MOVEMENT PATTERNS AND GROWTH OF AMERICAN EELS (ANGUILLA ROSTRATA) BETWEEN SALT AND FRESH WATER, BASED ON OTOLITH MICROCHEMISTRY

by

Heather M. Lamson

Bachelor of Science, University of Northern British Columbia, 2000

A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF

Master of Science

In the Graduate Academic Unit of Biology

Supervisors: D.K. Cairns, Ph.D., Department of Fisheries and Oceans, (Charlottetown, PEI) R.A. Curry, Ph.D., Canadian Rivers Institute, UNB (Fredericton)

Examining Board: S.C. Courtenay, Ph.D., Biology UNB (Fredericton), – Internal Examiner G.A. Benoy, Ph.D., Environment Canada – External Examiner

This thesis is accepted by the Dean of Graduate Studies

THE UNIVERSITY OF NEW BRUNSWICK November, 2005 © Heather M. Lamson, 2005 DEDICATION

For all those people who can put the sliminess aside and appreciate eels for the amazing creatures that they are.

ii ABSTRACT

Reconstructing lifelong movements of teleost fishes between salt and fresh water is possible through otolith strontium:calcium microchemistry. This method was employed to trace movements of American eels (Anguilla rostrata) captured in saltwater bays and adjoining freshwater ponds in Prince Edward Island, Canada. Three migratory contingents were identified: freshwater residents, inter-habitat shifters, and saltwater residents. A pond with a pool-and-weir salmonid fishway and drained by a low-gradient channel contained eels that had entered freshwater at all ages. Another pond with a 2.2 m vertical spillway contained only eels that had entered freshwater upon initial continental arrival. Saltwater residents dominated saltwater bays, which challenges the conventional paradigm of obligate catadromy for American eels. Growth rates increased with the amount of time eels spent in salt water. Annual growth of yellow eels that resided in salt water throughout their lives (94.0 mm yr-1) was double that of freshwater residents (45.2 mm yr-1). Eels that shifted between salinity zones had intermediate growth rates (60.5 mm yr-1). Population models estimating growth and escapement of silver eels from salt, brackish, and freshwater habitats need to account for divergent growth rates based on habitat salinity and realize that previously reported growth rates that have not sampled eels in salt water may be underestimated.

iii ACKNOWLEDGMENTS

Thanks to David Cairns, my supervisor, who provided the opportunity to study this fascinating species and for his guidance throughout the project. Jay Shiao, W.N. Tzeng and Yoshiyuki Iizuka were all instrumental by performing the otolith microchemical analysis. Thanks to my co-supervisor Allen Curry and supervisory committee member

Tillmann Benfey for their advice and direction. Corey Muttart, Valérie Tremblay,

Robbie Moore, Noella McDonald and Mark Grimmett provided greatly appreciated field and lab assistance. Fisheries and Oceans Canada in Charlottetown supported the project financially and through the use of equipment and office space. Thanks to Brian Jessop for guidance with growth analysis and to Jay Shiao for his assistance with ageing and analysis. This study received support from the National Science Council, ROC (NSC

91-2313-B-002-291 and 92-2313-B-002-057). I am grateful to my parents and family for all their support. Finally, I would like to thank my best friend and loyal field partner even though he is afraid of eels, my black lab, Angus.

iv TABLE OF CONTENTS

DEDICATION ...... ii

ABSTRACT...... iii

ACKNOWLEDGMENTS ...... iv

TABLE OF CONTENTS...... v

LIST OF TABLES ...... vii

LIST OF FIGURES ...... viii

1 GENERAL INTRODUCTION...... 1

1.1 Background ...... 1

1.2 Outline of Thesis...... 5

1.3 Statement Regarding Contribution of Co-Authored Articles...... 6

1.4 Literature Cited ...... 6

2 MOVEMENT PATTERNS OF AMERICAN EELS (ANGUILLA

ROSTRATA) BETWEEN SALT AND FRESH WATER IN A COASTAL

WATERSHED BASED ON OTOLITH MICROCHEMISTRY...... 10

2.1 Abstract ...... 10

2.2 Introduction...... 11

2.3 Methods...... 13

2.4 Results...... 16

2.5 Discussion ...... 18

2.5.1 Effect of obstacle type on upstream movement ...... 18

2.5.2 Movement patterns...... 19

2.5.3 Implications for conservation...... 21

v 2.6 Acknowledgements...... 22

2.7 Literature Cited ...... 23

3 GROWTH RATES OF AMERICAN EELS (ANGUILLA

ROSTRATA) IN RELATION TO HABITAT SALINITY BASED ON OTOLITH

MICROCHEMISTRY...... 39

3.1 Abstract ...... 39

3.2 Introduction...... 40

3.3 Methods...... 42

3.3.1 Growth Analysis ...... 44

3.4 Results...... 46

3.5 Discussion ...... 48

3.5.1 Length-at-age Analysis ...... 48

3.5.2 Back-calculation...... 49

3.5.3 Implications for Conservation...... 53

3.6 Acknowledgements...... 54

3.7 Literature Cited ...... 54

4 GENERAL DISCUSSION ...... 68

4.1 General Conclusions ...... 68

4.2 Summary of Findings...... 68

4.3 Management Implications and Suggested Research Needs...... 69

4.4 Literature Cited ...... 70

vi LIST OF TABLES

Table 2-1. Catch and catch rates of American eels in fyke nets in the Brackley- Covehead system...... 29 Table 2-2. Salinity history and rate of inter-habitat shifting of American eels captured in saltwater and freshwater sites. See text for definitions...... 30 Table 2-3. Habitat occupancy patterns of wild (non-stocked) yellow and silver Anguilla eels after arrival in continental waters, as inferred from otolith Sr:Ca ratios...... 31 Table 3-1. Mean annual growth (mm/year) of American eels under 15 years old, estimated by otolith radius back-calculation. SW= saltwater residents, IH= inter-habitat shifters and FW= freshwater residents...... 62

vii LIST OF FIGURES

Figure 2-1. Brackley and Covehead Bays, Prince Edward Island (a) and associated freshwater impoundments: McCallums Pond (b) and Cass and Marshalls Ponds (c)...... 32 Figure 2-2. Schematic profiles of pond outlets, showing the vertical spillway at McCallums Pond (a), the concrete salmonid fishway at Cass Pond (b) and the low-gradient channel at Marshalls Pond (c)...... 33 Figure 2-3. Mean within-year Sr:Ca ratios vs. age for American eels from five sites. For inter-habitat shifters (eels that moved between habitat types at least once after age 1), lines represent individual eels. For eels that remained resident in either salt or fresh water after age 1, symbols represent means, with “N” indicating the number of eels...... 34 Figure 2-4. Frequency distribution of Sr:Ca ratios measured in otoliths of age 1+ American eels captured in saltwater (a) and freshwater (b) sites...... 35 Figure 2-5. Rate of movement between salt and fresh water of eels sampled in Cass Pond which shifted between habitat types. Sample size (number of eel- years) is given at the top of the graph...... 36 Figure 2-6. Allocation of time in habitat categories during each year, by eels which shifted habitat at least once after age 1. Samples are from Covehead Bay (a), Cass Pond (b) and Marshalls Pond (c)...... 37 Figure 2-7. Percentage of sampled eels which showed movements out of the habitat type where they were captured, based on Sr:Ca analysis. The habitat type given in each panel is the habitat where captured. All species includes A. marmorata, A. australis and A. dieffenbachii. See Table 2-3 for data sources...... 38 Figure 3-1. Brackley and Covehead Bays, Prince Edward Island and associated freshwater impoundments: McCallums Pond and Cass and Marshalls Pond...... 63 Figure 3-2. a) Length-at-age of eels captured in saltwater Brackley and Covehead Bays and freshwater McCallums, Cass, Marshalls and Parsons Ponds and b) saltwater residents, freshwater residents, and freshwater residents under 15 years of age fitted with linear regression lines using an independent variable model...... 64 Figure 3-3. Relationship between total length (mm) and maximum otolith radius (mm)...... 65 Figure 3-4. Back-calculated lengths-at-age of saltwater residents, freshwater residents and inter-habitat shifters (mean +1 standard error)...... 66 Figure 3-5. a) Mean annual growth rate vs. time spent in salt water and b) Growth per year vs. percent of that year spent in salt water (F=150.931 p<.0001) of American eels under 15 years old in the Brackley-Covehead Watershed. . 67

viii

1 GENERAL INTRODUCTION

1.1 Background

The American eel (Anguilla rostrata) is a member of the family Anguillidae, which is comprised of eel species that follow a similar life cycle. Anguillid eels spawn in the open ocean and eggs hatch into a larval form called leptocephali. The leptocephali metamorphosize into the glass eel stage while drifting with currents towards continental waters. Glass eels typically begin to show pigmentation as they arrive in continental waters where they are called elvers. Once fully pigmented, they are termed yellow eels and will spend the majority of their lives feeding and growing across a range of habitats.

Upon maturity, a silvering of the skin prepares the “silver eel” for the oceanic migration back to their spawning area (Tesch 2003).

The catadromous life history of spawning in the ocean and rearing in fresh water is commonly assigned to anguillid eels which are often termed “freshwater eels.”

However, catadromy is a misleading descriptor for this genus because growth is not restricted to fresh water. Residency in fresh water is not obligatory for European eels,

A. anguilla (Limburg et al 2003), Japanese eels (A. japonica) (Tsukamoto and Arai

2001), short finned eels (A. australis) (Arai et al. 2004) or American eels (A. rostrata)

(Morrisson et al. 2003). Various patterns of movement across salinity gradients exist for yellow eels and include freshwater or saltwater residence and movements between salt and fresh water one or more times.

1

Various methods have been used to track yellow eel movements between salinity zones, including tagging (Oliveira 1997) and telemetry (Helfman et al. 1984). These methods, however, are limited in their ability to track individuals over long periods of time (Secor et al. 1995). The recent introduction of Sr:Ca otolith microchemistry methods have advanced anguillid eel research with the ability to determine lifetime movements across salinity boundaries.

Fish otoliths (ear stones) are composed primarily of calcite (CaCO3), which forms annual rings much like those of a tree due to slower growth during winter than summer.

During otolith formation, strontium (Sr) can substitute for calcium (Ca) due to similar ionic charges and radii. The ratio of Sr to Ca is approximately 100 times greater in salt water than in fresh. When a fish resides in salt water, this higher ratio is reflected in the deposition of Sr in its otoliths (Casselman 1982). Sr:Ca otolith microchemistry, by taking incremental measurements of Sr and Ca along a transect that dissects the annual rings of the otolith, has the power to uncover the past salinity history of a fish.

The American eel is one of two anguillid species that occur in the Atlantic Ocean (along with the European eel), and is distributed along the Atlantic coast during its yellow eel stage from Greenland to northern South America. There is much variability in traits across the species’ range though, based on our present knowledge, there is no genetic difference among the young arriving on continental shores (Avise et al. 1986, 1990).

2 American eel movements between salt and fresh water have been characterized by Sr:Ca otolith microchemistry in three previous studies (Jessop et al. 2002, Morrisson et al.

2003, Cairns et al. 2004) These studies all sampled eels in fresh or brackish water and determined that movement between fresh and brackish water is highly variable during the yellow eel life stage.

American eel abundances have declined since the mid 1980’s (Jessop 2000). Possible causes for the reduction in numbers include adverse oceanic conditions, habitat loss, excessive fishing pressure and obstruction to eel passage upstream. Dams and other obstacles can prevent or impede migration and adversely affect eel populations (Legault

1988, Feunteun et al. 2003). Eels are relatively weak swimmers and although they have been observed to move through grass and perform other migrational feats, obstacles generally impede movements. Previous results from Sr/Ca analysis on eels in a watershed with an impounded stream in eastern Prince Edward Island indicate that dams impede normal movements between fresh and salt water (Cairns et al. 2004).

The first objective of this study was to reconstruct movement patterns of American eels between salt and freshwater habitats in a small watershed in Prince Edward Island,

Canada, based on samples collected in full strength saltwater bays and freshwater ponds.

