**3.1 Lake substrates**

There are two main types of equipment now available for bottom classification but reviewers differ on the quality of bottom categorisation. The RoxAnn classification is based on energy calculations for first and second echosounder returns and the QTC view (version 5) calculates first echo shape parameters. QTC view provides automatic classifications and confidence estimates, while the RoxAnn relies on arbitrary manual calibration. The QTC bottom classes generally have consistent grain size and texture properties and follow grain size trends but RoxAnn classes are difficult to define. Both the RoxAnn and two versions of the QTC –View, series 5 were used to assess substrate.

*Depth and Substrate Hardness***:** Substratum information was collected using an American Pioneer V digital sonar system and a Trimble Pro XR submeter Global Positioning System. Sonar data was collected using a 120 kHz - 12 degree transducer. Information collected included depth, bottom hardness, and roughness. Bottom hardness is an interval measure of the magnitude of the sonar ping return signal. Larger byte range values indicate a harder substrate.

The QTC-View Series 5 classification system is based on the principle that the shape of an echo sounder's first echo discriminates seabeds or substrates. For example, the acoustical signal of a smooth, simple, muddy seabed absorbs a high amount of energy and exhibits a low degree of backscatter resulting in an echo trace with a relatively narrow peak and no tail. Energy reflected from a rough, complicated, gravel seabed exhibits a high degree of backscatter. This results in an echo with a wide peak and a tail. The QTC-View series collected all echos and then post-processes the data in QTC IMPACT. The echo and GPS data is merged and the poor quality echos are filtered out. After echo digitization and preprocessing, the datum is analyzed by algorithms which characterize the waveform by using energy and spectral components to generate a digital string of over 100 shape descriptors. This series of numbers constitutes a description of the echo shape. Statistical analysis determines the most useful elements or series of elements to best discriminate echo shape.

The depth and substrate was collected and calculation of available habitat was done on Idrisi for windows. A frequency distribution was created using the data from each pixel on the hardness and depth images. For most of the mapping ArcMap, digital elevation models and/or kriging were used.

*Sediment Analysis***:** Thirty-seven sediment grabs were collected so that comparisons could be made between sonar data and the sediment type. A substrate type; silt, sand, cobble, etc., could be related to hardness values, ranging from 90-145. Sampling sites were selected so that all possible substrate types were collected. Transects were run across the depth gradient in the lake running from the shallow sandy areas to the deeper silt/clay areas. Current was considered when selecting sample sites. Areas of high and low current were sampled (see section on current profiles)**.**

Buoys were placed at the spot to be sampled. The location of each sample was recorded using a GPS unit. An Eckman dredge was dropped at the buoyed site and the sample retrieved. The sample was placed in a bag and kept cool until it could be tested in the laboratory.

The total sample at each site was divided into three parts. 100 ml was taken for particle size analysis, 100 ml was taken for invertebrate sampling and the remaining portion (if any) was frozen.

There are two main types of equipment now available for bottom classification but reviewers differ on the quality of bottom categorisation. The RoxAnn classification is based on energy calculations for first and second echosounder returns and the QTC view (version 5) calculates first echo shape parameters. QTC view provides automatic classifications and confidence estimates, while the RoxAnn relies on arbitrary manual calibration. The QTC bottom classes generally have consistent grain size and texture properties and follow grain size trends but RoxAnn classes are difficult to define. Both the RoxAnn and two versions of

*Depth and Substrate Hardness***:** Substratum information was collected using an American Pioneer V digital sonar system and a Trimble Pro XR submeter Global Positioning System. Sonar data was collected using a 120 kHz - 12 degree transducer. Information collected included depth, bottom hardness, and roughness. Bottom hardness is an interval measure of the magnitude of the sonar ping return signal. Larger byte range values indicate a harder

The QTC-View Series 5 classification system is based on the principle that the shape of an echo sounder's first echo discriminates seabeds or substrates. For example, the acoustical signal of a smooth, simple, muddy seabed absorbs a high amount of energy and exhibits a low degree of backscatter resulting in an echo trace with a relatively narrow peak and no tail. Energy reflected from a rough, complicated, gravel seabed exhibits a high degree of backscatter. This results in an echo with a wide peak and a tail. The QTC-View series collected all echos and then post-processes the data in QTC IMPACT. The echo and GPS data is merged and the poor quality echos are filtered out. After echo digitization and preprocessing, the datum is analyzed by algorithms which characterize the waveform by using energy and spectral components to generate a digital string of over 100 shape descriptors. This series of numbers constitutes a description of the echo shape. Statistical analysis determines the most useful elements or series of elements to best discriminate echo

