Trout Research Associates
Your Subtitle text
Current Research Paper

Seasonal Movement, Habitat Use and Growth Rates of Brook Trout in the Upper Mersey River Watershed, Nova Scotia

Gary N. Corbett a, W. Reg Baird b, and Douglas G. Potter c

a. Box 103, Greenfield, Nova Scotia, Canada, B0T1E0
b. Box 22, Clementsvale, Nova Scotia, Canada, B0S1G0
c. Box 33, Clementsvale, Nova Scotia, Canada, B0S1G0

Abstract. - We used tags and radiotelemetry to monitor year round movements and habitat use of brook trout Salvelinus fontinalis in a 245 Km² area of a Nova Scotia watershed composed of a main river, stream tributaries and a lake. Between June 7th of 2005 and October 25th of 2006 twenty-two brook trout aged 3 to 5 years ( 230 mm - 350 mm) were implanted with radio tags and monitored for up to 333 days. Monitoring was done by foot, boat and by float plane. This information was supplemented with data from 121 recaptured trout from a sample of 1310 individuals marked with serial numbered tags between 2002 and 2007 in the same watershed. Brook trout moved extensively within the watershed with individuals moving as much as 85 km over a 303 day period. Daily movement patterns changed seasonally and varied from 100 m to 13 km over a 24 hr period. Age and size of sampled trout did not appear to be a factor in distance traveled or habitat used. Radio tagged trout wintered in Kejimkujik Lake; in the spring as early as April they began moving around the lake or into streams; some individuals made long distance journeys of up to 11 km over a 3 day period. By the first of May most trout were feeding in riffles and runs in the rivers or along the shore of the lake. When water temperatures approached 20E C they moved quickly to deeper water in the lake or to groundwater springs in the rivers covering as much as 13 km in one day. Summer refuges were found to have water temperatures below 20E C and as low as 9E C. Those in rivers had oxygen concentrations between 5 and 9.7 mg/L and they often had improved pH conditions over surface waters. Trout moved as much as 15 km to spawning areas in the rivers during the fall and to wintering sites by late December. Annual growth was largely confined to the months of April, May and June. Trout grew an annual average of 6.3 cm between age one and two, 4.0 cm between age two and three, 3.7 cm between age three and four and 4.4 cm between age four and five. Of the twenty-two radio tagged trout, one died shortly after surgery, 6 were predated by otter or mink, 4 known or suspected to have been killed by anglers and the fate of 4 was unknown. Premature battery failure was suspected in the 4 unknowns. Seven trout were still alive when transmitters failed after normal life span. The Mersey River system exhibits high summer water temperatures (> 25E C), has low pH (5.0) and feeding, spawning, summer and winter refuges often separated by long distances. Many brook trout must travel extensively to meet their physiological and seasonal habitat requirements. This study demonstrated that brook trout move in and out of Kejimkujik National Park and National Historic Site indicating that many trout cannot meet all of their annual requirements within the park. As landscapes outside of the park become more fragmented, some seasonal habitats may be jeopardized by incompatible and damaging land uses. In the longer term, climate change could jeopardize favorable conditions in summer refuges. Brook trout should be managed on a watershed level considering these seasonal movements and their critical habitats.

Introduction

   Salmonids are noted for traveling long distances. Salmon may travel hundreds of miles during seasonal migrations. Several studies have documented the mobility of brook trout Salvelinus fontinalis and journeys of up to 100 km by some individuals (Castonguay et al. 1982; Curry et al. 2002; Bonney 2007). Such movements are attributed to locating suitable foraging and spawning habitats. Many of these populations, however, are anadromous. Larger juvenile and adult anadromous brook trout typically travel from freshwater to estuaries and disperse over a few kilometers of the marine environment in spring and early summer, return to freshwaters by late summer, commonly returning to the estuary in fall and repeating this cycle over several years (Curry et al. 2002). Unfortunately there are few published studies on the daily and seasonal movements of wild non-anadromous brook trout. The most extensive research was conducted in the U. S. State of Maine. Studies of marked and radio-tagged trout by Maine hydro power and government biologists recorded distances of as much as 75 miles by individual brook trout. Movement was greatest in the spring when water temperatures were cool and flow was diminishing. Brook trout moved freely throughout the river while water temperatures remained cool, congregated in nearby lake environments when waters warmed and most spent the winter in lakes. They tended to remain in one area for an extended period and then move long distances in a short period (Bonney 2007).
   No brook trout have ever been radio-tagged and tracked in the Canadian province of Nova Scotia and few studies of tagged trout have studied the seasonal movements, habitat use and growth rates of a non-anadromous brook trout population in the province. Most large river systems in Nova Scotia flow into the Atlantic Ocean and brook trout can freely access their marine estuaries. An exception is the Mersey River system in southwestern Nova Scotia above Lake Rossignol. At least three hydroelectric dams prevent the upstream movement of fish because they lack fishways. Kejimkujik National Park and National Historic Site is located further upstream nearer the headwaters of the Mersey River. This 381 km² park covers 76 lakes and streams that protect a healthy non-anadromous brook trout population. This population, however, can access an additional 850 km² of watershed area that is under a variety of land uses, predominately forestry operations involving clear-cutting. Little research has been conducted on brook trout populations that move between protected and unprotected landscapes. The objective of this study was to characterize the seasonal movements, habitats used and growth rates of brook trout throughout the year in a population that does not exhibit anadromy. A further objective of this research was to study a population that mostly resides in a protected area but with access to outside areas where they could be impacted by development and land uses such as forestry.