As previous Sr:Ca studies on American eel movement and growth have only sampled eels from freshwater and brackish water sites, this was the first Sr:Ca study to examine

American eels captured in salt water. Secondly, this study examined the effect of various types of impoundment structures on eel movement between fresh and salt water.

3

The American eel is a long-lived species, spending 4 to 43 years in continental waters during the yellow eel life stage. Residency time to migration is closely linked to growth rate, as maturation and the silvering process depend on size, not age (De Leo and Gatto

1995). Most studies of eel growth in North America have looked at eels in fresh water, and have shown slow growth. Studies that have compared freshwater and brackish or saltwater growth (Morrison et al. 2003, Jessop et al. 2004, Cairns et al. 2004) have shown that growth rates of American eels that spend most of their time in fresh water are lower than those captured in waters of higher salinity. These findings indicate that data on eel growth may be skewed to show overall slower growth rates due to the preponderance of growth studies in fresh water. As growth rates are critical for population assessments, it is imperative that salt water growth data be included in the biological framework. Furthermore, it is important to include growth in saltwater because most fisheries in eastern North America target eels in salt or brackish bays and estuaries.

The third objective of this study was to relate growth rates to movement patterns between salt and fresh water and determine if growth is faster in fresh or saltwater habitats. I hypothesized that growth increases with the relative time spent in saltwater habitats. To test this hypothesis I related back-calculated growth at each age to the salinity the eel resided in as determined from Sr:Ca analysis. This was the first study to determine growth rates in relation to life history movement patterns of American eels captured in both saltwater and freshwater sites.

4

1.2 Outline of Thesis

Chapter 1 includes a general introduction, and outline of the thesis.

Chapter 2 examines the movement patterns of American eels between salt and fresh water. Otolith microchemistry was used to reconstruct the continental movement patterns of eels in a small watershed in Prince Edward Island consisting of adjoining saltwater bays, and inflowing watercourses with obstructions of various types at head of tide. The number of times eels shift between salt and fresh water was examined and related to the site of capture. The paper examines the colonization of upstream habitats in a small watershed and the implications of various types of obstruction to upstream movement.

This paper, titled “Movement patterns of American eels (Anguilla rostrata) between fresh and saltwater in a small coastal watershed, based on otolith microchemistry” was submitted to Marine Biology in July 2005.

Chapter 3 determines and compares growth rates of eels that inhabit fresh and salt water.

Growth was compared between eels that resided in fresh or salt water or moved between the habitats as established from otolith microchemistry. The paper proposes that growth rates of American eels are a function of the percent of time spent in either salt or fresh water, and will be higher during years spent in salt water than in fresh water. Back- calculated lengths were used to determine growth rates.

5

The paper titled “Growth rates of American eels (Anguilla rostrata) in relation to habitat salinity, based on otolith microchemistry” will be submitted to Marine

Ecology Progress Series.

Chapter 4 summarizes the major findings of the study and how this research is useful for conserving the species. Furthermore, it provides recommendations for future work in this field.

1.3 Statement Regarding Contribution of Co-Authored Articles

Heather Lamson was responsible for the logistical aspects of the research, data analysis and manuscript preparation. David K. Cairns (the supervisor) was responsible for securing funding, and providing advice on data interpretation and for overseeing preparation of article manuscripts. Jay C. Shiao, Yoshiyuki Iizuka, and Wann-Nian

Tzeng analyzed otoliths for Sr:Ca ratios and also provided input into the analysis and article manuscripts.

1.4 Literature Cited

Arai, T., Kotake, A., Lokman, P.M., Miller, M.J. and Tsukamoto K. 2004. Evidence of different habitat use by New Zealand freshwater eels Anguilla australis and A dieffenbachii, as revealed by otolith microchemistry. Mar Ecol Prog Ser 266:213 225.

6 Avise, J.C., Helfman, G.S., Saunders, N.C. and Hales, L.S. 1986. Mitochondrial DNA differentiation in North Atlantic eels: population genetics consequences of an unusual life history pattern. Proceedings of the National Academy of Sciences, USA 83: 4350-

4354.

Avise, J. C., Nelson, W.S., Arnold, J., Koehn, R.K., Williams, G.C. and Thorsteinsson,

V. 1990. The evolutionary genetic status of Icelandic eels. Evol. 44: 1254-1262.

Cairns, D.K., Shiao, J.C., Iizuka, Y., Tzeng, W.N., and MacPherson, C.D. 2004.

Movement patterns of American eels in an impounded watercourse, as indicated by otolith microchemistry. North Amer. J. Fish. Manage. 24: 452-458.

Casselman, J.M. 1982. Chemical analyses of the optically different zones in eel otoliths. Proc. 1980 N. Am. Eel Conf., pp. 74-82.

De Leo, G.A., and M. Gatto. 1995. A size and age structured model of the European eel (Anguilla anguilla). Can. J. Fish. Aquat. Sci. 52:1351 1367.

Feunteun, E., Laffaille, P., Robinet, T., Briand, C., Baisez, A., Olivier, J.-M., and Acou,

A. 2003. A review of upstream migration and movements in inland waters by anguillid eels: towards a general theory. In Eel biology. Edited by K. Aida, K. Tsukamoto, and

K. Yamauchi. Springer, Tokyo. pp. 191 213.

7 Helfman, G.S., Bozeman, E.L., and Brothers, E.B. 1984. Comparison of American eel growth rates from tag returns and length-age analyses. US Fish. Bull. 82:519–522.

Jessop, B.M. 2000. Estimates of population size and instream mortality rate of

American eel elvers in a Nova Scotia River. Trans. Am. Fish. Soc. 129:514-526.

Jessop, B.M., Shiao, J.C., Iizuka, Y., and Tzeng, W.N. 2002. Migratory behaviour and habitat use by American eels Anguilla rostrata as revealed by otolith microchemistry.

Mar. Ecol. Prog. Ser. 233:217-229.

Jessop, B.M., Shiao, J.C., Iizuka, Y, and Tzeng, W.N. 2004. Variation in the annual growth, by sex and migration history, of silver American eels Anguilla rostrata. Mar.

Ecol. Prog. Ser. 272:231-244.

Legault, A. 1988. Le franchissement des barrages par l'escalade de l'anguille: étude en

Sèvre Niortaise. Bull. Fr. Pêche. Piscic. 308:1 10.

Limburg, K.E., Wickstrom, H., Svedang, H., Elfman, M., and Kristiansson, P. 2003.

Do stocked freshwater eels migrate? Evidence from the Baltic suggests "yes." Amer.

Fish. Soc. Symp. 33: 275-284.

8 Morrison, W.E., Secor, D.H., and Piccoli, P.M. 2003. Estuarine habitat use by Hudson

River American eels as determined by otolith strontium:calcium ratios. Amer. Fish.

Soc. Symp. 33: 87-99.

Oliveira, K. 1997. Movements and growth rates of yellow-phase American eels in the

Annaquatucket River, Rhode Island. Trans. Amer. Fish. Soc. 126: 638-646.

Secor, D.H., Henderson-Arzapalo, A., and Piccoli, P.M. 1995. Can otolith microchemistry chart patterns of migration and habitat utilization in anadromous fishes?

J. Exp. Mar. Biol. Ecol. 192: 15-33.

Tesch, F.W. 2003. The eel. Blackwell Science, Oxford.

Tsukamoto, K., and Arai, T. 2001. Facultative catadromy of the eel Anguilla japonica between freshwater and seawater habitats. Mar. Ecol. Prog. Ser. 220: 265-276.

9 2 MOVEMENT PATTERNS OF AMERICAN EELS (ANGUILLA

ROSTRATA) BETWEEN SALT AND FRESH WATER IN A

COASTAL WATERSHED BASED ON OTOLITH

MICROCHEMISTRY

[Manuscript submitted for publication to Marine Biology July 2005)

2.1 Abstract

Otolith strontium:calcium ratios were used to trace lifetime movements of American eels (Anguilla rostrata) captured in saltwater bays and adjoining freshwater ponds in

Prince Edward Island, Canada. A pond with a pool-and-weir salmonid fishway and a pond drained by a low-gradient channel contained eels that had entered fresh water at all ages, but a pond with a 2.2 m vertical spillway contained only eels that had entered fresh water in the elver year. Saltwater residents were the dominant migratory contingent in saltwater bays (85% of 39), which overturns the paradigm of obligate catadromy for this species. Freshwater residency was the sole pattern found in the pond with the vertical spillway (100% of 12) and the majority contingent in the pond with the low-gradient channel (54% of 24). Inter-habitat shifting was the dominant migratory contingent in eels sampled from the pond with the pool-and-weir fishway (85% of 20). Resident eels were established in salt and freshwater habitats by the year after their arrival in continental waters. Eels that shifted between habitats increased their rate of inter-habitat

10 shifting with age. The high degree of plasticity in habitat use found in this study is consistent with worldwide Anguillid patterns as revealed by Sr:Ca.

2.2 Introduction

Anguillid eels are viewed as textbook catadromous species, spawning in the open sea, migrating to fresh water to rear and returning to the ocean to complete their life cycle.

There is, however, great variability in continental habitat use, and anguillid eels in this phase may be found in habitats ranging from full strength salt water to fresh water

(Daverat et al. 2004; Morrison et al. 2003).

Analysis of strontium:calcium ratios of eel otoliths has greatly expanded our knowledge of eel movement patterns, showing that some eels settle and remain in salt, brackish or fresh water during their continental lives, while others shift among these habitats

(Morrison et al. 2003; Tzeng et al. 1997, 2000; Tsukamoto and Arai 2001). In European

(Anguilla anguilla) and Japanese (A. japonica) eels, Sr:Ca data have revealed that some eels spend their entire lives in salt water (Tsukamoto et al. 1998; Arai et al. 2003).

These species can thus no longer be regarded as obligate catadromous fish, but instead use catadromy as a facultative life history option.

Sr:Ca studies on American eels (A. rostrata) at three locations have shown a variety of inter-habitat movement patterns. However, none of these investigations has sampled eels in full-strength salt water. Yellow American eels are commonly found in salt water,

11 but it has not been demonstrated that American eels can complete their life cycles in the marine environment.

Secor (1999) noted that estuarine and diadromous fishes commonly exhibit varying movement patterns among salt, brackish and fresh water. He termed these groups

"migratory contingents." Tosi et al. (1988) and Édeline and Élie (2004) found that glass eels sampled from marine waters showed distinct salinity preferences in laboratory trials, some choosing salt water while others preferred fresh. This implies that membership in migratory contingents may be determined before arrival at the coast.

Eels which attempt to ascend rivers may encounter natural or artificial obstacles. Eels under 10 cm long are able to creep up wet vertical surfaces (Legault 1988). Hence dams or waterfalls where water trickles down vertical walls may impose an age-dependent barrier to migration, with upstream movement possible only for young eels (Cairns et al.

2004).

We used the Sr:Ca technique to investigate the ontogeny, frequency and directions of

American eel movements between salt and fresh water in a small watershed in eastern

Canada. To permit the examination of effects of obstacle type on movement patterns, we chose a study area consisting of saltwater bays and adjacent freshwater ponds which were formed by dams of three different types: earthen dam with a vertical spillway, concrete dam with a salmonid fish ladder and earthen dam with a low-gradient outlet channel. Patterns of movement between salinity zones, as indicated by Sr:Ca profiles, were used to test the following predictions:

12 1. Ponds formed by dams with vertical water drops (salmonid fish ladder, vertical spillway) will contain only eels that entered at a small size, but the pond drained by a low-gradient channel will contain eels that entered at all sizes.

2. Eels will show a variety of migratory contingents, including residence in salt water, residence in fresh water, and movements between these habitats.

3. Some eels will show saltwater residence only, indicating an exclusively marine life cycle.

4. Eels choose their migratory contingents upon arrival in continental waters, so that colonization of salt water and of adjacent freshwater ponds occurs simultaneously.

