. Hydrological Processes. 14:369. Sih, A., B. G. Jonsson, and G. Juikart. 2000. Hab itat loss: ecological, ev olutionary and genetic consequences. Trends in Ecology and Evolution 15:132. Siler, E.R., J. B. Wallace, and S.L. Eggert. 2001. Long-term effects of resource limitation on stream invertebrate drift. Can. J. Fish. Aquat. Sci. 58:1624-1637. Silver, P., J.K. Cooper, M.A. Palmer, and E.J. Davis. 2000. The arrangem ent of resources in patchy landscapes: effects on dist ribution, survival, and resource acquisition of chironomids. Oecologia 124:216-224. Silver, P., C.B. McCall, and D. Wooster. 2004a Habitat partitioning by chironom id larvae in arrays of leaf patches in streams. Journal of the North American Be nthological Society 23:467279. Silver, P., D. Wooster, and M.A. Palmer. 2004b. Chironom id response to spatially structured, dynamic streambed landscapes. Journal of the North American Benthol ogical Society 23:69-77. Simonsen, J. F., and P. Harremoes. 1978. Oxygen and Ph fluctuations in rivers. W ater Research 12:477-489. Sliva, L. and D. D. Williams. 2001. Buffer Zone versus W hole Catchment Approaches to Studying Land Use Impact on River Water Qu ality. Water Research 35:3462-3472. Smock, L.A., G. M. Metzler, and J. E. Gladden. 1985. Role of debr is dam s in the structure and functioning of low-gradient h eadwater streams. Ecology 70:764-775. Snook, D. L. and A.M. Milner. 2002. Biological traits of m acroinvert ebrates and hydraulic conditions in a glacier-fed catchment (Frenc h Pyrnes) Archiv fr Hydrobiologie 153:245-271. Sderstrm,O. 1987. Upstream movements of invert ebrates in running water a review. Archiv fr Hydrobiologie 111:197. Sousa, W.P. 1984. The role of disturbance in natural comm unities. Annual Review of Ecology and Systematics 15:353-391.

PAGE 200

200 Southeast Regional Climate Center. Bai nbridge, Georgia Clim ate Information. http://climate.engr.uga.edu/bainbridge /index.h tml, last accessed August 8, 2007. Southwood, T.R.E. 1977. Habitat, the templet fo r ecological strategies. Journal of Anim al Ecology 46:337. Southwood, T.R.E. 1988. Tactics, strategies and templets. Oikos 52:3. Stamps, J.A., M. Buechner, and V. Krishnan. 1987. The effects of edge permeability and habitat geom etry on emigration from patches of habitat. American Naturalist 129:533. Stanley E.H., S.G. Fisher, and N.B. Grimm. 1997. Ecosystem expansion and contraction in streams. BioScience 47:427-435. Statzner B., B. Bis, S. Doledec, and P. Usse glio-Polatera. 2001. Perspectives for biomonitoring at large spatial scales : a unified measure fo r the functional compos ition of invertebrate communities in European running waters. Basic and Applied Ecology 2:73. Statzner B., S. Doledec, and B. Hugueny. 2004. Bi ological trait com position of European stream invertebrate communities: assessi ng the effects of various trai t filter types. Ecography 27: 470 488. Statzner B., P. Bady, S. Doledec, and F. Scholl. 2005. Invertebrate traits for the biomonitoring of large Eu ropean rivers: an initial assessment of trait patterns in least impacted river reaches. Freshwater Biology 50:2136. Stein, R. A. and J. J. Magnuson. 1976. Behavioral Response of Crayfish to a Fish Predator. Ecology 57:751-761. Stevens R. D., S.B. Cox, R.E. Strauss, and M.R. Willig. 2003. Patterns of functional diversity across an ex tensive environmental gradient: ve rtebrate consumers, hidden treatments and latitudinal trends. Ecology Letters 6:1099. Stone, M. K, and J. B. Wallace. 1998. Long-term recovery of a m ountain stream from clearcut logging: the effects of forest su ccession on benthic invertebrate co mmunity structure. Freshwater Biology 39:151. Stoddard, J.L., D.V. Peck, A.R. Olsen, D.P. Larse n, J. Van Sickle, C.P. Hawkins, R.M. Hughes, T.R. W hittier, G. Lomnicky, A.T. Herlihy, P.R. Kaufmann, S.A. Peterson, P.L. Ringold, S.G. Paulsen, R. Blair. 2005. Environmental monitori ng and assessment program: Western streams and rivers statistical summary. 1762 pp. U.S. En vironmental Protection Agency, Office of Research and Development, Washington, DC. EPA Stout, J., W. H. Taft, R. W. Merritt. 1985. Patte rns of m acroinvertebrate colonization on fresh and senescent alder leaves in two Mich igan streams. Fres hwater Biology 15:573.

