Standardized sampling of wild trout streams

VIti
Nzoo.G2-
MI
!C}97
,773

Standardized Sampling of Wild Trout Streams

by Jeffrey P. Durniak, Lee C. Keefer, and W. Ralph Ruddell
Georgia Department of Natural Resources Wildlife Resources Division Social Circle, Georgia

September 1997

This study was funded through the Federal Aid in Sport Fish Restoration Act under Dingell-Johnson Project F-25, Georgia.

FINAL REPORT

State: Georgia

Project Number: F-25-24

Project Title: Northern Region Fisheries Investigations

Study XXVII Title: Standardized Sampling of Wild Trout Streams

Period Covered: 1 July 1990 - 30 June 1997

Study Objective:

To establish and refine a long-term fish population monitoring program on selected wild trout streams to understand normal population variability and to detect population response to changes in habitat conditions and/or fisheries management strategies.

ABSTRACT
During 1991-95 fish populations were monitored in 27 identified stream reaches on 19 streams using depletion electrofishing techniques. Habitat, substrate and temperature variables were also estimated for each stream reach. Streams were grouped by the predominant trout species present. Comparisons were made among years on the same stream, and among streams for fish population differences within a group over the five-year study period. Comparisons were also made for relationships between various habitat and substrate variables and fish abundance using stepwise regression.
Little year-to-year variation was noted, but many differences among streams were observed. Statistically significant relationships were found between fish abundance and several habitat/substrate variables. Negative relationships were observed between combined sand and sediment and rainbow trout density, and positive relationships were observed between large woody debris (LWD) and the density of brook and brown trout. A significant relationship was observed between YOY abundance and the abundance of adult trout the following year for brown trout, but not for brook or rainbow trout.
A short term creel survey was conducted on one of the study streams and results were comparable to those seen on other southeastern trout streams.

1

INTRODUCTION Georgia's trout fishery supports several million trips annually. The fishery is restricted to higher elevation streams in the northern one-third of the state, where colder water temperatures allow year-round trout survival. While most Georgia trout anglers concentrate on stocked streams and lakes, a significant amount of effort is placed on wild rainbow (Onchorynchus mykiss), brook (Salvelinus fontinalis), and brown trout (Salmo trutta) fisheries. Wild trout biomass averages 10-30 kg/ha (Van Kirk 1969; Michaels 1974; England 1979; Fowler 1985; Dumiak and England 1986; Dumiak 1989), which is extremely low when compared with estimates from more fertile North American streams in the Northeastern and Western U.S., but is typical of the higher gradient freestone streams in the Southern Appalachian region (Habera and Strange 1993). The low natural productivity (total alkalinity < 10 mg/l) limits trout production, while the variable flow regime produces significant instability in annual trout abundance (England 1979; Dumiak and England 1986). Rainbow trout is the most common species inhabiting Georgia trout streams. Brown trout are commonly found in relatively low abundance, coexisting with rainbows in many streams, especially at lower elevations. Brook trout are generally confined to high elevation headwaters, upstream from barriers such as waterfalls that prevent encroachment by other fish species. An estimated 60 streams totaling 136 km contain wild brook trout. The north Georgia mountains face increasing pressure from human activity such as road construction, timber harvest, and residential development. Along with more
2

potential threats to their habitat, wild trout populations will also likely experience more fishing pressure from a rising number of anglers. Due to the popularity and the fragile nature of trout streams, there is considerable cause for concern.
Standardized fisheries sampling has been conducted by the Georgia Department of Natural Resources (GADNR) on most major reservoirs in Georgia since 1982 and has recently been expanded to include several river systems in south Georgia. This annual monitoring has allowed timely evaluation of fish population responses to environmental alteration, changes in fishing pressure, and specific fisheries management actions such as stocking or regulation change.
The contribution oflong-term studies to the understanding of trout population dynamics and subsequent optimum management strategies by biologists has been significant (McFadden et al. 1967; Seegrist and Gard 1972; Hunt 1974; Egglishaw and Shackley 1977; Waters 1983). However, no long-term research had been done in Georgia prior to the present study. Past studies of Georgia wild trout populations generally spanned only two to three years on an individual stream, were often non-quantitative, and addressed very specific management concerns. Recent evaluations of habitat conditions and current estimates of trout fishing pressure, harvest, and catch from wild trout streams were nonexistent. A monitoring program was needed to develop sound baseline fisheries and habitat data, characterize "normal"population response to environmental factors, and then to quantify positive or negative effects of man's activity on wild trout streams in Georgia. This study summarizes efforts made to establish and refine a long-term monitoring program for wild trout streams in northeast Georgia.
3

METHODS Stream and Site Selection
Existing file data, published and unpublished reports, and the personal knowledge of GADNR and U.S. Forest Service (USFS) fisheries biologists were used to select trout streams for study (Figure 1). The selected streams were a diverse cross-section of Georgia trout streams with a wide variation in resident trout species, trout population density, stream habitat and watershed conditions, angler interest, and existing angling regulations (Table 1). All streams were located on the Chattahoochee National Forest except for Dover and Metcalf creeks. Forest cover was generally second growth, mixed pine-hardwood in all watersheds. Unless noted otherwise, the streams had not been stocked for at least 20 years and were covered by the general trout fishing regulations: a creel limit of eight trout per day with no length restrictions, no bait restrictions other than the prohibition of live bait fish, and an open season extending from the last Saturday in March through October 31.
An attempt was made to place fish sampling locations near the upstream and downstream ends of the selected streams. The upper end was typically defined as the upstream segment of second order waters. The lower end was usually defined by USFS property boundaries and/or sampling goals or logistics considerations. Fish sampling stations were about 100 m in length, actual length depending on the habitat breaks (usually riffles) that provided appropriate locations for block net placement. This sample length usually provided for several riffle/pool sequences to be repeated within a site.
4

Fannin
-"'-----
.~
11

---T
\~ Towns

12 >
f 14 \ .jV)

Union
4
/-3

.I I,

'" \ 2 (

,,~:1

\
\

~

8",
Rabun <"r
J/
13~/
)

J

6
/
7-

/1 ~V\(
f

White \

9
\

\

(J"l

N

1. Charlies Creek

Lumpkin

-)

2. Chattahoochee River 3. Cooper Creek

\ \~~

4. Board Camp Creek

5. Dicks Creek

6. Dukes Creek

7. Dover Creek

8. Hoods Creek

9. Jones Creek

10. Moccasin Creek

11. Noontootla Creek

12. Reed Creek

13. Stonewall Creek

14. Totterypole Creek

Figure 1. Northesast Georgia study area showing location of trout stream segments sampled 1991-1995. Lines define county boundaries.

Table 1. Northeast Georgia trout streams included in the standardized sampling program, 1991-1995. Tributaries are denoted by an indentation.

Stream Name
Charlies Creek
Chattahoochee River Coopers Creek - Board Camp Creek Dicks Creek Dukes Creek - Bear Den Creek - Dover Creek Hoods Creek Jones Creek Moccasin Creek - North Fork - South Fork Noontootla Creek
- Chester Creek
Reed Creek - Hedden Creek - Ridley Creek Stonewall Creek Totterypole Creek

River System

Trout Species County Present 1

Angling Regulations

Tallulah

Towns

RB

General2 & Year Round Season

Chattahoochee Union

BK

General

Toccoa

Union

RB,BN BK

General General

Tallulah

Rabun

RB,BN,BK

General

Chattahoochee White

RB RB RB,BN

General General No harvesf

Chattooga

Rabun

BN,RB

Artificials only4

Etowah

Lumpkin BN

Artificials only

Tallulah

Rabun

RB,BN BK BN

General General General

Toccoa

Fannin

RB,BN

RB,BK

Artificials only, 406 mm minimum length limit Artificials only, 406 mm minimum length limit

Chattooga

Rabun

BN BN,BK BN,BK

General General General

Tallulah

Rabun

RB,BN

General

Chatooga

Rabun

BK

General

ITrout species are brook (BK), brown (BN), and rainbow trout (RB). 2General trout fishing regulations are eight fish creel limit, no size limit, natural baits (except live minnows) allowed, and season length from last Saturday in March through October.
3Closed to public angling by private landowner.
4Same as "General", except that natural baits are prohibited.

6

Three stations per location were selected to allow for variance estimation. Stations were often contiguous due to logistics; more time could be spent sampling rather than traveling between sites and preparing them for sampling. An attempt was also made to establish separate high-gradient and low-gradient stations within a fish sampling location, where practical, by visually assessing the differences in macrohabitat along each stream segment and determining appropriate cutoffs between stations. Brook Trout Streams
Four streams that historically supported allopatric brook trout populations were included in this study. Brief descriptions follow.
Board Camp Creek is a second order tributary to Logan Creek, which empties into Cooper Creek on the Cooper Creek Wildlife Management Area (WMA). This stream was selected because it contained an abundant wild brook trout population in the Toccoa River drainage, on the western side of the fisheries management region. A waterfall on Logan Creek prevents upstream colonization by other salmonids. Board Camp Creek is accessible to anglers only by foot since the unnumbered Forest Service access road to that drainage is closed to public use except during eight days of managed deer hunts, which are held after the close of trout season. Most fishing pressure on Board Camp Creek is believed to be from local anglers willing to walk a minimum of 4.2 km from the gate at Forest Service Road (FS) 39 to reach the creek. There are no known stocking records for Board Camp Creek.
Three contiguous 100 m fish sampling stations, beginning 0.6 km upstream from the confluence with Logan Creek, were selected on Board Camp Creek. The downstream
7

station was a higher gradient, bedrock dominated section, while the two upstream stations represented a gentler sloping and meandering reach.
The Chattahoochee River (headwaters) was selected because of its south-central location in the region and because of the existence of past sampling data (England 1979). Brook trout have existed in the uppermost 5.0 km of the Chattahoochee River, a third order stream above a barrier waterfall, since a renovation project in 1973. Rainbow trout were killed with rotenone and wild brook trout, collected via electrofishing from nearby streams, were restocked as a brood source. There has been no known stocking in this stream since this introduction. Road FS44B crosses the stream approximately 300 m above the barrier falls and dead-ends in a 0.5 ha wildlife opening maintained by GADNR as part of Chattahoochee WMA operations. Above the opening, the old, revegetated road bed continues to parallel the creek to its headwaters and apparently is used by hikers. The road is closed to public vehicles about 1.7 km downstream from the wildlife opening, at the far end of a popular, primitive campground maintained by the USFS.
Six sampling stations were established on the Chattahoochee River. Station one began approximately 10m upstream from the barrier falls at the lower end of brook trout water and covered a higher gradient reach with abundant bedrock. Stations two and three were contiguous, beginning about 100 m upstream from the upper end of station one and ending just upstream of the access road ford. They were lower gradient, meandering stream sections. Stations 1-3 are referred to as the lower Chattahoochee River. Stations four and five were contiguous and began about 300 m upstream from station three. Station six started approximately 100 m upstream from station five and was an old
8

9

limited population of wild brook trout. Before 1963, hatchery-reared fingerling brook trout were occasionally stocked by GADNR to bolster existing populations. During 1986, 199 wild brook trout fingerlings were collected from six other northeast Georgia streams and stocked into the Totterypole watershed to assess growth, survival, and effects on spawning/recruitment success (Durniak and Ruddell 1990). A system of unpaved USFS and county roads crosses the two streams in several locations. The stream was also being studied in a concurrent research project on the West Fork basin by the U.S. Forest Service Southeastern Experiment Station (Dr. C. Andrew Dolloff, pers. comm.).
Six fish sampling stations were established in the Totterypole watershed. Stations 1-3 were contiguous and began at the culvert on FS86. All three seemed similar in their habitat features. Station one was a former station ofDurniak and Ruddell (1990). Stations 1-3 are referred to as lower Totterypole creek. Stations 4-6 were located on Metcalf Creek and were reached via a small inholding of private property on the downstream side ofFS7, Hale Ridge Road. Station four, another sampling site from Dumiak and Ruddell (1990), started about 500 m downstream from the cabin at the end of the private access road and ended at a small barrier waterfall. Station five began 200 m upstream from station four, just above a very high gradient, bedrock-dominated reach, and covered the transition between steep and flat gradients. Station six began 300 m upstream from station five and covered a section of fairly flat, meandering stream. For reporting purposes, Metcalf Creek will be called upper Totterypole Creek.
10

Brown Trout Streams Three streams with salmonid populations dominated by brown trout were
incorporated into this study. Jones Creek is a third order tributai-y to the Etowah River on the Blue Ridge WMA in Lumpkin County. It supports a popular fishery for wild brown trout, and anglers are restricted to the use of artificial lures only. The last two stockings of Jones Creek occurred in 1982 (20 brook trout broodstock) and 1973 (1,000 brown trout fingerlings). Selected reaches have been used as either test or control sites for instream habitat improvements by USFS (Seehom 1992), with Trout Unlimited (TU) cooperating (M. Seehom, USFS, pers. comm.). It was selected because of its popularity with anglers, the availability of past sampling data (Van Kirk 1969a, England 1978, USFS unpublished file data), and the interest in maintaining annual monitoring of the habitat improvement sites. Angler access to Jones Creek is usually via a 0.3 -1.0 km hike from system roads FS77 and FS77A.
Four sample stations were selected on Jones Creek. Station one started approximately 1.6 km upstream from the USDA Natural Resources Conservation Service dry dam and represented a wide, meandering channel that appeared to contain natural trout habitat as pools, undercut banks, and instream woody debris. Station two started about 1.0 km upstream from station one, was wide and gently sloping, and contained lower quality habitat for adult trout than station one. It was a site planned for instream habitat improvement (M. Seehom, pers. comm.). Station three began about 300 m upstream from station two and represented a narrower, higher gradient reach with channel features heavily influenced by bedrock. Stations 1-3 are referred to as lower Jones Creek.
11

Station four (upper Jones Creek) started about 1.9 km above station three and was a USFS monitoring site for instream habitat improvement work. Channel modifications (cover logs, wedge dams, deflectors, and constrictors) were abundant throughout this station.
Reed Creek is a third order tributary to the Chattooga River in the extreme northeast comer of Georgia. It supports a naturalized brown trout population throughout most of the stream, with wild brook trout present in the extreme headwaters. Records show that the stream has not been stocked in at least 21 years. Fishing pressure seems light to moderate due to the relative remoteness of the watershed. Hedden Creek (third order) and Ridley Branch (first order) merge just upstream from Burrells Ford Road (FS646) to form Reed Creek. Hedden Creek is wider, steeper and features more bedrock, while Ridley Branch is smaller, more gently sloping, and contains abundant natural trout cover as undercut banks and instream woody debris jams. Hedden and Reed creeks carry heavy sediment loads which likely result from unpaved county and USFS roads in the watershed. Cooperative efforts among USFS, TU, and GADNR to control sediment inputs, install instream habitat improvements, and monitor fish populations were started in 1986. Reed Creek was chosen for monitoring because it represented a headwater brown trout stream in the northeastern part of the region, supported considerable interest by resource agencies and cooperators, and had been previously sampled.
One station was selected on each of the two tributaries to continue the monitoring protocol previously established. Both sample stations began at the confluence of Ridley Branch and Hedden Creek and continued upstream for 500 m.
12

The South Fork of Moccasin Creek (previously described) was chosen because it was centrally located, contained brown trout, and offered an opportunity to study all three trout species within one watershed. Instream habitat improvements by USFS and TU were done on part of the creek during 1989-1990, and this study offered a chance to evaluate the effectiveness of that work. Unimproved road FS165, which parallels the middle segment of this second order stream, is gated at FS26 and provides administrative access and an easy angler hike to the stream.
Three stations were established on South Moccasin Creek. The first began approximately 2.0 km downstream from the FS26 road crossing. This site was fairly high gradient and had little stream bank cover. Station two began about 100 m upstream from the first station, in a lower gradient reach, and included a number of USFS habitat structures. Station three started 150 m above station two and included no artificial habitat improvements. Natural cover, as deep pools, undercut banks, and instream woody debris jams, were abundant in this station. Rainbow Trout Streams
Five streams dominated by rainbow trout were monitored. Charlies Creek is a second order tributary to the Tallulah River in Towns County, near the North Carolina border. Catchable trout stocking in this stream was stopped in 1965 and fingerling stockings ceased after 1977. Trout population dynamics (Durniak and England 1986) and spawning activity (Couch 1985) in this stream have been previously investigated. Yearround trout fishing has been allowed on the Tallulah River and Charlies Creek since 1988. A rugged, unpaved county road parallels the lower half of the stream, and access
13

often requires the use of four-wheel drive vehicles. FS72, an unimproved forest road, leads into the headwaters.
Six fish sampling sites were used on Charlies Creek. Stations one and two were contiguous and started approximately 1.6 Ian upstream from the mouth of the creek, about 30 m above a popular primitive camp site along the rough county road. Those sample sites were in lower gradient, meandering reaches with a fair amount of instream woody debris jams. Station three began about 700 m above station two (200 m downstream from the county road ford) and was a swifter, shallower, bedrock-dominated reach. It had been previously sampled by Durniak and England (1986). Stations 1-3 are referred to as lower Charlies Creek. Stations 4-6 were contiguous and were located in the headwaters, approximately 1.7 Ian above station three. Stations four and six were also sampled by Dumiak and England (1986) and appeared to have a greater abundance of pool habitat and low gradient sections when compared to station five, which was dominated by fast, shallow riffles. Stations 4-6 are referred to as upper Charlies Creek.
Dicks Creek is a third order tributary to Lake Burton on the Lake Burton WMA, approximately 3.0 Ian north of Lake Burton Hatchery. FS164 parallels the stream for most of its length. The cut and fill slopes of US Highway 76 impact the intermittent and perennial tributaries flowing from the north. The stream appeared to carry a heavy sediment load, based on initial observation. The headwaters were renovated and restocked with wild brook trout in the early 1970s (England 1979), but creek chubs (Semotilus atromaculatus) survived the renovation and competed with the brook trout. The lower half of the stream supports a wild rainbow trout fishery, with the occasional
14

catch of brown trout reported. Hatchery-reared rainbow trout fingerlings were stocked in 1977, 1984, and 1986. The 1986 stocking of 2,025 fish was an experiment to improve depressed trout populations in the upper creek (Durniak and Ruddell 1990). Dicks Creek was selected for comparison to other streams, such as Charlies Creek, thought to have better habitat quality and higher trout densities.
Six fish sampling sites were located on Dicks Creek. Station one began approximately 1.8 km upstream from GA Route 197 and covered a wide, low gradient reach. Stations two and three were contiguous and started 200 m upstream from station one. Station three covered a higher gradient reach with large boulders and bedrock shaping the channel. Stations 1-3 are referred to as lower Dicks Creek. Headwater stations 4-6 were contiguous and started 50 m downstream from the mouth of Shook Branch. All three were moderate gradient reaches with limited cover for adult trout. Station six was a sample site used by Durniak and Ruddell (1990). Stations 4-6 are referred to as upper Dicks Creek.
Dukes Creek, a fourth order stream that feeds the Chattahoochee River in White County, held an abundant wild rainbow trout population in the main stream above Dukes Creek Falls and supported wild brook trout in several of its tributaries. The rainbow population had been studied previously (Couch 1985, Durniak and England 1986), while wild brook trout were successfully established in Bear Den Creek, one of the two headwater tributaries (Durniak and Ruddell 1990). The Richard Russell Scenic Highway (Route 348) bisects Dukes Creek and provides access to the Raven Cliffs Scenic Area, a very popular region for hiking and dispersed camping. System road FS 244 parallels the
15

stream below the highway, while a gated forest road follows Dukes Creek above the highway to its origin, the confluence of Bear Den and Little Low Gap creeks. Above that point, the old, revegetated road bed parallels Bear Den Creek and is traveled by hikers, campers, and anglers.
Six stations were established on Dukes Creek: three each on the main stream and on Bear Den Creek. Station one began about 0.5 km downstream from the FS244 bridge and covered a faster, bedrock-dominated reach. Station two began at the upper end of station one and started a transition into a lower gradient, meandering habitat. The third station began 40 m below the bridge and included a large pool below the bridge and a braided channel above it. This station was a sample site used by England and Durniak (1986). Stations 1-3 are referred to as lower Dukes creek. Contiguous stations 4-6 began approximately 500 m upstream from the mouth of Bear Den Creek. Site four was a lower gradient, meandering reach, while the two upper stations changed to a swifter, bedrockinfluenced habitat. For reporting purposes, Bear Den Creek will be considered as upper Dukes Creek.
Moccasin Creek has been described under the list of brook trout streams. A gravel road leads from the Lake Burton Hatchery into the WMA and ends next to the creek. From there, an old logging road parallels the main stream and is a very popular hiking trail. Three contiguous stations were set up on the main stream, starting approximately 1.2 km upstream from the hatchery intake. This stream reach features a deeply incised channel that runs through a small gorge. All three sample sites were in a swift, deep habitat dominated by bedrock ledges and large boulders.
16