Study Area

   The upper Mersey River (44E25N N, 65E15NW) is a section of the Mersey River in Nova Scotia, Canada which empties into Liverpool Bay on the Atlantic Ocean. The area of the entire watershed is 1963.2 km² with a main stem length of 109 km. The watershed has an extensive system of lakes and tributaries flowing into the main stem. The lower part of the Mersey River empties Lake Rossignol, the largest freshwater lake in Nova Scotia with a surface area of 180 km². This lake was impounded in 1929 with a dam for the purpose of providing a reservoir of water to feed 6 hydroelectric dams now located along the lower river. There are no fish ladders over the top two power dams effectively preventing all fish passage upriver and into lake Rossignol. Kejimkujik Lake ( 2453 ha ) is located above Lake Rossignol and joined by a 8.2 km section of the main stem. The Upper Mersey River and two smaller rivers, Little River and West River flow into Kejimkujik Lake along with a number of brooks. This part of the watershed is located within the boundaries of Kejimkujik National Park and National Historic Site. The headwaters on the main stem drain a 425.5 km² area above the park which enters the park at Maitland Bridge. The upper Mersey River brook trout population was chosen for this study because the population was thought to move seasonally in and out of the park. The park is protected from development and land use while the watershed above the park has been extensively logged and is fragmented with many roads. Some housing development and farming also occurs. The upper Mersey River width varies from 20 m to 100 m and .5 m to 3 m in depth. River gradient is generally gentle and the substrate is composed of boulder, cobble and gravel. Temperatures in the river typically reach 28E C or more in summer. The watershed has poor productivity with total dissolved solids as low as 20 mg/L and high acidity averaging pH 5. An active recreational trout fishery exists in the watershed and twelve fish species inhabit the area, however, predaceous smallmouth bass
 ( Micropterus dolomieui ) and chain pickerel ( Esox niger ), common in some adjacent watersheds do not occur. The only other salmonid species in the watershed are a small introduced population of Brown Trout ( Salmo trutta ) stocked between 1937 and 1967 and a remnant population of Lake Whitefish
 ( Coregonus clupeaformis ) introduced around 1900.

