Download This Document

BAY OF QUINTE REMEDIAL ACTION PLAN MONITORING REPORT #15

PROJECT QUINTE ANNUAL REPORT 2004

prepared by

Project Quinte members

in support of the Bay of Quinte Remedial Action Plan

Bay of Quinte Remedial Action Plan Kingston, Ontario, Canada.

November 2006

Editors Note: This report does not constitute publication. Many of the results are preliminary findings. The information has been provided to assist and guide the Bay of Quinte Remedial Action Plan. The information and findings cannot be used in any manner or quoted without the consent of the individual authors. Individual authors should be contacted prior to any other proposed application of the data herein.

FOREWORD

BAY OF QUINTE REMEDIAL ACTION PLAN MONITORING REPORT #15

2004 PROJECT QUINTE ANNUAL REPORT

In 1985, the Great Lakes Water Quality Board of the International Joint Commission (IJC) identified 42 Areas of Concern in the Great Lakes basin where the beneficial uses of the water were impaired. The Board recommended that the appropriate jurisdictions and government agencies prepare, submit and implement a Remedial Action Plan (RAP) in each area to restore the water uses.

The Bay of Quinte was designated as one of the Areas of concern. Ten of 14 beneficial uses described in Annex 2 of the Great lakes Water Quality Agreement (revised 1987) are impaired. The impaired uses include beach postings, eutrophication or undesirable algae, restrictions on fish consumption, taste and odour problems in drinking water, etc. The contributing factors are excessive phosphorus loadings, persistent toxic contaminants, bacteriological contamination, as well as alterations and destruction of shorelines, wetland and fish habitat.

The preparation of the Bay of Quinte RAP was initiated in 1986 with the formation of a joint federal-provincial Coordinating Committee. A Public Advisory Committee (PAC) was established in 1988. The PAC provided public commentary and community liaison to RAP development, programs and projects. This fruitful partnership contributed to the creation of RAP Stage 1 (1990) and Stage 2 (1993) reports, a series of technical documents, and numerous public outreach resources

In 1997, the Bay of Quinte RAP was reorganized. A new, locally managed implementation group was formed called the Restoration Council to provide a local voice to facilitate and oversee cleanup actions. The PAC also evolved and became a distinct incorporated entity known as Quinte Watershed Cleanup but still committed to on-going RAP progress.

The Bay of Quinte RAP Restoration Council is assisted by two organizations: (i) Quinte Watershed Cleanup (QWC) and (ii) Project Quinte scientists. The Bay of Quinte RAP program relies on the QWC for public input and Project Quinte scientists for ongoing scientific and technical advice. The contributions from both groups cannot be understated. The Bay of Quinte RAP has progressed smoothly, met the expectations of the public-at-large and incorporated an ecosystem approach because of the two organizations.

Project Quinte is a long-term, multi-agency research and monitoring project. The project's original objectives included studying, comparing and evaluating the Bay of Quinte limnological attributes (biological, chemical and physical) before and after phosphorus control was implemented at municipal sewage treatment plants thus reducing phosphorus loads to the

bay. Project Quinte began in 1972 and continues today. The time between 1972 and 1977 is referred to as the pre-phosphorus control period, while post-1977 is the post-phosphorus control period. Changes in water quality, as well as aquatic communities (e.g. phytoplankton, zooplankton, benthic and fish) between pre- and post-periods have been compared. Phosphorus control strategies were assessed. Finally, long-term ecosystem responses within the Bay of Quinte were examined and described. This information has been reported annually in a Project Quinte Annual Report. A summary of the research was also presented in Special Publication No. 86 of the Canadian Journal of Fisheries and Aquatic Sciences in 1986 and in many papers in the scientific literature.

A major focus of Project Quinte is to assess changes in lower trophic level productivity in response to the invasion of the bay by zebra mussels that began in 1994. The response to invasion of the bay by both zebra mussels and other non-native species of both zooplankton and fish and how changes are transferred up the food chain is currently being addressed. Clearly, managing both the fishery and phosphorus loadings during alterations in the food chain presents a challenge that requires the yearly productivity assessment provided by Project Quinte.

This report is the fifteenth in a series of water quality monitoring reports produced by the Bay of Quinte RAP. At this time, the information is background reference material. The information has been compiled and reported so that (1) the Bay of Quinte RAP Restoration Council and QWC can formulate abatement and remedial options, and (2) as a means to monitor remedial actions and report progress.

By way of this foreword, the Bay of Quinte RAP Restoration Council notifies potential users that this report does not constitute publication. Many of the results are preliminary findings. The information has been provided to assist and guide the Bay of Quinte Remedial Action Plan. The information and findings cannot be used in any manner or quoted without the consent of the individual authors. Individual authors should be contacted prior to any other proposed application of the data herein.

The Bay of Quinte RAP, when completed, will contain a summary of the findings and conclusions of these reports.

iii

PREFACE

This report has been prepared as part of the Bay of Quinte RAP under the auspices of the Canada-Ontario Great Lakes Remedial Action Plan program. Financial assistance and technical sponsorship for the investigations and research was provided by Fisheries and Oceans Canada, Environment Canada, the Ontario Ministry of the Environment and the Ontario Ministry of Natural Resources.

This report is the 15th in the Bay of Quinte RAP monitoring report series. It is part of the investigations, remedial action assessment and ecosystem monitoring conducted in support of the Bay of Quinte Remedial Action Plan. The report presents preliminary findings and data. The information presented does not necessarily represent the view or policies of the sponsoring agencies.

TABLE OF CONTENTS

INTRODUCTION...... 5

WATER TEMPERATURE MONITORING IN THE BAY OF QUINTE: 2001 TO 2004 ...... 7 K.E. Leisti and S.E. Doka

POINT SOURCE PHOSPHORUS LOADINGS 1965 TO 2004...... 8 P. Kinstler

THE MICROBIAL FOOD WEB OF THE BAY OF QUINTE, 2000-2004 ...... 11 M. Munawar and H. Niblock

PHOTOSYNTHESIS, CHLOROPHYLL A, AND LIGHT EXTINCTION ...... 23 M. Burley and E.S. Millard

NUTRIENTS AND PHYTOPLANKTON...... 40 K.H. Nicholls and E.S. Millard

ZOOPLANKTON IN THE BAY OF QUINTE ...... 52 K.L. Bowen and O.E. Johannsson

BENTHIC FAUNA IN THE BAY OF QUINTE...... 78 R. Dermott and R. Bonnell

BAY OF QUINTE FISH AND FISHERIES, 2004...... 89 J.A. Hoyle

ECOSYSTEM MODELLING: COMPARING THE BAY OF QUINTE AND ONEIDA LAKE ...... 100 M.A. Koops, E.S. Millard and C.K. Minns

2004 SUBMERGED MACROPHYTE SURVEY ...... 114 K.E. Leisti

EMPIRICAL DETERMINATION OF THE EPIPHYTE:MACROPHYTE RATIO FOR THE BAY OF QUINTE...... 128 K.E. Leisti and M. Burley

2004 LOWER TROPHIC LEVEL COMPARISON OF NEARSHORE AND OFFSHORE HABITATS IN THE BAY OF QUINTE, LAKE ONTARIO...... 141 K.L. Bowen, R. Dermott, O.E. Johannsson, M. Munawar and M. Burley

INTRODUCTION

Project Quinte is a co-operative, multi-agency, research and monitoring project between the federal (Department of Fisheries & Oceans) and provincial governments (Ontario Ministry of Natural Resources, Ontario Ministry of the Environment) that has investigated the long-term effects of the reduction in point-source phosphorus (P) loadings (Minns et al., 1986), food-chain influences (Nicholls and Hurley 1989), and more recently zebra mussel colonization (RAP monitoring reports Nos.7-12) on trophic dynamics of the entire bay ecosystem. The Bay of Quinte is one of 42 severely impaired ecosystems (Area of Concern - AOC) on the Great Lakes. Annex 2 of the Revised Great Lakes Water Quality Agreement of 1978 outlined a three stage Remedial Action Plan process that called for the identification of impaired beneficial uses (Stage I), their causes and a plan to be implemented (stage II) to restore these uses. The third and final stage of the RAP process requires monitoring to measure the success of RAP implementation that should ultimately lead to delisting the Bay of Quinte as an AOC. Project Quinte has been invaluable to stage three of the Remedial Action Plan (RAP), supporting an essential, continuing program of research and monitoring in the bay. Project Quinte has contributed long-term data that was used to produce comprehensive assessments of the status of two impaired beneficial uses, phytoplankton (Nicholls, Monitoring Report #11) and zooplankton (Johannsson and Nicholls, Monitoring Report #12) Project Quinte presented the first evidence of the impact of zebra mussels on the Bay of Quinte ecosystem in the 1995 monitoring report. Since that time we have observed variable impact, both spatially and inter-annually, of this invader on physical, chemical and biological properties of the bay (e.g. transparency, phosphorus, and lower trophic level biomass and production). There are no other Great Lakes studies that have the benefit of the extensive multi-trophic level database that exists for the Bay of Quinte to determine ecosystem level impacts of Dreissenid mussels. We stand to learn a great deal about the impacts of this exotic species on fisheries productivity in the Bay of Quinte that may be of use in other ecosystems where background data is not as extensive. The scope of the work in Project Quinte is rare in freshwater ecosystem research, now encompassing multi-year, multi-trophic level data from bacteria to fish. These data are being used to develop an ecosystem model for the bay that is being compared to a similar model being developed for Oneida Lake, another productive ecosystem that has undergone similar invasions by exotic species and declines in key fish species. See the chapter by Koops et al. in this report for a

synopsis of the second year’s progress on the ecosystem modeling project funded by the Great Lakes Fisheries Commission. Long-term data sets for biomass, species composition and production for all trophic levels provide a unique opportunity to model trophic interactions in the bay and determine how various factors are impacting important fish populations. An investigation on the spatial and seasonal extent of taste and odour and algal toxicity problems in the bay was conducted for the first time in 2003 and is reported here by Dr. Susan Watson of Environment Canada. Proliferation of the cyanonbacteria Microcystis has been a common occurrence in many parts of the Great Lakes including the Bay of Quinte since zebra mussels have been established. This algal produces the neuro-toxin microcystin, a potential health threat. Initiating a pilot study in 2003 on the extent of this problem in the bay was considered a high priority. This report provides a detailed summary of the results and highlights for the 2003 field season, prepared by various scientists studying the bay. Please contact these individuals directly if further explanation or detail is required.

Link to the Remedial Action Plan Progress towards restoration goals must be measured in order to delist impaired beneficial uses. The ongoing monitoring provided by Project Quinte is invaluable in assessing restoration progress as the federal government attempts to delist this AOC. The restoration goals originally set as part of the stage II development were considered to be the interim. Project Quinte has been instrumental in developing more realistic goals for phytoplankton, zooplankton, benthos and other related beneficial uses. Information on listing criteria and delisting goals has been summarized in the report, “Bay of Quinte RAP Monitoring and Delisting Strategy IBU Assessment Statements 2003” prepared for the Bay of Quinte Restoration Council under contract to Murray German Consulting and Fred Stride Environmental.

WATER TEMPERATURE MONITORING IN THE BAY OF QUINTE: 2001 TO 2004

K. E. Leisti and S.E. Doka

Great Lakes Laboratory for Fisheries and Aquatic Sciences Fisheries and Oceans Canada 867 Lakeshore Road, P.O. Box 5050 Burlington, Ontario, L7R 4A6

Since 2001, water temperatures have been recorded from April to November in the Bay of Quinte. In that year, Onset Tidbit loggers were deployed at a 1 m depth on the offshore mooring buoys used for Project Quinte at Belleville, Big Bay, Napanee, Hay Bay, Glenora and Conway. The following year, additional loggers at 10 m depth intervals were deployed at the deeper sites of Hay Bay, Glenora and Conway and this monitoring has continued into 2006. Nearshore (<1.5 m depth) water temperatures were also recorded at Muscote Bay, Big Bay, Napanee and Carnachan Bay from 2001 to 2004. As part of an IJC study, water temperature monitoring was also conduced in wetlands at Robinsons Cove and Hay Bay from 2002 to 2004. Temperatures were sampled at half hour intervals between April and November, but in 2003, loggers at selected nearshore and wetland locations were left out over the winter and following spring. Detailed information on this temperature monitoring program can be found in the Canadian Manuscript Report of Fisheries and Aquatic Sciences technical report entitled Water Temperature Monitoring of Selected Lake Ontario Wetlands and the Bay of Quinte, 2001 – 2004 by K.E. Leisti and S.E. Doka (In Prep.).

POINT SOURCE PHOSPHORUS LOADINGS 1965 TO 2004

Peter Kinstler

Ontario Ministry of the Environment, 133 Dalton Avenue, P.O. Box 820, Kingston, Ontario K7L 4X6

Total phosphorus loads to the Bay of Quinte from six municipal sewage treatment plants (STP) bordering the Bay of Quinte between 1965 and 2004 were derived from routine Ontario Ministry of the Environment (MOE) monitoring data (Tab. 1). Phosphorus loadings are reported as kilograms per day (kg d-1). Loadings are calculated as the monthly average effluent phosphorus concentration multiplied by the monthly average effluent flow from the STP for the period of record. In 2004, the estimated total phosphorus loads from STPs bordering the Bay of Quinte were 13.2 kg d-1 for the year and 1.2 kg d-1 for the May to October period. Since 1991, low total phosphorus inputs have been maintained by controlling variability in wastewater flow, enhancing phosphorus removal capabilities and expanding treatment capacity. To manage long term phosphorus loading, the MOE advised municipalities that a Bay of Quinte RAP “phosphorus load cap” would be applied to STPs bordering the Bay of Quinte. Initially, the “load cap” was to be calculated as the product of a total phosphorus effluent concentration of 0.3 mg l-1 and the STP hydraulic capacity for STPs bordering the Bay of Quite. This “load cap” would be used post 1995 to measure and assess performance. Since 1995, most of the municipal STPs have voluntarily reduced total phosphorus inputs or have had upgrades where the “load cap” has been incorporated into the STP’s operating Certificate of Approval. It is also noted that data has been obtained from the 8 Wing Trenton STP for 1998 through to 2004. Similarly, the MOE has advised municipalities that a Bay of Quinte RAP phosphorus load cap has been applied in the Bay of Quinte watershed in 1995. The watershed “load

cap” has been initially established as the product of a total phosphorus effluent concentration of 0.5 mg l-1 and the STP hydraulic capacity. Most STPs in the watershed have voluntarily reduced total phosphorus inputs or have had upgrades where the “load cap” has been incorporated into the STP’s operating Certificate of Approval. Complete bypassing events are not accurately reported in the data as presented. Appreciable bypassing continues to occur at some STPs adding to the phosphorus loading to the Bay of Quinte. Continued efforts are required to control wastewater influent flow variability and reduce bypassing to effectively “cap” phosphorus inputs to the Bay of Quinte.

Table 1. Phosphorus loadings from municipal sewage treatment plants bordering the Bay of Quinte (total P kgxd-1) from 1965 to 2004.

Period T CFB B D N P TOTAL 1965-72 53.0 13.0 110.0 2.7 20.0 15.0 214.0 1973-75 47.0 5.9 73.0 2.3 17.0 11.0 156.0 1976-77 43.0 2.1 39.0 2.3 63.0 9.1 158.0 1978-86 Jan-Dec 7.5 1.9 31.0 0.9 24.0 2.4 68.0 May-Oct 6.3 1.7 25.0 0.7 24.0 1.9 60.0 1987 Jan –Dec 5.9 2.7 16.0 0.6 11.0 2.1 38.0 May-Oct 6.4 2.0 15.0 0.5 6.2 1.8 32.0 1988 Jan-Dec 6.9 4.7 15.0 0.7 9.2 2.2 42.0 May-Oct 6.3 1.7 11.0 0.5 12.4 1.9 34.0 1989 Jan-Dec 6.0 * 13.0 0.7 5.6 1.4 26.7 May-Oct 5.3 * 11.0 0.6 5.9 1.0 23.8 1990 Jan-Dec 5.9 * 14.0 1.0 4.1 1.8 26.8 May-Oct 6.5 * 9.0 0.6 3.8 1.5 21.4 1991 Jan-Dec 6.8 * 9.2 1.0 2.6 1.7 21.3 May-Oct 5.2 * 4.6 0.8 2.3 0.9 13.8 1992 Jan-Dec 5.9 * 12.3 1.4 2.5 1.5 23.6 May-Oct 5.1 * 9.6 0.9 2.3 0.8 18.7

Table 1. Cont’d Period T CFB B D N P TOTAL 1993 Jan-Dec 6.1 * 12.1 0.9 3.0 1.0 23.1 May-Oct 4.6 * 10.1 0.3 2.8 0.7 18.5 1994 Jan-Dec 4.2 * 18.5 0.5 2.0 1.3 26.5 May-Oct 3.3 * 12.1 0.4 1.4 0.9 18.1 1995 Jan-Dec **4.2 * 14.3 0.3 1.5 1.2 21.5 May-Oct **3.3 * 14.2 0.2 1.1 0.9 19.7 1996 Jan-Dec 3.4 * 18.4 0.4 1.4 1.7 25.3 May-Oct 4.4 * 14.9 0.3 1.2 1.1 21.9 1997 Jan-Dec 4.0 * 16.7 0.5 1.8 1.4 24.3 May-Oct 3.2 * 13.2 0.4 1.4 0.6 18.8 1998 Jan-Dec 2.3 0.7 11.6 0.4 1.5 1.1 16.9 May-Oct 2.3 0.6 11.8 0.4 1.3 0.6 16.4 1999 Jan-Dec 3.3 0.6 6.8 0.7 1.6 1.3 13.6 May-Oct 3.5 0.5 7.0 0.6 1.3 0.8 13.2 2000 Jan-Dec 2.0 0.7 6.1 0.6 1.5 1.5 12.4 May-Oct 2.5 0.8 7.3 0.5 1.4 0.8 13.4 2001 Jan-Dec 1.5 0.5 4.8 0.2 1.3 1.8 10.1 May-Oct 1.4 0.5 3.5 0.1 1.3 1.0 7.8 2002 Jan-Dec 1.7 0.5 7.0 0.9 1.3 1.2 12.6 May-Oct 1.9 0.4 9.2 0.4 1.4 1.2 14.5 2003 Jan-Dec 2.2 0.7 13.0 0.2 4.0 2.2 22.3 May-Oct 2.8 0.8 4.0 0.2 1.0 0.5 9.3 2004 Jan-Dec 3.3 0.4 7.0 0.2 1.4 0.9 13.2 May-Oct 3.2 0.4 2.0 0.2 0.9 0.7 1.2 * CFB Trenton loadings not available. ** No 1995 records available, 1994 loadings substituted to estimate total loading. T=Trenton, CFB=Canadian Forces Base Trenton, B=Belleville, D=Deseronto, N=Napanee, P=Picton

THE MICROBIAL FOOD WEB OF THE BAY OF QUINTE, 2000-2004

M. Munawar and H. Niblock

Department of Fisheries and Oceans 867 Lakeshore Rd., Burlington, ON, Canada L7R 4A6

The importance of the microbial food web to aquatic ecosystems has been well established (e.g. Munawar et al., 1999, 2003). Principle components of the microbial food web include bacteria, autotrophic picoplankton (APP), phytoplankton, heterotrophic nanoflagellates (HNF), ciliates and zooplankton. Microbial food web samples have been collected at two stations in the Bay of Quinte since 2000. The upper bay station at Belleville was shallow (5 m), highly eutrophic with very high total phosphorus levels and was representative of the upper bay ecosystem. The second station at Conway was much deeper (32 m) and representative of the lower reaches. The methodology used for this microbial food web assessment, including identification and enumeration of each of the principle components as well as size fractionated primary productivity, followed standard techniques (see Munawar et al., 1999, 2005) and has been previously used in Remedial Action Plan (RAP) projects (Munawar and Niblock, 2005). Phytoplankton and zooplankton data used in this analysis have been provided courtesy of Scott Millard and Ora Johannsson respectively and published elsewhere in this report.

Results

Microbial loop (Bacteria, autotrophic picoplankton and heterotrophic nanoflagellates)

The seasonal variation of the microbial communities at Belleville and Conway is shown in Fig. 1a and 2a. Microbial biomass was low during spring and early summer with a single peak at Belleville and two peaks at Conway during late summer. Biomass was higher at

20000 -3 a) 10000 mg m mg 0

100%

80% ciliates

60% APP

40% HNF

microbial loop 20% Bacteria

15000

-3 10000 b)

mg m mg 5000

100% Dinophyceae 80% Cryptophyceae Diatomeae 60% Chrysophyceae

40% Chlorophyta

phytoplankton Cyanophyta 20%

8000

6000 -3 -3 c) 4000 Predators

mg m Herb+veliger 2000

100%

80% Calanoids 60% Cyclopoids

40% Cladocerans

Zooplankton Veligers 20%

0% Oct 5 Jun 1 Jun Sep 8 Sep Jul 13 Jul 29 Oct 20 Jun 15 Jun 29 Jun May 11 May 18 11 Aug 24 Aug Sep 21 Sep Figure 1. Seasonality of planktonic biomass Belleville, 2004 a) total microbial biomass (bacteria, heterotrophic nanoflagellates, autotrophic picoplankton and ciliates; mg m-3) and % composition b) phytoplankton biomass (mg m-3) and its taxonomic composition c) total zooplankton biomass (dry weight mg m-3) as predatory and herbivourous zooplankton and its breakdown by taxonomic group

a) 8000

-3 6000 4000

mg m 2000 0

100%

80% ciliates 60% APP HNF 40% Bacteria mircorbial loop 20%

0% b) 2000 -3 1000 mg m 0 100% Dinophyceae 80% Cryptophyceae 60% Diatomeae Chrysophyceae 40% Chlorophyta phytoplankton 20% Cyanophyta

c) 1500 Predators

-3 1000 Herb+veliger

mg m mg 500

0 100%

80%

Calanoids 60% Cyclopoids

40% Cladocerans

Zooplankton Veligers 20%

0% Jun 2 Oct 6 Jul 14 Jul 28 Jul Jun 16 Jun 30 Oct 21 Aug 10 Aug 25 Sep 10 Sep 22 May 12 May 19 May Figure 2. Seasonality of planktonic biomass Conway, 2004 a) total microbial biomass (bacteria, heterotrophic nanoflagellates, autotrophic picoplankton and ciliates; mg m-3) and % composition b) phytoplankton biomass (mg m-3) and its taxonomic composition c) total zooplankton biomass (dry weight mg m-3) as predatory and herbivourous zooplankton and its breakdown by taxonomic group.

Belleville than at Conway. As observed in previous years, the microbial community was overwhelmingly dominated by heterotrophic nanoflagellates (HNF).

Phytoplankton and zooplankton

During 2004, a phytoplankton peak was observed in early summer at Belleville (Fig. 1b). Later the biomass increased steadily throughout the sampling season. The early summer peak (6.37 g m-3) consisted mainly of Diatomeae (Fig. 1b) with species of Melosira contributing 80% of the total biomass (Tab. 1). The maximum concentration of phytoplankton biomass (10.95 g m-3) was observed at Belleville during the fall. At Belleville, algae were dominated by Diatomeae (35-95%) throughout the sampling season. The occurrence of a phytoplankton summer maximum coincided partially with the herbivorous plankton peak (Fig. 1c) which lasted for three sampling periods and was mainly composed of Cladocerans. Predatory zooplankton biomass remained relatively low throughout the sampling period but showed minor increases corresponding with increases of herbivorous zooplankters. At Conway, phytoplankton biomass (Fig. 2b) was approximately 10 times lower than at Belleville. Two distinct peaks were recorded, first in mid summer (1.68 g m-3) and second in early fall (1.46 g m-3; Fig. 2b). These peaks more or less corresponded with Diatomeae pulses of Aulacosira and Fragilaria spp. (Tab. 1). Phytoplankton composition at Conway was much more diverse than at Belleville. Herbivorous zooplankton, on the other hand, showed a peak in June which decreased significantly in the following sample. However, the herbivores showed a sustained peak for four sampling periods throughout late summer before declining in the fall (Fig. 2c). Predatory zooplankton remained low with minor peaks that followed the peaks of herbivores.

Table 1. Taxonomic composition plankton at P/B (production/biomass) peaks, Belleville and Conway stations, Bay of Quinte 2004. Productivity Biomass Peak in Peak in Station Date Main taxa present (m C m-3 h-1) (m m-3) Biomass? Productivity? 80% Aulacoseira spp. Jun 29, Belleville 128.8 6379 Yes Yes 7% Cryptomonas spp. 04 5% Aulacosira gran v. ang. F. Curvata 32% Aulacoseira (Melosira) spp. 13% Anabaena spp. 5% Stephanodiscus spp. Belleville Sep 8, 04 251.3 3345 No Yes 5% Stephanodiscus Binderana 5% Aulacoseira gran. v. ang. f. curvata Oct 20, 89% Aulacoseira (Melosira) spp. Belleville 98.9 10951 Yes No 04 4% Stephanodiscus Binderana 46% Cryptomonas spp. 17% Rhodomonas spp. Conway Jun 2, 04 16.4 399 No Yes 6% Fragilaria spp. 5% Stephanodiscus Binderana 40% Aulacoseira gran. v. ang. f. curvata Aug 11, Conway 28.3 1688 Yes Yes 27% Fragilaria spp. 04 10% Cryptomonas spp. 4% Stephanodiscus spp. 76% Aulacoseira (Melosira) spp. Conway Oct 6, 04 20.7 1468 yes Yes 4% Rhodomonas spp. 4% Fragilaria spp.

Size fractionated primary production

Total primary productivity at light optimum (Popt), % contributions to primary productivity by size class and corresponding Productivity to Biomass (P/B) ratios are shown in Fig. 3. At Belleville, primary productivity ranged from 10 to 251 mg C m-3 h-1 during the May to October period with two peaks during the summer (June: 128 and Sept: 251 mg C m-3 h-1) that corresponded with P/B peaks of 0.23 and 0.39 h-1 (Fig. 3a). The dominant taxa for the June peak were Aulacoseira and Cryptomonas spp. The September peak consisted of species of Aulacoseira, Anabaena, and Stephanodiscus (Tab. 1). The observed range of P/B from 0.08 to 0.39 was similar to that reported previously (Munawar and Niblock, 2005). The relative contribution of various size fractions of phytoplankton to primary productivity is shown in Fig. 3c. At Belleville, primary productivity was dominated by nanoplankton (2-20 μm) and netplankton (>20 μm) which corresponds to the large percentage of diatoms observed in the biomass.

a) b) Belleville Conway 30 0.30 300 0.45

0.40 25 0.25 250 0.35 20 0.20 200 0.30 -1 -1 h h -3

-3 0.25 150 15 0.15 0.20 P/B carbon mg C m C mg P/B carbon mg C m C mg 100 0.15 10 0.10

0.10 50 5 0.05 0.05

0 0.00 0 0.00 l l l t t y n g p t y n n n u u g p c c a un u u ep c a ay u Ju Ju J J u ug e ep J -J S M M J - - - - A A S S O -O M -J 3 -O - 1 5 9 3 7 - - - - 5- 0 1 Jun 1 27-Jul 4-Aug 8- 5 c)1 1 2 1 2 1 4 8 1 2 d) 8 15- 29 11-Au 2 21-Se 20-Oct 1 18 1 2 2 11-May 1 production P/B production P/B 100% 100%

80% 80%

60% 60%

40% 40%

20% contribution to production to contribution 20% contribution to production to contribution

0% 0% 1 Jun 5-Oct 13-Jul 27-Jul 8-Sep 2 Jun 6-Oct 15-Jun 29-Jun 20-Oct 14-Jul 28-Jul 11-Aug 24-Aug 21-Sep 11-May 17 May 21-Oct 16-Jun 30-Jun 10-Aug 25-Aug 12 May 19 May 10-Sep 22-Sep >20 μm 2-20 μm <2 μm >20 μm 2-20 μm <2 μm

Fig. 3. Seasonality of primary productivity (mg C m-3 h-1) and production to biomass ratio (h-1) during 2004 and corresponding percent size contribution to production at a) Belleville b) Conway.

Microbial loop: Comparison of 2003 versus 2004

Standard t-tests were used to compare the microbial loop among various seasons (spring, summer, and fall) and throughout the whole sampling period. During summer, bacterial biomass at Belleville was significantly lower in 2004 compared to 2003 (p<0.05) while the ciliate biomass was higher (p<0.05). When the whole sampling period was considered, ciliates were higher in 2004 than in 2003 (p<0.05). At Conway, the t-test for the whole sampling period indicated a significant decrease in bacterial biomass (p<0.05)

and HNF (p<0.001) in 2004. HNF also declined significantly during summer, 2004 (p<0.05).

Planktonic food web

Belleville Fig. 4 presents a holistic picture of the planktonic food web as fresh weight at Belleville (a) and Conway (b). At Belleville, the planktonic food web exhibited two well developed peaks, the first observed in late June and the second in late September. The June peak was equally dominated by HNF (40%) and Diatomeae (39%). The September peak was mainly composed of HNF (68%) and Diatoms (24%). The spring planktonic structure consisted predominantly of HNF, Diatoms, Cladocerans, Bacteria and Cryptophyceaen flagellates. In the summer, the planktonic food web was composed mainly of HNF, Diatoms, Cladocerans and Cyanophyta.