2.3 Methods

This study was conducted on the north shore of Prince Edward Island in Brackley and

Covehead Bays and associated ponds (Fig. 2-1). Both bays have full-strength salt water

(>28 ppt) and their combined watersheds total 81 km2. Four streams entering these bays are blocked by dams at head of tide, forming freshwater impoundments. Water exits

McCallums Pond to Brackley Bay by falling vertically 2.2 m from a wooden spillway set in an earthen dam (Fig. 2-2). Cass Pond on Covehead Bay has a 5-chamber pool- and-weir salmonid fishway through which water drops 0.9 vertical m over a horizontal distance of 12.2 m. Chambers are 1.8 m wide, 1.8-4.3 m long and 0.7-1.3 m deep.

Water also leaves Cass Pond over a vertical concrete spillway 5 m wide. Marshalls

Pond drains into Covehead Bay by a low-gradient channel with a rocky bottom that falls

5.0 vertical m over a horizontal distance of 303 m (1.7% slope). Parsons Pond connects to Covehead Bay via a culvert with wooden baffles on its floor that are intended to aid

13 fish movement. The sole unimpounded stream in the Brackley-Covehead system runs through the bed of the former McMillans Pond, whose dam washed out in 1996.

Eels were fished by fyke net in Brackley and Covehead Bays, associated ponds and the stream running through McMillans pond bed in May-November 2003. Eels were anaesthetized with clove oil, measured for total length, weighed and frozen until the otoliths were removed.

One otolith per eel was subjected to microchemical analysis. Otoliths were embedded in epofix resin, then ground and polished until the primordium was exposed. The otoliths were then carbon coated under a high vacuum evaporator prior to analysis with an electron probe microanalyzer (EPMA). Using a JEOL JXA-8900R system equipped with wavelength dispersive X-ray spectrometers, Sr and Ca concentrations were measured at 10 μm intervals from the primordium to the otolith edge. Beam conditions were an acceleration voltage of 15 kV, a current of 3nA and a 5 X 4 μm rectangular scanning beam. A synthetic aragonite (CaCO3) and strontiantite (SrCO3 NMNH

R10065) were used for standard calibrations. Sr concentrations were measured for 80 s at Sr Lα peak positions and 20 s at both the lower and upper sides of the baseline. Ca was measured for 20 s at the Ca Kα peak and for 10s at both sides of the baseline. After

Sr:Ca ratio analysis the otolith was polished to remove the carbon layer and etched for 1 to 2 min with 5% EDTA to reveal annular rings for age determination.

14 Sr:Ca reference levels were established to identify occupancy of fresh, salt and inter- habitat transition waters, on the basis of eels from the present study that occupied only one habitat type after age 1 (see Results). A five-point running mean was used to smooth short-term fluctuations in Sr:Ca ratios that are likely due to otolith surface flaws or analytic artifacts (Kotake et al. 2003). Smoothed Sr:Ca ratios that were less than the mean +2 SD of smoothed measurements of age 1+ eels from McCallums Pond were considered to indicate freshwater occupancy (F). Smoothed Sr:Ca ratios that were greater than the mean -2 SD of smoothed measurements of 15 saltwater resident age 1+ eels from Brackley Bay were considered to indicate saltwater occupancy (S). Sr:Ca ratios intermediate between these reference levels were considered to indicate transition between habitat types (T).

Eels were categorized as freshwater residents, saltwater residents or inter-habitat shifters on the basis of habitat occupancy patterns. An eel was categorized as a freshwater resident if its combined smoothed Sr:Ca ratios after age 1 were >97% F and <3% S, or

>92% F, <8% T and 0% S. Saltwater residents included eels whose post age 1 Sr:Ca ratios were >97% S and <3% F, or >92% S, <8% T and 0% F. Eels that fell into neither of these categories were termed inter-habitat shifters.

For each eel the proportions of smoothed measurements in F, S and T were calculated within each year, and then summed across all years to give eel-years for each category.

The mean Sr:Ca value within each year for each eel was also calculated from smoothed measurements. Changes in an eel’s smoothed Sr:Ca ratio between F and S after age 1

15 were considered to indicate inter-habitat shifts. Shifts did not include transfers to T.

Shifts were only counted if they were at least four Sr:Ca sample points (40 μm) past the year one marker.

2.4 Results

We caught eels at all fishing locations except McMillans pond bed (Table 2-1). We stopped fishing at Parsons Pond before obtaining sufficient eels for analysis due to excessive lethal bycatch of white perch (Morone americana). Otoliths from 95 eels captured in Brackley and Covehead Bays and McCallums, Cass and Marshalls Ponds were analysed for Sr:Ca ratios. Sr:Ca ratios peaked during the oceanic phase, fell as eels arrived in continental waters and then showed a variety of patterns (Fig. 2-3).

Overall frequency distributions of smoothed Sr:Ca ratios in age 1+ eels were bimodal in samples from both salt and freshwater sites (Fig. 2-4). Modal values were <0.5 x 10-3 and 5 x 10-3. The high mode dominated saltwater samples and the low mode dominated freshwater samples.

Some Sr:Ca trajectories of individual eels were stable near one of the two modes, while others varied widely (Fig. 2-3). All 12 eels from McCallums Pond showed low and stable Sr:Ca ratios, with a mean of 0.57 x 10-3 (SD=0.61 x 10-3). We considered these eels to be freshwater residents (F), and set the upper boundary for freshwater occupancy to be the McCallums mean + 2 SD (1.8 x 10-3). In Brackley Bay, 15 eels showed high and stable Sr:Ca ratios (mean=5.24 x 10-3, SD=1.28 x 10-3) and one eel showed a variable ratio. We considered the 15 Brackley eels with high and stable ratios to be

16 saltwater residents (S), and set the lower boundary for saltwater occupancy to be their mean ratio - 2 SD (2.7 x 10-3). Smoothed ratios between 1.8 x 10-3 and 2.7 x 10-3 were considered to indicate inter-habitat transition (T).

The proportion of eels aged 1+ that were freshwater residents, saltwater residents and inter-habitat shifters differed significantly among ponds (G-test, G=43.4, P<0.001) but not between bays (G=1.92, P>0.05) (Table 2-2). Saltwater residents dominated samples from both Brackley (93.8%) and Covehead Bays (78.3%). All eels sampled from

McCallums Pond were freshwater residents. Most (85%) eels sampled from Cass Pond were inter-habitat shifters. Marshalls Pond samples were 54.2% freshwater residents and 33.3% saltwater residents. When habitat use was measured in eel-years, a significant difference was found between the bays (G=12.31, P<0.01), though salt water use dominated samples from both bays (Brackley 98.9%; Covehead 87.5%). Freshwater use was the sole pattern in McCallums Pond samples. Eels from Cass Pond spent slightly more time in fresh water (52.5%) than in salt water (43.9%) and Marshalls Pond eels spent the majority (79.7%) of their time in fresh water. Proportion of time spent in the various habitats differed significantly among ponds (G=119.3, P<0.001).

Eels from all sites except McCallums Pond showed inter-habitat shifts (Table 2-2).

Mean rate of shifting did not differ between samples in the bays (Brackley 0.12 shifts per eel, Covehead 0.35 shifts per eel; ANOVA F=1.02, P>0.05). Rate of shifting differed among pond samples (ANOVA, F=28.2, P<.0001), with highest rates found in

Cass Pond (1.75 shifts per eel).

17

The number of shifts from salt to fresh water per eel-year for inter-habitat shifters sampled in Cass Pond increased significantly with age (logistical regression t-ratio =

2.04, P<0.05; Fig. 2-5). Rate of shifting from fresh to saltwater did not differ significantly with age (t-ratio=0.95, P>0.05). Overall rate of shifting between habitats increased with age (t-ratio=2.08, P<0.05). Fig. 2-6 plots the percent of time spent in salt, transition and fresh waters vs. age by inter-habitat shifters. For shifters captured in

Covehead Bay, percent of time in salt water increased, and percent of time in fresh water decreased, with age (salt: Spearman R=0.406, P<0.05; fresh: Spearman R=-

0.417, P<0.05; n=29 eel-years). For shifters from Cass Pond, percent of time in saltwater decreased, and percent of time in freshwater increased, with age (salt:

Spearman R=-0.305, P=0.003; fresh: Spearman R=0.263, P=0.001; n=29). Percent time allocated to salt and fresh habitats by shifters captured in Marshalls Pond did not show significant trends with age (salt: Spearman R=0.318, P>0.05; fresh: Spearman R=-

0.211, P>0.05; n=21).

2.5 Discussion

2.5.1 Effect of obstacle type on upstream movement

Eels cannot swim directly against strong currents, but those smaller than 10 cm can creep up damp vertical walls (Legault 1988). We therefore expected that eels would colonize ponds with vertical water drops (McCallums, Cass) only at young ages, while colonization of Marshalls Pond, which drains by a low-gradient channel, would occur at all ages. This was confirmed for McCallums Pond, where Sr:Ca ratios showed that all 18 captured eels had entered freshwater in their elver year (Fig. 2-3). The dam at this site, with its 2.2 m vertical water drop, evidently posed an obstacle to upstream migration of post-elver eels. Contrary to prediction, eels of all ages moved between salt water and

Cass Pond, indicating that the pool-and-weir salmonid fishway at that site did not impede upstream migration. Eels of all ages also moved between salt water and

Marshalls Pond, showing, as expected, that that pond's low-gradient channel could be readily ascended by eels.

2.5.2 Movement patterns

This study identified three migratory contingents of American eels. Saltwater residents dominated samples in two saltwater bays (84.6% of 39). Freshwater residents were the sole contingent found in McCallums Pond, which is accessible only to eels in their first continental year. Inter-habitat shifters were found in all study sites except McCallums

Pond. Shifters comprised of 85% of samples in Cass Pond and 12.5% in Marshalls

Pond, whose dams did not impede inter-habitat movement.

Sr:Ca life history data are now available for the continental phases of naturally-reared

Anguilla eels of six species at 39 locations worldwide (Table 2-3, Fig. 2-7). These studies show that plasticity of habitat usage is the norm among eels. For all species combined, 27 of 39 (69%) samples contained eels that had shifted from the habitat where they were captured. Fidelity to habitat type appears to be commonest in freshwater, where one half (7 of 14) of samples showed fresh water residency only. The exclusivity of this contingent at many locations may be due to long distances to other

19 salinity zones (Hudson River, USA, Morrison et al. 2003; Pearl River, China, Tzeng et al. 2003a), or to dams which impede upstream movements of eels at post-elver ages

(Cairns et al. 2004, this study). Most samples of eels from brackish and salt water contained eels that had used other habitats (brackish: 13 of 15, 87%; salt: 7 of 10, 70%).

Movement patterns shown by American eels in the Brackley-Covehead system reflect the high plasticity of Anguillid habitat use worldwide.

Sr:Ca studies have demonstrated that some European and Japanese eels never leave salt water (Tsukamoto et al. 1998; Arai et al. 2003). This study is the first to measure Sr:Ca ratios of American eels sampled in salt water. Our finding that most eels from Brackley and Covehead Bays had an exclusive saltwater Sr:Ca profile indicates that American eels can likewise complete their life cycle in the sea. The catadromy paradigm for the

American eel is thus overturned. Like European and Japanese eels, American eels must now be considered as species where catadromy is a facultative life history option. The high representation of exclusive saltwater residency in eels from bay samples (84.6%) also suggests that non-catadromy may be an important and common pattern for

American eels.

Tosi et al. (1988) and Édeline and Élie (2004) found that European glass eels have distinct individual salinity preferences. This implies that young eels separate into migratory contingents upon arrival on the coast, with salt-seeking eels remaining in marine waters while fresh-seekers ascend into fresh waters. The freshwater ponds in the

Brackley-Covehead system are adjacent to saltwater bays, and all are within 5 km of the

20 open sea. Given a salinity preference that is activated on arrival in coastal waters, and an upstream ascent rate of 0.6 km/day (White and Knights 1997), we hypothesized that there would be little time lag between invasion of salt and fresh waters. Consistent with expectation, saltwater and freshwater residents were established in their respective habitats by age 1 (Fig. 2-3). Early movement to settlement areas may have an adaptive basis. Because of their superior climbing ability, young eels can overcome vertical barriers to upstream migration (such as the dam at McCallums Pond) and reach habitats that are inaccessible at older ages.

