Stocked Trout Survival and Camera-based Angler Survey at Selected ACA Stocked Ponds

Stocked Trout Survival and Camera-based Angler Survey at Selected ACA Stocked Ponds

The Alberta Conservation Association is a Delegated Administrative Organization under Alberta s Wildlife Act.

Stocked Trout Survival and Camera-based Angler Survey at Selected ACA Stocked Ponds Kevin Fitzsimmons and Britt Keeling Alberta Conservation Association 101 9 Chippewa Road Sherwood Park, Alberta, Canada T8A 6J7

Report Editors PETER AKU KELLEY KISSNER Alberta Conservation Association 50 Tuscany Meadows Cres. NW 101 9 Chippewa Rd. Calgary, AB T3L 2T9 Sherwood Park, AB T8A 6J7 Conservation Report Series Type Data ISBN: 978-0-9949118-4-1 Disclaimer: This document is an independent report prepared by Alberta Conservation Association. The authors are solely responsible for the interpretations of data and statements made within this report. Reproduction and Availability: This report and its contents may be reproduced in whole, or in part, provided that this title page is included with such reproduction and/or appropriate acknowledgements are provided to the authors and sponsors of this project. Suggested Citation: Fitzsimmons, K., and B. Keeling. 2016. Stocked trout survival and camera-based angler survey at selected ACA stocked ponds. Data Report, D-2016-106, produced by Alberta Conservation Association, Sherwood Park, Alberta, Canada. 25 pp + App. Cover photo credit: David Fairless Digital copies of conservation reports can be obtained from: Alberta Conservation Association 101 9 Chippewa Rd. Sherwood Park, AB T8A 6J7 Toll Free: 1-877-969-9091 Tel: (780) 410-1998 Fax: (780) 464-0990 Email: info@ab-conservation.com Website: www.ab-conservation.com i

EXECUTIVE SUMMARY The Enhanced Fish Stocking (EFS) project creates recreational fisheries in areas of the province where such opportunities do not otherwise exist. Through EFS, we stock approximately 120,000 catchable-sized (i.e., 20 cm) trout into 60 ponds each year, creating put-and-take fisheries that allow anglers to harvest up to five fish per day. Most EFS ponds are situated close to urban centres, making them popular family destinations. However, recent evidence suggests some of these ponds may not be capable of supporting trout survival beyond mid-summer. Our data suggest that poor water quality, particularly high temperature and low dissolved oxygen (DO), and avian predation may be key contributing factors. During the summer of 2014, we assessed 12 representative EFS ponds to estimate survival of stocked trout and determine which water quality variables drive fish survival. As well, we conducted camera-based angler surveys to estimate fishing effort and harvest at these ponds. In the summer of 2015, we examined the extent of avian predation at four ponds (Bashaw, Vegreville and Windsor ponds, and Mirror Reservoir) and assessed the effectiveness of various avian deterrents at Mirror Reservoir. We obtained adequate angler survey data to estimate parameters at four ponds during the summer of 2014. The total number of estimated trips made by anglers to these ponds ranged from 537 at Westlock to 3,212 at Cipperley s. Angler effort ranged from 7 h/ha (95% CI = 3 13) at Mirror to 5,268 h/ha (95% CI = 4,727 5,804) at Cipperley s, and harvest of stocked trout ranged from 3.3% (95% CI = 3.0 3.7) at Mound Red to 72.7% (95% CI = 65.5 80.5) at Cipperley s. Mid-column water temperatures exceeded stress thresholds (19 C) for rainbow trout throughout the summer in most (7 of 12) ponds. Similarly, DO concentrations fell below the 5.0 mg/l limit at 7 of 12 ponds throughout the summer, but there was no clear association between high temperature and low DO. Chlorophyll-a concentrations varied across ponds (1.45 to 110 µg/l), and NH3 concentrations ranged from 0.05 to 0.69 mg/l, well above the Canadian Council of Ministers of the Environment limit (0.019 mg/l) for the protection of aquatic life, as well as above levels (0.04 mg/l) known to have negative impacts on rainbow trout health. ii

Fish survival varied considerably across the 12 ponds, ranging from 7.5% at Beaumont to 99.7% at Nuggent. At some ponds, low survival could be attributed to high angler harvest (e.g., Cipperley s 72.7%), indicating a direct benefit to anglers. Conversely, at other ponds, low survival rates more likely resulted from natural mortality, leading to large portions of unaccounted stocked fish (37.4% at Westlock to 79.1% at Mound Red). Trout survival also varied as a function of water quality. Maximum water temperature, minimum DO, NH3 and chlorophyll-a concentrations were key water quality variables that influenced rainbow trout survival. Avian predation also may have contributed to the high unexplained mortality of stocked fish in some ponds. In particular, we recorded large increases in cormorant presence following trout stocking at Mirror, where cormorant activity increased 14.9 to 39.7 times after stocking than before stocking. Cormorant abundance at Windsor and Bashaw were low compared with Mirror; however, their presence may indicate potential fish predation at these sites as well. No cormorants were documented at Vegreville. Avian deterrents installed at Mirror were not effective in reducing cormorant abundance at the pond. Key words: cormorants. stocked rainbow trout, trout survival, angler survey, water quality, iii

ACKNOWLEDGEMENTS We thank Alberta Conservation Association employees Cale Babey, Melissa Buskas, Andrew Clough, Troy Furukawa, John Hallett, David Jackson, Peter Jones, Jacob Rovere, Scott Seward and Zach Spence for assisting with data collection. We thank Alberta Environment and Parks for providing a camping area (Miquelon Lake Provincial Park). Andrew Paul and Mike Rodtka provided valuable comments on modelling survival. iv

TABLE OF CONTENTS EXECUTIVE SUMMARY... ii ACKNOWLEDGEMENTS... iv TABLE OF CONTENTS... v LIST OF FIGURES... vi LIST OF TABLES... vii LIST OF APPENDICES... viii 1.0 INTRODUCTION... 1 2.0 STUDY AREA... 2 3.0 MATERIALS AND METHODS... 5 3.1 Angler interviews... 5 3.2 Camera-based creel parameter estimates... 5 3.3 Water quality... 6 3.4 Survival estimates... 7 3.5 Predator estimates... 8 4.0 RESULTS... 9 4.1 Creel survey... 9 4.2 Water quality... 11 4.3 Survival estimates... 14 4.4 Avian predation... 19 5.0 SUMMARY... 22 6.0 LITERATURE CITED... 24 7.0 APPENDICES... 26 v

LIST OF FIGURES Figure 1. Figure 2. Figure 3. Enhanced Fish Stocking ponds assessed for trout survival, angler effort and water quality in 2014 and avian predatory activity in 2015... 4 Relative abundance of stocked rainbow trout from May to September at nine Enhanced Fish Stocking ponds... 16 Maximum likelihood estimates of cormorant hours two weeks pre- and post-stocking at Mirror Reservoir in 2014 and 2015... 21 vi

LIST OF TABLES Table 1. Table 2. Table 3. Table 4. Table 5. Table 6. Table 7. Table 8. Table 9. Name, location and characteristics of Enhanced Fish Stocking ponds evaluated for trout survival, angler effort and water quality parameters in 2014... 3 Name, location and physical characteristics of ponds evaluated for avian predation in 2015... 5 Summary of angler survey data from four Enhanced Fish Stocking ponds during the summer of 2014... 10 Summary of angling hours and angler effort collected with trail cameras from 12 Enhanced Fish Stocking ponds during the summer of 2014... 11 Mean, standard deviation around the mean, and maximum water quality parameters from composite water samples taken at Enhanced Fish Stocking ponds throughout the sampling period, 2014... 13 Number of fish stocked, percentage of initial stocking captured and fish surviving until mid-september from 12 Enhanced Fish Stocking ponds, 2014... 15 Summary of Cormack-Jolly-Seber mark-recapture models at nine Enhanced Fish Stocking ponds, 2014... 17 Cormorant counts and estimated activity at three Enhanced Fish Stocking ponds stocked with rainbow trout, 2015... 20 Summary of gill netting at Mirror (2015), including fishing effort, total catch and the associated catch-per-unit-effort... 22 vii

LIST OF APPENDICES Appendix 1. Total phosphorus concentrations at spring turnover (2011 2013) of Enhanced Fish Stocking ponds... 26 Appendix 2. Average depth of Enhanced Fish Stocking ponds... 27 Appendix 3. Waterbody maps indicating camera field of view... 28 Appendix 4. Appendix 5. Appendix 6. Appendix 7. Appendix 8. Program MARK.inp file formats used for survival estimates at Enhanced Fish Stocking ponds, 2014... 40 Candidate models for estimating survival and recapture probability at Enhanced Fish Stocking ponds... 49 Hourly dissolved oxygen and temperature at 12 Enhanced Fish Stocking ponds for the summer of 2014 with 19 C upper temperature and 5 mg/l lower DO reference points.... 50 Summary of fish captures and recaptures at 12 Enhanced Fish Stocking ponds during the summer of 2014.... 62 Estimates, standard errors and 95% confidence intervals for survival and recapture parameters at ponds where sufficient recapture data allowed for parameter estimation... 65 viii