Conway The planktonic food web at Conway is shown in Fig. 4b. There is a small peak in total biomass in spring (May) and well developed peaks in the mid summer (August) and late summer (September). HNF overwhelmingly dominated the microbial community (63-87% of planktonic biomass) with relatively large contributions by Diatoms at the summer peaks (Aug: 15% and Sep: 12%). In spring, the planktonic structure was overwhelmingly dominated by HNF. Other trophic groups such as Cladocerans, Bacteria and Cryptophyceae also contributed significantly. In the summer, HNF continued its dominance with contributions from Cladocerans and Diatomeae.

Planktonic carbon

The biomass of the complete planktonic food web was converted to organic carbon in order to assess linkages across various trophic levels. Conversion factors were previously reported in Munawar and Niblock (2005), and Munawar et al. (2005). Seasonally weighted

mean carbon for each trophic component (Fig. 5) indicated that at Belleville, HNF, phytoplankton and herbivorous zooplankton were likely the main vectors of energy transfer. No significant change was observed between 2003 and 2004 although the contribution of phytoplankton was slightly higher in 2004. At Conway, HNF were proportionally greater than Belleville corresponding with a reduced phytoplankton contribution. The three main components observed at Belleville (HNF, phytoplankton and herbivorous zooplankton) were also the main components at Conway.

a) b)

25000 10000

20000 8000

15000 -3 6000 -3 mg m mg m 10000 4000

5000 2000

0 0 100% 100%

75% 75%

50% 50%

25% 25%

0% 0% Jun 1 Jun Oct 5 Jul 13 Jul 29 Jul Sep 8 Jun 2 Jun Jun 15 Jun 29 Jun Oct 6 Oct Oct 20 Aug 11 Aug 24 Jul 14 Jul 28 Jul May 11 May 18 May Sep 21 Jun 16 Jun 30 Oct 21 Oct Aug 10 Aug 25 May 12 May 19 May Sep 10 Sep 22

APP Cyanophyta Chlorophyta Euglenophyta Chrysophyceae Diatomeae Cryptophyceae Dinophyceae Bacteria HNF Ciliates Dreissena veligers Cladocera Cyclopoida Calanoida

Figure 4. Seasonality of planktonic food web (mg m-3) showing the percent contribution by various planktonic components. a) Belleville and b) Conway.

Belleville Conway

2003

2004

phyto ciliates

Bacteria APP HNF herbivorous zooplankton predatory zooplankton

Figure 5. Seasonally weighted mean planktonic carbon contributions to the total biomass in 2003 and 2004 at Belleville and Conway respectively.

Summary

• As observed in previous years, the microbial communities at Bellville and Conway were overwhelmingly dominated by heterotrophic nanoflagellates (HNF). • At Belleville, phytoplankton was dominated by Diatomeae (35-95%) throughout the sampling season which coincided mostly with the dominance of cladecerans • At Conway, phytoplankton biomass was approximately 10 times lower than at Belleville. The observed peaks coincided with Diatomeae pulses comprised of Melosira and Fragilaria species. • Phytoplankton composition at Conway was much more diverse than at Belleville. Herbivorous zooplankton showed a sustained peak for four sampling periods during late summer before dropping in the fall. • Belleville primary productivity ranged from 10 - 251 mg C m-3 h-1 over the sampling season (May to October) with two peaks during the summer (June: 128 and Sept: 251 mg C m-3 h-1). These peaks corresponded with P/B peaks of 0.23 and 0.39 h-1. The bulk of primary productivity was contributed by nanoplankton (2-20 μm) and netplankton (>20 μm) which corresponded to high diatom biomass. • At Conway, total primary productivity ranged from 3 to 28 mg C m-3 h-1. Two peaks were observed, one in early summer (June: 16 mg C m-3 h-1) and one during mid summer (August: 28 mg C m-3 h-1). While the June peak in primary productivity corresponded with a peak in P/B (0.13 h-1), the August peak did not. The phytoplankton standing crop during the June peak was composed of species of Cryptomonas, Rhodomonas and Fragilaria. During the August peak, however, a slightly lower P/B ratio was observed (0.11 h-1). The phytoplankton standing crop included species of Aulacoseira, Fragilaria and Cryptomonas. The P/B peak at Conway was observed during September (0.26 h-1) when species of Aulacoseira, Fragilaria and Cryptomonas dominated. Size fractionated primary productivity results showed considerable variability.

• During spring, the planktonic food web at Belleville consisted mostly of HNF, Diatoms, Cladocerans, Bacteria and Cryptophyceae. During summer, the planktonic food web was composed mainly of HNF, Diatoms, Cladocerans and Cyanophyta. • At Conway, the spring planktonic food web structure was overwhelmingly dominated by HNF. Other trophic groups such as Cladocerans, Bacteria and Cryptophyceae also contributed. The summer was also dominated by HNF with lesser contributions from Cladocerans and Diatomeae. • The organic carbon pool at Belleville and Conway was overwhelmingly dominated by the heterotrophic nanoflagellates (HNF) with considerably lower contributions from phytoplankton and herbivorous zooplankton. Suggesting that a different route of energy transfer than normally believed. • The continued dominance of heterotrophic over autotrophic organisms in the planktonic food web of Quinte is a unique and noteworthy observation since the phosphorus abatement program was designed primarily to control eutrophication by controlling the excessive growth of autotrophic phytoplankton.

Acknowledgements

We thank Scott Millard, Ora Johannsson, Milos Legner (University of Toronto) Denis Lynn (University of Guelph) and Mark Fitzpatrick for various aspects of the project and preparation of this report. We also acknowledge several DFO personnel who assisted in the field and experimental work: Jocelyn Gerlofsma, Michele Burley, Silvina Carou, Peter Jarvis, and Bud Timmins.

References

Munawar, M., Niblock, H., 2005 Assessing and comparing the structure of the planktonic food web at a stressed station in the Bay of Quinte, 2002-2003 In: Monitoring report #14. Project Quinte Annual Report 2003, pp. 117-130. Bay of Quinte Remedial Action Plan, Kingston, Ontario.

Munawar, M., Munawar, I.F., Weisse, T., van Stam, H., Fitzpatrick, M., Lorimer, J., Wenghofer, C., Carou, S., 1999. Probing microbial food-web structure in Lake Erie. In: M. Munawar, T. Edsall, I.F. Munawar (Eds.), State of Lake Erie: Past, Present and Future, pp. 417-439. Backhuys Publishers, Leiden, The Netherlands.

Munawar, M., Legner, M., Munawar, I.F., 2003. Assessing the microbial food web of Lake Ontario. In: M. Munawar (Ed.), State of Lake Ontario, Past, Present and Future, pp. 135-170. Ecovision World Monograph Series. Aquatic Ecosystem Health and Management Society, Burlington, ON.

Munawar, M., Munawar, I.F., Fitzpatrick, M., Lynn, D., 2005. A preliminary assessment of the planktonic food web in Lake Michigan. In: T. Edsall, M. Munawar (Eds.), State of Lake Michigan: Ecology, Health and Management, pp. 183-203. Ecovision World Monograph Series. Aquatic Ecosystem Health and Management Society, Burlington, ON.

PHOTOSYNTHESIS, CHLOROPHYLL A, AND LIGHT EXTINCTION

Michele Burley and Scott Millard

Great Lakes Laboratory for Fisheries and Aquatic Sciences Laboratory for Fisheries and Aquatic Sciences, Department of Fisheries and Oceans 867 Lakeshore Road, P.O. Box 5050, Burlington, Ontario, L7R 4A6

Nutrients and lower trophic level sampling cruises 2004

The 2004 field season was conducted with a new vessel named Leslie J. This 26 foot Hourston launch is the namesake of John Leslie, a renowned larval fish biologist. John’s career at Fisheries & Oceans Canada has spanned over five decades and we are privileged that he continues his life’s work as an emeritus. The Leslie J. was stationed in Picton Bay and operated by scientific personnel from Fisheries and Oceans Canada (Burlington). Thirteen cruises were conducted on alternate weeks starting in early May and ending in late October. Stations Belleville (B), Napanee (N), Hay Bay (HB), Glenora (GL) and Conway (C) were visited on every cruise (Fig. 1). Due to circumstances the Murray J. vessel was used for two of the early cruises. The majority of the field work was carried out by DFO personnel: Michele Burley, Jocelyn Gerlofsma, Bud Timmins and Rachel Nagtegaal. Physical measurements such as station depth, mixing depth, Secchi disc depth and light extinction (400-700 nm) were measured. A Hydrolab minisonde 4a was used to determine depth profiles of temperature, dissolved oxygen, pH and conductivity at each station. Samples for nutrients, phytoplankton and zooplankton biomass and species composition, chlorophyll a (Chl) and seston were collected. Phytoplankton photosynthesis experiments were conducted in an incubator using a 14C-uptake technique to obtain photosynthetic parameters required as input for programs that compute depth and time-integrated photosynthetic rates. Seasonal (May 1-Oct 31), areal (g C m-2) and volumetric (g C m-3) rates of phytoplankton photosynthesis have been calculated. Photosynthesis experiments, chlorophyll and seston

filtration as well as nutrient processing were conducted at the Ontario Ministry of Natural Resources, Glenora Fisheries Research Station.

Figure 1. Project Quinte sampling stations B, N, HB, GL, and C.

Chlorophyll a

Seasonal mean chlorophyll levels in 2004 were very close to the 2003 values at stations B and C. The high seen in chlorophyll levels at Hay Bay in 2003 (16.6 mg m-3) contrasted with the low of 9.4 mg m-3observed in 2004. At station Napanee the seasonal mean chlorophyll values dropped 26%, while at station Hay Bay they dropped 45% from 2003 values. Reduced phytoplankton biomass due to zebra mussel grazing, using seasonal mean chlorophyll a (Chl) as an index, was first observed in 1995 (Tab. 1). Averaging the seasonal means indicates a decrease of Chl concentrations between a third to one half across all stations post zebra mussel introduction. With few exceptions, station Napanee (N) had consistently higher chlorophyll levels than Belleville (B) prior to 1994. This trend reversed from 1995 to 2004.

Table 1. Seasonal means and maximum chlorophyll a concentrations (mg m-3) at stations B, N, HB and C from 1977-2004. Chlorophyll analyses were done with GF/C filters ground in 90% acetone with concentrations uncorrected for phaeopigments. In previous reports the values were corrected for phaeopigments. Data using this method are not available prior to 1977. 1977 values are the only Pre-control data available using this method. See Nicholls for 1972-1976 chlorophyll data. B N HB C Year Mean Max Mean Max Mean Max Mean Max 1977 39.3 88.2 42.1 118.1 32.0 55.0 6.0 8.9 1978 20.9 48.4 29.0 71.1 23.8 51.8 6.2 13.2 1979 27.8 53.5 30.5 72.9 26.6 65.6 8.1 18.5 1980 21.7 48.8 29.3 50.9 28.4 48.5 10.1 17.7 1981 22.2 44.9 29.1 69.2 25.2 44.8 8.8 14.2 1982 21.8 47.8 26.8 45.3 24.7 73.3 6.1 9.9 1983 20.0 48.2 na na 24.1 60.9 8.9 23.9 1984 32.4 74.7 na na 37.1 65.4 8.6 27.3 1985 31.8 54.0 na na 28.3 48.2 8.8 19.3 1986 13.9 26.7 na na 24.9 43.5 8.0 13.9 1987 33.1 72.6 na na 25.0 52.9 7.4 17.9 1988 30.4 62.3 na na 22.3 50.0 6.4 9.3 1989 21.5 43.8 31.3 69.5 29.7 71.1 7.2 12.7 1990 24.5 51.1 25.0 50.5 21.2 42.3 7.1 12.0 1991 26.0 46.6 25.9 44.1 17.1 44.2 5.8 12.4 1992 20.7 43.6 28.4 69.1 22.8 46.0 6.4 11.5 1993 20.3 45.1 23.0 50.3 20.3 41.2 6.8 14.6 1994 25.8 58.8 25.9 57.1 22.2 48.3 5.5 10.9 Post P Control Means 24.4 51.2 27.7 59.1 24.9 52.8 7.4 15.2 1995 16.2 36.2 12.8 28.3 10.5 20.3 4.7 12.1 1996 9.5 38.2 8.9 26.5 8.5 24.2 3.2 9.4 1997 10.4 21.0 8.8 21.6 7.8 18.5 2.9 5.6 1998 16.7 44.4 10.5 22.0 12.0 24.3 4.0 14.8 1999 14.0 32.4 7.4 14.3 7.8 16.7 2.8 5.4 2000 11.5 21.7 8.0 22.1 9.3 18.8 3.7 6.9 2001 15.2 39.2 11.4 20.2 9.8 19.6 3.3 5.4 2002 12.2 26.9 10.0 22.6 10.1 18.5 3.2 5.4 2003 13.0 28.3 12.7 27.0 16.6 44.5 3.1 5.3 2004 13.7 25.5 9.4 16.5 9.2 18.4 3.2 7.4 1995-2004 Means 13.2 31.4 10.0 22.1 10.2 22.4 3.4 7.8

The coefficient of variation about seasonal means at all stations for the post zebra mussel years does not change appreciably with the addition of 2004 data. Hay Bay has the highest coefficient of variation (25.6%) for this time stanza. Interestingly the coefficient of variation at B has decreased in the post zebra mussel years by 6.5%. The lowest inter-annual variability from 2000 to 2004 is observed at station Conway. Chlorophyll levels at station C in the lower bay have shown a similar impact in all years since 1995 with levels varying between 3-5

mg m-3. The seasonal means, post zebra mussels, are typically half of pre invasion levels. The seasonal means from 2001-2004 have stayed close (3.1-3.4 mg m-3). The dramatic impact that zebra mussels have had on Chl since 1995 is very apparent when the grand mean of seasonal means since 1995 are compared to the post P control mean (1978- 94). Mean Chl for these periods show declines of about 60% at stations N and HB and 50% at stations B and C. A long-term, mean seasonal chlorophyll trend line was determined for the post phosphorus control period and before zebra mussels colonized the bay (1978-94) by calculating the mean across years for sequential two-week time periods throughout the season. Confidence limits for this line are high because of high inter-annual variability in driving variables such as climate, food-chain structure and phosphorus concentration (Nicholls and Hurley, 1989). For clarity, only the most recent years have been graphed in comparison to the 1978-1994 trends (Fig. 2). With few exceptions the Chl levels are below the long-term seasonal trend lines throughout the season at all stations. Two of these exceptions occurred at station B in 2004 in late June (21.4 mg m-3) and October (25.5 mg m-3). The other occurs at station C in mid July 2004 (6.8 mg m-3). The August 10th chlorophyll value at station C approximates the trend line at 7.43 mg m-3. Overall the peaks are lower than those observed in 2003 at B, N and HB. At Hay Bay the 2003 mid-August to early September chlorophyll values remain the highest observed since 1994.

Light extinction

Secchi disk transparency and vertical light extinction (extinction coefficient) are inversely related. Transparency increases (extinction coefficients decrease) from upper to lower bay, but the upper to lower bay gradient is less dramatic in post phosphorus versus pre phosphorus control years (Tab. 2). Prior to phosphorus control, Chl levels contributed significantly to light extinction coefficients (Millard, 1986). Presently Chl levels are low enough that background light extinction may be more significant than Chl in light attenuation, even in

the upper bay. Factors contributing to this include dissolved colouring and abiotic particulate matter.

Table 2. Seasonal means and maxima for the vertical light extinction coefficient (m-1) at stations B, HB, C from 1972-2004. B HB C Year Mean Max Mean Max Mean Max 1972 1.63 2.56 1.42 2.00 0.64 1.10 1973 1.83 2.71 1.36 2.00 0.64 0.89 1974 2.02 3.18 1.43 2.63 0.72 1.03 1975 1.96 3.07 1.44 2.05 0.68 0.99 1976 1.99 3.18 1.56 2.09 0.61 0.94 1977 1.82 3.07 1.39 2.22 0.53 0.76 Pre P Control Means 1.88 2.96 1.43 2.17 0.64 0.95 1978 1.44 2.19 1.27 1.84 0.53 0.96 1979 1.52 2.97 1.37 2.71 0.63 0.94 1980 1.47 2.25 1.38 1.81 0.66 0.83 1981 1.39 1.92 1.33 2.30 0.60 0.87 1982 1.45 2.30 1.28 2.56 0.55 0.75 1983 1.33 1.84 1.27 1.75 0.55 0.86 1984 1.50 2.42 1.45 2.70 0.55 1.03 1985 1.86 2.56 1.27 1.83 0.54 0.74 1986 1.18 1.33 1.12 1.53 0.54 0.79 1987 1.48 2.27 1.29 2.06 0.56 0.84 1988 1.74 2.35 1.21 1.78 0.60 0.91 1989 1.48 2.23 1.29 1.88 0.56 1.09 1990 1.46 2.14 1.05 1.48 0.54 0.67 1991 1.43 2.12 1.07 1.45 0.47 0.60 1992 1.22 1.95 1.13 1.68 0.47 0.81 1993 1.42 2.44 1.07 1.54 0.49 0.86 1994 1.46 2.33 1.11 1.74 0.42 0.67 Post P Control Means 1.45 2.19 1.19 1.84 0.53 0.82 1995 1.21 1.84 0.81 1.20 0.39 0.63 1996 1.10 1.71 0.83 1.27 0.40 0.67 1997 1.03 1.34 0.71 1.17 0.36 0.51 1998 1.31 2.82 0.82 1.47 0.38 1.08 1999 1.16 1.83 0.84 1.46 0.34 0.54 2000 0.89 1.16 0.76 1.04 0.38 0.55 2001 1.06 1.63 0.66 1.30 0.34 0.58 2002 1.00 1.64 0.72 0.94 0.41 0.50 2003 0.83 1.04 0.77 1.41 0.32 0.42 2004 1.05 1.46 0.76 1.69 0.36 0.55 1995-2004 Mean. 1.06 2.82 0.77 1.69 0.37 1.08

50 B N

) 30 -1

20 (µg l (µg

a 10

50 HB C 1978-94 40 2002 2003

Chlorophyll 2004

0 Apr May Jun Jul Aug Sep Oct Nov Dec Apr May Jun Jul Aug Sep Oct Nov Dec

Figure 2. Seasonal trends in chlorophyll a in the upper, middle and lower Bay of Quinte from 2002-2004 with comparison to the long- term seasonal trend line for the post-phosphorus control, pre-zebra mussel period (1978-94).

Annual variability in Chl at the levels now observed may consequently not be reflected in light extinction. Zebra mussels can also reduce light extinction (increase transparency) by grazing directly on tripton. The lower grand mean of all post zebra mussel seasonal means compared to the grand mean of all post P control seasonal means shows the significant impact of zebra mussels on water clarity in the entire Bay of Quinte. The seasonal mean light extinction at stations B and C decreased considerably from 2002 to 2003. This trend reversed in 2004, but the reversal was not as marked at C. Comparing grand seasonal means including the 2004 data, trends in light extinction at station N reflect a higher grazing impact compared to station B. Light extinction at Hay Bay was similar in 2003 (0.77) and 2004 (0.76). As in 2003, the seasonal mean extinction coefficient at HB is lower than at station N in 2004.

Table 3. Seasonal means and maximum Secchi disc depths (m) at stations B, N, HB and C from before (1990-1994) and after zebra mussel invasion (1995-2004). B N HB C Year Mean Max Mean Max Mean Max Mean Max 1990 1.3 1.8 1.2 1.8 1.6 2.5 2.9 3.5 1991 1.3 2.0 1.4 2.5 1.6 2.4 3.2 4.5 1992 1.3 1.8 1.2 1.8 1.5 2.3 3.1 4.5 1993 1.3 2.0 1.3 2.0 1.7 2.5 3.4 4.3 1994 1.3 2.0 1.3 2.3 1.6 2.0 4.0 5.0 Mean 1990-1994 1.3 1.9 1.3 2.1 1.6 2.3 3.3 4.4 1995 1.4 2.3 1.5 2.3 1.9 3.3 3.5 5.0 1996 1.9 2.5 2.4 4.0 2.5 3.5 4.9 7.0 1997 2.1 3.0 2.5 3.3 3.0 4.3 6.1 9.0 1998 1.8 4.2 2.1 4.0 2.4 4.2 5.8 9.5 1999 1.6 3.2 1.9 2.8 2.1 2.8 5.5 7.0 2000 1.8 2.8 2.4 4.5 2.4 3.8 5.3 8.3 2001 1.9 4.0 1.8 2.5 2.6 3.8 6.0 10.0 2002 2.1 3.3 2.2 3.5 2.3 3.3 6.0 10.3 2003 1.9 2.3 1.8 3.3 2.0 3.8 6.0 8.0 2004 1.7 4.2 2.0 5.0 2.2 4.0 5.9 9.5 Mean 1995-2004 1.8 3.1 2.1 3.4 2.4 3.6 5.5 8.2

Secchi disk measurements are not as sensitive to changes in water transparency as extinction coefficients because of the non-linear linear relationship between these parameters. From 1995-2000 station N had greater Secchi disk depths compared to B (Tab. 3). This is the same pattern noted for Chl and light extinction, suggesting a greater impact at station N by

zebra mussel grazing than at B. The difference in seasonal mean Secchi disk depths between these two stations has been less evident since 2001. The lower bay Secchi disc seasonal mean average remains close at 5.9-6.0 m. The post zebra mussel grand seasonal mean average Secchi disc at C has increased 40% from the pre zebra mussel grand seasonal mean. This pattern repeats itself at all stations and is next highest at station N where a 38% increase has occurred between these two time stanzas.

Table 4. Comparison of seasonal means for the light-saturated rate of photosynthesis Popt -3 -1 B -1 -1 B (mg C m h ), the photosynthetic parameters, P m (mg C mg Chl h ) and α (mg C -1 -1 -2 -1 -1 mg Chl E m ), chlorophyll (µg l ), and light extinction- εpar (m ) before (1990-94) and after zebra mussel invasion (1995-04). Number of observations = n. B B Stations Popt Pm Chl α εpar n B Mean 1990-1994 107 4.58 23.4 4.95 1.40 13 1995 78 4.59 16.2 4.59 1.21 13 1996 49 5.13 9.1 4.51 0.83 13 1997 64 5.54 10.4 6.40 1.03 13 1998 82 4.67 16.7 4.88 1.31 13 1999 116 7.33 14.0 7.21 1.16 13 2000 72 5.64 11.4 4.25 0.89 13 2001 87 5.10 15.2 4.75 1.06 13 2002 91 6.69 12.2 5.51 1.00 13 2003 88 6.54 13.0 5.19 0.83 13 2004 80 5.78 13.7 4.73 1.05 13 Mean 1995-2004 81 5.70 13.2 5.20 1.04 13 N Mean 1990-1994 118 4.54 25.5 4.77 1.43 65 1995 66 5.01 12.8 4.73 1.00 13 1996 50 5.23 8.9 4.25 0.59 13 1997 48 5.15 8.8 4.70 0.84 13 1998 63 5.66 10.5 5.46 0.98 13 1999 52 6.69 7.40 6.48 0.94 13 2000 49 6.06 8.0 4.49 0.83 13 2001 75 5.96 11.4 4.83 0.93 13 2002 71 6.26 10.0 5.30 0.90 13 2003 69 5.59 12.7 5.33 0.82 13 2004 43 4.32 9.4 3.99 0.94 13 Mean 1995-2004 59 5.59 9.99 4.96 0.88 13 HB Mean 1990-1994 90 4.10 20.9 4.77 1.10 64 1995 46 4.31 10.5 4.68 0.81 13 1996 41 4.68 8.5 4.99 0.48 13 1997 34 4.05 7.8 5.96 0.71 13 1998 57 4.42 12.0 4.85 0.82 13 1999 59 7.27 7.8 6.86 0.84 13

Table 4 Cont’d. B B Stations Popt Pm Chl α εpar n 2000 69 5.78 9.3 4.64 0.76 13 2001 48 4.75 9.8 4.10 0.78 13 2002 67 6.14 10.1 6.02 0.72 13 2003 80 5.13 16.6 5.47 0.77 13 2004 51 5.27 9.2 4.63 0.76 13 Mean 1995-2004 55 5.18 10.2 5.22 0.74 13 GL Mean 1992-1993 48 3.81 11.8 4.61 0.69 13 1996 25 4.07 7.4 4.60 0.57 13 1997 12 2.68 4.1 4.85 0.45 13 1998 22 3.39 6.6 4.87 0.47 13 1999 22 5.49 4.2 6.90 0.46 13 2000 26 3.80 6.2 3.64 0.51 13 2001 17 3.80 4.7 3.96 0.40 13 2002 22 3.99 5.2 4.25 0.51 13 2003 23 4.05 5.9 5.34 0.44 13 2004 24 3.66 5.7 4.46 0.49 13 Mean 1996-2004 21 3.88 5.6 4.76 0.48 13 C Mean 1990-1994 25 3.78 6.3 4.62 0.48 64 1995 18 3.58 4.7 4.46 0.39 13 1996 13 3.97 3.2 4.83 0.23 13 1997 7 2.43 2.9 4.25 0.36 13 1998 11 3.68 4.0 5.17 0.38 13 1999 10 3.65 2.8 6.46 0.34 13 2000 14 3.77 3.7 4.68 0.38 13 2001 12 3.60 3.3 4.50 0.37 13 2002 13 4.10 3.2 4.78 0.41 13 2003 16 4.84 3.1 6.11 0.32 13 2004 12 3.62 3.2 5.00 0.36 13 Mean 1995-2004 13 3.72 3.4 5.02 0.35 13

Photosynthetic parameters

Areal rates of phytoplankton photosynthesis (PP) integrate the effects of light extinction, B B B biomass (Chl) and the photosynthetic parameters P m and α over the euphotic zone. P m is an index of the maximum photosynthetic capacity of the phytoplankton community because it is measured at light levels that are optimal for photosynthesis. αB measures the photosynthetic efficiency of the phytoplankton community at sub-optimal light levels and determines photosynthetic rates at deeper parts of the water column. Both parameters are normalized on biomass as denoted by the superscript B.

B Seasonal mean P m decreased at all stations except HB in 2004 compared to 2003. The difference is notable at N and C with a decrease of 23% and 25% respectively compared to B -1 -1 the previous year’s means. The seasonal mean P m at HB (5.27 mg C mg Chl h ) is a slight increase from 2003 (5.13 mg C mg Chl-1 h-1) (Tab. 4). B The post zebra mussel P m grand means at stations B, N and HB are approximately B 20% higher than the pre zebra mussel P m grand means (Fig. 3). In the 2005 report it was B stated that seasonal mean P m at B, N and HB in all years post invasion have been higher than the pre-invasion means with only two exceptions (B 1995 and HB 1997). It is B interesting to note that the 2004 seasonal mean P m at N is also an exception in being lower B B than the pre-invasion grand mean P m. The P m pre and post zebra mussel invasion grand seasonal means for both lower bay stations (GL, C) show only a minimal difference. At B both GL and C the P m seasonal means decreased in 2004 versus 2003, most notably at Conway (25%). The field year 1999 stands out as having anomalously high αB values. Pre and post zebra mussel invasion grand seasonal mean αB’s were compared excluding 1999 data. This exclusion results in little difference between upper bay stations but increases are still observed at stations HB (6%) and C (5%). Most seasonal means at all stations have tended to be between 4.25 to 5.50 mg C mg Chl-1 E-1 m-2 with the mean for the period just prior to zebra mussel invasion (1990-94) also falling in this range. In 2004 the seasonal mean αB at all stations decreased compared to the 2003 values. This contrasts with the high αB observed at stations GL (5.34) and C (6.11) in 2003. B In 2004 26% of the variance in P m is due to mixing depth temperature at Belleville. The total phosphorus and extinction coefficient explain 13% and 17% respectively of the variance in αB at B in 2004.

30 6 120

25 5 100 ) -1 20 4 ) 80 h -1 -1 ) h -3 -3 15 3 60 (mg m a

10 2 (mg C m 40 Chl (mg C mg Chl mg (mg C opt m B P P

5 1 20

0 0 0 BNHBGLC BNHBGLC BNHBGLC

6 1.50

5 1.25 ) Before -1 ) 2 After m

-1 4 1.00 E -1

3 0.75

2 0.50 (mgmg C Chl B

1 (m Coefficient Attenuation Light 0.25 Alpha

0 0.00 BNHBGLC BNHBGLC Figure 3. Grand mean of seasonal means for chlorophyll a, photosynthetic parameters, Popt and light attenuation before (1990-94) and after (1995-2004) zebra mussel colonization.

Seasonal photosynthesis

Seasonal Area phytoplankton photosynthesis rates are generated using the modelling program of Fee (1990). In 2004 the SAPP at all stations were markedly lower than in 2003. SAPP decreased 35% and 57% at B and N respectively and 38% at HB. The largest decrease occurred at GL at 59%. Conway SAPP decreased 37% from the previous year. These SAPP values in 2004 are inconsistent when compared to all years in the study (Tab. 5). There are only two years at B where the SAPP is lower than the 2004 SAPP (1996, 1998). The 2004 SAPP at N, HB and GL are the lowest recorded for the study. The 2004 data is disparate with 2003. With the exception of B in 1999, SAPP in 2003 at B and HB were the highest values recorded at these stations in both the pre and post zebra mussel time frames. SAPP at GL in 2003 was the highest value recorded in the post zebra mussel time frame at this location. B Despite increased transparency and phytoplankton adaptations such as increases in P m and αB, the mean SAPP has declined throughout the bay during the post zebra mussel invasion period compared to pre-invasion means. The addition of the 2004 SAPP values widens this gap between the pre and post zebra mussel time stanza means. The empirical irradiance SAPP was calculated with solar irradiance measured at Pt. Petre. The cloudless irradiance SAPP was calculated with theoretical cloudless irradiance by the Fee software. Cloudless calculations allowed for inter annual comparison of SAPP without the impact of annual variation in solar irradiance. The ‘% of Cloudless’ in table 5 was the percentage of empirical SAPP with respect to cloudless SAPP. In 2004 the percentage of cloudless was anomalously low at all stations. The percentage of cloudless in 2004 ranged from 67% at Belleville to 72% at station Conway. These were the lowest percentages recorded for all stations for all years of the study.