Eels may choose among salinity zones, and they may also choose between sedentary and mobile lifestyles (Feunteun et al. 2003). Some eels in the Brackley-Covehead system shifted between habitats only once and then remained in the new habitat. This could be interpreted as searching for suitable habitat, and then settling there when they find it

(Fig. 2-3). However, others shifted continuously, suggesting that nomadic behaviour is an inherent property of some eels.

2.5.3 Implications for conservation

American eel abundances have declined since the mid 1980’s (Jessop 2000). Possible causes for the reduction in numbers include adverse oceanic conditions, habitat loss, excessive fishing pressure and obstruction to eel passage upstream. Dams and other obstacles can prevent or impede migration and adversely affect eel populations (Legault

1988, Feunteun et al. 2003).

21 Prince Edward Island has over 800 artificial impoundments (MacFarlane 1999), and most streams are blocked by one or several dams. Our results suggest that eels navigate dams equipped with pool-and-weir salmonid fishways or low-gradient channels. Ponds formed by dams with vertical spillways can be colonized, but only when eels are small and can climb vertical surfaces.

The declines in eel abundance indices have prompted efforts to devise management schemes that would assure adequate escapement to the spawning grounds (Richkus and

Whalen 2000, ICES 2003). Such schemes must recognise contributions of escaping silver eels from unfished, as well as fished areas. In much of eastern North America eel exploitation is restricted to coastal and estuarine waters. Our results suggest that fishing in marine waters may also affect eel populations in nearby fresh waters, due to movement between the two habitats. Neither marine populations nor those of adjacent fresh water are discrete. Population models that estimate escapement of silver eels from salt, brackish and freshwater habitats must account for these movements.

The relative importance of marine and fresh waters to the eel's life cycle, and patterns of migration between these habitats, are vital issues in eel biology and conservation. The

Sr:Ca technique will find wide employ as the key to their understanding.

2.6 Acknowledgements

This study received support from the National Science Council, ROC (NSC 91-2313-B-

002-291 and 92-2313-B-002-057). We thank Angus McLennan, Corey Muttart, Valérie

22 Tremblay, Noella McDonald and Robbie Moore for assistance in the laboratory and field, and Mark Grimmett for measuring chemical concentrations in water samples.

Tillmann Benfey and Allen Curry provided valuable advice at all stages and improved the manuscript with their comments.

2.7 Literature Cited

Arai, T., Kotake, A., Lokman, P.M., Miller, M.J., and Tsukamoto, K. 2004. Evidence of different habitat use by New Zealand freshwater eels Anguilla australis and A. dieffenbachii, as revealed by otolith microchemistry. Mar. Ecol. Prog. Ser. 266:

213-225.

Arai, T., Kotake, A. Ohji, M., and Miller, M.J. 2003. Occurrence of sea eels of

Anguilla japonica along the Sanriku Coast of Japan. Ichthyol. Res. 50: 78-81.

Cairns, D.K., Shiao, J.C., Iizuka, Y., Tzeng, W.N., and MacPherson, C.D. 2004.

Movement patterns of American eels in an impounded watercourse, as indicated by otolith microchemistry. North Amer. J. Fish. Manage. 24: 452-458.

Daverat, F., Élie, P., and LaHaye, M. 2004. Première caractérisation des histoires de vie des anguilles (Anguilla anguilla) occupant la zone aval du bassin versant

Gironde-Garonne-Dordogne: apport d'une méthode de microchimie. Cybium: 28(suppl.

1): 83-90.

23 Édeline, É., and Élie, P. 2004. Is salinity choice related to growth in juvenile eel

Anguilla anguilla? Cybium 28(suppl. 1): 77-82.

Feunteun, E., Laffaille, P., Robinet, T., Briand, C., Baisez, A., Olivier, J.-M., and Acou,

A. 2003. A review of upstream migration and movements in inland waters by anguillid eels: towards a general theory. In Eel biology. Edited by K. Aida, K. Tsukamoto, and

K. Yamauchi. Springer, Tokyo. pp. 191-213.

Gross, M.R. 1987. Evolution of diadromy in fishes. Amer. Fish. Soc. Symp. 1: 14-25.

Ibbotson, A., Smith, J., Scarlett, P., and Aprahamian, M. 2002. Colonization of freshwater habitats by the European eel Anguilla anguilla. Freshwater Biol. 47:

1696-1706.

ICES 2003. Report of the thirteenth session of the joint EIFAC/ICES working group on eels. EIFAC Occasional Paper no. 36.

Jessop, B.M. 1987. Migrating American eels in Nova Scotia. Trans. Amer. Fish. Soc.

116: 161-170.

Jessop, B.M. 2000. Estimates of population size and instream mortality rate of American eel elvers in a Nova Scotia River. Trans. Am. Fish. Soc. 129:514-526.

24 Jessop, B.M., Shiao, J.C., Iizuka, Y., and Tzeng, W.N. 2002. Migratory behaviour and habitat use by American eels Anguilla rostrata as revealed by otolith microchemistry.

Mar. Ecol. Prog. Ser. 233: 217-229.

Jessop, B.M., Shiao, J.C., Iizuka, Y, and Tzeng, W.N. Submitted. Migration between freshwater and estuary of juvenile American eels Anguilla rostrata as revealed by otolith microchemistry. Mar. Ecol. Prog. Ser.

Kotake, A., Arai, T., Ozawa, T., Nojima, S., Miller, M.J., and Tsukamoto, K. 2003.

Variation in migratory history of Japanese eels, Anguilla japonica, collected in coastal waters of the Amakusa Islands, Japan, inferred from otolith Sr/Ca ratios. Mar. Biol.

142: 849-854.

Kotake, A., Arai, T., Ohji, M., Yamane, S., Miyazaki, N., and Tsukamoto, K. 2004.

Application of otolith microchemistry to estimate the migratory history of Japanese eel

Anguilla japonica on the Sanriku coast of Japan. J Appl Icthyol 20:150-153.

Legault, A. 1988. Le franchissement des barrages par l'escalade de l'anguille: étude en

Sèvre Niortaise. Bull. Fr. Pêche Piscic. 308: 1-10.

Limburg, K.E., Wickstrom, H., Svedang, H., Elfman, M., and Kristiansson, P. 2003.

Do stocked freshwater eels migrate? Evidence from the Baltic suggests "yes." Amer.

Fish. Soc. Symp. 33: 275-284.

25

MacFarlane, R.E. 1999. An evaluation of the potential impacts of some Prince Edward

Island impoundments on salmonid habitat. M.Sc. thesis, Acadia University, Wolfville,

Nova Scotia.

Morrison, W.E., Secor, D.H., and Piccoli, P.M. 2003. Estuarine habitat use by Hudson

River American eels as determined by otolith strontium:calcium ratios. Amer. Fish.

Soc. Symp. 33: 87-99.

Richkus, W.A., and Whalen, K. 2000. Evidence for a decline in the abundance of the

American eel, Anguilla rostrata (LeSueur), in North America since the early 1980s.

Dana 12: 83-97.

Secor, D.H. 1999. Specifying divergent migrations in the concept of stock: the contingent hypothesis. Fish. Res. 43: 13-34.

Shiao, J.C., Iizuka, Y., Chang, C.W., and Tzeng, W.N. 2003. Disparities in habitat use and migratory behavior between tropical eel Anguilla marmorata and temperate eel A. japonica in four Taiwanese rivers. Mar. Ecol. Prog. Ser. 261: 233-242.

Tosi, L., Sala, L., Sola, C., Spampanato, A., and Tongiorgi, P. 1988. Experimental analysis of the thermal and salinity preferences of glass-eels, Anguilla anguilla (L.), before and during the upstream migration. J. Fish. Biol. 33: 721-733.

26

Tsukamoto, K., and Arai, T. 2001. Facultative catadromy of the eel Anguilla japonica between freshwater and seawater habitats. Mar. Ecol. Prog. Ser. 220: 265-276.

Tsukamoto, K, Nakai, I., and Tesch, W.V. 1998. Do all freshwater eels migrate?

Nature 396: 635-636.

Tzeng, W.N., Iizuka, Y., Shiao, J.C., Yamada, Y., and Oka, H.P. 2003a. Identification and growth rates comparison of divergent migratory contingents of Japanese eel

(Anguilla japonica). Aquaculture 216: 77-86.

Tzeng, W.N., Severin, K.P., and Wickstrom, H. 1997. Use of otolith microchemistry to investigate the environmental history of European eel Anguilla anguilla. Mar. Ecol.

Prog. Ser. 149: 73-81.

Tzeng, W.N., Shiao, J.C., and Iizuka, Y. 2002. Use of otolith Sr: Ca ratios to study the riverine migratory behaviors of Japanese eel Anguilla japonica. Mar. Ecol. Prog. Ser.

245: 213-221.

Tzeng, W.N., Shiao, J.C., Yamada, Y., and Oka., H.P. 2003b. Life history patterns of

Japanese eel Anguilla japonica in Mikawa Bay, Japan. Amer. Fish. Soc. Sym. 33:

285-293.

27 Tzeng, W.N., Wang, C.H., Wickstrom, H., and Reizenstein, M. 2000. Occurrence of the semi-catadromous European eel Anguilla anguilla in the Baltic Sea. Mar. Biol. 137:

93-98.

White, E.M., and Knights, B. 1997. Dynamics of upstream migration of the European eel, Anguilla anguilla (L.), in the Rivers Severn and Avon, England, with special reference to the effects of man-made barriers. Fish. Manage. Ecol. 4: 311-324.

28

Table 2-1. Catch and catch rates of American eels in fyke nets in the Brackley-

Covehead system.

Site Number Number Number

of of caught

eels gear-days per

caught gear-day

Brackley Bay 32 46 0.696

Covehead Bay 31 52 0.596

McCallums Pond 12 61 0.197

McMillans pond bed 0 44 0.000

Cass Pond 20 89 0.225

Marshalls Pond 30 50 0.600

Parsons Pond 6 5 1.200

29

Table 2-2. Salinity history and rate of inter-habitat shifting of American eels captured in saltwater and freshwater sites. See text for definitions.

Salinity history by numbera Salinity history by eel-yearsb Number of inter-habitat

Number Fresh water Inter-habitat Salt water Fresh Transition Salt Number shifts per eel

of residents shifters residents Eel-years % Eel-years % Eel-years % with edge Mean SD Range

c eels Number % Number % Number % mismatch

Brackley Bay 16 0 0.0 1 6.3 15 93.8 0.3 0.3 0.7 0.8 86.0 98.9 0 0.12 0.49 0-2

Covehead Bay 23 0 0.0 5 21.7 18 78.3 12.0 9.6 3.7 2.9 110.3 87.5 1 0.35 0.78 0-3

McCallums Pond 12 12 100.0 0 0.0 0 0.0 159.0 100.0 0.0 0.0 0.0 0.0 0 0.00 0.00 0-0

Cass Pond 20 2 10.0 17 85.0 1 5.0 66.2 52.5 4.5 3.6 55.3 43.9 3 1.75 1.12 0-4

Marshalls Pond 24 13 54.2 3 12.5 8 33.3 200.0 79.7 1.9 0.7 49.2 19.6 10 0.21 0.59 0-2 aProportion of eels with different salinity histories did not differ significantly among bays (G=1.92, P>0.05), but differed significantly among ponds

(G=43.4, P=0.005). bProportion of eel-years with different salinity histories differed significantly among bays (G=12.31, P<0.05) and ponds (G=119.3, P=0.005). cEels with edge mismatch are those whose otolith edge Sr:Ca ratio indicated fresh water but which were captured in salt water, and those whose otolith edge Sr:Ca ratio indicated salt water but which were captured in fresh water.

30 Table 2-3. Habitat occupancy patterns of wild (non-stocked) yellow and silver Anguilla eels after arrival in continental waters, as inferred from otolith Sr:Ca ratios.