PAGE 201

201 Stout, B. M., E. F. Benfield, And J. B. Webste r. 1993. Effects of a forest disturbance on shredder production in southern Appalachian head water stream s. Freshwater Biology 29:59. Strommer, J.L. and L.A. Smock. 1989. Vertical distribution and abundances of invertebrates within the sandy substrate of a low-gradient headwater stream. Freshwater Biology 22:263-274. Suberkropp, K. 1992. Interactions with invertebrates. In F. Barlocher, editor. The ecology of aquatic Hyphom ycetes. Pp.113. Springe r-Verlag, New Yo rk, New York, USA. Suberkropp, K. 1998. Effect of dissolved nutrien ts on two aquatic hyphom ycetes growing on leaf litter Mycological Research 102:998-1002. Sugihara, G. and R. May. 1990. Nonlinear forecas ting as a way of distinguishing chaos from measurement error in a data series. Nature 344:734. Summer W.B., C.R. Jackson, D.G. Jones, a nd M. Miwa. 2003. Characterization of hydrologic and sedim ent transport behavior of forested headwater streams in southwest Georgia. In : Proceedings of the 2003 Georgia Water Resour ces Conference, pp 157-160. Athens, GA. 23-24 April 2003. The Institute of Ecology: The University of Georgia, Athens, GA. Swank, W.T., J.M. Vose, and K.J. Elliot 2001. L ong-term hydrologic and water quality responses following commercial clearcutting of mixed hardwoods on a southern Appalachian catchment. Forest Ecology and Management 143:163. Sweeney, B.W. and R. L. Vannote. 1986. Growth and production of stream stonefly: Influences of diet and tem perature. Ecology 67:1396-1410. Swift, L.W. Jr and J.B. Messer. 1971. Forest cuttings ra is e temperatures of small streams in the southern Appalachians. Journal of Soil and Water Conservation 26:111-116. Taylor, B.R., D. Parkinson, and W. F. J. Parsons 1989. Nitrogen and lignin content as predictors of litter decay rates : A microcosm test. Ecology 70:97-104. Tett, P., C. Gallegos, M. G. Kelly, G. M. Hornberger, and B. J. Cosby. 1978. Relationships am ong Substrate, Flow, and Benthi c Microalgal Pigment Density in Mechums River, Virginia. Limnology and Oceanography 23:785-797. Tischendorf, L. and L. Fahrig. 2000. How should we m easure la ndscape connectivity? Landscape Ecology 15: 633. Tokeshi,M. and L.C.V. Pinder. 1985. Microhabita ts of stream inverteb rates on two subm erged macrophytes with contrasting leaf morphology. Holarctic Ecology 8: 313.

PAGE 202

202 Tolonen, K.T., H. Hmlinen, I.J. Holopain en, K. Mikkonen, and J. Karjalainen. 2003. Body size and sub strate association of littoral insects in relation to vegetation st ructure. Hydrobiologia 499:179. Towns, D.R. 1985. Limnological characteristics o f a South Australian intermittent stream, Brown Hill Creek. Australian Journal of Marine and Freshwater Research 36:821. Towns, D.R. 1991. Ecology of leptocerid caddisfly larvae in an interm ittent South Australian stream receiving Eucalyptus litter. Freshwater Biology 25:117. Townsend, C.R. 1989. The patch dynamics concept of stream comm unity ecology. Journal of the North American Benthological Society 8:36. Townsend, C.R. and A.G. Hildrew A.G. 1994. Species traits in relation to a habitat tem plet for river systems. Freshwater Biology 31:265. Townsend, C.R., S. Doledec, and M. Scarsbrook. 1997. Species traits in re lation to temporal and spatial heterogeneity in stream s: a test of th e habitat templet theory. Freshwater Biology 37:367 387. United States Department of Agriculture (USDA). 1939. Soil Survey of Decatu r County. United States Department of Agricultur e, Soil Conservation S ervice, Albany, GA. United States Environmental Protection Agency. 1999. Update of a mbient water quality criteria for ammonia. EPA 822-R-99-014. Office of Wate r, US Environmental Protection Agency, Washington, DC. United States Environmental Protection Agen cy. 2000. National water quality inventory: 1998 report to Congress. EPA 841-R-00-001. W ashi ngton, D.C.: U.S. Environmental Protection Agency. United States Environmental Protection Agen cy. 2003. Nonpoint-source pollution: T he nations largest water quality problem. Washington, D.C. : U.S. Environmental Protection Agency, Office of Water, Nonpoint Source Control Bran ch. www.epa.gov/owow/nps/facts/point1.htm. Vannote, R. L., G. W. Minshall, K. W. Cummins J. R. Sedell, and C. E. Cushing. 1980. River Continuum Concept. Canadian Journal of Fisheries and Aquatic Sciences 37:130-137. Viera, N.K.M, N.L. Poff, D.M. Carlisle, S.R. Moulton, M.L. Koski, and B.C. Kondratieff. 2006. A Database of Lotic Inv ertebrate Traits for Nort h America. U.S. Geological Survey Data Series 187. US Geological Survey, US Department of th e Interior, Reston, Virg inia. (Available from: http://pubs.usgs.gov/ds/ds187/) Vowell, J.L. Using stream bioassessment to mo nitor b est management practice effectiveness. Forest Ecology and Management 143:237.