Methods

   Between 2002 and 2007 thirteen hundred and ten brook trout were captured in fyke nets or by volunteer anglers and marked with serial numbered size 1 monel metal tags on the posterior edge of the operculum and released at the capture location. Each fish was measured for fork length (nearest mm) weighed (nearest g) and sex determined visually where possible. A sample of scales were removed for aging and exact capture location was recorded. Water depth from a gauge and temperature were recorded from a reference location in the river on a daily basis around precipitation events and regularly every 3 days when there were no rains. Levels of oxygen in the water, pH and conductivity were measured weekly in the river. Recaptured trout were measured, weighed, examined for condition, tag serial number and exact recapture location recorded.
   Between June 7th 2005 and October 25th 2006, twenty-two brook trout were captured in the study area and surgically implanted with Lotek model MBFT-4 radio tags having a weight of 7.7 g. or model MCFT-3 with a weight of 6.7 g. These units were programmed with a 50% duty cycle for a minimum life span of 293 days. The tags transmitted from 8 am to 8 pm for daytime monitoring . In June of 2005 seven were implanted with radio tags and in November, two more were implanted and released in the upper Mersey River. In May of 2006, seven more trout were implanted with an additional one completed in early June. Again in the fall, five were implanted in the upper Mersey River in late October. The radio-tagged trout were aged 3 to 5 (scale aged) and ranged in size from 230 mm to 350 mm FL and from 150 to 550 g. Surgical procedures were similar to McKinley et al. (1994) and Baird and Krueger (2003). Brook trout were anesthetized in a solution of 5 parts per million of clove oil in water. Trout were immobilized within 5 minutes and surgery varied from 3 to 10 minutes. Twelve trout were held in recovery boxes in the river for 24 hours, four for 4 hours and six for only 1 hour before release. There was one mortality shortly after surgery and before release.
   A Wildlife Materials model TRX-2000s receiver and AVM model LA12-Q receiver with 3 element yagi antennas were used to track the movements of trout. Tracking was done on foot, by truck with roof mounted antenna, from a canoe and boat as well as from the air. A 3 element yagi antenna was mounted vertically on the wing strut of a modified Taylorcraft floatplane and pointed at a 45 degree angle to the ground. Aerial tracking was done from an initial altitude of 366 m to locate the signal and altitudes down to 150 m to pinpoint the trout’s location. Tracking was done along rivers by following the main-stem and tributaries. Locations were found when there was a rapid weakening of the signal as the plane flew past the transmitter. Over lakes a grid pattern was flown on compass bearings as transmitter frequencies were scanned. Once a signal was received the transmitter location was found by triangulation. Locations were marked on maps and gps coordinates recorded as utms. Ground tracking was then done to obtain a more precise gps location. Tracking of each radio tag was done daily during the open water season and once per week during winter ice conditions. Data collected at trout locations included habitat type, physical measurements, water depth, temperature and in summer refuges included dissolved oxygen and pH. Trout locations and distances traveled were mapped and calculated in ArcGIS desktop 9.1. Movements were correlated with changes in water levels in the main river and temperature changes.
   Anglers were informed about the presence of trout marked with serial numbered tags or implanted with radio-tags. Notices were posted at common fishing locations throughout the national park as well as outside of the park. Anglers were given an informative pamphlet with their fishing licence and many were contacted on the river. They were asked to report the catch of tagged trout and collect data on tag serial numbers and exact location of the catch.

Results

Size at Age of Kejimkujik Brook Trout

   We compiled records of fork lengths and weights of 1889 aged brook trout recorded in the national park between 1994 and 2004 ( table 1 ). Trout were aged one to seven; only four were six and only one was seven years old since trout rarely live more than five years in the study area. Few trout aged one year were measured because most of the sample came from angler’s catches. Most trout caught by anglers were at least 2 years old.

Table 1. Average ( SD) fork length (cm) and weight (g) of brook trout recorded between 1994 and 2004 in Kejimkujik National Parks and National Historic Site (sample sizes in parentheses)

 

 

Age

Fork length

Weight

1

14.7 1.2 (25)

52.8 20 (5)

2

21 1.7 (539)

137.4 36.2 (348)

3

25 2.6 (612)

228.8 73.4 (497)

4

28.7 2.2 (284)

331.5 82 (229)

5

33.1 2.3 (75)

512.6 119.6 (60)

6

35.6 2.2 (4)

640 125.4 (4)

7

39 (1)

850 (1)

Growth Rates of Recaptured Trout

   We analyzed growth rates for 192 tagged and recaptured brook trout that were measured, weighed and scale aged. Marked individuals that were recaptured after varying periods of time permitted accurate seasonal and annual growth measurements. No standard rate of growth was observed as growth would be dependant upon the individual fish’s foraging ability, food quality and availability. Based on table 1., trout grew an annual average of 6.3 cm between age one and two, 4.0 cm between age two and three, 3.7 cm between age three and four and 4.4 cm between age four and five. Growth varied seasonally with the highest rates observed in spring and the lowest over the fall and winter period. Measurements of seventy-seven mark and recaptured trout during the spring foraging season from late April to early June permitted estimates of daily growth in length. Those trout averaged an increase of 0.57 mm in length per day (SD = 0.37) however, some exceptional individuals gained over 1.0 mm per day, one 2.1 mm per day while others showed no growth in length at all for several days. Measurements of forty-eight mark and recaptured trout during that same season also permitted estimates of daily growth in weight. Those trout averaged an increase of 3.6 grams in weight per day (SD = 3.5) with some individuals gaining more than 10 grams per day. A five year old trout gained 16.7 grams per day or 3.7% of it’s body weight per day over a six day period. A four year old gained 15 grams or 5.2% of body weight per day over a four day period. Growth appeared to be put on in spurts, varied by individuals and daily length versus weight gain was not highly correlated ( r = 0.32). During summer, trout exhibited little or no growth in length at all and most also lost weight. Annual growth was largely confined to the months of April, May and June. Some exceptional annual growth rates were also observed. Three trout gained 8.5 cm in length between age 2 and 3. One trout gained 7.5 cm between age 3 and 4 and a four year old trout gained 7.3 cm by the time it was five years old.