Stonewall Creek is a third order tributary to Tiger Creek, in the Tallulah River watershed in Rabun County. The watershed is primarily national forest land, but a significant portion of the extreme headwaters is privately owned, with an orchard and some limited residential development being the primary land uses. The stream appears to carry a heavy sediment load, and the trout fishery has traditionally been considered only fair, at best. Records show only one stocking in the last 20 years: 6,000 rainbow trout fingerlings in 1984. System road FS20 bisects the stream and provides access for a variety of outdoor recreationists. Many old, revegetated logging roads parallel the streams in this watershed and provide more access via foot travel. Anglers and conservation rangers report that wild rainbow trout are most commonly harvested, with brown trout caught occasionally. Like Dicks Creek, this stream was selected for study to compare with those streams supporting higher trout populations.
Six stations were established on Stonewall Creek. Contiguous stations 1-3 began 400 m downstream of the FS20 ford. The first station covered a series of bedrock ledges and the pools between them. Station two began at the top of the bedrock ledges and changed to lower gradient and a meandering channel. Station three was similar in appearance to station two. Stations 1-3 are referred to in this report as lower Stonewall Creek. Contiguous stations 4-6 (upper Stonewall Creek) were sited on smaller waters about 0.5 km upstream from the FS20 ford and were uniform in their habitat features. Gradient appeared moderate, bedrock shaped many channel features, and a heavy sand/silt bedload persisted.
17

Sympatric Trout Streams Cooper Creek is a fifth order tributary to the Toccoa River on the Cooper Creek
WMA in Union and Fannin counties. System roads FS33 and FS4 provide administrative and recreational access along portions of the creek, while the remaining reaches are rugged and remote. This stream is stocked with about 45,000 catchable trout annually, and supports a very popular fishery (Fatora 1983) for both stocked and wild trout. Since the late 1980s, instream habitat improvements have been done by USFS and TU in a 1 km section upstream from the mouth of Burnett Creek. Cooper Creek was selected for study because of its large size, its popularity, and an interest in evaluating habitat improvement efforts.
Two fish sampling stations were initially planned for Cooper Creek. They were on an accessible portion of the stream along FS33, where habitat improvements had occurred. This section of stream was low gradient and meandering. The first station started at the mouth of Bumett Creek and represented an unimproved reach that contained only natural cover. Station two began at the upstream end of station one, a riffle just upstream from an old mill dam site, and covered an "improved" section containing root wads, cover logs, deflectors, and channel constrictors. Station two ended about 200 m downstream from a spur road ford, which was a regular stocking site. Station three, in the Cooper Creek Scenic Area about 3.2 km below the first two stations was added in 1993 after a private inholding was acquired and the access road provided an easy route to this remote stream section. The sampling station started about 50 m upstream from Bryant Creek and ran along the lower half of a 1.5 ha wildlife opening.
18

Dover Creek is a third order tributary that enters Dukes Creek downstream from the study sites on the latter stream. The lower 500 m of Dover Creek flowed through private property (the Smithgall estate), which was acquired by GADNR in September 1995 as part of a statewide land preservation initiative. Dover Creek supported wild rainbow and brown trout through its lower reaches and wild brook trout in the headwaters. The landowner strictly limited angling pressure and harvest on the estate, arid agreed to prohibit both on his section of Dover Creek during this study. Dover Creek was restricted to catch-and-release angling after state acquisition. Supplemental feeding of trout in Dukes Creek was initiated by the landowner in the mid-1980s and maintained by the state after acquisition. This stream was chosen because it offered the best available site to study an unexploited trout population. Three contiguous sample stations were established in the uppermost 300 m on private land, just inside the wire mesh fence at the property line. All three sites appeared uniform, featuring moderate gradients and a mix of channel characteristics with no extreme bedrock dominance. Access was via an old, revegetated logging road that paralleled the creek.
Hoods Creek is a second order tributary to Warwoman Creek and part of the Chattooga River watershed in eastern Rabun County. The stream carries a heavy sand bedload. Hoods Creek and its tributary, Walnut Fork, are managed under an artificial lure only regulation. The stream supports low populations of wild brown and rainbow trout and is known more for the size rather than the number of trout harvested. Attempts to regulate flow in channel braids and to place log cribs full of clean gravel in the main channel apparently failed to enhance trout spawning and recruitment (GADNR,
19

unpublished file data). Log dams to enhance cover for adult trout were installed throughout the stream by GADNR in the late 1960s and early 1970s, and several of those structures are still intact. An old logging road allows vehicle access to just beyond the confluence of the two streams and provides a walking trail along the rest of Hoods Creek. Hoods Creek was chosen for study because of the availability of past sampling data that showed trout population densities much lower than many other streams.
Six fish sampling stations were chosen. Contiguous stations 1-3 began 250 m upstream from the mouth of Walnut Fork and ended just upstream of the first logging road ford. Those stations covered a moderate gradient reach with several large, deep pools created by bedrock ledges and woody debris jams. Contiguous stations 4-6 were sited at the upstream end of the second order reach, 2.1 km above station three, and appeared to cover a uniform section of moderate gradient and heavy bedrock influence.
Noontootla Creek is a third order stream that begins on the Blue Ridge WMA and flows into the Toccoa River in Fannin County. The stream is stocked with 3,500 catchable trout annually at six sites, with the closest site 1.7 km downstream from the WMA boundary. The creek on the WMA is managed under a 40 cm (16 inch) minimum size limit and artificial lures only restrictions (Fatora 1970, England 1978). The watershed on the WMA contains wild rainbow, brown, and brook trout, with the brook trout restricted to the extreme headwaters. It is a relatively large, steep stream with dominant bedrock features. Noontootla Creek was chosen for study because of its popularity and because the fishing regulations essentially result in an unharvested trout population.
20

Six fish sampling stations were established on Noontootla Creek. Water depth limited the use of backpack electrofishing gear over much of the creek's length, so the three stations on the main stream were picked due to their wadeability. The stations were about 2.6,6.2, and 6.5 km, respectively, upstream from the WMA boundary. All three typified the main stream, except they lacked extreme (>1.5 m) pool depth. Stations 1-3 are referred to in this report as lower Noontootla Creek. Contiguous stations 4-6 were on the upper end of Chester Creek, about 3.5 km above station three, and are referred to as upper Noontootla Creek in this report. Stations four and five were gently sloping and meandering, while station six changed to bedrock ledges on the upper third of the sample site. Fish Sampling
Fish sampling was conducted generally between July and October, with emphasis placed on returning to the same sites at about the same time each year. Individually labeled, metal tree tags were nailed to trees at the upstream and downstream ends of each sampling station to ensure that block nets were set at the same locations. The smaller, brook trout streams were sampled first. The larger streams were sampled later in the summer or fall to maximize collection efficiency during the annuallow-tlow period.
Fish were collected with gasoline-powered backpack electrofishing units that generated outputs of 450-600 volts AC. A three-pass removal depletion technique (White et al. 1982) was usually employed. Upstream and downstream ends of each sample site were blocked with small mesh minnow seines (or a natural barrier such as a waterfall) and three consecutive upstream passes were made by the electrofishing crew(s). Effort
21

was held constant among passes. In some sampling efforts on headwater streams, subsequent electrofishing passes were not made when the current pass captured zero or one fish, and no other escaping fish were observed by the electrofishing crew.
An electrofishing crew consisted of the backpack unit operator who also had a dip net as one electrode, an assistant with a dip net to help capture fish, and a bucket carrier. The number of electrofishing crews ranged from one to four, depending on stream width and habitat complexity, which affect capture success (Habera et al. 1992). Personnel from GADNR and the USFS, and volunteer help from interested anglers and TU members, cooperated in the sampling efforts.
Fish caught in each pass were removed from the sample site and recorded separately. Trout were anesthetized, measured to the nearest mm, weighed to the nearest g, and identified as either stocked or wild based on fin erosion, coloration, and known stocking history. Trout age class, either young-of-year (YOY) or adult, was assigned to individuals based on length-frequency histograms. Age class validation was done by marking obvious YOY fish with an adipose fin clip during 1993 samples and examining length-frequencies of recaptures in 1994 samples. Other fish were separated by species, counted, and bulk weighed. Fish were held in live cages outside the sample site until collections were done and were then released into the site.
Station length and width were measured during the first quantitative collection to determine the areas sampled. Length along the thalweg was determined with a tape measure, while stream width was measured at 10 m intervals after randomly picking a
22

starting point within the first 10m. Those initial length and width measurements were used in all subsequent calculations of fish density and biomass.
There were several exceptions to the 100 m sample site standard. Longer sampling sites were established on Cooper, Jones, and Reed creeks to incorporate existing sampling protocols and to obtain more representative reaches. Site lengths of approximately 300 m were established on Cooper Creek, the largest stream, to allow for multiple pool/riffle sequences within a site and to enable site two to contain an entire section ofUSFS habitat improvements (deflectors, cover logs, root wads, etc.). Site lengths of about 150 m were chosen for Jones Creek so that sample sites from an ongoing USFS/TU study on instream habitat improvement effects could be incorporated. In the Reed Creek watershed, the site of another USFS/TU habitat improvement project, the existing 500 m sample sites, one each on Hedden Creek and Ridley Branch, were also retained.
Fish populations were estimated using Microfish 2.2 (Van Deveter and Platts 1985), a maximum-likelihood estimator. Separate estimates were made for each species, with two exceptions. First, separate estimates were made for YOY and adult trout. Second, group estimates were made for closely related nongame species within a family when low sample sizes and non-descending catch patterns prohibited a maximum likelihood estimate by species. The resulting estimate was apportioned to each species based on its contribution to total catch for the group. For example, bluehead iliocomis leptocephalus) and creek chub catches were occasionally combined for a group population estimate. In all remaining cases, total catch was used as the sample population
23

estimate whenever mean capture probability of a species or age group was less than 0.1. Population estimates were multiplied by mean sample weight and divided by sample area to estimate biomass.
Habitat Assessment It was the original intention to survey microhabitat annually within the population
sampling sites. However, the literature showed mixed success of efforts to relate trout density and biomass to intensive microhabitat assessment (Loar et al. 1985; Durniak: and Ruddell 1990). Basinwide visual estimation techniques (BVET) (Hankin and Reeves 1988; Dolloff et al. 1993) appeared promising for rapid assessment of basinwide (overall) stream conditions, which may have a greater effect on trout populations than local (site) conditions. (Lanka et al. 1987; Kozel and Hubert 1989). A modified version of BVET was used to assess basinwide habitat conditions once during the study. Stream reaches containing the fish population sampling stations were assessed using BVET during the summer or fall.
Habitat classification was modeled after Bisson et al. (1981) and Platts et al. (1983). Macrohabitat units were assigned to one of five habitat types: pool, riffle, glide, cascade, or pocket water. Individual pools were also assigned to one of four types: dam, plunge, trench, or lateral scour. Pocket water was defined as a higher gradient riffle with abundant large boulders that created a complex of small dam and plunge pools within the riffle. Substrate was defined as either fine sediment 0.06 mm; referred to as sediment
24

in this text ), sand (0.07- 2 rnm), gravel (0.2 - 6.4 cm), cobble (6.5 - 25 cm), boulder (>26 cm), bedrock, or large woody debris (LWD).
Habitat assessment was done by two people. From a starting point at the downstream end of a macrohabitat unit, one person walked to the upstream end of the unit and measured unit length with a hip chain. The other person measured unit gradient with a clinometer and then waded upstream to his coworker. The two surveyors then agreed on a unit type, pool type, if applicable, and dominant substrate type, based on visual estimation. In those cases where a subdominant substrate type was evident, a weighting of 0.8 (dominant) and 0.2 (subdominant) was assigned to unit substrate classification. Working upstream, this process was repeated for each macrohabitat unit. Stream width was measured every 100 m to calculate an average width per km covered. Dominant landmarks such as tributaries were noted for reference.
Summer water temperatures were monitored in selected streams. Ryan model RTM recorders were programmed for hourly recordings and placed in an accessible site near the most downstream fish sampling station. Units were recovered during the fall. Hourly data were summarized via a computer spreadsheet to provide daily minima, maxima, and means, and a seven-day average of mean daily water temperature. Creel Survey
A non-uniform probability roving creel survey (Malvestuto 1983) was used to assess the trout fishery in Dukes Creek from March 26 through October 31, 1991. The 3.95 km (2.49 ha) segment of Dukes and Bear Den creeks from Dukes Creek Falls upstream, which contained the fish population sample sites, comprised the survey area.
25

Because of traditional heavy use of north Georgia trout streams on weekends (Fatora 1983, Durniak 1989), the survey was designed to emphasize weekend sampling. The survey duration was divided into eight four-week periods, with work weeks beginning Tuesday and ending Monday. A calendar day was defmed as two hours after sunrise to one hour after sunset. It was divided into equal AM and PM shifts, or "survey days." Survey days were stratified into weekday or weekend strata, with state and federal holidays and the Friday PM shift included in the latter category. Therefore, a calendar week without holidays contained five weekend and nine weekday survey days. For each week, one weekday and four weekend shifts were randomly selected for creel sampling.
The creel clerk parked at State Highway 356, which bisected the survey area, and flipped a coin to decide the starting direction. He walked along the stream bank to one end of the survey area, counting all anglers present. He then reversed course and slowly returned to the starting point, counting and interviewing anglers along the way. Upon return to the highway, he repeated this process for the remaining half of the survey area. Once the entire area was surveyed, the clerk remained at the parking area until the end of the shift and updated his interview sheets with completed trip data from any returning anglers.
The clerk recorded hours fished to the nearest 0.25 hr and number of each species caught, including all fish kept or reported released. Other data collected included total length (nearest mm) of each creeled fish, whether the trip was complete or incomplete, bait used, trip expenses, number of 1990 trout fishing trips, and home county. Anglers were also asked to rate their trip satisfaction on a scale of one (low) to ten (high) and to
26

estimate the number of trout fishing trips they took during the previous year. They were asked an open-ended question on what the most important factor was for them to have a "good" fishing trip. If the angler stated that he or she did not know, the clerk offered a choice among catching trout, keeping trout, scenery, and solitude.
Data were checked and transferred to microcomputer spreadsheet files. A BASIC microcomputer program then expanded the creel data by period, using the Cooperative Statistics Project methodology to provide estimates of total pressure and total harvest. Angler characteristics data were summarized directly from the computer spreadsheet. Data Analysis
Differences between fish density and biomass estimates among streams in a given stream group were evaluated by comparing the five-year average estimate for each stream using the Kruskal-Wallis one-way analysis of variance (Hollander and Wolfe 1973). Differences among years for the average annual density and biomass for all streams in the group were also evaluated using the Kruskal-Wallis test. Density and biomass estimates were compared for all YOY trout, adult trout, and total trout. Since the annual averages reported in the tables are weighted by sample size, the average of the values reported for each stream will not equal the overall annual average since the latter is based on individual station data. An analysis of the variation in density estimates ofYOY and adult trout among years within individual streams was also conducted using the KruskalWallis test.
Relationships between estimates of fish density and habitat, substrate and gradient (HSG) variables were evaluated using stepwise linear regression to develop a multivariate
27

model. Individual variables that were identified using the stepwise technique were also evaluated using simple linear regression. Fish densities were log transformed for use in all regression analyses.
Relationships were evaluated for data collected at individual fish sampling stations and for data measured over an entire stream reach. A reach was defined as starting approximately 300 m downstream of the lowest fish sampling station and extending approximately 300 m above the uppermost fish sampling station.
Adult brown trout densities in improved stream sections (Cooper station two, Hedden station one, South Moccasin station two, and Jones station four) were compared with estimates from one or more "unimproved" stations (Cooper station one, Ridley station one, South Moccasin stations one and three, Jones stations 1-3) within each respective stream system. The Wilcoxon signed-rank test was used to compare density differences between the two Cooper Creek stations and the Kruskal-Wallis test was employed on the remaining comparisons.
RESULTS Fish Populations
Species composition varied greatly among streams. Many streams (Board Camp, Charlies, Chattahoochee, North and South Fork Moccasin, Ridley, Hedden and Totterypole) contained only trout (Table 2). Other streams contained a varied species assemblage in addition to trout, ranging from one additional species in Dukes Creek to 17 in Cooper Creek. Chubs, sculpins, and suckers were the most important species by
28

Table 2. Fish species collected in quantitative electrofishing samples in selected northeast Georgia trout streams, 1991-1995. Common and scientific names are from Robins et al. (1990).

Stream Name

TAXA SALMONIDAE

Board Camp Charlies

Cooper

Chatt. River

Dicks

Dover Dukes

Hedden Hoods

Jones

Moccasin

N.

Main

S.

Fork Stream Fork

Noontootla Ridley Stonewall

Tottery Pole

Oncorthynchus mykiss,

X

X

X

X

X

X

X

X

X

Rainbow trout

X

X

Salmo trutta Linnaeus, Brown trout

X

X

X

X

X

X

X

X

X

X

X

Salvelinus fontinalis, Brook trout
N
1..0 CYPRINIDAE

X

X

X

X

X

X

X

X

X

X

X

X

Semotilus atromaculatus, Creek chub
Nocomis leptocephalus, Bluehead chub
Nocomis micropogon, River Chub
Hybopsis rubrifrons, Rosyface chub

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

Rhinichthys cataractae,

X

Longnose dace

Clinostomus funduloides,

X

Rosyside dace

Rhinichthys atratulus, Blacknose dace

Campostoma anomalum,

X

X

Stoneroller

X X

Table 2, continued.

Stream Name

TAXA CYPRINIDAE

Moccasin

Board

Chait.

Toltery

Camp Charlies Cooper River Dicks Dover Dukes Hedden Hoods Jones N.

Main

S. Noontootla Ridley Stonewall Pole

Fork Stream Fork

Luxilus coccogenis,

X

Warpaint shiner

Notropis leuciodus,

X

Tennessee shiner

Notropis lutipinnis, Yellowfin shiner

w

Luxilus zonistius,

0

Bandfin shiner

X

X

X

CASTOSTOMIDAE
Hypentelium nigricans, Northern hogsucker
Moxostoma rupiscartes, Striped jumprock

X

X

X

X

X

X

X

X

X

X

CENTRARCHIDAE

Lepomis macrochirs,

X

Bluegill

Micropterus salmoide,

X

Largemouth bass

Lepomis auritus,

X

Redbreast sunfish

Table 2, continued

Stream Name

TAXA
PERCIDAE
Etheostoma blennioides, Greenside darter Etheostoma inscriptum, Turquoise darter Etheostoma rufilineatum, Redline darter

Board Camp Charlies

Cooper

Chatt. River

Dicks

Dover Dukes

Hedden

Hoods

Jones

Moccasin

N.

Main

S.

Fork Stream Fork

Noontootla Ridley Stonewall

Tottery Pole

x

x

x

x

COTTIDAE

W

I-'

Cottus bairdi,

Mottled sculpin

x

x

xx

x

x

ICTALURIDAE

Ictalurus punctatus,

x

Channel catfish

number and weight in most streams (Table 3), but the relative importance of each species varied a great deal among streams. Y oung-of-Year Trout Density
The density ofYOY trout in brook trout streams varied from 41/ha in lower Totterypole Creek during 1992 to 659/ha in the upper Chattahoochee River during 1991 (Table 4). Upper Chattahoochee River also had the highest five-year average density, and was significantly different from the North Fork Moccasin and lower Totterypole estimates. Lower Totterypole had the lowest five-year average density. Mean annual density ofYOY trout in all brook trout streams combined ranged from 233/ha in 1993 to 361/ha in 1995, but the differences among years were not significant.
In brown trout streams, the density ofYOY trout ranged from zero in South Fork Moccasin Creek in 1991 to 1,258/ha in upper Jones Creek in 1993 (Table 5). Upper Jones Creek averaged 832 YOY trout/1m over the five-year study (the highest mean for all brown trout streams), but this was significantly different only from South Fork Moccasin. There were no significant differences among years for the average YOY density in brown trout streams.
In the rainbow trout stream group, the density of YOY trout ranged from zero observed in both upper and lower Stonewall Creek during several years to 1,226/ha in upper Dukes Creek in 1993 (Table 6). Over the five-year study, lower Charlies Creek had the highest YOY trout density, and was significantly different from the main stem of Moccasin Creek and both upper and lower Stonewall creeks. The annual average density of YOY trout in 1994 was significantly different from both the 1993 and 1995 estimates.
32

Table 3. Proportional abundance and biomass of nontrout species captured in selected Georgia trout streams, 1991-1995.