PAGE 203

203 Wallace J.B. and M.E. Gurtz. 1986. Response of Baetis m ayflies (Ephemeroptera) to catchment logging. American Midla nd Naturalist 115:25-41. Wallace, J. B., M. R.Whiles, S. Eggert, T. F. Cuffney, G. J. Lugthart, an d K. Chung. 1995. Longterm dynamics of coarse particulate organi c matter in three sma ll Appalachian Mountain streams. Journal of the North Am erican Benthological Society 14:217. Wallace, J. B., S. L. Eggert, J. L. Meyer, and J. R. W ebster.1999 Effects of resource limitation on a detrital-based ecosy stem Ecological Monographs 69:409. Ward, J. V., G. Bretschko, M. Brunke, D. Danielopol, J. Gibert, T. Gonser, and A.G. Hildrew. 1998. The boundaries of river systems: the m etazoan perspective. Freshwater Biology 40:531-569 Waters, T. F. 1972. Drift of stream insect s. Annual Review of Entom ology 17:253-278. Waters, T. F. 1995. Sediment in streams. Source s, biological effects, and control. Monograph 7. Am erican Fisheries Society, Bethesda, Maryland. Webster, J. R. and J. B. Waide. 1982. Effects of forest clearcu tting on leaf breakdown in a southern Appalachian stream Freshwater Biology 12:331. Webster, J. R. and E. F. Benfield. 1986. Vascular plant breakdown in freshwater ecosystem s. Annual Review of Ecology and Systematics 17:567. Webster, J.R., S. W. Golladay, E. F. Benfield, D. J. D' Angelo, and G. T. Peters.1990. Effects of forest disturbance on particulate organic matter budgets of small streams. Journal of the North American Benthological Society 9:120-140. Webster, J. R., S. W. Golladay, E. F. Benfiel d, J. L. Meyer, W. T. Swank, AND J. B. W allace. 1992. Catchment disturbance and stream responses: an overview of stream research at Coweeta Hydrologic Laboratory. Pages 231 In P. J. Boon, P. Calow, and G. E. Petts (editors). River conservation and management. John W iley and Sons Ltd., Chichester, UK. West, B. 2002. Water quality in the South. In Southern Forest Resource Assessm ent, 455-477. General Tech. Report SRS-53. D. N. Wear and J. Greis, eds. Asheville, N.C.: USDA Forest Service, Southern Research Station. Wharton, C.H. 1978. The Natural Environments of Georgia. Georgia Departm ent of Natural Resources. Wiegand, T., K.A. Moloney, and N. J, Kn auer. 1999. Finding th e m issing link between landscape structure and population dynamics: a spatially explic it perspective. American Naturalist 154:605. Wetherald R.T. and S. Manabe. 2002. Simulation of hydrologic changes associated