The depth and substrate was collected and calculation of available habitat was done on Idrisi for windows. A frequency distribution was created using the data from each pixel on the hardness and depth images. For most of the mapping ArcMap, digital elevation models

*Sediment Analysis***:** Thirty-seven sediment grabs were collected so that comparisons could be made between sonar data and the sediment type. A substrate type; silt, sand, cobble, etc., could be related to hardness values, ranging from 90-145. Sampling sites were selected so that all possible substrate types were collected. Transects were run across the depth gradient in the lake running from the shallow sandy areas to the deeper silt/clay areas. Current was considered when selecting sample sites. Areas of high and low current were sampled (see

Buoys were placed at the spot to be sampled. The location of each sample was recorded using a GPS unit. An Eckman dredge was dropped at the buoyed site and the sample retrieved. The sample was placed in a bag and kept cool until it could be tested in the

The total sample at each site was divided into three parts. 100 ml was taken for particle size analysis, 100 ml was taken for invertebrate sampling and the remaining portion (if any) was

**3. Sturgeon habitat assessment** 

the QTC –View, series 5 were used to assess substrate.

**3.1 Lake substrates** 

substrate.

shape.

and/or kriging were used.

section on current profiles)**.**

laboratory.

frozen.

*Laboratory analyses*: Gravel and larger particle sizes were separated individually by hand, sieve analysis was used to separate the portion of the sample in the very fine sand to gravel categories, and settling velocity was used for silt and clay particle sizes.

The sample was placed in a tray and mixed thoroughly. The sample was added to water and dispersent (Calgon-sodium hexametaphosphate), mixed, and let sit overnight. The sample was then mixed, frozen, thawed and mixed again. The process was repeated a second time to ensure that the particles were completely separated. After the sample was separated thoroughly by the above mentioned process the silt and clay particles were removed from the sand and gravel particles. The sample was wet sieved through a 4 phi sieve to remove the silt and clay from the sample. The sediment left in the 4 phi sieve was placed in a beaker and the sediment falling through the sieve was stored in a separate container. The fine particles and larger particles were then treated separately.

Sand samples were placed into a pyrite beaker and allowed to dry in the oven until the sample was completely dry. A pestle and mortar were used to break up any chunks formed by the drying process. The total sample was weighed and placed in the top of a series of sieves ranging from -1.5 phi to 4 phi (0.5 phi intervals). The sieves were then placed into a Ro-tap shaker and then shaken for ten minutes. The material left in each sieve was weighed and recorded.

Silt/clay samples were placed in a 4000 ml graduated cylinder. Water was added to the cylinder to make the total solution 4000 ml. The solution was then shaken thoroughly and allowed to settle for the appropriate amount of time necessary to separate silt from clay. Five phi (1/32 mm) particles are the largest size category and were removed first. Five phi particles fall at a rate of 4 mm/second. The 4000 ml suspension is 500 mm high therefore the suspension was allowed to settle for 6 minutes and 15 seconds to allow all 1/32 mm particles to fall the full length of the cylinder. The supernatant (water and particles smaller than 1/32 mm) was siphoned off and saved. The process was repeated twice more to remove any smaller particles that would settle from the bottom portion of the cylinder. The remaining precipitant after the third siphon was dried and weighed as the silt category. The siphoned water was also dried and weighed as the clay category.

The particle size classification used was Cummins (1962) modification of the Wentworth Scale. Thirty-seven sediment grabs were taken and analysed from Round Lake. GPS locations for the sediment grabs were compared to hardness values from the Round Lake map at the same GPS location to obtain a system of classifying hardness values obtained by sonar.

*Telemetry data analysis*: Data was presented in maps using Idrisi for Windows (Clark University, MA). Some maps were created in Idrisi for Windows. Acoustic telemetry data was imported from the Vemco system program.

Determination of depth selection and substrate selection was done by hand. Seven days were selected for analysis. Selection was based on movements to include the widest possible range of movement patterns. Days 206 and 221 were selected for sturgeon 4014.

Days 210 and 222 were selected for sturgeon 4015. Days 206, 211 and 219 were selected for sturgeon 4017 (Table 2). For each location on each day bottom depth was determined. Swimming depth minus bottom depth was calculated to determine all locations in which the fish was in contact with the substrate. All values of 1 or less were included. All figures and comments pertaining to substrate selection only include locations in which the fish was in contact with the bottom. Other location statements and figures use all positional data available.