1.0 INTRODUCTION In 1994, Alberta Environmental Protection and Alberta Fish Farmers Association initiated the Enhanced Fish Stocking Program (EFSP) to supplement the existing government fish stocking program, provide private industry with market opportunities and provide larger trout for put-and-take ponds. In 1998, Alberta Conservation Association (ACA) assumed responsibility for delivering the EFSP, hereafter referred to as Enhanced Fish Stocking (EFS), with the objective of providing increased angling opportunities to Albertans by stocking catchable-sized ( 20 cm) rainbow trout in the white zone of Alberta in ponds that frequently winterkill. ACA annually stocks approximately 115,000 rainbow trout and 5,000 brown trout into 61 ponds across Alberta. Most EFS ponds are located close to urban centres, making them popular family destinations. However, recent evidence suggests some of these ponds may not be capable of supporting trout survival beyond mid-summer. Results from our study of seven stocked waterbodies (including four EFS ponds) indicated that only 4.7% of stocked fish survived to the end of the summer; angler harvest accounted for 4%, while over 90% of stocked trout died of natural causes over the fishing season (Patterson and Sullivan 2013). A variety of factors likely contribute to the high mortality of stocked trout in EFS ponds; however, data collected by ACA suggest poor water quality, particularly high temperature and low dissolved oxygen (DO), may be key among these factors. During field surveys on Mirror Reservoir in the summer of 2014, we observed a substantial increase in the number of cormorants coincident with trout stocking dates, suggesting that avian predation may be yet another factor contributing to high mortality of stocked trout. Enhanced Fish Stocking ponds are typically small, shallow and prone to high temperatures, high nutrient concentrations, abundant algal growth and low oxygen levels. Although rainbow trout are more tolerant of a wider range of environmental conditions than other members of the salmonid family (Kerr and Lasenby 2000), physiochemical thresholds exists beyond which growth is limited and mortality occurs (Davis 1975; Cherry et al. 1997). During the summer of 2014, we carried out assessments at 12 representative EFS ponds to estimate survival of stocked trout, determine if water quality variables drive fish survival, and estimate angler effort and angler harvest at 1

these ponds. In addition to our direct observations on Mirror Reservoir, anecdotal evidence suggests the problem of avian predation may be widespread, possibly also occurring at Bashaw, Vegreville and Windsor ponds. In the summer of 2015, we examined the extent of avian predation at these four waterbodies and assessed the effectiveness of various avian deterrent methods on Mirror Reservoir. 2.0 STUDY AREA In 2014, study ponds were selected from a subset of ACA stocked ponds based on two criteria: 1) field staff must drive from a centrally located camp to study sites, conduct field work, and drive back within a 10-hour day; and 2) the study ponds should span a range of predicted survival rates. We predicted that shallow ponds with high total phosphorus (TP) levels would have lower fish survival than deeper ponds with lower TP (Table 1). To satisfy the first criterion, buffers of 100 km radius were placed around logical field-camp locations, and the density of ACA stocked ponds within the buffers was calculated. Among the buffers with the highest density of stocked ponds, the range of depth and TP was evaluated to satisfy the second criterion (Appendices 1 and 2). Both criteria for study ponds were met by selecting nine study ponds from the 100 km radius buffer around Miquelon Lake Provincial Park, with the addition of two ponds Nuggent and Westlock just outside of the buffer and Cipperley s north of Calgary (chosen because it represented the upper range of TP concentrations in ACA stocked ponds; Appendix 1). In 2015, four ponds were selected to monitor avian predation. These ponds were Mirror, Bashaw and Windsor, located within 35 km of one another, northeast of Red Deer (Figure 1, Table 2), and Vegreville pond located in the town of Vegreville, 100 km east of Edmonton (Figure 1, Table 2). 2

Table 1. Name, location and characteristics of Enhanced Fish Stocking ponds evaluated for trout survival, angler effort and water quality parameters in 2014. Waterbody Easting 1 Northing 1 Mean depth (m) Maximum depth (m) Total surface area (ha) Surface area in camera (ha) Total phosphorous (mg/l) Beaumont Pond 604434 5909623 3.0 8.3 2.5 1.1 0.016 Cipperley s Reservoir 561459 5728482 3.0 5.6 0.9 0.3 0.600 Daysland Pond 684104 5858842 2.7 5.7 1.6 0.9 0.250 Innisfree Trout Pond 723945 5919463 4.4 8.2 1.4 0.9 0.067 Irma Fish and Game Pond 752899 5867494 2.8 6.4 0.4 0.3 0.180 Lamont Pond 648255 5958523 4.4 6.7 5.6 3.6 0.210 Mirror Reservoir 626385 5811967 3.3 4.9 4.3 2.4 0.061 Mound Red Reservoir 569865 5855751 2.2 4.4 4.2 2.5 0.100 Nuggent Pond 546030 5846972 3.2 4.3 0.6 0.3 0.025 Radway Fish Pond 633741 5989375 2.0 3.6 0.7 0.1 0.034 Tees Trout Pond 614279 5812537 3.2 6.8 1.7 1.2 0.037 Westlock Recreational Pond 583083 5998080 3.4 5.3 1.4 0.5 0.094 1 Projected in NAD83 10TM. 3

Figure 1. Enhanced Fish Stocking ponds assessed for trout survival, angler effort and water quality in 2014 (black dots) and avian predatory activity in 2015 (red dots). 4

Table 2. Name, location and physical characteristics of ponds evaluated for avian predation in 2015. Waterbody Easting 1 Northing 1 Total Mean Maximum Surface area surface depth depth in camera area (m) (m) (ha) (ha) Bashaw Pond 366755 5825277 3.4 5.6 1.5 0.8 Mirror Reservoir 626385 5811967 3.3 4.9 4.3 2.4 Vegreville Pond 431306 5928760 2.5 4.4 1.4 1.0 Windsor Lake 358228 5826656 3.1 5.7 17 5.8 1 Projected in NAD83 10TM. 3.0 MATERIALS AND METHODS 3.1 Angler interviews We conducted a roving creel survey following a modified version of Pollock et al. (1994). We interviewed anglers upon the completion of their angling trip and collected data on trip length and rainbow trout catch (harvested and released). These data were then used to estimate the number of angler trips and fish harvest and release rates at each waterbody. 3.2 Camera-based creel parameter estimates To increase sampling effort and reduce survey costs, we supplemented angler interviews with trail camera data collected from May 17 to September 22. Digital cameras (Reconyx PC900 Hyperfire) were installed at each pond and programmed to take a photo hourly from 0600 to 2200 hours daily (2,193 hours in creel period). Cameras were installed on the south shore of each pond and facing north to reduce sun glare in photos, and were attached to a tree or post at a serviceable height (approximately 3 m) and locked to protect data. Data were downloaded from cameras and battery life was assessed approximately every two weeks. 5

In each photo, the number of anglers was counted, providing an instantaneous count of anglers on the pond, within the camera field of view. Photos where angler counts could not be determined (e.g., because fog, glare or insufficient light obscured the view) were not included in the analysis. Camera counts were then bootstrapped (10,000 replicates) to produce a distribution of mean counts, which when multiplied by the total number of hours in the survey period gave a distribution of mean effort (h) in the camera field of view during the survey period. We accounted for the portion of the pond the camera did not cover (each camera photographed only a portion of each pond) by spatially correcting camera-based estimates of angler effort (Appendix 3) using the surface area of the pond surveyed by the camera and the total surface area of the pond. Bootstrapped angler trip length and the spatially corrected camera angler effort data were used to estimate the mean number of angler trips (with 95% confidence intervals; 95% CI) for the survey period. Rainbow trout catch (harvested and released) was estimated using a total ratio estimate (Malvestuto 1983) of catch rate (total fish reported kept and released divided by total reported fishing hours) and the spatially corrected angling hours. Hooking mortality of released rainbow trout was estimated using a mean hooking mortality of 23.0% following Bartholomew and Bohnsack (2005). Fish mortality that could not be accounted for by either angler harvest or hooking mortality (i.e., unaccounted for fish) was calculated as total fish less survival, harvest and hooking mortality. Estimates of creel survey parameters were calculated using the R software package (R Core Team 2014). 3.3 Water quality To accommodate modelling survival along water quality parameters, we collected composite water samples during each site visit for analysis (Maxxam Analytics) of TP, chlorophyll-a (chl-a), ammonia (NH3) and turbidity. Samples were taken as late in the afternoon as possible to capture daily maximum ph and un-ionized NH3 concentrations. Temperature and DO were measured in the middle of the water column at one-hour intervals throughout the summer using digital data loggers that remained in place for the duration of the study. During each site visit, we also measured temperature, DO and ph profiles using a handheld YSI professional plus multi-parameter meter. These data will be used to fill gaps in existing datasets of water quality at ACA stocked ponds. 6

3.4 Survival estimates We estimated fish survival at two-week intervals from stocking date in mid-may to mid-september 2014. We captured fish with multi-panel gill nets set just after sunrise and left until approximately 12:00 p.m. Each net was 15.2 x 2.4 m and consisted of panels of different mesh sizes: 38, 63, 51 and 76 mm (stretched measure). To minimize fish mortality, nets were inspected every 10 to 15 min, and smaller mesh panels were replaced with larger ones over the summer period as fish grew larger. Each captured fish was uniquely marked with a passive integrated transponder (PIT) tag and released. If previously tagged, the fish was scanned, the PIT number was recorded, and then the fish was released. For the final gill netting during the week of September 22, 2014, we switched to two half Fall Walleye Index Netting (Morgan 2000) nets, set overnight to maximize capture of remaining stocked fish; each FWIN net was 30 x 1.8 m and consisted of eight 3.8 x 1.8 m panels of mesh sizes 25, 38, 51, 63, 76, 102, 127 and 157 mm. We derived estimates of trout survival (φ) and recapture probability (p) with the program MARK using the Cormack-Jolly-Seber live recaptures open population models with encounter history from mark-recapture data collected at each pond (Appendix 4). Because our study ponds winterkill annually, we considered all fish released at the time of stocking as encountered and released alive at the first encounter period. We estimated φ and p for 26 a priori models (Appendix 5). Models included a fully time-varying model, constrained models where φ and p vary with presumed environmental process or sampling design, and models where variation in φ and p were described as a function of environmental covariates. Environmental covariates included minimum DO and maximum temperature (three-hour moving average from data loggers over the time interval between fish encounters), and the laboratory results for chl-a, NH3 and turbidity from water samples at the time of the encounters. Although we used TP and depth to initially predict survival and select study waterbodies, we chose to model survival with temperature, DO, NH3, chl-a and turbidity because these parameters directly influence fish health and survival. We estimated model goodness of fit for the most saturated (most parametrized) model with adequate fit to the data, and we adjusted all other models with the ĉ adjustment for data over-dispersion (White et al. 2001). Candidate models were assessed and ranked relative to each other by model likelihood (standardized model 7