Eels sampled in salt water Eels sampled in brackish water Eels sampled in fresh water Sampling Mig. a N Percent occupancy history b Sampling Mig. a N Percent occupancy history b Sampling Mig. a N Percent occupancy history b Study location habitat S SB B BF F SBF habitat S SB B BF F SBF habitat SSBBBFFSBF Ref. c A. rostrata Gulf of St. Lawrence Estuary N 13 100 Impoundment N 15 100 1 Gulf of St. Lawrence Coastal bays N 39 85 15 Impoundments N 56 16 48 36 2 Nova Scotia River N 29 83 17 3 Nova Scotia River M 64 72 28 3 Nova Scotia River N 107 29 71 4 Hudson River Estuary N 29 35 65 River N 14 100 5

A. anguilla Sweden Coastal waters N 3 100 6 Sweden Coastal waters, estuaries N 18 72 28 7 Sweden Estuaries M 8 88 13 7 Baltic Sea exit Coastal waters M 63 6 62 16 16 8 Germany Open sea M,N 18 100 River N 9 100 9 France Bay N 10 100 Estuary N 12 8 42 50 River N 7 100 10

A. japonica Japan Bay N 19 74 26 11 Japan Bays M,N65017 33 12 Japan Coastal watersM25285220 13 Japan d Bay M 42 23 59 18 14 Japan Bay M 4522 2058

15 31 Japan River N 10 100 9 Japan Coastal waters M,N 39 28 56 15 Estuaries N 15 7 80 13 River M,N 7 14 86 16 East China Sea Open sea M 12 100 9 Pearl R., China River M 74 100 14 Taiwan Estuary M,N 58 5 21 60 14 River M,N 18 50 50 17 Taiwan Estuary M.N 58 5 21 60 14 River 6 100 18 Taiwan d Estuary M 18 13 80 7 14 Taiwan d Estuary N 33 9 64 27 14 A. marmorata Taiwan Estuary M,N 7 86 14 River M,N 79 25 75 17 A. australis New Zealand Coastal lagoon M 20 55 20 25 19 A. dieffenbachii New Zealand Coastal lagoon M 20 25 75 19 aM - on spawning migration, N - not on spawning migration bS -salt, B - brackish, F - fresh cReferences: 1 - Cairns et al. 2004, 2 - this study, 3 - Jessop et al. 2002, 4 - Jessop et al. submitted, 5 - Morrison et al. 2 003, 6 - Tzeng et al. 1997, 7 - Tzeng et al. 2000, 8 - Limburg et al. 2003, 9 - Tsukamoto et al. 1998, 10 - Daverat et al. 2004, 11 - Kotake et al. 2004, 12 - Arai et al. 2003, 13 - Kotake et al. 2003, 14 - Tzeng et al. 2003a, 15 - Tzeng et al. 2003b, 16 - Tsukamoto and Arai 2001, 17 - Shiao et al. 2003, 18 - Tzeng et al. 2002, 19 - Arai et al. 2004 dType B occupancy history may also include types SB, BF, and SBF

63o 10’W

Gulf of St. 012 Lawrence km

Covehe Brackley Bay ad Bay 46o a 25’N

Parsons Pond McCallums Pond

Cass Pond McMillans Marshalls pond bed Pond

b c

0 250 500 m Dam with fish ladder Dam with run-around Dam with outlet Cass vertical Pond overspill Marshalls Pond McCallums Pond

0 250 m

Figure 2-1. Brackley and Covehead Bays, Prince Edward Island (a) and associated freshwater impoundments: McCallums Pond (b) and Cass and Marshalls Ponds (c).

32 a

b 12.2 m

0.9 m

c 303 m 5.0m

Figure 2-2. Schematic profiles of pond outlets, showing the vertical spillway at

McCallums Pond (a), the concrete salmonid fishway at Cass Pond (b) and the low- gradient channel at Marshalls Pond (c).

33 7 -3 6 5 4 3 2 1

Mean Sr:Ca ratio (x10 ) 0

7 -3 6 N=18 Covehead Bay 5 4 3 2 1

Mean Sr:Ca ratio (x10 ratio Sr:Ca Mean ) 0 0 5 10 15 20

7 -3 6 McCallums Pond 5 4 3 2 1

Mean Sr:Ca ratio (x10 ratio Sr:Ca Mean ) 0 0 5 10 15 20

7 -3 6 5 4 3 2 1

Mean Sr:Ca ratio (x10 ratio Sr:Ca Mean ) 0 0 5 10 15 20

7 -3 6 N=8 Marshalls Pond 5 4 3 2 1

Mean Sr:Caratio (x10 ) 0 0 5 10 15 20 Age

Figure 2-3. Mean within-year Sr:Ca ratios vs. age for American eels from five sites. For

inter-habitat shifters (eels that moved between habitat types at least once after age 1),

lines represent individual eels. For eels that remained resident in either salt or fresh water after age 1, symbols represent means, with “N” indicating the number of eels. 34 300 Brackley Bay Covehead Bay

200

100

0 0246810 800 b McCallums Pond 600 Cass Pond Marshalls Pond

400

200 Number of measurements of Number measurements of Number 0 0246810 -3 Sr:Ca ratio (x 10 )

Figure 2-4. Frequency distribution of Sr:Ca ratios measured in otoliths of age 1+

American eels captured in saltwater (a) and freshwater (b) sites.

35 0.8 16 1616 16 13 9 4 3 0.6

0.4

0.2

0.0

0.2

0.4 Number of shiftsper eel-year To salt water salt To water fresh To 0.6 1 234 5 6 7 8 Age

Figure 2-5. Rate of movement between salt and fresh water of eels sampled in Cass

Pond which shifted between habitat types. Sample size (number of eel-years) is given at the top of the graph.

36 Fresh Transition Salt

6 100

80 b

60

40

20

0 12345678

Figure 2-6. Allocation of time in habitat categories during each year, by eels which shifted habitat at least once after age 1. Samples are from Covehead Bay (a), Cass Pond

(b) and Marshalls Pond (c). 37 A. 2 A. rostrata, 3 A. rostrata, 1 A. rostrata, 1 salt water rostrata, fresh water all habitats brackish 2 water 1 1 Number of studies 0 0 0 0 0 1-33 34-66 >67 0 1-33 34-66 >67 01-3334-66>67 0 1-33 34-66 >67

2 2 2 5 A. anguilla, A. anguilla, A. anguilla, A. anguilla, salt water brackish fresh water 4 all habitats water 3 1 1 1 2 1 Number of studies 0 0 0 0

0 1-33 34-66 >67 0 1-33 34-66 >67 01-3334-66>67 0 1-33 34-66 >67

A.japonica, all habitats 3 A. japonica, 2 A. 3 A. japonica, 5 salt water japonica, fresh water 4 brackish 2 2 3 1 water 2 1 1 1 Number of studies 0 0 0 0 0 1-33 34-66 >67 0 1-33 34-66 >67 01-3334-66>67 0 1-33 34-66 >67

4 All species, 5 All 12 All species,

salt water 6 10 all habitats 38 3 4 fresh water brackish 8 3 water 4 2 6 2 4 1 2 1 2 Number of studies 0 0 0 0 0 1-33 34-66 >67 0 1-33 34-66 >67 01-3334-66>67 0 1-33 34-66 >67 Percent of eels which showed movement out of the habitat type where they were captured

Figure 2-7. Percentage of sampled eels which showed movements out of the habitat type where they were captured, based on Sr:Ca analysis. The habitat type given in each panel is the habitat where captured. All species includes A. marmorata, A. australis and A. dieffenbachii. See Table 2-3 for data sources.

3 GROWTH RATES OF AMERICAN EELS (ANGUILLA

ROSTRATA) IN RELATION TO HABITAT SALINITY BASED

ON OTOLITH MICROCHEMISTRY

[Manuscript will be submitted for publication to Marine Ecology Progress Series]

3.1 Abstract

Growth of American eels (Anguilla rostrata) captured in salt and fresh water sites in a small coastal watershed in Prince Edward Island, Canada, was determined from length- at-age and back-calculations, and related to movement patterns revealed by Sr:Ca otolith microchemistry analysis. Annual growth of yellow eels that resided in salt water throughout their lives (94.0 mm yr-1) was double that of freshwater residents (45.2 mm yr-1). Eels that shifted between salinity zones had intermediate growth rates (60.5 mm yr-1). Growth was higher than that reported in previous studies, which sampled eels in fresh and brackish water sites. Growth rates are a product of various factors including habitat productivity and temperature. To avoid excessive cumulative mortality, freshwater fisheries for yellow eels must be managed more conservatively than fisheries in higher salinities where eels grow faster and may leave the system to return to the spawning grounds at an earlier age. Population models need to take into account variable growth rates related to variable coastal habitats to manage American eels effectively.

39

3.2 Introduction

Anguillid eels are facultatively catadromous; they spawn in the open ocean and grow and mature in waters in or along continents (Tsukamoto et al. 1998, Daverat et al. 2004,

Lamson et al. submitted). Movements across salinity gradients have been examined for various species including the American eel, Anguilla rostrata (Jessop et al. 2002,

Morrisson et al. 2003, Cairns et al. 2004, Lamson et al. submitted). These studies reveal a plasticity of life history patterns during the continental phase in which eels may reside only in fresh, brackish, or salt water, or shift among salinity zones once or multiple times.

Otolith microchemistry has been used to differentiate between marine and freshwater habitat use of diadromous fish including anguillid eels (Tzeng 1996, Tzeng et al. 1997).

Fish otoliths (ear stones) are composed primarily of calcite (CaCO3). During formation, strontium (Sr) can be substituted for calcium (Ca) due to similar ionic charges and radii.

The ratio of Sr:Ca is ~4.8 times greater in otoliths of fishes inhabiting marine environments than in fresh water (Campana 1999). Growth rates can be quantified by measuring the distance between the annual rings of growth along a transect from the otolith centre to its edge, and habitat salinity can be inferred by measuring Sr:Ca ratios along the same transect (Tzeng 1996). These methods, when combined, simultaneously reveal the growth and habitat history of the fish.

40 Eel density, water temperature and food availability are important factors in determining growth (Tesch 2003). The availability of food to an eel is influenced by productivity and the ability of an eel to capture and metabolize the food. Catadromy is theorized to have evolved in tropical regions where freshwater productivity is higher than in salt water (Gross 1987). In temperate regions where marine waters are typically more productive than fresh waters, the occurrence of variant habitat choices from solely salt water to brackish and fresh water during the continental growth phase may also occur.

Recent experiments by Édeline and Élie (2004) suggest that salinity may directly influence feeding behaviour and therefore growth. European glass eels (Anguilla anguilla) caught off a river mouth and reared for 66 days in salt or freshwater aquaria with uniform food availability showed higher growth rates in salt versus fresh water. In salt water, eels grew faster than those which had exhibited fresh water preferences.

Most American eel growth studies are based on eels sampled from fresh water because eels are presumed to be catadromous, growing only in freshwater habitats. Growth rates have recently shown to be higher in brackish than in fresh water (Morrison et al. 2003,

Cairns et al. 2004, Jessop et al. 2004). Sharp declines in some populations (Richkus and

Whalen 2000, Casselman 2003) have prompted concern for American eel conservation, and underline the importance of accurate population models to insure sustainable management (ICES 2001). Such models require growth data across the full range of salinities occupied by the species. Anguillid eels must attain a minimum size to mature

41 and therefore growth rate, not age, determines when an individual matures and returns to the ocean spawning ground (De Leo and Gatto 1995).

Cumulative exposure to commercial fisheries exploitation depends on the duration of the growth phase, and therefore its rate, as faster growing eels will embark on their spawning migrations earlier than slower growing eels. The bulk of the data on eel growth, derived mainly from fresh water sampling, may understate growth rates for the species. The American eel’s growth phase ranges from 4 to 43 years (Aoyama and

Miller 2003). This implies a very high variability in growth rate, perhaps due to variation in habitat salinity. With most eel fisheries in eastern North America being prosecuted in coastal bays and estuaries (Jessop 1997), there is a pressing need for growth data specific to these waters.