Movements of Trout Marked with Tags

   Between 2002 and 2007 thirteen hundred and ten brook trout were tagged in the study area with serial numbered monel metal tags. Most tagged trout were aged 2 to 5, however a small sample of twenty-five age 1 trout were tagged but none recaptured. A total of 121 tagged trout were recaptured in locations both inside and outside of the study area for a tag return rate of 9.2 %. Since recapture data depended largely upon angler reporting; the actual return could have been much higher as reports indicated that many anglers failed to report catching tagged trout. Reports of tagged trout from anglers were limited to the fishing season which ran from April 1st to August 31st (September 30th outside park). Trout capture by researchers continued until ice up in December but did not begin again until April. Sixty-three percent of trout tagged were recaptured at the same location within 38 days. Some individuals were recaptured several times. Sixteen percent were recaptured in the spring or fall of the following year in the same location. Recaptures of tagged trout, however, also demonstrated movement between the lower Mersey River, Kejimkujik Lake, upper Mersey River and Little River up to Big Dam Lake. Fifteen percent of recaptures were significant movements of between 8 and 27 km between waterbodies. One trout moved 20 km in 14 days between the lower and upper Mersey River and another moved 27 km in 35 days from Mill Falls to Big Dam Lake. No correlation between age or size and distance moved was apparent. Tags were generally retained for one year and scars were observed on some fish were the tag had dropped off. Recaptures of tagged trout confirm that many move seasonally and between tributaries in the watershed. The recapture of two trout in the Mersey River above the park indicated that they move in and out of the park. One was recaptured 13 km above the park boundary. The comparative results of the radio-tagged trout tracking seems to indicate that more tagged trout were probably caught by anglers outside of the park but were unreported.

Monitoring of Radio-Tagged Trout

   A clear signal from the radio-tagged brook trout could be received over a maximum distance of 5 km from an unobstructed vantage point such as a hill or scenic viewing tower that was located overlooking Kejimkujik Lake. At lower elevations or from a vehicle along the park roads, 2 km was possible. When tracking trout on the lake, a signal could be received from up to 2 km away. This was reduced to 1 km along rivers in a canoe and a maximum of 0.5 km when obstructed by vegetation. When a transmitter could not be located from the ground, a floatplane was used to relocate it. The maximum distance that a transmitter signal was received from the floatplane was 12 km at an altitude of 366 m above ground level (AGL). In this unique situation, the fish had been killed and eaten by an otter and the bare transmitter was laying in 25 cm of water. On average, radio-tagged trout signals were first received from a distance of 1.5 to 2.5 km at 366 m AGL. This average held true for lakes or rivers, although not for deep sections of lakes. If the fish were in deep areas of lakes the plane had to be within 0.5 km of the transmitter. When fish were moving rapidly such as the pre-spawn migration period, many hours of flying were often required to relocate them from their previous location. Although there were limited opportunities to monitor fish under the ice from the air, ice cover did not appear to hinder signal reception.