AICc [Akaike s Information Criterion corrected for small sample size] weights), with the model having a likelihood of 1.0 being the most probable or true model given the data. Models having <0.3 likelihood were not considered to have strong support. Where more than one model was selected, estimates of φ and p were obtained from model averaging (White et al. 2001). Where selected models had parameter estimates with very wide confidence intervals (likely due to estimates being near the upper boundary coupled with the additional parameters estimated in models), the next ranked model with fewer identifiable parameters was used to estimate φ and p. 3.5 Predator estimates Trail camera Avian predator activity at Mirror, Bashaw, Vegreville and Windsor ponds was collected in the summer of 2015 using digital trail cameras programmed to take a picture every 15 min from 0500 to 2200 hours. Similar data were collected at Mirror in 2014. Trail-camera data were downloaded at various times throughout the summer, and data were analyzed using Timelapse2 software (Greenburg 2015). Birds were classified as cormorants, pelicans, gulls, or other waterfowl, and were counted. Avian predator activity estimates were generated using the R software program (R Core Team 2014) through bootstrapping techniques and were spatially corrected for waterbody area not surveyed by cameras as previously described. Avian deterrents On April 15, 2015, we deployed three types of avian deterrents at Mirror to dissuade birds from landing on the water. Avian deterrent devices were purchased from Margo Supplies in High River, Alberta, and consisted of 14 scare balloons (40.5 cm diameter) and 3 oversized bald eagle replicas (91 cm in height) that were placed on the surface of the pond, and 6 raptor patterned kites (112 cm wing span) that were tethered and flown from 8.5 m poles around the perimeter of the lake. 8

Test Netting On May 22, 2015, at Mirror, 10 days after initial stocking, we used short-duration gill net sets to determine fish survival. Two nets were used: one with four panels (38, 51, 63 and 89 mm mesh sizes) and one with three panels (51, 63 and 89 mm mesh). Each net panel was 10 m long and 2 m high. Nets were set in various locations across the reservoir for approximately 20 min. If after 20 min no fish were caught, the nets were left another 35 min, and if again no fish were caught, nets were left overnight for up to 18 h. Subsequent netting was performed June 1 and July 8 using the same methods. All fish caught were measured and weighed. 4.0 RESULTS 4.1 Creel survey For the 2014 survey, we obtained adequate angler data at four ponds to derive estimates of creel parameters (Table 3). The total number of estimated trips made by anglers to these ponds ranged from 537 at Westlock to 3,212 at Cipperley s. Total catch rates ranged from 0.39 fish/h at Cipperley s to 1.22 fish/h at Westlock, angler harvest of fish ranged from 3.3% at Mound Red to 72.7% at Cipperley s, and angler release from 11.6% at Beaumont to 51.0% at Cipperley s (Table 3). Fish mortality from angler harvest (including an estimated 23% hooking mortality) ranged from 13.3% at Mound Red to 84.9% at Cipperley s. Using camera data, we estimated angler hours and effort ranged from 31 to 4,862 h and 7 to 5,268 h/ha, respectively (Table 4). The lowest angler hours and effort were recorded at Mirror and the highest at Cipperley s (Table 4). 9

Table 3. Summary of angler survey data from four Enhanced Fish Stocking ponds during the summer of 2014. Survey data Waterbody Beaumont Cipperley s Mound Red Westlock Number of trips Mean 1,066 3,212 1,116 537 95% CI 824 1,389 2,526 4,035 856 1,487 367 821 Catch rate (fish/h) Mean 0.50 0.39 0.56 1.22 95% CI NA NA NA NA Harvest (no. fish) Mean 795 1,091 100 887 95% CI 721 869 982 1,207 90 110 770 1,008 Release (no. fish) Mean 289 793 1,296 444 95% CI 262 316 714 878 1,165 1,429 385 504 Hooking mortality (no. fish) Mean 66 182 298 102 95% CI 45 91 124 252 202 410 67 145 No. fish released by anglers Mean 34.5 84.9 13.3 49.5 95% CI NA NA NA NA 1 Includes hooking mortality. NA = not available 10

Table 4. Summary of angling hours and angler effort collected with trail cameras from 12 Enhanced Fish Stocking ponds during the summer of 2014. Waterbody Total number of hours Angler effort (h/ha) Mean 95% CI Mean 95% CI Beaumont 2,176 1,974 2,380 888 805 971 Cipperley s 4,862 4,371 5,368 5,268 4,727 5,804 Daysland 439 315 581 273 196 362 Innisfree 1,271 1,122 1,428 878 775 986 Irma 612 517 717 1,390 1,172 1,627 Lamont 472 386 560 85 70 100 Mirror 31 13 56 7 3 13 Mound Red 2,480 2,232 2,732 596 536 657 Nuggent 505 415 600 856 703 1,016 Tees 1,148 937 1,364 1,559 1,273 1,853 Radway 520 431 615 308 255 364 Westlock 1,091 947 1,242 799 694 910 4.2 Water quality Rainbow trout begin to experience stress when temperatures increase above 19 C (Cherry et al. 1997) and when DO drops below 5 mg/l (Davis 1975). Mid-column water temperature at Innisfree, Lamont, Mound Red, Tees and Radway remained below 19 C throughout the summer, with the occasional short increase slightly above the threshold. Beaumont, Cipperley s, Daysland, Irma, Mirror, Nuggent and Westlock all had sustained periods (1 to 27 days) when water temperature rose above 19 C, typically beginning in late June or early July and lasting until late August (Appendix 6). At Innisfree, Irma, Lamont, Mound Red and Radway, DO fell below the 5 mg/l threshold and remained there consistently during part of June, July and August, often falling to 0 mg/l (Appendix 6). Beaumont, Daysland, Mirror and Cipperley s had DO concentrations that fluctuated around the 5 mg/l threshold for parts of the sampling period but never fell to 0 mg/l (Appendix 6). At Nuggent, DO remained above 5 mg/l for most of the sampling season, with a one-week window of consistent readings below 11

5 mg/l, but above 0 mg/l, in late August (Appendix 6). Dissolved oxygen concentrations at Tees were available for May and June only, but they remained above 5 mg/l (Appendix 5). Westlock experienced DO levels below 5 mg/l for the second half of June, falling to 0 mg/l, then fluctuating from the 0 to 5 mg/l for the remainder of the study period (Appendix 6). Chlorophyll-a concentrations varied across ponds (1.45 to 110 µg/l), with most ponds remaining below the threshold described by Barica (1975), who recommend stocking fish into lakes with maximum chl-a concentrations below 100 µg/l. Daysland was the only pond where chl-a concentrations rose above the threshold to 110 µg/l during one sampling event (Table 6). At all ponds, NH3 concentrations were consistently above 0.019 mg/l, an environmental quality guideline developed by the Canadian Council of Ministers of the Environment (CCME) for the protection of aquatic life (CCME 2010) (Table 5). Furthermore, all of our study ponds were above 0.04 mg/l, a concentration found to negatively impact rainbow trout health (Thurston et al. 1984), with concentrations ranging from 0.05 to 0.69 mg/l. Turbidity levels fluctuated at all ponds throughout the sampling period, but most levels remained below CCME guidelines of 8 NTU (CCME 2002) (Table 5). Daysland, Innisfree, Lamont, Mound Red and Westlock exceeded 8 NTU, although Lamont was the only pond where values remained high throughout the entire sampling period (Table 5). 12

Table 5. Mean, standard deviation around the mean, and maximum water quality parameters from composite water samples taken at Enhanced Fish Stocking ponds throughout the sampling period, 2014. Waterbody Chl-a (µg/l) NH3 (mg/l) Turbidity (NTU) Mean SD Max Mean SD Max Mean SD Max n Beaumont 6.61 4.09 12.60 0.05 0 0.06 1.61 0.94 3.60 8 Cipperley's 22.26 10.22 43.30 0.10 0.08 0.25 4.26 1.65 7.50 8 Daysland 46.41 29.81 110.00 0.28 0.16 0.46 7.79 3.86 15.00 7 Innisfree 13.54 4.49 19.80 0.14 0.17 0.56 7.83 2.57 11.00 8 Irma 5.10 5.39 18.00 0.15 0.22 0.69 1.84 1.2 4.60 8 Lamont 69.20 24.98 95.00 0.08 0.03 0.13 24.87 9.57 34.00 6 Mirror 10.87 3.41 14.80 0.05 0 0.06 4.23 1.45 5.90 3 Mound Red 42.91 25.85 84.80 0.10 0.08 0.26 8.59 3.61 13.00 7 Nuggent 4.99 1.36 8.00 0.05 0 0.05 1.56 0.43 2.40 8 Radway 8.00 5.08 19.60 0.10 0.06 0.22 1.59 0.52 2.80 8 Tees 11.84 8.20 21.30 0.05 0 0.05 2.10 1.22 3.50 3 Westlock 26.88 16.98 66.00 0.08 0.06 0.22 10.01 11.05 33.00 8 13