This study examined the relation between growth rates in saltwater bays and adjacent freshwater ponds using otolith Sr:Ca. We hypothesized that growth rates of American eels increase with the proportion of time spent in the more biologically productive salt waters.

3.3 Methods

This study was conducted on the north shore of Prince Edward Island in Brackley and

Covehead Bays and associated ponds (Fig. 3-1). These conjoined bays contain full- strength saltwater (>28 ppt) and have a combined watershed of 81 km2. Four streams entering these bays are blocked by dams at head of tide to form freshwater

42 impoundments. Water exits McCallums Pond to Brackley Bay by falling vertically 2.2 m from a wooden spillway set in an earthen dam. Cass Pond on Covehead Bay has a 5- chamber pool-and-weir salmonid fishway through which water drops 0.9 vertical m over a horizontal distance of 12.2 m. Cass Pond also has a vertical concrete spillway 5 m wide. Marshalls Pond drains into Covehead Bay by a low-gradient channel with a rocky bottom that falls 5.0 vertical m over a horizontal distance of 303 m (1.7% slope).

Parsons Pond connects to Covehead Bay via a culvert with wooden baffles on its floor that are designed to aid fish movement. The sole unimpounded stream in the Brackley-

Covehead system runs through the bed of the former McMillans Pond, whose dam washed out in 1996.

Eels were fished by fyke net and eel pots in Brackley and Covehead Bays, associated ponds, and the stream running through McMillans pond bed in May-November 2003. A backpack electroshocker was also used during this time to sample Cass Pond, Marshalls

Pond, McCallums Pond and the head of the low gradient channel draining Marshalls

Pond. Eels were anaesthetized with clove oil, measured for total length (to 1 mm), weighed (to 1 g), and frozen until the otoliths were removed.

A total of 164 eels from the Brackley-Covehead Watershed were captured. Of these, 95 eels were analysed for salinity history using otolith Sr:Ca ratios. One sagittal otolith per eel was subject to microchemical analysis. The ratio Sr:Ca was quantified at 10 μm increments along the radius of the otolith as described by Tzeng et al. (1997) to determine the salinity history of the eel. Sr:Ca reference levels were established to

43 identify occupancy of fresh (1.8 x 10-3), salt (2.7 x 10-3) , and inter-habitat transition (1.8 x 10-3 to 2.7 x 10-3) waters for each Sr:Ca point (Lamson et al. submitted). Eels were categorized as freshwater residents, saltwater residents or inter-habitat shifters on the basis of habitat occupancy patterns.

All 164 eels were independently aged a total of three times by two agers who marked the otolith annuli in years. Eels were considered to be age 0 in their year of arrival in continental waters, and year 1 was marked as the ring outside the elver check (Michaud et al. 1988, Shaio et al 2003). The distances from the primordium to each age marker, and to the edge of the otolith were measured (to 1 μm) along the longest axis using image analysis software (Rohlf 2004).

3.3.1 Growth Analysis

3.3.1.1 Length-at-age

Growth of freshwater residents and saltwater residents was first compared by analyzing length-at-age data. We used linear regression to test for the effect of age (log10 transformed) on growth (log10transformed). We tested for differences in the growth of freshwater versus saltwater residents by examining the interaction between salinity residence and age.

Because the point when eels leave the system to spawn depends on size, not age, faster growing eels will depart earlier than slow growing eels. Hence slow growing eels may be over represented in higher age categories in length-at-age plots. For this reason,

44 length-at-age regressions were calculated for eels under 15 years of age, and also eels of all ages. Age 15 was chosen to allow a large enough sample size for each category, but limited older eels that may be slow growers.

3.3.1.2 Back-calculation

Back-calculation uses length-at-capture and otolith radius measurements to estimate length at earlier ages (Francis 1990, 1995). We used the technique to increase the number of length-at-age data and to determine and compare growth of individual eels during time spent in freshwater and time spent in salt water.

To increase sample size, data from all eels captured in the Covehead-Brackley watershed were used to determine the relationship between total length and maximal otolith radius. A linear regression of fish length on otolith radius fitted to log10 transformed data was used in the body proportional hypothesis back calculation equation (BPH) to determine back calculated lengths at ages for eels that underwent

Sr:Ca analysis:

Log10Li = [(c+d*log10Oi) / (c+d*log10Oc)]*log10Lc

Where Li = length at capture, Oi = otolith radius at annulus i, Oc = otolith radius at capture, Lc = total length at capture, c = intercept and d = slope from the regression of total length on otolith radius.

Annual growth was estimated for each eel from age one to the last annual ring before the edge of the otolith by calculating the difference between length at age t+1 and at age t. We compared growth rates in years that eels spent in salt water with years spent in fresh water. Annual growth rates were compared at each site, for freshwater

45 residents, inter-habitat shifters and saltwater residents. Data were log transformed and only eels under 15 years were used. A linear mixed model was employed as repeated measurements of individuals were not independent. Individual eels were treated as the random effect and salinity residency and annual growth as fixed effects. All models were fit in R 2.1.1 (R Development Core Team 2004) statistical program using the nLME routine (Bates and Sarkar 2005).

3.3.1.3 Fulton’s Condition Factor

To analyze fish condition, Fulton’s Condition Factor (K) (Lagler 1956) was employed, following the equation:

K = [100,000 *W] / L3 where W is the weight (g) and L is total length (mm).

3.3.1.4 Chlorophyll

Mean chlorophyll a concentrations of eight freshwater impoundments and 19 bays/estuaries on PEI were derived from MacFarlane (1999), Raymond et al. (2002) and

PEI Dept.of Environment, Energy, and Forestry (2005). All chlorophyll assays were conducted in the same lab using the same methodology.

3.4 Results

The length-at-age relationships for eels captured in saltwater and freshwater were near- linear for younger ages but levelled off at older ages (Fig. 3-2a). Saltwater residents, as indicated by Sr:Ca ratios, (Fig. 3-2b) had significantly greater length-at-age regressions

46 than freshwater residents both when considering all eels (T=4.352, p<0.0001) and eels under 15 years (T=4.464, p<0.0001).

Regressions of total length vs. otolith radius showed no significant differences among inter-habitat shifters and either saltwater residents or freshwater residents (T=0.064, p=0.949). Therefore, data from all aged eels were used in the regression of total length on otolith radius. The regression equation was:

Log10TL (mm) = 1.1711xLog10 otolith radius (mm) + 2.5641 (Fig. 3-3).

This regression provided the coefficients for the BPH back calculation model to estimate lengths at ages for eels that underwent Sr:Ca analysis.

Mean back calculated length-at-age regressions produced slopes that were steepest for saltwater residents, intermediate for inter-habitat shifters and least steep for freshwater residents (Fig. 3-4). Mean annual growth rates based on back-calculations are presented by sample origin and salinity history in Table 3-1. Freshwater residents showed a mean annual growth rate of 45.2mm/year, inter-habitat shifters 60.5mm/year, and saltwater residents 94.0mm/year. Rates differed significantly among groups (F=57.8, df= 2,92, p<0.0001).

Eels under 15 years of age which had spent a higher proportion of time in salt water had higher average yearly growth rates (Fig. 3-5a) (F= 3.408 p<0.0001). Similarly, growth rates of eels under 15 years old within particular years increased with the percent of that year that was spent in salt water (Fig. 3-5b) (F=150.931 p<.0001).

47

Fulton’s Condition Factor was greatest for freshwater residents (x=0.187, SD=0.022) intermediate for inter-habitat shifters (x=0.177, SD=0.026) and least for saltwater residents (x=0.169, SD=0.030), and differed significantly among the three groups

(F=3.616, p=0.031). However, for eels under 15 years of age, condition factor did not differ significantly (F=2.023, p=0.139) between freshwater residents (x=0.178,

SD=0.030), inter-habitat shifters (x=0.164, SD=0.028) and saltwater residents (x=0.180,

SD=0.025).

Mean chlorophyll a concentrations were 4.56 µg/L (SD 2.28) in eight PEI freshwater ponds and 9.27 µg/L (SD 3.68) in 19 PEI bays and estuaries. Concentrations differed significantly between the habitat types (F=11.1, P=0.003). Chlorophyll a concentrations in Brackley and Covehead Bays averaged 9.60 µg/L, near the mean for PEI bays and estuaries. Chlorophyll data are not available for McCallums, Cass, and Marshalls

Ponds.

3.5 Discussion

3.5.1 Length-at-age Analysis

Faster growing yellow eels reach maturity and leave the system at an earlier age than slower growing eels; hence, previous studies of eel growth used age-restricted data subsets to reduce the possibility that slower growing eels may be over represented at older ages. Oliveira and McCleave (2002) compared growth histories of male and female eels and restricted the analysis to age 2–11 for males and age 2-14 for females. 48 Likewise, Jessop et al. (2004) restricted analysis to eels aged 1-13. Though we captured eels up to age 23, we restricted growth analysis from back-calculations to age 1-14. We tested length-at-age regressions for eels <15 years of age as well as for eels at all lengths and found in both analyses that saltwater residents had significantly higher lengths-at- age than freshwater residents.

All eels 15 years of age or older were freshwater residents, which is an indication that growth is slower in freshwater. Eels are fished in PEI in tidal waters only, and only eels longer than 50.8 cm may be retained. It is thus possible that fishing mortality proportionally reduced older eels in saltwater samples. Though there was a substantial representation of larger eels that were saltwater residents in our data set it is possible that eels that stay in the system longer (slow growers) have a higher chance of being removed by fisheries over time.

3.5.2 Back-calculation

Francis (1990, 1995) suggested that the differences in growth estimates of two different models can be used to evaluate imprecision of back-calculation. We used only the BPH back-calculation model as the regression of total length on otolith radius fit the data better than the Scale Proportional Hypothesis (SPH) which uses a regression of otolith radius on total length. The SPH model produced unrealistically low growth estimates that did not comply with observed growth measurements for age 1 eels from samples that we captured and from lengths reported at age 1 in other studies (Jessop et al. 2004).

49 3.5.2.1 Effect of salinity on growth

Length-at-age data for eels under 15 and for all eels showed that saltwater residents grew significantly faster than freshwater eels. Past Sr:Ca studies have shown that eels that had entirely freshwater life histories grew more slowly than those that migrated from fresh to brackish water or lived entirely in brackish water (Morrison et al. 2003,

Cairns et al. 2004). Hansen and Eversole (1984) sampled eels in brackish water and cited higher growth rates than previous studies that sampled in fresh water. Gray and

Andrews (1971), however, found that the growth of eels sampled in brackish water in

Newfoundland was generally slower than that of freshwater eels. Though these studies sampled in known salinities, actual salinity histories of individual eels were not known, so growth rates can not be associated with salinities of growth habitats with full confidence.

American eel growth rates in this study averaged 45.2mm/year for freshwater residents,

60.5mm/year for inter-habitat shifters and 94.0mm/year for saltwater residents. Jessop et al. (2004) estimated annual growth rates of 25.2mm/year for >age 1 for female eels using BPH back-calculations in the East River, Nova Scotia. Female American eels sampled in riverine habitats in Maine had growth rates of 30.9mm/year (Oliviera and

McCleave 2002, as reanalyzed by Jessop et al. 2004), and 39.8mm/year in Rhode Island as calculated from length at age analysis (Oliviera 1999). Anguillid eel growth rates are highly variable from eel to eel and between years within an individual (Tesch 2003).

Generally for most fish species, growth decreases gradually throughout life. Annual growth rates in Anguillid eels have also been shown to decrease with age (Oliviera and

50 McCleave 2002), but not always (Svedang et al. 1998), and are not always the highest during the first few years of life (Moriarity 1983).

Weight is highly variable in eels, due in part to variations in gut fullness (Tesch 2003).

We found that the condition factor was highest for freshwater residents, intermediary for inter-habitat shifters and lowest for saltwater residents when taking all eels into consideration. However, when only eels under 15 years of age were compared, there were no statistical differences. Growth may be directed towards weight in broadening instead of length expansion at older ages. Disregarding length analysis, these results suggest that food availability may not be different between salt and fresh water sites.