Stream Coopers
Dicks Dover Dukes Hoods
Jones Moccasin Noontootla
Stonewall

Family
Cyprinidae - chubs Cyprinidae - minnows Catostomidae - suckers Cottidae - sculpins Miscellaneous species
Totals Cyprinidae - chubs Catostomidae - suckers Cottidae - sculpins Miscellaneous species
Totals Cyprinidae - chubs Catostomidae - suckers
Totals Cyprinidae - chubs Cyprinidae - chubs Cyprinidae - minnows Catostomidae - suckers Cottidae - sculpins Miscellaneous species
Totals Cyprinidae - chubs Catostomidae - suckers Cottidae - sculpins
Totals Cyprinidae - chubs Cyprinidae - minnows Catostomidae - suckers Cottidae - sculpins
Totals Cyprinidae - chubs Cyprinidae - minnows Catostomidae - suckers Cottidae - sculpins Miscellaneous species
Totals Cyprinidae - chubs Catostomidae - suckers Miscellaneous species
Totals

Percent by Number
11.4 27.2 10.4 37.1 13.9 100.0 53.2 3.0 38.5 5.3 100.0 61.9 38.1 100.0 100.0 7.8 1.0 5.2 85.7 0.3 100.0 72.9 7.8 19.4 100.0 7.5 0.6 3.7 88.2 100.0 2.6 1.6 1.0 90.6 4.2 100.0 88.9 10.5 0.6 100.0

Percent by Weight
7.4 7.7 55.4 19.9 9.7 100.0 59.6 19.8 18.9 1.8 100.0 46.1 53.9 100.0 100.0 12.2 1.2 23.4 62.3 0.9 100.0 56.9 34.4 8.7 100.0 14.4 0.3 40.7 44.6 100.0 20.7 3.2 30.9 35.0 10.2 100.0 74.8 24.6 0.6 100.0

33

Table 4. Average density of young-of-year (YOY) and adult trout in Georgia brook trout streams sampled from 1991-1995. The Kruskal-Wallis test was used to detect differences (within age groups only) in annual values within each stream, five-year averages among all streams, and annual averages from combined streams. Within each of these three groups, values with the same superscript letter are not significantly different (a=O.05). Standard errors are in parentheses. ND = no data.

STREAM

Board Camp

Chattahoochee River, Lower

Chattahoochee River,

w
~

Upper

Moccasin, North Fork

Totterypole, Lower

Totterypole, Upper

Annual Average, All Streams

AGE GROUP
yay
Adult
yay
Adult
yay
Adult
yay
Adult
yay
Adult
yay
Adult
yay
Adult

1991
217' (116) 185" (27)
254' (92) 113- (83)
659' (216) 135" (20)
149" (25) 109" (32)
62" (7) 100" (48)
297" (37) 287" (79)
273" (59) 155b (24)

Average Density (fish/ha) by Year

1992

1993

1994

1995

588' (143) 177" (12)
88' (12) 140- 31)
354' (58) 235" (84)
163" (78) 161" (40)
41" (41) 93" (81)
242" (97) 313" (16)
246" (53) 187b (23)

362' (58) 591" (25)
285' (64) 63- (5)
243" (25) 142" (52)
166" (28) 225" (65)
86" (5) 79" (26)
256" (38) 431" (63)
233" (26) 255",b (49)

274' (15) 488" (67)
258' (124) 79- (22)
522" (126) 193" (19)
273" (62) 103- (69)
ND ND
232" (104) 333" (96)
312" (46) 239",b (47)

308' (60) 514- (77)
390' (102) 345- (71)
533" (56) 692" (160)
223" (32) 443" (135)
202" (14) 152- (53)
512" (132) 520" (60)
361" (41) 444" (53)

FIVE YEAR AVERAGE

YOY

ADULT

350a,b (48) 255a,b,c (42) 462" (59) 195b,c (23) 98c (21) 308a,b (44)

391- (50) 148b (34) 279",b (65) 208-,b (44) 106b (16) 377" (35)

Table 5. Average density of young-of-year (YOY) and adult trout in Georgia brown trout streams sampled from 1991-1995. The Kruskal-Wallis test was used to detect differences (within age groups only) in annual values within each stream, five-year averages among all streams, and annual averages from combined streams. Within each of these three groups, values with the same superscript letter are not significantly different (a=O.05). Standard errors are in parentheses. Where no standard error is shown, insufficient data were collected to calculate it.

STREAM

AGE GROUP

Hedden
Jones, Lower
Jones, Upper w
t.n
Moccasin, South Fork
Ridley

YOY
Adult
YOY
Adult
YOY
Adult
YOY
Adult
YOY
Adult

1991
19 19
128" (46) 86b (11)
424 424
0" (0) 161' (73)
85 164

Average Density (fish/ha) by Year

1992

1993

1994

302 19
223" (69) 135"b (4)
742 291
29" (19) 97' (53)
697 79

73 225
220" (59) 227' (6)
1,258 649
130" (33) 99' (39)
248 388

34 80
64" (22) 159"b (23)
556 662
76" (43) 86' (14)
61 218

1995
447 76
15S" (3) 157',b (41)
1,179 728
27" (13) 364' (36)
600 139

FIVE YEAR AVERAGE

YOY

ADULT

175a,b (85) 158a,b (24) 832" (166)
52b (16) 338a,b (132)

84b (38) 153b (15) 551' (83) 161b (33) 198"b (53)

Annual Average, All Streams

YOY
Adult

102" (47) ISO' (44)

278" (93) 121' (29)

292" (124) 249' (60)

119" (57) 188' (62)

308" (128) 279' (69)

Table 6. Average density of young-of-year (YOY) and adult trout for Georgia rainbow trout streams sampled from 1991-1995. The Kruskal-Wallis test was used to detect differences (within age groups only) in annual values within each stream, five-year averages among all streams, and annual averages from combined streams. Within each of these three groups, values with the same superscript letter are not significantly different (a=O.05). Standard errors are in parentheses.

STREAM

AGE GROUP

Charlies, Lower

Charlies, Upper

Dicks, Lower

Dicks, Upper

w
0'>

Dukes, Lower

Dukes, Upper

Moccasin, Main Stream

Stonewall, Lower

Stonewall, Upper

yoy
Adult
yay
Adult
yay
Adult
yay
Adult
yay
Adult
yay
Adult
yay
Adult
yay
Adult
yay
Adult

Annual Average, All Streams

yay
Adult

1991
673' (178) 577' (94)
365' (240) 245' (72)
187' (27) 98' (54)
80' (54) 71- (11)
278'b (69) 225b (57)
546"b (44) 195b (45)
82' (18) 149- (36)
0' (0) 20",b (13)
0' (0) 0- (0)
246',b (53) 175' (35)

Average Density (fish/ha) by Year

1992

1993

1994

1,149' (416) 646' (152)
588' (133) 267' (72)
591' (43) 81' (8)
221' (29) 54- (16)
670"b (104) 340-,b (65)
650"b (93) 366"b (38)
55' (9) 58- (12)
69' (54) 4b (4)
16' (8) 6- (6)

719' (124) 878- (203)
385' (69) 743- (167)
502' (157) 344- (90)
87- (16) 169- (91)
998' (192) 867- (70)
1,226' (88) 877",b (108)
132' (35) 159- (31)
43' (II) 19",b (6)
242' (94) 37- (10)

338' (89) 475- (221)
332' (109) 333' (102)
248' (63) 267- (82)
322' (141) 99- (40)
82b (59) 746"'b (63)
46b (26) 935- (83)
44' (16) 144- (30)
0- (0) 25-,b (10)
0' (0) 44- (22)

445"b (82) 203- (44)

481' (83) 455- 76)

157b (34) 341- (65)

1995
779' (Ill) 558- (150)
843' (171) 407' (75)
1,044' (325) 253- (101)
388' (107) 212' (56)
542"b (168) 615-,b (67)
1,079"b (146) 663-,b (90)
152' (75) 186- (50)
146' (52) 111' (36)
0' (0) 53- (17)
553' (86) 340- (48)

FIVE YEAR AVERAGE

YOY

ADULT

732' (108) 503' (78) 515"b (103) 220',b" (46) 514"b (98) 709' (1l7)
93b" (18) 51' (19) 52' (30)

627- (74) 399-,b (62) 209",b,., (39) 121b,."d (25) 559- (69) 607- (82) 139b,." .. (17)
36"" (12) 28" (8)

In the sympatric stream group, YOY trout density ranged from zero in upper Hoods Creek in 1991, 1992, 1994, and 1995, to 2,007/ha in Dover Creek in 1993 (Table 7). Over the five-year study, Dover Creek was significantly different from Cooper Creek, lower and upper Hoods, and lower Noontootla creeks. Upper Hoods averaged only one fish/ha, which was significantly less than estimated densities for all other streams in the sympatric group. There were no statistically significant differences among years for annual mean density ofYOY trout in the sympatric stream group.
Adult Trout Density Among brook trout streams, adult density ranged from 63/ha in the lower
Chattahoochee River in 1993 to 692/ha in the upper Chattahoochee in 1995 (Table 4). Over the five year study, Board Camp Creek had the highest mean density of adult brook trout and was significantly different from the lower Chattahoochee River and lower Totterypole Creek. For the average annual density of adult trout in all brown trout streams, differences between 1995 and both 1991 and 1992 were significant.
The density of adult trout in brown trout streams ranged from 19/ha in Hedden Creek during 1991 and 1992 to 388/ha in Ridley Branch during 1993 (Table 5). Over the five-year period, upper Jones Creek had a significantly higher mean density than Hedden, lower Jones and South Fork Moccasin creeks. There were no significant differences among years for the annual average of adult trout density in brown trout streams.
In rainbow trout streams, adult trout density ranged from zero in upper Stonewall in 1991 to 935/ha in upper Dukes Creek in 1994 (Table 6). Over the five- year study
37

Table 7. Average density of young-of-year (YOY) and adult trout in Georgia sympatric trout streams sampled from 1991-1995. The Kruskal-Wallis test was used to detect differences (within age groups only) in annual values within each stream, five-year averages among all streams, and annual averages from combined streams. Within each of these three groups, values with the same superscript letter are not significantly different (a=0.05). Standard errors are in parentheses.

STREAM
Coopers Dover Hoods, Lower Hoods, Upper Noontootla, Lower
cwo
Noontootla, Upper
Annual Average, All Sympatric Streams

AGE GROUP
yay
Adult
yay
Adult
yay
Adult
yay
Adult
yay
Adult
yay
Adult
yay
Adult

1991

136' 141"
244" 220'
60"b 151'
0' 25'
112' 192'
318' 491'

(108) (49)
(102) (36)
(27) (18)
(0) (6)
(27) (31)
(82) (90)

146'

(35)

207'

(40)

Average Density (fish/ha) by Year

1992

1993

1994

153' 200'
992"b 283'
211"" 66'
0' 19'
44' 196'
484' 593'

(4) (1)
(263) (52)
(6) (15)
(0) (0)
(11) (45)
(79) (17)

323'

(96)

227'

(49)

247' 221'
2,007' 607'
253' 144'
6' 12'
345' 242'
728' 575'

(84) (44)
(337) (76)
(22) (27)
(6) (6)
(183) (29)
(188) (48)

598' (173)

300'

(55)

30' 10'
1,037"b 589'
12b 53'
0' 19'
48' 168'
276' 461'

(7) (14)
(221) (132)
(6) (27)
(0) (Il)
(10) (25)
(113) (74)

234'

(97)

232'

(57)

1995

FIVE YEAR AVERAGE

YOY

ADULT

143' 89'
998"b 464'
91"b ISS'
0' 12'
62' 179'
525' 530'

(34) (25)
(259) (62)
(20) (51)
(0) (12)
(19) (24)
(243) (37)

303'

(99)

238'

(48)

141 b,. (30) 1,055' (177)
126b,. (25) Id (I)
122< (44) 466"b (73)

148<' (19) 433"b (51) 114" (17)
17 (3) 196b,< (14) 530' (26)

period, lower Charlies Creek had an average density of 627/ha, the highest of any rainbow trout stream, while upper Stonewall had the lowest. Differences among years for the average density of adult trout were not significant.
In streams with sympatric trout populations (Table 7), adult density varied from 12/ha in upper Hoods Creek in 1993 and 1995 to 607/ha in Dover Creek in 1993. Estimated adult density for upper Noontootla was significantly different from all other streams in this group except Dover Creek. No significant differences were detected among years for the density of adult trout. Total Trout Density
In brook trout streams, annual density estimates for all trout (yay and adults) ranged from 133/ha in lower Totterypole Creek in 1992 to 1,225/ha in the upper Chattahoochee River in 1995 (Table 8). Over the five-year period, densities in Board Camp Creek were significantly greater than those in the lower Chattahoochee, North Fork Moccasin, and lower Totterypole. For all brook trout streams combined, mean density for 1995 was significantly different from estimates for 1991 and 1992.
In brown trout streams, total trout density ranged from 38/ha in Hedden Creek in 1991 to 1,907/ha in upper Jones Creek in 1993 and 1995 (Table 9). Upper Jones Creek held significantly more trout than lower Jones, South Fork Moccasin and Hedden creeks. No differences among years were significant for total brown trout density.
Total trout density in rainbow streams (Table 10) ranged from zero in upper Stonewall Creek in 1991 to 2,102/ha in upper Dukes Creek in 1993. The average density of all trout in lower Charlies was significantly different from estimates for upper Dicks,
39

Table 8. Average density of all trout for Georgia brook trout streams sampled from 1991-1995. The Kruskal-Wallis test was used to detect differences in five-year averages among all streams, and annual averages from combined streams. Within each of these two groups, values with the same superscript letter are not significantly different (a=0.05). Standard errors are in parentheses.
N/D = no data.

STREAM
Board Camp (SE) Chattahoochee River, Lower (SE)
~
Cl Chattahoochee River, Upper (SE) Moccasin, North Fork (SE) Totterypole, Lower (SE) Totterypole, Upper (SE)
Annual Average, All Streams (SE)

Average Density (fish/ha) by Year 1991 1992 1993 1994 1995

402 (116)
368 (156)
795 (235)
258 (49)
162 (34)
585 (115)
428b (69)

765 (154)
228 (19)
589 (109)
324 (113)
133 (79)
555 (88)
433 b (64)

953 (39)
347 (69)
385 (61)
391 (93)
164 (19)
687 (101)
488a,b (67)

762 (78)
336 (146)
715 (108)
377 (131)
N/D
565 (129)

822 (33)
735 (170)
1225 (212)
667 (150)
354 (45)
1032 (190)

551 a,b (65)

806a (84)

FIVE YEAR AVERAGE
741 a (61) 403 b,c (66) 742a,b (96) 403 b,c (57) 203 c (34) 68y,b (69)

Table 9. Average density oftrout, nontrout, and all fish for Georgia brown trout streams sampled from 1991-1995. The Kruskal-Wallis test was used to detect differences (within species groups only) in five-year averages among all streams and annual averages from combined streams. Within each of these two groups, values with the same superscript letter are not significantly different (a=O.05). Standard errors are in parentheses. Where no standard error is shown, insufficient data were collected to calculate it.

STREAM

SPECIES GROUP

Hedden

Trout Nontrout All Fish

Jones, Lower

Trout Nontrout All Fish

Jones, Upper
.j:::> I-'

Trout Nontrout All Fish

Moccasin, South Fork Trout Nontrout All Fish

Ridley

Trout Nontrout All Fish

Annual Average, All Streams

Trout Nontrout All Fish

1991
38 0
38
215 (37) 848 (271) 1,063 (296)
848 0
848
161 (73) 0
161 (73)
249 0
249
252" (81) 170" (170) 534" (176)

Average Density (tishlha) by Year

1992
321 0
321

1993
298 0
298

1994
114 0
114

357 (69) 560 (207) 918 (267)

446 (59) 1,592 (880) 2,038 (938)

223 (42) 890 (320) 1,113 (344)

1,033 0
1,033

1,907 0
1,907

1,218 0
1,218

127 (40) 0
127 (40)

229 (29) 0
229 (29)

162 (50) 0
162 (50)

776

636

279

0

0

0

776

636

279

398" 106) 112" (112) 585" (154)

541" (178) 318" (318) 1,071" (403)

307" (117) 178" (178) 604' (198)

1995
523 0
523
313 (44) 475 (110) 788 (119)
1,907 0
1,907
391 (27) 0
391 (27)
739 0
739

FIVE YEAR AVERAGE

TROUT NONTROUT ALL FISH

258b (85) 0

259b (85)

311b (30) 873" (199)

1,184" (218)

1,383" (222) 0

1,383" (222)

2W (31) 0

536b 113)

535",b (113) 0

2l4",b (31)

587" (172) 95' (95) 745" (161)

Table 10. Average density of trout, nontrout, and all fish for Georgia rainbow trout streams sampled from 1991-1995. The Kruskal-Wallis test was used to detect differences (within species groups only) in five-year averages among all streams and annual averages from combined streams. Within each of these two groups, values with the same superscript letter are not significantly different (a=0.05). Standard errors are in parentheses.

STREAM

SPECIES GROUP

1991

Average Density (fishlha) by Year

1992

1993

1994

1995

FIVE YEAR AVERAGE

TROUT

NONTROUT

ALL FISH

Charlies, Lower
Charlies, Upper
Dicks, Lower
~
N
Dicks, Upper
Dukes, Lower
Dukes, Upper
Moccain, Main Stream

Trout Nontrout All Fish
Trout Nontrout All Fish
Trout Nontrout All Fish
Trout Nontrout All Fish
Trout Nontrout All Fish
Trout Nontrout All Fish
Trout Nontrout All Fish

1,250 1,250

(263) 0
(263)

610 (312) 0
610 (312)

285 3,096 3,381

(80) (236) (I94)

151 947 1,099

(60) (303) (361)

504 (46) 10 (10)
514 (41)

741 (78) 0
741 (78)

230 (53) 301 (98) 531 (102)

1,795 (529) 0
1,795 (529)
857 (200) 0
857 (200)
673 (35) 1,833 (72) 2,505 (107)
275 (45) 805 (157) 1,080 (200)
1,010 (152) 10 (10)
1,020 (162)
1,016 (97) 0
1,016 (97)
114 (20) 213 (56) 326 (74)

1,597 1,597

(325)
0 (325)

1,127 1,127

(230) 0
(230)

846 2,893 3,739

(241) (229) (205)

256 1,270 1,526

(102) (244) (345)

1,865 21
1,886

(254) (16) (256)

2,102 (52) 0
2,102 (52)

291 (25) 601 (34) 891 (50)

813 (309) 0
813 (309)

666 (210) 0
666 (210)

515 2,401 2,916

(125) (297) (406)

421 (179) 532 (73) 953 (208)

828 (120)

7

(7)

835 (127)

981 (94) 0
981 (94)

188 (23) 247 (20) 435 (36)

1,337 0
1,337

(253) (253)

1,250 0
1,250

(219) (219)

1,297 3,380 4,677
,
601 490 1,091

(397) (350) (189)
(75) (75) (48)

1,157 4
1,161

(182) (4)
(178)

1,741
0 1,741

(159) (159)

338 (60) 410 (129) 749 (184)

1,358' (160) 902"b (112) 723"b,o (123) 341 b,o,d (57)
1,066' (132) 1,316' (I 43)
232o,d (26)

0 1,358',b (160)

0 902b,o,d (112)

2,721 (175) 3,444' (219)
8090,b (106) 1,150b,o (Ill)
lid (4) 1,083"b,o (138)

0 1,316',b (143)

3

54

b ,<

(48)

587b,o,<I.. 68)

Table 10, continued.