Table 5. Seasonal† areal phytoplankton photosynthesis (g C m-2) in the Bay of Quinte 1988- 2004. Location Year Empirical Irradiance Cloudless Irradiance % of Cloudless‡ Trenton-T 1991 302 381 79 1992 239 317 76 1991-1992 271 349 78 Belleville-B 1988 160 204 78 1989 238 302 79 1990 234 311 75 1991 309 385 80 1992 267 354 75 1993 197 264 75 1994 247 336 74 1989-1994 249 325 76 1995 209 268 78 1996 120 160 75 1997 211 260 81 1998 152 189 80 1999 293 362 81 2000 224 287 78 2001 219 270 81 2002 248 312 80 2003 289 386 75 2004 189 282 67 1995-2004 215 278 78 Napanee-N 1990 234 312 75 1991 283 350 81 1992 294 385 76 1993 239 325 74 1994 254 349 73 1990-1994 261 344 76 1995 210 271 78 1996 133 178 75 1997 181 225 80 1998 194 241 81 1999 183 221 83 2000 157 201 78 2001 215 266 81 2002 207 260 80 2003 283 369 77 2004 122 179 68 1995-2004 188 241 78 Hay Bay-HB 1988 153 197 78 1989 319 405 79 1990 276 369 75 1991 282 349 81 1992 303 402 76 1993 245 336 73 1994 245 328 75

Table 5. Cont’d. Location Year Empirical Irradiance Cloudless Irradiance % of Cloudless‡ 1989-1994 278 365 77 1995 186 240 78 1996 158 204 77 1997 164 201 82 1998 222 275 81 1999 242 297 81 2000 203 259 78 2001 204 255 80 2002 282 358 79 2003 359 469 77 2004 136 200 68 1995-2004 216 276 78 Glenora-GL 1992 245 324 76 1993 230 308 75 1992-93 238 316 76 1996 147 194 76 1997 107 131 82 1998 187 229 82 1999 202 244 83 2000 148 191 78 2001 155 189 82 2002 146 181 81 2003 203 265 77 2004 83 118 70 1996-2004 153 194 79 Conway-C 1988 113 144 78 1989 169 213 79 1990 157 206 76 1991 186 234 79 1992 210 277 76 1993 190 252 75 1994 173 236 73 1989-1994 181 236 76 1995 165 212 78 1996 123 161 76 1997 85 104 82 1998 167 201 83 1999 146 176 83 2000 134 170 79 2001 147 181 81 2002 124 154 81 2003 188 245 77 2004 119 165 72 1995-2004 140 177 79 †Areal photosynthesis calculated for a standardized season length for all years of May 1 to Oct 31. ‡Seasonal areal photosynthesis calculated with empirical irradiance measured at Pt. Petre as a percentage of values calculated with theoretical cloudless irradiance at latitude 44°N. 1988 values were inexplicably low and were excluded from means for the pre zebra mussel period. See Millard et al. (1996).

Discussion

The 2004 field season was marked with misadventure on the first trip. The inaugural cruise of the Leslie J proceeded without incident to arrive at the Belleville station. Upon tying off at the buoy, the crew detected the odour of smoke and lifted the engine well compartment. It was at this point that fire was discovered in the stern. The crew used a fire extinguisher to expel the flames. The fire restarted a few times but was successfully extinguished by the crew each time. The Coast Guard was alerted and an auxiliary search and rescue boat arrived. The vessel was towed to the Belleville marina and the crew were unharmed. The Leslie J was returned to Burlington for repair where it was determined that faulty wiring was the cause of the fire. The vessel Murray J was used in the interim cruises. Given the anomalous nature of the SAPP data, the results were considered suspect and examined rigorously to verify their accuracy. The advantage of the modelling approach is that it allows the user to examine parameters such as solar radiation to isolate the driver of the changes observed in 2004. The solar data in 2003 was substituted in the 2004 data and the model run for station Belleville. This resulted in a restoration of SAPP and percentage of cloudless to 2003 levels. This affirmed that the driver of these changes was the solar radiation data. A concurrent program in Hamilton Harbour was in place in 2003 and 2004. The SAPP and percentage of cloudless for this study were examined for these field years. Both the SAPP and the percentage of cloudless dropped markedly in 2004 compared to 2003. The percentage of cloudless in Hamilton Harbour in 2004 was 66%, the same level observed in the Bay of Quinte. This pattern furthered confidence that the anomalous dataset was a result of low solar irradiance in 2004. Thus the 2004 dataset was accepted as accurate. It is highly unusual for annual and inter-annual variability in seasonal areal phytoplankton photosynthesis (SAPP) to be influenced by solar irradiance to this degree. B B Typically the primary variables are Chl, εpar, P m and α (see Millard and Burley, 2004; Millard et al., 1996). In 2004 reduced solar irradiance was the dominant factor affecting phytoplankton photosynthesis. It resulted in historically low levels of SAPP and percentage

of cloudless. This lowered the post zebra mussel seasonal means for these parameters. This contrasts with the 2003 field year. In 2003 SAPP at B (except 1999) and HB were the highest means recorded in the post phosphorous control time frame of the study. The grand seasonal areal phytoplankton photosynthesis means show a marked declined throughout the bay since zebra mussel invasion. This difference between pre and post zebra mussel time periods seasonal means is augmented with the addition of the 2004 data. From 1999 to 2003 chlorophyll seasonal concentrations at N and HB consistently increased. This contrasts with the decreases observed a these stations in 2004 for chlorophyll seasonal concentrations. In 2004 at station N transparency also decreased in addition to lower chlorophyll which would further impede phytoplankton production. The decline in light extinction at Napanee in 2004 despite lowered chlorophyll levels supports the hypothesis that background light extinction may be more significant than Chl in light attenuation even in the upper bay. B Seasonal mean P m has increased in the upper and middle bay stations in the post zebra mussel time period. The mixed depth temperature means from 1994-2004 do not indicate a B temperature increase to associate with higher P m’s. This increase is coupled with the dramatic decline in chlorophyll levels at these stations post zebra mussel invasion. This suggests a link between these trends and zebra mussel colonization. The changes in photosynthetic parameters may be a compensatory response of the phytoplankton community that maximizes production under the pressures of high grazing rates by mussels. Data from previous years suggests that phytoplankton community structure B is not the sole determinant for P m values. The gradient to lower values of this parameter from upper to lower bay has always been the result of the temperature gradient. Monitoring efforts must be maintained in order to properly evaluate ecosystem effects of invasive species such as the zebra mussel on the Bay of Quinte. This system has a long history of research and monitoring programs covering water quality to fisheries. More recent study efforts include the development of an ecosystem model. All of these aspects combine to provide a continuum of data which is uniquely poised to provide great insight into the ecosystem impacts of invasive species in the Bay of Quinte and similar ecosystems throughout the Great Lakes.

References

Fee, E.J. 1990. Computer programs for calculating in situ phytoplankton photosynthesis. Can. Tech. Rep. Fish. Aquat. Sci. 1740, 1-27.

Millard, E.S., 1986. Effect of light climate, photosynthetic capacity, and respiration on gross integral primary production and net primary production potential in the Bay of Quinte, Lake Ontario. In: C.K. Minns, D.A. Hurley, K.H. Nicholls (Ed.), Project Quinte: point- source phosphorus control and ecosystem response in the Bay of Quinte, Lake Ontario. Can. Spec. Publ. Fish. Aquat. Sci. 270, 128-138.

Millard, S., Burley, M., 2004. Photosynthesis, Chlorophyll a and Light Extinction. In: Monitoring Report #13. Project Quinte Annual Report 2002, pp. 21-38. Bay of Quinte Remedial Action Plan, Kingston, Ont., Canada.

Millard, E.S., Myles, D.M., Johannsson, O.E., Ralph, K.M., 1996. Phytoplankton photosynthesis at two index stations in Lake Ontario 1987-92: assessment of the long- term response to phosphorus control. Can. J. Fish. Aquat. Sci. 53, 1092-1111.

Nicholls, K.H., Hurley, D.A., 1989. Recent changes in the phytoplankton of the Bay of Quinte: the relative importance of fish, nutrients and other factors. Can. J. Fish. Aquat. Sci. 46, 770-779.

NUTRIENTS AND PHYTOPLANKTON

K.H. Nicholls1 and E.S. Millard2

1S-15 Concession 1, RR #1, Sunderland, Ontario L0C 1H0

2Great Lakes Laboratory for Fisheries and Aquatic Sciences, Fisheries and Oceans Canada, 867 Lakeshore Road, P.O. Box 5050, Burlington, Ontario L7R 4A6

Sampling and analyses1

Thirteen samplings at each of the four main Bay of Quinte stations (Belleville = B, Napanee = N, Hay Bay = HB, and Conway = C) during the period 11 May to 20 October, 2004, were undertaken by personnel of Fisheries and Oceans Canada (DFO) according to methods and at locations previously established by Project Quinte (see e.g. Millard and Burley (2004) for a map (their Fig. 1) and other details). Samples for nutrients and chlorophyll were again analyzed by the Ontario Ministry of the Environment, except for those collected 29 June for which no data could be found. The seasonal (May-October) means reported here for nutrient and chlorophyll data consequently were based on the 12 remaining dates for which we had data. In accordance with the enhanced analysis protocol of recent years, all 13 individual phytoplankton samples were analyzed from Stations B, N and C, and for the July through September period (seven samples) for the HB location. In addition, pooled samples resulting from the quantitative combining of aliquots of each of the 13 samples were analyzed for each of the four main Bay of Quinte stations to provide

1Staff of the Laboratory Services Branch, Ontario Ministry of the Environment (MOE) in Toronto performed the chemical analyses. Sample allocation for these analyses was provided by the Technical Support Section, Eastern Region, MOE. Most phytoplankton analyses over the 32- year period of Project Quinte were performed by MOE (Lucja Heintsch and Elaine Carney); E.C. has done all such analyses for the past several years with support from Fisheries and Oceans Canada.

an additional estimate of “average” May-October biovolume (but not used in this report, except for Station HB, for which the pooled sample provided the only estimate of May to October average phytoplankton biomass2.

Results and discussion

Total phosphorus

As has been typical over the entire period of Project Quinte, spring and early summer total phosphorus concentrations throughout the bay were relatively low in 2004, but then rose to peaks at all Bay of Quinte sampling stations during the mid- to late summer (Fig. 1). Maximum concentrations of 53 µg l-1 on 24 August, 49 µg l-1 on 24 August, 44 µg l-1 on 8 September and 19 µg l-1 were recorded at Stations B, N, HB and C, respectively. Secondary peaks of 44 µg l-1 at Station B and 19 µg l-1 at Station C were found on 5 October and 22 September at Stations B (Fig. 1a) and C (Fig. 1b), respectively. At Station C, concentrations at one metre above bottom (1 mob) were consistently lower than in the surface (euphotic zone) at this location (Fig. 1b). The 1 mob 2004 average of 9.8 µg l-1 contrasts sharply with the corresponding average of 13.3 µg l-1 found in the euphotic zone at this location and likely reflects the difference in phosphorus levels in the deep eastern basin of Lake Ontario (that influence directly the bottom waters at Station C) and those of the upper Bay of Quinte which are transported down to Station C by the normal flow regime and influence the surface (euphotic zone) samples collected at Station C, even after dilution with Lake Ontario water. May to October average total phosphorus concentrations in the upper and middle Bay of Quinte continued the trend to slightly but steadily increasing values observed since 1995 (Fig. 2). Mean euphotic zone concentrations at Stations B, N, HB and C during 2004 were 33.9, 31.1, 31.2, and 13.3 µg l-1, respectively. The overall decreases in phosphorus concentrations since the 1970's are consistent with documented decreases in the lower

2The terms “biovolume” and “biomass” are used here interchangeably, assuming unit density for planktonic algal cells; i.e. 1 mm3 has a mass of 1 mg.

Great Lakes over a similar time period (Nicholls et al., 2001) and coincide with improved and expanded wastewater treatment, including enhanced phosphorus removal processes at municipal sewage treatment plants.

Figure 1. Total phosphorus concentrations in the upper (a) and middle and lower (b) Bay of Quinte during May to October of 2004. All data represent euphotic zone samples except at Stn C where “1 mob” indicates samples from one metre above bottom.

Figure 2. Long term trends in May-October mean total phosphorus concentrations at Stations B, N, HB and C in the Bay of Quinte with reference to two major interventions/ perturbations indicated by vertical dashed lines. The interim RAP objective of 30 µg l-1 is also shown. No samples were collected at Stn N between 1983 and 1988, inclusive.

Chlorophyll a

Like total phosphorus, chlorophyll a concentrations in the Bay of Quinte rose to mid- or late summer peaks from spring lows (Fig. 3). The seasonal trends in chlorophyll reflected those observed in total phytoplankton (see below). Station B revealed the highest level (23.4 µg l-1) on 20 October (but a secondary peak of 18.2 µg l-1 had occurred on 8 September. Meanwhile levels at Stn N reached a seasonal high of 11.8 µg l-1 on 24 August and declined thereafter (Fig. 3a). St Stations HB and C, seasonal peaks of 13.6 and 5.7 µg l-1 occurred on 10 August, 2004 (Fig. 3b). Over long term, trends generally reflect the patterns observed in total phosphorus (Fig. 2) and total phytoplankton (see below). After steady declines through the late 1970's to the mid-1990's, concentrations appeared to level out during the most recent 7-8 years (Fig. 4). May to October means for all stations in 2004 were either the lowest (7.6 µg l-1 at Station N), the second-lowest (6.7 µg l-1 at Station HB and 2.7 µg l-1 at Station C) or the third- lowest (10.9 µg l-1 at Station B) on record for each of these respective sampling locations. The previously low chlorophyll year was 1996 - a probable impact of intensive filtration of chlorophyll-bearing planktonic algae by recently colonized zebra mussels.

Total nitrogen-to-total phosphorus ratios and nitrate trends

It has been instructive in the past, at 3-year intervals of Project Quinte reporting, to present 3-year averages of the total N-to-total P ratios for the upper bay, mainly with reference to changing nutrient resources for phytoplankton and the potential for shifts in composition related to nutrient availability (Nicholls et al., 2002). N-P ratios in the upper bay (Station B), middle bay (Station HB) and lower bay (Station C) increased dramatically from the mid-1970's to the mid-1990's, but has levelled off in recent years (Fig. 5). At least some of the reason for these long term changes relates to increases in nitrate levels in the bay, especially at Station C (Fig. 6), where concentrations have more than doubled in the past three decades. These increases are attributable to two main sources: (1) increased supply from land drainage and atmospheric deposition, and (2) decreased utilization by

phytoplankton as a result of phytoplankton biomass reductions induced by controls on the supply of phosphorus, the growth-limiting nutrient, and more recently, “grazing” (filter feeding) of phytoplankton by zebra mussels.

Figure 3. Euphotic zone total (uncorrected for phaeopigments) chlorophyll a concentrations at the upper (a) and middle and lower (b) Bay of Quinte sampling stations during May to October of 2004.

Figure 4. May to October mean total chlorophyll a concentrations (uncorrected for phaeopigments) at Bay of Quinte sampling stations, 1972-2004. No data are available for 1977 or for Station N between 1983 and 1988, inclusive.

Figure 5. Total nitrogen-to-total phosphorus ratios determined as May to October means of the ratio calculated for each sampling date ([nitrate-N + nitrite-N + total Kjeldahl- N]/total P) and further reduced here to three-year means for each of Stations B, HB and C, representing the upper, middle and lower Bay of Quinte, respectively.

Figure 6. May to October average nitrate-N concentrations at Station C, 1972-2004.

Figure 7. May-October average total phytoplankton biovolume (or biomass, assuming unit density for planktonic algal cells) for all years of Project Quinte, 1972-2004 for Stations B, N, HB and C. Note that Station N was not sampled during the period 1983-1988.

Figure 8. May to October average percentages of total phytoplankton biovolume contributed by the major classes of algae during the five-year period 2000-2004.

Phytoplankton

Based on statistical time-trend analyses, Nicholls et al. (2002) identified five categories of phytoplankton response in the Bay of Quinte to point-source phosphorus loading reductions in 1978 and to the establishment of zebra mussels in 1995. The elements of the phytoplankton community that changed after P reduction, but not after zebra mussel establishment (category 1) included: total phytoplankton, total diatoms, total cryptophytes, and total non-N-fixing blue-green algae. No similar intensive analysis of change was done on more recent data, but an overview May-October average phytoplankton composition and biomass during the most recent 5-6 years reveals very little change in these “category 1" components. May to October average total phytoplankton biovolumes for Stations B, N, HB and C of 3.80, 2.28, 1.85 and 0.56 mm3 l-1, respectively, in 2004 were consistent with similar values recorded since 1999 (Fig. 7). When compared with the much earlier post- phosphorus removal period however, the upper and middle Bay values in recent years of about 3 mm3 l-1 have averaged less than 50% of the biomass values recorded during the 1980's and early 1990's (Fig. 7). The domination of the phytoplankton community by diatoms (Bacillariophyceae) with 69.8%, 53%, 61.5%, and 44.8 % of the 2004 total at Stations B, N, HB, and C, respectively (Fig. 8), has been a continuing feature of the Bay of Quinte. Other important classes of algae in 2004 included the blue-greens (Cyanophyceae or Cyanobacteria) with 15.9% and 16.2% of the total at Stations N and HB, and the Cryptophyceae, with 14.1%, 11.1% and 26% of the total at Stations N, HB and C, respectively (Fig. 8). The seasonal distribution of total biomass in 2004 was different from that recorded in 2003, however. For example, values at Station B climbed steadily during late summer and fall to a peak of 11.0 mm3 l-1 in late October (Fig. 9). This contrasts sharply with the trend in 2003 at this location when biomass decreased steadily from a seasonal peak of 6.4 mm3 l-1 in early September to a late October value of only 2.4 mm3 l-1. Again, as has been typical over the entire three decades of Project Quinte, total phytoplankton values at Station C were consistently lower than at any of the other sampling locations (Fig. 9).

The October peak in total phytoplankton biomass mentioned above (11 mm3 l-1) was contributed almost entirely (10.5 mm3 l-1) by the class Bacillariophyceae (Fig. 10a) of which 9.8 mm3 l-1 was contributed by the genus Aulacoseira thus perpetuating the propensity, as observed in previous years, of this taxon to dominate the upper Bay phytoplankton. At Station HB in the third week of September (no samples from this location analyzed after this date), Aulacoseira biovolume was 3.9 mm3 l-1, or 91% of total diatom and 71% of total phytoplankton biovolume. Similarly, at Station N, this taxon had recorded biovolumes of 3.57 and 2.0 mm3 l-1 on 6 October and 21 October, respectively, representing 88% and 80% of total phytoplankton biomass. Cyanophyceae (blue-green algae) were virtually absent from the Bay of Quinte during May and June of 2004 (Fig. 10b), but quickly reached seasonal highs at Stations N and HB in July and August and at Station B in early September. The most important taxa at the upper and middle Bay stations included Microcystis, Anabaena and Chroococcus spp. The Cryptophyceae and Chlorophyceae exhibited very different seasonal distributions in 2004 with the cryptomonads achieving an early summer peak and declining thereafter (Fig. 10c), while the planktonic green algae (Chlorophyceae) increased steadily over the summer reaching their peak in early autumn (Fig. 10d). An exception may have been the lower Bay cryptomonads where both late spring and a mid-summer peaks were evident (Fig. 10c). In each case, the same genus (Cryptomonas) was responsible for most of the total class contribution. The Bay of Quinte Remedial Action Plan has included among its water quality and ecosystem rehabilitation objectives a stated desired reduction in total phytoplankton biomass as follows: “Reduce the average algal density from an existing level of between 7 and 8 to a new range between 4-5 mm3 l-1” (Bay of Quinte RAP, 1993). While the May- October average biomass in the upper and middle sections of the Bay have been consistently below the stated target range during the most recent years (Fig. 7), the establishment of zebra mussels in the Bay in 1995 may still represent a confounding factor in interpreting long term trends (Nicholls et al., 2002). The presence of zebra mussels in the Bay should perhaps be considered as a permanent element of the Bay’s biological community. While total phytoplankton biomass has generally been at acceptable levels

recently, there has apparently not been a major shift to the more desirable composition set out in the community target proposed by Nicholls et al. (2004). More specifically, the most important changes required in the composition of the upper Bay of Quinte phytoplankton to bring it more in line with the proposed “target” composition proposed by Nicholls et al. (2004) include large decreases in biomass of Aulacoseira and Microcystis - two dominant taxa in the 2004 samples. Until this happens, the phytoplankton community might be considered to be still very much a degraded, and “substandard” community, despite the achievement of generally acceptable total biomass values.

Figure 9. Seasonal changes in total phytoplankton biovolume, May to October of 2004 at the four main sampling stations in the Bay of Quinte.

a b

c d

Figure 10. Changes in the biovolume of the four major phytoplankton classes, (a) Bacillariophyceae (diatoms), (b) Cyanophyceae (blue-green algae or Cyanobacteria), (c) Cryptophyceae (cryptomonads), and (d) Chlorophyceae (green algae) during May to October of 2004. Samples from Station HB were analyzed only for the July-September period.

References

Bay of Quinte RAP, 1993. Time to Act. The Bay of Quinte Remedial Action Plan Stage 2 Report. Bay of Quinte RAP Coordinating Committee, September, 1993. Ontario Ministry of the Environment, Kingston, Ontario.

Millard, E.S., Burley, M., 2004. Photosynthesis, chlorophyll a, and light extinction. In: Project Quinte Annual Report for 2002, pp. 21-38. Bay of Quinte Restoration Council/Project Quinte. Bay of Quinte Remedial Action Plan. Kingston, Ontario.

Nicholls, K.H., Hopkins, G.J., Standke, S.J., Nakamoto, L., 2001. Trends in total phosphorus in Canadian near-shore waters of the Laurentian Great Lakes: 1976-1999. J. Great Lakes Res. 27, 402-422.

Nicholls, K H., Heintsch, L., Carney, E.C., 2002. Univariate step-trend and multivariate assessments of the apparent relative effects of P loading reductions and zebra mussels on the phytoplankton of the Bay of Quinte, Lake Ontario. J. Great Lakes Res. 28, 15- 31.

Nicholls, K H., Heintsch, L., Carney, E.C., 2004. A multivariate approach for evaluating progress towards phytoplankton community restoration targets: examples from eutrophication and acidification case histories. Aquatic Ecosystem Health & Management 7, 15-30.

ZOOPLANKTON IN THE BAY OF QUINTE

K.L. Bowen and O.E. Johannsson

Great Lakes Laboratory for Fisheries and Aquatic Sciences Department of Fisheries and Oceans 867 Lakeshore Road, P.O. Box 5050, Burlington, Ontario, L7R 4A6

Introduction

The zooplankton community in the Bay of Quinte is dynamic. It changes in structure down the length of the Bay from the shallow, eutrophic habitat in the upper bay, represented by the station at Belleville (B) and Napanee (N), through mesotrophic conditions in the section near Hay Bay (HB), to a meso-oligotrophic environment in the mouth of the bay at Conway (C). Station depth increases from 5 m at B and N, to 11 m at HB and 32 m at C. This means that thermal stratification occurs throughout the summer at C, sporadically at HB and not at B or N. Zooplankton community structure and productivity is controlled not only by the physical environment but also by the type and quantity of available food. This is determined by nutrient conditions, the structure of the fish community which is the primary source of mortality in this system, and the introduction of new species which either compete for food or alter predation on or within the zooplankton community. Monitoring of zooplankton in the Bay of Quinte tracks species composition, size structure, abundance, biomass and productivity in an effort to understand the response of the community to changes in the controlling factors and assess its ‘health’ as a component of the ecosystem. This report presents data from the 2004 field year in the context of previous data.

Methods

Zooplankton samples were collected at the three primary monitoring stations, B, HB, and C. Zooplankton were also collected at Napanee (N), a 5 m deep station located at the eastern end of the upper bay. Samples were collected on alternate weeks from the beginning of May until the end of October. As in previous years, discrete samples were collected through the water column with a 41-l Schindler-Patalas trap fitted with 64-µm mesh. Three depths were sampled at B (1 m, 2 m and 3 m), three at N (1 m, 2 m and 3 m), five at HB (1 m, 2 m, 3 m, 6 m and 9 m), and eight at C (1 m, 3 m, 5 m, 8 m, 10 m, 15 m, 20 m and 25 m). A single composite sample was constructed for each station-date by combining 50% of the sample from each depth. A minimum of 400 individual zooplankters and all loose eggs within a subsample aliquot were enumerated from each sample. The exotic predatory cladoceran Cercopagis pengoi, which was first seen in eastern Lake Ontario in 1998 (MacIsaac et al., 1999), was found at Belleville and Conway in 1999, and at all three stations in 2000. Each of the discrete depth samples were filtered through a 400-µm mesh, and all of the C. pengoi were enumerated. Cercopagis was not subsampled because individuals have a tendency to form clumps within the samples when their caudal spines become entangled. Veliger larvae were not counted in the zooplankton samples until 1995. Dreissenids were first found in the upper bay in 1995 and therefore veligers were not likely present in the plankton until that year. However dreissenids were abundant in the eastern basin of Lake Ontario in 1991, and veligers may have occasionally been present in the plankton at Conway before 1995. Seasonally-weighted mean (SWM) abundance, biomass and production were calculated over the May 1 to Oct 31 sampling season. Size was not originally measured. A later project examining the biomass size spectra of zooplankton communities provided the opportunity to go back and estimate the size structure of all the Quinte samples over the June 1 to Oct 6 period. Therefore, present and future estimates of cladoceran mean length were (or will be) calculated over this period. Biomass was calculated by multiplying the abundance estimates by mean species dry weights determined by Watson and Wilson for the Great Lakes, as has been done historically. Production was estimated by the egg-ratio

method of Paloheimo (1974) as described in Cooley et al. (1986). When a species’ seasonally- weighted, mean biomass was <50 mg m-2, production was estimated using P/B relationships as described in Johannsson et al. (2000). However in several cases, the egg- ratio production estimate was zero or close to zero, so the P/B production estimate was used despite a SWM biomass >50 mg m-2. It appears that P/B values estimated from the literature relationships overestimate production in the Bay of Quinte. Therefore, the Quinte P/B production estimates were corrected based on comparisons of P/B and egg ratio production values from Bay of Quinte between 1995 and 1999. These equations were updated in 2005 (Bowen and Johannsson, 2005).

Results

Belleville

With the exception of 2001 when seasonally-weighted mean (SWM) biomass reached over 500 µg l-1, biomass in the upper bay during the 1992-2004 period was lower than that observed in the 1980s (Fig. 1). SWM biomass was very similar in 2003 and 2004, averaging 220 and 218 µg l-1, respectively. There have been sporadic years of exceptionally low biomass: 1992, 1997, and 2000. Both 1992 and 2000 were cold during the summer months, which may help explain the low values (Fig. 2). Since the early 1990s, Belleville has been dominated by small and medium cladocerans (Bosmina, Eubosmina and Ceriodaphnia) and fewer copepods (Tab. 1, 2). In recent years, dominance has alternated between Bosmina (2000, 2002 and 2004) and the slightly larger Eubosmina (2001 and 2003). Eubosmina also dominated during most of the 1980s and 1990s. Abundance of the large cladoceran Daphnia galeata mendotae has also been erratic in recent years. Except for a thirty-year peak in 2001, its numbers have generally remained low since 2000. This followed a period of fairly stable, higher abundances in the 1990s. Another large cladoceran, the predatory Leptodora kindtii, has remained uncommon in the last decade. Leptodora is generally rare but an important species to follow because it is a large predatory cladoceran vulnerable to fish predation. Cercopagis pengoi is rare in the

upper bay and none have been observed at Belleville since 2001. This year-to-year change in the dominant cladoceran taxa has led to wide fluctuations in mean cladoceran length over the last five years (Fig. 3). Cladocera averaged only 0.33 mm at Belleville in 2000 and 2004, which is well below the values observed in the 1990s. These low values were largely because of the abundance of small Bosmina. Zooplankton biomass in the Bay of Quinte shows strong seasonal patterns, which are influenced by water temperature, phytoplankton abundance and fish predation. Biomass was very low at Belleville throughout the spring of 2004, and peaked dramatically in mid- June due to a sudden increase in Bosmina numbers (Fig. 4). Smaller Bosmina peaks again occurred in mid-August and early October. From late June to mid-July, Eubosmina, D. galeata mendotae, Daphnia retrocurva and Ceriodaphnia also contributed to zooplankton biomass. Veliger larvae peaked from early to mid-August, which was about a month later than the 2003 peak. Reasons for the zooplankton biomass crash in late July and subsequent recovery are unclear. Peak biomass in 2004 was much lower than in 2003 (248 vs. 663 µg l-1). Total SWM zooplankton production has remained very stable at Belleville between 2002 and 2004, averaging from 4551 to 4865 µg l-1 (Fig. 1). Cladocerans remained the dominant group of zooplankton in the upper bay, contributing 81.3% of SWM biomass and 88.6% of total production (94.5% of crustacean production; Tab. 3, 4). Of this total production, predatory cladocerans accounted for only 0.1%, whereas Dreissena veligers contributed 6.2%. Although the total contribution by cladocerans has remained similar over the last few years, the taxa that contributed most have fluctuated. In 2001 and 2003, D. galeata mendotae was the dominant contributor (50% and 36%, respectively). In alternate years, Bosmina and Ceriodaphnia dominated total production (62% in 2002 and 65% in 2004).