4.3 Survival estimates The percentage of fish that survived from initial stocking into mid-september ranged from 7.5% at Beaumont to 99.7% at Nuggent (Table 6). Mean fish survival across all ponds was 36.9% but varied substantially among ponds. Mound Red exhibited a precipitous decline in survival within two weeks of stocking (Figure 2). In contrast, few mortalities were recorded at Nuggent pond, and fish survival remained high throughout the study period. Irma, Daysland and Cipperley s ponds also exhibited relatively high survival throughout much of the summer, whereas the remaining ponds showed a comparatively steady decline and low fish survival through the summer (Figure 2, Table 6, Appendices 7 and 8). The percentage of stocked fish that we captured throughout the summer ranged from 0.7% at Mirror to 74.5% at Irma, with an overall mean of 23.2% (Table 6). We had few or no catches at Mirror, Lamont and Tees by mid-july, precluding estimates of survival at these three ponds. We found that the best supported models of rainbow trout survival from May to September (models having 0.3 likelihood; Table 7) showed evidence for survival being constant through time (seven models) and being dependent on time in the first encounter interval (two models). We also found evidence that fish survival was a function of maximum water temperature (six models), minimum water DO (five models), NH3 or chl-a concentrations (four models each) (Table 7). Given the parameters we modelled, our results show that survival was influenced most strongly by temperature at Cipperley s, Irma and Mound Red and by DO at Innisfree and Mound Red. At Beaumont, Cipperley s and Westlock, we found that the concentration of NH3 was a predictor of fish survival. At Daysland and Irma, survival varied as a function of chl-a, and turbidity was associated with survival at Radway pond. 14

Table 6. Number of fish stocked, percentage of initial stocking captured and fish surviving until mid-september from 12 Enhanced Fish Stocking ponds, 2014. Waterbody Number of trout stocked No. fish captured Percent fish captured Fish surviving until mid-september No. Percent 95% CI Beaumont 2,500 392 15.7 188 7.5 3.5 10.1 Cipperley s 1,500 298 19.9 783 52.2 32.5 56.1 Daysland 400 286 71.5 256 64.1 10.3 84.5 Innisfree 1,600 268 16.8 310 19.4 17.1 21.1 Irma 400 298 74.5 177 44.2 29.1 46.2 Lamont 2,000 117 5.9 NA 1 NA 1 NA 1 Mirror 3,000 21 0.7 NA 1 NA 1 NA 1 Mound Red 3,000 114 3.8 228 7.6 2.4 8.4 Nuggent 1,000 377 37.7 997 99.7 99.7 100 Tees 1,000 18 1.8 NA 1 NA 1 NA 1 Radway 1,000 209 20.9 242 24.2 5.1 33.4 Westlock 2,000 175 8.8 262 13.1 7.0 17.6 1 Estimates not available because fish captures were too low. 15

Figure 2. Relative abundance (%) of stocked rainbow trout from May to September at nine Enhanced Fish Stocking ponds. Estimates derived from modelled survival parameters. 16

Table 7. Summary of Cormack-Jolly-Seber mark-recapture models at nine Enhanced Fish Stocking ponds, 2014. Models describe survival (φ) between encounters and recapture probability (p) at encounters with or without time dependence (. and t respectively), as a function of ammonia, chlorophyll-a concentrations and turbidity the day of the encounter occasion (NH3, chl-a and turb, respectively), or a function of minimum dissolved oxygen and maximum temperature (three-hour moving average) over the time interval between fish encounters (DO and temp, respectively). Candidate models with likelihood <0.10 are not shown. Model AICc 1 AICc AICc Wt. Likelihood # Par 2 Beaumont Φ(NH3) p(t1,t8) 692.69 0.00 0.32 1.00 5 Φ(.) p(t) 694.06 1.38 0.16 0.50 9 Φ(DO) p(t) 695.87 3.18 0.07 0.20 10 Φ(chl-a ) p(t) 695.91 3.23 0.06 0.20 10 Φ(t1) p(t) 695.99 3.30 0.06 0.19 10 Φ(temp) p(t) 696.01 3.32 0.06 0.19 10 Φ(NH3) p(t) 696.01 3.33 0.06 0.19 10 Φ(t1,t8) p(t1,t8) 696.45 3.77 0.05 0.15 6 Cipperley s Φ(temp) p(t1,t8) 1,053.59 0.00 0.38 1.00 5 Φ(NH3) p(t1,t8) 1,053.59 0.00 0.38 1.00 5 Φ(.) p(t) 1,057.24 3.65 0.06 0.16 9 Daysland Φ(chl-a) p(t) 1,200.80 0.00 0.19 1.00 9 φ(.) p(t) 1,200.96 0.16 0.18 0.92 9 Φ(DO) p(t1,t8) 1,202.25 1.45 0.09 0.48 5 Φ(DO) p(t) 1,202.56 1.76 0.08 0.41 10 Φ(temp) p(t) 1,202.98 2.18 0.07 0.34 10 φ(t1) p(t) 1,203.01 2.21 0.06 0.33 10 φ(t1) p(t) 1,203.01 2.21 0.06 0.33 10 Φ(chl-a) p(t1,t8) 1,203.15 2.35 0.06 0.31 4 φ(.) p(t1,t8) 1,203.91 3.11 0.04 0.21 4 Φ(temp) p(t1,t8) 1,204.71 3.91 0.03 0.14 5 Φ(NH3) p(t1,t8) 1,205.03 4.23 0.02 0.12 5 Φ(NH3) p(nh3) 1,205.04 4.25 0.02 0.12 4 φ(t1,t8) p(t) 1,205.07 4.28 0.02 0.12 11 Φ(DO) p(do) 1,205.25 4.45 0.02 0.11 4 17

Table 7. Continued. Model AICc 1 AICc AICc Wt. Likelihood # Par 2 Innisfree Φ(DO) p(t1,t8) 707.75 0.00 0.28 1.00 5 Φ(chl-a) p(t1,t8) 707.96 0.22 0.25 0.90 5 Φ(.) p(t1,t8) 709.26 1.52 0.13 0.47 4 Φ(chl-a) p(t) 710.62 2.87 0.07 0.24 8 Φ(NH3) p(t1,t8) 710.62 2.87 0.07 0.24 5 Φ(.) p(t) 711.39 3.65 0.05 0.16 9 Φ(t1,t8) p(t1,t8) 711.74 4.00 0.04 0.14 6 Irma Φ(chl-a) p(t1,t8) 1,331.84 0.00 0.38 1.00 5 Φ(temp) p(t1,t8) 1,331.84 0.00 0.38 1.00 5 Φ(.) p(t) 1,335.09 3.25 0.07 0.20 9 Mound Red Φ(temp) p(t1,t7) 412.69 0.00 0.16 1.00 5 Φ(DO) p(t1,t7) 412.78 0.09 0.15 0.96 5 Φ(temp) p(temp) 412.83 0.14 0.15 0.93 4 Φ(DO) p(do) 413.08 0.39 0.13 0.82 4 Φ(chl-a) p(t1,t7) 413.27 0.58 0.12 0.75 5 Φ(NH3) p(t1,t7) 414.29 1.60 0.07 0.45 5 Φ(t1,t7) p(.) 414.53 1.84 0.06 0.40 4 Φ(chl-a) p(chl-a) 416.09 3.40 0.03 0.18 4 Φ(t1,t7) p(turb) 416.14 3.45 0.03 0.18 5 Φ(t1,t7) p(t1,t7) 416.98 4.29 0.02 0.12 6 Nuggent Φ(.) p(t) 1,257.50 0.00 0.28 1.00 9 Φ(DO) p(t) 1,259.53 2.03 0.10 0.36 10 Φ(NH3) p(t) 1,259.53 2.03 0.10 0.36 10 Φ(chl-a) p(t) 1,259.53 2.03 0.10 0.36 10 Φ(temp) p(t) 1,259.53 2.03 0.10 0.36 10 Φ(DO) p(t) 1,259.53 2.03 0.10 0.36 10 Φ(t1) p(t) 1,259.53 2.03 0.10 0.36 10 Φ(DO) p(t1,t8) 1,261.27 3.77 0.04 0.15 5 Φ(t1,t8) p(t) 1,261.56 4.06 0.04 0.13 11 18