STREAM

SPECIES GROUP

Stonewall, Lower Stonewall, Upper

Trout Nontrout All Fish
Trout Nontrout All Fish

.po
w Annual Average, All Streams

Trout Nontrout All Fish

1991
20 (13) 241 (63) 261 (58)
0 (0) 172 (57) 172 (57)
421" (85) 530' (190) 837' (193)

Average Density (fishlha) by Year

1992

1993

1994

73 (58) 394 (134) 467 (84)

22

(2)

311 (158)

333 (157)

648"b (122) 396' (114) 930' (157)

62 (14) 364 (15) 426 (22)
279 (94) 333 (99) 612 (95)
936"b (152) 609' (179) 1,394' (209)

25 (10) 156 (46) 181 (38)
25 (10) 242 (219) 292 (213)
496"" (78) 398' (147) 802' (167)

1995
257 (70) 263 (61) 520 (116)
53 (17) 233 (69) 286 (82)
892' (122) 531' (204) 1,326' (255)

FIVE YEAR AVERAGE

TROUT

NONTROUT

ALL FISH

87d (28) 76d (32)

284"'< (37)

371"d" (44)

258<,d (53)

339<1.0 (64)

Moccasin, and upper and lower Stonewall Creeks. Differences among annual mean densities of all streams were significant only between 1991 and 1995.
Sympatric trout streams (Table 11) had total trout densities ranging from 12/ha in upper Hoods Creek in 1995 to 2,614/ha in Dover Creek in 1993. Five-year mean trout densities for upper Noontootla and Dover Creeks were significantly different from the other streams. There were no significant differences among the five yearly estimates of mean annual trout density for sympatric streams.
Density of Nontrout Species In brook trout streams no other fish species were captured, with one exception.
Wild rainbow trout were captured in upper Totterypole Creek during sampling in 1992 (one fish), 1993 (four), 1994 (seven), and 1995 (five). These fish apparently resulted from a recent, illegal stocking. They were removed from the stream to prevent further colonization and recruitment. The rainbow trout data were not included in any analyses.
Only trout were captured in all brown trout streams except lower Jones Creek, where densities ranged from 475/ha in 1995 to 1,592/ha in 1993 (Table 9). Due to variability in the data, there were no significant differences in nontrout abundance among years in this stream .
In rainbow trout streams the density of nontrout ranged from zero at both upper and lower Charlies and upper Dukes Creeks for all years, to 3,380/ha in lower Dicks Creek in 1995 (Table 10). Over the five-year study period, the density of nontrout in lower Dicks Creek was significantly different from all other rainbow trout streams except
44

Table 11. Average density of trout, nontrout, and all fish for Georgia sympatric trout streams sampled from 1991-1995. The Kruskal-Wallis test was used to detect differences (within species groups only) in five-year averages among all streams and annual averages from combined streams. Within each of these two groups, values with the same superscript letter are not significantly different (a=O.05). Standard errors are in parentheses.

STREAM

SPECIES GROUP

Coopers

Trout Nontrout All Fish

Dover

Trout Nontrout All Fish

Hoods, Lower
.p:.
(J'l
Hoods, Upper

Trout Nontrout All Fish
Trout Nontrout All Fish

Noontootla, Lower

Trout Nontrout All Fish

Noontootla, Upper

Trout Nontrout All Fish

Annual Average, Sympatric Streams

Trout Nontrout All Fish

1991
277 (60) 945 (21) 1,216 (34)
464 (134) 91 (26) 556 (160)
211 (16) 862 (261) 1,073 (277)
25 (6) 0 25 (6)
304 (59) 95 (52) 399 (41)
809 (130) 311 (85) 1,120 (214)
352' (69) 398' (l08) 703b (122)

Average Density (fish/ha) by Year

1992

1993

1994

352 (3) 1,154 (24) 1,506 (27)
1,274 (314) 54 (44)
1,328 (341)
277 (18) 424 (69) 701 (77)
19 (0) 0 19 (0)
240 (53) 54 (30) 294 (24)
1,077 (69) 554 (195)
1,631 (262)

468 (126) 1,949 (257) 2,417 (131)
2,614 (410) 65 (19)
2,679 (404)
398 (48) 1,663 (468) 2,060 (502)
18 (10) 0 18 (10)
587 (210) 655 (391) 1,242 (420)
1,303 (235) 1,004 (252) 2,306 (408)

133 (19) 1,351 (194) 1,484 (188)
1,626 (350) 59 (13)
1,685 (356)
65 (33) 262 (66) 327 (77)
19 (II) 0 19 (II)
216 (34) 136 (88) 352 (112)
738 (186) 612 (142) 1,350 (321)

551' (128) 371' (106) 878"b (167)

898' (221) 1,000' (214) 1,787' (253)

466' (149) 454' (129) 870"b (175)

1995
232 (53) 1,497 (132) 1,729 (80)
1,462 (288) 6 (6)
1,468 (285)
246 (58) 526 (202) 772 (227)
12 (12) 0 12 (12)
241 (35) 185 (158) 427 (186)
1,056 (235) 910 (230)
1,965 (465)
542' (139) 586' (151) 1,062"b (193)

FIVE YEAR AVERAGE

TROUT

NONTROUT

ALL FISH

289b (44) 1,430' (1l9)

1,488' (219)

55" (12)

2 lOb" (30)

747' (165)

18' (4) 318b (54)

0 225b (94)

996' (88)

677' (99)

1,718' (129) 1,543' (219)
987"b (189) 18' (4)
543b" (124) 1,674' (173)

upper Dicks Creek. When streams were combined by year, there were no significant differences among years in the annual mean density of nontrout.
The density of nontrout in sympatric trout streams ranged from zero in upper Hoods Creek for all years to 1,949/ha in Cooper Creek in 1993 (Table 11). Mean nontrout density in Cooper, lower Hoods, and upper Noontootla creeks was significantly different from that for upper Hoods, Dover and lower Noontootla Creeks. For all sympatric streams combined, the differences among the five yearly estimates for nontrout were not significant. Total Fish Density
Since no nontrout were captured in brook trout streams, total fish density was the same as total trout density. In brown trout streams, total density ranged from 38/ha in Hedden Creek in 1991 to 2,038/ha observed in lower Jones Creek in 1993 (Table 9). Both upper and lower Jones Creek were significantly different from Hedden and South Fork Moccasin creeks over the five year period. For brown trout streams combined by year, the differences in total density were not statistically significant.
In rainbow trout streams, total fish density ranged from 172/ha in upper Stonewall Creek in 1991 to 4,677/ha in lower Dicks Creek in 1995 (Table 10). The average density observed in lower Dicks Creek was significantly different from upper Charlies, upper Dicks, Moccasin and upper and lower Stonewall creeks. For rainbow trout streams combined by year, no significant differences were observed among yearly means.
In the sympatric trout stream group, the total fish density ranged from 12/ha in upper Hoods in 1995 to 2,679/ha in Dover Creek in 1993 (Table 11). Mean densities in
46

Cooper, Dover, and upper Noontootla creeks were significantly different from upper Hoods and lower Noontootla creeks. Significant differences in total density existed between 1991 and 1993 when estimates for all years were combined.
Young-of-Year Trout Biomass YOY biomass in brook trout streams ranged from 0.33 kg/ha in lower Totterypole
Creek in 1992, to 5.2 kg/ha in the upper Chattahoochee River in 1991 (Table 12). The densities observed in the upper Chattahoochee were significantly different only from lower Totterypole Creek.
Biomass estimates for YOY trout in brown trout streams ranged from zero in South Fork Moccasin Creek in 1991 to 11.02 kg/ha in upper Jones Creek in 1993 (Table 13). The average biomass ofYOY trout over the five-year study period in South Fork Moccasin Creek was significantly different from upper Jones Creek. There were no significant differences among yearly means of YOY biomass.
In rainbow trout streams (Table 14), YOY biomass ranged from zero in 1991 and 1994 in lower Stonewall and 1991, 1994 and 1995 in upper Stonewall, to 5.72 kg/ha in lower Charlies Creek in 1992. When annual estimates were averaged for each stream, YOY biomass in lower and upper Charlies and lower Dukes creeks were significantly different from main stem Moccasin and upper and lower Stonewall creeks. Mean YOY biomass estimates for 1992, 1993 and 1995 were significantly different from that for 1994.
47

Table 12. Average biomass of young-of-year (YOY) and adult trout for Georgia brook trout streams sampled from 1991 - 1995. The Kruskal-Wallis test was used to detect differences (within age groups only) in five-year averages among all streams and annual averages from combined streams. Within each of these two groups, values with the same superscript letter are not significantly different (a=O.05). Standard errors are in parentheses. ND = no data.

STREAM

AGE

Average Biomass (kg/ha) by Year

GROUP

1991

1992

1993

1994

1995

FIVE YEAR AVERAGE

YOY

ADULT

Board Camp

YOY
Adult

Chattahoochee River, YOY

Lower

Adult

Chattahoochee River, YOY

Upper

Adult

..f::> 00

Moccasin, North Fork YOY

Adult

Totterypole, Lower

YOY
Adult

Totterypole, Upper

YOY
Adult

1.81 (1.08) 5.02 (0.76)
2.51 (0.88) 3.32 (2.26)
5.20 (1.54) 3.69 (0.45)
1.21 (0.11 ) 2.53 (0.53)
0.53 (0.01) 2.28 (0.57)
2.34 (0.07) 8.90 (2.75)

3.39 (1.09) 5.71 (0.48)
0.77 (0.10) 4.82 (1.38)
2.25 (0.56) 7.24 (2.39)
1.26 (0.51) 5.75 (1.32)
0.33 (0.33) 1.80 (0.96)
1.79 (0.79) 9.93 (1.29)

1.22 (0.18) 12.80 (0.96)
2.21 (0.63) 3.23 (1.22)
1.71 (0.24) 5.16 (1.87)
1.01 (0.14) 9.59 (3.61)
0.67 (0.06) 2.98 (0.65)
1.35 (0.27) 11.87 (1.08)

0.85 (0.09) 11.58 (1.27)
1.67 (0.78) 3.06 (0.76)
2.73 (0.79) 5.63 (0.72)
1.37 (0.36) 3.01 (2.07)
NO NO
NO 10.39 (2.62)

1.06 (0.21) 12.14 (1.50)
1.69 (0.44) 12.66 (3.87)
1.97 (0.15) 15.70 (3.88)
1.l0 (0.14) 10.22 (3.67)
0.96 (0.07) 4.85 (0.24)
1.70 (0.41) 14.97 (2.58)

1.67a,b,c (0.36) 1.77a,b (0.29) 2.77' (0.46) 1.19b,c (0.12) 0.62c (0.10) 1.62a,b (0.21)

9.45a,b (0.98) 5.42b.c (1.29) 7.48a,b,< (1.42) 6.22b,< (1.29) 2.98c (1.45) 11.21' (1.00)

Annual Average, All Streams

YOY
Adult

2.27' (0.46) 4.29b (0.76)

1.63' (0.33) 5.87b (0.77)

1.36' (0.16) 7.61 a,b (1.16)

1.51' (0.27) 1.41' (0.13) 6.73a,b (1.15) 11.76a (1.34)

Table 13. Average biomass of young-of-year (YOY) and adult trout for Georgia brown trout streams sampled from 1991 - 1995. The KruskalWallis test was used to detect differences (within age groups only) in five-year averages among all streams and annual averages from combined streams. Within each of these two groups, values with the same superscript letter are not significantly different (a=O.05). Standard errors are in parentheses. Where no standard error is shown, insufficient data were collected to calculate it.

STREAM

AGE GROUP
1991

Average Biomass (kg/ha) by Year

1992

1993

1994

1995

FIVE YEAR AVERAGE

YOY

ADULT

Hedden

YOY
Adult

Jones, Lower

YOY
Adult

.j::> Jones, Upper
~

YOY
Adult

Moccasin, South Fork

YOY
Adult

Ridley

YOY
Adult

0.34 1.28
1.25 (0.57) 8.27 (1.24)
.56 39.56
0 9.80 (4.27)
0.88 15.11

3.74 2.52
1.47 (0.47) 16.52 (0.96)
6.68 29.43
0.38 (0.25) 5.35 (3.38)
5.47 6.12

0.48 14.54
1.52 (0.40) 18.76 (3.21)
11.02 46.70
2.44 (0.66) 14.27 (4.88)
1.15 17.84

0.13 8.46
0.48 (0.14) 11.96 (2.50)
3.24 51.45
0.75 (0.38) 12.32 (3.47)
0.25 13.35

2.27 7.51
0.51 (0.01) 14.21 (3.86)
5.07 63.98
0.17 (0.09) 15.46 (7.41)
2.29 11.13

1.39a,b (0.70) 1.05a,b (0.17) 6.31" (1.30) 0.75b (0.27) 2.01a,b (0.93)

6.81b (2.39) 13.96,b (1.38) 46.22' (5.79) 11.44" b (2.09) 12.71,b (1.98)

Annual Average, All Streams

YOY
Adult

1.17" (0.60) 2.38" (0.80) 12.24' (3.83) 11.52" (3.11)

2.73" (1.08) 19.80' (3.83)

0.81" (0.33) 16.23' (4.59)

1.30" (0.55) 19.19' (6.17)

Table 14. Average biomass of young-of-year (YOY) and adult trout for Georgia rainbow trout streams sampled from 1991 - 1995. The Kruskal-Wallis test was used to detect differences (within age groups only) in five-year averages among all streams and annual averages from combined streams. Within each of these two groups, values with the same superscript letter are not significantly different (a=O.05). Standard errors are in parentheses.

STREAM

AGE GROUP

Charlies, Lower

Charlies, Upper

Dicks, Lower

Dicks, Upper

U1 0

Dukes, Lower

Dukes, Upper

Moccasin, Main stream

Stonewall, Lower

Stonewall, Upper

yay
Adult
yay
Adult
yay
Adult
yay
Adult
yay
Adult
yay
Adult
yay
Adult
yay
Adult
yay
Adult

Annual Average, All Streams

yay
Adult

1991
3.69 (1.03) 21.12 (1.73)
2.40 (1.62) 13.93 (2.91)
1.19 (0.17) 13.57 (10.63)
0.87 (0.64) 2.93 (0.76)
2.70 (0.65) 8.30 (2.23)
4.17 (0.26) 8.79 (1.74)
0.65 (0.13) 7.56 (2.42)
0 1.94 (1.13)
0 0
1.74a.b (0.35) 8.67 (1.65)

Average Biomass (kg/ha) by Year

1992

1993

1994

5.72 (1.47) 20.47 (5.01)
3.33 (1.00) 11.38 (2.00)
2.90 (0.28) 4.49 (1.18)
2.11 (0.15) 2.13 (0.90)
4.19 (0.70) 10.44 (1.85)
3.20 (0.37) 11.06 (1.20)
0.38 (0.09) 5.00 (1.72)
0.91 (0.71) 0.10 (0.10)

3.18 (0.34) 21.43 (5.87)
2.04 (0.49) 24.86 (8.20)
2.26 (0.69) 11.27 (3.98)
0.85 (0.14) 9.84 (5.72)
5.00 (0.83) 20.71 (1.64)
7.06 (0.44) 24.06 (5.29)
0.70 (0.17) 13.10 (4.88)
0.54 (0.13) 1.97 (0.40)

2.46 (0.89) 15.93 (7.47)
2.81 (0.93) 16.98 (4.46)
1.28 (0.34) 17.83 (6.68)
2.14 (0.65) 6.94 (3.79)
0.51 (0.39) 19.49 (1.43)
0.21 (0.14) 27.51 (0.49)
0.26 (0.08) 15.49 (8.97)
0 2.33 (0.46)

0.23 (0.12) 1.38 (1.38)

2.89 (0.83) 2.19 (1.19)

0 1.85 (0.96)

2.55" (0.39) 7.38 (1.34)

2.72" (0.42) 14.38" (2.11)

1.08b (0.25) 13.82 (2.08)

1995
3.97 (0.38) 17.27 (5.33)
4.27 (0.94) 18.82 (4.13)
3.35 (1.06) 9.27 (4.60)
1.59 (0.42) 7.72 (2.47)
3.02 (0.74) 20.74 (2.11)
3.54 (0.51) 23.59 (2.71)
0.45 (0.25) 12.29 (6.40)
1.62 (0.51) 7.95 (3.82)
0 2.00 (0.08)
2.42" (0.33) 13.30" (1.72)

FIVE YEAR AVERAGE

YOY

ADULT

3.80" (0.45) 2.97" (0.45) 2. I9a.b (0.32) 1.51a.b,c (0.23)

19.22" (2.13) 17.20" (2.20) 11.29",b. (2.65) 5.91b,c,d (1.47)

3.08" (0.48) 3.64" (0.60) 0.49b,c (0.07) 0.61 b,c (0.22) 0.62c (0.34)

15.94",b (1.61) 19.00" (2.29) 10.69",b,C(2.33) 2.86c,d (0.99) 1.48d (0.41)

In streams with sympatric trout populations (Table 15), YOY biomass ranged from zero in upper Hoods Creek in four of five years, to 7.93 kg/ha in Dover Creek in 1993. Over the five-year study period, Dover YOY biomass was significantly different from Cooper, lower and upper Hoods and lower Noontootla Creeks. YOY Biomass estimates for 1992 and 1993 were significantly different from the 1994 estimate. Adult Trout Biomass
In brook trout streams (Table 12), adult biomass ranged from 1.80 kg/ha in lower Totterypole Creek in 1992 to 15.70 kg/ha in the upper Chattahoochee River in 1995. For the five-year study period, mean biomass in upper Totterypole was significantly different from lower Chattahoochee, North Fork Moccasin, and lower Totterypole. For all brook trout streams combined, mean annual biomass for 1995 was significantly different from estimates for 1991 and 1992.
The biomass of adult trout in brown trout streams ranged from 1.28 kg/ha in Hedden Creek in 1991 to 63.98 kg/ha in upper Jones Creek in 1995 (Table 13). Adult biomass over the five-year study period in upper Jones was significantly different from Hedden and South Fork Moccasin Creeks. There were no significant differences among yearly means of the biomass of adult trout.
In rainbow trout streams, adult trout biomass ranged from zero in upper Stonewall in 1991 to 27.51 kg/ha in upper Dukes in 1994 (Table 14). Adult biomass estimates over the five-year period for lower and upper Charlies and upper Dukes creeks were significantly different from those for upper Dicks and upper and lower Stonewall creeks.
51

Table 15. Average biomass of young-of-year (YOY) and adult trout for Georgia sympatric trout streams sampled from 1991 - 1995. The Kruskal-Wallis test was used to detect differences (within age groups only) in five-year averages among all streams and annual averages from combined streams. Within each ofthese two groups, values with the same superscript letter are not significantly different (a=O.05). Standard errors are in parentheses.

STREAM

AGE GROUP

Coopers

Dover

Hoods, Lower

U1 N

Hoods, Upper

Noontootla, Lower Noontootla, Upper

yay
Adult
yay
Adult
yay
Adult
yay
Adult
yay
Adult
yay
Adult

Average Biomass (kg/ha) by Year

1991

1992

1993

1994

1.33 (1.01) 21.87 (11.63)
2.55 (1.12) 29.88 (8.27)
0.55 (0.19) 12.32 (3.54)
0 6.87 (2.11)
0.85 (0.20) 18.32 (2.20)
2.61 (0.61 ) 13.49 (2.36)

1.55 (0.04) 14.87 (0.70)
4.54 (1.23) 21.99 (3.74)
1.66 (0.08) 11.27 (6.62)
0 5.87 (1.73)
0.16 (0.02) 18.45 (5.17)
2.06 (0.53) 18.18 (1.56)

1.52 (0.49) 18.50 (2.65)
7.93 (1.46) 29.96 (2.86)
1.61 (0.10) 8.65 (1.52)
0.Q7 (0.07) 0.27 (0.14)
1.22 (0.48) 21.08 (2.09)
1.77 (0.31) 17.09 (1.86)

0.17 (0.03) 9.56 (1.27)
3.85 (1.02) 22.96 (4.59)
0.12 (0.09) 3.54 (1.88)
0 4.31 (2.49)
0.16 (0.02) 14.30 (1.61)
0.72 (0.27) 12.20 (1.61)

1995
0.69 (0.19) 12.63 (2.50)
3.56 (0.90) 27.06 (5.29)
0.35 (0.08) 15.90 (4.46)
0 3.09 (3.09)
0.22 (0.07) 21.64 (2.36)
2.49 (1.12) 16.45 (1.89)

FIVE YEAR AVERAGE

YOY

ADULT

0.99b,c (0.22) 4.49" (0.66) 0.86b,c (0.18) O.Old (0.01) 0.52c,d (0.15) 1.93a,b (0.30)

15.04b,c (1.96) 26.37" (1.89) 1O.33cd (1.89) 4.08d (1.02) 18.76",b (1.32) 15.48",b, (0.93)

Annual Average, All Streams

yay
Adult

1.41a,b (0.31) 16.85" (2.58)

1.96" (0.39) 15.12" (1.96)

2.61" (0.62) 15.92" (2.39)

0.86b (0.36) 11.14" (1.81)

1.27a,b (0.37) 16.13" (2.16)

There were no significant differences among yearly means of the average annual biomass of adult trout.
In streams with sympatric trout populations (Table 15), adult trout biomass ranged from 0.27 kg/1m in upper Hoods Creek in 1993 to 29.96 kg/ha in Dover Creek the same year. Biomass in Dover was significantly different from both lower and upper Hoods and Cooper creeks over the five year study period. No significant differences among years were detected for the annual average biomass in combined sympatric streams.
Total Trout Biomass Trout biomass in brook trout streams ranged from 2.13 kg/ha in lower Totterypole
Creek in 1992 to 17.67 kg/ha in the upper Chattahoochee River in 1995 (Table 16). Over the five-year study period, differences in mean trout biomass between upper Totterypole and lower Totterypole, North Fork Moccasin and the lower Chattahoochee River were significant. When comparing years, the difference in mean biomass between 1991 and 1995 was significant.
In brown trout streams (Table 17), trout biomass ranged from 1.62 kg/ha in Hedden Creek in 1991 to 69.05 kg/ha in upper Jones in 1995. The five-year mean estimate for trout biomass in upper Jones was significantly different from Hedden and South Fork Moccasin creeks. When all brown trout streams were combined, no significant differences among yearly means were observed.
53

Table 16. Average biomass of total trout for Georgia brook trout streams sampled from 1991-1995. The Kruskal-Wallis test was used to detect differences in five-year averages among all streams and annual averages from combined streams. Within each of these two groups, values with the same superscript letter are not significantly different (a=O.05). Standard errors are in parentheses. ND =no data.