Movements of Radio-Tagged Trout

   Twenty-two brook trout aged 3 to 5 years ( 230 mm - 350 mm) were captured, implanted with radio-tags between June 7th 2005 and October 25th 2006 and released in the upper Mersey River. Depending on their fate, the trout were tracked for 161 to 333 days. Five trout tracked for 303 to 333 days covered a distance of 46.5 to 85 km. Four trout followed for 201 to 224 days moved between 18.5 and 76.7 km. One trout that was only tracked for 167 days covered a distance of 68 km. There was no correlation between distance moved and the age or size of radio-tagged trout. Of the twenty-two radio tagged trout, one died shortly after surgery, 6 were predated by otter or mink, 4 known or suspected to have been killed by anglers and the fate of 4 was unknown. Premature battery failure was suspected in the 4 unknowns. Seven trout were still alive when transmitters failed after normal life span.
   Seventeen radio-tagged trout tracked through at least two seasons used a variety of habitats scattered throughout the study area. In April, May and Early June they foraged for food in riffle, runs and pools in the rivers or along the shore of Kejimkujik Lake. When water temperatures approached 20EC they moved quickly covering as much as 13 km per day to a deep thermocline in Kejimkujik lake or to ground water thermal refuges in the rivers. Individual trout were observed to move as much as 38 km to be located in small site-specific cold water springs indicating memory and a strong homing instinct towards those sites. Three trout moved outside of the national park into summer refuges adjacent to forestry clear-cuts. These were located 22, 12 and 8 km above the park. Trout stayed in those sites throughout the summer months or ventured outside for only brief periods of time. In September, when water temperatures in the river fell below 20EC the radio-tagged trout left the summer refuges and from late September and October moved gradually into spawning areas. Eleven trout moved into the Mersey River spawning areas. At least five of those traveled upriver to spawning areas above Kejimkujik National Park and National Historic Site. Two trout moved into Rogers Brook and one into Little River. Two other trout may have spawned in Kejimkujik Lake. By mid-November nine radio-tagged trout moved into Kejimkujik Lake for the winter. Some trout in the Mersey River moved as much as 17 km in 2 days to reach the wintering area. One trout moved into and spent the winter in Grafton Lake. Trout moved little during the winter. In January of both 2006 and 2007 the ice thawed on Kejimkujik Lake providing relatively open water and allowed for floatplane aerial monitoring. Open water, however, did not appear to stimulate much movement as all except one fish remained in or close to the winter location. The one that moved, swam 2 km from it’s winter site and remained there until ice-out on March 29th. In April, trout movement resumed as soon as ice left the wintering area.
   Seasonal movements were not correlated with age or size of trout but movement in and out of summer refuges was correlated with temperature. Changes in water level in the Mersey River was not correlated with any movement pattern however trout seemed to move more during rising levels especially during movement into the fall spawning areas.

Summer Refuge Habitats

   Radio-tagged trout spent the warm summer months in deep areas of Kejimkujik Lake with a distinct thermocline, coves influenced by cold spring water or streams with ground and spring water seepage. Summer refuges of eleven trout were examined. The sites were scattered around the study area and separated by a much as 30 km. Several trout moved between sites that were close to each other ( < 1 km) during the summer; two trout used two different sites in the Mersey River in the park while two used two sites in the river upstream from the national park. Another moved in and out of a site in the West River when conditions were suitable. Two trout moved twice between the thermocline and a spring site in Kejimkujik Lake. Movement into and out of refuges was correlated with changes in water temperature. We measured surface water temperatures as high as 27 EC and pH as low as 4.4 adjacent to these sites. The physical and chemical characteristics of the refuges ( Table 2 ) varied from marginal for trout survival to high quality conditions. Springs were the best with temperatures near source as low as 9 EC and 100 % oxygen saturation. Improvements in pH by as much as a full point above surface waters were recorded in springs. Areas of ground water seepage in stream sites had good to marginal conditions and varied from 60 to 3000 m² in water surface area. A 10,000 m² deep basin in Kejimkujik Lake formed a thermocline during the summer months. Radio-tagged trout appeared to suspend themselves where temperatures and oxygen levels were favorable.

Table 2. Physical and chemical conditions of known summer refuges in the upper Mersey River watershed

 

 

Location

Surface area (m²)

Max depth (m)