Movements and Habitat Use by Lake Sturgeon (*Acipenser fulvescens*)

Table 3. Sediment classification scheme for Round Lake.

**4. Lake sturgeon movements** 

and in the water column.

values. Hardness values range from 95 (clay) to 150 (rock, see Fig. 11).

of the lake. Movement on day 220 was restricted to the river outlet.

frequency of 16 percent.

in an Unperturbed Environment: A Small Boreal Lake in the Canadian Shield 383

and 135 (medium sand) respectively (Fig. 12). Substrate hardness of 110 (fine sand) had a

**Phi Particle Size (mm) Category Hardness**  9 <0.0039 clay 95 5,6,7,8 0.0039 – 0.0625 silt 100 4 0.00625 - 0.125 Very fine sand 105 3 0.125 – 0.25 Fine sand 110 2 0.25 – 0.5 Medium sand 115 1 0.5 - 1 Coarse sand 120 0 1 – 2 very coarse sand 125 -1,-2,-3 2 – 16 gravel 130 -4,-5 16 - 64 Pebble 135 -6,-7 64 - 256 Cobble 140 -8 >256 Boulder 145

Thirty–seven sediment grabs were taken to compare with the hardness values obtained from the sonic data. Table 3 lists the substrate classification given to each range of hardness

The biological data for the nine lake sturgeon tagged with acoustic tags are listed in Table 2. The nine fish were tracked for 27 days and 15,446 locations were obtained. Movements ranged from individuals that were mostly sedentary to highly mobile individuals. Daily movements were variable between fish as well as by the same fish on different days. Figure 13 shows the locations of fish 4015 on four separate days. Movement was confined to the inlet to Round Lake on day 210. Movement increased on days 211 and 212 and covered most

A comparison of the movements of juvenile and adult lake sturgeon is shown in Fig. 14. Movements of the juvenile fish were focused at the inlet and outlet and in the deep hole (~16 m). Movements of the subadult and adult lake sturgeon were also associated with the inlet and outlet but the movements were more widespread around the lake. The channel where water entered the lake was a preferred site as was the outlet from the lake. Figure 15 shows the swimming depth of sturgeon 4014 on day 206 relative to the bottom depth. Note the day 206 is based on January 1 being day 1. Sturgeon 4014 was on the bottom 30% of all locations on day 206. During the hours from midnight to 5 AM sturgeon 4014 was in the water column the majority of the time. From 5 AM to 11 PM more time was spent on the bottom. After 11 PM lakes sturgeon movements shifted to the water column. Figure 16 shows sturgeon 4015 on day 221 where 53% of all locations were on the bottom on day 221. Sturgeon 4017 on day 211 was on the bottom for the entire day but periodically swam to the surface (Fig. 17**).** Figure 18 shows the overall distribution of each lake sturgeon fitted with a pressure tag and the total distribution of all fish on the bottom

Round Lake.

Fig. 9. Digital elevation model of Fig. 10. Depth availability in Round Lake

Round Lake obtained from sonar data. Round Lake.

Fig. 11. Hardness map of Pigeon River at Fig. 12. Hardness (substrate) availability in

*Substrate and Depth***:** Maximum depth of Round Lake is 16m. A depth map of Round Lake is shown in Fig. 9**.** Two deep holes, one off the Northeast corner of each island, are found in the lake. The general structure is bowl shaped. Depth availability is shown in Fig. 10. Two and three meters depths are available 33 and 16% respectively. Seven, 8, and 9 meter depths are available 5, 6, and 8% respectively.

Substrate hardness of the lake is shown in Fig. 11. Substrate was generally related to depth. The deeper areas of the lake had softer substrates with a high percentage of silt. The shallow sections along the shoreline to about 10m depth had sandy substrates. Cobble and rock substrate predominated in areas of high flow at the inlet and outlet. Availability of substrate hardness was 11, 25, and 17 percent for hardness values of 125 (coarse sand), 130 (gravel),


and 135 (medium sand) respectively (Fig. 12). Substrate hardness of 110 (fine sand) had a frequency of 16 percent.

Table 3. Sediment classification scheme for Round Lake.

Thirty–seven sediment grabs were taken to compare with the hardness values obtained from the sonic data. Table 3 lists the substrate classification given to each range of hardness values. Hardness values range from 95 (clay) to 150 (rock, see Fig. 11).