STREAM
Board Camp (SE)
Chattahoochee River, Lower (SE) Chattahoochee River, Upper (SE)
tTl
~ Moccasin, North Fork (SE) Totterypole, Lower (SE)
Totterypole, Upper (SE)
Annual Average, All Streams (SE)

Average Biomass (kg/ha) by Year

1991

1992

1993

1994

1995

6.83 (0.96)
5.83 (2.99)
8.89 (1.98)
3.74 (0.64)
2.81 (0.58)
11.25 (2.80)
6.56 b (0.96)

9.10 (1.53)
5.59 (1.29)
9.49 (2.51)
7.01 (1.80)
2.13 (1.24)
11.71 (1.41)
7.50 a,b (0.95)

14.02 (0.86)
5.44 (1.85)
6.88 (1.87)
10.61 (3.75)
3.65 (0.68)
13.22 (1.33)
8.97a,b (1.17)

12.43 (1.34) 4.73 (1.43) 8.36 (0.09) 4.38 (2.42)
ND
ND
8.24a,b (1.12)

13.20 (1.32)
14.35 (4.23)
17.67 (4.01)
11.32 (3.74)
5.81 (0.21)
16.67 (2.96)
13.17a (1.43)

FIVE YEAR AVERAGE
11.12a,b (0.86) 7.19 b,c (1.38)
10.26a,b (1.37) 7.41 b,c (1.33)
3.60c (0.53)
12.83 8 (1.05)

Table 17. Average biomass of trout, nontrout, and all fish for Georgia brown trout streams sampled from 1991-1995. The Kruskal-Wallis test was used to detect differences (within age groups only) in five-year averages among all streams and annual averages from combined streams. Within each of these two groups, values with the same superscript letter are not significantly different (a=O.05). Standard errors are in parentheses. Where no standard error is shown, insufficient data were collected to calculate it

STREAM

SPECIES GROUP

1991

Average Biomass (kg/ha) by Year

1992

1993

1994

1995

FIVE YEAR AVERAGE

TROUT

NONTROUT

ALL FISH

Hedden
Jones. Lower
Jones, Upper
Moccasin, South Fork
CJ1 CJ1
Ridley
Annual Average, All Streams

Trout Nontrout All Fish
Trout Nontrout All Fish
Trout Nontrout All Fish
Trout Nontrout All Fish
Trout Nontrout All Fish
Trout Nontrout All Fish

1.62 0
1.62
9.52 (1.11) 12.81 (3.20) 22.33 (2.14)
45.12 0
45.12
9.80 (4.27) 0
9.80 (4.27)
15.99 0
15.99

6.26 0
6.26
17.99 (0.61) 6.51 (0.98) 24.50 (0.95)
36.11 0
36.11
5.73 (3.17) 0
5.73 (3.17)
11.59 0
11.59

15.02 0
15.02
20.28 (3.52) 10.74 (2.92) 31.22 (5.36)
57.72 0
57.72
16.71 (4.50) 0
16.71 (4.50)
18.99 0
18.99

8.59 0
8.59
12.44 (2.61) 9.12 (1.05) 21.56 (3.55)
54.69 0
54.69
13.07 (3.29) 0
13.07 (3.29)
13.60 0
13.60

9.78 0
9.78
14.79 (3.87) 5.69 (0.82) 20.48 (3.52)
69.05 0
69.05
15.63 (7.32) 0
15.63 (7.32)
13.42 0
13.42

13.41' (4.33) 4.27' (2.33) 17.68' (4.41)

13.90' (3.48) 2.17' (1.12) 16.07' (3.94)

22.52' (4.74) 3.65' (2.01) 26.17' (5.01)

17.05' (4.88) 3.04' (1.55) 20.09' (4.81)

20.39' (6.56) 1.90' (0.98) 22.29' (6.41)

8.25' (2.19) Ob
15.00,b (1.44) 9.01' (1.06)
52.54' (5.61) Ob
12.19' (2.09) Ob
14.72' (1.28) Ob

8.25b (2.19) 24.02'b (1.66) 52.54' (5.61) 12.19b (2.09) 14.72'b (1.28)

In rainbow trout streams (Table 18), trout biomass estimates ranged from 1.01 kg/ha in lower Stonewall in 1992 to 31.12 kg/ha in upper Dukes Creek in 1993. Fiveyear average biomass estimates for both lower Charlies and upper Dukes creeks were significantly different from those for upper Dicks and upper and lower Stonewall creeks. There were no significant differences among annual means for total trout biomass.
In streams with sympatric trout populations, annual trout biomass estimates ranged from 0.49 kg/ha on upper Hoods to 37.89 kg/ha on Dover in 1993 (Table 19). Average trout biomass in Dover creek was significantly different from all other streams in this group except lower Noontootla. There were no significant differences in the annual average biomass for all trout among years. Nontrout Biomass
Nontrout species were not captured in brook trout streams. In the brown trout stream group, nontrout were only captured in one of 5 streams sampled, lower Jones Creek. Estimates ranged from 5.69 kg/ha in 1995 to 12.81 kg/ha in 1991 and the differences among annual means were not significant (Table 17).
In rainbow trout streams, nontrout were captured in six of 9 streams sampled (Table 18). Biomass estimates for nontrout in streams where they occurred in this group ranged from 0.05 kg/ha in lower Dukes in 1991, to 27.85 kg/ha in lower Dicks Creek in 1991 .. The highest five-year average biomass in lower Dicks Creek was significantly different from upper and lower Charlies, upper and lower Dukes and upper Stonewall. There were no significant differences among years for the average annual biomass of nontrout in rainbow trout streams.
56

Table 18. Average biomass of trout, nontrout, and all fish for Georgia rainbow trout streams sampled from 1991-1995. The KruskalWallis test was used to detect differences (within species groups only) in five-year averages among all streams and annual averages from combined streams. Within each of these two groups, values with the same superscript letter are not significantly different (a=O.05). Standard errors are in parentheses.

STREAM Charlies, Lower
Charlies, Upper
Dicks, Lower
-U...1..
Dicks, Upper
Dukes, Lower
Dukes, Upper
Moccasin, Main Stream

SPECIES GROUP
Trout Nontrout All Fish
Trout Nontrout All Fish
Trout Nontrout All Fish
Trout Nontrout All Fish
Trout Nontrout All Fish
Trout Nontrout All Fish
Trout Nontrout All Fish

1991
24.70 (2.32) 0
24.70 (2.32)
16.34 (4.33) 0
16.34 (4.33)
14.76 (10.79) 27.85 (3.78) 42.61 (11.96)
3.80 (1.39) 11.35 (4.04) 15.15 (5.37)
11.00 (1.79) 0.05 (0.05) 11.05 (1.74)
12.97 (1.78) 0
12.97 (1.78)
8.21 (2.54) 6.82 (1.96) 15.03 (4.03)

Average Biomass (kglha) by Year

1992

1993

1994

26.19 (5.81) 0
26.19 (5.81)

24.60 (6.20) 0
24.60 (6.20)

18.39 (8.34) 0
18.40 (8.34)

14.70 (2.98) 0
14.71 (2.98)

26.90 (8.68) 0
26.90 (8.68)

19.80 (5.13) 0
19.80 (5.13)

7.39 (1.05) 20.75 (1.60) 28.14 (0.61)

13.53 (4.58) 26.81 (2.78) 40.34 (2.51)

19.10 (6.86) 24.00 (1.87) 43.10 (8.43)

4.24 (1.05) 8.86 (2.79) 13.10 (3.64)

10.69 (5.72) 8.00 (3.21) 18.69 (8.82)

9.08 (4.42) 5.21 (1.60) 14.29 (5.67)

14.63 (2.32) 0.14 (0.14) 14.77 (2.46)

25.72 (2.47) 0.51 (0.40) 26.23 (2.58)

20.00 (1.75) 0.19 (0.19)
20.19 (1.92)

14.26 (1.52) 0
14.26 (1.52)

31.12 (5.13) 0
31.12 (5.13)

27.72 (0.47) 0
27.72 (0.47)

5.38 (1.63) 3.51 {1.20) 8.89 (2.24)

13.80 (4.74) 7.29 (0.65) 21.09 (5.22)

15.75 (8.91) 2.38 (0.56) 18.13 (8.63)

1995
21.23 (5.70) 0
21.23 (5.70)
23.08 (4.48) 0
23.08 (4.48)
12.62 (5.14) 19.58 (1.07) 32.20 (4.94)
9.32 (2.18) 6.05 (2.95) 15.37 (4.81)
23.76 (2.27) 0.12 (0.12) 23.88 (2.20)
27.13 (2.78) 0
27.13 (2.78)
12.74 (6.32) 2.61 (0.73) 15.36 (6.18)

FIVE YEAR AVERAGE

TROUT NONTROUT

ALL FISH

23.03' (2.39) 0

23.03'b (2.39)

20.17'b (2.38) 0

20.17,b (2.38)

13.48,b.e (2.66) 23.80' (1.26)

7.43 b.e.d (1.51)

7.89'" (1.28)

37.28' (3.10) IS .32b.e.d(2.29)

19.02'b (1.68) 00,

19.22'b (1.72)

22.64' (2.27) O

22.64'b (2.27)

1. 18'b.e.d(2.31 )

4.52'b (0.71)

15.70b.e (2.40)

Table 18, continued.

STREAM

SPECIES GROUP

Stonewall, Lower Stonewall, Upper

Trout Nontrout All Fish
Trout Nontrout All Fish

Annual Average, All Streams

Trout Nontrout All Fish

1991
1.94 (1.l3) 2.22 (0.61) 4.16 (1.39)
1.38 (1.38) 2.19 (0.91) 3.57 (2.26)

Average Biomass (kg/ha) by Year

1992

1993

1994

1.01 (0.81) 2.60 (0.91) 5.00 (1.81 )

2.51 (0.51) 3.21 (0.44) 5.72 (0.21)

2.33 (0.46) 1.52 (0.18) 3.85 (0.62)

2.57 (2.23) 2.66 (0.97) 5.24 (2.95)

3.05 (1.l3) 2.02 (1.11) 6.90 (1.90)

2.23 (0.85) 2.09 (1.76) 4.32 (1.83)

1995
9.57 (4.32) 4.58 (2.86) 14.15 (7.15)
1.30 (0.65) 2.27 (0.78) 3.56 (1.26)

10.57' (1.83) 5.61" (1.79) 16.18' (2.59)

10.04' (1.66) 4.28 (1.30) 14.48' (1.72)

16.88' (2.40) 5.32" (1.65) 22.40' (2.53)

14.93' (2.16) 3.93 (1.46) 18.87' (2.69)

15.64' (1.96) 3.91 (1.23) 19.55' (1.26)

U1 CO

FIVE YEAR AVERAGE

TROUT

NONTROUT

ALL FISH

3.47'd (1.l3) 2.83,b" (0.59)

6.58'd (1.64)

2.11 d (0.55)

2.25b,< (0.44)

4.72d (0.87)

Table 19. Average biomass of trout, nontrout, and all fish for Georgia sympatric trout streams sampled from 1991-1995. The Kruskal-Wallis test was used to detect differences (within species groups only) in five-year averages among all streams and annual averages from combined streams. Within each of these two groups, values with the same superscript letter are not significantly different (a=O.05). Standard errors are in parentheses.

STREAM

SPECIES GROUP

Coopers

Trout Nontrout All Fish

Dover

Trout Nontrout All Fish

Hoods, Lower
U1
~
Hoods, Upper

Trout Nontrout All Fish
Trout Nontrout All Fish

Noontootla, Lower

Trout Nontrout All Fish

Noontootla, Upper

Trout Nontrout All Fish

Annual Average, All Streams

Trout Nontrout All Fish

1991
23.20 (10.62) 20.29 (1.02) 43.49 (9.60)
32.43 (8.44) 6.83 (3.26)
39.26 (9.30)
12.87 (3.36) 8.49 (2.23) 21.36 (4.58)
6.87 (2.11 ) 0
6.87 (2.1\)
19.17 (2.40) 2.53 (2.27) 21.71 (2.05)
16.10 (2.74) 1.57 (0.30) 17.68 (3.04)

Average Biomass (kg/ha) by Year

1992

1993

1994

16.42 (0.67) 18.90 (0.11) 35.32 (0.56)
26.53 (4.97) 3.52 (3.03) 30.05 (7.20)
12.94 (6.61) 6.08 (1.59) 19.01 (5.67)
5.87 (1.73) 0
5.87 (1.73)
18.60 (5.18) 4.15 (3.81) 22.75 (2.14)
13.16 (5.63) 2.22 (0.91) 15.38 (5.95)

20.02 (3.13) 24.29 (2.45) 44.30 (0.75)
37.89 (1.85) 5.03 (1.10) 42.93 (0.79)
10.26 (1.61 ) 14.25 (2.84) 24.52 (4.08)
0.49 (0.29) 0
0.49 (0.29)
22.30 (2.21) 3.45 (1.85) 25.76 (0.62)
18.86 (2.15) 4.23 (1.15) 23.09 (1.40)

9.73 (1.28) 12.74 (0.71) 22.46 (1.19)
26.81 (5.61) 3.44 (1.47) 30.25 (6.42)
3.66 (1.92) 2.05 (0.80) 5.71 (2.1\)
4.31 (2.49) 0
4.31 (2.49)
14.46 (1.62) 2.62 (2.27) 17.07 (3.86)
12.92 (1.72) 2.19 (0.48) 15.1\ (2.09)

1995
13.32 (2.58) 16.34 (0.67) 29.67 (3.25)
30.62 (4.80) 0.16 (0.16) 30.78 (4.70)
16.24 (4.48) 5.21 (2.27) 21.45 (5.79)
3.09 (3.09) 0
3.09 (3.09)
21.86 (2.43) 1.66 (1.31)
23.52 (3.69)
18.94 (1.71) 3.65 (0.91) 22.59 (1.60)

18.16' (2.69) 5.80' (1.67) 23.98' (3.55)

15.54' (2.34) 5.04' (1.56) 20.58' (2.86)

18.30' (2.86) 8.54' (2.10) 26.85' (3.60)

11.98' (2.12) 3.84' (1.07) 15.82' (2.49)

17.35' (2.34) 4.50' (1.42) 21.85' (2.58)

FIVE YEAR AVERAGE

TROUT

. NONTROUT

ALL FISH

16.03b (2.00) 30.86' (2.37) 11.20b., (1.90)
4.12' (1.01) 19.28'b (1.37) 16.00b (1.38)

18.34' (1.30) 3.80b (1.01) 7.21" b(1.34)
Ob 2.88b (0.95) 2.77b (0.41)

34.38' (2.79) 35.32' (2.64) 18.41 b (2.49)
4.12' (1.01) 22.16"b (1.29) 18.77b (1.54)

In the sympatric trout stream group, nontrout were captured in all streams except upper Hoods Creek (Table 19). The five-year average non trout biomass for Cooper Creek was significantly different from all other streams in this group except lower Hoods. Comparisons of the annual average biomass of nontrout in this group indicated no significant differences among years.
Total Fish Biomass Since no nontrout species were captured in brook trout streams, the total fish
biomass is the same as the total biomass previously reported. In brown trout streams total fish biomass ranged from 1.62 kg/ha in Hedden Creek in 1991 to 69.05 kg/ha in upper Jones Creek in 1995 (Table 17). The five-year average biomass in upper Jones was significantly different from all other streams in this group, except lower Jones. No significant differences were detected among years for estimates of total biomass.
In rainbow trout streams (Table 18), total biomass ranged from 2.19 kg/ha in upper Stonewall Creek in 1991 to 43.10 kg/ha in lower Dicks in 1994. The five-year average biomass in lower Dicks was significantly different from upper Dicks, main stem Moccasin and upper and lower Stonewall creeks. Comparisons of the annual average fish biomass for all streams in this group indicated no significant differences among years.
Total biomass within sympatric trout streams (Table 19) ranged from 0.49 kg/ha in upper Hoods in 1993 to 44.30 kg/ha in Cooper Creek also in 1993. The five-year average biomass estimates in Dover and Cooper Creeks were significantly different from upper and lower Hoods and upper Noontootla Creeks. Comparisons of the annual average
60

total biomass over all sympatric trout streams indicated no significant differences among years. Trout Population Structures
Length frequency summaries were developed for streams that produced at least 50 adult (l0 cm or larger) fish of a particular species. All brook trout populations were dominated by small fish (Figure 2), as only 5.2-16.5% of all fish collected exceeded 15 cm (Table 20). Only three streams produced fish 20 cm or larger, and only one stream, Chattahoochee River, produced any brook trout 25 cm or larger.
Brown trout streams contained more large trout (Table 21). Generally, about 3040% of each sample exceeded 15 cm and 3-10% exceeded 25 cm. One exception was Noontoola Creek, where 44% of the fish were 25 cm or larger (Figure 3).
Most rainbow trout populations were also dominated by small fish, with 92 to 97% of the samples composed of individuals <20 cm total length (Table 22). Three exceptions were Stonewall, Noontootla, and Cooper creeks, where 11.8%, 16.4%, and 31 % respectively of rainbow trout were <20 cm. A majority (60%) of those larger fish from Cooper Creek were stocked trout. Dukes Creek had the lowest proportion of adult trout exceeding 20 cm (Figure 4). YOY vs. Adult Trout Abundance
Brown trout streams were the only group in which a significant relationship existed between changes in YOY abundance and changes in adult abundance the
following year. Linear regression showed a significant positive relationship (R2 = 0.67).
61

c
.0-
ro+oJ

::J 00 (L

-+oJ
::J

"'0
0'>

N

~

0

+oJ
C
Q)
U
'-Q)
(L

Length Class
10-14.9 em ~15-19.9 em 100
80
60
40
20
o Board Camp Chattahoochee Noontootla N. Moccasin Stream

Ridley

Q
Totterypole

Figure 2. Size class distribution of adult brook trout in GADNR electrofishing samples from selected trout streams during 1991-1995.

Table 20. Length frequency of brook trout collected in quantitative electrofishing samples on selected Georgia streams during 1991-1995. Sample size is in parentheses.

Upper Limit of Length Group (mm)

Board Camp Creek (393)

<59

4.1

69

16.0

79

11.7

89

7.1

99

4.6

109

7.9

119

12.5

129

10.9

139

9.4

149

6.4

159

5.9

169

2.0

179

1.0

189

0.3

199

0.3

209

0.0

219

0.0

229

0.0

239

0.0

249

0.0

259

0.0

269

0.0

Select size groups:

>149

9.5

>199

0.0

>249

0.0

Percent of Sample by Stream

Chattahoochee River (719) 0.6 5.9 17.8 18.1 13.2 7.2 4.0 7.9 6.7 7.4 4.8 3.3 2.0 0.6 0.4 0.0 0.0 0.1 0.0 0.0 0.1 0.0

Noontootla Creek (132) 4.6 14.4 18.9 10.6 2.3 6.8 15.2 12.1 3.8 4.6 3.8 2.3 0.0 0.8 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

N.Moccasin Creek (337) 0.0 3.0 13.1 17.2 10.4 5.0 4.5 8.6 13.7 8.9 7.1 4.2 2.1 1.5 0.0 0.6 0.3 0.0 0.0 0.0 0.0 0.0

11.3

6.9

15.8

0.2

0.0

0.9

0.1

0.0

0.0

Ridley Branch
(77) 0.0 14.3 32.5 9.1 6.5 2.6 2.6 11.7 7.8 7.8 2.6 0.0 1.3 1.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

Totterypole Creek (483) 2.5 7.3 9.5 12.4 8.9 6.4 8.9 9.9 8.5 9.1 6.6 5.0 2.3 1.2 0.6 0.6 0.0 0.0 0.2 0.0 0.0 0.0

5.2

16.5

0.0

0.8

0.0

0.0

63

Table 21. Length frequency of brown trout collected in quantitative electrofishing samples on selected Georgia streams during 1991-1995. Sample size is in parentheses.