Temperature

E C min

pH

Dissolved oxygen mg/L

Kejimkujik Lake 1

10,000

thermocline

n/a

22.1 @ 8 m

14.5 @ 10 m

4.9

5.8 @ 8 m

3.2 @ 10 m

Kejimkujik Lake 2

800

spring

7.5

13.5 @ 5 m

5.7

5.6 @ 5 m

Kejimkujik Lake 3

2500

cove

1.7

18.6

4.9

8

Mersey River 1

400

2.5

18.6

5.7

5.2 @ 2 m

Mersey River 2

300

2.8

13.7

4.9

5.3 @ 2 m

Mersey River 3

60

1.5

20.2

6

5.6

Mersey River 4

3000

3.9

13.4

5.7

6.8

Mersey River 5

100 spring

1

10.9

5.8, surface 4.9

9.7

Tributary Br 1

2000

2.3

15.4

5.3

5

Tributary Br 2

2000

1.7

15.9

5.3

6.4

Little River

2385

2.5

13.4

4.8

9.1

West River

360

2.5

9.2

5, surface 4.4

9.6

Discussion

   Seasonal movements of brook trout in the upper Mersey River were generally similar to previous studies of non-anadromous trout populations. They used a large lake or deep pools in the rivers as winter refuges and moved from them in early spring into river reaches or along lake shores to forage. In summer they used deep lake areas or springs in the river as thermal refuges moving into the rivers and brooks in the fall to spawn. Previous studies documenting the complex seasonal pattern of movement in a non-anadromous brook trout population have been largely confined to the U.S. State of Maine. A study of 24 radio-tagged brook trout on the Rapid River in western Maine found that they moved freely throughout the river while water temperatures remained cool. Brook trout congregated in nearby lake environments when waters warmed and most spent the winter in lakes (Bonney 2007). Another study involving 36 radio-tagged brook trout on the Kennebec River in Maine (FLPE 2000) found that they tended to remain in one area for an extended period and then move long distances in a short period.
   Spring movement from wintering sites by Mersey River brook trout began soon after ice left the lakes. A 26.7 cm radio-tagged brook trout moved 11 km from Kejimkujik Lake and up the West River over a period of 3 days from March 31st to April 3rd in 2006. Three other radio-tagged trout moved about in Kejimkujik Lake before transmitter battery life was expended (implanted 10 months previous). Results from marked trout and tracking of radio-tagged trout indicate that during late April, May and early June most trout forage for food in riffle, runs and pools in the rivers while a small percentage of them forage along the shores of Kejimkujik Lake. Few recaptures of marked trout were made by anglers in Kejimkujik Lake and only two radio-tagged trout spent the spring moving around the lake. When water temperatures approached 20EC in late spring, eleven radio-tagged trout moved quickly covering as much as 13 km per day to distant thermal refuges. Individual trout were observed to move as much as 38 km to be located in small site-specific cold water springs indicating some memory of them and a strong homing instinct towards those sites. Four radio-tagged trout moved downstream and across Kejimkujik Lake to a deep section with a distinct thermocline. Four other radio-tagged trout traveled downstream and across Kejimkujik Lake and up the West River or Little River to find thermal refuges. Three trout moved upstream outside of the national park into summer refuges adjacent to forestry clear-cuts. These were located 22, 12 and 8 km above the park. Because brook trout require relatively cool summer water temperatures, cool water refuges may be important for survival in streams that have lethal water temperatures (>25EC) during the summer (Baird and Krueger 2003). Curry et al (2002) reported that large migrant trout in the Kennebecasis River occupy deep pools and runs with overhead cover where groundwater maintained stream temperatures at about 15E C. Torgersen et al (1999) states that during high temperature events, fish often seek out thermal refugia near cold water springs or at the mouth of groundwater tributaries. Salmonids are able to use specific orientation cues to return precisely to their home section following displacement and field studies have shown that homing accuracy can be very high for individual fish displaced hundreds or thousands of metres (see review by Bélanger and Rodriguez 2001 ). Trout stayed in those sites throughout the summer months or ventured outside for only brief periods of time. Venturing outside the refuge was consistent with Baird and Krueger (2003) who found that individual brook trout may not constantly occupy cool water areas but instead may periodically move out of refuge areas to feed and then return. Occasionally, in their study, they were observed on feeding expeditions in water temperatures up to 25.4E C.
   Carline and Machung (2001) found that the mean critical thermal maximum for wild brook trout in their Pennsylvania study area was 28.7E C. Summer water temperatures in the Mersey River and surface waters of lakes in the area can often reach or exceed this value. Mersey River water is stained brown from tannic acids in the many bogs found in the headwaters and the river is very shallow especially during low summer water level period. Sun exposure and warm weather can quickly increase water temperatures to critical values for trout. We studied the physical and chemical conditions of thermal refuges used by radio-tagged brook trout (table 2). Conditions varied from excellent to marginal; springs at the source had temperatures as low as 9.2E C and saturated oxygen levels. Groundwater sites in the rivers had temperatures between 13E C and 20E C and oxygen levels between 7 and 5 mg/L, the latter being marginal for brook trout. The size of these sites varied from 60 m² and only 1 m depth to 3000 m² and a maximum of 4 m in depth. Sites in Kejimkujik Lake were found to be very marginal with trout being squeezed between a thin layer of 22E C water with 5.8 mg/L of oxygen during the warmest summer periods. Tannic acids in the Mersey River also acidify the waters to an average of pH 5. We found improvements of as much as half the acidity (pH 6) within some thermal refuges in the river.