Differences in fish growth with salinity are commonly attributed to higher productivity in brackish and salt water at temperate latitudes (Gross 1987, Gross et al. 1988, Jonsson and Jonsson 1993). Catadromy is theorized to have evolved in tropical regions where freshwater productivity is higher than saltwater productivity (Gross 1987) and residing in fresh water during the growing years is a more advantageous strategy. In temperate regions, where marine waters are more productive than fresh water, the occurrence of contingents that occupy fresh water to varying degrees may be observed instead of a strict catadromous lifestyle. In PEI, mean chlorophyll a concentrations in bays and estuaries are about double those of freshwater ponds, which implies higher primary productivity in marine habitats. Therefore, positive relation between salinity and growth reported herein and elsewhere is consistent with the notion that growth rate depends on food availability as influenced by productivity.

51

However, the relationship between growth and productivity may not be so simple.

Higher estuarine than freshwater growth rates have been observed in Japanese eels from sub-tropical Taiwan (Tzeng et al. 2003). Factors that affect growth in eels include temperature (Graynoth and Taylor 2000), eel density (DeLeo and Gatto 1996, Graynoth and Taylor 2004), availability of food items (Tesch 2003), and possibly the enhanced production of growth hormones in salt water (Édeline and Élie 2004), each of which may or may not interact with productivity.

In a study on European eels captured in a lake, growth was higher in eels whose stomachs contained invertebrates rather than fish (Moriarity 1973). Tesch (2003) suggested that if invertebrates are absent due to high competition among predators or to low productivity, eels will switch to piscivory. Prey species in salt and brackish water habitats have been shown to differ, with crabs, fishes and various crustaceans dominating prey species in salt water (Tesch 2003), crabs, mussels, fish, isopods, polychaetes and insects in brackish water (Wenner and Musick 1975) and insects, snails, chironomids, fish and various other organisms in fresh water (Ogden 1970, De Nie

1987). Tesch (2003) suggested that prey available in fresh water is more suited to smaller eels, which implies that movements to salt water should increase with age.

However, we found that movements towards fresh water increased with age, though there was much variability (Lamson et al. submitted). Furthermore, from stable isotope analyses in lakes of New Brunswick, American eels were found to be among the top predators (R.A. Curry, unpublished data).

52

Temperatures affect growth by influencing prey abundance, consumption, digestive rates and energy outputs (Graynoth and Taylor 2004). Eel growth rates have been shown to peak at temperatures ranging from 20-27°C and cease at temperatures below

9°C when they are dormant (Graynoth and Taylor 2000). High summer temperatures near optimal eel growth ranges are indicative of the shallow bays and ponds in Prince

Edward Island, contributing to the high annual growth rates observed in this study.

Édeline and Élie (2004) suggested that as growth hormone regulates osmoregulation in salt water, appetite and growth may be intrinsically higher and more dependent on salinity than the availability of food.

3.5.3 Implications for Conservation

Most fisheries for American eels in Atlantic Canada and the eastern United States target yellow eels in sheltered coastal and estuarine waters (Jessop 1997). Our findings indicate that eels grew twice as fast in salt water. Because eel growth increases from fresh to brackish to salt water (see also Morrison et al. 2003, Cairns et al. 2004, Jessop at al. 2004), the length of time needed to reach the size of sexual maturation decreases with salinity of growth habitat. Sexual maturation in eels is size rather than age dependent and varies geographically and according to sex (Oliviera 1999). Female silver eels exiting freshwater ponds a few km east of Brackley and Covehead Bays had a mean length of 69.6 cm (SD=5.4, N=264; Cairns et al. submitted). Given a minimum legal catch size of 50.8 cm and length of 69.6 cm at spawning migration, our regressions of growth on age of mean back calculations indicate that eels growing in salt water

53 spend 2.5 years in the exploitable size range, while eels rearing in fresh water spend 4.7 years in the same range.

Because duration of exposure to exploitation is longer in fresh water, freshwater fisheries for yellow eels should be managed more conservatively than those in marine waters, to avoid excessive cumulative mortality. Salinity-specific growth rates also need to be incorporated into population models that seek to establish sustainable management of American eels (ICES 2001).

3.6 Acknowledgements

This study received support from the National Science Council, ROC (NSC 91-2313-B-

002-291 and 92-2313-B-002-057). We thank Angus McLennan, Corey Muttart, Valérie

Tremblay, Noella McDonald and Robbie Moore for assistance in the laboratory and field. Tillmann Benfey and Allen Curry provided valuable advice at all stages and improved the manuscript with their comments.

3.7 Literature Cited

Aoyama, J., and Miller, M.J. 2003. The silver eel. Pp. 107-117 in Aida, K.,

Tsukamoto, K., and Yamauchi, K. (eds.). Eel biology. Springer, Tokyo.

Bates, D. and Sarkar, D.. 2005. Linear mixed effects models using S4 models. LME 4 package for R.

54 Cairns, D.K., Omilusik, D.L., Leblanc, P.H., Atkinson, E.G. and Moore, D.S.

Submitted. American eel abundance indicators in the southern Gulf of St. Lawrence.

Can. Data Rep. Fish. Aquat. Sci.

Cairns, D.K., Shiao, J.C., Iizuka, Y., Tzeng, W.N., and MacPherson, C.D. 2004.

Movement patterns of American eels in an impounded watercourse, as indicated by otolith microchemistry. North Amer. J. Fish. Manage. 24: 452-458.

Campana, S.E. 1999. Chemistry and composition of fish otoliths: pathways, mechanisms and applications. Mar. Ecol. Prog. Ser. 188:263-297.

Casselman, J.M. 2003. Dynamics of resources of the American eel, Anguilla rostrata: declining abundance in the 1990s. Pp. 255-274 in K. Aida, K. Tsukamoto, and

Yamauchi, K. (Eds.). Eel biology. Springer, Tokyo.

Daverat, F., Tomas, J., Lahaye, M., Palmer, M. and Elie, P. 2004. Tracking continental habitat shifts of eels using otolith Sr/Ca ratios: validation and application to the coastal, estuarine and riverine eels of the Gironde–Garonne–Dordogne watershed. Mar. Fresh.

Res. 56: 619–627.

De Leo, G.A., and Gatto, M. 1995. A size and age-structured model of the European eel (Anguilla anguilla). Can. J. Fish. Aquat. Sci. 52:1351-1367.

55 De Leo, G.A.; and Gatto, M. 1996. Trends in vital rates of the European eel: Evidence for density dependence? Ecol. Appl. 6:1281-1294.

De Nie, H.W. 1987. Food, Feeding Periodicity And Consumption Of The Eel Anguilla

Anguilla L. In The Shallow Eutrophic Tjeukemeer (The Netherlands). Arch.

Hydrobiology 109: 421-443.

Édeline, E. and Élie, P.. 2004. Is salinity choice related to growth in juvenile eel

Anguilla anguilla? Cybium. 28 suppl.:77-82.

Francis, R.I.C.C. 1990. Back-calculation of fish length: a critical review. J. Fish. Biol.

36:883-902.

Francis, R.I.C.C. 1995. The analysis of otolith data - A mathematician’s percpective

(What, precisely is your model?) In Recent Developments in Fish Otolith Research. The

Belle W. Baruch Library in Marine Science Number 19: 81-95.

Gray, R.W., and Andrews, C.W. 1971. Age and growth of the American eel (Anguilla rostrata) (LeSueur) in Newfoundland. waters. Can. J. Zool. 49: 121–128.

Graynoth, E. and Taylor, M. 2000. Influence of different rations and water temperatures on the growth rates of short finned and long finned eels. J. Fish. Biol. 57:681-699.

56 Graynoth, E. and Taylor, M. 2004. Growth of juvenile eels (Anguilla spp.) in lowland streams in New Zealand. Fish. Res. 66: 95-106.

Gross, M.R. 1987. Evolution of diadromy in fishes. Am. Fish. Soc. Symp. 1:14-25.

Gross, M.R., Coleman, R.M., McDowall, R.M. 1988. Aquatic productivity and the evolution of diadromous fish migration. Science 239:1291-1293.

Hansen, R.A. and Eversole, A.G. 1984. Age, growth and sex ratio of American eels in brackish-water portions of a South Carolina river. Trans. Am. Fish. Soc. 113:744-749.

ICES. 2001. Report of the EIFAC/ICES Working Group on eels. ICES CM

2001/ACFM:03.

Jessop, B.M. 1997. An overview of European and American eel stocks, fisheries, and management issues. Pp. 6-20 in R.H. Peterson, ed. The American eel in eastern

Canada: stock status and management strategies. Can. Tech. Rep. Fish. Aquat. Sci. No.

2196.

Jessop, B.M., Shiao, J.C., Iizuka, Y., and Tzeng, W.N. 2002. Migratory behaviour and habitat use by American eels Anguilla rostrata as revealed by otolith microchemistry.

Mar. Ecol. Prog. Ser. 233: 217-229.

57 Jessop, B.M., Shiao, J.C., Iizuka, Y, and Tzeng, W.N. 2004. Variation in the annual growth, by sex and migration history, of silver American eels Anguilla rostrata. Mar.

Ecol. Prog. Ser. 272:231-244.

Jonsson, B., and Jonsson, N. 1993. Partial migration: niche shift versus sexual maturation in fishes. Rev. Fish Biol. Fisheries 3:348-365.

Lagler, K.F. 1956. Freshwater fishery biology. Brown, New York. 421 p.

Lamson, H.M., Shaio, J.C., Iizuka, Y., Tzeng, W.N., and Cairns, D.C. Submitted.

Movement patterns of American eels (Anguilla rostrata) between fresh and saltwater in a small coastal watershed, based on otolith microchemistry. Mar. Biol.

MacFarlane, R.E. 1999. An evaluation of the potential impacts of some PEI impoundments on salmonid habitat. MSc. Thesis. Acadia University.

Michaud, M., Dutil, J.-D. and Dodson, J.J. 1988. Determination of young American eels, Anguilla rostrata, in fresh water, based on otolith surface area and microstructure.

J. Fish Biol. 32: 179-189.

Moriarty, C. 1973. Studies of the eel Anguilla anguilla in Ireland. 2. In Lough Conn,

Lough Gill and North Cavan Lakes. Irish Fisheries Investigations, Series A 13: 3-13.

58 Moriarity, C. 1983. Age determination and growth rate of eels, Anguilla anguilla (L.).

Journal of Fish Biology 23: 257–264.

Morrison, W.E., Secor, D.H., and Piccoli, P.M. 2003. Estuarine habitat use by Hudson

River American eels as determined by otolith strontium:calcium ratios. Amer. Fish.

Soc. Symp. 33: 87-99.

Ogden, J. C. 1970. Relative abundance, food habits, and age of the American eel,

Anguilla rostrata (LeSueur), in certain New Jersey streams. Trans. Am. Fish. Soc.

99:54-59.

Oliveira, K. 1999. Life history characteristics and strategies of the American eel,

Anguilla rostrata. Can. J. Fish. Aquat. Sci. 56:795-802.

Oliveira, K, and McCleave, J.D. 2002. Sexually different growth histories of the

American eel in four rivers in Maine. Environ. Biol. Fishes. 59:141-151.

PEI Dept. Energy and Environment. 2005. http://www.gov.pe.ca/envengfor/waterdata/app.php. Downloaded November 2005

R Development Core Team. 2004. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-

00-3, URL http://www.R-project.org

59

Raymond, B.G., C.S. Crane, and Cairns, D.K. 2002. Nutrient and chlorophyll trends in

Prince Edward Island estuaries. Pp. 142-153 in D.K. Cairns (Ed.). Effects of land use practices on fish, shellfish, and their habitats on Prince Edward Island. Can. Tech. Rep.

Fish. Aquat. Sci. no. 2408.

Richkus, W.A., and Whalen, K. 2000. Evidence for a decline in the abundance of the

American eel, Anguilla rostrata (LeSueur), in North America since the early 1980s.