Upper Limit of Length Group (mm)
<59 69 79 89 99 109 119 129 139 149 159 169 179 189 199 209 219 229 239 249 259 269 279 289 299 >299 Select size groups: >149 >199 >249

Cooper Creek (434)
0.9 6.5 11.8 15.0 10.4 7.4 5.5 2.5 3.2 1.2 1.2 1.6 3.2 5.3 4.9 4.6 3.0 1.6 0.9 0.5 0.5 1.2 1.6 0.5 0.7 4.6
35.9 19.7 9.1

Dover Creek (277)
2.9 11.6 25.3 12.3 4.7 2.2 2.2 0.7 2.9 4.0 4.0 5.4 3.3 2.9 0.4 2.5 1.1 3.6 1.8 1.8 0.4 0.4 1.1 0.7 0.0 2.2
31.6 15.6 4.8

Percent of Sample by Stream

Hedden Creek (241)
1.7 5.8 16.6 10.4 9.5 5.0 4.6 3.3 0.8 0.8 1.7 6.6 7.5 5.8 6.2 3.7 2.9 1.7 0.4 2.1 0.8 0.0 1.2 0.0 0.4 0.4

Ridley Branch (293)
2.7 10.6 17.1 11.3 10.2 2.7 2.4
1.0 1.4 4.4 6.5 5.8 5.1 3.8 2.4 2.1 2.7 2.4 0.7 1.4 0.7 0.7 1.0 0.7 0.3 0.0

Hoods Creek (157)
0.0 6.4 8.3 16.7 14.7 5.1 0.6 1.9 0.6 2.6 2.6 2.6 3.2 3.2 3.8 3.9 2.6 2.6 2.6 1.9 2.6 3.2 1.3 0.6 1.9 5.1

Jones Creek (1,031)
0.8 6.5 9.4 12.8 11.7 7.9 2.7 1.6 0.9 2.7 6.1 5.7 4.6 5.0 3.0 3.6 3.4 3.2 1.7 1.3 1.9 1.0 1.3 0.3 0.2 1.0

Noontootla Creek (84) 1.2 1.2 3.6 10.7 3.6 4.8 0.0 1.2 2.4 1.2 2.4 2.4 3.6 1.2 1.2 2.4 2.4 3.6 3.6 3.6 8.3 8.3 0.0 7.1 3.6 16.7

S. Moccasin Creek (131) 0.0 0.0 0.0 2.3 5.3 10.7 15.3 13.7 8.4 5.3 3.8 2.3 1.5 2.3 3.8 2.3 4.6 2.3 2.3 2.3 0.8 3.8 0.8 2.3 1.5 2.3

41.4

36.3

43.7

43.3

70.4

39.0

13.6

12.7

28.3

18.9

59.6

25.3

2.8

3.4

14.7

5.7

44.0

11.5

64

I

Length Class

~12-14.9 em ~15-19.9 em ~20-24.9 em ~25+ em
100

c:

..0...
co

80

co:::J-
0....... 60

:::J

(J) tJ1

"'C
'+-

40

..0...

c:

0)

u
t-

20

O)

0..

o Cooper

Dover

Hedden

Ridley

Hoods

Jones

Noontootla S. Moccasin

Stream

Figure 3. Size class distribution of adult brown trout in GADNR electrofishing samples from selected trout streams during 1991-1995.

Table 22. Length frequency of rainbow trout collected in quantitative electrofishing samples on selected Georgia streams during 1991-1995 Sample size is in parentheses.

Upper Limit of Length
Group (mm)

Charlies Creek (1,312)

<59

0.5

69

8.0

79

18.2

89

20.3

99

6.2

109

3.0

119

6.7

129

8.3

139

8.3

149

4.6

159

4.0

169

2.8

179

2.5

189

2.0

199

1.2

209

1.2

219

0.5

229

0.5

239

0.4

249

0.2

259

0.2

269

0.1

279

0.1

289

0.2

299

0.0

>299

0.2

Selected size groups:

>149

16.1

>199

3.6

>249

0.8

Cooper Creek (529)
2.5 3.8 8.3 8.9 7.4 6.1 3.4 2.1 1.5 2.5 2.5 4.9 4.8 6.4 4.2 5.9 5.7 6.2 3.8 1.9 3.0 0.8 1.1 0.8 1.1 0.7
53.8 31.0 7.5

Percent of Sample by Stream

Dicks Creek (770)
6.4 15.2 20.7 15.7 6.8 3.8 2.0 4.4 3.5 4.3 3.1 3.1 3.4 1.2 1.4 1.6 1.2 0.9 0.5 0.4 0.3 0.0 0.3 0.1 0.0 0.0

Dover Creek (774) 12.1 22.0 23.4 11.2
3.8 1.3 2.5 3.1 4.3 3.9 3.0 2.1 1.6 1.7 0.9 0.9 0.4 0.5 0.5 0.1 0.3 0.1 0.0 0.1 0.0 0.4

Dukes Creek (2,200)
1.9 6.7 16.6 13.2 7.2 6.2 7.6 9.0 8.8 6.2 5.6 4.4 2.8 1.1 1.2 0.7 0.3 0.2 0.0 0.1 0.1 0.1 0.1 0.1 0.0 0.0

Hoods Creek
(56) 0.0 0.0 8.9 16.1 8.9 5.4 1.8 1.8 8.9 12.5 8.9 3.6 3.6 10.7 1.8 3.6 0.0 3.6 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

Moccasin Creek (281) 2.5 9.6 9.3 8.2 7.5 1.1 2.1 5.0 7.8 9.6 10.3 5.3 5.7 6.1 2.5 1.1 1.8 2.5 0.4 0.7 1.1 0.0 0.0 0.0 0.0 0.0

Noontootla Creek (748) 10.8 11.0 7.6 3.1 3.7 4.3 5.0 4.0 5.6 7.1 6.8 4.4 4.7 3.5 2.4 3.5 2.7 3.6 2.4 1.9 0.9 0.3 0.4 0.3 0.03 0.1

Stonewall Creek (146) 0.0 0.7 0.0 3.4 15.1 22.7 18.5 12.3 4.1 3.4 2.1 0.7 2.1 1.4 2.1 0.7 1.4 2.7 1.4 1.4 0.7 0.0 0.7 1.4 0.0 1.4

17.5

12.6

16.8

35.8

37.5

38.2

20.2

5.3

3.3

1.7

7.2

7.6

16.4

11.8

0.7

0.9

0.4

0.0

1.1

2.3

4.2

66

c 100
0

~ Length Class
~10-14.9 em ~15-19.9 em ~20+ em

~

-CO
:J

80

0-

0

0...
~

60

:J

O'l

"'C
'+-

40

F==-1

~

---<

'-I

0

c~

au> 20

L-

a>

0...

o

Charlies Cooper Dicks

Dover Dukes Hoods Moccasin Noontootla Stonewall

Stream

Figure 4. Size class distribution of adult rainbow trout in GADNR electrofishing samples from selected trout streams during 1991-1995.

No significant relationships between YOY and adult abundance were detected for brook trout streams, rainbow trout streams or streams with sympatric trout populations. Habitat Composition
Habitat varied a great deal among streams (Table 23) but pools and riffles were the dominant habitat types in most streams. Total pool area varied from 23.2% on the upper Chattahoochee River to 57.4 % on South Moccasin. Riffle area ranged from 28.5% on lower Stonewall to 74.1 % on the upper Chattahoochee River. Cascade was another important habitat type, and it varied from zero in several streams to 10.6 % in upper Hoods Creek. Gradient also varied a great deal among streams, with upper Cooper Creek having the highest percent of gradient < 2% and the lowest percent of gradient> 4%. The upper Chattahoochee had the lowest percentage of gradient < 2%, and upper Hoods had the highest percentage of gradient >4%.
LWD was the least abundant substrate in most streams, with the highest levels occurring in Board Camp and upper Totterypole Creeks with 3.3%. Cobble was the most common substrate overall, being most abundant in upper Dicks Creek where it composed 83.0% ofthe total (Table 24). Sand and sediment were also important substrates, and they varied from zero in several streams to 18.7 % for sediment in lower Stonewall to 31.6% for sand in Hedden Creek.
. Maximum temperatures for the seven-day period of highest mean water temperatures ranged from 16.4 0 C in the Chattahoochee River to 21.0 0 C in Cooper Creek (Table 25). Mean daily water temperature never exceeded 19.6 0 C. Water
68

Table 23. Relative habitat composition of Georgia trout stream reaches containing 1991-1995 fish monitoring stations. Stations shown are contained within the reach indicated. Habitat types are reported as percent of total area.

Stream

Board Camp

Lower Chattahoochee

Upper Chattahoochee

Lower Charlies

Upper Charlies

Lower Cooper

Upper Cooper

Lower Dukes

Upper Dukes

Dover

Lower Dicks

Upper Dicks

m

\.0

Lower Hoods

Upper Hoods

Lower Jones

Upper Jones

Moccasin Proper

Moccasin North Fork

Moccasin South Fork

Lower Noontootla

Upper Noontootla

Hedden

Ridley

Lower Stonewall

Upper Stonewall

Lower Totterypole

Upper Totterypole

Sample Year 1994 1994 1994 1990 1990 1994 1994 1994 1994 1990 1990 1990 1990 1990 1994 1994 1990 1990 1990 1994 1994 1994 1994 1994 1994 1994 1994

Reach (Meters)
729 683 919 869 918 720 887 1,145 689 781 1,052 738 820 896 2,203 598 1,485 893 898 1,055 1,467 777 949 951 757 740 616

Station Cascade

1-3

0.7

1-3

3.7

4-6

0.0

1-3

2.2

4-6

0.0

0.0

1-2

0.0

1-3

0.0

4-6

0.0

1-3

0.1

1-3

3.5

4-6

0.0

1-3

0.2

4-6

10.6

1-3

0.0

4

0.0

1-3

4.2

1-3

'\.5

1-3

1.3

1-3

0.0

4-6

0.0

0.0

0.0

1-3

9.6

4-6

0.0

1-3

8.5

4-6

6.8

Glide
0.4 1.0 0.0 0.0 0.0 4.2 6.5 3.4 1.0 0.0 0.0 0.0 0.0 0.0 2.6 1.4 0.0 0.0 0.0 0.8 1.3 2.9 3.3 3.3 2.2 0.0 1.2

Rime
61.2 66.9 74.1 47.7 70.6 50.4 45.1 47.5 70.1 43.3 44.6 61.4 35.7 43.5 45.4 55.0 31.6 55.8 41.3 40.4 60.9 47.5 52.8 28.5 41.5 43.5 50.3

Trench
15.6 19.6 19.5 20.1 11.9 40.8 26.8 22.6 13.9 28.2 23.9
1.2 12.0 7.61 20.4 20.1 26.3 15.9 8.1 27.3 21.9 7.3 7.5 10.9 31.9 17.0 6.9

Lateral Scour 4.8 6.2 0.0 8.6 9.3 0.0 20.6 16.6 9.2 6.5 15.3 28.2
1.0 0.8 17.7 14.3 0.8 9.3 37.1 0.0 \.5 5.5 11.8 4.3 0.0 4.4 6.0

Pool Plunge
4.0 0.3 2.2 8.7 0.9 0.0 0.0 1.3 3.6 3.1 6.6 3.6 7.3 7.3 1.4 2.9 Il.l 2.1 2.8 3.5 0.9 2.4 2.4 10.8 0.4 0.0 2.6

Dam
12.1 1.2 \.5
10.5 7.3 4.6 0.9 8.7 2.3 6.5 1.7 5.7 37.0 26.4 11.3 6.4 1.0 15.5 9.5 24.8 13.5 34.3 22.3 25.6 20.5 26.8 25.0

Total Pool .,1).5 27.2 23.2 47.9 29.4 45.4 48.4 49.1 28.9 44.3 47.5
38.6 57.3 42.1 50.8 43.7 39.3 42.8 57.4 55.9 37.8 49.5 44.0 57.0 52.8 48.1 40.5

Pocket Water
1.2 1.2 . 2.7 2.2 0.0 0.0 0.0 0.0 0.0 12.3 4.5 0.0 6.9 3.8 1.2 0.0 24.9 0.0 0.0 3.1 0.0 0.0 0.0 1.6 3.5 0.0 1.3

<2%

Gradient 2-4 %

29.8

25.7

29.3

36.0

23.3

43.2

42.4

19.7

27.2

49.2

62.1

34.2

66.5

30.2

47.0

29.3

25.4

54.3

51.2

30.2

53.4

28.5

23.6

22.2

42.9

25.2

38.7

9.9

54.8

34.3

44.8

29.4

39.7

24.3

44.3

30.5

35.3

32.7

37.3

37.9

54.5

22.6

31.0

29.3

47.6

32.8

52.5

29.4

64.2

11.9

54.1

27.7

32.3

21.8

>4%
44.6 34.7 33.5 38.0 23.5
3.8 3.3 23.8 20.3 18.6 18.1 54.2 31.8 51.3 10.9 25.8 36.0 25.2 32.0 24.8 23.0 39.6 19.6 18.1 23.9 18.2 45.9

Table 24. Relative substrate composition of trout stream reaches containing 1991-1995 fish monitoring stations.

Stream
Brook Trout Streams Board Camp Lower Chattahoochee Upper Chattahoochee Moccasin N. Fork Lower Totterypole Upper Totterypole Mean
Brown Trout Streams
Lower Jones Upper Jones Moccasin S. Fork Hedden Ridley
Mean

Year Sampled
1994 1994 1994 1990 1994 1994
1994 1994 1990 1994 1994

Reach Length Meters
729 683 919 893 740 616
2,203 598 898 777 949

Rainbow Trout Streams
Lower Charlies Upper Charlies Lower Dicks Upper Dicks Lower Dukes Upper Dukes Moccasin Proper Lower Stonewall Upper Stonewall
Mean Sympatric Trout Streams
Lower Cooper Upper Cooper Dover Lower Hoods Upper Hoods Lower Noontootla Upper Noontootla
Mean

1990 1990 1990 1990 1994 1994 1990 1994 1994

869 918 1,052 738 1,145 689 1,485 951 757

1994 1994 1990 1990 1990 1994 1994

720 887 781 820 896 1,055 1,467

Station
1-3 1-3 4-6 1-3 1-3 4-6
1-3 4
1-3
1-3 4-6 1-3 4-6 1-3 4-6 1-3 1-3 4-6
1-2 1-3 1-3 4-6 1-3 4-6

Fines Sand

0.1

11.1

0.1

0.1

0.0

1.2

6.2

2.5

6.2

16.6

3.8 26.7

2.7

9.7

2.6 12.4

0.8

9.4

7.1

3.3

2.5 31.6

8.6 27.5

4.3

16.8

0.0

8.1

2.9

2.1

12.4

0.7

12.2

1.4

4.3

1.5

0.3

1.4

11.2

0.0

18.7

17.5

7.2

17.4

7.7

5.6

0.0

8.2

0.0

3.5

5.2

2.7

7.3 24.5

0.6 21.4

0.8

8.2

1.0 12.9

2.1

11.6

Percent Substrate Type Gravel Cobble Boulder

19.5

39.4

14.5

7.1

69.0

12.1

11.1

58.8

17.2

24.4

36.0

4.9

3.5

14.9

21.1

16.0

16.8

8.0

13.6

39.2

13.0

9.6

45.0

7.6

16.3

56.0

11.6

24.1

55.3

3.9

5.9

31.4

8.2

15.0

37.9

0.7

14.2

45.1

6.4

15.3

40.0

19.2

15.8

64.6

12.3

10.8

39.3

9.1

3.3

83.0

0.2

5.9

69.0

7.6

4.2

59.6

18.8

1.4

29.3

29.8

2.1

16.9

14.4

5.8

18.4

27.5

7.2

46.7

15.4

0.5

23.9

37.0

8.0

52.7

29.3

5.2

48.9

32.1

3.7

26.2

10.0

4.3

27.1

9.1

1.6

13.5

47.8

16.7

39.0

9.7

5.7

33.0

25.0

Bedrock Wood

12.3

3.3

11.4

0.3

11.0

0.7

26.6

0.3

36.0

1.7

25.5

3.3

20.5

1.6

22.1

0.8

3.7

2.2

0.2

0.3

20.2

0.3

9.8

0.5

11.2

0.8

16.8

0.6

1.9

0.5

27.8

0.0

0.0

0.0

10.5

1.2

15.1

0.7

26.4

0.0

29.6

0.9

23.6

0.0

16.9

0.4

30.4

0.0

6.4

0.1

6.1

0.0

27.0

1.3

37.3

0.2

27.7

0.3

19.5

1.2

22.1

0.4

70

Table 25. Means of minimum (Min), maximum (Max), and mean daily water temperatures for the seven-day period with the highest average daily mean water temperature recorded on selected Georgia trout streams.

Stream
Charlies Charlies Charlies Chattahoochee Chattahoochee Chattahoochee Cooper Cooper Cooper Dicks Dicks Dicks Dukes Dukes Noontootla Noontootla Noontootla Stonewall Stonewall Totterypole Totterypole Totterypole

Year

Period of Record

7-Day Period With Highest Mean Water Temperature

1993 June 24-0ctober 4 June 24-Ju1y 30

1994 June 22-0ctober 27 July 17-July 23

1995 June 6-0ctober 2

July 25-July 31

1993 June 25-0ctober 4 July 24-July 30

1994 June 30-0ctober 27 July 22-July 28

1995 June 9-0ctober 2

August 18-August 24

1993 June 26-0ctober 5 July 24-July 30

1994 July I-October 30 July 8-July 14

1995 June 7-0ctober 2

August 19-August 25

1993 June 24-0ctober 4 July 24-July 30

1994 June 14-0ctober 27 July 21-July 27

1995 June 6-July 25

July 19-July 25

1993 July 2-0ctober 4

July 23-July 29

1995 June 10-October 1 August 18-August 24

1993 June 29-0ctober 5 July 25-July 31

1994 June 24-0ctober 30 August l4-August 20

1995 June 7-July 25

July 19-July 25

1993 June 24-0ctober 4 July 23-July 29

1994 June 22-0ctober 27 July 14-July 20

1993 June 24-0ctober 4 July 24-July 30

1994 June 10-0ctober 27 July 19-July 25

1995 June 6-July 25

July 19-July 25

Min
16.3 15.4 16.2 17.8 15.5 17.9 18.5 16.8 18.5 17.9 16.3 17.3 18.4 18.1 17.7 16.0 17.0 18.0 16.9 17.4 16.2 17.2

Max Mean
17.8 17.0 16.2 15.8 17.2 16.8 19.3 18.4 16.4 15.9 18.8 18.5 21.0 19.6 18.2 17.6 20.1 19.4 19.3 18.6 17.4 16.9 18.7 17.6 20.2 19.2 19.0 18.7 19.2 18.6 16.9 16.5 18.3 17.4 19.5 18.7 18.1 17.4 18.9 18.0 17.4 16.8 18.6 17.6

71

temperature was therefore judged not to be a critical factor for these streams, and no further evaluation of water temperatures was conducted.
FishlHabitat Relationships Trout Density/ Habitat Relationships Among Sampling Stations
For all stream groups except brook trout streams, significant relationships were observed between YOY trout densities and habitat characteristics among fish sampling stations (Table 26). In brown trout streams, the combination of large woody debris and riffle produced a significant regression model. Riffle was the more influential factor of the two. In rainbow trout streams, combined sand/sediment (SASED) and glide were significantly related to YOY densities, with SASED as the more important variable. In sympatric trout streams, cascade and riffle variables produced a significant regression, with cascade as the primary model component.
Adult trout densities were also significantly related to station habitat factors. For brook trout streams, LWD was the only significant factor. In brown trout streams, LWD and sediment variables produced a significant model, with LWD as the primary influence. In rainbow trout streams, a significant stepwise regression included glide and SASED, with SASED being more influential. In sympatric trout streams, the best regression model included dam pool, cascade, and glide variables.
Total trout densities also related to station characteristics. LWD was the sole variable significantly related to total trout density in both brook and brown trout streams. In rainbow trout streams, the combination of glide and SASED produced a significant
72

Table 26. Results of stepwise linear regression of the density of YaY, adult and total trout against habitat, substrate and gradient variables in each fish sampling station. Variables evaluated include percent substrate as bedrock, boulder, cobble, gravel, sand, wood sediment, and SASED (sand and sediment), the percent composition of cascade, dam pool, plunge pool, lateral scour pool, total pool, glide, pocket water, and riffle, the percent of stream length with a gradient of less than 2%, between 2 and 4%, and greater than 4%, and the density of nontrout species in a reach. Variables indicated were selected by the stepwise linear regression technique as part of a multiple regression model using a.:5 .05 as criteria for variables to enter or exit the model. Each stepwise model includes the variables indicated, and adjusted R2 values are shown in bold. Other adjusted R2 values are from simple linear regressions for the variable indicated, and are shown as an indication of the relative importance of each variable in the stepwise model.