PAGE 204

204 with global warming. Journal of Geophysical Research 107:4379-4394. Wiens, J. A. 1976. Population responses to patchy environm ents. Annual Review of Ecology and Systematics 7:81-120. Wiens, J. A., J. T. Rotenberry, and B. Vanhorne 1985. Territory size variations in shrubsteppe birds. Auk 102:500-505. Wiens, J., J. Addicott, T. J. Case, and J. Diam ond. 1986. Overview: The importance of spatial and temporal scale in ecologica l investigations. Pages 145-153 In J. Diamond and T. J. Case (editors). Wiens, J. A., N. C. Stenseth, B. Vanhorne, and R. A. Ims. 1993. Ecological mechanisms and landscape ecology. Oikos 66:369-380. Wiens, J. A., R. L. Schooley, and R. D. Week s. 1997. Patchy landscapes and anim al movements: Do beetles percolate? Oikos 78:257-264. Wiggins, G.B. 1996. Larvae of the north americ an caddisfly genera. 457 pp. University of Toronto Press. Wiggins B., R.J. Mackay, and I.M. Smith. 1980. Evolutionary and ecolo gical strategies of anim als in annual temporary pools. Archiv fur Hydrobiologie Supplement 58:97-206. Wilcove, D. S. 1985. Nest predation in forest tracts and the decline of m igratory songbirds. Ecology 66:1211-1214. Williams, D.D. 1987. The Ecology of Temporary Waters. Timber Press, P ortland, Oregon. Williams, D.D. 1996. Environmental constraint s in tem porary fresh waters and their consequences for the insect fauna. Journal of the North American Benthological Society 15:634650. Williams, T. M., D. D. Hook, D. J. Lipscomb, X. Zeng, and J. W Albiston. 1999. Effectiveness of best management practices to protect water qu ality in the South Carolina Piedmont. In Proc. Tenth Biennial Southern Silvicultural Research Conference, 357-362. General Tech. Report SRS-30. T. A. Waldrop, ed. Ashevi lle, N.C.: USDA Forest Service, Southern Research Station. Wipfli, M.S. and D.P. Gregovich. 2002. Export of invertebrates and detritus from fishless headwater streams in southeas tern Alaska: implications for downstream salmonid production. Freshwater Biology 47:957-969. Wissinger, S.A. 1997. Cyclic colonization in pred ictably ephem eral habitats: a template for biological control in annual crop systems. Biologi cal Control 10:4.

PAGE 205

205 With, K.A. and T.O. Crist. 1996. Translating acro ss scales: simulating species distributions as the aggregate response of individuals to heterogeneity. Ecolog ical Modelling 93:125 137. Wood, P. J. and P. D. Armitage. 1997. Biologica l effects of fine sedim ent in the lotic environment. Environmental Management 21:203. Wootton, J.T.and M.E. Power. 1993. Productivity, consumers, and the structure of a river food chain. Proceedings of the Nationa l Academ y of Sciences 90:1384. Wright J.F., R.J.M.Gunn, J.M.Winder, J.H. Blackburn, and R. W iggers. 2001. The response of chalk stream invertebrates to a prolonged drought: the value of a long-term dataset. Verhandlungen der Internationalen Vereinigung fur Theoretische und Angewandte Limnologie 27:912. Yamamura, K., M. Kishita, N. Arakaki, F. Kawamura, and Y. Sadoyama. 2003. Estimation of dispersal distance by m ark-recapture experiments us ing traps: correction of bias caused by the artificial removal by traps. Population Ecology 45:149-155. Zah, R., P. Burgherr, S. M. Bernasconi, and U. Uehlinger. 2001. Stable isotope analysis of m acroinvertebrates and their food sources in a glacier stream Freshwater Biology 46:871. Zobel, M. 1997. The relative roles of species po ols in determ ining plant species richness: an alternative explanation of species coexistenc e? Trends in Ecology and Evolution 12:266-269. Zollner, P.A. and S.L. Lima. 1999. Search strategi es for landscape-level interpatch movem ents. Ecology 80:1019. Zollner, P.A. 2000. Comparing the landscape level pe rceptual abilities of forest sciurids in fragm ented agricultural la ndscapes, Landscape Ecology 15:523.

PAGE 206

BIOGRAPHICAL SKETCH Marcus Griswold was born in Baltimore, Maryland, on September 30, 1978. He pursued a B.S. in biology at the Univers ity of Maryland at College Park. His m asters work took him to the University of Florida to work on predato r-prey dynamics of larval mosquitoes under the direction of Phil Lounibos. His interest in aquatic ecology and background in Entomology led him to Thomas Crisman to pursue a PhD in envi ronmental engineering sciences, with a focus on riparian zone management in aquatic ecosystems. During this time, his work was funded by the U.S. EPA, Sigma Xi, and the Friends of the Osa. He has worked in a variety of stream systems in the southeastern U.S. and Costa Rica, from prim ary tropical forests to degraded urban streams. His goal is to utilize his knowledge of aquatic stressors to properly manage aquatic ecosystems, balancing human needs and maintenance of ecosystem function and biodiversity.