Table 7. Continued. Model AICc 1 AICc AICc Wt. Likelihood # Par 2 Radway Φ(.) p(turb) 389.70 0.00 0.27 1.00 3 Φ(temp) p(temp) 390.73 1.03 0.16 0.60 4 Φ(DO) p(do) 391.35 1.65 0.12 0.44 4 Φ(t1,t8) p(turb) 391.92 2.23 0.09 0.33 5 Φ(temp p(t1,t8) 392.38 2.68 0.07 0.26 5 Φ(.) p(t) 392.56 2.87 0.06 0.24 9 Φ(NH3) p(t1,t8) 392.63 2.93 0.06 0.23 5 Westlock Φ(NH3) p(nh3) 328.09 0.00 0.36 1.00 4 Φ(.) p(.) 329.69 1.60 0.16 0.45 2 Φ(temp) p(temp) 331.65 3.56 0.06 0.17 4 Φ(.) p(turb) 331.69 3.61 0.06 0.16 3 Φ(DO) p(do) 331.95 3.86 0.05 0.14 4 Φ(t1,t8) p(.) 332.03 3.94 0.05 0.14 4 Φ(.) p(t1,t8) 332.28 4.19 0.04 0.12 4 Φ(DO) p(t1,t8) 332.63 4.55 0.04 0.10 5 1 AICc= Akaike s Information Criterion corrected for small sample size. 2 Number of parameters specified by the model design. 4.4 Avian predation Trail-camera photos indicated that cormorants were the most abundant avian predator that visited study ponds. Estimated cormorant activity was highest at Mirror, followed by Windsor and Bashaw (Table 8). No cormorants were documented at Vegreville. Relative to the surface area of the waterbody, there was an increase in cormorant activity of 35.6, 36.1, 3.3 and 1.3 h/ha at Mirror Reservoir (first stocking), Mirror Reservoir (second stocking), Windsor Lake and Bashaw Pond, respectively, after stocking. For Mirror Reservoir, this translates to an increase for first and second stockings of 14.9 and 39.7 times, respectively. At Mirror Reservoir, large groups of cormorants were observed shortly after each stocking date and typically persisted for up to 10 days (Table 8). Estimated cormorant activity for 2015/16 did not vary greatly from 2014/15 estimates (Figure 3), when no deterrents were used, suggesting we were not effective at preventing birds from foraging at Mirror. Despite relatively intensive netting of Mirror Reservoir, 19

very few fish were captured (Table 9). We attribute this extremely low abundance of rainbow trout two weeks post-stocking to cormorant predation. Table 8. Cormorant counts and estimated activity (hours) at three Enhanced Fish Stocking ponds stocked with rainbow trout, 2015. Waterbody Mirror Reservoir 1st Stocking Mirror Reservoir 2nd Stocking Cormorant counts (2-week period) Prior to stocking Post stocking Prior to stocking Cormorant hours 2-weeks Post stocking Mean 95% CI Mean 95% CI 18 253 11 2 24 164 85 257 6 217 4 0 10 159 115 209 Windsor Lake 0 77 0 0 56 25 94 Bashaw Pond 0 3 0 0 2 0 4 20

Figure 3. Maximum likelihood estimates of cormorant hours two weeks pre- and post-stocking at Mirror Reservoir in 2014 and 2015. Hours shown for 2015 are the combined hours from the two stocking events that occurred in 2015. 21

Table 9. Summary of gill netting at Mirror (2015), including fishing effort, total catch and the associated catch-per-unit-effort (CPUE). Date No. days after stock Fishing time (min) Gill net area (m 2 ) Total fish caught CPUE (fish/100 m 2 24 h -1 ) 22 May 2015 10 276 640 1 0.82 01 June 2015 19 2,518 500 2 0.23 08 July 2015 13* 2,838 420 2 0.24 *After second stocking event. 5.0 SUMMARY Fish survival ranged considerably across the 12 EFS ponds, with estimates ranging from 7.5% at Beaumont to 99.7% at Nuggent. Angler effort varied from 7 h/ha at Mirror Reservoir to 5,265 h/ha at Cipperley s, resulting in a range of trout harvest (where estimable) from 3.3% at Mound Red Reservoir to 72.7% at Cipperley s. At some ponds, low survival could be attributed to high angler harvest (e.g., Cipperley s 72.7%), indicating a direct benefit to anglers. Conversely, at other ponds, the low survival rates more likely resulted from natural mortality, leading to large percentages of unaccounted stocked fish (37.4% at Westlock to 79.1% at Mound Red). While these estimates of natural mortality are slightly lower than those previously reported for other stocked ponds (91.2%; Patterson and Sullivan 2013), they still remain concerning. In some cases, survival was a function of water-quality parameters. Modelling results suggest that maximum water temperature and minimum DO, NH3 and chl-a concentrations were key water-quality variables that influenced rainbow trout survival. At the outset of our study, we chose ponds that spanned a predicted range of fish survival based on TP concentrations and average depth, with ponds with lower TP concentrations and deeper depths predicted to experience greater survival. However, survival did not follow these predicted trends, and no defined relationship between TP and depth was found at our ponds. This result could be due to the nature of the TP data we used to predict survival, which were instantaneous measurements and not necessarily representative of maximum or mean TP levels. At Cipperley`s, estimates of fish survival were moderate even though we predicted mortality to be highest based on previously recorded TP levels. Water quality at this pond is improved through 22

subsurface aeration, suggesting that poor water quality may be mediated, leading to greater fish survival and availability to anglers. Some portion of unexplained mortality of stocked fish could be attributed to avian predation. We recorded large increases in cormorant presence following trout stocking at Mirror and Windsor. Gill netting at Mirror following stocking events resulted in very few fish captures, suggesting high trout mortality within a short time after stocking. Cormorants have the capacity to consume large numbers of stock-sized trout (Derby and Lovvorn 1997), which could explain the removal of substantial numbers of stocked fish given the large number of birds recorded at these ponds. The variety of bird deterrents we deployed were ineffective at reducing cormorant abundance at Mirror, suggesting that alternative methods may be required if stocking is going to be used successfully. At some ponds, such as Tees and Mound Red, escapement of fish via outflow may also have resulted in significant losses of stocked fish and lower estimated survival rates. Anecdotal evidence from Mound Red suggests that substantial angling effort was focused on pools below the reservoirs outfall, suggesting losses are substantial. Choosing ponds with no outflow or mitigating outflows to prevent loss of fish would help retain fish in the ponds for anglers. Expanding our knowledge of stocked fish survival at individual ponds is helpful to assess our current stocking practices. As we have demonstrated here, EFS ponds continue to be popular destinations for anglers. We identified that water quality plays a role in stocked trout survival, in addition to angler harvest. In ponds where causes for mortality were identified, measures should be taken to try to increase fish survival and availability to anglers. This includes, but is not limited to, focusing efforts on deeper ponds where temperature and DO will be more amenable to trout, improving water quality through efforts such as aeration and riparian enhancement, and stocking alternate species or strains of fish that are more tolerant of higher temperatures and lower DO levels. Where sources of mortality remain unknown, further investigation is required. Overall, this study provides guidance for selecting ponds for future stocking and will be beneficial in the future development of the EFS program. 23

6.0 LITERATURE CITED Barica, J. 1975. Summerkill risk in prairie ponds and possibilities of its prediction. Journal of the Fisheries Board of Canada 32: 1283 1288. Bartholomew, A., and J.A. Bohnsack. 2005. A review of catch-and-release angling mortality with implications for no-take reserves. Reviews in Fish Biology and Fisheries 15(1 2): 129 154. Canadian Council of Ministers of the Environment (CCME). 2002. Canadian water quality guidelines for the protection of aquatic life: Ammonia. In: Canadian environmental quality guidelines, 1999. Canadian Council of Ministers of the Environment, Winnipeg, Manitoba, Canada. Canadian Council of Ministers of the Environment (CCME). 2010. Canadian water quality guidelines for the protection of aquatic life: Total particulate matter. In: Canadian environmental quality guidelines, 1999. Canadian Council of Ministers of the Environment, Winnipeg, Manitoba, Canada. Cherry, D.S., K.L. Dickson, J. Cairns Jr., and J.R. Stauffer. 1997. Preferred, avoided, and lethal temperatures of fish during rising temperature conditions. Journal of Fisheries Research Board of Canada 34: 239 246. Davis, J.C. 1975. Minimal dissolved oxygen requirements of aquatic life with emphasis on Canadian species: A review. Journal of Fisheries Research Board of Canada 32: 2295 2332. Derby, C.E., and J.R. Lavvorn. 1997. Predation on fish by cormorants and pelicans in a cold-water river: A field and modeling study. Canadian Journal of Fisheries and Aquatic Sciences 54(7): 1480 1493. Greenberg, S. 2015. Timelapse2: An image analyser for camera traps. Includes software and manuals. Available online at http://saul.cpsc.ucalgary.ca/timelapse/ [Accessed September 26, 2015]. 24

Kerr, S., and T.A. Lasenby. 2000. Rainbow trout stocking in inland lakes and streams: An annotated bibliography and literature review. Fish and Wildlife Branch, Ontario Ministry of Natural Resources, Peterborough, Ontario. 220 pp + App. Malvestuto, S.P. 1983. Sampling the recreational fishery. Pages 397 430. In: L.A. Neilson and D.L. Johnston, editors. Fisheries techniques (First edition). American Fisheries Society, Bethesda, Maryland, USA. 371 pp. Morgan, G.E. 2000. Manual of instructions: fall walleye index netting (FWIN). Ontario Ministry of Natural Resources, Fish and Wildlife Branch, Peterborough, Ontario. 34 pp. Patterson, W.F., and M.G. Sullivan. 2013. Testing and refining the assumptions of put-and-take rainbow trout fisheries in Alberta. Human Dimensions of Wildlife 18(5): 340 354. Pollock, K.H., C.M. Jones, and T.L. Brown. 1994. Angler survey methods and their applications in fisheries management. American Fisheries Society, Bethesda, Maryland, USA. 371 pp. R Core Team. 2014. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Available online at http://www.r-project.org/. Thurston, R.V., R.C. Russo, R.J. Luedtke, C.E. Smith, E.L. Meyn, C. Chakoumakos, K.C. Wang, and C.J.D. Brown. 1984. Chronic toxicity of ammonia to rainbow trout. Transactions of the American Fisheries Society 113: 56 73. White, G.C., K.P. Burnham, and D.R. Anderson. 2001. Advanced features of program MARK. Pages 368 377. In: Wildlife, land, and people: Priorities for the 21st century. Proceedings of the Second International Wildlife Management Congress. The Wildlife Society, Bethesda, Maryland, USA. 25