YOYDensity

Adult Density

Total Trout Density

RZ

Factor (+/-)

a

RZ

Factor (+/-)

a

RZ

Factor (+/-)

Brook Trout Streams 18 Stations

No relationships

0.26 Large woody debris (+)
0.26 Stepwise regression model

0.02 0.21 Large woody debris (+) 0.21 Stepwise regression model

Brown

0.37 Large woody debris (+)

Trout Streams 0.43 Riffle (+)

'-l 9 Stations
W

0.81 Stepwise regression

model

0.05 0.62 Large woody debris (+) 0.03 0.13 Sediment (+)
0.83 Stepwise regression model

0.01 0.65 Large woody debris (+) 0.02
0.65 Stepwise regression model

Rainbow

0.44 SASED (-)

Trout Streams 0.29 Glide(-)

27 Stations

0.53 Stepwise regression

model

<0.01 0.41 SASED (-) <0.01 0.35 Glide (-)
0.54 Stepwise regression model

<0.01 0.45 SASED (-) <0.01 0.34 Glide (-)
0.57 Stepwise regression model

Sympatric

0.54 Cascade (-)

Trout Streams 0.18 Riffle (+)

18 Stations

0.65 Stepwise regression

model

<0.01 0.41 Cascade (-) 0.06 0.46 Dam pool (-) 0.06 Glide (-) 0.03 Large woody debris
0.81 Stepwise regression model

<0.01 0.40 Cascade (-) <0.01 0.49 Dam pool (-)
0.03 0.06 Glide(-) 0.03 0.06 Plunge pool
0.83 Stepwise regression model

a
0.03
0.01
<0.01 <0.01
<0.01 <0.01
0.03 0.01

model. For sympatric trout streams, the four component model of dam pool, cascade, plunge pool, and glide produced the best regression model.
Trout Density/ Habitat Relationships Among Stream Reaches Stepwise regression of mean trout density for adjacent stations on stream reach
habitat variables showed many relationships (Table 27). yay brook, brown, and rainbow trout densities showed correlations to habitat features. In brook trout streams three variables, gradient <2%, gravel, and pocket water, produced a significant relationship. For brown trout streams, boulder and total pool appeared to explain much variation in yay density. In rainbow trout streams, only SASED showed a correlation to yay abundance, while in sympatric trout streams, cascade was the only significant variable detected.
Adult trout densities showed relationships to reach HSG characteristics as well. In brook trout streams, plunge pool showed the only significant relationship to abundance. For brown trout streams, no significant relationships were found for the density of adult trout. In rainbow trout streams, SASED and LWD variables produced a significant model, with SASED as the more important factor. Cascade was the only significant factor related to adult density in sympatric trout streams.
Reach attributes also showed significant relationships to total trout densities. In brook trout streams, stepwise regression suggested that four factors, gradient <2%, plunge pool, pocket water and sediment are related to trout abundance. Gradient < 2% appeared to be the most important variable in that relationship. In brown trout streams, riffle and
74

Table 27. Results of stepwise linear regression of the density of YaY, adult and total trout against habitat, substrate and gradient variables in a stream reach extending from about 300 m downsteam to about 300 m upstream of fish sampling stations within that reach. Variables evaluated include percent substrate as bedrock, boulder, cobble, gravel, sand, wood sediment, and SASED (sand and sediment), the percent composition of cascade, dam pool, plunge pool, lateral scour pool, total pool, glide, pocket water, and riffle, the percent of stream length with a gradient of less than 2%, between 2 and 4%, and greater than 4%, and the density of nontrout species in a reach. Variables indicated were selected by the stepwise linear regression technique as part ofa multiple regression model
a.s using .05 as criteria for variables to enter or exit the model. Each stepwise model includes the variables indicated, and adjusted R2 values are shown in bold.
Other adjusted R2 values are from simple linear regressions for the variable indicated, and are shown as an indication of the relative importance of each variable in the stepwise model.

YOY Density

Adult Density

R2

Factor (+/-)

a

R2

Factor (+/-)

Brook Trout Streams 6 Reaches

0.89 -0.05 0.62
1.00

Gradient <2% (-) Gravel (+) Pocket water (+)
Stepwise regression model

0.01 <0.01
0.03

0.82 Plunge pool (+)
0.82 Stepwise regression model

-.J
(j1

Brown

-0.20 Boulder (+)

0.04

Trout

0.97 Total pool (-)

<0.01

Streams

9 Reaches

1.00 Stepwise

regression model

No significant relationships.

Rainbow Trout Streams 27 Reaches

0.67 SASED (-)
0.67 Stepwise regression model

<0.01

0.80 SASED (-) 0.05 LWD(+)
0.89 Stepwise regression model

Sympatric

0.83 Cascade (-)

0.01

0.71 Cascade (-)

Trout

Streams

0.83 Stepwise

0.71 Stepwise regression

6 Reaches

regression model

model

a
0.01
<0.01 0.04

Total Trout Density

R2

Factor (+/-)

0.71 Gradient <2% (-) 0.58 Plunge Pool (+) 0.43 Pocket water (+) 0.28 Sediment (-)

1.00 Stepwise regression model
0.21 Dam pool (-) 0.85 Riffle (+)

1.00 Stepwise regression model
0.76 SASED (-)

0.76 Stepwise regression model

a
<0.01 <0.01
0.05 <0.01
<0.01 0.02
<0.01
<0.01

0.29 0.71 Cascade (-)

0.02

0.71 Stepwise regression model

dam pool produced a significant model. Trout abundance in rainbow and sympatric trout streams were each related to single factors, SASED and cascade, respectively.
Stream Structures For Cooper, Hedden, and South Moccasin creeks, stream structures did not
increase adult brown trout populations over those found in adjacent, unimproved stream sections during this study. Cooper station two averaged 48/ha for the five-year period, which was not significantly different from station one density (40/ha). Hedden Creek, which was improved, held consistently fewer adults than Ridley Branch (Table 5). No significant difference was detected among the three South Moccasin stations. Station two (improved) averaged 215/ha for the five-year period, which was higher than stations one (l61/ha) and three (l09/ha), but the improved station held the highest density in only two of five years.
In Jones Creek, the improved station held more than three times the number of adult brown trout than were found in the other three stations. In 1984, just before installation of stream structures, station four contained 199/ha. Samples in 1985, 1987, 1988, and 1990 averaged 616/ha after improvement (Monte Seehom, U.S. Forest Service, unpublished data). Creel Survey
The creel survey on Dukes Creek ended after only seven periods due to statewide budget cuts. During the survey, 189 anglers were interviewed in the study area and completed trip information was obtained for 122 (65%) of those interviews (Table 28).
76

Table 28. Summary statistics for the seven-month roving creel survey on Dukes Creek, Georgia, from March 30 to October 7, 1991. Standard errors are in parentheses.

Period March 26 to April 22

Estimated Effort
Anglers (hrs)

61

496

Trout Harvest
Rate (n/hr)

Estimated Trout Harvest

0.67

332

% Successful
Anglers
66

April 23 to May 20

9

68

1.07

72

56

May 21 to June 17

32

201

1.29

260

71

June 18 to July 15

25

351

0.89

314

77

July 16 to August 12

34

234

1.01

236

65

August 13 to September 9

19

145

1.07

156

63

September 10 to October 7

9

85

0.19

16

78

Trout Catch Rates (n/hr)

All Successful Anglers Anglers

2.1

3.1

(0.3)

(0.4)

1.7

3.0

(0.6)

(0.5)

2.1

2.9

(0.4)

(0.4)

2.9

3.8

(0.5)

(0.5)

2.1

3.2

(0.4)

(0.5)

2.1

3.2

(0.6)

(0.7)

2.1

2.7

(0.7)

(0.7)

TOTAL SE

189

1,580

(381)

0.88 (0.12)

1,386 (283)

68

2.1

3.2

(0.2)

(0.2)

77

Total effort during the survey was estimated at 1,580 hr (632 hr/ha). Effort was highest during the first period, as the trout fishing season opened. Angler hours dropped precipitously during the second period, most likely due to an unusually rainy May when heavy thundershowers appeared to discourage trips. Effort recovered during subsequent periods before tailing off during late summer/early fall.
Interviewed anglers reported a catch of 689 trout, which included the harvest of 314 rainbow and 81 brook trout observed by the clerk. Seventy-eight percent of the rainbow harvest and 53% of the brook harvest were composed offish between 160 and 209 mm (Figure 5). Of the harvest, five rainbow trout and two brook trout exceeded 250 mm, with the largest of each species measuring 302 mm and 303 mm, respectively.
Angler success, defined as the reported catch of a least one trout during a trip, averaged 68% during the survey and stayed above 60% for six of the seven periods. The overall catch rate for all anglers was 2.1/hr, while successful anglers reported a catch rate of 3.2/hr. Catch rates remained fairly consistent over the entire survey.
The expanded estimate of total trout harvest was 1,386 fish. The overall harvest rate was 0.88/hr and harvest rates remained near or above 1.0/hr for the middle five of the seven survey periods. Information from angler interviews and fish population surveys suggested that approximately 28% of the trout observed in the creel (all brook trout and 30 rainbow trout) were caught in tributaries outside the survey area. If total harvest estimates are reduced to account for fish creeled beyond the survey boundaries, then 998 rainbow trout (400/ha) were creeled from the study area during the survey period.
78

~RBT Shocked.(229) ~RBT Creeled (314) mmBKT Creeled(81) I

1

I

/

/

/

/

c 25
...0.-...
-CO 20
::J

/
f~

0.
0 15

>

'-J

(L

/

\.D

"t-
..O.... 10

C

Q)

U
"-

5

Q)

(L

o

a
~~

<59 109 159 209 259 >299

Upper Limit of Length Interval (mm)

Figure 5. Length frequency of rainbow trout in 1991 electrofishing samples from Dukes Creek (RBT Shocked) and of rainbow (RBT Creeled) and brook trout (BKT Creeled) observed in angler creels during a 1991 survey of that stream. Sample size is in parentheses.

Angler characteristics data showed a diverse constituency. Counties of origin included 30 Georgia counties and two other states (FL, NC). Forty-one percent of interviews were obtained from local anglers, residents of either White or Hall counties. The average length ofa completed trip was 2.0 hr. The majority (66%) of anglers used natural baits, while 28% used artificial lures and 6% used artificial flies. Anglers reported spending an average of $21.09 (SE =1.69) on their current trip and estimated taking an average of24 (SE = 1.3) trout fishing trips in 1990. More than 21 % of anglers rated their trip satisfaction as a nine or a 10, with 56% scoring their trips as a seven or eight, 19% scoring five or six, and less than 4% of anglers rating their trip satisfaction under a score of five. Anglers most commonly reported "catching fish" (46% of respondents) and "getting away" (40%) as the most important factors contributing to their trip success. The difference in mean catch rates between the "catching fish" (1.7 troutlhr) and "getting away" (2.4/hr) groups of respondents was significant (one tailed t-test, P= 0.03). A higher proportion (46%) of the "catching fish" group reported catching no fish when compared with the "getting way" group (17%).
DISCUSSION Fish Population Trends
Part of the variation in fish density estimates, even within a stream, is due to at least two factors unrelated to fish population dynamics. The first is the variation in efficiency among different sampling crews or among individuals within a crew. Although efforts were made to standardize sampling effort as much as possible,
80

differences in capture efficiency probably remained. The second and more important factor is the expansion of population estimates to density estimates. For perspective, a 100/ha difference relates to only 10 fish in a 100 m sampling station that is 10m wide. Therefore, relatively small differences in total catch, due to sampling error, true population differences, or other factors, will result in large differences in density estimates. This should help keep comparisons of density and biomass estimates among streams and years in perspective.
Evaluation of trout density and biomass estimates shows that most of the observed variation occurred among streams and not among years. Relative ranking of streams within a group was remarkably consistent from year to year, with high or low density streams generally remaining at their particular end of the range. The only streams with significant annual variation in adult trout densities were upper and lower Dukes Creek in the rainbow group, and lower Jones in the brown group. Annual variations in yay density were observed in upper and lower Dukes Creek in the rainbow group, and Dover and lower Hoods in the sympatric group. Similar variations in adult density were not observed in the sympatric group, however.
Most yearly variations appeared to be stream-specific, which showed that the causes ofthe variation were probably stream-specific and not region-wide. One exception to this occurred in 1994 when seven of nine rainbow streams appeared to have a relatively low density of yay. These low yay densities in 1994 were not reflected in low adult densities in 1995, however.
81

The results of comparing yay with subsequent adult densities indicate that yay abundance is not usually a limiting factor for brook or rainbow trout, and that some other factor such as food or adult habitat is limiting during most years. Strange and Haberra (1995) also failed to find a significant relationship between yay rainbow and adult abundance in similar Tennessee streams. During years with poor reproductive conditions such as flooding, low yay abundance may then limit adult abundance, but such conditions did not occur during this study. It appears that yay abundance in brook and rainbow streams is not a reliable predictor of adult abundance.
In contrast, in brown trout streams the evidence is fairly strong that YOY abundance was a significant limiting factor, and that issue needs to be addressed in any future management plans developed for brown trout streams. Browns seem able to survive in limited numbers in some streams in which brook and rainbow do not, particularly those with high levels of sand, sediment and bedrock, where spawning habitat may be scarce. Supplemental stocking of fingerling browns in such streams may prove beneficial if reproduction is truly limiting. The Walhalla brown trout strain currently in the GADNR hatchery system would be acceptable, since it is basically the same genetic stock as that found in wild populations in Georgia (Norgren et al. 1990). Trout/Habitat Relationships
When trout abundance was compared to habitat variables among stream reaches, several variables consistently showed significant relationships. SASED (sand and sediment combined) and cascade appeared often as negative factors. These results suggest an adverse impact by finer substrate particles on trout populations (Eaglin and
82

Hubert 1993, Alexander and Hansen 1983). The negative influence of cascades on trout abundance is also believable, given the general unsuitability of cascades as fish habitat.
One surprising result was the negative relationship of gradient < 2% with YOY brook
trout, given this species' apparent preferences for low gradient streams in other parts of its range (Kozel and Hubert 1989a). In Georgia trout streams, low gradient often relates to high levels of sand and sediment (Beisser 1996), which would result in poor incubation habitat and explain the statistical relationship we observed. Small sample sizes also limited the power of the analyses.
The correlations of pocket water with YOY trout and plunge pools with adult trout were two relationships that appeared unique to brook trout. Just why this is so is unknown, but R2 values were high, (0.62 and 0.82 respectively) and the data points appeared to be well distributed among streams, so real relationships may exist between these habitat features and brook trout abundance in these streams.
When troutlhabitat relationships were examined at the station level, the results were somewhat different from those seen in reach analysis. This is understandable since some factors such as LWD probably only have a local impact (microhabitat for individual fish), whereas other factors like sand, sediment, or cascades would likely have reach- or stream-wide impacts by influencing spawning and recruitment or by inhibiting migration.
Of particular interest in the station-level analysis was the fact that LWD appears to be a significant positive factor for brook and brown streams, but not for rainbow streams or sympatric streams. Cover was not measured in this study, but LWD as a substrate type probably provided a good index of the relative abundance of cover. In
83

rainbow streams, SASED was again the most important factor, while cascade was the most important factor in sympatric streams.
It appears that trout populations in Georgia streams can be impacted by a variety of variables most of which can change both temporally and spatially. A particular factor that may be limiting trout abundance in one stream during a given year may not be limiting in another stream or in the same stream during another year (Hergen and Hubert 1996). Managers' efforts to "fix" a limiting factor to increase trout densities sometimes fail, because the factor being "fixed" is not limiting on that particular stream during the time it was tested. For example, stocking fingerling trout into a stream that is not currently recruitment limited will probably not add significantly to the fishery. However, if the stream experiences a catastrophic event that wipes out an entire year class, then stocking fingerlings may provide significant benefits. Stream Improvement Structures
The lack of prior population estimates limited a true determination of the effects of instream habitat improvements on trout abundance, as only one year of pre-treatment data was available for Jones Creek. Comparisons to adjacent stream sections were done with the knowledge that habitat variability among stations limited the applicability of this approach. The lack of creel data also limited the assessment, since angler harvest could have prevented detection of trout population responses to habitat manipulation.
Given these constraints, the data still appear to show some significant trends. Brown trout in Cooper, Hedden, and South Moccasin creeks did not respond as well as expected to stream improvement because recruitment, not adult habitat, appeared to be
84

the limiting factor in those streams. Either spawning, incubation, or fry rearing habitat was poor, as mean YOY brown trout densities ranged from 52/ha in South Moccasin Creek to 175/ha in Hedden Creek. In contrast, recruitment was very good in upper Jones Creek (832/ha) and adult habitat may have been the factor limiting brown trout populations there. The effects of higher recruitment versus the effects of habitat improvement on adult brown trout density in upper Jones could not be separated, but the data and the literature (Binns 1994) suggest positive effects from habitat improvement in Jones Creek. This hypothesis can be further tested by either stocking fingerling brown trout into other improved stations or by adding stream improvements to unimproved stations, which would have at least five years of pretreatment data available for companson. Creel Survey
The creel survey showed that the Dukes Creek watershed supported a popular and productive wild trout fishery when compared with other streams in the Southeast. Fishing pressure on Dukes Creek (632 hr/ha) was only slightly below that estimated for the Chattooga River (718 hr/ha; Durniak 1989) and the Lake Lanier Tailwater (802 hr/ha; Martin 1985), two large, popular trout streams. The Chattooga supports both wild and stocked fish, while the Lanier Tailwater relies exclusively on stocked fish. The Dukes Creek estimate would probably have approached those two higher figures if May 1991 had not been an unusually rainy month.
Angling pressure on Dukes Creek exceeded that on Jones Creek, a wild brown trout stream, in the 1960s (438 hr/ha; Van Kirk 1969) and 1970s (156 hr/ha; England
85

1978) and the Chattahoochee (422 hr/ha) and Coleman (363 hr/ha) rivers, wild brook trout streams, in 1977 (England 1979). Angling use of Dukes Creek appeared to fall far short of the effort estimated on stocked trout streams (1,000-3,000 trips/km; Fatora 1983) and on Waters Creek (>5,000 hr/ha; unpublished file data), a trophy trout stream that is supplementally fed. Dukes Creek pressure exceeded the 1995 estimate of 557 hr/ha for the North River, a popular wild rainbow trout stream in Tennessee (Strange and Habera 1996).
The absolute values for Dukes Creek trout catch rates are questionable because they are based on unverified angler reporting. However, this is typically how catch rates are determined. If one assumes equal reporting bias by trout anglers among creel studies, the relative differences in catch rates should hold merit for fisheries managers. Given this assumption, Dukes Creek catch rates of2.1/hr for all anglers and 3.2/hr for successful anglers were high when compared to other Southeastern trout fisheries. Rainbow trout catch rates on Waters Creek (GA) have ranged between 0.3 and 1.0/hr (GADNR, unpublished file data). Combined catch rates of rainbow and brown trout on the Chattooga River (GA&SC) averaged O.32/hr (Dumiak 1989). Fatora (1970) reported a wild trout catch rate ofO.88/hr from Noontootla Creek (GA) during 1964-1969. On eight heavily stocked northeast Georgia streams where most of the catch was likely creeled, harvest rates averaged 0.97/hr (Fatora 1983). Trout catch rates on stocked streams in northwest Georgia averaged 0.98/hr in 1992 and 0,36/hr in 1993 (Beisser 1996). The Dukes Creek estimates were less than the catch rates of3.16/hr and 6.54/hr measured on two stocked North Carolina trout streams managed under delayed harvest regulations
86

(Borawa et al. 1993). The consistently high catch rates reported by Dukes Creek anglers over the entire season were likely the main reason for high satisfaction ratings by anglers.
The survey question on angler motivation confirmed that catching trout was the most important element of a successful fishing trip to Dukes Creek. Those anglers who caught zero or few fish were forthright in stating this. More successful anglers claimed that "getting away" to a north Georgia trout stream was the primary motivation. We believe that if those successful anglers experienced lower catch rates, their responses to our question would have been different. Unfortunately, anglers interviewed failed to distinguish between catch and harvest as motivating factors, so the importance of harvest to trip success could not be measured. Beisser (1996) found that "catching fish" and "catching large fish" were primary motivations for stream selection by northwest Georgia anglers.
Maintaining high trout catch rates should be a major consideration for most Georgia trout management strategies. A brief creel survey on selected weekends during a fishing season may be an economical means to monitor success of mountain trout fisheries. Overall catch and harvest rates, catch and harvest rates of successful anglers, length distribution of the harvest, and anglers' ratings of trip satisfaction may be an adequate set of parameters to monitor the quality of a fishery.
The comparison of the instream trout population versus fish harvest showed that Dukes Creek anglers selectively caught the larger fish in the stream (Figure 5). The population was not estimated in 1990, prohibiting a more accurate assessment of exploitation. The harvestable trout population in 1991 was estimated by adding the 1991
87

harvest to the mean adult density (210 fishlha) in 1991. Recruitment was assumed to offset natural mortality in the adult population. The estimated harvest of400 fishlha would then represent an exploitation rate of 66%.
Although the reproductive capacity of this population is not adversely affected by this harvest rate (Durniak and England 1986), the size structure of the population is impacted (Figure 5). This is not necessarily detrimental when trout harvest defines a "quality" fishing experience, because wild rainbow trout in Southeastern freestone streams appear to have high natural mortality rates and do not "stockpile" or grow much larger than 23-25 cm (Cada et al. 1987; Ensign et al. 1990; Strange and Habera 1995) However, if "quality angling" is defined by users as high catch rates of the larger (20-25 cm) fish in the population, then management opportunities may exist.
On streams where a significant majority of anglers are willing to trade harvest for improved catch rates of adult trout, special regulations may allow the recycling of many "quality" size trout that are typically creeled early in the season. This may be the case in Noontootla Creek, where restrictive regulations maintain essentially a catch-andrelease fishery. The issue is a social one that depends upon characterizing trout anglers' perceptions of "quality" fishing, quantifying the relative demand for each type of trout fishing experience, and diversifying management approaches to meet the expectations of different angler groups (Fatora 1977). A statewide trout angler survey would provide this vital customer information and show where management changes are needed.
88