   In September, when water temperatures in the river fell below 20EC the radio-tagged trout left the summer refuges and from late September through October moved gradually into spawning areas. Eleven tracked trout moved into the upper Mersey River and five of those traveled further upriver to spawning areas above the national park. Two other trout stayed in the park and moved into Rogers Brook and one moved into Little River. Two other radio-tagged trout may have spawned in Kejimkujik Lake. We were unable to describe the physical characteristics of individual spawning sites due to high water levels and inclement weather. Our observations from the shore, however, indicate that these were areas of ground-water discharge in the main river near brooks or steep hills. By mid-November nine of the radio-tagged trout moved into Kejimkujik Lake for the winter. Some tracked trout in the Mersey River moved as much as 17 km in 2 days to reach the wintering area. One trout moved into and spent the winter in Grafton Lake. Trout moved little during the winter. In January of both 2006 and 2007 the ice thawed on Kejimkujik Lake providing relatively open water and allowed for floatplane aerial monitoring. Open water, however, did not appear to stimulate much movement as all except one fish remained in or close to the winter location. The one that moved, swam 2 km from it’s winter site and remained there until ice-out on March 29th. During the course of this study, five trout tracked for 303 to 333 days covered a distance of 46.5 to 85 km. Four trout followed for 201 to 224 days moved between 18.5 and 76.7 km. One trout that was only tracked for 167 days covered a distance of 68 km.
   In our small sample of 22 radio-tagged brook trout, most of them foraged during spring in the upper Mersey River river reaches but sought refuge from extreme summer water conditions in a variety of habitats in widely separated locations. In the fall, many of them regrouped back in the river for spawning and later for overwintering in Kejimkujik Lake. A defining characteristic of the brook trout species is it’s plasticity of life history tactics (Power 1980). In a population, such as the one in the Mersey River possibly limited by a scarcity of thermal refuges, the ability to rapidly travel relatively long distances coupled with an accurate homing ability may be a critical survival tactic. We were unable to determine how many trout utilize these sites nor to describe any competition for a position in these sites that might occur. Our radio-tagged trout, however, were able to remain in the sites over the summer and only one individual was observed to move between small adjacent refuges in the upper Mersey River. This individual was aged 3 years and the smallest trout that we radio-tagged (23 cm, 150 g).
   Growth rates of 192 tagged and recaptured brook trout were analyzed. Scale aging enabled us to determine mean annual growth and measurements of seventy-seven marked and recaptured trout during the spring foraging season from late April to early June permitted estimates of daily growth in length. Compared to other mainland Nova Scotia non sea-run brook trout populations studied by MacMillan and Crandlemere (2005), Mersey River brook trout had annual length growth rates 30% higher at age 1 and 12% higher from age one to age two. Mersey River brook trout averaged about the same length at age 3 but fell behind by 7% from age 3 to age 4. Age 3 Mersey River brook trout, however, averaged 1.5 cm longer than age 3 brook trout in Halifax and Guysborough counties. No data on age 5 non sea-run brook trout were available for comparison but sea-run 5 year old brook trout were significantly longer (0 38.8 cm FL) than Mersey River trout. Daily growth rates of Nova Scotia brook trout were previously unreported. During the spring foraging season from late April to early June brook trout averaged an increase of 0.57 mm in length per day (SD = 0.37) however, some exceptional individuals gained over 1.0 mm per day for several days, one gained 2.1 mm per day while others showed no growth in length at all for several days. Measurements of forty-eight mark and recaptured trout during that same season also permitted estimates of daily growth in weight. Those trout averaged an increase of 3.6 grams in weight per day (SD = 3.5) with some individuals gaining more than 10 grams per day. A five year old trout gained 16.7 grams per day or 3.7% of it’s body weight per day over a six day period. A four year old gained 15 grams or 5.2% of body weight per day over a four day period. Growth appeared to be put on in spurts, varied by individuals and daily length versus weight gain was not highly correlated ( r = 0.32). During summer, trout exhibited little or no growth in length at all and most also lost weight. Annual growth was largely confined to the months of April, May and June. Some exceptional annual growth rates were also observed. Three trout gained 8.5 cm in length between age 2 and 3. One trout gained 7.5 cm between age 3 and 4 and a four year old trout gained 7.3 cm by the time it was five years old.
   We demonstrated that brook trout move seasonally in and out of Kejimkujik NP&NHS to seek foraging, summer thermal refuges and spawning areas. Some summer refuges and spawning sites are adjacent to forested areas that were clear-cut in the past or may be cut again in the future. Further, the river section within the national park is not far below forestry operations. Land uses near trout habitats or long term impacts such as climatic change could affect trout populations both within and outside of Kejimkujik. A rich literature on research conducted in the United States and Canada in forests under even-aged management (i.e., clear-cutting) suggests that it can alter the landscape in ways that adversely affect aquatic habitat ( VanDusen and Huckins 2005). Bonney (2007) states that land uses such as agriculture, forestry and development may alter natural flow regimes and clear-cutting greatly exacerbates flooding. He further states that flooding may eliminate entire brook trout year classes through destruction of eggs and fry. The transport and deposition of sediment can also blanket and destroy invertebrate populations that trout depend upon for food. Climate change is a major long term threat to brook trout populations. Results of research by MacMillan et al. (2005) indicate that climate warming could potentially have the most profound impact on cold water habitat structure in the future for Nova Scotia streams. A 2EC increase in summer average water temperature could result in a 50% loss in the number of cool water sites. Bonney (2007) states that coldwater refuges will become more important, but brook trout crowded into these areas will suffer from physiological and competitive stress, and be more vulnerable to diseases and predation. The impact of forestry operations on cold water habitat in the upper Mersey River and climate change on thermal refuges in Kejimkujik Lake require further examination and long term monitoring. These and the protection of cold water habitat will be important steps toward preserving brook trout populations in the Mersey River watershed.