7.0 APPENDICES Appendix 1. Total phosphorus concentrations at spring turnover (2011 2013) of Enhanced Fish Stocking ponds. Ponds selected for this study are shown in grey. McVinnie Res. Bow City East Pond Enchant Park Pond Beaumont Pond Waskasoo Pk Pond Fort Lions Pond Granum Pond Gibbons Pond Goldspring Pk Pond Viking Pond Nuggent Pond Niemela Res. Bashaw Pond Stirling Pond Boyle Pond Hermitage Pk Pond Foremost Res. Legal Res. Radway Pond Tees Reservoir Oyen Res. Echo Dale Park Wetaskiwin Pond Two Hills Pond Two Hills Pond Morinville Lake Mirror Res. Morinville Pond Windsor Lake Anderson Dam Innisfree Pond McQuillan Res. Lougheed Pond Pleasure Island Emerson Hansens Res. Telegraph Pk Pond Westlock Pond Mound Red Res. Gooseberry Pk Pond Kraft Pond Vegreville Pond Strathmore Pond Wallace Pk Pond Irma Pond Bud Miller Pond Bonnyville Pond Lamont Res. Centennial Pk Pond Oyen Town Daysland Pond Vermillion Pk Pond Midway Res. Proalta Pond Dewitts Pond Cipperleys Res. Castor Pond 0.0 0.2 0.4 0.6 0.8 1.0 Total phosphorous (mg/l) 26

Appendix 2. Average depth of Enhanced Fish Stocking ponds. Ponds selected for this study are shown in grey. Granum Pond Magrath Pond Bonnyville Pond Midway Res. Niemela Res. Dewitts Pond Tees Reservoir Echo Dale Park Gooseberry Pk Pond Strathmore Pond Hermitage Pk Pond Castor Pond Wallace Pk Pond Mound Red Res. Boyle Pond Centennial Pk Pond Oyen Res. Enchant Park Pond Vegreville Pond Proalta Pond McVinnie Res. McQuillan Res. Waskasoo Pk Pond Oyen Town Daysland Pond Two Hills Pond Kraft Pond Two Hills Pond Irma Pond Beaumont Pond Cipperleys Res. Anderson Dam Legal Res. Windsor Lake Pleasure Island Goldspring Pk Pond Viking Pond Nuggent Pond Radway Pond Hansens Res. Mirror Res. Bashaw Pond Westlock Pond Lougheed Pond Bud Miller Pond Emerson Bow City East Pond Lamont Res. Innisfree Pond Wetaskiwin Pond Foremost Res. Fort Lions Pond Telegraph Pk Pond 0 1 2 3 4 5 6 Average depth (m) 27

Appendix 3. Waterbody maps indicating camera field of view. Beaumont Pond 28

Cipperley`s Reservoir 29

Daysland Pond 30

Innisfree Trout Pond 31

Irma Fish and Game Pond 32

Lamont Pond 33

Mirror Reservoir 34

Mound Red Reservoir 35

Nuggent Pond 36

Tees Trout Pond 37

Radway Fish Pond 38

Westlock Recreational Pond 39

Appendix 4. Program MARK.inp file formats used for survival estimates at Enhanced Fish Stocking ponds, 2014. /* Beaumont Pond Data, Live Recaptures, 8 capture occasions plus stocking day = 9 events, 1 group, time interval = 1.5,1,1,1,1,1,1,3 with 1= 2 week interval, 2500 fish stocked */ 100000000 2153 ; 100000010 7 ; 100000100 8 ; 100001000 26 ; 100010000 16 ; 100100000 79 ; 100100001 1 ; 100100010 1 ; 100101000 2 ; 100110000 1 ; 101000000 153 ; 101000001 1 ; 101000010 2 ; 101000110 1 ; 101001000 2 ; 101010000 2 ; 101100000 3 ; 101100010 1 ; 110000000 36 ; 110001000 3 ; 110100000 2 ; 100000010-1 ; 100000100-4 ; 100001000-6 ; 100010000-2 ; 100100000-2 ; 101000000-37 ; 101000010-1 ; 101001000-2 ; 101010000-1 ; 101100000-1 ; 110000000-4 ; 110001000-1 ; 40

/* Cipperleys Pond Data, Live Recaptures, 8 capture occasions plus stocking day = 9 events, 1 group, time interval = 1,1,1,1,1,1,1,3 with 1= 2 week interval */ 100000000 1240 ; 100000001 16 ; 100000010 18 ; 100000100 19 ; 100000110 1 ; 100001000 12 ; 100010000 53 ; 100010001 2 ; 100100000 53 ; 100100010 1 ; 100100100 2 ; 100101000 1 ; 100110000 8 ; 101000000 52 ; 101000100 2 ; 101001000 1 ; 101010000 4 ; 101010010 1 ; 101100000 5 ; 110000000 7 ; 110010000 2 ; 100000010-2 ; 100000100-2 ; 100001000-1 ; 100010000-7 ; 100100000-1 ; 100110000-1 ; 41

/* Daysland Pond Data, Live Recaptures, 8 capture occasions plus stocking day = 9 events, 1 group, time interval = 1.5,1,1,1,1,1,1,3 with 1= 2 week interval */ 100000000 199 ; 100110100 1 ; 100000010 9 ; 101000010 2 ; 100000100 25 ; 101000101 1 ; 100001000 18 ; 101010000 2 ; 100010001 1 ; 101100000 2 ; 100010100 3 ; 110000100 1 ; 100011000 3 ; 110010000 2 ; 100100001 4 ; 110100000 1 ; 100100100 6 ; 111000001 1 ; 100101001 1 ; 100000010-1 ; 100110010 1 ; 100001000-5 ; 101000000 39 ; 100100000-2 ; 101000100 2 ; 100110100-1 ; 101001000 3 ; 101000100-1 ; 101010001 1 ; 110000000-2 ; 110000000 13 ; 110100010-1 ; 110001000 1 ; 110010000 2 ; 110011000 1 ; 110100000 1 ; 110100010 1 ; 111000001 1 ; 111000100 1 ; 100000010-1 ; 100000100-6 ; 100001000-5 ; 100010000-1 ; 100100000-2 ; 100110000-1 ; 100110100-1 ; 101000000-8 ; 101000100-1 ; 101010000-1 ; 110000000-2 ; 110010000-2 ; 110100010-1 ; 100000001 12 ; 100000011 1 ; 100000111 1 ; 100010000 15 ; 100010010 1 ; 100010110 1 ; 100100000 15 ; 100100010 1 ; 100101000 2 ; 100110000 6 ; 42

/* Innisfree Pond Data, live recaptures, 8 capture occasions plus stocking day = 9 events, 1 group, time interval = 1,1,1,1,1,1,1,3 with 1= 2 week interval */ 100000000 1373 ; 100000001 7 ; 100000010 9 ; 100000100 13 ; 100001000 25 ; 100010000 50 ; 100010010 2 ; 100010100 1 ; 100011000 1 ; 100100000 35 ; 100100010 1 ; 100100100 1 ; 100101100 1 ; 100110000 5 ; 100111000 1 ; 101000000 57 ; 101000010 3 ; 101001000 1 ; 101010000 1 ; 110000000 12 ; 111000000 1 ; 100000010-1 ; 100001000-3 ; 100010000-12 ; 100010010-1 ; 100100000-5 ; 100100010-1 ; 101000000-5 ; 101000010-1 ; 110000000-1 ; 43

/* Irma Pond data, live recaptures, 8 capture occasions plus stocking day = 9 events, 1 group, time interval = 1,1,1,1,1,1,1,3 with 1= 2 week interval */ 100000000 192 ; 100101010-1 ; 100000001 10 ; 101100100-1 ; 100000010 9 ; 110000100-1 ; 100000011 1 ; 110001000-1 ; 100000100 23 ; 100001000 33 ; 100001001 1 ; 100001100 1 ; 100010000 31 ; 100010100 2 ; 100010101 1 ; 100011000 4 ; 100100000 25 ; 100100001 1 ; 100100010 3 ; 100100100 4 ; 100101000 3 ; 100101010 2 ; 100110000 3 ; 101000000 19 ; 101000010 3 ; 101000100 2 ; 101000110 1 ; 101001000 2 ; 101010000 3 ; 101010001 1 ; 101100000 3 ; 101100001 1 ; 101100100 1 ; 101101000 1 ; 110000000 5 ; 110000100 3 ; 110001000 2 ; 110010000 2 ; 110010010 1 ; 110100000 1 ; 100000100-3 ; 100001000-8 ; 100010000-3 ; 100100000-2 ; 100101000-1 ; 44

/* Mound Red Reservoir data, live recaptures, 7 capture occasions plus stocking day = 8 events, group, time interval = 1,1,1,1,1,2,3 with 1= 2 week interval */ 10000000 2899 ; 10000001 14 ; 10000010 8 ; 10000100 4 ; 10000101 1 ; 10000110 1 ; 10000111 1 ; 10001000 11 ; 10001001 1 ; 10001100 1 ; 10010000 10 ; 10010001 1 ; 10011101 1 ; 10100000 23 ; 10100001 1 ; 10110000 1 ; 11000000 22 ; 10000010-5 ; 10000100-1 ; 10000110-1 ; 10001000-2 ; 11000000-2 ; 45