Fishing Regulation Effects Comparison of trout density, biomass, and size structure among streams showed
that fishing regulations appeared to have varying levels of effectiveness. Charlies Creek was managed under the most liberal regulations, including a year-around season, but still maintained high rainbow trout density and biomass relative to other study streams. Abundance and standing crop were also comparable to 1981-1983 estimates for Charlies Creek (Durniak and England 1986) which were obtained before season liberalization in 1988. Population structure was also very similar to that found in most other rainbow trout streams (Figure 4, Table 22) and to length distributions observed by Durniak and England (1986). No negative biological effects of a year-around trout season on Charlies Creek were observed. This is consistent with current thought among Southeastern trout biologists, who believe there is no biological justification for a closed season (Habera and Strange 1993).
The effects of the artificial lures restriction on Hoods and Jones Creek trout populations were unclear. On Hoods Creek, this regulation did not promote stockpiling of trout, as low population levels, similar to those measured in the 1950s and 1960s (GADNR 1970), persisted. There were higher proportions of larger rainbow and brown trout in Hoods relative to other streams, but this was likely a result of low trout density. Adult brown trout density and biomass estimates for lower Jones Creek were very similar to the 131lha and 18.1 kglha reported for the Chattooga River above Burrells Ford (Durniak 1989), which is managed under general trout regulations. Upper Jones Creek was affected by high yay recruitment and habitat improvement, so regulation effects
89

could not be discerned. Brown trout population structure in Jones Creek was comparable with that in other monitored streams and no clear advantage of the regulation was evident.
Catch-and-release regulations showed some success on Dover Creek. Adult trout density was lower than that in Dukes Creek, but biomass was higher due to a larger average fish size. This could have been due to the influence of brown trout, which accounted for 25% of the Dover catch. Within the adult age class, the relative abundance of rainbow trout >20 cm was four times greater in Dover than in Dukes (Figure 4), which appeared to have better habitat for large fish.
Trout population structures in Noontootla Creek showed that catch-and-release regulations may have benefitted adult rainbow and brown trout. The 20-25 cm size class of rainbow trout appeared protected from harvest and available for catch-and-release. Brown trout>25 cm were also disproportionately represented in samples, compared with other streams. Brook trout in Chester Creek (upper Noontootla), the smaller headwaters, showed no response to the regulation, as no fish> 18.9 cm were captured. The data suggest that lower Noontootla Creek, with large, deep pools relative to most Georgia trout streams, can maintain larger rainbow and brown trout with the aid of restrictive harvest regulations. With few exceptions, the upper limits to growth in this stream may be about 28 cm for rainbow trout and 35 cm for brown trout. Only two rainbow and four brown trout larger than these respective limits were collected. The regulation clearly cannot increase the harvest offish above the 40.5 cm size limit, as many anglers mistakenly believe, due to inherent limits in stream productivity.
90

Comparison to Basinwide Visual Estimation Because variations in stream habitat strongly influence the abundance and size
structure of salmonids (Boussu 1954, Lanka et al. 1987, Scarnecchia and Bergersen 1987, Kozel et al. 1989, Kozel and Hubert 1989a, Kozel and Hubert 1989b, Newman and Waters 1989, Bozek and Hubert 1991, Larscheid and Hubert 1992, Nelson et al. 1992, Hubert and Kozel 1993, Herger and Hubert 1996) researchers have long recognized the limitations of the representative reach approach to sampling stream fish populations (Hankin and Reeves 1988, Dolloff et al. 1993, Dolloff et al. 1997). In general, estimates based on representative reach sampling are accurate for that particular reach, but cannot be extrapolated to represent the entire stream due to the natural variation in stream habitat. Habitat units in streams tend to be unevenly distributed and tend to change as linear factors such as elevation and stream size change. As a result of these limitations, the basinwide visual estimation technique (BVET) of evaluating streams has been developed (Dolloff et al. 1997, Hankin and Reeves 1988).
BVET consists of two phases, a visual estimation phase in which researchers visually estimate habitat fish populations and other stream features for an entire watershed, followed by a verification phase in which the visual estimates are verified and calibrated by making more elaborate measurements of habitat and fish abundance on a small subsample of the habitat units identified.
Representative reach estimates, as conducted for this study, are useful monitoring tools that can detect changes in stream conditions that are basinwide in nature such as flooding, drought, etc. The use of several stations per stream reduces, but does not
91

eliminate, the chance that some localized change will be mistakenly interpreted as a basinwide event. Representative reach estimates do not provide an accurate picture of the status of an entire stream, but provide only a snapshot of the reach sampled.
BVET provides a much more accurate evaluation of the entire stream, and thus is much more useful when attempting to evaluate the relative condition of different streams. The relative costs of each technique need to be considered when developing a sampling plan, but the inclusion of some BVET in future stream sampling would be desirable. Management Implications
Because of the wide variability in fish abundance among streams in this study, the usefulness of extrapolating monitoring station results to other Georgia trout streams is limited. In most cases, local habitat conditions and weather events appear to affect streams individually. Fish population and habitat sampling on each individual stream of interest may be needed in order to adequately evaluate them. The BVET technique may be an efficient way to accomplish this task.
Representative reach sampling still has merit. The relative abundance, biomass, and population structure of Georgia trout streams, as determined in long-term reach sampling, should remain valuable for comparison to streams within and beyond the Southeast region. The effects of a regional climatic event, such as a severe drought or flood,may be extrapolated to other streams if effects are consistent across monitoring sites. Consistencies in year class strength across streams in response to large climatic events have been observed previously (Dumiak and England 1986; Strange and Habera 1995). Electrofishing is still the method of choice to validate BVET estimates, and may
92

indeed be a more efficient means of monitoring cover-seeking fish species. Withinstation comparisons are still valid and may allow a more rigorous assessment of management actions such as instream improvement, regulation changes, or sediment abatement programs.
Creel survey information is valuable. Sampling logistics and budget constraints have limited and will continue to limit the application of "traditional" surveys on wild trout streams in Georgia. Even with a well-designed survey, the undetected harvest of relatively few fish can significantly affect creel estimates, given the low abundance of adult trout. A more cost-effective approach may be to sample spring and early summer weekends and record catch rates, length distribution of harvest, and angler satisfaction.
One side benefit from the study was the educational value of cooperative field sampling. Some USFS staff members and interested anglers gained a better understanding of trout population dynamics, limiting factors, and trout management strategies through participation in fish sampling. Managers and users could relate much better to GADNR data on trout streams after assisting with field collections. Sampling also provided opportunities to communicate, which led to a better understanding of each group's role in trout management and a higher level of credibility for the management agencies.
The presence of rainbow trout in the Totterypole Creek watershed was confirmed. Renovation of that stream and restocking with wild brook trout, once their genetic status is determined (McCracken et al. 1993), should be considered. The threat of another
93

illegal introduction is high because of easy access through the watershed, so costlbenefit analysis of renovation should consider this factor.
Our results strongly suggest that fine sediment and sand adversely impact Georgia trout populations. More work is needed to 1) identify natural sedimentation levels versus increased levels due to man's activities in the watersheds, and 2) determine threshold levels that adversely impact Southeastern fish populations. In the meantime, more effort by land managers is needed to control erosion and reduce sediment inputs into trout streams.
RECOMMENDATIONS 1. Inform anglers, landowners, and land use regulators of the adverse effects of sand and sediment on wild trout populations in Georgia. Support initiatives to protect riparian buffer zones and to control erosion and sedimentation from land disturbing activities. 2. Concentrate future sampling on streams that have not been sampled recently in order to better understand the entire trout resource. Sampling should, however, continue on a small subset of streams used in this study in order to better understand long term trends. Other streams should be sampled on a rotating basis so that all significant wild trout streams will be sampled over a 10-15 year period. 3. Evaluate the streams in this study at least once using the full BVET procedure, including validation. After establishing a baseline, an abbreviated method might be used to monitor trends. Snorkeling as an alternative to electrofishing should be evaluated.
94

4. Stock fingerling brown trout in Hoods and Hedden creeks to test whether recruitment is truly limiting and to increase the sport fishery potential of those streams. Do not stock Ridley Branch and South Moccasin Creek so they can serve as reference sites. 5. Add stream improvement structures to another station on Jones Creek to test the response of brown trout to habitat enhancement. Attempt habitat improvement at selected stations on rainbow and brook trout streams. 6. Conduct a statewide trout angler survey to understand customer demands on the trout fishery. Consider changes to the wild trout management program to accommodate those demands that fall within the realm of biological limits.
REFERENCES CITED
Alexander, G.R. and E.A. Hansen. 1983. Sand sediment in a Michigan trout stream. Part II. Effects of reducing sand bedload on a trout population. North American Journal of Fisheries Management 3:365-372.
Beisser, G.S. 1996. Development ofa stream classification system for evaluating trout stocking in Georgia. Final Report, Dingell-Johnson Project F-36. Georgia Department of Natural Resources, Fisheries Management Section.
Binns, N.A. 1994. Long-term responses of trout and macrohabitats to habitat management in a Wyoming headwater stream. North American Journal of Fisheries Management 14:87-89.
Bisson, P.A., J.L. Nielsen, R.A. Palmason, and L.E. Grove. 1982. A system of naming habitat types in small streams, with examples of habitat utilization by salmonids during low streamflow. In N.B. Armantrout, (ed.). Acquisition and utilization of aquatic habitat inventory information. American Fisheries Society, Western Division, Bethesda, Maryland. Pp 62-73.
95

Borawa, J.C., 1.H. Mickey, and M.S. Davis. 1993. First year assessment of delayed harvest trout regulations. Proceedings of the 47th Annual Conference of the Southeastern Association ofFish and Wildlife Agencies 47:642-649.
Boussu, M.F. 1954. Relationship between trout populations and cover on a small stream. J. Wildl. Manage. 18:229-239.
Bozek, M.A., and W.A. Hubert. 1991. Segregation of resident trout in streams as predicted by three habitat dimensions. Can. 1. Zool. 70:886-890
Cada, G.F., J.M. Loar, and M.J. Sale. 1987. Evidence of food limitation of rainbow and brown trout in southern appalachian soft-water streams. Transactions of the American Fisheries Society 116:692-702.
Couch, W.S. 1985. Physical and temporal factors associated with spawning of naturalized populations of rainbow trout in selected headwater streams of Georgia. Master's thesis, University of Georgia, Athens, Georgia.
Dolloff, C.A., D.G. Hankin, and G.H. Reeves. 1993. Basinwide estimation of habitat and fish populations in streams. General Technical Report SE-83. USDA Forest Service, Southeastern Forest Experiment Station, Asheville, NC. 25 pp.
Durniak, J.P. and R.H. England. 1986. Dynamics of rainbow trout populations in small headwater streams. Final report, Dingell-Johnson Project F-25-13. Georgia Department of Natural Resources, Game and Fish Division, Atlanta, Georgia.
Durniak, J.P. 1989. A fisheries survey of the upper Chattooga River. Final Report, Dingell-Johnson Project F-25. Georgia Department of Natural Resources, Wildlife Resources Division, Social Circle, Georgia.
Durniak, J.P. and W.P. Ruddell. 1990. Evaluation of trout fingerling stocking in small headwater trout streams. Final Report, Dingell-Johnson Project F-25. Georgia Department of Natural Resources, Game and Fish Division, Atlanta, Georgia.
Eaglin, G.S. and W.A. Hubert. 1993. Effects oflogging and roads on substrate and trout in streams of the medicine bow national forest, Wyoming. North American Journal of Fisheries Management 13:844-846.
Egglishaw, H.L. and P.E. Shackley. 1977. Growth, survival, and production ofjuvenile salmon and trout in a Scottish stream, 1966-75. Journal ofFish Biology 11:647672.
96

England, R. H. 1978. Special trout management regulations. Final Report, DingellJohnson Project F-25-4. Georgia Department of Natural Resources, Wildlife Resources Division, Social Circle, Georgia.
England, R.H. 1979. Management regulations for native brook trout streams. Final report, Dingell-Johnson Project F-25-5. Georgia Department of Natural Resources, Game and Fish Division, Atlanta, Georgia.
Ensign, W.E., RJ. Strange, and S.E. Moore. 1990. Summer food limitation reduces brook and rainbow trout biomass in a southern Appalachian stream. Transactions ofthe American Fisheries Society 119:894-901.
Fatora J.F. 1970. Noontootla - a sixteen year creel and use history of a southern Appalachian trout stream under changing management regulations. Proceedings of the 24th Annual conference of the Southeastern Association of Game and Fish commissioners 24:622-637.
Fatora, J.F. 1977. Stream trout fishery management in the southeastern United States. Proceedings of the 30th Annual conference of the Southeastern Association of Game and Fish commissioners 30:280-284.
Fatora, IF. 1983. Creel survey of intensively stocked trout streams. Final Report, Dingell-Johnson Project F-25-10. Georgia Department of Natural Resources, Wildlife Resources Division, Social Circle, Georgia.
Fowler, D.L. 1985. Fish assemblage characteristics of acid-sensitive streams in the southern Appalachian mountains. Master's thesis. University of Georgia, Athens, Georgia.
GADNR. 1970. Annual Progress Report. Statewide Fisheries Investigations, Project F21-2. Georgia Department of Natural Resources Game and Fish Division, Atlanta, Georgia.
Habera, IW., RJ. Strange, and S.E. Moore. 1992. Stream morphology affects trout capture efficiency of an AC backpack electrofisher. Journal of the Tennessee Academy of Science 67:55-58.
Habera, J.W. and R.J. Strange. 1993. Wild trout resources and management in the southern Appalachian mountains. Fisheries 18(1):6-11.
Hankin, D.G. and G.B. Reeves. 1988. Estimating total fish abundance and total habitat area in small streams based on visual estimation methods. Canadian Journal of Fisheries and Aquatic Sciences 45:834-844.
97

Hergen, L.G. and W.A. Hubert. 1996. Comparison of habitat composition and cutthroat trout abundance at two flows in small mountain streams. North American Journal of Fisheries Management 16:294-301.
Hubert, W.A. and S.J. Kozel. 1993. Quantitative relations of physical habitat features to channel slope and discharge in unaltered mountain streams. 1. Freshwater Ecol. 8:177-183.
Hunt, R.L. 1974. Annual production by brook trout in Lawrence Creek during eleven successive years. Technical Bulletin No. 82. Wisconsin Department of Natural Resources, Madison, Wisconsin.
Kozel, SJ. and W.A. Hubert. 1989a Factors influencing the abundance of brook trout (salvelinus fontinalis) in forested mountain streams. 1. Freshwater Ecol. 5:113122.
Kozel, S.J. and W.A. Hubert. 1989b Testing of habitat assessment models for small trout streams in the Medicine Bow National Forest, Wyoming. North American Journal of Fisheries Management 9:458-464.
Kozel, S.J., W.A. Hubert, and M.G. Parsons. 1989. Habitat features and trout abundance relative to gradient in some Wyoming streams. Northwest Science Vol. 63, No.4:175-182.
Lanka, R.P., W.A. Hubert, and T.A. Wesche. 1987. Relations of geomorphology to stream habitat and trout standing stock in small, Rocky Mountain streams. Transactions of the American Fisheries Society 116:21-28.
Larscheid,1. and W.A. Hubert. 1992. Factors influencing the size structure of brook Trout and brown trout in southeastern Wyoming mountain streams 12: 109-117.
Loar, J.M., Editor. 1985. Application of habitat evaluation models in southern Appalachian trout streams. Environmental Sciences Division Publication 2383. Oak Ridge National Laboratory, Oak Ridge, Tennessee.
Malvestuto, S.P. 1983. Sampling the recreational fishery. Pages 397-419 in L.A. Nielson and D.L. Johnson, editors. Fisheries Techniques. Southern Printing Company, Inc., Blacksburg, Virginia.
Martin, C.R. 1985. Hatchery and field performance of six trout strains stocked as catchables in the Chattahoochee River, Georgia. Final Report, Dingell-Johnson Project F-26. Georgia Department of Natural Resources, Wildlife Resources Division, Social Circle, Georgia.
98

McFadden, J.T., G.R. Alexander, and D.S. Shetter. 1967. Nwnerical changes and population regulation in brook trout, Salvelinus fontinalis. Journal of the Fisheries Research Board of Canada 24:1425-1459.
McCracken, G.F., C.R Parker, and S.Z. Guffey. 1993. Genetic Differentiation and Hybridization between Stocked Hatchery and Native Brook Trout in Great Smoky Mountains National Park. Transactions of the American Fisheries Society 122:533-542.
Michaels, RA. 1974. Brook trout production in renovated headwater streams of North Georgia. Master's thesis. University of Georgia, Athens, Georgia.
Nelson, RW., W.S. Platts, D.P. Larsen, and S.E. Jensen. 1992. Trout Distribution and Habitat in Relation to Geology and Geomorphology in the North Fork Hwnboldt River Drainage, Northeastern Nevada. Transactions of the American Fisheries Society 121 :405-426.
Newman, RM. and T.F. Waters. 1989. Differences in brown trout (Salmo trutta) production among contiguous sections of an entire stream. Canadian Journal of Fisheries and Aquatic Sciences 46:203-213.
Nongren, K.G., RA. Dunham, and O. Smitherman. 1990. Identification of brown trout populations in Georgia. Final Report, Dingle Johnson Project F-42. Georgia Department of Natural Resources, Game and Fish Division, Atlanta, Georgia.
Platts, W.S., W.F. Megahan, and G.W. Minshall. 1983. Methods for evaluating stream, riparian, and biotic conditions. General Technical Report INT-138. USDA Forest Service, Intermountain Forest and Range Experiment Station, Boise, Idaho.
Scamecchia, D.L. and E.P. Bergersen. 1987. Trout production.and standing crop in Colorado's small streams, as related to environmental features. North American Journal of Fisheries Management 7:315-330.
Seegrist, D.W. and R. Gard. 1972. Effect of floods on trout in Sagehen Creek, California. Transactions of the American Fisheries Society 101 :478-482.
Seehorn, M.E. 1992. Stream Habitat Improvement Handbook. USDA Forest Service Technical Publication R8-TP 16. Atlanta, Georgia.
Strange, RI and IW. Habera. 1993. Wild Trout Project: 1992 Annual Report. Tennessee Wildlife Resources Agency and the University of Tennessee, Knoxville. Publication No. RI1-2216-76-001-93. 134 pp.
99

Strange, R.J. and J.W. Habera. 1995. Wild Trout Project: 1994 Annual Report. Tennessee Wildlife Resources Agency and the University of Tennessee, Knoxville. Publication No. Rll-2217-39-001-95. 214 p.
Strange, R.J. and J.W. Habera. 1996. Wild Trout Project: 1995 Annual Report. Tennessee Wildlife Resources Agency and the University of Tennessee, Knoxville. Publication No. Rll-2217-93-001-96. 223 p.
Van Deveter, J.S. and W.S. Platts. 1985. A computer software system for entering, managing, and analyzing fish capture data from streams. Research Note INT-352. USDA Forest Service, Intermountain Forest and Range Experimant Station, Ogden, Utah.
Van Kirk, R.R. 1969a The effects of the construction of a dry dam on a brown trout population. PhD dissertation. University of Georgia, Athens, Georgia.
Van Kirk, R.R. 1969. Trout Habitat Study in Selected Streams on the Chattahoochee National Forest. University of Georgia, Athens, Georgia.
Waters, T.F. 1983. Replacement of brook trout by brown trout over 15 years in a Minnesota stream: production and abundance. Transactions of the American Fisheries Society 112:137-146.
White, G.C., D.R. Anderson, K.P. Burnham, and D.L. Otis. 1982. Capture-recapture and removal methods for sampling closed populations. Technical publication LA8787-NERP. Los Alamos National Laboratory, Los Alamos, New Mexico.
100