Acknowledgments

We wish to thank Kejimkujik National Park and National Historic Site, the Mersey Tobeatic Research Institute, Nova Scotia Department of Fisheries and Aquaculture and Trout Nova Scotia for logistic and financial support. Kejimkujik Ecologist Chris McCarthy, Database Manager Sally O’Grady and NSCC student Mike Townsend assisted in the compilation and analysis of data and production of Figure 1. Hedy Armour provided research papers. Many anglers volunteered their time to catch trout for tagging. In particular we wish to thank Robin Olsen, Barbara Baird, Robert Emin, Roy Bertaux and Trout Nova Scotia volunteers. Summer students Cody Joudry , Nicholas Whynot and Liam McNeil assisted in field work.

 
References

Baird, O. and C. Krueger. 2003. Behavioral thermoregulation of brook and rainbow trout: comparison of summer habitat use in an Adirondack river, New York. Transactions of the American Fisheries Society 132:1194-1206

Bélanger, G. and M. A. Rodriguez. 2001. Homing behaviour of stream-dwelling brook charr following experimental displacement. Journal of Fish Biology 59:987-1001

Bonney, F. 2007. Squaretails: biology and management of Maine’s brook trout. Maine Department of Inland Fisheries and Wildlife, Augusta, Maine.

Carline, R. F. and J. F. Machung. 2001. Critical thermal maxima of wild and domestic strains of trout. Transactions of the American Fisheries Society 130:1211-1216

Castonguay, M., G. J. Fritzgerald, and Y. Cote. 1982. Life history and movements of anadromous brook char, Salvelinus fontinalis in the St. Jean River, Gaspe, Quebec. Canadian Journal of Zoology 60:3084-3091Curry, R. A., D Sparks, and J. Van de Sande. 2002. Spatial and temporal movements of a riverine brook trout population. Transactions of the American Fisheries Society 131:551-560.

Florida Power and Light Energy (FPLE). 2000. Tributary investigations Fall 2000. FPLE biological section report 6 p.

Hokanson, K. E. F., J. H. McCormick, B. R. Jones, and J. H. Tucker. 1973. Thermal requirements for maturation, spawning, and embryo survival of the brook trout, Salvelinus fontinalis. Journal of the Fisheries Research Board of Canada 30:975-84.

MacMillan, J. L., and T. J. Crandlemere. 2005. Woodens River and Tangier Grand Lake trout study 2005. Nova Scotia Department of Agriculture and Fisheries, Inland Fisheries Division, Pictou, N.S.

MacMillan, J. L., D. Caissie, J. E. LeBlanc and T. J. Crandlemere. 2005. Characterization of summer water temperatures for 312 selected sites in Nova Scotia. Can. Tech. Rep. Fish. Aquat. Sci. 2582: 43p.

McKinley, R. S., G. Power, and H. E. Kowalyk. 1994. Transmitter attachment/implant procedures. Environmental Research Department, Ontario Hydro research Division, Report 4-100, Toronto

O’Connor, J. F., and G. Power. 1973. Homing of brook trout ( Salvelinus fontinalis ) in Matamek Lake, Quebec. J. Fish. Res. Board Can. 30:1012-1014

Power, G. 1980. The brook charr, Salvelinus fontinalis. Pages 141-203 in E. K. Balon, editor. Charrs: salmonid fishes of the genus Salvelinus. Dr. W. Junk, The Hague, The Netherlands.

Torgerson, C. E., D. M. Price, H. W. Li and B. A. McIntosh. 1999. Multiscale thermal refugia and stream habitat associations of Chinook salmon in Northeastern Oregon. Ecological Applications 9 (1):301-319

VanDusen, P. J. and C. J. F. Huckins. 2005. Associations among selection logging history, brook trout, macroinvertebrates, and habitat in northern Michigan headwater streams. Transactions of the American Fisheries Society 134:762-774.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Website Builder