/* Nuggent Pond data, live recaptures, 8 capture occasions plus stocking day = 9 events, 1 group, time interval = 1,1,1,1,1,1,1,3 with 1= 2 week interval */ 100000000 677 ; 100000001 4 ; 100000010 11 ; 100000100 23 ; 100001000 44 ; 100010000 35 ; 100010001 1 ; 100010100 1 ; 100011000 2 ; 100100000 64 ; 100100100 3 ; 100101000 6 ; 100110000 1 ; 100110100 1 ; 101000000 62 ; 101000010 1 ; 101000100 5 ; 101001000 3 ; 101010000 3 ; 101100000 2 ; 101110000 1 ; 110000000 37 ; 110000010 1 ; 110000100 1 ; 110001001 1 ; 110010000 1 ; 110100000 1 ; 110101000 1 ; 110110000 1 ; 110111000 1 ; 111000000 5 ; 100000100-3 ; 100001000-6 ; 100010000-5 ; 100100000-7 ; 100100100-1 ; 101000000-1 ; 101001000-1 ; 46

/* Radway Pond data, live recaptures, 8 capture occasions plus stocking day = 9 events, 1 group, time interval = 0.5,1,1,1,1,1,1,3 with 1= 2 week interval */ 100000000 810 ; 100000001 2 ; 100000010 10 ; 100000100 17 ; 100001000 27 ; 100010000 51 ; 100011000 1 ; 100100000 15 ; 100110000 2 ; 101000000 43 ; 101000100 1 ; 101001000 2 ; 101010000 3 ; 101100000 1 ; 110000000 15 ; 100000100-2 ; 100001000-5 ; 100010000-20 ; 100100000-1 ; 100110000-1 ; 47

/*Westlock Pond data, live recaptures, 8 capture occasions plus stocking day = 9 events, 1 group, time interval = 1,1,1,1,1,1,1,3 with 1= 2 week interval */ 100000000 1836 ; 100000001 1 ; 100000010 1 ; 100000100 16 ; 100001000 36 ; 100001100 1 ; 100010000 6 ; 100011000 1 ; 100100000 25 ; 100100100 1 ; 101000000 33 ; 101100000 1 ; 110000000 39 ; 111000000 3 ; 100000010-1 ; 100000100-4 ; 100001000-5 ; 100010000-1 ; 100100000-2 ; 48

Appendix 5. Candidate models for estimating survival (φ) and recapture (p) probability at Enhanced Fish Stocking ponds. Model φ (t) p(t): both parameters fully time varying φ (.) p(.): both parameters no time effect φ (t) p(.): φ fully time varying, p no time effect φ (.) p(t): φ no time effect p fully time varying φ (t1,8) p(t1,t8): time effect where φ1,8 and p1,8 all others φ (t1,t8) p(t): φ1,8 all others and p fully time varying φ (t1,t8) p(.): time effect at φ1,8 all others and p no time effect φ (.) p(t1,t8): φ no time effect and p1,8 all others φ p(temp+do+chl-a+nh3+turb): φ, p function of max temperature, min DO over time periods and chl-a +NH3+turbidity at time φ (temp) p(temp): φ, p function of max temperature over time periods φ (DO) p(do), φ: p function of min DO over time periods φ (chl-a) p(chl-a): φ p function of chl-a at time φ (NH3) p(nh3): φ, p function of ammonia at time φ (temp) p(t1,t8): φ, function of max temperature over time periods, p1,8 all others φ (DO) p(t1-,8): φ, function of min DO over time periods, p1,8 all others φ (chl-a) p(t1,t8): φ function of chl-a at time, p1,8 all others φ (NH3) p(t1,t8): φ, function of ammonia at time, p1,8 all others φ (t1,t8) p(turb): φ time effect at φ1,8 all others and p function of turbidity at time φ (t7=t8) p (t): φt7,t8 = all others and p fully time varying φ (t1) p(t), φ t1 all others and p fully time varying φ (.) p(turb): φ no time effect and p function of turbidity at time φ (t) p(t1,t8): φ time effect and p1,8 all others φ (temp) p(t), φ, function of max temperature over time periods, p fully time varying φ (DO) p(t): φ, function of min DO over time periods, p fully time varying φ (NH3) p(t): φ, function of ammonia at time, p fully time varying φ (chl-a) p(t): φ, function of chl-a at time, p fully time varying 49

Appendix 6. Hourly dissolved oxygen (DO) and temperature (temp) at 12 Enhanced Fish Stocking ponds for the summer of 2014 with 19 C upper temperature and 5 mg/l lower DO reference points. Beaumont Pond 50

Cipperley s Reservoir 51

Daysland Pond 52

Innisfree Trout Pond 53

Irma Fish and Game Pond 54

Lamont Pond 55

Mirror Reservoir 56

Mound Red Reservoir 57

Nuggent Pond 58

Radway Fish Pond 59

Tees Trout Pond 60

Westlock Recreational Pond 61

Appendix 7. Summary of fish captures and recaptures at 12 Enhanced Fish Stocking ponds during the summer of 2014. NA = not available. Waterbody Date Fish captures Recaptures of tagged fish Beaumont Pond 01/05/2014 2,500 * NA Beaumont Pond 23/05/2014 41 NA Beaumont Pond 04/06/2014 167 0 Beaumont Pond 17/06/2014 92 6 Beaumont Pond 02/07/2014 20 3 Beaumont Pond 16/07/2014 33 7 Beaumont Pond 29/07/2014 9 1 Beaumont Pond 12/08/2014 13 7 Beaumont Pond 24/09/2014 17 2 Cipperley s Reservoir 12/05/2014 1,500 * NA Cipperley s Reservoir 28/05/2014 11 NA Cipperley s Reservoir 08/06/2014 65 0 Cipperley s Reservoir 22/06/2014 70 5 Cipperley s Reservoir 06/07/2014 73 15 Cipperley s Reservoir 19/07/2014 14 2 Cipperley s Reservoir 01/08/2014 25 4 Cipperley s Reservoir 19/08/2014 22 4 Cipperley s Reservoir 25/09/2014 18 2 Daysland Pond 06/05/2014 400 * NA Daysland Pond 23/05/2014 22 NA Daysland Pond 03/06/2014 57 2 Daysland Pond 18/06/2014 43 4 Daysland Pond 02/07/2014 42 14 Daysland Pond 16/07/2014 34 12 Daysland Pond 29/07/2014 47 18 Daysland Pond 13/08/2014 18 11 Daysland Pond 24/09/2014 23 16 Innisfree Pond 13/05/2014 1,600 * NA Innisfree Pond 24/05/2014 14 NA Innisfree Pond 05/06/2014 64 1 Innisfree Pond 20/06/2014 44 0 Innisfree Pond 03/07/2014 62 7 Innisfree Pond 17/07/2014 30 5 Innisfree Pond 30/07/2014 18 4 Innisfree Pond 14/08/2014 15 6 Innisfree Pond 26/09/2014 7 0 62

Appendix 7. Continued. Waterbody Date Fish captures Recaptures of tagged fish Irma Pond 09/05/2014 400 * NA Irma Pond 24/05/2014 14 NA Irma Pond 05/06/2014 41 0 Irma Pond 18/06/2014 63 7 Irma Pond 03/07/2014 52 10 Irma Pond 17/07/2014 50 15 Irma Pond 30/07/2014 41 16 Irma Pond 14/08/2014 21 14 Irma Pond 26/09/2014 16 9 Lamont Reservoir 20/05/2014 2,000 * NA Lamont Reservoir 26/05/2014 11 NA Lamont Reservoir 07/06/2014 24 0 Lamont Reservoir 21/06/2014 68 1 Lamont Reservoir 15/07/2014 6 1 Lamont Reservoir 13/08/2014 7 1 Lamont Reservoir 26/09/2014 1 0 Mirror Reservoir 25/06/2014 3,000 * NA Mirror Reservoir 01/07/2014 9 NA Mirror Reservoir 15/07/2014 6 1 Mirror Reservoir 12/08/2014 1 0 Mirror Reservoir 23/09/2014 7 0 Mound Red Reservoir 16/05/2014 3,000 * NA Mound Red Reservoir 25/05/2014 22 NA Mound Red Reservoir 06/06/2014 26 0 Mound Red Reservoir 19/06/2014 13 1 Mound Red Reservoir 04/07/2014 14 1 Mound Red Reservoir 18/07/2014 9 3 Mound Red Reservoir 15/08/2014 10 2 Mound Red Reservoir 26/09/2014 20 9 Nuggent Pond 16/05/2014 1,000 * NA Nuggent Pond 26/05/2014 51 NA Nuggent Pond 07/06/2014 83 5 Nuggent Pond 19/06/2014 83 7 Nuggent Pond 06/07/2014 49 12 Nuggent Pond 19/07/2014 58 17 Nuggent Pond 31/07/2014 34 12 Nuggent Pond 15/08/2014 6 3 Nuggent Pond 26/09/2014 13 2 63

Appendix 7. Continued. Waterbody Date Fish captures Recaptures of tagged fish Parlby Reservoir 07/05/2014 1,000 * NA Parlby Reservoir 22/05/2014 2 NA Parlby Reservoir 03/06/2014 2 0 Parlby Reservoir 16/06/2014 1 0 Parlby Reservoir 14/07/2014 7 0 Radway Pond 21/05/2014 1,000 * NA Radway Pond 27/05/2014 15 NA Radway Pond 08/06/2014 50 0 Radway Pond 22/06/2014 18 1 Radway Pond 05/07/2014 63 5 Radway Pond 20/07/2014 31 3 Radway Pond 01/08/2014 19 1 Radway Pond 16/08/2014 11 0 Radway Pond 23/09/2014 2 0 Westlock Pond 15/05/2014 2,000 * NA Westlock Pond 25/05/2014 42 NA Westlock Pond 06/06/2014 37 3 Westlock Pond 20/06/2014 28 1 Westlock Pond 04/07/2014 7 0 Westlock Pond 18/07/2014 39 1 Westlock Pond 31/07/2014 19 2 Westlock Pond 17/08/2014 1 0 Westlock Pond 23/09/2014 2 0 * Number of fish stocked into waterbody. 64