Dana 12:83-97.

Rohlf, F.J. 2004. tpsDig, digitize landmarks and outlines, version 2.0. Department of

Ecology and Evolution, State University of New York at Stony Brook.

Shiao, J.C., Iizuka, Y., Chang, C.W., and Tzeng, W.N. 2003. Disparities in habitat use and migratory behavior between tropical eel Anguilla marmorata and temperate eel A. japonica in four Taiwanese rivers. Mar. Ecol. Prog. Ser. 261: 233-242.

Svedang, H., Wickstrom, H., Reizenstein, M., Holmgren, K., and Florenius, P. 1998.

Accuracy and precision in eel estimation, using otoliths of known and unknown age. J.

Fish Biol. 53: 453-464.

Tesch, F.W. 2003. The eel. Blackwell Science, Oxford.

60 Tsukamoto, K, Nakai, I., and Tesch, W.V. 1998. Do all freshwater eels migrate?

Nature 396: 635-636.

Tzeng, W.N. 1996. Effects of salinity and ontogenic movements on strontium:calcium ratios in the otoliths of the Japanese eel, Anguilla japonica Temminck and Schlegel. J.

Exp. Mar. Biol. Ecol. 199:111-122.

Tzeng, W.N., K.P. Severin, and Wickstrom, H. 1997. Use of otolith microchemistry to investigate the environmental history of European eel Anguilla anguilla. Mar. Ecol.

Prog. Ser. 149:73-81.

Tzeng, W.N. Iizuka Y., Shaio, J.C., Yamada Y. and Oka, H.P. 2003. Identification and growth rates comparison of divergent migratory contingents of Japanese eel (Anguilla japonica). Aquaculture 216:77-86.

Wenner, C. A., and Musick, J. A. 1975. Food habits and seasonal abundance of the

American eel, Anguilla rostrata, from the lower Chesapeake Bay. Chesapeake Sci.

16:62.

61

Table 3-1. Mean annual growth (mm/year) of American eels under 15 years old, estimated by otolith radius back-calculation. SW= saltwater residents, IH= inter-habitat shifters and FW= freshwater residents.

Salt water bays Fresh water ponds All All All Brackley Covehead Both bays McCallums Cass Marshalls All ponds SW FW IH SW IH SW IH SW IH FW SW FW IH SW FW IH SW FW IH Mean 104.6 78.4 88.6 59.2 98.2 55.8 44.0 69.6 68.1 69.2 75.3 42.1 41.1 75.2 45.2 61.7 94.0 45.2 60.5 SD 34.0 40.0 30.5 25.9 33.4 24.4 14.7 17.7 33.1 24.1 24.9 15.9 20.8 22.7 18.2 26.6 32.9 18.2 26.2 Number of eels 15 1 18 5 33 6 12 1 2 17 8 13 3 9 27 20 42 27 26 Total eel years 85 6 78 27 163 33 148 11 18 77 31 170 32 42 336 109 205 336 142 F-value 2.998 13.139 19.866 0.034 11.44 15.32 57.8 p-value 0.1053 0.0016 1.00E-04 0.9666 4.00E-04 <.0001 <.0001 62

Figure 3-1. Brackley and Covehead Bays, Prince Edward Island and associated freshwater impoundments: McCallums Pond and Cass and Marshalls Pond.

63

100 90 80 a 70 60 50 40 Length (cm) Salt (n=92) 30 Fresh (n=72) 20 10 0 0 5 10 15 20 25

100 90 80 b 70 60 50 40

Length (cm) Salt water residents 30 20 Fresh water residents 10 Fresh water residents under 15 years 0 0 5 10 15 20 25 Age (years)

Figure 3-2. a) Length-at-age of eels captured in saltwater Brackley and Covehead Bays and freshwater McCallums, Cass, Marshalls and Parsons Ponds and b) saltwater residents, freshwater residents, and freshwater residents under 15 years of age fitted with linear regression lines using an independent variable model.

64

3 Log10TL = 1.1711Log10 radius + 2.5641 R2 = 0.7903 2.8

2.6

2.4 total length (mm) 10 2.2 Log

2 -0.4 -0.2 0 0.2 0.4

Log10 otolith radius (mm)

Figure 3-3. Relationship between total length (mm) and maximum otolith radius (mm).

65

800 700 600 500 400 300

Length (mm) Inter-habitat shifters 200 Freshwater residents 100 Salt water residents 0 0 5 10 15 20 Age (yr)

Figure 3-4. Back-calculated lengths-at-age of saltwater residents, freshwater residents and inter-habitat shifters (mean +1 standard error).

66

) 140 y = 0.435x + 46.386 R2 = 0.5411 120 a

100

80

60

40

Mean growth annual (mm

20

020406080100 Percent of lifetime spent in saltwater

200

) 180 y = 0.4418x + 44.211 160 b R2 = 0.4217

140 120

100 80

60 40

GrowthAnnual (mm/year 20 0

0 20406080100

Percent of year spent in salt water

Figure 3-5. a) Mean annual growth rate vs. time spent in salt water and b) Growth per year vs. percent of that year spent in salt water (F=150.931 p<.0001) of American eels under 15 years old in the Brackley-Covehead Watershed.

67

4 GENERAL DISCUSSION

4.1 General Conclusions

Otolith microchemistry analysis has enabled researchers to describe the lifelong movements of fish that travel between salt and fresh water. This method has provided the imperative to re-categorize anguillid eels from strictly catadromous to facultative catadromous, or in another description, from spawning in the ocean and growing in fresh water, to spawning in the ocean and growing in continental waters, which may not include freshwater inhabitation.

4.2 Summary of Findings

Three life history patterns, termed migratory contingents, were identified for American eels in my study area. After reaching continental waters as elvers, some eels were found to solely inhabit fresh water (freshwater residents) or salt water (saltwater residents), while others shifted between salt and fresh water (inter-habitat shifters).

In the study site, freshwater ponds connected to either of two saltwater bays with varying degrees of obstruction to movement. Of the eels captured in a pond with a pool-and-weir salmonid fishway connection, inter-habitat shifters were the dominant migratory contingent. Freshwater residents dominated a pond connected to the bay by a low-gradient channel. A pond with a 2.2 m vertical spillway contained only freshwater residents that had entered the pond in their elver year. Saltwater residents were

68

dominant in saltwater bays, which overturns the paradigm of obligate catadromy for this species.

Growth was fastest in saltwater residents, intermediate in inter-habitat shifters and slowest in freshwater residents. Mean annual growth increased with the percent of time an eel resided in salt water. Similarly, annual growth within a given year increased with the proportion of that year that was spent in salt water.

4.3 Management Implications and Suggested Research Needs

American eel abundance indices have recently shown sharp declines throughout the species range (ICES 2003). Possible causes for the reduction in numbers include adverse oceanic conditions, habitat loss, excessive fishing pressure and obstruction to movement. Dams and other obstacles can prevent or impede migration and adversely affect eel populations (Legault 1988, Feunteun et al. 2003). With over 800 artificial impoundments (MacFarlane 1999), most streams in Prince Edward Island are blocked by one or several dams. Our results suggest that eels can readily navigate dams equipped with pool-and-weir salmonid fishways or low-gradient channels. Ponds formed by dams with vertical spillways can also be colonized, but only when eels are small.

Efforts to devise management plans that would assure adequate escapement to the spawning grounds are needed (Richkus and Whalen 2000, ICES 2003). Such plans must recognize contributions of migrating silver eels from unfished, as well as fished, areas. In much of eastern North America eel exploitation is restricted to coastal and

69

estuarine waters. Our results suggest that fishing in marine waters may also affect eel populations in nearby fresh waters, due to movement between the two habitats. Neither marine populations nor those of adjacent fresh water are discrete. Population models that estimate escapement of silver eels from salt, brackish and freshwater habitats must account for these movements. Further research is needed to determine relative proportions of silver eels exiting various systems which are freshwater residents, saltwater residents and inter-habitat shifters.

The principal fisheries for American eels in Atlantic Canada and the eastern United

States target yellow eels in sheltered coastal and estuarine waters. Our findings indicate that eels reared in salt water grew 2.1 times faster than eels reared in fresh water.

Because eel growth increases from fresh to brackish to salt water (see also Morrison et al. 2003, Cairns et al. 2004, Jessop et al. 2004), the length of time needed to reach the size of sexual maturation decreases with salinity of growth habitat. Because duration of exposure to exploitation is much longer in fresh water, freshwater fisheries for yellow eels should be managed more conservatively than those in marine waters, to avoid excessive cumulative mortality. Salinity-specific growth rates also need to be incorporated into population models that seek to establish sustainable management of

American eels (ICES 2001).

4.4 Literature Cited

Cairns, D.K., Shiao, J.C., Iizuka, Y., Tzeng, W.N., and MacPherson, C.D. 2004.

Movement patterns of American eels in an impounded watercourse, as indicated by otolith microchemistry. North Amer. J. Fish. Manage. 24: 452-458.

70

Feunteun E., Laffaille, P., Robinet, T., Briand, C., Baisez, A., Olivier, J-M., and Acou,

A. 2003. A review of upstream migration and movements in inland waters by anguillid eels: towards a general theory. In: Aida, K., Tsukamoto, K., Yamauchi, K. (eds) Eel biology. Springer, Tokyo, pp 191 213.

ICES. 2001. Report of the EIFAC/ICES Working Group on eels. ICES CM

2001/ACFM:03.

ICES. 2003. Report of the thirteenth session of the joint EIFAC/ICES working group on eels. EIFAC Occasional Paper no 36.

Jessop, B.M., Shiao, J.C., Iizuka, Y, and Tzeng, W.N. 2004. Variation in the annual growth, by sex and migration history, of silver American eels Anguilla rostrata. Mar.

Ecol. Prog. Ser. 272:231-244.

Legault, A. 1988. Le franchissement des barrages par l'escalade de l'anguille: étude en

Sèvre Niortaise. Bull Fr Pêche Piscic 308:1 10.

MacFarlane, R.E. 1999. An evaluation of the potential impacts of some PEI impoundments on salmonid habitat. MSc. Thesis. Acadia University.

71

Morrison, W.E., Secor, D.H., and Piccoli, P.M. 2003. Estuarine habitat use by Hudson

River American eels as determined by otolith strontium:calcium ratios. Amer. Fish.

Soc. Symp. 33: 87-99.

Richkus, W.A., and Whalen, K. 2000. Evidence for a decline in the abundance of the

American eel, Anguilla rostrata (LeSueur), in North America since the early 1980s.

Dana 12:83-97.

72

Curriculum Vitae

Heather M. Lamson Mailing Address: Department of Biology Canadian Rivers Institute University of New Brunswick PO Box 45111 Fredericton, NB E3B 6E1 Phone: (506) 447-3373 Fax: (506) 453-3583 Email: [email protected]

Home Address: 95 Kingswood Drive Fredericton, New Brunswick E3B 6Z8 Phone: (506) 454-9160

Universities Attended:

2000 B.Sc. in Fisheries Biology, University of Northern British Columbia

Publications:

Lamson, H., Shiao, J. C., Iizuka, Y., Tzeng, W.N. and D.K. Cairns. Submitted. Movement patterns of American eels (Anguilla rostrata) between fresh and saltwater in a small coastal watershed, based on otolith microchemistry. Marine Biology.

Conference Presentations:

Lamson, H., Shiao, J. C., Iizuka, Y., Tzeng, W.N. and D.K. Cairns. 2004. “Movement patterns of American eels (Anguilla rostrata) between fresh and saltwater in a small coastal watershed, based on otolith microchemistry”. Southern Gulf of St. Lawrence Coalition on Sustainability. Charlottetown, Prince Edward Island.

Lamson, H., Shiao, J. C., Iizuka, Y., Tzeng, W.N. and D.K. Cairns. 2005. “Movement patterns of American eels (Anguilla rostrata) between fresh and saltwater in a small coastal watershed, based on otolith microchemistry”. U.N.B.’s Graduate Student Association Conference, Fredericton, N.B