600 Belleville Veligers Calanoids 500 Cyclopoids Cladocerans )

-1 400

300

200 Biomass (ug.L

100

0 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003

14000 Belleville

12000

) 10000 -1

8000

6000

Production (ug.L Production 4000

2000

1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 Figure 1. Trends in volumetric seasonally-weighted mean dry biomass and seasonal production at Belleville.

Table 1. Seasonal mean abundances of major zooplankton species at Station B (nos m-3), 1975-2004. Species 1975 1976 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 Bos 5006 61713 2469 3666 26373 32848 125122 90673 26764 19829 26724 22793 11299 19315 Eubos 48732 29155 49471 51067 41015 103837 59667 58274 90417 72121 70351 47818 60133 60050 D. ret 12632 10742 22971 10350 13535 28695 32717 27919 17335 22262 17866 3678 8778 6312 D.g,m, 285 14 416 3471 646 4927 1866 838 7006 2162 6115 11168 12208 13185 Cerio 490 5750 844 648 2355 3555 6575 13288 10706 11363 9424 1242 645 2573 Chy 41206 54664 43466 20367 11670 36805 46441 64978 56434 22974 51566 50943 28395 46525 Diaph 3270 2122 2268 2317 1243 1160 4262 1929 630 1095 5449 917 1295 572 Lepto 351 381 157 113 51 169 215 74 0 80 274 0 44 41 Diac 24 140 83 57 157 59 63 232 209 174 180 301 111 510 Vern 1518 1134 396 541 468 1142 1546 753 6988 1512 1463 2176 4192 2292 Tropo 55 1710 1885 592 153 275 2011 7856 1951 415 5175 143 553 289 Meso 185 241 582 688 190 1387 1262 1155 1767 705 1644 1338 1945 5161 Species 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 Bos 7087 10966 8418 10167 15506 22353 13707 15058 7682 24624 25704 40343 10135 57811 Eubos 60005 16718 47555 46325 13769 21018 22321 44748 31078 1677 42439 7547 28793 5451 D. ret 9967 4949 7478 13440 2168 11934 6252 21172 8839 1583 5518 4795 3627 3458 D.g,m, 5961 4575 5347 6417 8943 8961 5607 4340 4885 178 32192 689 6601 1779 Cerio 2098 1014 1731 457 1925 2218 818 1911 2699 6632 2210 16884 6291 6 Chy 52620 10602 19716 52786 9150 15003 715 13861 15980 705 6559 2490 5050 3217 Diaph 1776 352 1976 3648 913 1526 2384 3456 2070 433 2317 981 1553 99 Lepto 103 0 82 146 24 34 112 8 93 6 71 39 48 23 Cerco 11 18 1 0 0 0 Diac 165 101 855 144 255 142 82 16 4 19 36 32 26 55 Vern 971 52 147 85 425 963 753 334 217 51 396 81 97 0 Tropo 1757 285 423 57 222 299 522 780 2004 403 2765 653 2269 960 Meso 1557 215 277 801 421 808 909 2921 1727 56 2386 200 601 141 Veligers 22631 45536 23729 15752 20683 33313 5844 34518 22110 19351 *Other species seen in 1997-2004: Alona sp., Ceriodaphnia quadrata, Holopedium gibberum, Polyphemus pediculus, Pleuroxus sp., Scaphaloberis kingi, Eucyclops agilis, Eurytemora affini, Leptodiaptomus sicilis, L. ashlandi, L. siciloides and Skistodiaptomus oregonensis. Copepod juveniles (nauplii and copepodids) are not included in the above totals.

Table 2. Abbreviations of the species names in the abundance tables. Scientific Name Abbreviated Name Bosmina longirostris Bos Eubosmina coregoni Eubos Daphnia retrocurva D. ret Daphnia galeata mendotae D.g,m, Ceriodaphnia lacustris Cerio Chydorus sphaericus Chy Diaphanosoma birgei Diaph Leptodora kindtii Lepto Cercopagis pengoi Cerco Diacyclops thomasi Diac Cyclops vernalis Vern Tropocyclops extensus Tropo Mesocyclops edax Meso Dreissena veligers Veligers

Table 3. Distribution of zooplankton SWM biomass (May 1 - Oct. 31 2004) amongst the taxonomic groups in the Bay of Quinte at the four monitoring stations. Taxon Feeding Percentage of Total Percentage of SWM Biomass Guild SWM Biomass without Veliger Larvae C HB N B C HB N B Cladocerans 72.1 76.1 68.5 81.3 77.0 83.2 84.3 87.5 Herbivorous 68.9 76.1 68.2 81.1 73.5 83.2 84.0 87.3 Carnivorous 3.2 0.0 0.3 .2 3.4 0.0 0.4 0.2 Cyclopoids 18.1 13.3 10.1 9.6 19.3 14.6 12.4 10.4 Calanoids 3.4 2 2.6 2.0 3.7 2.2 3.2 2.1 Veligers 6.3 8.6 18.7 7.1

SWM Biomass (µg l-1) 69.3 342.3 210.6 218.1 64.9 312.9 171.2 202.6

Table 4. Distribution of zooplankton SWM production (May 1 - Oct. 31 2004) amongst the taxonomic groups in the Bay of Quinte at the four monitoring stations. Taxon Feeding Percentage of Total Percentage of Production Guild Production without Veliger Larvae C HB N B C HB N B Cladocerans 77.9 80.8 80.6 88.6 87.4 91.6 80.8 94.5 Herbivorous 69.3 80.8 80.3 88.5 77.8 91.5 80.5 94.4 Carnivorous 8.6 0.1 0.3 0.1 9.6 0.1 0.3 0.1 Cyclopoids 9.9 6.7 4.1 4.5 11.2 7.6 4.1 4.8 Calanoids 1.3 0.7 0.7 0.6 1.4 0.8 0.7 0.7 Veligers 10.9 11.7 14.6 6.2

Production (µg l-1) 732 4570 4935 4572 721 4558 4921 4566

Belleville 24 June - September

20 Pinatubo eruption Mean Temperature (C) Mean Temperature

18 1975 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003

Figure 2. June to September mean whole water column temperature at Belleville, 1975 to 2004.

Mean Cladoceran Length 0.6

0.55

0.5

0.45

0.4

0.35 Body Length (mm) Body Length 0.3 Belleville Hay Bay 0.25 Conway 0.2 1976 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004

Figure 3. Trends in the June 1 – Oct. 6 time-weighted mean length of Cladocera at the three Bay of Quinte stations: Belleville, Hay Bay and Conway

500 Belleville Veligers 2004 Calanoids Hay Bay 400 Cyclopoids 2004 Other Herb. Clad.

) Bosminids -1 300 Daphnia g.L μ

200 Biomass ( Biomass

100

0 500 Napanee 2004 400 Conway 2004 ) -1

g.L 300 μ

200 Biomass ( Biomass

100

0 11- 18- 01- 15- 29- 13- 29- 11- 24- 08- 21- 05- 20- May May Jun Jun Jun Jul Jul Aug Aug Sep Sep Oct Oct 13-Jul 29-Jul 01-Jun 15-Jun 29-Jun 05-Oct 20-Oct 11-Aug 24-Aug 08-Sep 21-Sep 11-May 18-May Figure 4. May to October 2004 seasonal trends in volumetric dry biomass (based on measured weights) at Belleville, Napanee, Hay Bay and Conway in the Bay of Quinte. Bosminids include Eubosmina and Bosmina, and Daphnia includes D. galeata mendotae and D. retrocurva. “Other Herb. Clad.” represents the remaining herbivorous cladocerans, particularly Ceriodaphnia.

Overall, cyclopoid copepods provided only 4.5% of total production at Belleville in 2004. This group has contributed on average 6.8% of the total since dreissenids invaded in 1995. In comparison, cyclopoids represented 14.5% of production in the decade prior to invasion. They were also a less important component between 1975 and 1983 (4.3%). Percent cyclopoids, in terms of biomass, has also declined since the 1980s (Fig. 5). This decline has been particularly noteworthy in both Diacyclops thomasi and Acanthocyclops vernalis (Fig. 6). D. thomasi, which is the most common copepod in the Great Lakes (Hudson et al., 1998), tends to reproduce mainly in the spring, with a second egg-producing period in late summer. Egg-ratio production (based on actual egg counts in the samples) of this species peaked in May in the 1980s, but shifted later in the summer in the 1990s. Inexplicably, it has fallen to virtually zero since 1997. Similarly, the egg-ratio production estimates of both A. vernalis and Mesocyclops edax fell markedly in the mid- to late1980s, and have also dropped to zero since 2000. Both are generally considered to be summer- reproducing warm water taxa. Tropocyclops, now the dominant copepod at Belleville, also reproduces rapidly throughout the summer. It is among the smallest cyclopoids in the Great Lakes (Hudson et al., 1998). Tropocyclops is the only cyclopoid with relatively unchanged (but variable) biomass and production estimates over the course of the study.

Percent Cyclopoids Dreissena invasion 60 by Biomass

50 Cercopagis introduction 40

30 Percent 20

0 1975 1978 1981 1984 1987 1990 1993 1996 1999 2002

Belleville Hay Bay Conway

Fig. 5. Percent cyclopoid copepods by biomass in the Bay of Quinte, 1975 to 2004.

0.8 5.0 Diacyclops - Belleville 0.7 Tropocyclops

) 4.0 -1 0.6 1970s g.L

μ 0.5 1980s 3.0 0.4 1990s 0.3 2000s 2.0 Mean Daily 0.2

Production ( 1.0 0.1 0.0 0.0

14 8 12 Cyclops Mesocyclops ) 7 -1 vernalis 10 6 g.L μ 8 5 4 6

Mean Daily Daily Mean 3 4 2 Production ( Production 2 1 0 0 May June July Aug Sep Oct May June July Aug Sep Oct

Figure 6. Mean daily production estimates of four cyclopoid taxa from May to October at Belleville, for four time periods (1970s to 2000s).

Hay Bay

SWM biomass at Hay Bay has slowly climbed since 2001, and in 2004 reached the highest level since 1996 (342 µg l-1; Fig. 7). Increases in both cladocerans and veligers were largely responsible for this upward trend. Despite this gradual climb, biomass during the last decade has remained substantially lower than that of the peak years of the 1980s. On average, veligers were the most abundant species at Hay Bay in 2004 (Tab. 5). However, given their small size, they represented only 7.1% of biomass. Cladocerans were the most dominant group, representing 81.3% of biomass. D. galeata mendotae was the most dominant taxon in terms of biomass, followed by Ceriodaphnia, D. retrocurva, Eubosmina and Bosmina. Never abundant, Cercopagis peaked in 2002 at 84 individuals m-3 and declined to only 6 individuals m-3 in 2004. SWM cyclopoid biomass has remained consistently low since 1996, averaging about 42 µg l-1. In comparison, values between 1982 and 1992 were generally between 100 and 200

μg l-1. Tropocyclops and M. edax were the most common cyclopoids in 2004. As at Belleville, calanoid densities and biomass were about an order of magnitude lower than those of cyclopoids, and this group comprised a very minor part of the zooplankton community. Zooplankton biomass at Hay Bay increased gradually throughout the spring of 2004 (Fig. 4). D. galeata mendotae populations began climbing in mid-July, peaked in late July, and reached a second, smaller peak in mid-August. Most of the fall Daphnia were D. retrocurva. However, as with Belleville, the 2004 maximum was much smaller than in 2003. Bosminds peaked several times throughout the season, and were an important component in the fall. Veligers were most abundant in mid-June, whereas cyclopoids were most common in the early summer and early fall. Zooplankton production in 2004 increased somewhat relative to 2003 (4570 and 4009 µg l-1, respectively; Fig. 7). Relative to other years in the post-Dreissena period, production was at an intermediate level in 2004. However, it was substantially lower than the peak years of the 1980s and early 1990s. Overall, cladocerans contributed the most to total production in 2004 – 80.8%, or 91.6% of crustacean production (Tab. 4). Ceriodaphnia was the dominant zooplankton producer in 2004 (22.4% of the total). Although production and biomass of D. galeata mendotae was lower in 2004 compared to 2003, it still represented 20.0% of the total. Veligers contributed 11.7% - a substantial drop compared to the record high of 22.9% in 2003. Cyclopoids contributed 6.7%, and calanoids only 0.7% of the total. The June 1 – Oct 6 cladoceran mean length was 0.432 mm, which was similar to other mean values at Hay Bay since 1999 (Fig. 3). Mean size during this period was substantially lower than the peak observed in the mid- to late1990s.

Conway

SWM zooplankton biomass at Conway in 2004 was 69.3 µg l-1 (Fig. 8). In comparison, 2003 SWM biomass was 108.7 µg l-1 - the highest value since 1996. This 2003 high was primarily due to the abundance of the large cladoceran D. galeata mendotae and veligers.

Table 5. Seasonal mean abundances of major zooplankton species at Station HB (no.s m-3), 1975-2004. Species 1975 1976 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 Bos 40091 47891 21799 18826 145948 62053 70523 78658 48005 71423 45936 53867 29631 33781 Eubos 16387 29344 25494 27008 16870 35561 63739 55498 46275 24668 54345 27621 57740 48805 D. ret 5577 6383 8024 8435 7153 10060 29490 21469 22540 10536 10248 17719 17969 10256 D.g,m, 36 29 126 1752 392 1907 2389 1355 1166 2599 5775 2002 8255 6472 Cerio 2570 2766 2983 604 2528 7504 30520 16492 12978 15887 35395 26623 8758 4267 Chy 19785 29725 41747 27538 29264 50062 85786 55082 28605 34637 65071 21743 39362 32219 Diaph 1234 604 564 729 470 476 2958 755 355 1122 812 2888 541 223 Lepto 162 195 146 149 61 83 58 20 24 30 0 0 0 4 Diac 95 512 701 417 577 869 1107 939 448 657 1537 1144 1068 544 Vern 342 321 262 337 420 307 576 3663 1213 964 680 1017 1625 804 Tropo 527 9460 4883 1992 1167 2799 6925 14208 9765 2136 4613 1953 5520 1403 Meso 74 107 340 423 281 328 1517 2745 1276 1722 1017 3140 2614 899 Species 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 Bos 319382 9201 21360 14455 25769 16612 19928 18516 42175 35695 30968 36972 8262 26122 Eubos 24418 16259 19636 24755 13983 11332 14652 15544 28314 14698 8819 19341 15917 19823 D. ret 13301 8021 7913 13951 5250 7873 10285 10023 16206 9244 7600 9851 8040 9532 D.g,m, 3844 9747 5601 7603 5689 16256 7117 10511 3786 6041 3008 4723 12097 9779 Cerio 12769 2229 6348 2769 6183 3710 3471 4784 15741 14970 7181 11861 1790 11198 Chy 20091 17350 17665 23001 14830 28372 6258 16861 10692 4537 5866 3950 35922 9968 Diaph 596 24 632 1082 152 615 1158 1121 1200 431 590 1418 1033 997 Lepto 0 0 34 0 86 30 13 10 5 0 9 25 46 5 Cerco 0 8 27 1 0 6 Diac 782 145 523 517 740 406 979 745 141 13 138 84 561 222 Vern 282 52 20 391 98 540 735 246 46 207 96 90 69 0 Tropo 2931 56 352 436 479 844 976 860 4348 1849 2474 1306 2141 4159 Meso 2988 14 569 663 1150 978 1965 2291 1756 831 576 634 1427 1159 Veligers 12741 19389 14383 14392 10668 50068 5393 35104 55304 36655 Other species seen in 1997-2004: Alona sp., Alona affinis, Chydorus piger, Cercopagis pengoi, Diaphanosoma brachyura, Holopedium gibberum, Polyphemus pediculus, Sida crystallina, Eurytemora affini, Leptodiaptomus sicilis, L. siciloides, Limnocalanus sp., Skistodiaptomus oregonensis

800 Hay Bay Veligers 700 Calanoids Cyclopoids 600 Cladocerans ) -1 500

400

300 Biomass (ug.L Biomass 200

100

0 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003

12000 Hay Bay

10000 ) -1 8000

6000

4000 Production (ug.L Production

2000

0 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003

Figure 7. Trends in volumetric seasonally-weighted mean biomass and seasonal production at Hay Bay.

Despite a decrease relative to its 2003 peak of 3522 individuals m-3, D. galeata mendotae averaged 1475 individuals m-3 in 2004 (Tab. 6). The only other years when this taxon exceeded 1000 individuals m-3 were 2002, 1996 and 1994. D. retrocurva and veligers also decreased in 2004, but numbers of Bosmina, Eubosmina and Holopedium remained fairly similar. Cercopagis abundance peaked in 2002 and 2003 (207 and 202 individuals m-3, respectively), and declined to only 76 individuals m-3 in 2004. Although total biomass in 2004 was higher than the lowest period between 1997 and 2001, it was still substantially lower than during the 1980s and early 1990s when cyclopoids were abundant. In 2003, both calanoid and cyclopoid mean biomasses remained low, and as a percent of total biomass, cyclopoids continued to be much less dominant than they were prior to Dreissena invasion (Fig. 5). Between 1980 and 1994, cyclopoid SWM biomass averaged 42.3% of the total community, compared to only 21.7% since 1996. Cyclopoid biomass reached its lowest value in 2000 (5.2 µg l-1) and has since recovered somewhat in 2003 and 2004. Zooplankton production rates at Conway were much lower in 2004 than in 2002 and 2003 (Fig. 8). Production in 1996, 2002 and 2003 were unusually high for the post- dreissenid period. Cladocerans comprised 77.9% of total production at Conway in 2004, of which D. galeata mendotae, Eubosmina and Holopedium were the largest contributors (13.5%, 13.2% and 12.4%, respectively). Veliger production in 2004 was only 80.2 µg l-1; the lowest level since 1995 when dreissenids were first recorded in the bay. Cyclopoid production increased from only 2.7% in 2002 to 9.9% in 2004 (Tab. 4). Production of Diacyclops, the dominant cyclopoid at Conway, generally peaks in August and September (Fig. 9). However, since 1999, production estimates of this taxon using the egg-ratio method have fallen dramatically. Tropocyclops, typically another common cyclopoid at Conway, has also exhibited a very low egg-ratio production most years since 1992 (Fig. 9). Production of both taxa also appears to have shifted to slightly later in the season. Similarly, Mesocyclops virtually disappeared between 1995 and 2001. However, it may still have been represented by the unidentified juvenile stages. Mesocyclops’ egg-ratio production estimates were also very low or zero during this period, although it recovered in 2003 and 2004. In fact, the 2003 estimate was about an order of magnitude higher than any

Table 6. Seasonal mean abundances of major zooplankton species at Station C (nos m-3), 1975-2004. Species 1975 1976 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 Bos 13734 14104 14604 15233 30829 78046 41902 41881 13800 33413 28724 23815 28174 18211 Eubos 4608 2242 7633 5885 3844 2290 8810 4987 8266 8440 7726 4861 10712 18301 D. ret 2133 1044 3275 3174 2862 3691 8239 4154 6654 5362 7927 4212 6481 5927 D.g,m, 59 2 235 75 17 25 182 138 164 397 219 38 406 566 Cerio 1665 656 930 757 1726 3136 5451 2298 836 9030 6397 2736 3788 6488 Chy 2365 2941 9616 3135 4355 2304 3628 1574 1751 2420 1537 1693 3999 2543 Diaph 229 40 313 287 189 79 859 192 74 249 217 76 213 168 Lepto 53 32 46 30 15 21 27 14 15 3 31 44 40 20 Diac 1258 1551 4168 2519 3014 3339 2710 2067 1414 2926 2646 1275 2319 1147 Vern 120 84 62 69 71 36 76 75 48 117 46 59 86 98 Tropo 12 1437 1541 956 1156 1051 7054 2807 3353 2103 5020 1683 2058 748 Meso 102 143 133 184 383 112 247 344 236 184 147 178 510 142 Species 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 Bos 45307 6359 6814 12892 5382 11823 6841 10102 5486 4007 4669 10293 5014 5453 Eubos 4452 5164 3219 5062 2403 1485 1288 3642 4913 1390 982 5625 3588 3041 D. ret 4132 1930 1931 3287 3149 4674 2890 5608 4078 839 2538 2773 3562 1915 D.g,m, 145 460 721 1227 619 2552 556 554 211 343 310 1266 3522 1475 Cerio 2241 649 2658 2742 2073 10624 1870 2837 3201 742 1229 4365 1437 1021 Chy 1493 6304 677 1078 1547 6833 223 2969 2323 600 208 1186 2138 1999 Diaph 279 25 76 175 135 181 216 162 302 61 293 673 357 294 Lepto 10 5 0 11 28 0 9 0 26 7 21 18 25 19 Cerco 36 73 118 207 202 76 Diac 1608 1272 593 1054 1369 1954 2398 915 423 191 395 232 675 489 Vern 37 10 0 63 29 83 33 9 16 16 24 8 8 0 Tropo 1018 1369 1009 181 784 253 254 78 629 121 278 27 279 217 Meso 132 47 31 242 182 226 215 107 124 56 141 129 382 145 Veligers 12575 19666 37940 16661 9883 6291 9710 12272 16686 5476 * Other species seen in 1997-2004: Alona sp., Cercopagis pengoi, Eurycercus lamellatus, Holopedium gibberum, Polyphemus pediculus, Sida crystallina, Cyclops scutifer, Epischura lacustris, Eurytemora affini, Leptodiaptomus sicilis, L. siciloides, L. minutus, Skistodiaptomus oregonensis

250 Conway Veligers Calanoids Cyclopoids 200 Cladocerans ) -1 150

100 Biomass (ug.L

0 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003

3000 Conway

2500 ) -1 2000

1500

1000 Production (ug.L Production

500

0 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003

Figure 8. Trends in volumetric seasonally-weighted mean biomass and seasonal production at Conway.

other year of the study. Calanoid adults were not a significant component of the community. In 2004, mean cladoceran size was 4.22 mm, which is similar to the mean sizes observed during most of the post-Dreissena period (Fig. 3). In comparison, the 2003 mean of 0.536 mm was the highest level at Conway since the study began in 1975. This was due to domination by large D. galeata mendotae. In contrast, the mean size was only 0.390 mm in 2000, which was more typical of the 1985 to 1994 period. Unlike the other stations, no dramatic mid-summer peak in Cladocera biomass was observed at Conway (Fig. 4). A small peak in Bosmina biomass was observed in mid-to late June, with a similar peak in Daphnia in late July. Cyclopoids were most abundant in mid- May and late summer. Compared to the other stations, total biomass remained low at Conway for the entire sampling season.

2.0 1.8 Diacyclops - Conway 1.8 1.6 Tropocyclops ) 1.6 -1 1970s 1.4 1.4 g.L 1.2 μ 1.2 1980s 1.0 1.0 1990s 0.8 2000s 0.8

Mean Daily 0.6 0.6 0.4 0.4 Production ( Production 0.2 0.2 0.0 0.0

0.05 0.4 0.05 Cyclops Mesocyclops ) 0.3

-1 0.04 vernalis 0.04 g.L 0.3 μ 0.03 0.2 0.03 0.02 0.2

Mean Daily 0.02 0.1

Production ( Production 0.01 0.01 0.1 0.00 0.0 May June July Aug Sep Oct May June July Aug Sep Oct

Figure 9: Mean daily production estimates of four cyclopoid taxa from May to October at Conway, for four time periods (1970s to 2000s).

Napanee

2003 was the first year that Napanee was sampled since the early 1980s. In 2004, total SWM biomass was 210 µg l-1 at Napanee, which was similar to Belleville (218 µg l-1) but lower than Hay Bay (342 µg l-1; Tab. 7). However, in 2003, biomass was intermediate at Napanee (286.3 µg l-1) relative to Belleville (219.9 µg l-1) and Hay Bay (321.7 µg l-1). In terms of location in the Bay of Quinte, Napanee falls between these two stations. Cladocerans comprised 68.5% of the total zooplankton biomass at Napanee, compared to 18.7% for veligers and 10.1% for cyclopoids. In both 2003 and 2004, veliger larvae were the most commonly encountered taxon (Tab. 8), and the relative proportion of veligers (by biomass) was higher at Napanee than any other station sampled in the Bay of Quinte. The high numbers of veligers at this station supports the observation that dreissenids appear to be more abundant at Napanee than at Belleville (Millard and Burley, 2004). In terms of biomass, the most dominant taxon in 2004 was D. galeata mendotae, followed by veligers, Bosmina and Ceriodaphnia. Production at Napanee was very similar in 2003 and 2004 (5104 and 4935 µg l-1, respectively; Tab. 4). During both years, this station supported the highest zooplankton production in the Bay of Quinte. Cladocerans were responsible for 80.6% of this production, compared to 14.6% for veligers and 4.1% for cyclopoids. In 2004, mean cladoceran length was highest at Napanee relative to the other three stations, averaging 0.450 mm. Total biomass began rising in mid-June and peaked in mid-July due to a dramatic rise in Daphnia biomass (Fig. 4). Daphnia largely vanished by mid-August, and total biomass remained low but relatively constant until early October. Veligers appeared in mid-May, peaked in mid-June and showed a second smaller peak in August.

Table 7. Comparison of zooplankton parameters at Napanee and Belleville, 2003 and 2004.

2003 2004 Parameter N B N B

Most abundant Cladoceran Eubosmina Eubosmina Bosmina Bosmina

Dominant Cladocerans by biomass D. galeata D. galeata D. galeata Bosmina D. ret Eubosmina Bosmina Cerio Eubosmina D. ret Cerio. D. galeata

SWM Total Density (no. m-3) 150,982 126,011 114,156 128,231 SWM Veliger Density (no. m-3) 55,307 22,110 49,289 19,351 SWM D. galeata mendotae Density (no.m-3) 10,999 6,601 7,721 1,779 Mean Cladoceran Size (mm) 0.491 0.471 0.450 0.328

SWM Biomass (µg l-1) 286.3 219.9 210.6 218.1 % Cladocerans by biomass 71.9 76.3 68.5 81.3 % Copepod by biomass 12.7 15.6 12.7 11.6 % Veliger by biomass 15.5 8.0 18.7 7.1

SWM Seasonal Production (µg l-1) 5104 4865 4935 4572 P:B Ratio 17.8 22.1 23.4 21.0 Number of Taxa 24 22 20 22

Table 8. Seasonal mean abundances of major zooplankton species at Station N (nos m-3), 2003 and 2004. Species 2003 2004 Bos 8814 18745 Eubos 16424 5709 D. ret 9395 1548 D.g,m, 10999 7721 Cerio 2247 5368 Chy 12466 4277 Diaph 599 93 Lepto 50 23 Cerco 1 0 Diac 118 0 Vern 65 0 Tropo 1698 1947 Meso 709 307 Veligers 55307 49289

Other comparisons also indicate differences in the zooplankton community between the two upper bay stations. In 2004, D. galeata mendotae dominated at Napanee and Bosmina at Belleville. Furthermore, the seasonal pattern of zooplankton abundance peaks and community composition also differed between the two stations.

Discussion

In 2004, SWM zooplankton biomass in the Bay of Quinte continued its trend of recovery from the very low levels observed during the cold summer of 2000. Biomass was relatively unchanged between 2003 and 2004 at Belleville and Hay Bay, whereas it dropped somewhat in 2004 at both Napanee and Conway. The only other exception to this upward trend was the very high zooplankton biomass at Belleville in 2001. Over the last three years, zooplankton production remained fairly stable in the upper and middle bays, but declined at Conway relative to the previous two years. However, biomass still remained lower than during the peak period of 1982 to 1991. It is likely that this drop has been largely caused by the invasion of Dreissena in 1995 and subsequent food web alterations. Reduced phytoplankton biomass due to zebra mussel grazing was first observed that year (Millard and Burley, 2004). Predation by the invasive cladoceran Cercopagis pengoi has also potentially lowered zooplankton biomass in the lower Bay of Quinte. This exotic zooplankton species was first identified in the Bay of Quinte in 1999. Never abundant in the upper and middle bays, it was virtually absent from the Belleville, Hay Bay and Napanee samples in 2004. Although Cercopagis gradually increased between 1999 and 2002 in the lower bay, it declined in 2004. Benoit et al. (2002) found that peak abundances of Cercopagis corresponded seasonally with declines in cyclopoid nauplii and copepodids at Conway. In Johannsson et al. (2003), it was noted that cyclopoids in general, and most particularly Diacyclops thomasi, have declined in the bay in recent years. It was hypothesized that this decline was caused by a drop in productivity caused by dreissenids, possibly coupled with predation of juvenile copepods by Cercopagis (Johannsson et al., 2004). Furthermore, egg production

has also declined in most cyclopoid taxa in the Bay of Quinte since the late 1990s. However, as the loss of egg-ratio production was most dramatic at Belleville, it is unlikely a Cercopagis effect. The indirect impact of dreissenids would be mediated through reduced food sources for young cyclopoids (rotifers, algae, small cladocerans), or through a nutrient deficiency in the new algal diet. The 2004 biomass data, combined with the change in egg- ratio production estimates, further support the Dreissena hypothesis. Another possibility is that YOY fish (alewife and perch) are selectively preying on egg-bearing copepods. This loss of reproductive copepods needs to be examined further. It is noteworthy that the cladoceran Holopedium only became an important part of the zooplankton community following invasion by dreissenids, and that it was most abundant in 2003 and 2004 at Conway. This species generally prefers cold, oligotrophic waters (Pennak, 1989), and it is an effective grazer of nannoplankton about 20 µm in length (McNaught, 1978). Its adaptation to feed on low-density phytoplankton may give it an advantage over other cladocerans in the presence of Dreissena grazing pressure. Mean water temperatures may also influence zooplankton biomass in the Bay of Quinte. Over the last twenty years, the two years with the lowest abundances (1992 and 2000) also had the lowest June-September mean temperatures in that time period. The cool weather in 1992 can be explained by the eruption of Mount Pinatubo in the Philippines in June 2001 (Newhall and Punongbayan, 1996). The unusually cool weather in 2000 was likely connected to an extended La Nina event in the Pacific Ocean (NOAA). Although the summer of 2004 was quite rainy and appeared cooler than normal, when we examined the Belleville data, August was the only month with a mean water temperature below the thirty-year average. However, this short-term temperature drop appeared to have little effect on Belleville’s 2004 mean biomass. Changes in fish species composition and planktivorous fish biomass also help determine abundance, biomass and mean sizes of zooplankton. The mean length of cladocerans is interpreted as an indication of predation pressure from fish on the zooplankton community, as fish preferentially consume the larger zooplankton individuals (Cooley et al., 1986). In 2001, the high densities of D. galeata mendotae at Belleville, the resulting large mean length, and high zooplankton biomass suggests that fish predation in

the upper bay was low that year. These observations were also made throughout the bay in 2003, again indicating low predation rates. Low levels of planktivory in the lower Bay of Quinte in recent years is further supported by observations that alewife biomass and densities fell in Lake Ontario between the early 1990s and 2002, and have remained consistently low through 2004 (Mr. R. O’Gorman, USGS, Oswego, N.Y., pers. comm.; Shaner and LaPan, 2004). Alewife is the dominant planktivorous fish in the pelagic zone of Lake Ontario, and was one of the most abundant fish in the bay between the early 1970s and 2001 (Hoyle, 2001; Hurley, 1986). Dominance by small cladocerans and the resulting drop in mean size at Belleville in 2002 and 2004 indicate higher predation in the upper bay for those years. D. galeata mendotae (and subsequently mean cladoceran size) declined somewhat at Hay Bay and Conway in 2004 relative to the previous year. This suggests that planktivorous fish may have recovered somewhat in the Bay of Quinte in 2004, but overall, predation remained lower than during the 1999 to 2002 period. Several years ago, concerns were raised that the Belleville station alone may no longer adequately represent the upper Bay of Quinte. For example, macrophytes have proliferated in the nearshore areas around Belleville following the Dreissena invasion. To address this question, the Napanee station was added in 2003 and 2004. There were some differences in the dominant cladoceran taxa, seasonal mean biomass and production between Belleville and Napanee, but many differences were not consistent from one year to the next (Tab. 7). In general terms, however, zooplankton production, mean cladoceran size and veliger biomass were higher at Napanee. These observations suggest that the zooplankton community is somewhat heterogeneous in the upper Bay of Quinte, especially in terms of dreissenid abundance, zooplankton production and impacts of planktivorous fish. In summary, monitoring the Napanee station may help us better understand the status of zooplankton in the bay, and efforts should be continued if resources allow.

References

Benoit, H.P., Johannsson, O.E., Warner, D., Rudstam, L., Sprules, G., 2002. Assessing the Impact of a Recent Predatory Invader: A Study of the Population Dynamics, Vertical

Distribution and Potential Prey of Cercopagis pengoi in Lake Ontario. Limnol. Oceangr. 47, 626-635.

Bowen, K.L., Johannsson, O.E., 2005. Zooplankton in the Bay of Quinte. In: Monitoring Report #14. Project Quinte Annual Report 2003, pp. 47-68. Bay of Quinte Remedial Action Plan, Kingston, Ont., Canada.

Cooley, J.M., Moore, J.E., Geiling, W.T., 1986. Population dynamics, biomass and production of the macrozooplankton in the Bay of Quinte during changes in phosphorus loadings. In: C.K. Minns, D.A. Hurley, K.H. Nichols (Eds.), Project Quinte: point- source phosphorus control and ecosystem response in the Bay of Quinte, Lake Ontario. Can. Spec. Publ. Fish. Aquat. Sci. 86, 166-176.

Hoyle, J.A. 2001. Fisheries Assessment Activities in the Bay of Quinte. In: Monitoring Report #11. Project Quinte Annual Report 1999-2000, pp. 101-118. Bay of Quinte Remedial Action Plan, Kingston, Ont., Canada.

Hurley, D.A., 1986. Fish populations of the Bay of Quinte, Lake Ontario, before and after phosphorus control. In: C.K. Minns, D.A. Hurley, and K.H. Nichols (Eds.), Project Quinte: point-source phosphorus control and ecosystem response in the Bay of Quinte, Lake Ontario. Can. Spec. Publ. Fish. Aquat. Sci. 86, 201-214.

Johannsson, O.E., Dermott, R., Graham, D.M., Dahl, J.A., Millard, E.S., Myles, D.D., LeBlanc, J., 2000. Benthic and Pelagic Secondary Production in Lake Erie after the Invasion of Dreissena spp. with Implications for Fish Production. J. Great Lakes Res. 26, 31-54.

Johannsson, O.E., Gerlofsma, J., Bowen, K.L., 2003. Zooplankton. In: Monitoring Report #12. Project Quinte Annual Report 2001, pp. 50-67. Bay of Quinte Remedial Action Plan, Kingston, Ont., Canada.

Johannsson, O.E., Gerlofsma, J., Bowen, K.L., 2004. Zooplankton in the Bay of Quinte. In: Monitoring Report #13. Project Quinte Annual Report 2002, pp. 39-54. Bay of Quinte Remedial Action Plan, Kingston, Ont., Canada.

Laxson, C.L., McPhedran, K.N., Makarewicz, J.C., Telesh, I.V., 2003. Effects of Cercopagis pengoi on the lower foodweb of Lake Ontario. Freshwat. Biol. 48, 2094- 2106.

MacIsaac, H.J., Grigorovich, I.A., Hoyle, J.A., Yan, N.D., Panov, V.E., 1999. Invasion of Lake Ontario by the Ponto-Caspian predatory cladoceran Cercopagis pengoi. Canadian Journal of Fisheries and Aquatic Sciences 56, 1-5.

Newhall, C.G., Punongbayan, R.S., 1996. Fire and Mud: Eruptions and Lahars of Mount Pinatubo, Philippines. University of Washington Press, Seattle.

Paloheimo, J.E., 1974. Calculation of instantaneous birth rate. Limnol. Oceanogr. 19, 692- 694.

Pennak, R.W., 1989. Freshwater Invertebrates of the United States: Protozoa to Mollusca. John Wiley and Sons, New York, NY, USA.

Shaner, T., LaPan, S., 2004. Pelagic Planktivores. Lake Ontario Annual Report 2003. New York State Department of Environmental Conservation.

BENTHIC FAUNA IN THE BAY OF QUINTE

Ronald Dermott and Robert Bonnell

Great Lakes Laboratory for Fisheries & Aquatic Sciences Fisheries & Oceans Canada 867 Lakeshore Road, P. O. BOX 5050, Burlington, Ontario, L7R 4A6

These are the results of benthic sampling in the Bay of Quinte during August 2004, conducted by the Department of Fisheries and Oceans. The nature of the benthic community changes down the length of the bay in response to depth, average water temperature, and food supply. The locations of the index sites which were established in the 1970's, occur along this gradient (Fig. 1). Only limited observations were made in 2004 on the zebra mussels (Dreissena spp.) living on the soft muds at the 4 Index sites. For information on zebra mussel abundance in the Bay of Quinte see the technical report of Dermott et al. (2003). Again in 2004, samples were collected at the four benthic index sites: Big Bay (7 m), Glenora (22 m), Conway (32 m) and Lake Ontario outside the Upper Gap (LOX site at 31 m depth), using a McKee launch. Each sample was collected with a 9 inch Ekman grab (0.05 m-2 area) which was sieved on a #30 (0.58 mm) screening bucket and preserved initially in 10 % formalin, and within 2 weeks transferred into alcohol. Later the species were identification, enumeration and the biomass determined for the taxa present. The invertebrates were blotted to remove external water then weighed on a balance, the mollusc wet weight values included their shells. Important findings in 2004 were the collection of two taxa new to the Bay, and the continuing decrease in the abundance of the native fingernail clams. At Big Bay, the small oligochaete Dero flabelliger (Naididae) was present at a low density of 0.6 per Ekman (12 m-2). If verified, this is the first record of this south-eastern species in the Great Lakes. One specimen of the small chironomid Microchironomus nigrovittatus was found in the Conway samples. Two specimens of the New Zealand mud snail (Potamopyrgus

antipodarum) were also found again at Conway. This species first appeared in the lower bay near Conway in 1994, probably introduced from uncleaned sampling gear used by the CSS Lauzier off the Niagara Bar where the snail is common, and used again at Quinte during the same cruise. This snail is known to have been spread in western US rivers from mud stuck to waders and construction equipment. (Hosea and Finlayson, 2005). The snail was collected in Quinte only during 1994 and again in 2001 so it appears to be spreading slowly.

Bay of Quinte

Belleville x ay Trenton nw Big Bay Co x x x L.O.X a nor N Gle 12 km

Eastern Lake Ontario

Figure 1. Location of benthic index sites in the Bay of Quinte.

It is now 10 years since the amphipod Diporeia and the isopod Caecidotea began to disappear in the lower Bay of Quinte between Glenora and the Upper Gap. In 2004, all the amphipods found at the index sites were the shallow-water species Gammarus fasciatus. Gammarus had increased in response to the abundance of Dreissena and increased macrophytes. However since 2001, density of Gammarus decreased at Big Bay, possibly due to increased predation by the goby populations (Fig. 2). Although the exotic amphipod Echinogammarus ischnus is present among the zebra mussel beds in the Bay of Quinte, no specimens were collected at the offshore index sites in 2004. Similar to the amphipod Diporeia, the isopod Caecidotea has also been absent at all the index sites since 1993, including Big Bay. Isopods were favoured by eutrophic conditions,

so their decrease after 1977 may partly be due to improving water conditions. Both crustacea species, Diporeia and Caecidotea were very abundant in the lower Bay of Quinte following the die-off of white perch in 1977.

Amphipod -2 mg m Wet

Pisidium -2 mg m Wet

Gastropods -2 mg m Wet m mg

Year 1968 - 2001 Figure 2. Trends in wet biomass (shell-free) of amphipods, Pisidium and gastropods at the Big Bay site.

Table 1. Density of benthic invertebrates at Bay of Quinte Index Sites, number per Ekman sample (0.05 m2) Oligochaetes Sphaeriidae Chironomids Amphipods Isopods Dreissena 1972 Big Bay 66.3 0.8 62.5 0. 0. Glenora 62.1 64.1 30.6 0.5 0.7 Conway 68.5 34.4 15.4 6.0 1.7 LOX 45.8 47.8 9.6 132.3 13.5 1977 Big Bay 117.3 0 122.3 0. 0. Glenora 996.8 359.3 36.5 3.0 0. Conway 678.3 160.3 11.5 9.0 52.3 LOX 99.0 203.1 8.5 480.6 2.7 1982 Big Bay 35.8 0.0 21.6 0.0 0. Glenora 200.6 14.3 40.0 32.6 2.3 Conway 38.8 5.8 10.6 210.0 2.6 LOX 27.3 3.5 171.3 317.8 1.5 1983 Big Bay 80.0 0.0 14.5 0.3 0. Glenora 35.6 13.5 74.2 33.0 2.0 Conway 12.5 5.2 20.2 518.3 0.7 LOX 28.6 22.2 0.8 154.3 7.8 1992 Big Bay 53.2 2.0 67.6 0.8 0. Glenora 67.0 34.5 77.2 77.4 4.0 Conway 59.2 27.2 3.0 445.0 29.6 LOX 46.6 183.4 4.6 135.4 4.8 1996 Big Bay 107.4 34.5 28.2 1.5 0. 0. Glenora 99.2 24.0 53.6 110.2 0. 0.0. Conway 28.6 68.4 10.0 11.0 0. 0. LOX 139.8 24.0 2.4 0.6 0. 1997 Big Bay 44.4 35.2 60.0 3.8 0. 0. Glenora 61.6 7.4 32.8 26.4 0. 0.2 Conway 91.2 33.2 5.0 0. 0. 0.4 LOX 195.4 27.4 2.6 0.2 0. 0.

Table 1 (cont’d)

Oligochaete Sphaeriidae Chironomids Amphipods Isopods Dreissena 1998 Big Bay 109.0 7.8 44.6 1.2 0. 0.2 Glenora 26.6 27.4 37.8 4.4 0. 4.4 Conway 177.0 28.2 2.4 0.2 0. 8.2 LOX 162.2 21.0 9.8 0. 0. 0.6 1999 Big Bay 220.6 53.2 75.6 1.2 0. 0.2 Glenora 108.0 14.5 37.8 0.8 0. 0.5 Conway 141.2 10.0 21.6 1.2 0. 76.8 LOX 130.2 9.6 11.6 0.8 0. 284.4 2000 Big Bay 132.8 54.3 46.8 4.5 0. 0.5 Glenora 78.3 8.8 45.0 1.5 0. 0.0 Conway 168.3 11.5 74.8 2.3 0. 351.5 LOX 311.8 6.5 33.2 2.3 0. 34.5 2001 Big Bay 88.0 74.6 116.8 8.0 0. 284 Glenora 17.8 14.4 65.0 2.0 0. 0.8 Conway 58.2 5.2 25.8 2.8 0. 21.2 LOX 211.8 6.0 7.0 0 0. 15.8 2002 Big Bay 108.8 51.2 11.4 0.8 0. 0.4 Glenora 96.8 8.2 20.8 0.2 0. 2.0 Conway 105.6 8.2 32.8 3.2 0. 51.4 LOX 285.4 4.2 37.6 1.0 0. 81.4 2003 Big Bay 37.4 12.6 15.0 0.4 0. 4.6 Glenora 30.6 0.0 68.6 0.0 0. 0.4 Conway 90.0 7.8 21.0 0.4 0. 232.0 LOX 225.8 4.8 21.8 0.0 0. 34.2 2004 Big Bay 39.2 10.6 16.4 0.2 0. 0.2 Glenora 48.4 0.4 45.0 0.0 0. 0.0 Conway 113.8 1.8 7.8 0.0 0. 35.2 LOX 248.6 0.8 6.8 0.0 0. 48.2

Quinte Big Bay wet wts. g m-2 Oligochaetes

c c

02 04 06

Figure 3. Trends in wet biomass of oligochaetes and chironomids at the Big Bay site.

Average bottom fauna densities the Bay of Quinte during 2004, were similar to that in 2002 and 2003. Oligochaete density and biomass remained highest at the Lake Ontario exit site (LOX; Tab. 1). The biomass of the oligochaetes at Big Bay was lower in 2004 than that over the last 2 years, but greater than the biomass during the period 1990 to 2000 (Fig. 3). Oligochaete abundance decreased following improved sewage treatment and nutrient reduction in the bay, as the amount of decaying detritus and sediment bacteria decreased. At the Lake Ontario site LOX, oligochaetes have increased following the disappearance of the amphipod Diporeia in 1996 (Tab. 1). The settling detritus previously consumed by the amphipods is now consumed by the worms. Although not quantified, the abundance of the filamentous bacteria Thioploca has also decreased at both LOX and Conway since 2001. The oligochaete composition has remained very similar at the Big Bay, and Glenora sites

since 1977. At Conway, the worms Spirosperma ferox and Potomothrix vejdovski are now more common than in the 1980s (Johnson and McNeil, 1986). These worms replaced Quistadrilus multisetosus, which had been common in the Bay during the eutrophic 1970's but has since become very rare as water quality improved. However, the oligotrophic indicator worm Stylodrilus heringianus, which is sensitive to low oxygen, has become very rare in the lower bay and the LOX site since 1997. Oligochaete worms were the dominant portion of the non-Dreissena biomass at Conway and especially at the Lake Ontario site (Tab. 2). The chironomid midges still formed the greatest biomass of the invertebrates at Big Bay and Glenora (Tab. 2). The large red Chironomus midges are much more abundant in the upper bay and Glenora. The largest Chironomus plumosus were common only at the Big Bay site, this species was replaced by the smaller Chironomus atritibia at the lower bay sites and at the LOX site. The density of smaller chironomids at the Conway and LOX site has decreased in recent years.

Table 2. Wet Biomass (mg /Ekman +shells) of benthic invertebrates in the Bay of Quinte. Year Oligochaete Sphaeriidae Chironomids Amphipods Isopods Dreiss. 1991 Big Bay 70.3 ± 19.8 4.2 ± 1.7 283.1 ± 63.0 4.6 ± 2.9 0.0 ± 0.0 Glenora 18.7 ± 5.3 42.6 ± 3.3 741.9 ± 46.7 206.5 ± 24.5 78.9 ± 10.3 Conway 233.5 ± 43.3 107.6 ± 8.7 69.2 ± 24.7 730.3 ± 72.9 577.9 ± 25.5 LOX 48.2 ± 5.3 221.1 ± 35.8 19.4 ± 8.7 1409.6 ±67.2 12.9 ± 6.0 1992 Big Bay 22.1 ± 4.2 1.0 ± 0.3 66.5 ± 8.9 2.9 ± 2.7 0.0 ± 0.0 Glenora 20.4 ± 4.1 84.2 ± 8.6 292.6 ± 42.5 80.5 ± 14.3 0.6 ± 0.3 Conway 42.4 ± 6.3 27.7 ± 6.1 22.6 ± 9.9 392.8 ± 21.5 118.3 ± 32.1 LOX 39.9 ± 13.6 113.6 ±14.5 5.3 ± 1.2 167.6 ±28.4 9.7 ± 5.1 1993 Big Bay 70.8 ± 5.7 5.7 ± 1.6 325.0 ± 32.9 2.4 ± 1.1 0.1 ± 0.1 Glenora 12.2 ± 7.6 146.5 ± 34.2 206.1 ± 29.2 20.2 ± 3.3 0.9 ± 0.2 Conway 57.6 ± 13.0 93.2 ± 11.1 19.1 ± 6.5 0.0 ± 0.0 0.0 ± 0.0 LOX 60.2 ± 8.23 147.2 ± 47.5 8.6 ± 2.2 173.5 ± 68.9 6.9 ± 6.9 1996 Big Bay 42.4 ± 12.3 14.9 ± 3.7 29.9 ± 9.9 2.8 ± 0.2 0. ± 0. 0.0 ± 0.0 Glenora 31.5 ± 2.8 36.6 ± 6.3 154.5 ± 18.8 137.2 ± 8.1 0. ± 0. 0.0 ± 0.0 Conway 41.4 ± 11.7 61.3 ± 19.3 13.4 ± 3.4 7.7 ± 1.7 0. ± 0. 0.0 ± 0.0 LOX 430.7 ± 186.2 11.7 ± 3.6 2.3 ± 0.9 0.2 ± 0.2 0. ± 0. 0.0 ± 0.0

Table 2 (cont’d)

Year Oligochaete Sphaeriidae Chironomids Amphipods Isopods Dreiss. 1997 Big Bay 20.2 ± 5.4 32.1 ± 5.0 40.4 ± 7.1 1.4 ± 0.7 0. ± 0. 0.0 ± 0.0 Glenora 19.9 ± 5.1 12.9 ± 3.6 130.8 ± 15.3 36.9 ± 6.6 0. ± 0. 0.2 ± 0.2 Conway 44.5 ± 12.6 49.9 ± 8.7 5.1 ± 1.0 0.0 ± 0.0 0. ± 0. 39.9 ± 39.9 LOX 282.7 ± 71.7 18.9 ± 5.7 0.1 ± 0.1 0.1 ± 0.1 0. ± 0. 0.0 ± 0.0 1998 Big Bay 21.2 ± 6.9 3.5 ± 2.0 35.1 ± 26.3 0.4 ± 0.3 0. ± 0. 0.1 ± 0.1 Glenora 14.5 ± 6.5 24.0 ± 9.8 103.0 ± 19.2 11.5 ± 9.5 0. ± 0. 18.2 ± 11.5 Conway 127.0 ± 30.3 31.5 ± 10.1 1.1 ± 0.6 0.1 ± 0.1 0. ± 0. 65.6 ± 62.6 LOX 271.7 ± 61.6 12.5 ± 2.6 9.4 ± 2.4 0.0 ± 0.0 0. ± 0. 9.4 ± 4.9 1999 Big Bay 10.5 ± 2.1 16.8 ± 3.3 32.4 ± 10.3 0.5 ± 0.4 0. ± 0. 0.1 ± 0.1 Glenora 24.7 ± 5.5 13.6 ± 3.5 106.2 ± 10.4 1.3 ± 1.1 0. ± 0. 0.2 ± 0.2 Conway 163.5 ± 22.4 10.9 ± 4.6 10.9 ± 3.2 0.5 ± 0.3 0. ± 0. 9550 ± 7249 LOX 213.1 ± 67.8 11.4 ± 7.3 11.9 ± 4.5 0.4 ± 0.2 0. ± 0. 1334 ± 839 2000 Big Bay 74.8 ± 16.8 20.5 ± 4.9 199.1 ± 68.1 1.2 ± 0.5 0. ± 0. 5.3 ± 3.2 Glenora 18.2 ± 2.9 14.0 ± 4.9 139.4 ± 12.8 0.4 ± 0.3 0. ± 0. 0.0 ± 0.0 Conway 234.9 ± 31.2 46.9 ± 33.7 34.9 ± 7.8 1.8 ± 0.8 0. ± 0. 15052± 3272 LOX 516.0 ± 2.9 ± 0.9 27.2 ± 4.1 0.4 ± 0.3 0. ± 0. 697.7 ± 576 107.0 2001 Big Bay 53.8 ± 10.1 31.4 ± 9.0 65.9 ± 14.3 2.2 ± 0.4 0. ± 0. 52.1 ± 19.6 Glenora 9.1 ± 3.2 9.9 ± 1.8 130.3 ± 9.7 0.4 ± 0.3 0. ± 0. 2.9 ± 1.1 Conway 64.4 ± 4.8 7.9 ± 1.8 20.3 ± 2.1 1.9 ±1.4 0. ± 0. 10108±3174 LOX 152.1 ± 29.5 7.4 ± 3.5 11.3 ± 2.9 0.0 ± 0.0 0. ± 0. 339.3 ±145.6 2002 Big Bay 88.1.± 23.1 25.6 ± 8.1 90.7 ± 29.3 0.1 ± 0.1 0. ± 0. 0.1 ± 0.1 Glenora 15.6 ± 3.5 12.0 ± 4.8 60.3 ± 8.1 0.1 ± 0.1 0. ± 0. 2.3 ± 1.0 Conway 69.2 ± 14.4 3.6 ± 1.4 45.6 ± 12.3 2.5 ±1.8 0. ± 0. 15991±4725 LOX 457.1 ± 32.1 3.6 ± 1.0 23.0 ± 3.6 0.2 ± 0.2 0. ± 0. 8171 ± 4513 2003 Big Bay 88.7 ± 11.9 9.5 ± 3.9 108.9 ± 23.2 1.7 ± 1.7 0. ± 0. 79.3 ± 56.3 Glenora 11.4 ± 4.6 0 ± 0 205.3 ± 1.7 0 ± 0 0. ± 0. 0.1 ± 0.1 Conway 99.0 ± 21.1 6.2 ± 2.0 28.7 ± 3.8 0.1 ± 0.1 0. ± 0. 20624±5221 LOX 399.1 ±189.2 4.6 ± 1.0 31.6 ± 3.9 0 ± 0 0. ± 0. 1457 ± 609 2004 Big Bay 58.9 ± 20.3 3.2 ± 1.8 149.4 ± 44.0 0.1 ± 0.1 0. ± 0. 0.1 ± 0.1 Glenora 12.6 ± 4.6 0.1 ± 0.1 111.7 ± 30.0 0 ± 0 0. ± 0. 0. ± 0 Conway 41.4 ± 4.4 1.3 ± 0.5 15.5 ± 4.5 0 ± 0 0. ± 0. 16985±6536 LOX 446.4 ±135.7 1.0 ± 0.7 88.7 ± 53.4 0 ± 0 0. ± 0. 4770 ± 2762

Since 1998, density of the native fingernail clams (Sphaeriidae; mostly Pisidium casertanum) has been much lower at all sites except Big Bay. Pisidium spp. (pea clams) was most abundant at Big Bay, where their density peaked in 2001. Density of the fingernail clams continued to decrease in the lower bay. The larger fingernail clams Sphaerium corneum and Musculium spp., which had been common on the sandier sediments during the 1980's, have become very rare since the arrival of the Dreissena mussels. All 4 genera are filter-feeding bivalves which complete for algae and suspended matter, but the smaller Pisidium spp. can supplement their needs by burrowing into the sediments in search of bacteria. Quagga mussels (Dreissena bugensis) continue to be abundant on the soft mid-channel muds at the Conway and LOX sites, forming a wet biomass (including shells) of 16.9 gm per sample (339 g m-2). Still, almost all the Dreissena at Big Bay were zebra mussels (D. polymorpha) in 2004. The young mussels at Big Bay and Glenora do not usually survive over the winter possibly due to being buried by re-suspended sediments during the fall and winter storms. Substrate at the Lake Ontario site is fine silty sand over hard clay which may provide the quagga mussels with a more stable substrate. At Conway, the substrate is fine silt with a few clumps of mussels. Once a clump of shells forms on the bottom, succeeding generations can colonize these clumps and begin to spread out on the bottom. Again in 2004, most of the Dreissena spp. at the Conway site were young recently settled mussel with lengths less than 5 mm (18.2 per Ekman). However, almost all the biomass (16.9 gm per Ekman) was large mussels while the young created only 0.001 gm per Ekman. Total non-Dreissena biomass in the Bay has displayed a decreasing trend since the arrival of Dreissena in the early 1990's. This is most evident in the lower bay at the Conway and LOX sites (Fig. 4). The disappearance of the large deepwater amphipod population (Diporeia) in 1995 from the lower bay and eastern Lake Ontario, caused the greatest drop in total biomass. The slow progressive decrease in native clams (Pisidium and Sphaerium spp.) has continued the reduction in total non-Dreissena biomass. The quagga mussels at Conway and the LOX site make up much more biomass than was lost in the other benthic fauna. This emphasizes the dominating role the introduced mussels are having on the community of the Bay of Quinte.

Total Non-Dreissena Wet Biomass (+shells) gm m-2

60 O

50 C Conway (Lower Quinte) -2 40 O L.O. Eastern L. Ontario log L.O. r 2 =0.55 CO log Con r 2 =0.52 30 O C O

gm. wet mwet gm. 20

C CO C 10 O O O C O O O O O O C CO CO O C C C C CO C C 0 C C 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 Year Figure 4. Trends in decreasing non-Dreissena biomass in the lower Bay of Quinte.

Table 3. Benthic Diversity (Margalef) in the Bay of Quinte. Year Big Bay Glenora Conway L. Ontario (7 m) (21 m) (32 m) (31 m) 1967 1.17 1.99 1.39 1.33 1972 1.05 1.32 1.39 0.99 1982 0.78 1.66 1.27 1.42 1985 1.14 2.04 1.13 1.51 1992 1.79 1.54 1.10 1.08 1995 2.33 2.39 1.85 1.44 1999 3.87 2.74 2.62 1.23 2000 3.30 2.49 2.82 2.16 2001 3.06 2.99 2.45 1.88 2002 2.00 3.05 2.78 1.85 2003 2.81 2.28 2.63 2.35 2004 2.18 2.98 2.30 1.50

Invertebrate diversity at all but the Lake Ontario Index sites was very similar (2.1-3.0), indicating a moderately diverse benthic community (Tab. 3). Diversity at the LOX site (1.5) decreased in 2004, but has remained similar to that in the 1980’s. At that site the benthic community was limited to 3 species of oligochaete, 2 species in each of the chironomids, Pisidium, ostracods, as well as quagga mussels, and nematodes. Levels of this Margalef

Diversity Index (Margalef, 1958) measures number of species (s-1) / natural log of total individuals. This index ranges from less than 1 in stressed communities dominated by a few species to a value of over 3 in diverse communities.

References

Dermott, R., Bonnell, R., Carou, S., Dow, J., Jarvis, P., 2003. Spatial distribution and population structure of the mussels Dreissena polymorpha and Dreissena bugensis in the Bay of Quinte, Lake Ontario, 1998 and 2000. Can. Tech. Rep. Fish. & Aquat. Sci. 2479, 58 pp.

Hosea, R.C., Finlayson, B., 2005. Controlling the spread of New Zealand mud snails on wading gear. California Department of Fish and Game, Office of Spill Prevention and Response. Admin. Report 2005-02.

Johnson, M.G., McNeil, O.C., 1986. Changes in abundance and species composition in benthic invertebrate communities of the Bay of Quinte, 1966-84. Can. Spec. Publ. Fish. Aquat. Sci. 86, 177-189.

Margalef, R., 1958. Information theory in ecology. Gen. Syst. 3, 36-71.

BAY OF QUINTE FISH AND FISHERIES, 2004

J. A. Hoyle

Ontario Ministry of Natural Resources Lake Ontario Management Unit Glenora Fisheries Station R.R. #4, Picton, Ontario, K0K 2T0

Introduction

This report highlights three major fisheries assessment activities conducted by the Lake Ontario Management Unit during 2004 on the Bay of Quinte—fish community index netting surveys, commercial harvest reporting, and recreational fishing surveys.

Fish community index gillnetting, trawling, and nearshore trapnetting surveys

Annual summer index gillnetting and bottom trawling are used to detect long-term changes in the Bay of Quinte fish community (Fig. 1). By providing trend-through-time indices of species abundance, these projects have also routinely delivered stock-specific information to fisheries managers. The projects have run for over 40 years in the case of gillnetting and for 30 years in the case of trawling (Casselman and Scott, 1992; Hurley, 1992; Hoyle, 2004). In addition, a provincially standardized trapnet project (NSCIN) was initiated in and conducted annually since 2001 to assess an expanding nearshore fish community. The nearshore fish community expanded after dreissenid mussel invasion and resultant increases in water clarity and aquatic vegetation in the mid-1990s. Catch summaries for the three gear types—gillnets, trawling and nearshore trapnetting—for 2004 are presented in Tab. 1, 2 and 3 respectively.

Table 1. Species-specific catch per gillnet set in the Bay of Quinte, 1992-2004. Shown are the average catches in 1-3 gillnet gangs set at each of 1-5 depths during each of 2-4 visits to each of 3 sites (Big Bay, Hay Bay and Conway). The total number of sets each year is indicated. Year Species 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 Mean 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.0 0.0 0.0 0.0 0.0 0.0 Lake sturgeon 0.0 Longnose gar 0.9 5.5 0.2 3.8 0.7 1.4 0.0 5.9 0.6 1.6 1.5 0.2 1.2 1.8 Alewife 315.6 248.5 347.2 224.5 85.5 183.8 121.7 8.5 54.9 58.3 23.8 25.2 68.3 135.8 Gizzard shad 1.8 34.1 5.3 27.4 0.5 1.2 1.8 22.7 2.5 3.1 10.1 2.3 0.4 8.7 Chinook 0.2 0.9 0.0 0.0 0.0 0.0 0.4 0.2 0.0 0.2 0.0 0.2 0.4 salmon 0.2 Atlantic 0.0 0.0 0.2 0.0 0.0 0.0 0.0 0.2 0.0 0.0 0.0 0.0 0.0 salmon 0.0 Brown trout 6.6 4.7 1.3 1.6 0.0 0.8 0.2 0.2 0.0 0.4 0.2 1.4 0.4 1.4 Lake trout 22.3 8.8 7.1 4.1 15.3 9.1 5.0 0.6 5.3 2.7 8.4 7.2 7.9 8.0 Lake whitefish 8.0 6.6 2.6 0.0 6.1 2.1 7.2 2.1 1.2 1.8 0.9 2.9 0.4 3.2 Cisco (Lake 1.1 4.7 1.5 1.9 10.8 21.6 23.2 0.8 4.5 2.2 0.2 0.0 0.2 herring) 5.6 Coregonus sp. 0.0 0.0 0.0 0.3 0.0 0.6 0.2 0.0 0.0 0.0 0.0 0.0 0.2 0.1 Rainbow smelt 1.3 0.6 1.6 0.8 0.0 0.6 1.8 1.1 0.0 0.7 0.4 0.0 0.2 0.7 Northern pike 2.7 4.1 6.8 1.9 2.6 1.2 0.9 1.3 1.6 1.6 0.4 0.8 0.2 2.0 Mooneye 0.0 0.0 0.0 0.0 0.0 0.4 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 White sucker 33.1 30.1 30.9 36.0 26.1 29.6 20.6 23.8 22.0 25.4 27.2 14.5 19.7 26.1 Silver sedhorse 0.0 0.0 0.0 0.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Moxostoma sp. 0.0 0.3 0.0 0.0 0.0 0.6 0.0 0.0 0.0 0.2 0.0 0.2 0.0 0.1 Common car 1.5 2.5 1.3 0.0 0.0 1.2 0.4 0.0 0.2 0.0 0.0 0.2 0.2 0.6 Spottail shiner 0.0 0.0 0.0 0.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Brown 6.4 32.6 11.5 7.1 2.8 4.3 10.1 10.6 6.8 11.3 8.2 2.9 3.9 bullhead 9.1 Channel catfish 0.5 3.3 1.1 0.3 0.2 0.6 0.7 0.4 0.2 0.2 0.4 0.2 0.4 0.7 Stonecat 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.2 0.2 0.0 0.0 0.0 Burbot 0.0 2.2 0.0 0.3 0.0 0.2 0.7 0.0 0.0 0.0 0.0 0.0 0.0 0.3 Trout-perch 0.0 0.4 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 White perch 221.7 282.9 276.0 130.8 40.2 49.5 65.3 101.0 43.0 32.9 61.2 85.7 184.2 121.1

Table 1. Continued.

Year Species 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 Mean White bass 0.5 0.0 0.0 0.3 0.0 0.0 0.0 0.0 0.0 0.0 0.2 0.0 0.0 0.1 Rock bass 14.8 24.7 4.6 8.2 3.8 8.8 11.2 11.0 5.1 1.6 3.3 0.6 0.6 7.6 Pumpkinseed 0.0 6.6 0.0 0.5 1.9 3.1 21.3 18.3 11.7 26.7 13.7 2.1 8.3 8.8 Bluegill 0.0 0.0 0.0 0.0 0.2 0.8 2.2 1.1 1.4 10.4 5.5 0.6 0.4 1.7 Smallmouth 2.9 3.8 0.5 0.8 2.1 7.4 3.7 4.5 1.6 1.1 0.2 0.0 0.0 2.2 bass Largemouth 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.2 0.0 0.0 0.0 0.0 0.0 0.0 bass Black crappie 0.4 1.1 0.0 0.0 0.0 0.0 0.7 0.0 0.0 0.4 0.5 0.4 0.2 0.3 Yellow perch 725.1 948.1 513.0 747.0 547.5 624.8 667.1 896.6 752.5 728.8 714.5 493.2 388.7 672.9 Walleye (Yellow 84.2 131.9 54.5 77.4 60.2 32.9 31.4 29.5 24.5 13.9 21.9 22.3 16.6 46.2 pickerel) Round goby 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 4.0 43.3 120.9 12.9 Freshwater 16.6 17.5 15.9 17.5 21.9 19.5 12.9 13.2 15.8 31.6 15.7 11.0 21.3 17.7 drum Total catch 1468.4 1806.6 1282.9 1293.3 828.2 1006.5 1010.8 1153.9 955.6 957.4 922.5 717.3 845.2 1096.0 Number of sets 36 21 36 24 28 32 30 31 32 36 36 34 34

Table 2. Species-specific catches by site in the 2004 fish community index bottom trawling program in the Bay of Quinte. Catches are the total number of fish observed for the number of trawls indicated. Trawls distances were a ¼ mile. Approximate site depths are indicated in brackets. Trenton Belleville Big Bay Deseronto Hay Bay Conway Species (4 m) (5 m) (6 m) (5 m) (7m ) (23 m) Silver lamprey 0 0 0 0 0 1 Alewife 69 923260 443 537 23 Gizzard shad 0 4720 467 11 1 0 Lake trout 0 0 0 0 0 5 Lake whitefish 0 0 0 0 0 9 Cisco (Lake herring) 0 0 0 0 0 1 Rainbow smelt 0 0 0 0 0 43 White sucker 9 1 11 10 66 89 Common carp 0 0 3 0 0 0 Spottail shiner 10 31 162 12 515 0 Brown bullhead 9 143 190 296 83 0 Channel catfish 0 0 0 1 0 0 Burbot 0 00 0 0 1 Trout-perch 0 115 25 36 620 520 White perch 646 15441 11993 24609 3963 0 White bass 0 29 2 13 1 0 Rock bass 0 0 0 0 1 0 Pumpkinseed 454 33 293 156 37 0 Bluegill 7 0 6 0 0 0 Smallmouth bass 4 0 0 10 0 0 Largemouth bass 1 5 2 2 0 0 Black crappie 0 2 3 0 0 0 Lepomis sp. 0 19 0 0 0 0 Yellow perch 1492 530 1602 1217 4422 704 Walleye 12 21135 121 25 12 Johnny darter 3 0 0 0 0 0 Logperch 34 1 0 0 1 0 Round goby 68 371 126 165 28 1015 Freshwater drum 3 35 125 125 84 0 Total 2821 2158918405 27228 10384 2422 Number of trawls 8 8 8 8 8 12

Table 3. Species-specific NSCIN trapnet catches in the upper and lower Bay of Quinte, 2001- 2004. No netting was completed in the lower Bay of Quinte in 2001. The numbers of trapnet sets are indicated. Upper Bay Lower Bay Total Species 2001 2002 2003 2004 2002 2003 2004 2002 2003 2004 Longnose gar 9 12 41 70 13 16 2 25 57 72 Bowfin 13 5 21 19 24 36 12 29 57 31 Gizzard shad 40 52 72 2 27 18 7 79 90 9 Lake trout 0 0 0 1 0 0 0 0 0 1 Northern pike 37 21 31 25 42 36 28 63 67 53 Mooneye 1 0 0 0 0 0 0 0 0 0 White sucker 37 53 62 45 107 141 92 160 203 137 Silver sedhorse 0 0 25 29 0 3 2 0 28 31 Shorthead redhorse 0 0 3 17 0 0 0 0 3 17 Greater redhorse 0 0 8 2 0 0 0 0 8 2 River redhorse 2 0 5 6 0 0 0 0 5 6 Moxostoma sp. 23 15 3 0 0 0 0 15 3 0 Goldfish 0 0 00100 1 00 Common carp 3 4 10 3 12 11 8 16 21 11 Golden shiner 1 0 1 0 3 1 3 3 2 3 Rudd 0 0 0 0 1 0 0 1 0 0 Brown bullhead 6036 3450 1344 750 2501 2844 1254 5951 4188 2004 Channel catfish 78 78 54 48 41 50 43 119 104 91 American eel 16 5 0 1 6 6 2 11 6 3 White perch 79 104 277 132 39 270 84 143 547 216 White bass 2 5 4 4 1 6 4 6 10 8 Rock bass 33 24 23 21 51 149 34 75 172 55 Pumpkinseed 3218 2631 970 552 4087 745 660 6718 1715 1212 Bluegill 6105 5135 2385 2707 453 253 299 5588 2638 3006 Smallmouth bass 34 60 13 59 28 38 25 88 51 84 Largemouth bass 89 220 285 219 181 92 124 401 377 343 Black crappie 353 540 368 580 209 187 155 749 555 735 Yellow perch 135 123 70 30 117 50 60 240 120 90 Walleye 114 89 80 92 164 295 202 253 375 294 Freshwater drum 229 119 137 77 186 252 190 305 389 267 Effort (Number of net sets) 36 36 36 36 36 36 36 72 72 72

Bay of Quinte Big Bay Deseronto Belleville

Trenton Conway Hay Bay

Lake Ontario

N Lake Ontario

Figure 1. Map of the Bay of Quinte showing fish community index gillnetting (bars) and bottom trawling (circles) locations.

The top five most abundant species in Bay of Quinte gillnets were yellow perch, white perch, round goby, alewife, and freshwater drum (Tab. 1). Of these species, white perch, round goby, alewife and drum were more abundant in 2004 than 2003 while yellow perch were less abundant. Round goby has increased exponentially since its arrival in the late- 1990s. The top five most abundant species in Bay of Quinte trawls were white perch, yellow perch, gizzard shad, alewife, and round goby (Tab. 2). Seventy-two random trapnet sites were sampled from September 7 to October 5 in a variety of nearshore habitat types (Tab. 3). Nearly 8,800 fish comprising 26 species were captured. The five most abundant species by number were bluegill, brown bullhead, pumpkinseed, black crappie and largemouth bass. The five most abundant species by weight were brown bullhead, walleye, freshwater drum, channel catfish and bluegill. The centrarchid family of fish (bluegill, pumpkinseed, black crappie, largemouth bass, rock bass and smallmouth bass) comprised a total of 62% by number and 23% by weight of the catch.

1-3 Bay of Quinte 1-4

44o N 1-1 1-2

1-8 Lake Ontario

78o W

Figure 2. Map of commercial harvest quota zones on the Canadian waters of Lake Ontario, the Bay of Quinte, and the St. Lawrence River.

Commercial harvest reporting

Lake Ontario supports a relatively small but locally important commercial fish industry. The commercial harvest comes primarily from the Canadian waters of Lake Ontario east of Brighton and the Bay of Quinte (Fig. 2). The total harvest of all species in 2004 was 404,236 lb with an associated value of $249,444 (Tab. 4). The top five species in terms of landed value were yellow perch, lake whitefish, sunfish, brown bullhead, sunfish and walleye.

Recreational angling surveys

Bay of Quinte recreational angling surveys are conducted annually during the walleye angling season (January 1 to February 28 and first Saturday in May to December 31). Angling effort is measured using aerial counts during ice fishing surveys, and a

combination of aerial counts and on-water counts during open-water surveys. On-ice and on-water angler interviews provide information on catch/harvest rates and biological characteristics of the harvest.

Table 4. Commercial fish harvest (lb) and value ($) for fish species in the Canadian waters of Lake Ontario, 2004. Quota zone Price Harvest Species 1-1 1-2 1-3 1-4 1-8 per lb (lb) Value American eel 4 306 $3.95 309 $1,222 Black crappie 22 14 5,133 4 7 $2.20 5,179 $11,394 Bowfin 788 3,799 3 3 $0.30 4,593 $1,378 Brown bullhead 6,547 1,013 79,350 3,166 707 $0.34 90,784 $30,867 Channel catfish 21 $0.43 21 $9 Common carp 3,338 2,686 5,747 760 $0.18 12,531 $2,256 Freshwater drum 10 1,777 20,601 8,924 85 $0.09 31,397 $2,826 Lake herring 2 132 481 308 $0.25 923 $231 Lake whitefish 3,300 80,949 14,178 2,734 $0.46 101,161 $46,534 Rock bass 1,399 715 3,260 407 38 $0.46 5,818 $2,676 Round goby 8 8 $ - Suckers 174 10,634 164 453 $0.10 11,425 $1,143 Sunfish 2,341 151 42,790 63 $0.95 45,343 $43,076 Walleye 672 1,992 4,226 8 $2.00 6,897 $13,794 White bass 1 5 3 112 74 $0.73 195 $142 White perch 8 137 8,337 3,729 37 $0.30 12,249 $3,675 Yellow perch 1,366 22,479 22,701 28,572 286 $1.17 75,404 $88,222 Total 16,453 112,875 213,956 58,472 2,479 404,236 $249,444

Ice fishery

Fishing effort in 2004 (79,767 angler hours) was the highest since 2000. Numbers of walleye caught and harvested were 8,413 and 4,075 respectively. Walleye fishing success (number of walleye caught and harvest per hour were 0.105 and 0.051 respectively) was much higher than the previous year (Tab. 5).

Table 5. Summary of total fishing effort (virtually all fishing effort is targeted at walleye), numbers of fish harvested and caught, and walleye angling success (CUE and HUE are the numbers of walleye caught and harvested, respectively, per hour) during the Bay of Quinte ice fishery, 1993-2004. 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 Fishing Effort (angler hours): Total All Anglers 271,088 300,049 215,518 392,602 220,263 117,602 140,363 139,047 77,074 37,129 16,237 79,767 Number of Walleye: Caught 21,326 31,060 28,939 58,468 42,315 11,167 23,293 9,949 982 2,601 321 8,413 Harvested 14,816 8,557 17,445 20,972 22,631 6,089 15,285 9,240 938 2,468 70 4,075 Walleye Angling Success: CUE 0.079 0.104 0.134 0.149 0.192 0.095 0.166 0.072 0.013 0.070 0.020 0.105 HUE 0.055 0.029 0.081 0.053 0.103 0.052 0.109 0.066 0.012 0.066 0.004 0.051

Table 6. Summary of fishing effort (expressed in angler hours separately for all anglers and those targeting walleye), numbers of fish harvested and caught, and walleye angling success (CUE and HUE are the numbers of walleye caught and harvested, respectively, per hour by anglers targeting walleye) during the Bay of Quinte open-water recreational fishery, 1993-2004. 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 Fishing Effort (angler hours): Total All Anglers 644,477 693,731 519,276 665,436 544,476 481,553 379,012 309,259 247,537 177,092 219,684 241,700 Anglers Targeting Walleye 637,401 689,543 512,054 660,005 539,276 475,678 374,128 296,841 222,052 154,570 194,168 203,082 Number of Fish Harvested: Northern Pike 2,279 1,717 375 1,228 1,501 1,539 1,413 2,561 1,658 7,084 818 1,356 Smallmouth Bass1 778 519 704 1,075 Largemouth Bass1 4,890 2,340 4,333 6,808 Yellow Perch 8,205 5,226 14,587 33,609 31,462 41,313 35,102 17,630 7,768 3,876 4,588 3,440 Walleye 145,383 145,642 98,537 117,931 82,790 52,844 33,575 22,811 28,078 17,903 34,905 24,277 Number of Fish Caught: Northern Pike 10,318 11,691 2,964 5,884 7,912 7,950 11,577 15,809 10,835 7,084 5,134 7,834 Smallmouth Bass1 6,347 2,884 3,453 4,052 Largemouth Bass1 19,675 11,387 15,002 22,946 Yellow Perch 141,424 80,699 102,433 298,677 402,216 20,849 391,708 260,029 143,530 104,071 125,129 70,369 Walleye 266,638 262,760 166,229 209,280 134,651 70,527 47,562 28,024 40,734 29,459 70,471 39,251 Walleye Angling Success CUE 0.417 0.378 0.320 0.317 0.250 0.148 0.127 0.094 0.182 0.186 0.344 0.193 HUE 0.227 0.209 0.189 0.179 0.154 0.111 0.090 0.077 0.126 0.113 0.178 0.119 1The number of smallmouth and largemouth bass are for the last Saturday in June (opening day of bass season) to November 30, and are only available for the past four years.

Open-water fishery

Fishing effort in 2004 (241,700 angler hours for all anglers and 203,082 hours for anglers targeting walleye) was higher than the previous two years. Numbers of walleye caught and harvested were 39,252 and 24,277 respectively. Walleye fishing success (number of walleye caught and harvest per hour by anglers targeting walleye were 0.193 and 0.119 respectively) declined compared to the previous year but was still better than the 1998-2002 time-period. Over 80% of harvested walleye were age-3; from the 2001 year-class. Other species caught included over 70,000 yellow perch and about 24,000 each of largemouth bass and white perch (Tab. 6).

References

Casselman, J.M., Scott, K.A., 1992. Research project: fish community dynamics of the Outlet Basin of Lake Ontario. In: Lake Ontario Fisheries Unit, 1991 Annual Report, Section 18, pp. 18-1 to 18-8.Ontario Ministry of Natural Resources, Picton, Ontario.

Hoyle, J.A., 2004. Eastern Lake Ontario Fish Community Index Netting Protocol, 2004. Lake Ontario Fisheries Unit, Internal Report, LOA 06.04. Ontario Ministry of Natural Resources, Picton, Ontario.

Hurley, D.A., 1992. Research project: fish community studies of the Bay of Quinte. In: Lake Ontario Fisheries Unit, 1991 Annual Report, Section 19, pp. 19-1 to 19-7. Ontario Ministry of Natural Resources, Picton, Ontario.

100

ECOSYSTEM MODELLING: COMPARING THE BAY OF QUINTE AND ONEIDA LAKE

Marten A. Koops, E. Scott Millard, and Charles K. Minns

Great Lakes Laboratory for Fisheries and Aquatic Sciences, Fisheries and Oceans Canada 867 Lakeshore Road, Burlington, Ontario, L7R 4A6

Historic milestones thought to influence the ecosystem dynamics of the Bay of Quinte, Lake Ontario, and Oneida Lake, New York, have a similar progression. Following increased phosphorus loadings and cultural eutrophication of the ecosystems, the Great Lakes Water Quality Agreement led to phosphorus control measures being implemented and reductions in the levels of nutrients in both systems. In general, the result of phosphorus control measures were observed in the primary producers with decreases in phytoplankton abundance and an increase in macrophytes, however, this pattern is more strongly observed in the Bay of Quinte than Oneida Lake. Both ecosystems were invaded by non-native species such as zebra mussels (Dreissena polymorpha) and cormorants (Phalacrocorax auritus). Generally, these species followed an exponential increase in abundance, except for zebra mussels in Oneida Lake which appear in large quantities and stabilize very rapidly. Zebra mussels affect an ecosystem’s physical properties by clearing the water column, and water clarity increased in both ecosystems following invasions by zebra mussels. Increased water clarity led to rapid increases in macrophyte abundance in the Bay of Quinte. These milestones led to the identification of three time periods in the 1972-2002 long term data: a pre-phosphorus control time period, a post-phosphorus control time period, and a post-zebra mussel invasion time period. Though the start and end of each of these time periods differs between the two ecosystems (Tab. 1), these time periods serve as a context for comparison between the Bay of Quinte and Oneida Lake.

101

Table 1. Time periods for the Bay of Quinte and Oneida Lake Ecopath models. Years Model Covers Time Period Bay of Quinte Oneida Lake Pre-Phosphorus Control (Pre-P) 1972 – 1977 1972 – 1985 Post-Phosphorus Control (Post-P) 1978 – 1994 1986 – 1991 Post-Zebra Mussel Invasion (Post-ZM) 1995 – 2002 1992 – 2002

The Bay of Quinte and Oneida Lake appear to have responded to phosphorus controls in a similar manner. Walleye (Sander vitreus) abundances peaked in both ecosystems in the mid to late 1980’s (middle of the post-phosphorus control time period), and then exhibited sustained declines through the 1990’s. Explanations for the walleye decline fall into one of three general categories: habitat changes caused by the invasion of zebra mussels, increased predation mortality from cormorants, and increased fishing mortality. We built three Ecopath models of each ecosystem, one model as a snapshot of each time period, to better understand the ecosystem changes driven by phosphorus control and invasion by zebra mussels. The Ecopath with Ecosim modelling approach is a two-step process (Christensen and Walters, 2004, Christensen et al., 2005). The first step (Ecopath) is the construction of a mass balance model of the ecosystem, the second step (Ecosim) uses the balanced model to run time dynamic simulations that can be used to explore the potential ecosystem impacts of changes and management strategies. Here we report on comparisons of the Ecopath models among time periods and between ecosystems.

Methods

Mass balance in Ecopath is achieved by solving the master equation (Christensen et al., 2005):

(1) Error! Objects cannot be created from editing field codes.

This equation is solved simultaneously for each group in the model. For each group i, Pi is the total production, M2i is the total predation as determined by the consumption of group i

102

by each predator specified in the diet matrix, Yi is the total fishery catch, BAi is the biomass accumulation and Ei is the net migration. M0i represents all other mortality and is calculated from the ‘ecotrophic efficiency’ term specifying the proportion of a group’s production used within the ecosystem, i.e. the proportion of Pi that is accounted for in the model by M2i, Yi, BAi and Ei. Diet composition, biomass accumulation, fishery catches and net migration must be specified for each group. In addition, three of the following four parameters must be specified: production to biomass ratio (P/B), consumption to biomass ratio (Q/B), biomass (B), and ecotrophic efficiency (EE). By only specifying three parameters, one parameter is free to be estimated by the software to bring the model into mass balance. Usually, but not always, the EE parameter is left unspecified as it is often the most difficult to estimate, and since EE can only range between 0 and 1, it provides an easy check for mass-balance. Initial inputs of parameter values will not result in a balanced model. Therefore, it is necessary to adjust parameter values to bring the model into mass-balance. When adjusting parameter values, having estimates of uncertainty about the input values can ensure that these values are not adjusted unrealistically. A balanced model is one subset of the possible combinations of parameter values that can represent the ecosystem. The Ecopath models were manually balanced using an iterative process where input parameters were adjusted based on their estimated uncertainty (Koops et al., In Prep.). Overestimation of P/B ratios are known to build unrealistic resilience into an ecosystem model, so adjustments to estimated P/B ratios were resisted. Diets were generally the most uncertain estimates, so diets were often adjusted first and the final balanced diets for the post-zebra mussel invasion models were validated.

103

Table 2. Species and functional groups used in the Bay of Quinte and Oneida Lake Ecopath models. Category # Groups Groups Birds 1 Cormorants Fish 23 Alewife, carp, benthivores, drum, gizzard shad, gobies1, Lake sturgeon2, panfish (2 stanzas: YOY, 1+), piscivores, planktivores, smallmouth bass, trout-perch, walleye (6 stanzas: YOY, ages 1, 2, 3, 4, 5 and older), white perch (2 stanzas: YOY, 1+), yellow perch (2 stanzas: YOY, 1+) Benthos 6 Amphipods-isopods, bivalves, dreissenids, gastropods, oligochaetes- chironomids, other benthos Zooplankton 5 Cercopagis1, copepods, herbivorous zooplankton, predatory cladocerans, rotifers Primary 4 Epiphyton, macrophytes, periphyton, phytoplankton Producers Detrital 3 DOC, pelagic detritus, sedimented detritus Pools 42 1 Only present in the Bay of Quinte. 2 Only present in Oneida Lake.

Results

A total of 42 species or functional groups were identified between the two ecosystems (Tab. 2). Of these groups, 39 are common to both systems. The two species unique to the Bay of Quinte are aquatic invasive species (AIS) which have invaded since 2000, the predatory cladoceran Cercopagis pengoi and the round goby (Neogobius melanostomus). The species unique to Oneida Lake is lake sturgeon (Acipenser fulvescens), which were historically native to both systems, and were re-introduced in Oneida Lake in 1996. These unique species represent only 0.011% of the total biomass in the Bay of Quinte and 0.017% of the total biomass in Oneida Lake. However, based on other parts of the Great Lakes, it is expected that round gobies will increase substantially in abundance, and the biomass of lake sturgeon has exhibited exponential increases since their re-introduction to Oneida Lake.

104

-20 -20 A -22 C

-24 -24 Trout perch Trout-perch -26 -28 -28

C Signature (‰) -30 13

δ -32 -32 -34 -36 -36 -38 -34 -32 -30 -28 -26 -24 -22 -34 -32 -30 -28 -26 -24 -22 13 Actual δ13C Signature (‰) Actual δ C Signature (‰) 18 19 B Piscivores D 16 Walleye 17

14 15

13 N Signature (‰)10 Predicted 15 δ 11 8

6 9 4 7 Predicted 5 7 9 11 13 15 8 10 12 14 16 18 Actual δ15N Signature (‰) Actual δ15N Signature (‰)

Figure 1. Predicted stable isotope signatures for each fish group from the balanced Ecopath diet matrix versus field-measured stable isotope signatures from the Bay of Quinte (A and B) and Oneida Lake (C and D). Solid lines show regressions and 95% confidence intervals. Dashed lines show 1:1 lines. Groups falling on or outside the confidence intervals are identified.

105

Diet validation

Diet composition was developed on the basis of multiple lines of evidence: literature reports of species trophic interactions, expert opinion, and quantitative examination of species gut contents (Koops et al., In Prep.). Each identified prey was assigned a proportion in the diet, and then these initial values were modified during mass balancing of the models (Koops et al., In Prep.). A diet-based theoretical isotopic signature for each fish group was calculated by multiplying species-specific δ13C and δ 15N isotope values by the proportion specified in the balanced model diet matrix and adjusted for trophic fractionation. The diet- based theoretical isotope signatures were then compared to the isotope signatures measured by Bowman (2005). The majority of Ecopath predicted values fall within the 95% confidence intervals for the field measured values. Of the Ecopath values that do not correspond to the associated field-measured isotope values, 43% are for functional groups (e.g. planktivores) or young- of-the-year groups. Regressions of the Ecopath diet-based theoretical isotope signatures on the field measured isotope signatures exhibited good correspondence (Fig. 1), with only the δ 15N signatures for Oneida Lake exhibiting a slope significantly different from 1 (Fig. 1D, t = 2.49, df = 27, p = 0.019). Rank order correlations of all δ 13C and δ 15N signatures for both 13 ecosystems were significant (Spearman rank correlations, Quinte δ C rs = 0.73, Quinte 15 13 15 δ N rs = 0.92, Oneida, δ C rs = 0.75, Oneida δ N rs = 0.73, all df = 28, all p < 0.05), indicating excellent correspondence between the predicted and measured cardinal ranking of the studied taxa.

Ecopath comparisons among time periods and across ecosystems

The Bay of Quinte and Oneida Lake exhibit similar levels of consumption, total system throughput, production, connectance and system omnivory, based on the balanced Ecopath models (Tab. 3 and 4). Total system throughput, as a measure of the size of the ecosystem in terms of flow, does not differ between the Bay of Quinte and Oneida Lake in the pre- phosphorus control or post-zebra mussel invasion time periods, but does differ during the

106

post-phosphorus control time period, with Oneida Lake exhibiting greater total system throughput. This is indicative of the timing of the increase in total system throughput, with the increase occurring in Oneida Lake between the pre and post phosphorus control time periods, whereas the increase occurred in the Bay of Quinte post-zebra mussel invasion. Phosphorus control had many similar effects on both ecosystems, increasing consumption, decreasing the ratio of total primary production to total biomass, increasing the ratio of total biomass to total throughput, and increasing the total biomass present in the ecosystems (Tab. 3 and 4). Invasion by zebra mussels continued the direction but vastly accelerated these changes, affecting the Bay of Quinte greater than Oneida Lake. Through the pre- phosphorus control and post-phosphorus control time periods, Oneida Lake contained a greater total biomass, but this difference reversed in the post-zebra mussel invasion time period. Most of this difference is due to the higher biomass of macrophytes and zebra mussels in the Bay of Quinte relative to Oneida Lake.

Table 3. Ecosystem statistics for each time period calculated from the balanced Ecopath models for the Bay of Quinte. Parameter Pre-P Post-P Post-ZM Units Sum of all consumption 1305.3 1718.6 3568.4 t·km-2·yr-1 Sum of all exports 3541.1 2949.7 2821.9 t·km-2·yr-1 Sum of all respiratory flows 261.7 336.5 811.9 t·km-2·yr-1 Sum of all flows into detritus 7314.6 6116.3 7173.6 t·km-2·yr-1 Total system throughput 12423 11121 14376 t·km-2·yr-1 Sum of all production 4111 3252 4668 t·km-2·yr-1 Mean trophic level of the catch 3.04 3.28 3.33 Gross efficiency (catch/net p.p.) 0.00034 0.00051 0.00032 Calculated total net primary production 3778 2815 3966 t·km-2·yr-1 Total primary production/total 14.4 8.4 4.5 respiration Net system production 3516.3 2478.6 3154.6 t·km-2·yr-1 Total primary production/total biomass 71.4 51.1 11.0 Total biomass/total throughput 0.004 0.005 0.025 Total biomass (excluding detritus) 52.9 55.1 359.0 t·km-2 Total catches 1.29 1.45 1.26 t·km-2·yr-1 Connectance Index 0.203 0.211 0.206 System Omnivory Index 0.134 0.147 0.114

107

Table 4. Ecosystem statistics for each time period calculated from the balanced Ecopath models for Oneida Lake. Parameter Pre-P Post-P Post-ZM Units Sum of all consumption 1203.3 1343.3 2938.0 t·km-2·yr-1 Sum of all exports 3205.9 3905.0 3288.7 t·km-2·yr-1 Sum of all respiratory flows 321.3 354.5 700.4 t·km-2·yr-1 Sum of all flows into detritus 7076.2 8750.0 8334.7 t·km-2·yr-1 Total system throughput 11807 14353 15262 t·km-2·yr-1 Sum of all production 3727 5128 4917 t·km-2·yr-1 Mean trophic level of the catch 3.51 3.34 3.43 Gross efficiency (catch/net p.p.) 0.000062 0.000044 0.000051 Calculated total net primary production 3449 4816 4369 t·km-2·yr-1 Total primary production/total 10.7 13.6 6.2 respiration Net system production 3127.6 4461.9 3668.3 t·km-2·yr-1 Total primary production/total biomass 49.8 47.7 15.0 Total biomass/total throughput 0.006 0.007 0.019 Total biomass (excluding detritus) 69.3 100.9 290.5 t·km-2 Total catches 0.21 0.21 0.22 t·km-2·yr-1 Connectance Index 0.224 0.202 0.224 System Omnivory Index 0.146 0.13 0.123

100

60 / Total PP (%) PP / Total

20 Phytoplankton P Phytoplankton 0 Pre-P Post-P Post-ZM

Time Period Figure 2. Percentage of primary production attributable to phytoplankton production in the Bay of Quinte (●) and Oneida Lake (■) as calculated from mass balanced Ecopath models for each time period.

108

A greater proportion of the primary productivity originates from pelagic phytoplankton in the Bay of Quinte than in Oneida Lake (Fig. 2), even though the Bay of Quinte contains more macrophyte biomass than Oneida Lake. Phosphorus control and then zebra mussel invasion shifted primary production away from phytoplankton in both ecosystems (Fig. 2), mostly toward macrophytes and epiphytes. While the effect of phosphorus control was similar for the two systems, the effect of zebra mussel invasion was greater in the Bay of Quinte than Oneida Lake.

100

40 % Phytoplankton% 20 Production Consumed

0 Pre-P Post-P Post-ZM Figure 3. Percentage of phytoplankton production which is consumed in the Bay of Quinte (●) and Oneida Lake (■) as calculated from mass balanced Ecopath models for each time period.

Consumption of phytoplankton production increased mildly with phosphorus control in the Bay of Quinte (Fig. 3: 25.4% to 46.9%), and held constant in Oneida Lake (Fig. 3: 32.3% to 32.7%). However, consumption of phytoplankton production increased dramatically after the invasion of zebra mussels, exceeding 90% of phytoplankton production in both ecosystems (Quinte: 99.2%; Oneida: 94.0%). Most of the increased consumption of phytoplankton production that occurred after phosphorus control measures were implemented can be attributed to the shift in primary production away from

109

phytoplankton that occurred in both ecosystems (Fig. 2), however, the further decline in phytoplankton production post zebra mussel invasion can not explain the increase in consumption of phytoplankton production. The same suite of phytoplankton consumers is present in these two ecosystems (Fig. 4), with herbivorous zooplankton, rotifers, and oligochaetes-chironomids ranking as the top three phytoplankton consumers in the pre- phosphorus control and post-phosphorus control time periods in both ecosystems. With the invasion of zebra mussels, an additional phytoplankton consumer is added to both ecosystems, and zebra mussels alone consume greater than 50% of the phytoplankton production in both systems (Fig. 4), explaining the large increases in consumption of phytoplankton production (Fig. 3).

AB 2% 3%

4% 5% 14% 6% 7% 16%

CD3% 3% 1%

6% 9% 29% 4% 7% 17%

2% 2% EF

17% 51% 27% 2% 58% 6% 24%

Figure 4. Phytoplankton consumers, with percentage of phytoplankton production consumed, in the Bay of Quinte (A, C, E) and Oneida Lake (B, D, F) as calculated from mass balanced Ecopath models for the pre-P (A, B), post-P (C, D), and post-ZM (E, F) time periods. Unlabelled consumption represents <1% of phytoplankton production.

110

Phosphorus control had relatively little impact on invertebrate biomass and production in either ecosystem. The relative contribution from benthic versus pelagic sources for invertebrate biomass and production held fairly constant in the Bay of Quinte and shifted very slightly toward pelagic biomass and production in Oneida Lake (Fig. 5). The invasion of zebra mussels drastically increased both the benthic invertebrate biomass and production relative to the pelagic biomass and production in both ecosystems (Fig. 5 A and B). However, when zebra mussel biomass and production is removed from these ratios, the shift from pelagic to benthic invertebrate biomass and production still appears in the Bay of Quinte, but Oneida Lake continues its pre- to post- phosphorus control trend toward increasing pelagic invertebrate biomass and production (Fig. 5 C and D).

60 3.5 A C

3 50

2.5 40

2 30 1.5

20 1 Benthos:Zooplankton B

10 0.5

0 0 Pre-P Post-P Post-ZM Pre-P Post-P Post-ZM

2 0.75 B D

1.5 0.5

0.25

Benthos:Zooplankton P 0.5

0 0 Pre-P Post-P Post-ZM Pre-P Post-P Post-ZM Time Period Time Period Figure 5. Ratio of total benthic invertebrate to zooplankton (A) biomass (t·km-2) and (B) production (yr-1) and the ratio of non-dreissenid benthic invertebrate to zooplankton (C) biomass (t·km-2) and (D) production (yr-1) in the Bay of Quinte (●) and Oneida Lake (■) as calculated from mass balanced Ecopath models for each time period. Pre-P and Post- P data are unchanged (A to C, B to D), but note the change of scale.

111

0.6 A

0.5

0.4

0.3

0.2 Predator:Prey B Fish

0.1

0 Pre-P Post-P Post-ZM

0.3 B 0.25

0.2

0.15

0.1 Predator:Prey P Fish

0.05

0 Pre-P Post-P Post-ZM Time Period Figure 6. Ratio of predator (trophic level ≥ 3.5) to prey (trophic level < 3.5) fish (A) biomass (t·km-2) and (B) production (yr-1) in the Bay of Quinte (●) and Oneida Lake (■) as calculated from mass balanced Ecopath models for each time period.

At the upper trophic levels, a comparison of the biomass or production of predatory (trophic level ≥ 3.5) to prey (trophic level < 3.5) fishes shows a consistent response across the three time periods in Oneida Lake, but not in the Bay of Quinte (Fig. 6). Oneida Lake exhibits a continuous shift from predator to prey fish biomass through phosphorus control

112

to the present where fish biomass at trophic levels less than 3.5 outweighs predatory fish biomass by a factor of 4:1. The Bay of Quinte exhibits a similar end point, but started with a very low predator to prey fish biomass ratio, as identified by Hurley and Christie (1977). Predator to prey fish production exhibit similar patterns (Fig. 6B). Not surprisingly fish production is biased toward lower trophic level fishes in both ecosystems, though there is greater bias post-zebra mussel invasion compared to the pre-invasion, post-phosphorus control time period.

Discussion

The Ecopath models suggest that the Bay of Quinte and Oneida Lake ecosystems are similar. Many of the Ecopath statistics and responses to phosphorus control and zebra mussel invasion are very comparable. Phosphorus control and zebra mussel invasion have affected the Bay of Quinte and Oneida Lake in similar ways. Primary production has shifted from phytoplankton towards macrophytes and epiphytes, partially, but not entirely due to the high consumption of phytoplankton by zebra mussels. Invertebrate (benthic and planktonic) biomass and production has shifted from the pelagic to benthic zones. In the Bay of Quinte, this shift toward benthic production is still present when the influence of zebra mussels is removed. Fish production and biomass has shifted from higher trophic level fishes toward lower trophic level fishes. Overall, both ecosystems have shifted away from pelagic toward benthic portions of the food webs, though the Bay of Quinte remains a more pelagic system than Oneida Lake.

Acknowledgements

The hard work of the Quinte-Oneida project team is appreciated including: Kelly Bowen, Chirstine Brousseau, John Casselman, Jeremy Coleman, Ron Dermott, Jay Dietrich, Dean Fitzgerald, Kristen Holeck, Jim Hoyle, Brian Irwin, Randy Jackson, Ora Johannsson, Kathy Leisti, Chris Mayer, Jennifer MacNeil, Ed Mills, Heather Niblock, Mohi Munawar, Michael Power, Bob Randall, Lars Rudstam, Tom Stewart, and Bin Zhu.

113

References

Bowman, J.E., 2005. The use of stable isotopes to elucidate feeding relationships in key fish species in the Bay of Quinte and Oneida Lake. M.Sc. thesis, Department of Biology, University of Waterloo, Waterloo, ON.

Christensen, V., Walters, C.J., 2004. Ecopath with Ecosim: methods, capabilities and limitations. Ecol. Model. 172, 109-139.

Christensen, V., Walters, C.J., Pauly, D., 2005. Ecopath with Ecosim: a User’s Guide. Fisheries Centre, University of British Columbia, Vancouver. November 2005 edition, 154 p. (available online at www.ecopath.org)

Hurley, D.A., Christie, W.J., 1977. Depreciation of the warmwater fish community in the Bay of Quinte, Lake Ontario. J. Fish. Res. Board Can. 34, 1849-1860.

Koops, M.A., Millard, E.S., Mills, E.L. (Eds.), In Prep. Documentation for mass balance models of the Bay of Quinte (Lake Ontario) and Oneida Lake (New York). Can. Tech. Rep. Fish. Aquat. Sci. 2616.

114

2004 SUBMERGED MACROPHYTE SURVEY

K.E. Leisti

Great Lakes Laboratory for Fisheries and Aquatic Sciences Fisheries and Oceans Canada, 867 Lakeshore Road, P.O. Box 5050 Burlington, Ontario, L7R 4A6

This was the second macrophyte survey to be conducted after the 1994 invasion of zebra mussels (post-zm) into the Bay of Quinte. The 2000 macrophyte survey detected a substantial increase from the post-phosphorus period (1978-1994; post-P) in both cover and SAV bed extent at some locations, particularly in the upper bay (Seifried, 2001). For detailed information on the 2004 macrophyte survey, including historical data from 1972, refer to Leisti et al. (2006).

Sampling and analysis

Sampling occurred over 3 weeks beginning August 16th, September 7th and September 13th, 2004. Sampling was conducted at the same survey locations since 1972 (Fig. 1; Crowder and Bristow, 1986). Sampling consisted of echosounding and point sampling along a total of 30 transects; 3 transects on the north shore and 3 on the southern shores at Trenton, Belleville, Big Bay, Hay Bay and Conway. Using a BioSonics DT4000 with a 430 kHz 6.8° transducer, echosounding was conducted along transects that started near the shore at approximately 1 m water depth and ran perpendicular to the shoreline to a minimum of 100 m beyond the edge of the submerged aquatic vegetation (SAV) bed. Depending on the length of the transect, a minimum of 3 and maximum of 6 points were sampled for SAV along each transect. These locations corresponded to the biomass sampling points of the 2000 macrophyte survey. In 2004, operational restrictions were placed on any type of in-water sampling (e.g. diving), so rake tosses were used to retrieve

115

SAV into the boat (a 17’ McKee), where they were quickly sorted either to genus or species and visual estimates of density (sparse, moderate, dense) and species composition were made. SAV biomass was not measured in the 2004 survey, as it was in 2000, due to the inability to maintain a constant sample area using the rake toss methodology.

Figure 1. Transect sites for the 2004 macrophyte survey in the Bay of Quinte. Also shown are water depths to IGLD85 datum.

The echosounding data was analyzed using BioPlant 1.0 software (BioSonics, 2001) to determine water depth, percent cover and plant height. This data is geo-referenced, enabling mapping of SAV cover using ArcView 3.3. The extent of the SAV bed at each transect was determined by measuring in ArcView, the distance between the start of the transect to the last plant on the transect as established by the BioSonics data. The rake toss data was compiled to determine species richness and the frequency of occurrence of species per transect.

Results and discussion

SAV Cover and bed extent

Maps for each of the locations were produced of percent cover resulting from the analysis of the BioSonics data. Only the Trenton (Fig. 2) and Conway north (Fig. 3) data

116

are shown here for illustration purposes. Maps for the remaining 21 transects from the 2004 survey can be found in Leisti et al. (2006). Included on these maps are the IGLD85 water depths since angiosperms are generally not found beyond the euphotic zone thus basin morphometry has a considerable effect on the ultimate extent of the SAV beds. The SAV bed extent varies considerably at Trenton (Fig. 2), with the Trenton north transects substantially longer than those of Trenton south. Examination of the 3 m depth contour at Trenton reveals that this contour is much farther offshore on the north than the south shore indicating that basin morphometry influences these differences. Basin morphometry is restrictive at Conway (Fig. 3), where there is only a narrow strip of substrate with sufficiently shallow water to support SAV growth prior to a rapid descent into much deeper waters.

Figure 2. 2004 SAV cover for Trenton transects based on the analysis of the BioSonics acoustical data. Water depths are at IGLD85 datum.

117

Figure 3. 2004 SAV cover for Conway north transects based on the analysis of the BioSonics acoustical data. Water depths are at IGLD85 datum.

There are many abiotic and biotic factors that impact SAV growth. Abiotic factors include irradiance and water clarity, sediment composition, slope, exposure, temperature, water chemistry, hydrostatic pressure and water level fluctuation. Biotic factors include periphyton abundance, grazing by invertebrates and waterfowl, uprooting by fishes such as carp, plant diseases and invasion by exotic SAV. Depending on basin morphometry, processes affecting SAV growth associated with the shallow nearshore may include wave action and ice scouring while in the deepest waters, these processes are typically light availability, temperature and hydrostatic pressure. Additionally, SAV growth may be anthropogenically controlled through mechanical, chemical or biological means. Response to these stressors can vary by SAV species due to differences in reproduction, life histories and morphologies. Single or multiple stressors can result in patchiness along transects, as seen at TN5 (Fig. 2). SAV density, as indicted by percent cover and bed extent should be examined concurrently to obtain a clearer understanding of SAV occurrence and establishment. As seen in Fig. 4, although some transects may have high SAV density, their SAV bed extent

118

is insubstantial. When averaging the 3 transects at each location, mean cover in the Bay of Quinte varied from 33 to 64% while the weed bed extent ranged from 70 m to 1.2 km. Trenton north had the most extensive beds (mean = 1.2 km) in the survey with a combined mean cover of 61%, the second highest SAV density. Belleville south ranked second in terms of SAV bed extent (525 m) and fourth in cover (52%), with transect BS5 the major contributor for this location. The transects at Belleville north were third in SAV bed extent (442 m), but had the second lowest mean cover at 42%. Hay Bay east had the highest mean cover of the survey at 64% and is fourth in SAV bed extent at 305 m. The Conway transects experienced some of the highest cover values with means of 61 and 58% for CN and CS respectively, but have the second lowest SAV bed extents at 133 and 93 m. Big Bay had some of the lowest cover values, at 45 and 33% for BBS and BBN respectively, with mean SAV beds ranging from 157 to 274 m.

1500 100 a 1000 67

500 33 Percent SAV Cover SAV Percent SAVBed Extent (m)

0 0 CS1 CS3 CS5 TS1 TS3 TS5 BS1 BS3 BS5 CN1 CN3 CN5 TN1 TN3 TN5 BN1 BN3 BN5 BBS1 BBS3 BBS5 HBE1 HBE3 HBE5 BBN1 BBN3 BBN5 HBW1 HBW3 HBW5 Transect Transect Length to Last Plant Cover

Figure 4. Percent SAV cover and SAV bed extent from the start of the transect to the last plant for all 2004 transects. Error bars on cover represent ± 1 SE.

119

SAV Cover and Bed Extent, 1972 to 2004

In a historical context, SAV beds that increased substantially from the post-P to the post-zm period kept the bulk of their beds intact between 2000 and 2004, although cover values have been reduced slightly (Fig. 5). Macrophyte response to these 2 major perturbations in the Bay of Quinte has varied depending on their location. The upper bay generally had the greatest response in terms of both cover and SAV bed extent, the middle bay showed a moderate response in terms of SAV bed extent, while the lower bay had limited response in both metrics. In the upper bay, SAV cover in the pre-P time stanza was sparse, although the SAV bed extent was larger than that found in the post-P period. This is somewhat misleading because several transects had very sparse cover at or near the end of the SAV bed. If cover values less than 10% near the end of these transects were excluded, the mean SAV bed extent would be reduced by half. After reduction in point source phosphorous loadings at sewage treatment plants, the upper bay experienced a 35% reduction in total phosphorous (Nicholls and Millard, 2005) and a 13% increase in Secchi depths. Cover values increased into the lower portion of the moderate range, but the SAV bed extent decreased in this post-P period. The limited SAV response to the reduction of TP could be due to the reflux of phosphorous from the sediments, which would decline slowly through the accumulation of new sediments (Minns et al., 1986). There was a substantial macrophyte response in the upper bay after the invasion by zebra mussels, both in terms of SAV density and distribution. This type of response post- zebra mussel invasion has been recorded elsewhere (Skubinna et al., 1995). Mean SAV cover along all 6 reference transects increased 1.7 times and the mean SAV bed width tripled between the 1994 and 2000 surveys. The transect at Trenton increased by a factof of 3.5, from 340 m in 1994 to 1.2 km in 2000. As water clarity increased due to zebra mussel filtration, SAV was able to increase in density and also move further offshore, especially in gently sloping sites at some locations that allowed for a rapid expansion into large areas that were only slightly deeper.

120

In the upper bay, in the post-zm time stanza, there was a 16% decrease in cover on average between the years 2000 and 2004. However, single transects can experience high inter-annual variability due to a combination of physical and biotic factors described above. Transect lengths in the upper bay remained relatively constant between the two years. In the pre-P period in the middle bay, cover was already in the moderate category and fluctuated between moderate and dense in the post-P period. There was not a substantial difference in SAV bed extent between the pre and post-P periods although the 1994 SAV bed at 180 m was the largest recorded to that date. Mean SAV density for the 2 reference transects in the middle bay during the post-zm period did not change appreciably when compared to the post-P period, although the SAV bed extent did expand by 130 m. Slight decreases in both SAV density and distribution occurred between 2000 and 2004. Throughout the entire time series, environmental parameters such as Secchi depth and total phosphorous in the lower bay were substantially different than the other two bays (Nicholls and Millard, 2005; Burley and Millard, 2005). During both the pre-P and the post-P period, mean cover in the lower bay fluctuated between moderate and dense with SAV beds never extending more than 145 m offshore. Increased water clarity in the Conway area during the post-zm period has not had a substantial impact on SAV as cover values and SAV bed extents were similar to those found throughout most of the time series. Between 2000 and 2004, the SAV beds expanded slightly at both Conway transects with cover decreasing very slightly. The steep drop off in this area restricts a substantial increase in the SAV beds and the variation in density may be partly due to SAV attempting to establish on steeper slopes with subsequent sloughing in these areas.

121

a) 400 100 Pre-P Post-P Post-ZM 300 67

200

33 100

0 0

b) 400 100 Percent SAVCover

300 67

200

33 100

0 0 SAV Bed Extent (m) Extent Bed SAV

c) 400 100

300 67

200

33 100

0 0 72 74 76 78 80 82 84 86 88 90 92 94 96 98 00 02 04 Year Transect Length to Last Plant SAV Cover

Figure 5. Mean percent SAV cover and SAV bed extent to the last plant for the: a) upper bay, b) middle bay and c) lower bay the years 1972 to 2004. (Note: for all years, this data was compiled from reference transect 3 only. Crowder’s knotted line dataset is used from 1972 to 1988 with cover based on SAV presence/absence at 1 m intervals and SAV bed extent the product of the number of sub-transects with plants by sub- transect length. For the remaining years, cover is based on echogram interpretation and transect length represents the distance between the start of the transect to the last plant, either scaled from the echograms (1994 to 2000) or, in 2004, directly measured in ArcView.) The cover error bars represent ± 1 SE.

122

2004 maximum depth of colonization

The deepest water depth that supported plant growth was determined for each transect, (Fig. 6). The maximum depth was similar for transects in the upper and middle bays, ranging from 3.0 to 4.6 m. SAV growth extended much deeper at Conway, between 5.1 and 7.5 m. Secchi disc depths measured at the point sampling locations were also deeper at Conway, ranging from 3.9 to 5.4 m while the upper and middle bays had Secchi depths between 1.2 and 2.4 m.

4 Depth (m)

0 TS1 TS3 TS5 BS1 BS3 BS5 CS1 CS3 CS5 TN1 TN3 TN5 BN1 BN3 BN5 CN1 CN3 CN5 BBS1 BBS3 BBS5 HBE1 HBE3 HBE5 BBN1 BBN3 BBN5 HBW1 HBW3 HBW5 Transect Max SAV Secchi

Figure 6. Maximum depth of SAV colonization and Secchi depth for all transects in the 2004 survey.

123

a) 100

b) 100

0 Percent Frequency

100 c)

0 BS1 BS3 BS5 CS1 CS3 CS5 TS1 TS3 TS5 BN1 BN3 BN5 CN1 CN3 CN5 TN1 TN3 TN5 BBS1 BBS3 BBS5 BBN1 BBN3 BBN5 HBE1 HBE3 HBE5 HBW1 HBW3 HBW5 Transect

Figure 7. Percent frequency of occurrence by transect for the 3 most frequent species in the 2004 survey: a) Vallisneria americana, b) Heteranthera dubia and c) Myriophyllum spp. (composed of M. spicatum and M. exalbescens)

2004 species frequency and richness

Of the 17 species found in the Bay of Quinte during the 2004 survey, Vallisneria americana, Heteranthera dubia and Myriophyllum spp. (comprised of M. spicatum and M. exalbescens) were most frequent (Fig. 7). V. americana, H. dubia and M. spicatum are all widely distributed and tolerant of turbidity. An introduced species, M spicatum had

124

invaded the Bay of Quinte prior to 1971 (Crowder and Bristow, 1986) and other studies have shown that it expands quickly, peaks in 5 to 10 years, then subsequently declines (Trebitz et al., 1993; Painter and McCabe, 1988). V. americana was found on all transects and was very common (>67%) on 19 of the 30 transects. H. dubia was absent on both short HBW transects, but very common on 57% of the transects, including all those of Big Bay and Conway south, and sparse (<33%) on BS3 and CN1 transects. Myriophyllum spp. was not found at BS3, HBE1 and CN5, but was very common on 30% of the transects. It was sparse on 5 of the 27 transects where it was present, including the remaining Hay Bay east transects, 2 transects in Belleville north, and at the BBN3 transect. Species richness varied from 2 on transects BS3 and HBW5, whose transect lengths were 6 and 19 m respectively, to 11 species on transects BS5 and BBN1 (Fig. 8). Conway had the highest richness, closely followed by Trenton, whereas the least number of species was found in Hay Bay. It should be noted that sampling effort varied across transects with the number of sites sampled generally increasing with the length of the transect. Richness on the three very short transects (< 20 m length) was between 2 and 3 and these sites had between 1 and 3 point samples. Conversely, CS5 recorded 6 species using a single sample point. The next shortest transect was TS1 at 57 m where 7 species were found, so it is unlikely that sample size was an issue, however it is difficult to extricate bed width from sampling effort without more detailed statistical analysis. There was generally little change in the Bay of Quinte between the 2000 and 2004 macrophyte surveys in terms of overall density and extent. The increases in SAV bed extent recorded in 2000 after the invasion by zebra mussels for the upper bay had been maintained into 2004, although the transects saw a 16% decrease in SAV cover. Slight decreases in both SAV density and extent were recorded for the middle bay between 2000 and 2004 while the lower bay experienced a slight increase in SAV bed extent. As in the 2000 survey, the most frequent species in 2004 were V. americana, and H. dubia. Between 2000 and 2004, species richness increased on two, decreased on 2 and remained the same on 1 of the 5 reference transects. For detailed information regarding the 2004 macrophyte survey, including the historical context back to 1972, refer to Leisti et al., 2006.

125

12 8 7 10 6 8 5 6 4

Richness 3 4 2 Sites # Sample 2 1 0 0 TS1 TS3 TS5 BS1 BS3 BS5 CS1 CS3 CS5 TN1 TN3 TN5 BN1 BN3 BN5 CN1 CN3 CN5 BBS1 BBS3 BBS5 HBE1 HBE3 HBE5 BBN1 BBN3 BBN5 HBW1 HBW3 HBW5 Richness # Sample Sites

Figure 8. Species richness by transect for all transects surveyed in 2004. The number of sample points for each transect varied depending on transect length thus sampling effort per transect is included.

Acknowledgements

I am very grateful to both Rachel Nagtegaal and Bud Timmins for all their help in the field. I would also like to thank S.E. Doka, C.K. Minns and C. Bakelaar for their advice.

References

Burley, M., Millard, E.S., 2005. Photosynthesis, Chlorophyll A, and Light Extinction. In: The Project Quinte Annual Report 2003, pp. 27-46. Bay of Quinte RAP Restoration Council, Belleville, Ontario.

126

Crowder, A., Bristow, M., 1986. Aquatic macrophytes in the Bay of Quinte, 1972-82. In: C.K. Minns, D.A. Hurley, K.H. Nicholls (Eds.), Project Quinte: point-source phosphorous control and ecosystem response in the Bay of Quinte, Lake Ontario. Can. Spec. Publ. Fish. Aquat. Sci. 86, 114-127

Leisti, K.E., Millard, E.S., Minns, C.K., 2006. Assessment of Submergent Macrophytes in the Bay of Quinte, Lake Ontario, August 2004, Including Historical Context. Can. MS Rpt. Fish. Aquat. Sci. 2762, ix+81.

Minns, C.K., Owen, G.E., Johnson, M.G., 1986. Nutrient Loads and Budgets in the Bay of Quinte, Lake Ontario, 1965-81. In: C.K. Minns, D.A. Hurley, K.H. Nicholls (Eds.), Project Quinte: point-source phosphorous control and ecosystem response in the Bay of Quinte, Lake Ontario, Can. Spec. Publ. Fish. Aquat. Sci. 86, 59-76

Nicholls, K.H., Millard, E.S., 2005. Nutrients and Phytoplankton. In: The Project Quinte Annual Report 2003, pp. 13-26. Bay of Quinte RAP Restoration Council, Belleville, Ontario.

Painter, D.S., McCabe, K.J., 1988. Investigation into the disappearance of Eurasian water milfoil from the Kwaratha lakes. J. Aquat. Plant Manage. 26, 3-12

Seifried, K.E., 2001. An Overview of Previous Aquatic Macrophyte Surveys and the 2000 Survey Results. In: The Project Quinte Annual Report 1999 – 2000, pp. 43-52. Bay of Quinte RAP Restoration Council, Belleville, Ontario.

Skubinna, J. P., Coon, T. G., Batterson, T.R., 1995. Increased abundance and depth of submersed macrophytes in response to decreased turbidity in Saginaw Bay, Lake Huron. Journal of Great Lakes Research 21, 476-488.

127

Trebitz, A.S., Nichols, S.A., Carpenter, S.R., Lathrop, R.C., 1993. Patterns of Vegetation Change in Lake Wingra Following a Myriophyllum-Spicatum Decline. Aquatic Botany 46, 325-340.

128

EMPIRICAL DETERMINATION OF THE EPIPHYTE:MACROPHYTE RATIO FOR THE BAY OF QUINTE

K.E. Leisti and M. Burley

Great Lakes Laboratory for Fisheries and Aquatic Sciences Fisheries and Oceans Canada 867 Lakeshore Road, P.O. Box 5050 Burlington, Ontario, L7R 4A6

Ecosystem modelling is currently being conducted for the Bay of Quinte in order to evaluate the response of fish populations to ecosystem changes (Koops et al., 2005). As part of the modelling process, separate macrophyte models were developed for the three bays in Quinte by Seifried (2002) to predict the density and distribution of submerged aquatic vegetation (SAV). Relationships were developed between SAV cover and biomass from the 1988, 1994 and 2000 macrophyte survey data to convert cover to biomass for input into the Echopath model (Minns et al., In Prep.). Epiphytes contribute to primary productivity and needed to be accounted for in the primary production portion of the Ecopath model. GLLFAS conducted a literature review (Bernard, 2003) that indicated epiphyte biomass was approximately twice the biomass of its macrophyte host. This ratio was believed to be high, so empirical verification in the Bay of Quinte was recommended. The objective of the epiphyte sampling was to determine the ratio of epiphyte to macrophyte biomass and assess variability of the ratio in terms of SAV species and location. Epiphyte biomass in the Ecopath model could then be estimated if required based on the Quinte data.

129

Sampling and analysis

The literature suggests that temporal variation in epiphyte composition and biomass are related to their physiochemical environment (Morin and Kimball, 1983; Cattaneo and Kalff, 1978). Studies have found that epiphyte biomass is associated with the surface areas of its macrophytic host (Cattaneo and Kalff, 1980; Lalonde and Downing, 1991). Epiphytic abundance and production has also been shown to vary by depth (Cattaneo and Kalff, 1980; Lalonde and Downing, 1991; Sheldon and Boylen 1975; Morin and Kimball, 1983) and exposure (Strand and Weisner, 1996). The sampling design attempted to address potential variability from 3 factors: macrophyte morphology, geographic location and exposure. The four most dominant macrophyte species in terms of biomass in the 2000 Bay of Quinte survey were selected for sampling (Seifried, 2001). These were Vallisneria americana (27%), Ceratophyllum demersum (25%), Heteranthera dubia (22%) and Myriophyllum spp. (17%). Combined, these four species represented 92% of the total biomass sampled. The species also represented different morphologies: V. americana being a broad, ribbon leaf rosette, C. demersum a rootless plant with dissected leaves, H. dubia has narrow, ribbon leaves borne singly on the stem, and Myriophyllum spp., a predominantly upper canopy plant with finely dissected leaves.

Figure 1. 2004 epiphyte sampling locations.

130

There were operational restrictions on in-water sampling, so all sampling had to occur from the McKee launch and was thus limited to the upper water column. Studies suggest that this is generally the depth with the highest biomass of epiphytes (Lalonde and Downing, 1991; Morin and Kimball, 1983). Since a trophic gradient exists within the Bay (Minns et al., 1986), sampling sites were selected in areas with high (near Trenton) and lower (Big Bay area) nutrient loading, Fig. 1. Sample codes, site coordinates and water depth at the sample site are located in Tab. 1. Within these sites, attempts were made to sample at the edge of weed beds or areas with sparse cover and in the middle of weed beds which had dense cover. Epiphyte growth may vary between these two locations due to higher water velocities in sparse cover (Strand and Weisner, 1996) or self shading in dense cover. A series of 3 replicates were scheduled for each species/location/density for a total of 48 samples. Epiphytes were sampled on August 26 – 27, 2004. In the sampling protocol, care was taken to capture both the tightly and loosely adhering epiphytes since Haines et al. (1987) indicated that the loose component accounts for a significant portion of total epiphyte biomass and production. Lake water, filtered through a 20 μm mesh to remove any phytoplankton, was used to fill 2 l jars. The lids were placed on the jars which were inverted in the water near the macrophyte to be sampled. The lid was removed and an approximate 15 cm section of macrophyte and associated epiphytes were gently placed into the jar. The macrophyte stem was then pinched off, the lid replaced and the jar brought onto the boat. Excess water was decanted through a 20 μm mesh and the filter was backwashed with a formalin aceto- alcohol solution (FAA). The FAA consisted of two parts formalin, 10 parts ethanol, 1 part glacial acetic acid and 7 parts water.

131

Table 1. Epiphyte sampling site information. Included is the sample date, location, species, SAV density, site coordinates and water depth. Sample code comprised of site code (BB = Big Bay, T = Trenton) & species code (C = C. demersum, H = H. dubia, M = Myriophyllum spp., and V = V. americana) & density code (D = dense and S = sparse) & replicate number.

Date Sample Species SAV Water (2004) Code Location Code Density Latitude Longitude Depth (m) 26-Aug BB_V_D1 Big Bay V D 44.1353833 77.1899167 2.1 26-Aug BB_V_D2 Big Bay V D 44.1353833 77.1899167 2.1 26-Aug BB_V_S1 Big Bay V S 44.1347667 77.2379833 1.1 26-Aug BB_V_S2 Big Bay V S 44.1347667 77.2379833 1.1 26-Aug BB_V_S3 Big Bay V S 44.1347667 77.2379833 1.1 26-Aug BB_M_D1 Big Bay M D 44.1381167 77.1907667 2.5 26-Aug BB_M_D2 Big Bay M D 44.1381167 77.1907667 2.5 26-Aug BB_M_D3 Big Bay M D 44.1353833 77.1899167 2.1 26-Aug BB_M_S1 Big Bay M S 44.1347667 77.2379833 1.1 26-Aug BB_M_S2 Big Bay M S 44.1347667 77.2379833 1.1 26-Aug BB_M_S3 Big Bay M S 44.1347667 77.2379833 1.1 26-Aug BB_C_D1 Big Bay C D 44.1347667 77.2379833 26-Aug BB_C_D2 Big Bay C D 44.1344 77.18915 1.6 26-Aug BB_C_D3 Big Bay C D 44.1344 77.18915 1.6 26-Aug BB_H_D1 Big Bay H D 44.1381333 77.1907833 2.5 26-Aug BB_H_D2 Big Bay H D 44.1354167 77.1899167 2 26-Aug BB_H_D3 Big Bay H D 44.1354167 77.1899167 26-Aug BB_H_D4 Big Bay H D 44.1354167 77.1899167 1.6 26-Aug BB_H_S1 Big Bay H S 44.1347667 77.2379833 1.1 26-Aug BB_H_S2 Big Bay H S 44.1347667 77.2379833 1.1 26-Aug BB_H_S3 Big Bay H S 44.1347667 77.2379833 1.1 27-Aug T_C_D1 Trenton C D 44.0959833 77.555 3.1 27-Aug T_C_D2 Trenton C D 44.0959833 77.555 3.1 27-Aug T_C_D3 Trenton C D 44.0959833 77.555 3.1 27-Aug T_C_S1 Trenton C S 44.0967167 77.5446 3.5 27-Aug T_C_S2 Trenton C S 44.0967167 77.5446 3.5 27-Aug T_C_S3 Trenton C S 44.0967167 77.5446 3.5

132

Table 1. Cont’d

Sample Species SAV Water Date Code Location Code Density Latitude Longitude Depth (m) 27-Aug T_M_D1 Trenton M D 44.0959833 77.555 3.1 27-Aug T_M_D2 Trenton M D 44.0959833 77.555 3.1 27-Aug T_M_D3 Trenton M D 44.0959833 77.555 3.1 27-Aug T_V_D1 Trenton V D 44.0959833 77.555 3.1 27-Aug T_V_D2 Trenton V D 44.0959833 77.555 3.1 27-Aug T_V_D3 Trenton V D 44.0959833 77.555 3.1 27-Aug T_H_D1 Trenton H D 44.0959833 77.555 3.1 27-Aug T_H_D2 Trenton H D 44.0959833 77.555 3.1 27-Aug T_H_D3 Trenton H D 44.0959833 77.555 3.1 27-Aug T_H_S1 Trenton H S 44.0967167 77.5446 3.5 27-Aug T_H_S2 Trenton H S 44.0967167 77.5446 3.5 27-Aug T_H_S3 Trenton H S 44.0967167 77.5446 3.5

The FAA method (Gough and Woelkerling, 1976) was chosen not only as a chemical stripper to break down the mucilaginous bonds but also as a preservative of the samples as well. This method appeared to be optimal for our research compared to others such as scraping, agitation and microscopy (Aloi, 1990). Scraping is only possible on macrophytes with a tough, durable exterior. Agitation (shaking or stirring) to dislodge the epiphyton may not remove tightly attached epiphytes. Microscopy does allow direct observations of epiphytes but the expense and time required was prohibitive for our study. The samples were stored on ice and then refrigerated at 4 °C in the lab until processed the following week. Field conditions restricted the number of samples that were collected. C. demersum was not located at the Big Bay sparse site and a heavy thunderstorm cut short the sampling at Trenton so Myriophyllum spp. or V. americana were not collected. Laboratory analysis involved separating the ephiphytes from the macrophytes using filters and measuring the wet weight of each component. Upon inspection of the samples it was determined that while stripping was effective, agitation would enhance removal of epiphytic material from the macrophytes. The macrophyte section with the FAA solution

133

was placed in a beaker and any visible invertebrates were removed. Larger sections of macrophytes were cut into small pieces and all material was agitated with a magnetic stirrer between 10 to 60 minutes. In this method, it was assumed that anything remaining on a 500 μm mesh was macrophytic while ephiphytes were subsequently captured on a 20 μm mesh. After agitation the sample was poured and rinsed through a wetted and pre-weighed 500 µm mesh. The supernatant was collected and then poured and rinsed through a wetted and pre-weighed 20 µm mesh. The meshes were then suctioned for 1-5 minutes to remove excess water and reweighed. The epiphytic material was then preserved in FAA and placed in cold storage. Processing of 68% of the samples occurred the week after they were collected in the field. The remaining samples were left until Nov 22nd and at that time, fungal growth was discovered in the samples so unfortunately they had to be discarded.

Results and Discussion

A total of 28 samples were collected and processed with the resulting ratios listed in Tab. 2. Missing samples include all Trenton sparse Myriophyllum spp. and Vallisneria americana and sparse Ceratophyllum demersum from Big Bay. Pooling the data by either SAV species, location and density, the boxplots in Fig. 2 show that Ceratophylum demersum had the lowest median epiphyte to SAV ratio while Vallisneria americana had the highest. By location, Big Bay had a higher ratio than Trenton, but this may be biased due to the incomplete dataset. SAV located within dense beds appear to have higher biomass of epiphytes than found on sparse SAV. Pooling all the data together and taking the median resulted in an epiphyte to macrophyte ratio of 0.16. This was the ratio that was subsequently input into the Bay of Quinte Ecopath model.

134

Table 2. Epiphyte to macrophyte ratio for all 2004 samples analyzed. Site code comprised of location (BB = Big Bay, T = Trenton) + species (C = C. demersum, H = H. dubia, M = Myriophyllum spp., V = V. americana) + plant density (D = dense, S = sparse) + replicate number.

Sample Processing Wet Date Date Plant weight Wet weight Epi/SAV (2004) (2004) Site Code Location Species Density SAV (g) epiphytes (g) Ratio 26-Aug 03-Sep BBCD2 Big Bay C D 9.34 2.26 0.24 26-Aug 03-Sep BBCD3 Big Bay C D 12.34 0.44 0.04 26-Aug 02-Sep BBHD4 Big Bay H D 7.72 0.40 0.05 26-Aug 02-Sep BBHD3 Big Bay H D 2.36 0.75 0.32 26-Aug 03-Sep BBHS1 Big Bay H S 1.48 0.24 0.16 26-Aug 03-Sep BBHS2 Big Bay H S 1.56 0.25 0.16 26-Aug 01-Sep BBMD4 Big Bay M D 5.67 0.44 0.08 26-Aug 02-Sep BBMD3 Big Bay M D 2.65 0.87 0.33 26-Aug 02-Sep BBMD2 Big Bay M D 3.47 1.77 0.51 26-Aug 03-Sep BBMD1 Big Bay M D 4.11 1.29 0.31 26-Aug 02-Sep BBMS1 Big Bay M S 1.34 0.34 0.25 26-Aug 02-Sep BBVD2 Big Bay V D 0.45 0.12 0.27 26-Aug 03-Sep BBVD1 Big Bay V D 0.06 0.03 0.50 26-Aug 03-Sep BBVS3 Big Bay V S 0.50 0.21 0.42 27-Aug 02-Sep TCD2 Trenton C D 18.75 0.14 0.01 27-Aug 02-Sep TCD1 Trenton C D 7.55 0.34 0.05 27-Aug 02-Sep TCS1 Trenton C S 16.49 0.10 0.01 27-Aug 03-Sep TCS2 Trenton C S 8.03 0.74 0.09 27-Aug 03-Sep TCS3 Trenton C S 2.89 0.12 0.04 27-Aug 01-Sep THD1 Trenton H D 1.95 0.15 0.08 27-Aug 03-Sep THD2 Trenton H D 0.28 0.15 0.54 27-Aug 02-Sep THS2 Trenton H S 6.62 0.41 0.06 27-Aug 02-Sep THS1 Trenton H S 5.58 0.17 0.03 27-Aug 01-Sep TMD3 Trenton M D 3.07 0.29 0.09 27-Aug 03-Sep TMD2 Trenton M D 2.42 0.34 0.14 27-Aug 03-Sep TVD2 Trenton V D 0.48 0.10 0.21 27-Aug 03-Sep TVD3 Trenton V D 0.62 0.12 0.19

135

0.6 0.6 a) b) 0.5 0.5

0.4 0.4

0.3 O 0.3

A 0.2 R 0.2

0.1 0.1

0.0 0.0 a at C H M V D Big Bay Trenton ll A LOCATION BOXPLOT_SP

0.6 c) 0.5

0.4

O 0.3

A R 0.2

0.1

0.0 Dense Sparse PLANTDENS

Figure 2. Boxplots of the epiphyte/macrophyte ratio by a) species, b) location, and c) plant density. Species codes are: C = C. demersum, H = H. dubia, M = Myriophyllum spp. and V = V. americana

The Ecopath project has determined primary productivity in the upper Bay of Quinte for the 3 time stanzas (pre-phosphorous control, post phosphorous control and post zebra mussel invasion) (Minns et al., In Prep.). Initially an epiphyte to macrophyte ratio of 4.016 was determined through a literature review (Bernard, 2003). The field evidence outlined above suggests the ratio should be much lower and the impact of the change in ratio has been examined, Fig. 3.

136

20 143 40 82

15 30 -2 -2

10 20

Biomass g m Biomass 5 10 Phytom g Biomass

0 0 PreP PostP Post ZM

Mac EpiLit EpiField PeriH PeriS Phyto

Figure 3. Biomass of the primary producers for the upper Bay of Quinte during the 3 time stanzas. Pre P is prior to phosphorous point source control (1972 to 1978), post p is after implementation of phosphorous control (1979 to 1994) and post zm is following the invasion by zebra mussels (1995 to 2000). Shown are biomass estimates for macrophytes (Mac), epiphytes calculated with the ratio obtained from the literature (EpiLit), epiphytes determined by the ratio obtained from Quinte field data (EpiField), periphyton on hard substrates (PeriH), periphyton on soft substrates (PeriS) and phytoplankton (Phyto).

When using the literature value for the epiphyte/macrophyte ratio, phytoplankton biomass dominates in the pre-p time stanza, accounting for 71% of the total primary producer biomass, with epiphytes contributing 16% and SAV 8%. With the slight increase in SAV cover during the post-p period, the SAV contribution increases by 7% to 15%, the epiphytes almost double to 30% while phytoplankton decreases to 49%. In the post-zm period, the upper bay experienced increases in SAV density in combination with the

137

substantial expansion of some of the macrophyte beds. Epiphytes now dominate, contributing 57% of the total primary producer biomass, with SAV at 33% and phytoplankton reduced to 9%. The scenario is quite different if the empirically derived epiphyte/macrophyte ratio is used to calculate epiphyte biomass. In the pre-p period, phytoplankton once again dominates at 84% with the contribution of SAV at 10%. Epiphytes now contribute the least of all the primary producers, less than 1% while soft periphyton contributes 4%. In the post-p period, phytoplankton contribution remains relatively high at 69% while SAV has increased to 21%. The soft periphyton increases by 3% to 7% with epiphytes contributing 1.7%, slightly higher than the 1.3% for hard periphyton. In the post-zm time stanza, SAV now dominates at 74%, with phytoplankton contributing 20%. Epiphyte biomass now accounts for 5% of the total primary producers biomass while soft periphyton decreases to 1.6%. The epiphyte to macrophyte ratio obtained from the literature was calculated from data derived from temperate freshwater lakes (Bernard, 2003). These data were in several different units, including g chl a m-2 and Bernard used a relationship between macrophyte biomass and surface area developed by Armstrong et. al. (2003) to convert m2 of macrophytes into dry weight biomass. Additionally, assumptions regarding the carbon content of ash free dry weight plant material and the wet:dry weight ratios of epiphytes were made for the final conversion to a wet biomass ratio (Minns et al, In Prep.). The requirement to convert the literature values into standardized units may account for some of the difference between the literature and empirically derived ratios. The empirical ratio may actually be an overestimate for epiphyton due to the limited effort involved in separating the algae from other organic and inorganic material. An epiphyte/macrophyte survey using the same protocols as in Quinte 2004 will be conducted in Hamilton Harbour in August 2006 to determine variability in the ratio when the protocol has been consistent.

138

Acknowledgements

The authors would like to thank M.A. Koops, C.K. Minns and E.S. Millard for their assistance. They are indebted to Rachel Nagtegaal for her able assistance in both the field and laboratory work.

References

Aloi, J.E., 1990. A Critical Review of Recent Freshwater Periphyton Field Methods. Canadian Journal of Fisheries and Aquatic Sciences 47, 656-670.

Armstrong, N., Planas, D., Prepas, E., 2003. Potential for estimating macrophyte surface area from biomass. Aquatic Botany 75, 173-179.

Bernard, A., 2003. Literature Summary of Production to Biomass Ratios of Periphyton and Epiphyton in Temperate Freshwater Systems. GLLFAS internal document. Burlington.

Cattaneo, A., Kalff, J., 1978. Seasonal changes in the epiphyte community of natural and artificial macrophytes in Lake Memphremagog (Que. & Vt.). Hydrobiologia 60, 135- 144.

Cattaneo, A., Kalff, J., 1980. The relative contribution of aquatic macrophytes and their epiphytes to the production of macrophyte beds. Limnology and Oceanography 25, 280-289.

Gough, S.B., Woelkerling, W.J., 1976. On the removal and quantification of algal aufwuchs from macrophyte hosts. Hydrobiologia 48, 203-207.

139

Haines, D.W., Rogers, K.H., Rogers, F.E.J., 1987. Loose and firmly attached epiphyton: their relative contributions to algal and bacterial carbon productivity in a Phragmites marsh. Aquatic Botany 29, 169-176.

Koops, M.A., Millard, S.E., Minns, C.K., 2005. Ecosystem Modelling: Comparing the Bay of Quinte and Oneida Lake. In: Project Quinte Annual Report 2003, pp. 3-9. Bay of Quinte RAP Restoration Council, Belleville.

Lalonde, S., Downing, J.A., 1991. Epiphyton biomass is related to lake trophic status, depth and macrophyte architecture. Canadian Journal of Fisheries and Aquatic Sciences 48, 2285-2291.

Minns, C.K., Owen, G.E., Johnson, M.G., 1986. Nutrient Loads and Budgets in the Bay of Quinte, Lake Ontario, 1965-81. In: C.K. Minns, D.A. Hurley, K.H. Nicholls (Eds.) Project Quinte: point-source phosphorous control and ecosystem response in the Bay of Quinte, Lake Ontario, Can. Spec. Publ. Fish. Aquat. Sci. 86, 59-76.

Minns, C.K., Leisti, K.E., Chu, C., Bernard, A., Bakelaar, C.N. In Prep. Estimates of Macrophyte, Epiphyton, and Periphyton Biomass and Production for the Bay Of Quinte, Lake Ontario. Canadian Manuscript Report of Fisheries and Aquatic Sciences.

Morin, J.O., Kimball, K.D., 1983. Relationship of macrophyte-mediated changes in the water column to periphyton composition and abundance. Freshwater Biol. 13, 403- 414.

Seifried, K.E., 2001. An overview of previous aquatic macrophyte surveys and the 2000 survey results. In: Project Quinte Annual Report 1999-2000. Bay of Quinte RAP Restoration Council, Belleville.

140

Seifried, K.E., 2002. Submerged Macrophytes in the Bay of Quinte: 1972 to 2000. M.Sc. Thesis, University of Toronto, Toronto.

Sheldon, R.B., Boylen, C.W., 1975. Factors affecting the contribution by epiphytic algae to the primary productivity of an oligiotrophic freshwater lake. Applied Microbiology 30, 657-667.

Strand, J.A., Weisner, S.E.B., 1996. Wave exposure related growth of epiphyton: Implications for the distribution of submerged macrophytes in eutrophic lakes. Hydrobiologia 325, 113 – 119.

141

2004 LOWER TROPHIC LEVEL COMPARISON OF NEARSHORE AND OFFSHORE HABITATS IN THE BAY OF QUINTE, LAKE ONTARIO

K.L. Bowen, R. Dermott, O.E. Johannsson, M. Munawar, and M. Burley

Fisheries and Oceans Canada, Canada Centre for Inland Waters, 867 Lakeshore Rd., Burlington, ON, Canada

Introduction

To better understand long-term changes in the Bay of Quite, the ecosystem modeling software Ecopath with Ecosim has been used to create a mass-balance model of the upper bay (Millard and Minns, 2001). This work was undertaken by the Department of Fisheries and Oceans’ Great Lakes Lab for Fisheries and Aquatic Sciences (GLLFAS) in Burlington, Ontario, in conjunction with Ontario Ministry of Natural Resources’ Glenora Fisheries Research Station. However, one limitation of the extensive Belleville data set is that it contains only offshore data. As littoral and wetland habitats cover a large extent of the upper bay, morphometric corrections are needed to extrapolate offshore data to the whole bay. To better refine the Ecopath model, a multi-trophic level comparison of nearshore and offshore habitats was initiated in 2001 (Bowen et al., 2003). This involved sampling lower trophic levels and fish in three habitat types in the Bay of Quinte. These included an offshore area, a littoral low-density macrophyte zone and a littoral high-density macrophyte zone. The 2001 study suggested that the lower food web in the nearshore is quite different than that found at the offshore stations, particularly during mid-summer when littoral macrophyte growth is at its peak. For example, the offshore data probably overestimate the importance of phytoplankton production in the nearshore areas. Zooplankton taxonomic

142

composition was also very different and more diverse in the macrophyte beds relative to offshore areas, although total biomass appeared to remain similar. Offshore benthic invertebrate data dramatically underestimated Dreissena biomass, particularly in rocky areas. Offshore sediments also supported lower numbers of other benthic invertebrates, especially plant-associated species such as amphipods and isopods. Finally, fish abundance appeared to be highly variable among the areas, depending on date. Biomass of bluegills, pumpkinseed and yellow perch may be higher in the littoral weedy areas, particularly in spring and fall. Following the 2001 study, it became apparent that we still lack data from the intermediate depth zones within the upper bay (e.g., depths between 3 and 5 m). This zone represents 41% of the area and 55% of the upper bay volume, and supports sparse to moderate macrophyte growth (Seifried, 2002). It is unclear whether the macrophyte beds in the 3-5 m depth zone should be treated as equivalent to those in the 0-3 m zone, or whether this area should be treated as offshore. Another potential gap in the existing Bay of Quinte data is that lower trophic level sampling has only been conducted during the day. For example, some zooplankton taxa (e.g., Cladocera) may be under-represented by daytime sampling. These taxa may seek refuge from predation during the day within the macrophytes or on the bottom, and move out in the water column at night to graze (Timms and Moss, 1984). To address these questions, a second inshore-offshore survey was conducted in July 2004. This study included an offshore area, a littoral low-density macrophyte zone and a littoral high-density macrophyte zone. To address diurnal variation, the same stations were sampled both during the day and night.

Methods

The inshore-offshore field work was conducted in Big Bay, a large embayment east of Belleville in the Bay of Quinte (Fig. 1). It is bordered by Prince Edward County’s Big Island to the south. Big Bay is approximately 7.5 km long and 5 km wide, with a southwest

143

Figure 1. Location of July 2004 Lower Trophic Level Inshore-Offshore sampling stations in Big Bay, Bay of Quinte.

144

to northeast orientation. Although there are a few rocky shoals, most of the shallow areas in Big Bay are colonized by submerged aquatic macrophytes (Seifried, 2002). A more detailed description of the study area and sampling methods are given in Bowen et al. (2003). The three study areas in 2004 matched those of the 2001 inshore-offshore study. The 2004 locations were as follows:

1) Inshore (about 1.5 m deep) with high densities of macrophytes (HM). 2) Inshore gradient (about 1.5 m to 4.5 m deep) with few macrophytes (LM). 3) Offshore (5 m deep) with no macrophytes.

The three high macrophyte (HM) transects were located in a sheltered embayment at the east end of Big Island (BQ22, BQ23 and BQ2K1). Each transect measured 100 m long and followed the 1.5 m depth contour. Sampling was conducted during the day, and repeated later that same night. An integrated water sample was collected from surface to a depth of 1.25 m at each end and the middle of each transect and pooled together. Subsamples were taken for ciliates, phytoplankton, rotifers and chlorophyll. At each of the three points on the transect, zooplankton samples were taken by pumping a total of 60 l of water from the surface to just off bottom. Zooplankton were collected on a 64 µm net and preserved in buffered 4% formalin. Chlorophyll, ciliates and phytoplankton were preserved or filtered using the same methods as the Bay of Quinte lower trophic level biweekly sampling program (e.g., Burley and Millard, 2005; Bowen and Johannsson, 2005; Munawar and Niblock, 2005). Secchi depth, water temperature and light extinction (LICOR) were also measured at the middle point of each transect during daylight hours. The three low macrophyte (LM) transects (BQ29, BQ31 and BQ32) were placed along the wave-exposed northern shore of Big Island. In 2001, the three transects ran parallel to shore along the 1.5 m depth contour, with three points sampled and pooled along each. In 2004, additional transects were sampled at depths of 2.5, 3.5 and 4.5 m. The bottom type at a depth of 1.5 m or less generally consisted of cobble, gravel and sand. The substrate at depths of 2 m or greater consisted mostly of silt and detritus, and supported low to moderate densities of Valisneria, Myriophyllum and Richardson’s pondweed. Zooplankton,

145

rotifers, ciliates and phytoplankton were sampled from the surface to just off bottom at each of the three points along the transects, and pooled. These samples were collected both day and night as in the high macrophyte area. Chlorophyll, LICOR, Secchi and surface and bottom temperature was also recorded once during daylight hours. The offshore zone in Big Bay ranges between 5 and 6 m deep, with a fairly uniform mud bottom. This area is too shallow and turbulent to become thermally stratified. Once in the morning, afternoon and evening on each of three days, one offshore station (BQ2) was sampled. Integrated water was collected from 4 m to the surface for rotifers, ciliates and phytoplankton. Zooplankton were also screened from a 60 l water sample pumped from 4 m to the surface. In the afternoon only, integrated water was also taken for chlorophyll. LICOR, Secchi and surface and bottom temperature was also recorded once each day.

Results

To date, only the chlorophyll samples have been analyzed. Budget constraints have not yet allowed for the analyses of zooplankton, rotifer, ciliate or phytoplankton samples. Furthermore, high algal biomass in the zooplankton samples will make counting difficult. However, all samples have been archived and are available for analyses if resources permit.

References

Bowen, K., Millard, E.S., Dermott, R., Johannsson, O., Munawar, M., 2003. Lower Trophic Level Comparison of Nearshore and Offshore Habitats in the Bay of Quinte, Lake Ontario. In: Monitoring Report #12. Project Quinte Annual Report 2001, pp. 79-109. Bay of Quinte Remedial Action Plan, Kingston, Ontario.

Bowen, K.L., Johannsson, O.E., 2005. Zooplankton in the Bay of Quinte. In: Monitoring Report #14. Project Quinte Annual Report 2003, pp. 47-68. Bay of Quinte Remedial Action Plan, Kingston, Ontario.

146

Burley, M., Millard, E.S., 2005. Photosynthesis, Chlorophyll a and Light Extinction. In: Monitoring Report #14. Project Quinte Annual Report 2003, pp. 27-46. Bay of Quinte Remedial Action Plan, Kingston, Ontario.

Millard, S., Minns, C.K., 2001. Overview of the development of a Bay of Quinte ecosystem model using Ecopath with Ecosim. In: The Project Quinte Annual Report 1999-2000, pp. 8-14. Bay of Quinte RAP Restoration Council, Belleville, Ontario.

Munawar, M., Niblock, H., 2005. Assessing and Comparing the Structure of the Planktonic Food Web at a Stressed Station in the Bay of Quinte, 2002 – 2003. In: Monitoring Report #14. Project Quinte Annual Report 2003, pp. 117-130. Bay of Quinte Remedial Action Plan, Kingston, Ontario.

Seifried, K., 2002. Submerged macrophytes in the Bay of Quinte: 1972 to 2000. M.Sc. Thesis, Department of Zoology, University of Toronto, Toronto, Ontario.

Timms, R.M., Moss, B., 1984. Prevention of growth of potentially dense phytoplankton populations by zooplankton grazing, in the presence of zooplanktivorous fish in a shallow wetland ecosystem. Limnol. Oceangr. 29, 472-486.

147

Editorial Committee: J. Lorimer H. Niblock M. Munawar