Appendix 8. Estimates, standard errors and 95% confidence intervals for survival (φ) and recapture (p) parameters at ponds where sufficient recapture data allowed for parameter estimation. Beaumont Pond: Estimates, standard errors, and 95% confidence intervals for survival (φ) and recapture (p) parameters. Estimates derived from model averaging of the top two models: Φ(NH3) p(t1,t8), Φ(.) p(t). Parameter Estimate Standard error Lower 95% CI Upper 95% CI Φ (t1-t2) 0.962 0.056 0.556 0.998 Φ (t3-t8) 0.658 0.168 0.309 0.892 p (t1) 0.016 0.006 0.008 0.032 p (t2) 0.070 0.019 0.041 0.116 p (t3) 0.056 0.012 0.036 0.085 p (t4) 0.044 0.024 0.015 0.122 p (t5) 0.049 0.019 0.023 0.102 p (t6) 0.021 0.032 0.001 0.312 p (t7) 0.028 0.050 0.001 0.524 p (t8) 0.011 0.060 0.000 0.997 Cipperley s Reservoir: Estimates, standard errors and 95% confidence intervals for survival (φ) and recapture (p) parameters. Estimates derived from the Φ(.) p(t) model. Parameter Estimate Standard error Lower 95% CI Upper 95% CI Φ (t1-t8) 0.922 0.080 0.573 0.991 p (t1) 0.006 0.004 0.002 0.019 p (t2) 0.051 0.013 0.030 0.084 p (t3) 0.061 0.019 0.032 0.112 p (t4) 0.071 0.028 0.033 0.148 p (t5) 0.015 0.009 0.005 0.048 p (t6) 0.028 0.017 0.008 0.090 p (t7) 0.027 0.019 0.007 0.101 p (t8) 0.027 0.026 0.004 0.158 65

Daysland Pond: Estimates, standard errors and 95% confidence intervals for survival (φ) and recapture (p) parameters. Estimates derived from model averaging of the top three models: Φ(chl-a) p(t), φ(.) p(t), Φ(DO) p(t1,t8). Parameter Estimate Standard error Lower 95% CI Upper 95% CI Φ (t1) 0.994 0.025 0.946 1.0 Φ (t2) 0.992 0.028 0.937 1.0 Φ (t3) 0.983 0.043 0.300 1.0 Φ (t4) 0.989 0.032 0.229 1.0 Φ (t5) 0.987 0.034 0.279 1.0 Φ (t6) 0.952 0.088 0.305 0.999 Φ (t7) 0.952 0.098 0.226 0.999 Φ (t8) 0.747 0.296 0.120 0.985 p (t1) 0.063 0.015 0.039 0.101 p (t2) 0.146 0.026 0.102 0.205 p (t3) 0.115 0.023 0.077 0.168 p (t4) 0.111 0.024 0.072 0.168 p (t5) 0.093 0.027 0.052 0.160 p (t6) 0.128 0.032 0.077 0.205 p (t7) 0.067 0.034 0.024 0.172 p (t8) 0.460 0.456 0.023 0.969 Innisfree Trout Pond: Estimates, standard errors and 95% confidence intervals for survival (φ) and recapture (p) parameters. Estimates derived from the Φ(.) p(t1,t8) model. Parameter Estimate Standard error Lower 95% CI Upper 95% CI Φ 0.815 0.043 0.716 0.884 p (t1) 0.010 0.006-0.001 0.021 p (t2-t7) 0.012 0.006 0.004 0.032 p (t8) 0.071 0.018 0.042 0.115 66

Irma Trout Pond: Estimates, standard errors and 95% confidence intervals for survival (φ) and recapture (p) parameters. Estimates derived from model averaging of the top two models: Φ(chl-a) p(t1,t8), Φ(temp) p(t1,t8). Parameter Estimate Standard error Lower 95% CI Upper 95% CI Φ (t1-t6) 1.000 0.000 1.000 1.000 Φ (t7) 0.464 0.117 0.256 0.684 Φ (t8) 0.950 0.059 0.626 0.995 p (t1) 0.038 0.011 0.021 0.068 p (t2) 0.117 0.009 0.102 0.135 p (t3) 0.117 0.009 0.102 0.135 p (t4) 0.117 0.009 0.102 0.135 p (t5) 0.117 0.009 0.102 0.135 p (t6) 0.117 0.009 0.102 0.135 p (t7) 0.117 0.009 0.102 0.135 p (t8) 0.102 0.049 0.039 0.245 Mound Red Reservoir: Estimates, standard errors and 95% confidence intervals for survival (φ) and recapture (p) parameters. Estimates derived from model averaging of the top seven models: Φ(temp) p(t1,t7), Φ(DO) p(t1,t7), Φ(temp) p(temp), Φ(DO) p(do), Φ(chl-a) p(t1,t7), Φ(NH3) p(t1,t7), Φ(t1,t7) p(.). Parameter Estimate Standard error Lower 95% CI Upper 95% CI Φ (t1) 0.294 0.135 0.105 0.598 Φ (t2) 0.530 0.162 0.240 0.802 Φ (t3) 0.734 0.151 0.376 0.926 Φ (t4) 0.814 0.155 0.369 0.970 Φ (t5) 0.956 0.071 0.440 0.998 Φ (t6) 0.943 0.078 0.492 0.996 Φ (t7) 0.900 0.144 0.279 0.995 p (t1) 0.031 0.019 0.009 0.098 p (t2) 0.053 0.029 0.018 0.147 p (t3) 0.055 0.029 0.019 0.149 p (t4) 0.057 0.031 0.019 0.160 p (t5) 0.060 0.038 0.017 0.193 p (t6) 0.060 0.038 0.017 0.191 p (t7) 0.149 0.260 0.003 0.907 67

Nuggent Pond: Estimates, standard errors and 95% confidence intervals for survival (φ) and recapture (p) parameters. Estimates derived from model averaging of the top seven models: Φ(.) p(t), Φ(DO) p(t), Φ(NH3) p(t), Φ(chl-a) p(t), Φ(temp) p(t), Φ(DO) p(t), Φ(t1) p(t). Parameter Estimate Standard error Lower 95% CI Upper 95% CI Φ (t1-t5) 1.000 0.000 1.000 1.000 Φ (t6) 1.000 0.002 0.996 1.004 Φ (t7) 1.000 0.000 1.000 1.000 Φ (t8) 0.997 2.705-4.304 6.299 p (t1) 0.049 0.011 0.032 0.074 p (t2) 0.082 0.013 0.059 0.112 p (t3) 0.088 0.014 0.064 0.119 p (t4) 0.052 0.011 0.034 0.078 p (t5) 0.064 0.012 0.044 0.092 p (t6) 0.038 0.009 0.023 0.061 p (t7) 0.013 0.006 0.006 0.030 p (t8) 0.006 0.051-0.093 0.105 68

Radway Pond: Estimates, standard errors and 95% confidence intervals for survival (φ) and recapture (p) parameters. Estimates derived from model averaging of the top four models: Φ(.) p(turb), Φ(temp) p(temp), Φ(DO) p(do), Φ(t1,t8) p(turb). Parameter Estimate Standard error Lower 95% CI Upper 95% CI Φ (t1) 0.952 0.068 0.523 0.997 Φ (t2) 0.934 0.074 0.578 0.993 Φ (t3) 0.933 0.073 0.585 0.993 Φ (t4) 0.898 0.100 0.511 0.987 Φ (t5) 0.715 0.234 0.209 0.960 Φ (t6) 0.787 0.222 0.217 0.980 Φ (t7) 0.826 0.181 0.286 0.983 Φ (t8) 0.699 0.278 0.148 0.969 p (t1) 0.024 0.010 0.010 0.055 p (t2) 0.035 0.010 0.020 0.061 p (t3) 0.036 0.010 0.021 0.063 p (t4) 0.091 0.041 0.037 0.210 p (t5) 0.065 0.046 0.016 0.237 p (t6) 0.057 0.031 0.019 0.157 p (t7) 0.046 0.031 0.012 0.158 p (t8) 0.051 0.040 0.010 0.213 69

Westlock Pond: Estimates, standard errors and 95% confidence intervals for survival (φ) and recapture (p) parameters. Estimates derived from model averaging of the top two models: Φ(NH3) p(nh3), Φ(.) p(.). Parameter Estimate Standard error Lower 95% CI Upper 95% CI Φ (t1) 0.940 0.096 0.360 0.998 Φ (t2) 0.640 0.145 0.341 0.859 Φ (t3-t5) 0.940 0.096 0.360 0.998 Φ (t6-t8) 0.640 0.145 0.341 0.859 p (t1) 0.029 0.009 0.016 0.051 p (t2) 0.029 0.009 0.016 0.052 p (t3) 0.022 0.009 0.010 0.050 p (t4) 0.017 0.012 0.004 0.065 p (t5) 0.027 0.008 0.015 0.049 p (t6) 0.029 0.009 0.016 0.052 p (t7) 0.029 0.009 0.016 0.052 p (t8) 0.029 0.009 0.016 0.052 70

Alberta Conservation Association acknowledges the following partner for its generous support of this project: