The INTERCAFE Field Manual

Research methods for , fishes, and the interactions between them Photograph — Shutterstock

The INTERCAFE Field Manual

Research methods for Cormorants, fishes, and the interactions between them Photograph — Shutterstock

Dave Carss, Rosemarie Parz-Gollner & Josef Trauttmansdorff the intercafe field manual

ACKNOWLEDGEMENTS

This publication has been produced Russell, Josef Trauttmansdorff, During the drafting process, the by INTERCAFE’s Work Group 1, Mennobart van Eerden, Stef van Field Manual was reviewed by the the ‘Ecological Databases and Rijn, Catarina Vinagre, and Stefano Work Group and was also made Analyses’ group, which comprised Volponi. Additional text was available to all INTERCAFE 24 researchers from 22 countries provided by Stuart Newson (British participants for comment and across and beyond (for full Trust for Ornithology, UK). correction. Numerous participants details of WG1 participants, please from INTERCAFE’s other two see Appendix Three). Helpful comments and corrections Work Groups made contributions on various drafts were provided by to the WG1 meetings, and these This publication was drafted the above and also by: Ian Winfield meetings were also informed by Thomas Bregnballe, Dave (Centre for Ecology & Hydrology, on occasions by discussions Carss, Henri Engström, Manfred Lancaster, UK). with invited experts and local Enstipp, Susana Franca, Marijan stakeholders. Wherever appropriate, Govedic, Robert Gwiazda, A number of people also kindly information from these sources Reinhard Haunschmid, Sven- provided photographs for inclusion has also been incorporated into the Hakon Lorentsen, Petr Musil, and individual sources are INTERCAFE Field Manual. Ivailo Nikolov, Jean-Yves Paquet, acknowledged under each picture. Rosemaire Parz-Gollner, Ian

[2] the intercafe field manual

CONTENTS

1 PREFACE 5 7 INTRODUCTION Part Two — Working with Fishes 74 2 INTRODUCTION 7.1 European fish diversity 75 Part One — Working with Cormorants 9 7.2 Fish distribution and movements 76 2.1 The Great (Phalacrocorax carbo) 9 7.3 Fish communities and assemblages 76 2.2 Some important words of warning 12 7.4 Ecosystem-level factors affecting fishes 77 7.5 Defining fish populations 78 3 COUNTING CORMORANTS 14 7.6 Estimating fish abundance 79 3.1 Introduction 14 3.2 Breeding colony counts 15 8 ESTIMATING ABUNDANCE OF FISH 80 3.4 How to use numbers? 26 8.1 Estimating abundance of marine and 3.5 Estimating breeding success 30 coastal fish species 80 8.2 Estimating abundance of freshwater fish 4 CORMORANT DIET, FOOD AND ENERGY species 84 INTAKE 35 8.3 Problems in estimating the abundance of 4.1 Introduction 35 fish 89 4.2 Feeding observations 37 4.3 Pellet collection and analysis 37 9 INDICATORS OF CORMORANT DAMAGE 92 4.4 Stomach contents analysis 42 9.1 Wounding of fish 92 4.5 A note on regurgitations 44 9.2 Damage at fish farms 93 4.6 Reference collection 45 9.3 Damage to wild stocks 94 4.7 Bioenergetics 46 10 ATTEMPTING TO INTEGRATE CORMORANT 5 SEXING, AGEING AND MEASURING AND FISH DATA 96 CORMORANTS 54 10.1 Introduction 96 5.1 Introduction 54 10.2 Interpreting understandings of 5.2 Plumage colouration — age determination 56 Cormorant ‘impact’ at fisheries 97 5.3 Body measurements — sex determination 58 10.3 Introduction to the ecological 5.4 Large Samples — sex and age discrimination 59 perspective 101 5.5 Racial identity (carbo, sinensis) 62 10.4 Review of some key considerations for 5.6 Measurements from nestlings 64 Cormorant-fisheries interactions 103 10.5 The effect of Cormorant predation on 6 MOVEMENTS 66 fish — some words on data requirements 104 6.1 Introduction 66 10.6 Predator-prey relationships — some 6.2 Metal-rings 67 concepts 106 6.3 Colour-rings 68 10.7 Using fish stock assessments as a basis 6.4 Reading colour-rings on Cormorants 69 for assessing Cormorant impact 107 6.5 Reporting sightings of ringed Cormorants 70 10.8 Natural regulation in fish populations 108 6.6 Some ideas for using colour-rings for new 10.9 The effect of natural variation on studies 72 production and catches 109 6.7 More general information, selection of 10.10 Why do biologists find it difficult to useful links 72 measure a Cormorant ‘impact’ at fisheries? 110

www.intercafeproject.net [3] the intercafe field manual

10.11 Manipulative field experiments and 12 APPENDIX TWO: NATIONAL AND understanding complex ecological INTERNATIONAL CORMORANT COUNTS, processes 113 EXAMPLES OF DATA FORMS 135 10.12 How best to make progress? Data requirements and methodology 115 13 APPENDIX THREE: WORK GROUP 1 10.13 Concluding remarks — integrating MEMBERSHIP 140 perspectives 119

REFERENCES 121

11 APPENDIX ONE: RELATED CORMORANT SPECIES IN EUR0PE 131 11.1 Pygmy Cormorant (Phalacrocorax pygmeus) 131 11.2 (Phalacrocorax aristotelis) 133

[4] the intercafe field manual

1 PREFACE

This publication is supported by the context of the interdisciplinary Norway, Serbia) and the Middle COST. It is one of the outputs of the management of human-wildlife East (). In addition to these INTERCAFE COST Action (635). conflicts at local to international 28 countries, Ukraine and Croatia COST (European Cooperation in levels across Europe. It also were also associated with the Science and Technology) is the aimed at delivering a coordinated Action. INTERCAFE held a series longest-running intergovernmental information exchange system of eight meetings, each themed network for cooperation in research and improved communication around a topic particularly relevant across Europe. between stakeholders. To this end, to the host country: INTERCAFE attempted INTERCAFE — Conserving to address:- 1. Gdansk, Poland, April biodiversity: interdisciplinary 2005 — ‘Cormorant ecology, initiative to reduce pan-European i. The fundamental distrust commercial fishing and Cormorant-fishery conflicts — was between the main stakeholder stakeholder interaction’ awarded funding for four years groups which was compounded 2. Saxony, Germany, September (2004–2008). COST Actions are by the disparate and 2005 — ‘Commercial carp charged with directing European uncoordinated nature of aquaculture’ science and do not pay for available sources of information, 3. Hula Valley, Israel, January researchers’ time. Instead, funding ii. the necessity of applying an 2006 — ‘Cormorant-fishery was available for INTERCAFE integrated interdisciplinary conflict management in the Hula to organise and run a series of research approach (biological, Valley, Israel’ international meetings, drawing social, legal) to Cormorant- 4. Bohinj, Slovenia, October together researchers from a fishery conflicts (as these are 2006 — ‘Angling and EU number of disciplines (- as much a matter of human legislation’ related and broader ecology, interests as they are of biology 5. Hanko, Finland, April fisheries science and management, or ecology), thus recognising 2007 — ‘What to do when the sociology, social anthropology the need for different Cormorant comes’ and international law) and other perspectives in the development 6. Po Delta, Italy, September experts (very often connected of collaborative strategies, and 2007 — ‘Extensive aquaculture with fisheries production, harvest iii the lack of an integrated systems and relationships and management, or to regional/ understanding of the between stakeholder national policy and decision- interdisciplinary factors at the perspectives and different making). Under INTERCAFE’s heart of Cormorant-fisheries spatial and institutional levels’ coordination, interested parties, conflicts that precludes the 7. South Bohemia, from local stakeholders to provision of useful and practical Czech Republic, April international policy-makers, were information and advice to all 2008 — ‘Management practices thus offered a unique opportunity interested/affected parties. in a complex habitat mosaic and to address European Cormorant- at local, regional and national fisheries issues. The INTERCAFE network levels’ comprised almost seventy 8. Paris, France, September The main objective of researchers from all 27 EU Member 2008 — ‘The management of INTERCAFEwas to improve States (except Luxemburg, Malta Cormorant-fisheries conflicts in European scientific knowledge of and Spain) and other countries France and the wider European Cormorant-fisheries interactions in in continental Europe (Georgia, context’

www.intercafeproject.net [5] the intercafe field manual

At each meeting, INTERCAFE as possible given the budgetary an integrated synthesis (ISBN participants worked in one of three and time constraints on all of 978-1-906698-065) at http://www. Work Groups, covering the broad INTERCAFE’s participants. intercafeproject.net aims of the Action: The INTERCAFE publications are: Drawing on INTERCAFE’s ▪▪ Work Group One — Ecological ability to develop a network of Databases and Analyses ▪▪ Cormorants and the European researchers and the Action’s ▪▪ Work Group Two — Conflict Environment: exploring privileged opportunity to see and Resolution and Management Cormorant status and distribution hear about Cormorant-fishery ▪▪ Work Group Three — Linking on a continental scale. issues across Europe and beyond, Science with Policy and Best (ISBN 978-1-906698-07-2) ‘The INTERCAFE Field Manual’ Practice ▪▪ The INTERCAFE Field is a review of the methods used by Manual: research methods for researchers to study the ecology of Most meetings included a field Cormorants, fishes, and the both Cormorants and fishes, and visit to allow participants to see interactions between them. an exploration of how such bird Cormorant-fishery conflicts at first- (ISBN 978-1-906698-08-9) and fish datasets can be combined hand. In addition, wherever possible ▪▪ The INTERCAFE Cormorant and compared in a useful and the INTERCAFE budget was also Management Toolbox: methods biologically meaningful way. used to invite appropriate local, for reducing Cormorant regional, national or international problems at European fisheries. The INTERCAFE Field Manual experts to these meetings. Through (ISBN 978-1-906698-09-6) has evolved organically over these discussions and interactions, ▪▪ Cormorant-fisheries conflicts a long gestation period. The INTERCAFE participants tried to at Carp ponds in Europe and original idea to write a ‘Cormorant understand the diverse Cormorant- Israel — an INTERCAFE Manual’ emerged during late-night fishery conflicts in Europe and overview. meetings of several participants beyond. (ISBN 978-1-906698-10-2) at the 4th European Conference ▪▪ Essential social, cultural on Cormorants in Bologna, Italy, This publication is one of a series and legal perspectives on in November 1995. Here, there of INTERCAFE outputs aimed Cormorant-fisheries conflicts. was a workshop on methods of at providing readers with an (ISBN 978-1-906698-11-9) studying diet and feeding ecology overview of European Cormorant- Chaired by DNC. This meeting was fishery conflicts and associated Highlights from these publications inspirational and brought together issues, which is as comprehensive will be available in INTERCAFE: 28 researchers from 12 countries

Photograph — Shutterstock

[6] the intercafe field manual

across Europe, north Africa and demography, movements and Cormorant researchers do the work Israel who sat for many hours migration could be described in a they do and how they understand around a large table and drafted ‘Cormorant Manual’. Neither the the inevitable limitations of their a paper that was subsequently aftermath of the WI-CRG Bologna work and the level of confidence published as Carss et al. (1997). meeting nor the REDCAFE project they have in their ‘results’. yielded ‘the Manual’ but the idea After the meeting, several people was not forgotten. It is to Josef Second, much Cormorant research focussed on writing specific texts Trauttmansdorff’s eternal credit across Europe and beyond is which were then circulated to that he resurrected the notion of undertaken (and funded, if funds are all for comments before being the Manual in the early days of available) because the increasing sent to independent reviewers for INTERCAFE and — alongside numbers and expansion of the refereeing. Finally, the sections Rosemarie Parz-Gollner in ’ geographical range across were combined together with a particular but several other key Europe has raised serious concerns running narrative and the final writers (authorship is given amongst representatives of many review paper emerged. The aim of alongside each chapter) — worked fisheries sectors. Indeed, many this paper was to ‘get our house tirelessly towards its fruition. who catch fish for commercial in ’ by discussing the various reasons and/or pleasure believe that methods of diet and food intake The original idea of the Cormorants are having harmful assessment. Manual — to summarise the effects on their business and way methods used by Cormorant of life (van Eerden et al. 2003: To one of us (DNC), the writing researchers to (i) show others 11). Thus the idea was developed process behind this move ‘towards clearly how the research is done to include fisheries methods in a consensus view’ demonstrated and (ii) to try to standardise the Field Manual to complement that a relatively large number of these methodologies amongst that of Cormorant research. This international researchers (from the research community — was information was not needed very diverse biological, ecological developed by INTERCAFE in here in relation to standardising and national backgrounds) could three important ways. methods as much has already work together very effectively been accomplished in the world and, importantly, that the final First, quite a bit of background of fisheries research. Instead, the product was greater than the sum was added to the texts on the emphasis here is on explaining to of its constituent parts. It was from methods used to study Cormorant the interested reader how fisheries this collaboration that the seed to ecology. The idea here was that researchers do the work they develop the REDCAFE Concerted even if readers had no intentions do and how they understand the Action (2000–01) was born, a of undertaking any studies inevitable limitations of their work piece of work that led logically to themselves, they could gain and the level of confidence they the INTERCAFE COST Action a strong understanding of the have in their ‘results’. This fisheries 3 years later. Whilst various issues complexities facing researchers research information is described of Cormorant ecology — and the who were involved in such and discussed throughout much of rigorous study of it — were being work. The questions researchers Part Two of the Field Manual. discussed in Bologna, there was are trying to answer are often much discussion on scientific very simple but the scientific, Third, and perhaps most difficult, methodologies. One oft-voiced methodological, practical, temporal was the desire to explore and wish during those days was that and financial constraints they discuss how Cormorant and it would be extremely useful for inevitably face are seldom so. Thus, fisheries data are integrated. the research community (both as well as an attempt to standardise Such integration is fundamental established researchers and those methodological best practice (to to an understanding of how starting to work on Cormorants) minimise bias, improve validity, Cormorant predation might if all the standard methods used and make studies comparable), ‘impact’ on fisheries interests and, to study Cormorant numbers, Part One of the Field Manual also as pointed out by van Eerden et diet, food intake, energy budgets, became an attempt to explain how al. (2003), this impact has been

www.intercafeproject.net [7] the intercafe field manual

the driver of much research effort. ‘starting point’, nevertheless, it is to integrate information from Nevertheless, such integration of a serious attempt (i) to describe numerous different sources and to bird and fish data is notoriously the current methods used by collate knowledge and experience difficult to undertake with Cormorant researchers and to help on a pan-European (and even a high degree of ‘scientific’ and encourage others in this field, global) scale and this is undoubtedly confidence. Thus the final chapter (ii) to discuss the most commonly a unique strength of the work of the of ‘The INTERCAFE Field used research methods in relevant INTERCAFE Action. Manual — research methods areas of fisheries science, and (iii) for Cormorants, fishes, and to describe the not insignificant INTERCAFE’s Work Group 1 the interactions between them’ problems researchers face when produced two main outputs, this explores attempts to combine attempting to integrate such data. publication provides an overview of Cormorant and fish data, suggests common methods and techniques how best to make progress in this Like all such endeavours, this used in Cormorant-fish studies and field, and discusses how these two publication is the result of the joint which formed the basis of much of perspectives might be integrated. efforts of a number of researchers the quantitative data collated and during the INTERCAFE Action. discussed in the other WG1 output. The contents of the Field Manual The aggregation of information was The other WG1 output ‘Cormorants are not carved on stone. It is possible both in the field by meeting and the European Environment: anticipated that current methods local and national experts from a exploring Cormorant ecology on may be refined and new ones diverse range of sites across Europe a continental scale’ (van Eerden, developed as technology constantly during INTERCAFE meetings (see van Rijn, Volponi, Paquet & Carss improves and becomes available. above), as well as from the literature 2012) provides the results of the Like Carss et al (1997), this and unpublished data. Of particular main analysis of ecological aspects publication should be seen as a benefit was the opportunity in relation to European Cormorants.

[8] the intercafe field manual

2 INTRODUCTION Part One — Working with Cormorants D N Carss and J Trauttmansdorff

This part of the INTERCAFE an absolute estimate of numbers population dynamics — how Field Manual is devoted to of Cormorants (or fishes, see Cormorant numbers change a critical review of field and chapter 8) is required or whether a over both time and geographical laboratory methods used in ‘relative index’ is sufficient. Both range, including how researchers scientific research on Cormorants. can measure changes through investigate the movements of the It aims to offer the reader the best time but only an absolute estimate birds. The following sections give practical advice on how to use tells researchers how many (a) a brief overview of the ecology these methods, understand their are in a particular site. of the and its limitations, and maximise their Field workers must also consider relationship with humans and usefulness. As well as promoting what the data they collect will (b) some very important words the use of standardised research be used for — a reliable index of warning about Health and methodologies, it is also hoped that is more useful to researchers Safety and other practical issues the reader will also gain a better than a poorly conducted count. associated with much of the field understanding of the research It may not be necessary to count and laboratory work described in work required to answer such all the Cormorants (or fishes) at this Field Manual. deceptively simple questions as a site to obtain an estimate of the ‘How many Cormorants are there?’ total size of the population. Thus and ‘What do Cormorants eat?’. field workers need to consider 2.1 The Great Cormorant sampling — it may often be more (Phalacrocorax carbo) This Field Manual is seen as being effective to survey representative a starting point for standardising samples of the population and The Great Cormorant (Phalcrocorax and improving Cormorant research to extrapolate these results to carbo) is one of 65 bird species methods rather than being the obtain an estimate of the entire from six families that make up the last word. All of the methods population. Clearly, samples order Pelecaniformes. This order documented here are currently should then be selected by non- includes the , used by researchers but sometimes subjective means, as those selected and , Cormorants and in slightly different forms. Many for convenience (for example) shags, the and the of them are based on common can have serious drawbacks. It (‘snakebirds’), and sense, detailed understanding of is thus recommended that field tropic birds. Full descriptions of all the species’ natural history and a researchers consult some of the species in this order, as well as their basic knowledge of the sorts of widely available texts on biological comparative biology, full ecological things that are feasible in the field. sampling and statistical analyses and behavioural reviews discussions Furthermore, methods are usually wherever necessary. of their relationships with man devised so that the subsequent are given in two excellent books: data collected can be analysed in In addition, the research techniques Cormorants, and Pelicans some statistically meaningful way. described in this part of the Field of the World (Johnsgard 1993) and One of the important questions Manual explain the intensive work Pelicans, Cormorants and their from this perspective is whether required to investigate Cormorant relatives (Nelson 2005). Readers

www.intercafeproject.net [9] the intercafe field manual

wishing to discover more about In the first half of the 20th these fascinating birds are urged Century, the number of to consult these comprehensive Cormorants in Europe was greatly volumes. reduced but, primarily due to conservation legislation and The Great Cormorant is near resulting actions, the population cosmopolitan in its range, which is in Europe has grown substantially ‘wider than any other Pelecaniform: (for example, see van Eerden et al. from Iceland to New Zealand 2003). However, this ‘conservation through Eurasia and Malaysia, success’ for some has caused great tolerating astonishing range of concern to others. Representatives topography, climate, and habitat in sectors such as fish farming, from tropical to arctic, sea level sports fishing and commercial to 3,400 m, coast to continental fisheries have all voiced their interior, all continents except South displeasure and stressed the America’ (Nelson 2005: 413). harmful effect of Cormorant Despite this vast geographical predation on their business and Figure 2.1 Catching fish with Cormorants in eastern Asia (from range and tolerance for diverse way of life. This conflict of Hoffman 1858). aquatic and wetland habitats, interests between Cormorants the Great Cormorant’s numbers and fisheries was the subject globally are ‘fairly modest and far of an extensive synthesis at the exceeded by even single colonies pan-European level (Carss 2003, of several other pelecaniforms’ Carss & Marzano 2005) which (Nelson 2005: 413). ultimately formed the foundations for the INTERCAFE COST Great Cormorants have been hunted Action. for their flesh for millennia and are still taken as food in Norway for General description of the instance where some 10,000 birds Great Cormorant may be taken per year. Elsewhere Three species of Cormorant the species’ fishing skills have commonly occur in Europe, been exploited on Dojran lake in the Great Cormorant, Pygmy Macedonia where fisherman use Cormorant (P. pygmeus) and the wing-clipped Cormorants and other Shag (P. aristotelis), the last two fish-eating birds to drive fishes into species being described further in specially constructed traps (Gabriel Appendix One. There are thought Figure 2.2 Adult Great Cormorant. et al. 2005). More commonly, to be six races of Great Cormorant Photo courtesy of J Trauttmansdorff. especially in China and Japan, around the world but only two, Cormorants can be domesticated Phalacrocorax carbo carbo (the and trained to catch fish for human ‘Atlantic’ race) and P. c. sinensis Whilst carbo birds tend to be larger consumption (see Nelson 2005: (the ‘Continental’ race), occur in than those of the sinensis race, 98). However, the Cormorant’s Europe. Carbo Cormorants tend there is some overlap (both in size fish-catching abilities are usually to be restricted to the Atlantic and geographical distribution) viewed in a more negative light. coasts of Norway, Britain, Ireland and so there is the possibility of For years, probably for centuries, and northern France, whilst the confusion in the field (though see there has been discussion about sinensis race is far more numerous section 5.5 for further details). Cormorants as fish-eating predators and is distributed across much of Similarly, in the field it is often and potential competitors with the rest of continental Europe and difficult to distinguish the sexes man for food in both marine and Scandinavia, through Asia east to despite overall differences in freshwater environments. Japan and south to Sri Lanka. average body size (mass) and in the

[10] the intercafe field manual

shape of the head (see also sections the gills and are often manipulated Cormorants adapt their foraging 5.3 and 5.4). on the water surface before being behaviour to the food available in swallowed head-first. Smaller fish the different types of waters and Cormorants are very well adapted may even be swallowed whilst so, in big water bodies (e.g. the to their way of life, catching the bird remains underwater. The IJsselmeer in the Netherlands) and feeding on fishes almost average length of fish taken lies with large schools of small fish, exclusively. The species forages between 10 and 25 cm but birds they form feeding flocks, whilst in a very wide range of aquatic can take both shorter and longer on many smaller rivers where habitats in both salt and fresh individuals, their ultimate size individual fish tend to be more waters, preferring relatively being limited by the gape of the widely dispersed, Cormorants tend sheltered and shallow coastal bird. The daily amount of fish taken to hunt individually. areas, lakes, reservoirs, lagoons, by an individual Cormorant varies floodlands, open water in swamps throughout the year, depending Over the year, Great Cormorants and other wetlands, deltas, on things like the bird’s breeding can be observed in suitable habitats estuaries, saltpans, rivers and canals state, whether it is feeding young, across the whole of Europe (and (Cramp & Simmons 1980). or on the time of year — birds may even into Northern Africa during require more food (i.e. energy) the winter). During spring, most While swimming, the Cormorant’s during the wintering period. On Cormorants breed along European body often lies deep in the water average (though see section 4.6), coasts from the Netherlands up and the tail is flat on the water most birds probably consume to the Scandinavian and Baltic surface. The hollow bones of around 500 g of food per day, countries, although landlocked the Cormorant have less air in though the Daily Food Intake can countries (or those with little coast) them than many other birds and be highly variable, depending on also hold numerous inland colonies so the species has a relatively the time of year and the birds’ (see chapter 7 of van Eerden et al. high specific weight. For diving, behaviour. For instance, flying 2012). After rearing their chicks, Cormorants use their feet as one hour probably requires the birds start migrating around the paddles and keep their wings close consumption of about 50 g of fresh colonies. In autumn, when the to the body whilst underwater, fish. water temperatures are dropping, steering with the tail feathers. and fishes often move to deeper The plumage of the Cormorant Cormorants are social birds, regions and are not so available to is less waterproof than that of breeding and roosting in large the birds, Cormorants generally other waterfowl and it absorbs aggregations and, under certain start migrating southwards. They water relatively quickly whilst the conditions, foraging in large disperse very widely during the bird is diving. On one hand this flocks of up to 2,000 individuals. winter and can be found in all makes diving easier (by increasing the bird’s specific weight) but on the other hand it shortens the maximum diving time (exposure) and forces the birds to dry the feathers after diving. This wing- spreading posture is very typical of the species and, as well as helping to dry the plumage, it is also suggested by some to be an aid to the digestion of large, cold fish.

The sharp hooked bill of the Cormorant is an efficient means of catching and holding on to its fish prey. Fish are usually held behind Figure 2.3 Swimming Cormorant. Photo courtesy of J Trauttmansdorff.

www.intercafeproject.net [11] the intercafe field manual

regions where open water bodies with enough food are available (see chapter 8 of van Eerden et al. 2012).

Great Cormorants usually breed for the first time when they are two or three years old. The nesting place is occupied by the male who displays there for the attention of a female. Later both birds build the nest together. Here, the female lays 3–4 (occasionally 5) eggs with a laying interval of 1–2 days. Both parents incubate the eggs alternately and after 28–29 days the young hatch in laying sequence. Both adults then feed the nestlings, first a ‘soup’ of pre-digested fish and later whole fish, which the chicks take out of the throat of their parents. After 4–5 weeks the young start to leave the nest and successively explore their surroundings. In total it takes between 50–60 days for the young Cormorants to able to fly but they continue to be fed by their parents for another 3–4 weeks before becoming fully independent.

Figure 2.4 Breeding Cormorant. Photo courtesy of J Trauttmansdorff. 2.2 Some important words of warning accept responsibility for accidents All species are to some extent This Field Manual includes that occur whilst undertaking any affected by human disturbance information on numerous field of the work described herein. All (directly or indirectly) and many and laboratory techniques that fieldworkers should consult local are particularly vulnerable during can be used to conduct research experts with relevant experience the breeding season. Thus, ‘as into Cormorants, fishes and the before undertaking any fieldwork. little disturbance as possible’ is interactions between them. However, More thorough texts on Health a good general guideline for all many of the techniques require and Safety are available from most fieldwork. InALL circumstances, skilled application and some are research institutions and NGOs permission should be obtained potentially dangerous and pose very and must be consulted wherever before gaining access to private real (i.e. potentially fatal) safety necessary. Moreover, training from land. Furthermore, for many risks. Much of the information, and an experienced researcher will often research activities (bird catching certainly the sentiment, of this section be necessary before any field work and ringing or many forms of fish is taken from Gilbert et al. (1998). is attempted. The following safety sampling for example), official This Field Manual is not a key source reminders are largely common (national or regional) licences are of Health and Safety advice, nor is sense suggestions for what may be required. Moreover, many of these it a Health and Safety handbook and difficult tasks undertaken in difficult require some form of intensive the compilers of this Manual do not conditions. training before they are issued.

[12] the intercafe field manual

Advice on these matters will be carry a map, compass, whistle available from national authorities (for attracting attention in an or established researchers in the emergency), waterproof watch, relevant field. Before undertaking torch (and spare batteries) and, any fieldwork, no matter how always, a first-aid kit. simple it may seem, it is essential for researchers to think carefully Fieldworkers should also always about all the potential risks leave details of their planned involved, including the task itself, location with a responsible person, the environment, the weather, including date and time of departure, the time of day or night, and means of transport (including any other special circumstances. vehicle identification details if It should also be noted that appropriate), itinerary, any potential working in the field alone can hazardous techniques to be used, significantly increase the level of expected time of leaving the site and risk that a researcher is exposed to. returning to base. This responsible Researchers also need to consider person should be told what to do and and be aware of potentially very who to contact in the event of the serious occupational diseases that fieldworker not returning. they might be exposed to such as Tetanus, Lyme disease (tick- This section is just a very brief borne) and Leptospirosis (including overview of some of the key Health Weil’s disease). Field researchers and Safety issues that should be should consider the most suitable considered by workers before clothing for the habitats and season stepping out into the field. It is they are working in and also carry not comprehensive nor is it any extra clothing if necessary. They replacement for any specialist should also consider the need for training required or the help and a ‘survival bag’ and appropriate advice of a local experienced expert Figure 2.4 Breeding Cormorant. Photo courtesy of J Trauttmansdorff. food and drink supplies, and in the field.

www.intercafeproject.net [13] the intercafe field manual

3 COUNTING CORMORANTS T Bregnballe, D N Carss, S-H Lorentsen, S Newson, J Y Paquet, R Parz-Gollner, S Volponi

This chapter focuses on Cormorant extent on why information on bird of potential ‘new’ sites to visit. population counts for both summer numbers is being sought in the The choice to move to a new (i.e. breeding) and winter (i.e. first place. However, whatever the site may also be influenced by migration, winter roosts) seasons. It reasons for counting Cormorants such things as the bird’s previous also explains differences in the data at a particular place and time, experience there, and its distance collected from undertaking ‘day’ consideration has to be given away — flight is an energetically versus ‘roost’ counts, gives some to the bigger picture — of both costly business. definitions of the term ‘numbers’, geographical area and Cormorant and presents two examples of behaviour and ecology — if the Cormorants are highly mobile how numerical data can be used resulting counts are to have any birds and in any one area they will to calculate ‘Cormorant days’ and biological meaning. The term certainly move between foraging breeding success. ‘biological meaning’ is important and resting (loafing) sites, and and we will discuss it here. between these and a communal night roost. As Cormorants are 3.1 Introduction An observer standing on the bank moving about within a given Across Europe there is considerable of a river or a lake or fish pond area, a count of individuals variation in the numbers of may see many Cormorants on the at one particular site does not Cormorants, their breeding and water. However, over a relatively necessarily mean a lot in relation wintering aggregations, and short period of time, new birds to the actual number of birds in migration patterns. Ecologically, may arrive and others leave. Over the area. But what do we mean the population dynamics and the course of a single day there by ‘area’? Generally, it is thought migratory and foraging behaviour can often be considerable variation that Cormorants range 5–25 km of the birds is complex. Moreover, in numbers at any particular between roosts and foraging areas Cormorants forage in a wide site, tending to peak in the early each day, although this range could variety of habitat types, taking a morning and again in the late extend to 40–60 km. However, diverse range of prey species. afternoon. An individual bird’s Cormorants can also make wider decision to remain in a particular excursions — perhaps leaving an Cormorants are large birds and place could depend on such factors area for several days and travelling are often very conspicuous and as the amount of food it has eaten, some hundreds of kilometres before relatively easy to spot in the the prevailing weather conditions, returning to their ‘core’ area. On landscape. However, counting them, the time of day, the tidal cycles (in top of these ‘local’ movements, or more importantly, counting them an estuary for instance), and the Cormorants also exhibit seasonal in an accurate and reliable way — in level of disturbance from humans migration (see chapter 6 of van a way that the numbers produced or competition and interference Eerden et al. 2012). Birds breeding are biologically meaningful — is far from other birds foraging at the in the north of Europe may move from a simple task. same site. A bird’s decision to south across the continent during remain at a particular site may autumn to their wintering quarters The techniques used to count also depend on the site or sites it in the south. In late winter and Cormorants will depend to a great visited previously and the range early spring they will make the

[14] the intercafe field manual

return journey north, often stopping three other races of the species be defined as belonging to the for short periods of a few days are known in north-western, west same colony if they are located (perhaps more) in a number of and south Africa, and Australia, within 2,000 m of the nearest regions or countries before arriving Tasmania and New Zealand neighbouring group of nests. once again at their breeding (Nelson 2005). The nominate race Similarly, visibly separate groups colonies. P. c. carbo breeds around the coasts of nests (but still within 2,000 m of north-western Europe (Norway, of one or more other groups of Thus, given the complexities of Great Britain and Ireland, and in nests) should be referred to as ‘sub- Cormorant numbers on both a daily northern France). Though mainly colonies’. Finally, a single occupied and an annual basis, the concept coastal during the breeding season, nest is sufficient to be classified of ‘area’ (in terms of how many this race frequently occurs in as a separate colony if it is located Cormorants it holds) is a flexible freshwaters outside the breeding more than 2,000 m from any other one. It is also very clear that the season. The carbo race is almost nests. area associated with any particular entirely ground-nesting, including Cormorant count should always be coastal cliffs and rock stacks Cormorants breeding on the defined as accurately as possible, offshore. The P. c. sinensis race ground tend to nest in discrete and that the potential errors in such breeds from southern Norway and and well-defined groups but the counts be fully understood. Finland in the north throughout exact location of these groups may Central and Southern Europe, shift from year to year. Care must The logistics, labour, and mostly in brackish and fresh water therefore be taken during nest coordination required to count systems. The sinensis race mostly counts in ground nesting colonies birds over large geographic areas breeds on aerial structures (mostly to check all potentially suitable are considerable and should not trees but also shrubs and man-made sites for the presence of isolated, or be underestimated. Similarly, structures, but also on the ground in newly-established, groups of nests. the complexities of Cormorant reed beds, bare rocks, beaches, on behaviour and the flexible nature of islands or sand dunes). The method Definition of colony size the ‘area’ being used by individual used to count nests thus depends Cormorant nests can vary from birds at any particular time of day partly on whether they are in trees little more than a depression in the or season are always on the mind or on the ground. ground with little or no additional of those Cormorant ecologists nest material, to large structures or birdwatchers and others who Definition of a colony of sticks and debris — often attempt to count the birds. In this Historically, the definition of a containing 100s or 1,000s of twigs section we describe how best to ‘colony’ has varied somewhat and other material and growing count Cormorants in each of these among countries, as nests often over the years to around 1m wide situations — at breeding colonies, occur in discrete groups of varying and 0.5 m high. At the start of the at night roosts and on foraging size and can be spread over a breeding season, potential nesting grounds. Throughout, the aim is considerable area whilst groups sites (some with and some without to understand the limitations of of nests may frequently change nesting material) are advertised to common methods used to count in size and specific location potential mates by male birds. It is birds and recommend those between breeding seasons. Thus mainly the male that brings nesting methods that produce the most for biological reasons, and in order material to the nest when building accurate, reliable and repeatable to improve future possibilities for new nests or refurbishing ones used figures. making comparisons over time and previously. A complete nest may be between countries, we recommend built from scratch in less than five use of the following definitions: days. 3.2 Breeding colony counts a colony should be considered separate from another one if it There has been some variation Two sub-species or races of the is isolated from other group(s) among and within countries with Great Cormorant (Phalacrocorax of nests by at least 2,000 m. respect to the precise definition carbo) breed in Europe but at least Therefore, groups of nests should of colony size, mainly because of

www.intercafeproject.net [15] the intercafe field manual

different approaches to including (or not including) partially completed nests in nest counts. To standardize counts, we strongly recommend that colony size be defined as the number of apparently occupied nests (often referred to as ‘AON’). An apparently occupied nest is a nest that is in use and sufficiently completed to hold one or more eggs (i.e. nests without eggs or chicks are included if they are presumed to be occupied by a nesting pair).

The number of apparently occupied nests can then be taken as a minimum estimate of the number of breeding pairs within the colony. Figure 3.1 Seasonal change in Cormorant nest numbers given as the percentage of Whilst this method is probably the the maximum number of nests counted in each 10-day period in a section of the Vorsø most reliable for estimating the colony in 1983. total minimum number of pairs of birds breeding at a colony in any one season, it clearly does not The final (minimum) breeding the maximum number of nests represent the possible maximum count (of pairs of Cormorants) is are occupied. If the colony is number of birds associated with the thus the maximum AON count. visited, and nests counted, several colony as it does not include young times during the same season, the birds that prospect for breeding Timing of the count convention is to use the maximum opportunities, or others that might Clearly, the most accurate count count of AON as the size of the have attempted but failed to find a of Cormorant nests (and hence, colony in that particular year. mate and breed during the season. colony size) should be made when However, the ability of the observer

Table 3.1 Best estimate of the timing of maximum nest numbers in Great Cormorant colonies for different countries in Europe.

Country Period when maximum nest Comments numbers occurs Norway1 1 May–15 June Coastal colonies of P. c. carbo Denmark2 25 April–10 May Coastal as well as inland colonies England3 12 April–17 May Inland colonies of mixed sinensis/carbo Wales3 10 May–7 June Coastal colonies of P. c. carbo Italy4 15 April–30 May New colonies settle later Czech Republic5 25 April–5 May South Bohemia The Netherlands6 March Inland colonies The Netherlands6 May IJsselmeer colonies (1978–2000) The Netherlands6 April IJsselmeer colonies (2001–2005) The Netherlands6 May/beginning of June Coastal colonies Germany7 First half of May 1 Røv, N, and Lorentsen, S unpublished, 2 Eskildsen, J, Gregersen, J, Sterup, J, and Bregnballe, T. Unpublished; 3 Newson, S. unpublished; 4 Volponi, S. unpublished; 5 Martincová, R, and Musil, P. unpublished; 6 van Rijn, S, and van Eerden, M R; 7 Knief, W. unpublished.

[16] the intercafe field manual

to make a count at the time when important not to count when it is 70%, whilst one made in June would nest numbers are at their maximum raining or during very cold or very have underestimated it by about 10% will often be constrained by several warm weather to avoid chilling or and one in July by as much as 50%. factors. An observer rarely knows heat stress to either eggs or chicks. exactly when during the season In Table 3.1 we give a best estimate nest numbers can be expected to be The timing of maximum nest for the timing of maximum nest at their maximum. In planning the numbers within a colony may vary numbers in different countries, based timing of a nest count, the observer between years and from colony to on the experience of those involved also needs to take into account the colony. Several publications present in monitoring Cormorant colonies. fact that (1) the number of nests information on the variation in These periods can thus be taken into tends to reach a maximum later timing of egg-laying and hatching, account when planning the timing of in the season in relatively newly- events which are both correlated counts in each country. The timing founded colonies than in older, with the timing of nest building (e.g. of maximum nest numbers in Europe long-established ones, and (2) Newson et al. 2005, Kopciewicz is partly related to location of the that the visibility of nests in trees et al. 2003). However, few studies colony in relation to a North-South declines during the season as leaves have presented information on the and East-West gradient (see sketch emerge and hide nests from view. seasonal variation in nest numbers in Figure 3.2), but it also varies within a single colony. An example locally, partly depending on seasonal Observers may also be forced to of such seasonal variation in nest variation in food availability. count at a sub-optimal time for numbers is shown here for an area For example, in Great Britain numerous reasons, including poor within the Vorsø colony in Denmark Cormorants breeding on the coast weather conditions and/or visibility (Figure 3.1). Clearly, the maximum initiated breeding several weeks at the time of the planned count. number of apparently occupied nests later than those breeding in inland Furthermore, in cases where the (N = 315) was counted on the 8th colonies (Newson et al. 2005). counting of nests can be expected to visit (during April), a count on the lead parents to temporally abandon 4th visit (during March) would have To sum up, although there is a eggs and/or chicks, it is very underestimated this total by around broad pattern in geographical

Figure 3.2 Sketch of the general time shift in breeding Cormorants from north to Figure 3.3 Ground nesting colony south and west to east across Europe. Denmark. Photo courtesy of T Bregnballe.

www.intercafeproject.net [17] the intercafe field manual

variation in timing of breeding (as necessary) and count the nests from indicated in the table), there is also there. Registering nest content is not considerable annual variation and essential for counting nests, but if variation within regions. Ideally, possible a general assessment of the counts of nests in a colony should stage of the breeding cycle should be carried out when the maximum be given (i.e. record approximate number of nests are occupied. proportion of nests with eggs, <1 However, this is sometimes not week-old chicks etc.). If repeated possible because observers very counts are performed, the highest often only have one chance to number recorded should be used as visit a colony during the breeding the total number of AON’s for the Figure 3.4 Counting of ground- season. Unfortunately, there is no colony. In some colonies it may not nesting colony. Photo courtesy of easy way to ‘correct’ or ‘adjust’ possible to see all parts from the T Bregnballe. nest counts to account for the fact vantage point(s) selected. Keep a that they were not made at the note and a map of the parts of the time of maximum nest numbers. colony that are not visible and try nights should be avoided. For these reasons, many counts to estimate (minimum-maximum) are often an underestimate of the number of AON’s likely to be In cases where observers walk maximum AON. hidden, based on numbers in the through the colony counting nests visible sectors of the colony. When and recording nest contents (eggs Finally, the number of pairs reporting the results of the count of and chicks), we recommend that attempting to breed in a colony will nests, make clear how many were information is dictated into a in most cases be higher than the counted directly and how many are tape recorder. This enables the number of nests counted even at the of unknown reliability. observer to keep a better track of time when nest numbers reach their which nests have (and have not) maximum (e.g. Harris & Forbes Entering the colony may be the been counted and it minimises 1987, Walsh et al. 1995). For only option available in some the duration of disturbance. example, a nest built by a pair that ground-nesting sites because When recording nest contents, we gave up early in the season may vantage points are unavailable. In recommend recording for each be taken over later by a new pair, this case, sticks can be placed in the nest the number of eggs, number and nests may disappear before ground within the colony or spray of chicks, and estimated age of the (and new nests may be built after) paint used on selected nests to keep oldest chick in the brood (give age the maximum in nest numbers is track of the parts of the colony in estimated days since hatching, reached in the colony - as has been where nests have been counted. see section 5.6 on biometrics for shown for Shags Phalacrocorax The duration of disturbance can age determination). aristotelis (Harris & Forbes 1987). be reduced by having two or three persons carry out the count Cormorants breeding on the 3.2.1 Nest counts in ground- together. The extent of disturbance ground tend to nest in discrete and well-defined groups but the nesting colonies will, in some areas, also be lower if the count is carried out during the exact location of these groups Counts from the ground night. This is frequently done in may shift from year to year. Care Care should be taken to minimise Norway where summer nights are must therefore be taken to check disturbance. This must be given long and darkness is not a problem all potentially suitable sites for high priority in order to reduce the (N. Røv pers. comm.). Counting the presence of isolated, or newly- exposure of eggs and nestlings to nests at night in northern latitudes established, groups of nests both weather and predators. We has the advantage that adults tend therefore recommend that entering to remain on their nest for longer Using aerial photographs the colony is avoided if possible. and that gull predation is lower The best method, and in some Instead, the observer should find a than during daytime disturbance. areas the only one suitable for suitable vantage point (or several if Obviously, counts during dark counting ground-nesting colonies,

[18] the intercafe field manual

is to photograph the colony from all nests and keep track of those may be mistaken for those of the air. Subsequently, large prints that have (and have not) been Cormorants. of the photos should be made, or counted. It is helpful to make maps slides projected onto a large sheet of the colony and to use features 3.3 Roost counts of paper or the wall. Nests can in the landscape (e.g. certain thus be easily counted by marking individual trees) to keep track of Standard waterbird counts — well them off. Be aware of double- the sections of the colony that have established in the European counting sites if several photos been counted. birdwatching community to from the same colony are used. collect bird-census-data in wetland It is recommended that several Nest counting in tree-nesting habitats — are normally conducted observers count nests from the colonies will often cause extensive during the day time. However, same photo and that the mean of disturbance to the colony. Be this counting methodology is not these counts is used as the size of aware that incubating Cormorants appropriate when assessing the the colony. suddenly detecting a person in total numbers of Great Cormorants the colony can flush from the nest in a specific region during the 3.2.2 Nest counts in tree- immediately causing one or more winter or the migration period. This is because Cormorants frequently nesting colonies eggs to fall out of the nest. A nest count in a tree-nesting colony will move between foraging and loafing In planning the timing of a count, frequently lead to the exposure sites during their daily activities it should be remembered that the of eggs and small nestlings to and there is a strong risk of either visibility of nests usually declines predation from crows and magpies. missing birds or double-counting quite rapidly (within a few days) The loss of eggs and chicks can be individuals. Cormorants also tend as buds burst and leaves grow on minimised by moving around in to use a variety of water habitats trees. Before counts commence, the colony in a way that minimises including small rivers or lakes that it should be determined whether the number of nests disturbed are generally not taken into account some sectors of the colony can within any given time period. during standard waterbird counts. be counted from outside, thereby Finally, be aware that Cormorants As a result, it has been calculated minimising disturbance. However, sometimes breed in mixed colonies that counting Cormorants using the it is usually necessary to walk with other species like herons and standard waterbird count technique through the entire colony to count that the nests of these other species can underestimate regional numbers of birds by at least 30%, and that this counting error varies greatly according to specific local situations (Newson et al. 2005, Worden et al. 2004).

We therefore strongly recommend that workers seeking an accurate count of Cormorants take advantage of the communal roosting habit of the species, which allows accurate counting in most of European wintering and migration situations. In practice, this means that Cormorant counts must be made by ‘controlling’ (i.e. counting the numbers of birds at) roost sites in a coordinated and simultaneous Figure 3.5 Tree colony in Lepelaarplassen in the IJsselmeer area, Netherlands. way in the late afternoon before Photo courtesy of Mervyn Roos. dusk. Text Box 3.1 gives the

www.intercafeproject.net [19] the intercafe field manual

of islands) and the geographic ▪▪ Where to count: count latitude. However, regardless of Cormorants on night-roosts. its location, any roost where birds ▪▪ When to count: simultaneous congregate and rest is easy to evening counts (late afternoon detect due to the white colour of before dusk). the guano that is visible even in late ▪▪ Data collection: use autumn when most trees are still standardized protocols to take covered with leaves. records. ▪▪ Frequency of counts: once a ‘Day’ and ‘night’ roosts month at least during main Cormorants gather at roosts both migration period (depending during the day and at night. After on field workers’ capacity, foraging, Cormorants tend to form geographical area with diurnal roosts (‘day roosts’), close respect to phenology, and to their feeding areas. Generally, large-scale movements). day roosts contain only a few ▪▪ Coordination of counts: dozen birds but some may be regional, national, used by hundreds of individuals. international levels. Counts from day roosts should Text Box 3.1 Recommended methods not be used to estimate regional for counting migrating Cormorants at Cormorant numbers because birds inland sites. may switch between locations Figure 3.6 Roost site with wintering or use several places during the Cormorants (in January), mixed flock day for resting and drying their with juvenile and adult birds. recommended methods for plumage after a foraging bout. In Photo courtesy of J Trauttmansdorff. counting migrating Cormorants at order to distinguish a diurnal roost inland sites. from a night roost it is necessary to wait until nightfall to see whether can sometimes be up to 30–40 km. Communal night roosts Cormorants leave a particular roost Consequently, in winter, any given At the end of the day during the at dusk and move to a different one night roost generally holds most, non-breeding season, and especially to spend the night. Night roosts if not all, of the Cormorants that in the winter, all Cormorants from a can also be occupied by a variable have spent the preceding day given area aggregate at a communal number of birds throughout within a 30 km radius. Cormorant night roost - a traditional site daylight, indeed most ‘traditional’ numbers at night roosts are lowest used as a resting site night after night roosts (used year after year) in the early morning but increase night and year after year. These are also known to serve as day gradually during the day to a night- roosts are always located adjacent roosts. time maximum, as birds return to a significant area of water in from foraging trips. relatively quiet places (preferably At the end of the day however, islands or undisturbed shorelines). every Cormorant will congregate 3.3.1 Counting methods: Most roosts are located on riverside at a night roost that can hold where, when and how? trees (dead or alive, deciduous or anything from a few birds to coniferous) but roosts can also be around a thousand individuals. Very Where to count located on sand or gravel banks, rarely, such night roosts may hold Depending on the area to be artificial structures, small pools, several thousand individuals. As covered, winter Cormorant counts rocks in the water, and even on flying is energetically very costly need to be coordinated by a cliffs. Choice of roost site depends for wintering Cormorants, the regional, national, or international on the nature of the waterbody distances between night roosts and coordinator responsible for (e.g. marine coastline, river, lake, foraging sites are generally only a collating individual roost counts man-made inland waters, presence few kilometres, though the distance from various locations to produce

[20] the intercafe field manual

vantage points from which to count temporal development of wintering birds and details of the Cormorant’s numbers can be produced (i.e. behaviour in relation to the roost numbers increasing to a mid-winter site. peak before declining as birds begin to return to the breeding If the traditional roosting sites are grounds) and a maximum mid- not known by local observers in winter estimate of Cormorant the area to be counted, then some numbers can be derived from the preparatory field work is needed to series of counts undertaken. locate them before any scheduled large-scale count. Interrogating the However, in most European local network of observers, through countries, only one officially internet forums or regional bird agreed counting date is used for watching journals may help. a general winter count (e.g. that for the International Waterfowl When to count Census). This is because of the Cormorant night roost counts should logistical difficulties of having be coordinated in time within a many people in the field counting certain area. This means that all simultaneously several times during counts have to be carried out on the winter months. Traditionally, the same day simultaneously by all the agreed date for a mid-winter observers. This is not a trivial task census on a large geographic scale and requires a lot of preparatory occurs around mid-January (on the regional, national, or international work. To ensure simultaneous Sunday closest to 15th of January) totals. Most large-scale counts (e.g. counts on a national level, a list of when most birds moving over regional, national, international) specific counting dates should be the European continent would be require a dedicated network of agreed amongst the counting team expected to have arrived in their teams of field workers in order to involved well before the migration wintering quarters. ensure full coverage. For example, period starts, and several months the first coordinated winter count of before count(s) are expected to take The considerable coordination (and Cormorants across Europe (January place. In order to produce population field) skills of such simultaneous 2003) required the efforts of over numbers from a national census, (i.e. ‘agreed date’) counts can 3,000 individual observers counting counts should cover the whole not be underestimated. The birds at night roosts. known migration period relevant logistical and practical issues for the specific country. Thus, when to be considered are immense. Before starting to count, an winter roost counts are undertaken Cormorant flocks can frequently inventory — including the at a national (or regional) level, shift between adjacent night roosts geographical coordinates — of all observers need to take into account locally, or over a wider area, and the known night roosting sites in the stage of migration in that so simultaneous counts on the the area to be counted should be particular area or country. same date require detailed national prepared. As well as detailing the coordination. Similarly, close scale of the task, this inventory In order to keep track of the international coordination and will indicate the number of field ongoing migration of Cormorants collaboration is crucial between workers required to undertake the across Europe (from the breeding neighbouring countries as they counts. For a coordinated winter grounds in the north to the southern often share roost sites or roosting count, at least one person should wintering grounds), national counts locations along national borders be given responsibility for each should be scheduled at least once (often along river courses). known roost. Ideally, this person a month (and ideally every second should be familiar with his or her weekend) during the migration On a pan-European level, the actual particular roost and know the best period. In this way, a picture of the date of the counts in relation to the

www.intercafeproject.net [21] the intercafe field manual

timing of the wintering migration is through binoculars or a telescope ▪▪ During the counting period, actually less critical. This is based and a second records the running record every bird entering the on the assumption that, regardless commentary from his/her colleague roost, paying particular attention of the phase of migration in any on bird numbers, flight direction, to large flocks. Record both the particular region at the time of flock-size, age composition and so time of arrival at the roost and the simultaneous count, ‘all’ birds on. flight direction of the birds if at will be recorded because counts all possible. This information are undertaken simultaneously Undertaking an evening count will help regional coordinators on a large-scale across the whole Ideally, birds using communal identify any Cormorants shifting continent at the same time (day or night roosts should be counted between simultaneously- weekend). at the end of the afternoon/early counted neighbouring roosts. evening. The observer should be ▪▪ Just before complete darkness, How to count in place around two hours before make a final count of all the Field workers normally use dusk, but this period can be birds present. Remember to binoculars or a telescope to count shorter if the roost is small (up to stay until complete darkness Cormorants. Observers should 300 birds) and/or the observer is and record the time the count is work from a sheltered position that familiar with the birds’ behaviour completed. Write down the time offers a good view of the roost but at the specific location. However, when you leave the observation does not go closer than the bird’s it is absolutely crucial that the point. ‘reaction’ distance (i.e. no closer observer continues to count until than around 200 m, but reaction it is completely dark and that the Undertaking a morning count distance can be shorter in places time he or she leaves the counting Depending on the locality and the where Cormorants are not subject spot is recorded. Cormorants size of the roost, morning counts to shooting and so are more tolerant rarely enter roosts in the few can also produce reliable figures for of human presence) so that the minutes before complete darkness the number of Cormorants using birds can enter or leave the roost but if neighbouring roosts are them. Morning counts are generally without disturbance. The more disturbed, or birds have experience less accurate than those made in the birds that are present at a roost, the of harassment actions, flocks of evening and this should be taken more difficult it becomes to collect disturbed birds can be expected to into account when interpreting and record additional information arrive at roosts even after sunset particular counts. on a small scale (e.g. position of when it is getting very dark. For individual birds, age ratios). At these reasons, it is important Nevertheless, morning counts can large roosts, flight movements to proceed with counts in the provide accurate information on the might happen so fast, and in such following recommended way: numbers of Cormorants at specific big waves, that here observers need roosts under certain circumstances. to concentrate all their efforts on ▪▪ On arriving at the counting For instance, where roosts are just counting the numbers of birds spot (write down starting relatively small and morning with perhaps only little time left to time), start with a preliminary and evening counts have been take short notes on the flock size of overview counting every bird systematically compared through arriving birds and their direction(s) already present in the roost a series of repeated counts by the of flight. (and, if possible, estimate same observer. In these cases, the the age-composition, see age observer(s) should be in position In general, counting birds moving determination paragraph at the well before the first light of the day in big flocks, requires considerable end of 3.3.2 below and also (i.e. no later than 30 minutes before training and practice which can section 5.2). sunrise), as some individual birds only come through field experience ▪▪ Take records regularly every invariably leave the roost before and dedication. Such counts 10–15 minutes, write records the mass departure of roosting may be made easier through the down on a protocol list Cormorants to their foraging use of counting teams where chronologically (see Appendix grounds. Mass departure from the one observer watches the birds Two for example). roost usually occurs when it is still

[22] the intercafe field manual

too dark to accurately count the employs deliberate disturbance of provided by national or regional birds, although in some areas some the birds its use must be kept to a coordinators (see Appendix Two roosting birds tend to wait until minimum. for national examples). Many the rising sun warms them before countries have standard national leaving. If either of these types of count is forms for collecting waterfowl data used, this fact must be recorded, and these can be used, or adapted, Sub-optimal counting methods and observers must be aware that in for the specific needs of Cormorant Under special circumstances, general they will get less accurate roost site counts. If such counts are standard roost counts as described results from these methods. to be carried out across Europe, above may not be possible. If this guidelines and counting forms is the case, as a last resort, workers Counts at coastal areas, aerial should be translated into national may consider one of the following surveys languages. As an example of an considerably less accurate The counting methods described international count, the standard methods. previously are generally most form for the 2003 pan-European appropriate for inland roost sites, winter roost count is shown in If the roost is not visible from accessible to observers on foot and Appendix Two. any good vantage point, or is on have been well tried and tested private property or some other in many locations. However, it is What to record? inaccessible place, it might be considerably more difficult to count Most counting forms require possible to undertake a departure Cormorant roosts along marine observers to make records by using count. From whatever vantage shorelines or on islands, where either (a) tick-boxes to choose point available, the observer boats are usually needed to carry between various factors (e.g. type counts birds during their mass observers to ‘control’ (i.e. count of roost, climatic conditions during departure from the roost. This birds at) roosts. the count, estimate of the accuracy relies on the fact that the observer has good knowledge of the local The aerial survey is a frequently departure flyways. If several used method for estimating ▪▪ Country, name of départment, major departure flyways are used, waterbird flocks in general, as or province there may be a need for several well as for counting breeding ▪▪ Name of observer (at specific observers. Counting a roost of populations (Laursen et al. 2008, location) birds using a flyway to/from the Pihl & Frikke 1992). To date, only ▪▪ Date of counting roost should only be used in the a few European countries (e.g. ▪▪ Precise point in time of single morning at the moment of ‘mass Finland, Denmark) have used observation and/or time departure’. Arrival of Cormorants this method to count Cormorants period (duration) of longer at the roost site can generally at their roosts during the winter observation occur throughout the day and months. As described previously, ▪▪ Name of waterbody and roost some birds may even stay at the aerial surveys are much more location (e.g. 100 m upstream roost the whole day long, except commonly used to control of village X on River Y) for a short moment after the mass breeding status, the development ▪▪ Geographical coordinates, departure. of breeding sites or to detect the or indication of the roost establishment of new colonies location on a map If vegetation or other obstacles across large areas or ones that are ▪▪ Total number of Cormorants hide roosting birds, several difficult to access. present at the end of the day observers (certainly more than (= number of birds staying one) may consider undertaking a 3.3.2 Data collection, at that specific roost site overnight) disturbance count. In this case, one counting protocols (forms) or more people disturb the birds Text Box 3.2 The absolute minimum by approaching the roost whilst All data records should be made information needed when counting an observer attempts to count the on specific counting forms (usually Cormorants at night roosts must include flying Cormorants. As this method following standard protocols), the following details.

www.intercafeproject.net [23] the intercafe field manual

Table 3.2 Additional data collection (see also Appendix Two for examples of national counting forms).

Details about … Free text and/or tick boxes to record … Environment ▪▪ Type of water body: e.g. river, lake, sea, impoundment area ▪▪ Location of roost: e.g. island, river bank ▪▪ Type of roost: e.g. tree, ground, artificial structure, poles, cliffs Counting Conditions ▪▪ Local climatic conditions: e.g. wind, rain, snowfall, ice cover, fog, visibility ▪▪ Accuracy of counting /data record: e.g. 100%, 75%, 50%, <50% ▪▪ Additional comments: instances of disturbance, traffic, hunting Bird Details ▪▪ Flock size ▪▪ Arrival time of single birds or flocks (e.g. 1 x 10, 2 x 40....) ▪▪ Flight directions ▪▪ Activity of birds: e.g. ‘16.20h flock of 20 birds, flying upstream/coming from west etc., all birds land in water, swimming and diving; etc.16.30h–20 birds perch on trees…’ ▪▪ Presence of ringed birds (metal and/or color rings) ▪▪ Age ratio: number of individuals in the flock or estimated percentage of counting), or (b) to write specific In practice, the collection of this counters use the same definitions. comments in a series of separate kind of information (see Table 3.2) Frequent counts and records that boxes. Text Box 3.2 shows the should be standardized in some distinguish between adult and absolute minimum information way (e.g. make use of tick boxes in juvenile birds can give valuable needed when counting Cormorants reporting forms) to guarantee that information on the ‘quality’ of a at night roosts.

Additional information Additional notes about environmental parameters and Cormorants can be recorded during the roost count survey to give a more detailed description of roost ‘quality’ (e.g. type and position of trees used, general accessibility for humans, nearby roads or waterways being used by whom, frequency and kind of disturbances), flock composition (e.g. age composition) or Cormorant behaviour (e.g. circling, diving, resting on the water/on the ground, showing alert behaviour, comfort behaviour, sleeping = head under wings etc.) under the specific local conditions. In combination these details can be useful in helping to identify the requirements of roosting Cormorants, which may contribute towards a better understanding as to how range expansion is likely to Figure 3.7 Immature Cormorant proceed under given environmental (above, J Trauttmansdorff) conditions. and adult birds (right, T Bregnballe).

[24] the intercafe field manual

roost as well as on the turnover the observer and their breast and advise the counting teams of migrating flocks (i.e. how long colouration can not be seen), it organized locally. People counting Cormorants stay at particular sites). can be difficult to distinguish in the field should use forms to between adult and juvenile birds. compile their counting results Age determination — important Under these circumstances, such either directly into a computer note: only completely ‘black’ birds should be recorded as being or should fill out the form by birds (i.e. black plumage on of ‘undetermined’ age class (for hand and send results back to their front/ventral ‘breast’ side further details see chapter 5). their national coordinator for as well as the back) should be processing. The collection of recorded as ‘adult’ birds; juvenile forms, summing-up and analysis or immature birds are pale on 3.3.3 Data aggregation and their ventral side (completely or synthesis partly white, showing various In the long run, the opportunity Maximum numbers are white-black spotted patterns) for data aggregation of wintering generally used and reported and are browner on their back. Cormorant numbers collected on when dealing, for example, with From early January onwards, different spatial levels should be breeding pairs or the maximum identification of adult birds is the ultimate goal of observers. It number of mid-winter migrants made easier by the presence of is evident that climatic conditions, observed (per month or any other white patches on the thighs and on various geographic levels, defined time period) in a defined sides of the head. However, in are the driving force influencing area. many cases (for instance when the migration of Cormorants birds are facing away from over the European continent. In some cases, perhaps when The successful survival of birds only a single (or opportunistic) depends on quick reactions and count has been made at a flexible behaviour on a daily particular place/time, it is basis in relation to changing impossible to know how environmental conditions, representative this count is especially during severe winter with respect to phenology. In situations. So the crucial such cases, the count should be point — to achieve and combine/ reported as a time- and site- aggregate Cormorant numbers specificsingle record. on a large geographic scale — is the harmonization of a national/ Mean numbers are often international counting date and the regularly reported too. In use of methods for simultaneous practice the term ‘mean’ or data collection (as a response to ‘average number’ is used the high mobility of Cormorants). most commonly in relation to Cormorant numbers. The Data collection on a national ‘mean number’ of birds can be level calculated from regular counts National coordinators (at the carried out in a standardized country level) are key players way. Numbers presented thus in organisation and information should give the best possible transfer. They should be estimate of birds present in any responsible for the distribution spatially-defined area (e.g. site, of, as well as the collection of, region or on the national level) counting forms to/from observers. within a given time period. Regional coordinators should be involved in the often lengthy Text Box 3.3 Conventions and process of assisting to build-up definitions.

www.intercafeproject.net [25] the intercafe field manual

pan-European wide Cormorant and definitions used by researchers When discussing ‘Cormorant roost site count. To do this, one dealing with Cormorant numbers. numbers’ with respect to can take advantage of, and gain wintering birds, a number of support through, one of several associated specific pieces of 3.4.1 Counting Cormorants existing international waterbird information should also be census networks (e.g. Wetlands on their foraging grounds given in order to make the International). Named European and calculating ‘Cormorant numbers meaningful. These coordinators must be identified and are (i) the reference area under days’ they are responsible for an allocated consideration, (ii) the relevant list of countries from which they have This type of data will be of interest time period of data collection, to find (and keep in close contact for assessing Cormorant numbers (iii) the frequency of counts (i.e. with) national coordinators. Final using any kind of spatially- a single or multiple count), and counting results on a national level, defined unit, particularly for (iv) the counting method used together with a map indicating the property-related issues and fishery (i.e. a day or roost count?). So, distribution of the roost site locations for example, when talking about should be addressed to European Cormorant numbers in ‘Europe’, Generally, two distinct methods coordinators, who can ultimately the names of the relevant can be used to collect data on produce a European synthesis. Member States, the time period Cormorant numbers at very under discussion, and the time specific sites. Both address of year (e.g. breeding or mid- slightly different issues or winter) should be indicated. 3.4 How to use numbers? address different questions in relation to different spatial ‘How many Cormorants do scales. When considering Cormorant you have?’ is certainly the most numbers on a wider geographical- frequently asked question in the Day-counts provide data on scale, a single figure or count ongoing public debate over potential the bird numbers and their result obtained during one census ‘impact’ of Cormorants at fisheries. activity pattern (e.g. in feeding done in mid-winter (e.g. mid- It seems to be a simple question but areas). There are two possible January) should not be used in practice needs to be more specific methods: (1) the observer stays to calculate the mean number before a ‘correct’ answer or ‘best in one location and monitors of birds present over a whole estimate’ can be given. For example, the activity of birds over a winter-migration period (i.e. which time of year is being given time period, and (2) over several months) within any considered? What geographical area the observer moves around specified area. This is because, in do the numbers refer to? Are we ‘controlling’ (i.e. counting and any given area, numbers fluctuate considering breeding or wintering recording birds) in a specific throughout the year, in relation to birds? Do the numbers available area, a linear transect, or a the annual cycle or in response to refer to the numbers of individuals defined number of locations, in temporary climatic conditions. or to breeding pairs? Are counts a standardised way. available for specific races? Text Box 3.4 Area under consideration and the frequency of counts. Roost-counts provide data on Depending on (1) the goal of the the number of birds present research, (2) the counting effort, or aggregating from a defined of count results on a country level and (3) the protocols for data area (activity range). This is the should be the responsibility of the collection, both the ‘quality’ of the preferable method for use during national coordinators. data as well as its interpretation the migration period, and it also can differ in relation to a variety describes the phenology of birds Data collection on a European of questions. The following three on a wider scale. level Text Boxes (3.3, 3.4 and 3.5) intend Much more effort and organisational to give brief overviews of some of Text Box 3.5 Basic methodology — work is needed to undertake a the frequently used conventions ‘day’ versus ‘roost’ counts.

[26] the intercafe field manual

management discussions on the than once. To this end it is always ensure that counts are standardised ‘impact’ of feeding Cormorants on worthwhile to record additional and comparable (even if they do fish stocks. features while counting — for not give the actual ‘total’ number of instance, the numbers (and flock birds involved). Most people come into contact size) and directions of birds moving with Cormorants when the birds are within the site, and the locations Intuitively, the logical step after seen at foraging sites. As foraging of known roost or resting (loafing) an accurate Cormorant count sites are the places where feeding sites — in order to accommodate is obtained (particularly from a Cormorants frequently come into these in the final ‘best estimate’ of fisheries perspective) is to convert conflict with commercial and/ bird numbers. this count to some measure (or or recreational fisheries interests, index) of Cormorant use of the this is the starting point for our Even such basic counts can be site. At its simplest, this could be discussion on counting Cormorants. time consuming and may take a simple calculation of the number At a small site, counting is several hours, a factor that must be of Cormorants counted and the relatively straightforward, if an considered when planning counts. number of days they visit (and observer finds a good vantage point It is also necessary to consider at presumably feed) at a site. This overlooking the whole site. which time of day counts should be ‘predation pressure’ is sometimes undertaken. In general, maximum calculated as: The waterbody should be scanned counts at the same site will vary systematically and a cumulative between morning, mid-day and Number of Cormorants x days count made. At any time, some afternoon periods and this variation of presence at a site birds will be arriving at the site may be unpredictable. Maximum and some leaving, whilst others numbers can occur during a There are two possible ways of will be diving underwater and different time period on different calculating this, as follows: excluded from counts. Therefore days or even during the same day. it is necessary to undertake several For instance, a study at Loch Leven 1. By considering the number systematic counts of the water in Scotland (where Cormorants of birds present on the count and take the average number, were counted three times a day for day as a constant and then the maximum, or the modal (i.e. 106 days) showed that some counts multiplying it for the number the most frequently counted at one time of day were up to 40% of days elapsed until the total) count. With care, the most higher than other counts made at following count, or commonly counted total will be the different times on the same day same as the maximum count — but (Wright 2003: 347). 2. By linear interpolation of whatever this ‘final’ count is, care two consecutive counts should be taken to note how it was Prior ecological knowledge of (n1 + n2) derived. how the birds use a specific site x d 2 may allow an observer to choose As waterbodies increase in size, it to count birds at the time of day becomes more difficult for a single when their numbers are greatest. where n1 and n2 are the number observer to make complete, accurate However, this is not always the of birds counted at time 1 and 2, counts. In these cases, a single case. The timing of counts will, and d is the number of days that observer (or several) could attempt of course, be constrained by have elapsed between the two to count all birds from different other factors affecting the time consecutive counts (Im & Hafner vantage points and amalgamate their period available for an observer to 1984, 1985). Essentially, this counts on a final total. However, undertake them. Nevertheless, the just means that for days between care must be taken that counts most important requirement is that successive counts, it is assumed are synchronised to reduce the multiple counts undertaken over a that the average number of birds chances of birds moving between specified time period at a specific (from the 2 counts) were present the observation ranges of the site are undertaken at the same time on the days when counts were not observer(s) and being counted more of day each time. This will at least made.

www.intercafeproject.net [27] the intercafe field manual

Example A In Example A, both methods of estimating Cormorant-days 60 underestimate the actual number by varying degrees (22% 50

r underestimate for method [1] and 14% underestimate for method [2]). 40 Conversely, in Example B, both methods of estimating Cormorant- 30 days overestimate the actual number by varying degrees (29% 20

Maximum cormorant numbe overestimate for method [1] and 12% overestimate for method [2]). 10 In both examples, Cormorant-days estimated by Method (2) which 0 12345678910 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 attempts to be more ‘biologically Day number meaningful’ — by taking an Figure 3.8 Hypothetical daily Cormorant counts (maximum number observed) at a average of each subsequent pair of foraging site. Birds were counted by a trained observer at a standard time of day and counts — still either underestimates weather conditions remained similar throughout the 36 day counting period. or overestimates the number of Example A Table 3.3. Cormorant-days by over 10%. Although such levels of accuracy are common in ecological datasets, However, each method has its maximum numbers of Cormorants it must be remembered that these limitations as can be seen in the recorded were 7, 11, 56, 11, 7 and 9 ‘best estimates’ are just that — best following hypothetical examples (on each day) and so the cumulative estimates. where Cormorant numbers are number of Cormorant-days derived known on a daily basis for a period from these 6 counts estimated by As these examples demonstrate, of 36 days (Figures 3.8 and 3.9) but methods (1) and (2) are 653 and the level of fluctuation (i.e. the we assume, in the first instance, that 565 individuals, respectively (see difference between the lowest and birds have only been counted once a also 7-day data in Table 3.3). highest counts in the dataset), as week, on days 1, 8, 15, 22, 29 and 36. Example B For this full 36-day count, the actual cumulative number of 70 Cormorant days is 997. Taking 60 the counts obtained on days 1, 8, r 15, etc., the maximum numbers 50 of Cormorants recorded were 6, 40 22, 8, 23, 48 and 30 (on each day) and so the cumulative number of 30

Cormorant-days derived from these 20

6 counts estimated by methods Maximum cormorant numbe (1) and (2) are 779 and 857 10 individuals, respectively (see also 0 7-day data in Table 3.3). 123456789101112131415161718192021222324252627282930313233343536 Day number

For this full 36-day count, the actual Figure 3.9 Hypothetical daily Cormorant counts (maximum number observed) at a cumulative number of Cormorant foraging site. Birds were counted by a trained observer at a standard time of day and days is 506. Taking the counts weather conditions remained similar throughout the 36 day counting period. obtained on days 1, 8, 15, etc., the Example B Table 3.3.

[28] the intercafe field manual

Table 3.3 Maximum daily Cormorant count every 7 days, actual cumulative The potentially large movement Cormorant-days (for daily counts) and estimated cumulative Cormorant-days from 2 (or ‘turnover’) of Cormorants at different methods. See text for full details. specific sites is difficult to capture in most counts. Being familiar with Example A the local situation and phenology Day number 1 8 15 22 29 36 helps to quantify this aspect in Maximum daily Cormorant 6 22 8 23 48 30 planning counting regimes. count Actual cumulative number 6 140 331 397 715 997 For example, at Loch Leven in of Cormorant days Scotland, based in part on the (1) Estimate of cumulative 6 64 204 275 461 779 movements of radio-tracked Cormorant days Cormorants, it was estimated 22% underestimate that the actual number of birds (2) Estimate of cumulative 6 112 210 329 593 857 Cormorant days passing through the site was 14% underestimate probably ten times the mean Example B number counted there at any one time (Wright 2003). Similarly, Day number 1 8 15 22 29 36 using re-sightings of ringed birds Maximum daily Cormorant 7 11 56 11 7 9 and concurrent systematic counts count of Cormorants at a roost in Lake Actual cumulative number 7 75 223 380 452 506 Geneva, Frederiksen et al. (2003) of Cormorant days estimated that the site was actually (1) Estimate of cumulative 7 60 182 529 602 653 used by 66% more birds over the Cormorant days 29% overestimate season than were recorded there during the peak count. These (2) Estimate of cumulative 7 61 262 463 517 565 Cormorant days two labour-intensive studies 12% overestimate perhaps give the main message of this section — that counts of Cormorants at any particular site well as the gradient of the curve Cormorants, and the variation in are, in effect, just snapshots of the (i.e. the ‘speed’ at which numbers their daily presence there. As a actual situation there. Clearly, the are increasing or decreasing) general rule, it is recommended more frequently these snapshots are can easily influence the result in to schedule counts as follows. taken, particularly during the main either direction. Clearly, the most First, once every week or every periods of Cormorant movement, accurate assessment of Cormorant ten days at fisheries/feeding the more accurate the picture of numbers (and hence, of Cormorant- areas of little apparent interest to Cormorant numbers at any one site. days) comes from standardised Cormorants, or where Cormorant Similarly, even ‘complete’ counts daily counts but such frequent presence is known to be pretty at a site are merely a piece of a counts are not always possible and constant. Second, two or three much larger jigsaw — the ultimate so most field data are a compromise times per week at fisheries/feeding size of which is determined by between the time available for areas where Cormorant presence the frequent short- and large-scale its collection and the best (i.e. is heavy or when Cormorant movements of birds on both a daily most biologically meaningful) presence is highly variable (e.g. and seasonal basis. interpretation. during migration periods). Finally, counts should be conducted every It is important to note that the Bearing this in mind, the day at fisheries/feeding areas ‘turnover’ of Cormorants may well frequency at which counts are where damage prevention is have little influence on the total carried out should be adapted to carried out by means of lethal or number of ‘Cormorant-days’, or the requirements of any study, non-lethal techniques in order to ultimately on the estimated ‘impact’ the use of the waterbody by assess effectiveness. on fish stocks, calculated for a

www.intercafeproject.net [29] the intercafe field manual

foraging site (see above) because study, arguably the most useful laying pair is to follow individually birds leaving are likely to be merely single parameter to obtain (e.g. for recognisable nests from egg-laying replaced by other individuals and estimating total production of young (or early incubation) until all chicks thus the maximum number on the site and for inclusion in population have reached fledging age. at any time will remain fairly stable. models) is the number of young that fledge per egg-laying pair, as this Where multiple visits can be made However, these examples highlight perhaps best represents reproductive to a colony, nest failures (i.e. that day-counts at foraging sites are output per breeding female. We those breeding attempts for which not suitable or recommended for define ‘fledging’ here as young success = zero) can be identified assessing Cormorant ‘population’ birds leaving the nest, and not the and included in the calculations. numbers on a wider scale, where subsequent independence of the Therefore, multiple visits allow coordinated simultaneous roost chicks from the adults which can estimates to be produced that are counts may be more appropriate. often be problematic to determine. a good reflection of reproductive output per nest or clutch and, as Given the intrinsic difficulties such, allow for reliable between-year described here of attempting to and between-colony comparisons. count highly mobile birds such as Cormorants, there are two other Minimising disturbance situations where Cormorants can All unnecessary visits inside a be counted. Cormorants (at least, colony during egg-laying and sexually active adults) congregate incubation should be avoided in colonies to breed and during because disturbance at this stage the rest of the year all birds gather can result in nest desertion. on roosts to spend the night. The Disturbance also increases the previous sections describe how probability of egg and chick to take advantage of the social predation (e.g. from gulls Laridae behaviour of the birds in order to or crows Corvidae) as well as best assess their numbers. Figure 3.10 Nest with newly- mortality through chilling or heat hatched chicks and eggs. stress. Visits inside a colony should Photo courtesy of T Bregnballe. always be kept to an absolute 3.5 Estimating breeding minimum, with parents disturbed success In the following we describe for no longer than 30 minutes, methods that workers should use to preferably less. Where there are In this section, we discuss some obtain an estimate of the number chicks of four weeks of age or central ecological issues relating to of young fledged per egg-laying older, disturbance could potentially Cormorant population dynamics. pair (referred to as ‘fledgling force premature fledging, so unless These include such questions as production’) or of brood size around it is necessary to enter the colony, How many ‘active’ nests are there the time of fledging (referred to as observations should be made from in a colony? How many eggs/young ‘brood size’). The description and a distance using a telescope. If data are there in the nests? How many recommendations are mainly based cannot be obtained without entering young survive to leave the nest (i.e. on Newson & Bregnballe (2003). a colony and you have no previous ‘fledge’)? experience ask for advice or read 3.5.1 Fledgling Blackburn (1999). An important aim of any study of avian breeding success should production — Multiple Whenever possible, visits (where be to obtain estimates that are observations of individually disturbance is unavoidable) should comparable both between colonies recognisable nests be made at dawn or mid-late and between years. Whilst the afternoon before, or after feeding, most relevant parameters to record The best method to estimate the to reduce food loss from chicks depend on the specific aims of the number of young fledged per egg- by regurgitation (see Hughes et

[30] the intercafe field manual

al. 1998). Some researchers have this time, and go on to fledge one may be necessary to provide visual found that disturbance can be kept or more young). However, if the markers within the colony to enable low by visiting ground nesting colony is visited at a time when individual nests or colony sections to colonies at night. nestlings are at an advanced stage, be re-identified on subsequent visits. the oldest chicks in a brood may Timing and number of visits already have fledged and be sitting Sample size It may be useful to visit the colony away from the nest and so not be If the fate of only a few individually prior to the expected egg-laying included in records of nest contents. recognisable nests is being followed stage to estimate the best date for until chicks fledge, it may also be the first recording. This can be Monitoring nests at an advanced helpful to record brood size for assessed by examining the stage stage may also be difficult at a larger number of nests during of breeding birds (i.e. whether ground-nesting colonies where the colony visits prior to fledging. they are nest building, incubating older chicks are likely to wander Although, it is not possible to etc.). In this way the worker could from their nests, making estimation combine the two data sets and also gather additional information of brood-size difficult. There is analyse them statistically, it can about the timing of breeding in that obviously a trade-off between provide useful information on the particular year. Timing of breeding collecting reliable data and concordance between the mean can vary significantly between keeping researcher effort and brood size at/near fledging in the years for individual colonies and colony disturbance to a minimum. nests monitored throughout the between colonies in a particular However, visiting a colony at about breeding season and the mean year (e.g. Newson et al. 2005). ten-day intervals should result in brood size at/near fledging in nests little loss of data. that were not monitored. The colony should be visited regularly from egg-laying to Mapping of nests Specific areas within a colony are fledging. If this is not done, Breeding attempts are sometimes likely to differ in their ‘quality’ information about the losses of initiated later in the breeding season and as such, experienced and both clutches and broods can not be some time after most pairs have inexperienced birds are not likely to collected. Even where regular visits already started to breed, often be distributed at random (van Eerden to a colony throughout the season by younger or less experienced et al. 1991, Bregnballe & Gregersen can be standardised, factors out of breeding birds. In order to identify 2003, Kopciewicz et al. 2003, Krag the researcher’s control (e.g. nest and include any such breeding 2003). For this reason care should collapse rates, which are likely to attempts, it is necessary to map be taken to sample from several differ between colonies and years all nests in the sample area and subsections within the colony. due to changing environmental highlight the ones being monitored Ideally, one should aim to collect conditions) can have a large affect (see Figure 3.11). In ground- data from at least 30 nests from any on production estimates even if nesting colonies in particular it discrete section of the colony. the actual fledging production of successful nests is constant. What to record during visits? However, if monitoring broods Position of nests. For tree-nesting at regular intervals throughout colonies it can be relevant to development is possible, this record where on the tree the nest is should reduce the risk of such located. This is because the quality errors. of different areas of the colony and/ or birds present in different areas Obviously, visiting the colony at a of a colony may differ. For ground- time when the majority of nestlings nesting colonies it can be relevant are close to fledging will provide Figure 3.11 Sketch from a section of to record whether or not the nest the best estimate of fledgling one nesting tree within a Cormorant is in the periphery of the colony production for successful nests (i.e. colony. Drawing courtesy or near to its core. This is because nests that have not failed before of J Gregersen. some studies have shown that nests

www.intercafeproject.net [31] the intercafe field manual

in the periphery are more exposed 1 Figure 3.12 Chicks of different to predation than those closer to age — from just hatched to just the centre (Siegel-Causey & Hunt fledged. Photos courtesy of Stef van Rijn. 1981, Quintana & Yorio 1998). 1. Eggs and nestling of 0–1 days old (colony of Enkhuizen, IJsselmeer Clutch size. Collecting information area, the Netherlands 24 April 2006), on clutch size (i.e. number of eggs 2. Nestlings of 5–7 days old (colony per nest) can become relevant of Enkhuizen, IJsselmeer area, the if workers find differences in Netherlands 24 April 2006), brood sizes at fledging and want 3. Nestlings of 10–12 days old (colony of 2 to know whether these are likely De Kreupel, IJsselmeer area NL 14 April to be consequence of differences 2006), in the number of eggs laid or a 4. Nestlings of 18–20 days old (colony of consequence of differences in Kivilaid, Estonia 17 June 2007), partial chick loss. However, if 5. Nestlings of 23–26 days old (colony of the aim of research is merely to Pohja Malusi, Estonia 19 June 2007). estimate the number of young fledged per egg-laying pair, clutch-size data do not need to be number of predators. In tree-nesting collected. 3 colonies it is common that Magpies Pica pica and Crows predate eggs. Ageing of chicks. Where hatch Foxes Vulpes vulpes and Martens dates are not known, the age Martes spp. may also be important of Cormorant chicks can be predators in some colonies. determined in the field according to both their size and feather In ground-nesting colonies, gulls development (van Rijn et al. 2003). are usually the main predators of The downy feathers start growing eggs and nestlings. Great Black- from about the 6th day after hatch. backed Gulls Larus marinus (but At 10–14 days the chicks are 4 also Herring Gulls Larus argentatus covered by brown-blackish woolly and Lesser Black-backed Gulls down, and growth of the tail and Larus fuscus) are particularly flight feathers starts from about known to take Cormorant chicks. 14–20 days of age (Olver & Kuyper Great Black-backed Gulls may 1978; del Hoyo et al. 1992). The take chicks up to an age of at least chicks stay in the nest till they are 35 days. Foxes may also predate about 50 days old and fledged. eggs and chicks if they are present on nesting islands or if they can Predators. To help interpret colony 5 gain access to it when Cormorants and/or year-to-year differences in are breeding. In some countries estimates of fledgling production, (e.g. Finland, Russia, Estonia, it is often very helpful to have Germany, Sweden and Denmark) knowledge of predators in the White-tailed Sea Eagles Haliaeetus vicinity of the colony and the albicilla may occur as predators extent of their predation. It is of young, and occasionally adult, therefore useful to record the Cormorants (Koryakin & Boyko presence of potential predators 2005, Leihikoinen 2006). and any incidents or indications of predation. Cormorant eggs and Weather. It is often also helpful chicks are potentially exposed to a to make notes about the weather

[32] the intercafe field manual

up to, and during, the breeding 3.5.2 Fledgling production: been lost (cold eggs or dead chicks season. For example, a particularly a single visit to the colony may still be present in the nest); harsh winter may explain the (b) the chicks have fledged and late onset of the breeding season, Some researchers judge this method left the nest, and (c) eggs or small whilst heavy rain and/or high to be inappropriate for estimating chicks are present (but not visible) winds may cause the sudden loss mean fledgling production per nest and the parents are absent. Be of clutches. or clutch. Nevertheless, where aware that adults may sit in empty multiple visits to a colony are not nests (e.g. deserted nests and nests Analysis of data possible, data collected on a single from which chicks have fledged Where multiple visits are made visit can still provide an important earlier in the season). to a colony, the easy option is to indication of breeding success in a calculate and report (1) the number particular year. Estimates, assumptions and of young raised to fledging (or as decisions near to fledging as possible) per One of the uncertainties of this To estimate the number of young egg-laying pair; (2) the proportion method is that mean brood size fledged per egg-laying pair or per of clutches that were unsuccessful estimated from a single visit will nest, the estimated proportion of (i.e. the proportion from which no be strongly dependent on the stage nests that produced at least one young fledged); (3) the brood size of nestling development at the time fledged chick (i.e. the proportion around the time of fledging (only of the visit. Therefore, the age of of successful nests) is multiplied including nests that had at least chicks should be estimated for by the estimated mean number one young); (4) the proportion of nests where brood size is recorded. of chicks assumed to fledge per unsuccessful nests that were so Furthermore, even though the successful nest. because all eggs where lost and the number of empty nests is recorded proportion that were unsuccessful in the section of the colony A number of decisions have to because all chicks died before sampled, it will not be possible be made when estimating the fledging. Another possibility is to obtain a precise estimate of proportion of successful nests. to apply a modification of the the number of complete losses of Mayfield Method (Mayfield 1961, eggs or chicks. Thus nests where For the apparently empty nests 1975) or a similar approach, to all eggs were lost or all chicks (see a-c above), a decision must be account for partial brood losses (as died may have disappeared prior made as whether or not to assume in Newson & Bregnballe 2003). to the visit and some empty nests that nests appeared empty because Improved methods now exist for may be empty because the young (i) the chicks had fledged and left relaxing the assumptions of the had fledged and left the nest (see the nest; (ii) all eggs or chicks had Mayfield method (such as assumed below). been lost; or (iii) eggs or chicks constant survival over a specified were present but invisible to the period) and for accounting for Field method observer. The proportion of empty potentially important sources of The observer should aim to visit nests that were so because the variation in nest-survival data the colony when the majority of chicks had fledged and left the nest (e.g. Rotella et al. 2007). Use of chicks are 30–40 days old (see may be estimated from observations the Mayfield Method increases above concerning minimising of presence of fledged chicks (i.e standardisation and comparability disturbance). Brood size should those not sitting in nests) inside of fledging production estimates. be noted for as many nests as and outside the colony. It may be When reporting results, it is possible and mean age of the chicks reasonable for some colonies and important to state the limitations estimated for each brood. Within years to assume that the majority of the data (e.g. sample size, selected sectors of the colony, of the late clutches or broods with representativeness of sample) all nests within the following small chicks will be lost before and any assumptions made (e.g. categories should also be recorded. fledging. Some studies show that assumed constant nestling survival few of the very late clutches in over specified period) in the Those apparently empty nests Cormorant colonies produce fledged analyses. where (a) all eggs or chicks have young (Bregnballe 1996).

www.intercafeproject.net [33] the intercafe field manual

For the nests containing eggs mentioned in full when reporting data when reporting results. For or chicks, a decision must be the results of the estimates. example, if mean brood size is made as to how best to estimate recorded prior to fledging, it is the probability that the breeding Analysis and reporting of data important to report the ranges of attempt will be successful (i.e. that When analysing data, workers ages included (e.g. 35–45 day old at least one of the chicks in the nest should combine an estimate of the chicks), the mean age of chicks and will fledge). A decision must also proportion of nests that produced the variance around this measure. be made as to whether mean brood no fledged chicks with an estimate As timing of breeding will vary size (e.g. of 35–45 day old chicks) of the number of chicks fledged in between years and colonies, it can provides a fair estimate of brood nests where at least one chick did be difficult to obtain comparable size at fledging. These decisions fledge. As always, it is important estimates of brood size from a (basic assumptions) must be to state the limitations of the single visit each year.

[34] the intercafe field manual

4 CORMORANT DIET, FOOD AND ENERGY INTAKE D N Carss, M van Eerden, M Enstipp, M Govedicˇ, R Gwiazda, J Trauttmansdorff

4.1 Introduction of each, following Carss et al. This chapter is presented in two (1997) carefully. This paper (from parts. The first covers methods From their prey choice as assessed Wetlands International’s Cormorant of assessing diet, the second from dietary studies, Cormorants Research Group) attempts to reach takes this dietary information a are known to be more or less consensus on methods and is used step further and considers the generalist predators. This means by researchers across Europe and conversion of fish as food to fish as that they usually consume the most beyond. In this chapter, we pick out energy — the ultimate requirement abundant fishes or those that are the salient parts of the Carss et al. of all living things. Energetic most easy to catch at a particular (1997) review but would also urge considerations are important for foraging location. Usually the interested readers to consult the several reasons, perhaps most composition of the prey taken by original paper directly. Despite the importantly they will influence Cormorants and that present in the fact that the review was published both the ‘quality’ of an individual foraging location are fairly similar over a decade ago, much, if not (i.e. its capacity to fly, move and so the composition of the birds’ all, that it says on both the notion between habitats, breed etc.) and diet can give a reasonable picture of of rigorous scientific data on how much prey a Cormorant eats, the fish stock and its composition in Cormorant diet and food intake (and its daily food intake. This is an the waters where the Cormorant has hence, the need for standardised important biological issue when foraged (e.g. Hald-Mortensen 1997, methodologies), and the rationale considering predation at fisheries Carss & Marquiss 1997). for obtaining such data in relation but such information also has to the ongoing public debate about political value (see Text Box 4.1). There are several different methods Cormorant ‘impacts’ at fisheries, available to those who want to is still highly relevant today. Given In relation to Cormorant diet, the assess Cormorant diet. Birds can this, we have occasionally quoted choice of method will depend on be observed foraging, cast oral sections of Carss et al. (1997) the aim of the research, the time pellets (containing hard, undigested verbatim in this chapter but always available and the amount of funding prey remains) can be collected and with the relevant citation. available (see also Table 4.1). analysed, as can regurgitations (containing partially-digested ‘meals’) collected at roosts or, most Table 4.1 Basic information provided by different methods of Cormorant diet commonly, from nestlings during assessment. Yes = Y, No = N, Potential bias = ?; further methods concerning daily visits to the colony. Finally, the food intake can be found in section 4.7 (Bioenergetics). stomach contents of dead (usually Observations Pellets Stomachs Regurgitates shot) birds can also be examined. Prey species (list) ? Y Y Y Here, we discuss these different Species composition N Y Y Y methods and highlight the Fish length/weight Y (?) Y (?) Y Y (?) advantages and disadvantages Daily food intake N Y (?) Y (?) N

www.intercafeproject.net [35] the intercafe field manual

To observe Cormorants while they Sometimes claims of ‘impact’ are contested. Thus, as well as being are foraging needs, above all, time, of scientific interest, knowledge of Cormorant diet (linked — through but the method is restricted to an understanding of fish biology and fisheries science to — ‘impact’) recording only those fish that the also has political value. This is particularly true where there are calls to bird brings to the surface to swallow control bird numbers in order to protect fisheries. As Carsset al. (1997: as small fish are often swallowed 198) point out: underwater (Strod et al. 2003). For those fish brought to the surface, it ‘There are calls to devise a pan-European management plan for may be possible to identify both the Cormorants and to ‘control’ birds by culling, despite the scarcity of species involved and to estimate its scientific evidence for any detrimental effects of fish predation on length. The use of such methods is natural water bodies. Attempts to provide such evidence, for any somewhat limited but if an observer fishery type, in any country, necessarily require rigorous estimates wanted to assess the hunting of diet, including such things as fish species composition and size success of Cormorants on a water distribution. Moreover, knowledge of foraging ecology is also body with a known fish stock (e.g. essential when assessing possible impact, particularly information on a Carp Cyprinus carpio pond) this daily food (and energy) requirements, diet shifts, prey selection, the could be a good method. It also has influence of foraging habitat, and any seasonal and annual variation. the advantage that it should not be a Thus knowledge of Cormorant diet, food and energy intake has both disturbance to the birds. scientific and political value.’ Collecting pellets under roosts Text Box 4.1 Knowledge of Cormorant diet is an essential element to assessing any or colonies is quite easy, but it potential impact of these birds at fisheries. takes considerable time to analyse them and a reference collection of fish bones and other hard parts is per day (Zijlstra and van Eerden older birds as a defence reflex when almost certainly required. There 1995), the method has been used to disturbed — can be treated in the are serious concerns over possible reconstruct daily food intake. same way as stomach contents. The biases in this method, as some assumption being that the bird has hard parts may not survive the In old, established colonies or regurgitated the full contents of digestion process (being digested roosts, sample collection should its stomach. Possible bias occurs completely or severely eroded be restricted to the early morning as regurgitated material may by it). Nevertheless, pellets are because during the day other consist of the larger fish taken. still a useful sampling technique mammal (e.g. Foxes Vulpes vulpes, Partially-digested material from because they potentially sample a Wild Boar Sus scrofa) or bird earlier feeding trips may not be large number of individual birds species (e.g. Coot Fulica atra, included in regurgitations. Stomach through time. One sample may Gulls Laridae, Herons Ardeidae) contents analysis obviously thus contain pellets originating can feed on the leftovers from the requires dead birds (most often from several days before. Bearing Cormorants. The collection of these shot), and these may not always in mind the potential biases in the samples causes no, or very little, be available. The work involved method, pellet analysis can give a disturbance for the Cormorants. In in regurgitation or stomach broad, relative (at least qualitative) many situations, for example large, contents analysis is slightly more overview of which fish species extensive lakes or large rivers, where onerous (and smelly) than that are eaten in a particular area. It is observing foraging birds is difficult, for pellets. However, because possible to estimate the length of areas where birds cannot be shot, or stomachs and regurgitations often fishes from their remains in pellets during non-breeding period, pellets contain relatively fresh food but this requires considerable may be the only means of assessing material, this method — if used effort and is restricted only to the food of Cormorants. carefully — probably gives the best those hard parts that show little assessment of Cormorant diet in sign of digestion (erosion). As In many ways, regurgitations — terms of both prey composition and each bird produces one pellet often produced by nestlings or size-distribution of fishes.

[36] the intercafe field manual

4.2 Feeding observations However, in captivity, experiments a useful method of obtaining a have demonstrated incontrovertibly description of their diet in (at least) As Carss et al. (1997: 209–210) that Cormorants can swallow fish qualitative terms. It is possible point out, ‘Direct field observation (of maximum length 8 cm in this to estimate the daily food intake of foraging Cormorants to record case) underwater (Strod et al. of birds by this method but only the prey they consume has several 2003). Thus, given this potentially if one can quantify the rate of advantages. It is non-destructive serious error in diet assessment, erosion of material recovered and can provide large amounts it is not possible to determine from pellets. Care must be taken of data with little disturbance to Cormorants’ daily food intake from when interpreting the results of the birds. A major advantage over direct observations. pellet analyses as there are serious other methods is that spatial and potential biases, erosion of otoliths temporal variation in diet might be and other hard parts being one of assessed with some accuracy (Davis the main potential errors. & Feltham, 1996), with data on 4.3 Pellet collection and prey and precise feeding locations analysis being collected simultaneously. 4.3.1 Collection and Cormorants regurgitate the In addition to dietary information, undigested parts of their prey conservation observations can also provide data (e.g. bones, scales or ear-stones on diving behaviour and an index Pellets should be collected as soon ‘otoliths’) as oral pellets, covered of foraging performance. It may as possible after they are produced in mucus and produced from be difficult to get close enough to the stomach lining (Figure 4.1). birds to enable the identification Pellets are relatively easy to collect of all prey items caught during a under night roosts or in breeding feeding bout. Biases will be further colonies and are sometimes the compounded if particular prey only available means of assessing species are more, or less, easy to the food of Cormorants. This is recognise. This problem may be particularly acute at fisheries with a diverse fish fauna. It may be possible to reduce such potential biases by categorising fish on their type (e.g. Cyprinid/flatfishes, Davies & Feltham, 1996) or body shape. Errors could also occur when attempting to estimate the size of fish caught, especially when judging the lengths of smaller prey items, which generally have shorter handling times. The sizes of fish caught by birds during observations are usually estimated in relation to bill or head length. Despite the common use of this technique, there have been few attempts to quantify observer bias under experimental conditions. The possibility of birds swallowing items underwater must be considered. The frequency at which this occurs has yet to be fully Figure 4.1 Two Cormorant pellets, the undigested fish bones can clearly be quantified.’ seen within the mucus coating of the pellet. Photos courtesy of S van Rijn.

www.intercafeproject.net [37] the intercafe field manual

by the birds. Ideally they should be chewing pads, lower jaws or collected as soon as the Cormorant vomers are picked out manually flies off. The area must be inspected using fine tweezers. Fish scales thoroughly in order to recover the should only be picked out if there small pellets as well as it is thought are no other structures present. that large pellets are often more Within individual fish there are commonly found and collected actually three pairs of otoliths by researchers as they tend to be and some of these chalky-looking easier to locate. Pellets should also structures are relatively easy to be intact, especially if collected identify, count and measure when from beneath trees where the fall extracted from pellets. In the case may cause then to disintegrate. of Cyprinid (Carp family) fishes, During collection, each pellet the large otoliths recovered in must be kept separately in a plastic pellets are either the asterisci or bag and they can then be held in lapilli, whereas the otoliths from a deep freezer for a longer period other groups of fishes are usually of time. This is necessary when the sagittae. The eye lenses of one wishes to compare individual fishes are also often found in meals of individual birds. Mixed pellets. Although it is seldom samples can be collected if only a possible to identify the fish species qualitative description of the prey from these structures, lenses can be species composition and size of counted (to estimate the minimum a wider ‘population’ of birds is numbers of fish) if they outnumber needed. It is possible to collect the other undigested key remains. pellets at daily or longer intervals It is not necessary to disinfect the Figure 4.2 A dissected Cormorant pellet. depending on the research question. otoliths or most other hard parts Photo courtesy of J Trauttmansdorff. but the pharyngeal bones can 4.3.2 Preparation and be dipped into a 3% solution of the pellet and left until the next hydrogen peroxide (H O ) for a day (for 15–20 hours) by which analysis 2 2 few seconds and dried afterwards if time the mucus has dissolved and Pellets, previously kept frozen required. becomes transparent. The solution at - 20°C, can be placed at room is then passed through a 0.3 mm temperature for a few hours to In some instances, pellets may be sieve and all the remaining otoliths thaw. They should then be removed difficult to tease apart and so they and other undigested remains can from the plastic bag and placed into can be dried at room temperature be picked out. Care must be taken a Petri dish and the bag washed for a few days before being with the use of NaOH as it is out with water passed through a ‘crumbled’ carefully into a Petri capable of dissolving or softening 0.3 mm sieve, any remains can then dish. Alternatively, they can be some important structures such as be added to the rest of the pellet in soaked in an enzymic washing chewing pads and bones. the Petri dish. This process is time- powder or NaOH solution to consuming but is important because dissolve the binding organic mucus Pellets will almost certainly contain some pellets fall apart during matter. This approach avoids the the remains of several fishes thawing and can get stuck inside possibilities of smaller otoliths or and it is often possible to both the plastic bag. The distinctive other remains being overlooked identify these to species level and undigested items in pellets (Figure in any folds in the mucus pellet. to estimate their original size by 4.2) must then be examined under a A 9% NaOH solution (90 g of careful examination of key bones stereomicroscope. granules dissolved in 1,000 ml of recovered from the pellets. Many water) or a saturated solution of of these key bones occur in pairs Otoliths and other bony identifiable ‘biological’ (enzymic) washing within individual fish (i.e. right-and parts such as pharyngeal bones, powder should be poured over left-handed bones from either side

[38] the intercafe field manual

Figure 4.3 Dried contents of a Cormorant pellet extracted from mucus. Photo courtesy of J Trauttmansdorff.

Figure 4.4 Dried contents of Cormorant pellets (mostly otoliths) awaiting identification, measurement, and counting. Photo courtesy of S van Rijn.

four small fish and two large ones (i.e. six individual fish in total). In all cases, the number of fish recorded in a pellet is determined on the basis of the most numerous structure found in the undigested remains. It is not always possible to of the fish). The commonest key the original length of the fish by determine the species represented bones recovered from pellets are use of standard bone length:fish in pellets by examination of their the pharyngeal teeth of Cyprinids, length equations, although precise otoliths, particularly if these are lower jaws (dentaries) of other errors associated with this are partly dissolved (eroded through groups, and the ear stones (otoliths) currently unknown (Carss et al. the action of digestion). And so (in of all species. Other remains 1997). With practice, these key the absence of such species-specific commonly recorded include bones can be picked out of pellets, structures such as pharyngeal teeth) vertebrae, vomers (bones from the where necessary they can be these records should be reported roof of the mouth in many species), paired (i.e. right and left bones as ‘unidentified Cyprinid’ for chewing pads from Cyprinids, eye of the same species and size and example. However, species can de lenses, and fish scales. so assumed to be from the same determined by careful examination individual). For example, where of pharyngeal bones and chewing The size of many of these hard a pellet contains four small left pads (see section 4.3.3). parts can be measured when otoliths, three small right ones, and dry (Figure 4.3), and these a further two large right otoliths, Ultimately, the minimum measurements used to estimate the record for this pellet should be numbers of each fish species

www.intercafeproject.net [39] the intercafe field manual

can be recorded for each pellet. represented in the pellet sample can structure — the basioccipittale Thus for each specific sample of be assessed, and each species can bone (see Veldkamp 1995 for more pellets collected, a cumulative be assigned a proportion of this. details). The length (CPL) and the total of the numbers of individual width (CPW) of the chewing pad fish of different species can 4.3.3 Recognisable should be measured. It should be be recorded. There are several possible to measure the maximum structures — key bones in ways of presenting the results width and the maximum length of of pellet analysis, each giving pellets the chewing pad, bearing in mind a slightly different picture of The pharyngeal bone is the ossified that this structure has specific Cormorant diet. At its simplest, it fifth gill arch of Cyprinids and shapes in different taxonomic is possible to record the presence/ loaches (Cobitidae), and its shape, groups (e.g. Chub). absence of different species size, and the position of teeth on in the pellets. These data can it are species-specific. There are Otoliths (or statoliths) are pairs be presented as frequency of several published keys to aid of structures in the inner ear of occurrence, either ‘percentage identification (e.g. Horoszewicz bony fish. The lapilli are found frequency’ (the proportion of 1960, Rutte 1962, März 1987, in the utricle, the sagittae in the pellets containing a particular Leopold et al. 2001). Several of suculus, and the asterisci in the species) or ‘relative frequency’ these keys also include advice on lagena within the skull. As fish (the number of occurrences of a how best to measure these bones are digested, the skull breaks particular species as a percentage and which equations to use to down and the otoliths become of all recorded prey items). Both convert these measurements to separated and are clearly visible frequency of occurrence methods estimates of original fish length. within pellets (as small white give a picture of Cormorant diet, Some examples are shown in ‘stones’), whilst most (if not all) either in terms of the proportion Figure 4.5. of the other bones of the skull are of pellets containing species X completely digested. The sagittae or the proportion of all identified The pharyngeal bones (PB) can be are generally the biggest of the fishes that comprised species Y. classified as left (PBl) and right three types of otolith in most fish For percentage frequency, the pharyngeal (PBr) and the lengths species, except for the Cyprinids total used is the number of pellets PBL1 and PBL2 can be measured where the asterisci are the largest and the frequencies will not sum for both. PBL1 is the length otoliths. The asterisci otoliths to 100% (as pellets can contain between the ventral and dorsal tips from Cyprinids (OLA) can also be remains of more than one species). of the bone and PBL2 is the length classified as either left or right and, For relative frequency, the total (width) between the dorsal tip with the help of measurements, can used is the number of recorded and the lateral process. With these be paired-up to represent single occurrences and the frequencies measurements it is possible to pair- fish. Otoliths must be examined will sum to 100% (see e.g. up right and left pharyngeal bones very carefully to exclude any that Dirksen et al. 1995; Keller, 1995; and thus to estimate the minimum show signs of erosion and those Veldkamp, 1995). number of fish represented by showing no signs of digestion can pharyngeal teeth in pellets. The be measured from the anterior to Researchers also often derive fish additional measurement PBL2 can posterior ends. The lapilus otoliths body length from the measurement also be used because pellets often of Cyprinids (OLL) can not be of key bones found in pellets, using contain pharyngeal bones that have categorised as being either left or either published equations or ones broken ventral tips, and so relying right as they are often too small. derived from local samples of only on the PBL1 measurement Nevertheless, they can be counted reference fish. They then convert can lead to an underestimate of the (and this number divided by two to fish size estimates to ones of fish original fish length. obtain the minimum fish number) weight (again using published and measured as described above. or reference equations). In this The chewing pad (CP) is an odd manner (see section 4.3.4), the structure found only in Cyprinids, For other taxonomic groups (e.g. total estimated biomass of fish being part of the pharyngeal Brown Trout Salmo trutta, Grayling

[40] the intercafe field manual

3.9, also e.g. Horoszewicz 1960, Rutte 1962, Wise 1980, März 1987, Libois & Hallert-Libois 1987, Libois et al. 1988. Petrova & Zivkov 1989, Mehner 1992, Leopold et al. 2001, Schulz-Mirbach & Reichenbacher 2006). There is also an AFORO online database for otolith identification — see www. cmima.csic.es/aforo/index.jsp

During pellet analysis and subsequent interpretation of data, it must be remembered (Carss et al. 1997: 199–200) that ‘pellet analysis is a useful method of obtaining a rough index of Cormorant diet in qualitative terms but there is serious doubt as to whether it can be used to derive quantitative information on, for example, the species composition or size-range of fishes taken. With reservation (in particular relating to the under-recovery of small fish), pellets may be used to investigate spatial or temporal variation in the relative frequencies of particular prey species of varying provenance. However, great care must be taken when interpreting the results of Figure 4.5 Measurements of biometric parameters of ‘key bones’ used in diet such studies as there are serious analysis. A–pharyngeal bone (PB) of a Nase Chondrostoma nasus, B–phary- potential biases.’ Nevertheless, ngeal bone of a Chub Leuciscus (Squalius) cephalus: PBL1 and PBL2–lengths of some studies (e.g. Dirksen et al. the PB; C–chewing pad (CP) of a Danube Roach Rutilus virgo, D–chewing pad 1995) have used the results of of a Chub; CPW–width of the CP, CPL–length of the CP; E–sagitta of a Grayling Thymallus thymallus, F–sagitta of a Brown Trout Salmo trutta: OSL–length of pellet analysis to estimate total fish the sagitta; G–lapilus of a Cyprinid: OLL–length of the lapilus; H–asteriscus consumption by Cormorants. of a Cyprinid (Chub): OLA–length of the asteriscus. Throughout, one unit of measurement corresponds to 1 mm. Photo courtesy of M Govedicˇ. 4.3.4 From otoliths and other bones to diet Thymallus thymallus, Pike Esox (Sander lucioperca) which can be composition lucius, Perch Perca fluviatilis) used for species identification, as the sagittae otoliths (OSL) can be can the opercular bones of Perch In order to reconstruct fish length classified as either left or right and and Pikeperch, and the vomers from measured otoliths or other the whole length of the measured as of Trout and Grayling. Specific bones it is necessary to have above. Other commonly recorded keys for the identification and regression formulae of bone length to key bones include the lower jaws measurement of these bones have fish length. Subsequently fish mass of Trout, Perch, Pike and Pikeperch been published (see Figures 3.8 and can be estimated from fish length

www.intercafeproject.net [41] the intercafe field manual

using other regression equations. ‘the following methods apply it contains. Sample size may be All species of fish have different to the stomach contents of dead further reduced if stomachs are regression formulae, often depending birds but also to regurgitations empty, but this can be reduced on season and local forms. If one and samples obtained by stomach by shooting birds later in the day is carrying out a major study on flushing. Daily food intake should after they have had a chance to diet, it is best is to base this analysis not be determined from stomach feed (although this adds a bias if on regression equations derived contents analyses because it is not some prey are digested quickly or directly from local fish. Short-term known whether a bird had stopped if there are diurnal variations in the or incidental studies tend to use feeding for the day (when it was prey selected.’ regression formulae taken from the shot). Maximum values are overly literature (e.g. Govedicˇ et al. 2002, influenced by subjective judgement 4.4.1 Collection and Przybylski 1996, Wise 1980). of what constitutes a full stomach, conservation and might overestimate daily intake The bony structures used for species by ignoring lower values.’ If researchers and hunters recognition are linearly correlated cooperate (e.g. within a to fish length and in a semi log ‘Stomachs (regurgitations and management program), and way to fish weight. The lengths of stomach flushings) often contain carcasses are delivered for analysis, fish should be measured up to the relatively fresh material but there it is advisable to prepare a standard nearest 1 mm and the body mass are also well-established methods identification label, which can to 0.1 g precision. Three lengths for dealing with well-digested prey. be filled out with the necessary should be measured: standard The differences between relatively information and fixed to the bird. length (SL), fork length (FL) and fresh fish and the undigested but This ensures that all the necessary total length (TL) (Ricker 1979), heavily eroded items at the bottom information is recorded as soon but only the total length (TL) is of the stomach are clear. Some of as the bird has been killed, and to be used to derive the regression the more serious errors associated that this information remains equation. To measure the total with such well-digested material with the bird from the time it was length of the caudal fin, it should be (e.g. as in pellet analysis above) shot. The more information that squeezed and the length at maximal can be avoided. Sometimes, can be recorded for each bird, the extension measured (Ricker 1979). stomach contents may be the only greater potential of interpreting way of assessing Cormorant diet the subsequent results of stomach As discussed in section 4.3.2, it is if pellets cannot be collected or contents analysis. Table 4.2 shows a good idea to explore the size- direct observations are difficult. the minimum information that distribution of fish eaten as well Stomach contents samples can should be recorded for each bird as to examine the diet based on be accompanied by age, sex, and by the hunter and the person numerical and mass proportions parasite infestation information for examining the carcass in the by species. This is especially each bird, and they are also site- laboratory. Needless to say, time important when trying to compare specific (i.e. foraging grounds are and place of shooting Cormorants the ‘importance’ of particular fish known) or can be implied from the will determine their diet and species — numerically, small 0+ location at which the bird was shot.’ this can also lead to bias. Dead (the young of the year) fish may Cormorants can also be obtained be the most frequently eaten prey ‘There are several disadvantages through cooperation with fishermen but, based on mass, a small number to stomach contents analysis, who use gill nets. Such nets can of larger (i.e. older) fish may most obviously the necessity of kill (drown) Cormorants as they are contribute most to overall diet. killing birds. Licences are required foraging, so the method is useful in to kill Cormorants in European some circumstances. 4.4 Stomach contents analysis countries and so in most cases the samples available for analysis Shot or other dead Cormorants Carss et al (1997: 204–205) are small. The sampling unit for should be examined as soon after discuss stomach contents analysis such analysis is an individual death as possible or stored frozen in considerable detail, stating that stomach, not the number of fish (-20˚C) as soon as possible.

[42] the intercafe field manual

Table 4.2 Minimum information to the body cavity with sharp scissors from the hindgut. The stomach be recorded by hunters and scientists making an incision in the flank and the foregut including the in relation to dead birds made available below the sternum which passes proventriculus should be opened for stomach contents analysis. through the sternum and clavicle. from the top to bottom (Figure 4.6). The skin of the throat is then By hunter By scientist opened by cutting across the gullet Some researchers then weigh the Date of shooting Sex of bird just below the head and down the stomach and its contents, and Time of shooting Age of bird length of the neck to join the top of then also weigh the contents on Location of Total weight of the initial body cavity incision. The removal. This latter measurement shooting bird body cavity incision is continued can be subtracted from the mass Type of water body down the remaining length of the of the intact bird to give a more body to the vent. The trachea and precise measurement (i.e. of the 4.4.2 Dissection various internal organs are then whole carcass minus the weight of separated from the stomach, which any meal in its stomach) (see Table After thawing and taking biometric can now be lifted out of the body 5.1 and description of ‘net weight’ measurements (see chapter 5) open cavity in its entirety and severed of body mass in section 5.4).

4.4.3 Analysis Any whole fish should be removed carefully, identified and measured (length, weight). Carss et al. (1997: 207–208) propose a standard method for processing stomach contents, an edited extract of which is quoted here. ‘All remaining partially-digested material should be flushed out, using a water bottle, into a storage container. A saturated solution of biological washing powder should then be added which completely covers the stomach contents. The washing powder digests all the remaining flesh from partially-digested fish after about 4–6 days. The process can be speeded up by placing samples in an oven at about 37°C and stirring occasionally. Once samples are ready they should be poured into a fine sieve and rinsed well with cold water to clean and dislodge stubborn pieces of flesh or cartilage, and to disarticulate skeletons or skulls.’

‘Once thoroughly rinsed, bones Figure 4.6 Two examples of opened Cormorant stomachs showing partially- should be transferred to filter paper digested food material. Photos courtesy of J Trauttmansdorff. for drying. Care should be taken to

www.intercafeproject.net [43] the intercafe field manual

ensure that all bones are collected, been shown for both Goosander Very often different species of as fine vertebral spines and very (M. merganser) and Cormorant parasites (mostly nematodes) can small pharyngeal teeth sometimes stomach contents data. In Scotland be found in the pellets and in the become lodged in fine-meshed (where the fish community Cormorant stomachs (see Figure sieves. Samples should be air dried comprises relatively few species) it 4.6). Take at least a sample of for 1–3 days before examination, has been concluded that ‘adequate’ them and if you are not able to alternatively they can be placed estimates of diet were possible identify them immediately, put in paper packets after drying from samples of 12–15 stomachs them in alcohol (70%) in a well- for storage and later analysis. containing food but more analysis sealed vial. Biological washing powder does is required from large samples not appear to damage bones and air of stomachs containing diverse Text Box 4.2 Parasites. drying does not lead to significant fishes.’ shrinkage of material over time stomach. Similarly, bone lengths periods of a few days-weeks.’ Care must be taken when can be converted to estimated interpreting the records of all fish lengths, and estimated fish ‘Analysis of the dried remains the (fresh and digested) material lengths to estimated fish weights involves placing them in a Petri recorded in stomach contents. As through a series of regression dish for examination under a low Carss et al. (1997) state ‘In theory, equations (for example, see power binocular microscope. the more digested the contents the Appendix 1 in Carss & Ekins Key bones (described in section greater the potential for bias, as 2002). Data are then usually 4.3.3) can then be extracted. As some items will be less resistant reported, on a species by species with remains from pellets, the to digestion than others and may basis, as percentages — both by appearance of these key bones can thus be digested relatively quickly. number and by mass. This dual be used for identification, and their Some workers have tried resolving approach to data presentation is size to estimate fish lengths and this problem by examining only important as often Cormorant diet hence fresh weights from a series intact material. However, it has comprises many small fish which of regression relationships. For been shown for Goosander stomach contribute a lot to numbers but the most accurate assessments of contents, that smaller fish were little to biomass. Conversely, most diet, key bones must be robust and under-represented by this method, of the biomass in the diet can be resistant to digestion, relatively probably because the digestion of the result of a small number of easy to identify, and diagnostic. fish is so rapid that small items large fish. Through the careful Thus it is recommended that care can disintegrate while larger ones measurement of key bones and is taken when selecting potential resist disintegration for longer. the accurate use of regression key bones to record. For example, Studies examining diet from intact equations, it is also possible to Feltham (1990) did not use otoliths items alone will underestimate the present dietary data in terms of from the stomach contents of proportion of small fish species the length-frequencies of different Red-breasted Mergansers (Mergus and overestimate the mean size prey species. Such information can serrator) because their presence for some larger ones.’ Thus data be enhanced further by reference did not correspond well with the from whole, intact, undigested to additional information from presence (or size) of other bones fish should be included with those the location where the birds were from the same fish species.’ determined from the examination shot (e.g. Russel et al. 2003, of key bones from partially- Trauttmansdorff 2003). ‘Finally, it should be remembered digested material. that the sampling unit for stomach contents analysis is not As with pellet analysis, key bones 4.5 A note on regurgitations the number of fish recorded in a can be identified to species level, stomach but the stomach itself. counted (as pairs if necessary) When Cormorants are disturbed Similarly, sample size is important and measured in order to estimate at a roost or in a breeding colony, as it may affect the accuracy the minimum number of fish they often vomit whole or partly of diet assessments — as has of different species within the digested fish which can be

[44] the intercafe field manual

collected from the ground below. Both whole and partially-digested fish can be treated in the same way as described for stomach contents analysis but there are potentially serious problems with interpretation. The samples collected from the ground may tend to be larger specimens (these are easier to find in dense undergrowth) and it is often not possible to determine which fishes were regurgitated by which bird (i.e. which fish correspond to the same ‘meal’) and this compromises the sample size. This is clearly not such a problem if regurgitations are collected at ground-nesting colonies or if they are collected from nestlings if trees are climbed to examine nest contents. With regurgitations produced by nestlings there is a further theoretical bias associated with the fact that adults might eat low quality food themselves and feed higher quality food (of different species and/or size) to their young. However data are currently unavailable to test this idea. Regurgitated fish can be used to calculate the regression lines needed to correlate otoliths and other key bones to total fish length (and mass, if specimens are completely intact).

4.6 Reference collection Before analysing the diet of Cormorants in any detail, it will almost certainly be necessary to establish a reference collection of the key bones, what will allow the Figure 4.7 Pictures of key bones identification of the fish remains Figure 4.8 Pictures of key bones from Chub Leuciscus Leuciscus (scales, found in pellets, stomachs, or from Perch Perca fluviatilis (scales, praeoperculum, operculum, clavicle, regurgitations. Not all keybones praeoperculum, operculum, lower pharyngeal bones). will be well-described in the jaw, clavicle). Photo courtesy of J Trauttmansdorff. literature and so making a reference Photo courtesy of J Trauttmansdorff.

www.intercafeproject.net [45] the intercafe field manual

0.5 mm intervals for small ones). how the act of feeding (as well Such material will be invaluable as the type and ‘quality’ of prey) when relating the length of bones contributes to both the more found in diet samples to the original general behaviour and ecology length of ingested fish (see earlier of the birds in the environment comments on regression equations). throughout the year, but also how The most commonly prepared much energy an individual bird bones and other structures in a spends on capturing a meal and reference collection are probably how much energy it obtains from scales, otoliths, praeoperculum, digesting and assimilating it. operculum, lower jaws, clavicles, This complex inter-relationship pharyngeal bones (Cyprinidae, between an individual’s Cobitidae), chewing pads physiology and its more general (Cyprinidae), vomers (Salmonidae, ecology and behaviour is Thymallidae) and vertebrae. As termed ‘bioenergetics’. This is examples, key bones from Chub, a complicated subject but one Perch and Brown Trout are shown that is important to understand in Figures 4.7–4.9. as we consider the ‘daily food intake’ (DFI) and ‘daily energy expenditure’ (DEE) of 4.7 Bioenergetics Cormorants.

How much fish does a Cormorant All living organisms must obtain eat each day? This is a vital food from the environment from question in relation to the which they derive chemical energy potential impacts of Cormorants to perform work, to maintain their at a fishery (see also chapter structural integrity, and, ultimately, 10). Here we must first have to reproduce. Hence, the quest rigorous estimates of both the for food is a fundamental driving number of birds feeding there force in determining (not necessarily the same as the behaviour. One highly relevant number counted there at any one aspect of behaviour in relation to time, see chapter 3, section 3.4) Cormorant-fishery interactions will and their diet there (in terms be foraging behaviour and foraging of biomass and the proportion site choice: the choice a bird makes that is of particular interest, see in relation to where to feed, the section 4.4.3). Crucially, we also length of time it will spend fishing Figure 4.9 Pictures of key bones need good quality data on the in one place, and decisions on from Brown Trout Salmo trutta species composition, availability (praeoperculum, operculum, lower how far it is ‘prepared’ to fly when or abundance of their prey in the jaw, clavicle). switching to a new foraging site. It waters where they forage (see Part Photo courtesy of J Trauttmansdorff. is therefore critical to understand Two of the Field Manual). both how much energy Cormorants require and how this energy is collection allows workers both to acquired. become familiar with bones and 4.7.1 Introduction other structures from particular Turning to how much fish is In the past, Cormorants were species, and also to prepare a consumed by Cormorants, this accused of devouring huge number of ‘type’ specimens of is a deceptively easy question to amounts of fish valuable to different lengths (perhaps at 1or ask but a much more complex one humans, and several early studies 2 cm intervals for large fishes and to answer. We need to consider estimated that they consumed more

[46] the intercafe field manual

Table 4.3 Estimated daily food intake (DFI, g/day) for both European races of Great and the estimated daily food intake Cormorant (different sexes, ages, and activities) and mean body mass values (and (differences of almost 400 g/day methods) from which these have been derived (from Carss et al. 1997). These figures for sinensis and 100 g/day for have then been used to determine the estimated DFI as a percentage of the mean carbo). Obviously a considerable body mass. amount of this difference in Race Age, (season Mean Estimated Estimated DFI Method estimated daily food intake will or activity) body daily food as percentage used be due to the different energy (and mass intake of mean body hence food) demands on the birds (g) g/day mass engaged in different activities. sinensis Adult 2,275 707 31% 31% Taking DFI as a percentage of (wintering) body mass, estimates vary but sinensis Immature 2,079 662 32% 32% are commonly around 25–30% (wintering) for sinensis and 30% for carbo. sinensis Adult 2,230 251 11% 11% Further estimates of daily energy (incubating) expenditure and daily food intake sinensis Adult (rearing 2,230 334 15% 15% are given in section 4.7.5 (Table small chicks) 4.5). sinensis Adult (rearing 2,230 621 28% 28% downy chicks) One possible way to investigate the sinensis Adult 1,915 502 26% 26% food consumption of animals in (unknown) the wild is to study their energetic sinensis Adult 2,122 522 25% 25% requirements. The energetic costs (wintering) for an animal to live throughout a carbo Adult 2,901 843 29% 29% day following its normal routine (wintering) are defined as its daily energy carbo Immature 2,675 790 30% 30% expenditure (DEE). Once an (wintering) estimate of DEE is established, this carbo Adult males 2,870 890 31% 31% can be converted to an estimate of (chick-rearing) daily food intake if information carbo Adult females 2,870 800 28% 28% on diet, its energy density, and (chick-rearing) the digestive efficiency of the consumer are available. Here we will briefly discuss what than the equivalent of their own (1997, see Table 4.3 above) determines Cormorant energy body mass in fish per day (e.g. tabulated current (at that time) requirements, the factors affecting 187% of body mass; Black 1964). estimates of daily food intake them, and the methods available to However, more recent studies for individuals of both carbo and estimate energy requirements and, have shown that a Cormorant’s sinensis races Great Cormorants, ultimately, food consumption. As daily food requirements, even based on predictive equations, Carss et al. (1997: 214) emphasise during energetically challenging time budgets (using a variety ‘pellets, the stomach contents periods (e.g. during chick rearing of techniques, assumptions and of shot birds, and regurgitations or during winter at low ambient estimates) and for a number of cannot be used to derive good temperatures), are considerably situations (times of year). estimates of DFI because of the lower than this. In fact, several associated biases in estimating studies (e.g. Grémillet et al. Overall (see Table 4.3), for both diet’ (discussed in section 4.3). 2003) reported that daily food these races of Great Cormorant Similarly, estimates of DFI derived requirements of Great Cormorants there is variation between the from feeding captive Cormorants are actually no different from studies in both the body mass or from food requirements of well- several species of of (over 300 g difference within grown chicks almost invariably similar body mass. Carss et al. sinensis, over 200 g within carbo) underestimate DFI as the flying

www.intercafeproject.net [47] the intercafe field manual

and/or swimming components usually make up a relatively small 10 m were on average 22% greater of the bird’s energy budget are portion of their daily time-activity than when birds performed shallow excluded in such calculations. budget (see Grémillet et al. 2003 horizontal dives (at depths of The solution to these problems and Grémillet et al. 2005b). around 1 m). is thus to base estimates of DFI on considerations of the energy As pursuit divers, Cormorants The daily energy expenditure of requirements of wild birds. submerge to depth, where they Cormorants might vary to some will search, pursue and ultimately degree from day to day, depending 4.7.2 Metabolic rate and its capture their prey. Water on a multitude of biotic and abiotic modifiers temperature will therefore be an factors that should be taken into important factor to consider when consideration. Most importantly, The energy contained in the food trying to estimate dive costs in however, it will vary during the that animals ingest becomes these birds (Enstipp et al. 2005, annual cycle of an animal. For available during digestion through Enstipp et al. 2006a). The depth example, energy expenditure the stepwise transformation of to which birds descend will also might be greatly increased during larger molecules (into ultimately modify dive costs. Buoyancy in the breeding season, especially

CO2 and water) in a process Cormorants is lower than that of during chick-rearing, when birds called oxidation. The sum of all other aquatic birds because of have to catch sufficient fish not chemical reactions that occur in their partially wettable plumage only for themselves but also for an organism over a specific period (Lovvorn & Jones 1991, Grémillet their chicks. Accordingly, birds is defined as its ‘metabolic rate’. et al. 2005a). This reduces the will increase their foraging effort Various components contribute amount of work that birds have (e.g. increase flight times, dive to the overall metabolic rate of to do in order to counteract bout duration and/or dive depth), an animal. The minimal rate their buoyancy during diving, which will lead to a greater of energy expenditure in an decreasing overall locomotor costs energetic demand and, therefore, endotherm (essentially, ‘warm- considerably. With increasing water food intake. Another potentially blooded’ birds and mammals) pressure, the birds’ plumage air challenging period for Cormorants under defined conditions is layer becomes compressed as it is the winter, where low air and called basal metabolic rate dives deeper, reducing buoyancy water temperatures might greatly (BMR) and only varies with work even further. However, birds increase thermoregulatory costs. time of day or season. Apart also rely on the air layer trapped Seasonal or long-term changes from these basal requirements, within their plumage for insulation, in prey abundance (perhaps adult birds also spend energy so that heat loss to the environment due to climatic fluctuations or for thermoregulation, digestion, at depth will be greatly increased. overfishing for example) might moult, and locomotion (e.g. Enstipp et al. (2006a) found that also force Cormorants to increase flying, diving). The daily energy dive costs in Double-crested their foraging effort. Indeed, requirement of an animal will Cormorants (Phalacrocorax Enstipp et al. (2007) showed that thus depend on which activities auritus) when diving vertically to foraging success of Double-crested it engages in and for how long. It furthermore depends on the energetic costs associated with Table 4.4 Activity specific energy expenditures (W kg-1) forPhalacrocorax carbo these activities and on a multitude carbo (mean body mass: 3.2 kg). of environmental factors that will NOTES: a modified from measured value reported by Storchet al. (1999), b estimated from measurements in modify these costs (e.g. climatic double-crested Cormorants (P. auritus) (Enstipp et al. 2006a), c modified from measured value reported by d conditions such as air and water Grémillet et al. (2003). from Grémillet et al. (2003), who used the aerodynamic model of Pennycuick (1989) to estimate flight costs. Water temperature during measurements in water was 13°C. Air temperatures during temperatures, solar radiation, and measurements were within the thermoneutral zone (TNZ) of Cormorants. wind exposure). For Cormorants, Rest (day)a Rest (night)a Rest (on Diving Diving (10m)b Flyingd flying and diving are the most water)b (1m)c costly activities and it is not surprising that these activities 4.4 3.7 11.7 22.8 27.1 78.8

[48] the intercafe field manual

Cormorants depends greatly on wing-spreading was estimated at sparing them from having to spend prey density. Hence, birds might 6.2 W kg-1. additional energy to generate heat try to buffer any reduction in prey (i.e. through shivering or activity). abundance by increasing their When ingesting cold fish, birds In Double-crested Cormorants there foraging effort in multiple ways, will have to expend energy to is evidence for such a mechanism leading to an increase in both their warm the meal (in addition to (Enstipp et al., 2008). DEE and DFI. the energy reported in the table for ‘resting’) so that further 4.7.4 Methods to estimate digestion can take place effectively 4.7.3 Activity-specific energy requirements (Grémillet & Schmid, 1993). energetic costs of Great During digestion, the bird’s Estimating the energy requirements Cormorants metabolic rate will be elevated of free-ranging animals is a due to the metabolic processes challenging task. Today, there The estimated energetic costs of digestion, assimilation, and are three different approaches to associated with specific activities nutrient interconversion. As these studying the field energetics of of Great Cormorants are listed processes are accompanied by the animals and each is discussed in in Table 4.4. All values were release of heat, this phenomenon detail below. measured via respirometry with has become known as the heat the exception of flight costs, increment of feeding (HIF). The (1) Establishing a time-energy which were estimated using the magnitude and duration of the budget (TEB) by combining a aerodynamic model of Pennycuick HIF depend on meal size and detailed time-activity-budget (TAB) (1989). Respirometry uses food type, a diet rich in protein with laboratory measurements of measurements of the animal’s and a large meal size having activity-specific metabolic rates, respiratory exchange (oxygen the greatest effect. For Double- while also taking into account a uptake, CO production) to 2 crested Cormorants, Enstipp et al. variety of biotic and abiotic factors calculate oxygen consumption (2008) found that the energetic (e.g. temperature) that influence rates. These can then be converted costs attributable to the digestion the energy requirements. From into metabolic rate when the of a single 100g Herring Clupea the knowledge of how long an substrate metabolised is know harengus were 51 kJ (this animal engages in each activity (lipid, protein, carbohydrate, includes the costs of heating the on a daily basis (and the energetic as indicated by the respiratory fish to body temperature), which costs associated with each activity), exchange ratio or RER), using the represents about 5.5% of the daily energy expenditure can be appropriate energetic equivalents. gross-energy content of the fish calculated (Goldstein, 1988). It is and which was associated with an obvious that the quality of a time- Cormorants also spend energy elevated metabolic rate for some energy budget can only be as good for activities not listed but their 5–6 hours after feeding. as the components from which metabolic rate during these it was derived. In this context it activities is usually just slightly While the energetic costs should be noted that the accuracy higher than during the resting associated with digestion should of the time-activity budget (TAB) is state and they typically account be incorporated into the daily crucial and will have the strongest for only a minor portion of their energy budget, it is not clear yet, impact on the overall estimate of daily energy requirements. The to what degree the heat released energy requirements. costs of wing-spreading, for during digestion might potentially example, were estimated by be used by birds to substitute for A variety of techniques can be Grémillet et al. (2000) for Great thermoregulatory costs. In other combined to establish a detailed Cormorants, based on respiratory words, under some circumstances TAB. In the past researchers measurements in Double-crested (i.e. at low ambient temperatures) often used a combination Cormorants (Hennemann, 1985). birds might be able to use this of direct observation and For a 3.2 kg Great Cormorant excess heat generated during bird instrumentation (e.g. energy consumption during digestion for thermoregulation, radiotelemetry) to distinguish

www.intercafeproject.net [49] the intercafe field manual

Table 4.5 Estimates of DEE and DFI in Cormorants. All DEE and DFI values during the breeding season do not take into account chick energy requirements.

Species Body Location Period Method DEE (kJ Conversion Fish energy DFI (g fish Reference mass day-1) efficiency density (kJ g-1 day-1) and as (kg) (%) wet mass) % of body mass P.c. sinensis 2.23 Germany Incubation TEB 760 77.0 4.0 247 (11%) Grémillet et al. (1995) P.c. sinensis 2.23 Germany Chick-rearing TEB 997 77.0 4.0 324 (14%) Grémillet et al. (1995) P.c. sinensis 2.12 Germany Winter DLW 2094 77.65 5.0 539 (25%) Keller & Visser (1999) P.c. carbo 3.2 France Incubation TEB 2131 77.0 4.0 692 (22%) Grémillet et al. (2000) P.c. carbo 2.3 France Incubation TEB 1532 77.0 4.0 497 (22%) Grémillet et al. (2000) P.c. carbo 3.2 France Chick-rearing TEB 2500 77.0 4.0 812 (25%) Grémillet et al. (2000) P.c. carbo 2.3 France Chick-rearing TEB 1797 77.0 4.0 583 (25%) Grémillet et al. (2000) P.c. carbo 3.2 Scotland Winter TEB 2779 77.6 5.33 672 (21%) Grémillet et al. (2003) P.c. carbo 3.5 Greenland Winter TEB 3632 77.6 4.0 1170 (33%) Grémillet et al. (2005b) Other Cormorant species P. aristotelis 1.78 Scotland Chick-rearing TEB 2249 81.0 5.4 514 (29%) Enstipp et al. (2006b) between time spent at the colony be a challenging problem and of Europe (Carss 2002, Carss & and that spent foraging. With explains the general bias in field Marzano 2003). recent technological developments studies towards the breeding a number of miniaturized season when it is generally easiest A further problem with time- electronic devices are now to catch and re-catch birds as they energy budget calculations is available that can be attached to are highly faithful to their nest that the separation of the various Cormorants and which enable us sites and young and unlikely to activities and their associated to accurately record the amount of abandon them. In this context it energetic costs overlooks the time that birds engage in a range is noteworthy that three studies fact that they are not necessarily of activities. For example, Daunt have actually estimated DEE independent from each other. and colleagues used a combination of Great Cormorants during the Many components of the TEB of compass loggers and time- winter period (Keller & Visser, might interact and so modify the depth-recorders in European 1999; Grémillet et al., 2003, resulting total energy expenditure Shags Phalacrocorax aristotelis and Grémillet et al., 2005b). of an animal. For example, heat to distinguish between phases of These studies are especially generated through physical or (i) rest on land or at sea from (ii) important as many of the physiological (e.g. digestive) flight and (iii) diving (Enstippet conflicts with Cormorants, and activity might be used for al., 2006b). One general problem hence considerable demand for thermoregulation (i.e. ‘keeping here - and one which concerns all rigorous estimates of daily energy warm’), sparing birds from having three methods - is the necessity expenditure and daily food intake, to spend additional energy to of capturing and recapturing occur during the winter (i.e. non- generate heat. This could be an individual birds. This can often breeding period) in many parts important factor during the winter,

[50] the intercafe field manual

when ambient temperatures in most commonly used today for implantation it has great potential most cases will be below the lower measuring daily energy expenditure for the future, especially since critical temperature of Cormorants. in seabirds and has also been used further miniaturization of memory At these times birds might have for Great Cormorants (Keller & chips will allow data collection to spend significant additional Visser 1999). However, there are a over extended periods. This is energy for thermoregulation but number of important assumptions particularly important in the context this is usually not accounted for in and requirements for this method of DEE variations throughout time-energy budget calculations. to produce a reliable estimate the annual cycle of Cormorants. Hence, energetic measurements of daily energy expenditure (see Given the logistic constraints, most that investigate the interaction Speakman 1997). There are field studies on energetics of the various biotic and abiotic also some drawbacks. A major have been carried out during the factors during certain activities disadvantage is that accuracy of breeding season, an energetically are of great importance and some individual measurements is so low challenging time. Consequently, investigations have been conducted that it can not be reliably used to DEE estimates are biased and in Cormorants. For example, estimate the energy demands of might (or might not) be lower Enstipp et al. (2005) and Enstipp an individual bird. The reasons outside the breeding season. et al. (2006a) investigated the behind this remain, as Butler et effects of a number of factors (air al. (2004) put it, ‘obscure’ and With the development of and water temperature, dive depth, DLW studies therefore require miniaturized, implantable nutritional state) on the diving relatively large sample sizes (e.g. data-loggers, the recording of energetics of European Shags and 9–10 individuals). Another big physiologically relevant variables Double-crested Cormorants. A drawback with this method is that (e.g. heart rate) for extended further area of great importance is it only provides a mean value of periods (currently up to one year, the energetic costs associated with energy expenditure over the entire see Grémillet et al., 2005b) has flight. Estimates in time-energy study period. In other words, become possible. Heart rate, budget calculations are usually it is not possible to assess the when properly calibrated against based on the aerodynamic model costs of specific activities. Some oxygen consumption (see Butler of Pennycuick (1989). However, studies have tried to overcome this 1993 and Butler et al., 2004 for this model produces extremely last problem by simultaneously details), might serve as a proxy for high flight costs for Cormorants deploying activity recorders to energy expenditure. Hence, with and there is a clear need for gather time-activity data, which the deployment of this technique in further research, perhaps using then enable back-calculation of the future, might allow researchers doubly-labelled water (see below), specific activity costs (e.g. flight). to gain a better understanding of to validate these predictions or However, the latter approach has how energy expenditure varies produce revised estimates for the met with criticism (see Wilson & throughout the annual cycle of energetic costs of flight. Culik 1993). Cormorants.

(2) The doubly-labelled water (3) The heart rate method (Butler 4.7.5 Daily energy (DLW) method (Lifson et al. 1993) exploits the physiological expenditure (DEE) in Great 1955) estimates the rate of CO relationship between heart rate (ƒ ) 2 H Cormorants O production from the difference and oxygen consumption rate (V 2) in the rate of loss of so-called and requires calibration of both For Great Cormorants, only the ‘labelled hydrogen’ (2H or 3H) and variables against each other under first two methods described above oxygen (18O) from the animal’s controlled conditions. Oxygen have been used to estimate daily body. CO2 production can then be consumption rate can then be energy expenditure. Results from converted to a measure of daily estimated from heart rate recorded these studies are compiled (Table energy expenditure through a in free ranging animals (for a 4.5) and it is clear that estimates well-known ‘respiratory quotient’ comparison of DLW and heart rate of DEE vary considerably with (see Carss et al., 1997 for methods see Butler et al. 2004). location and period of the study. further details). It is the method While this method requires surgical The high energetic expenditure

www.intercafeproject.net [51] the intercafe field manual

during chick-rearing most likely Diet information can be obtained for Channel Catfish(Ictalurus reflects increased activity as a in numerous ways, through direct punctatus). In other words, the consequence of having to deliver observation (section 4.2), pellet birds only ‘absorbed’ about 75– sufficient food for the chicks. High analysis (section 4.3) or from 80% of the actual energy content energy expenditures during the analysis of stomach contents or of the fish that they ate, and this winter are probably a consequence regurgitations (sections 4.4 and ‘absorption rate’ clearly differed for of low ambient temperatures and 4.5). The energetic density of fish different fish species. the associated increased energy taken should ideally be measured required to ‘keep warm’. The directly via bomb calorimetry. 4.7.7 Models to estimate study by Grémillet et al. (2005b) If this is difficult, values might represents an extreme case, alternatively be found in the daily energy expenditure where birds winter at the limit of literature. However, one should (DEE) and daily food intake their northern distribution range, be aware that energy density (DFI) in Cormorants near the Arctic Circle. Here, varies considerably throughout Great Cormorants encounter the annual cycle of fish and might Grémillet et al. (2003) established air temperatures well below also differ between geographic a computer model (Excel zero (down to -30°C) and water regions (see for example Hislop et spreadsheet) that can be used to temperatures as low as -1°C, while al. 1991) and this can potentially calculate the DEE and DFI of also increasing their dive depth introduce a big error into the DFI Great Cormorants from detailed throughout the winter (mean dive estimate. time-activity data. It combines depth = 18 m, with some dives time-activity data (which can be exceeding 40 m). From the knowledge of diet changed according to situation) composition and its energetic with activity-specific metabolic density, a ‘composite energy rates (representing the state of 4.7.6 Converting daily density’ value can be calculated knowledge in 2003). It is based on energy expenditure (DEE) to and DEE can be converted to observations of Great Cormorants daily food intake (DFI) DFI. However, not all the energy wintering in Scotland and contained in a fish can be converted incorporates the energetic costs If we want to convert the estimate and retained by the bird swallowing associated with resting, flight, of energy expenditure into fish and digesting the fish. Hence, we diving, and wing-spreading. It also mass required to satisfy these also need to know what proportion takes into account a number of energetic needs, we need detailed of the total energy held within a fish factors influencing dive costs, such information on three things. can be retained by a Cormorant. as water temperature, dive depth, This has to be accounted for in the and dive-pause ratio. 1. The diet composition, that is DFI estimate and will consequently which fish species are taken increase our original estimate. For Enstipp et al. (2006b) used a and in what proportions? (see Great Cormorants no estimates similar model to establish a time- sections 4.3.2 and 4.4.3); of energy conversion efficiency energy budget and to estimate daily 2. the ‘energetic density’ of the exist. However, Brugger (1993) energy expenditure and daily food fish species consumed. In other investigated the ‘digestibility’ intake for European Shags during words, the calorific content of three fish species by Double- chick-rearing in Scotland. While of the prey, ‘oily’ fish such crested Cormorants. The these models are quite useful, as Herring and Eel Anguilla proportions of total energy within they can only be as good as their anguilla being more energy- fish that were actually obtained by input data and even then important rich than Cod Gadus morhua or the birds — the so-called ‘nitrogen assumptions and restrictions should flatfishes (e.g. Pleuronectidae) corrected metabolizable energy not be overlooked. One should not for instance, and; coefficients’ were about 75% for be overwhelmed by the apparent 3. the efficiency with which Bluegill (Lepomis macrochirus), ease with which parameters can Cormorants convert the ingested about 78% for Gizzard Shad be calculated with these models energy contained in the fish. (Dorosoma cepedianum), and 79% and they should also be improved

[52] the intercafe field manual

as further knowledge becomes Double-crested Cormorant dive More recently, Ridgway (2010) available. For example, while the costs via respirometry and found builds on many of the issues and effect of depth on the energetic that the physical model used by studies described above and reviews costs of diving is incorporated Grémillet and colleagues greatly estimates of daily energy expenditure in the model by Grémillet et al. overestimates this effect. Hence, and food intake in cormorants (2003), it is merely based on new measurements should be (Phalacrocorax spp.) and makes a physical model of heat loss. incorporated into Grémillet et al’s a number of recommendations for Enstipp et al. (2006a) measured overall model, but this remains to estimating DEE and DFI at the the effect of dive depth on be done. population level.

www.intercafeproject.net [53] the intercafe field manual

5 SEXING, AGEING AND MEASURING CORMORANTS T Bregnballe, D N Carss, M van Eerden, S Newson, R Parz-Gollner and S van Rijn

5.1 Introduction Sexing and ageing Cormorants can be performed on live birds, Cormorants are often casually based on characteristics relating referred to as ‘big black birds’. to the appearance of the bird However, within well-defined (Millington 2005, Newson et al. ranges of course, they come in 2004). Breeding birds can also often a variety of structural sizes and be sexed based on observations plumage colourations (see Figure of behaviour (displaying and 5.1 and others in this chapter). copulations). With some experience These differences in plumage the sex of a bird can be estimated colouration and structural size, based on its general structure, for if carefully and systematically example the size and appearance of recorded, can give valuable insight the bill and head. Measuring birds is into several important aspects of most often carried out when ringing the ecology of the species, for nestlings but also when birds are example in relation to timing and caught or shot under licence. strategy of migration pattern and Figure 5.1 Two ring recoveries breeding. Plumage colouration Observations of ringed birds (of showing differences in body size and can be used to determine the age known age) give the opportunity plumage characteristics (birds shot in of birds — at least categorising to check if a field estimate of age Austria January 2006). Left: (No. 382) them as ‘juvenile’, ‘immature’ based upon plumage characteristics ringed in Denmark, adult, breeding or ‘adult’ (i.e. sexually mature). is correct. Colour-ringed birds of plumage, male, age 3 years and 8 This can be used (in conjunction known sex (based on behaviour months. Right: (No. 379) ringed in with roost counts, see section 3.3) in the breeding season) make it Finland, male, immature, age 1 year to record the proportional (i.e. possible to recognise sexes based and 8 months. adult:immature) age-composition on a comparison of structural size Photo courtesy of R Parz Gollner. of birds at roosts for instance or with other birds. the use of particular roosting or foraging sites by different age- Carcasses can be studied by known age) give the opportunity to classes of birds. Structural size of describing the plumage and by check if an informed guess of age- birds can help to determine the an examination of the bird’s related to plumage is correct. sexes of the birds in the field. This reproductive organs through knowledge, also from dissected dissection. Analysing carcasses gives When analysing carcasses, sex individuals also contributes to a the opportunity to relate structural and age-related differences in better understanding of timing of size to the sex of individuals. plumage colouration of birds also migration and strategies of birds. Observations of ringed birds (of allow researchers the opportunity

[54] the intercafe field manual

to examine Cormorant diet (see of individual birds. This ecological 4.5). Whilst there is still much to chapter 3) in relation to different segregation might also affect the learn about the genetics of Great sex and age classes. Plumage behaviour of males and females Cormorants across the geographic colouration, particularly the within flocks of social fishing range of the species, biometric development of breeding plumage, individuals as well as the distribution measurements allow the majority can also be used to compare the diet of both sexes in wintering areas (van of P. c. carbo and P. c. sinensis of breeding birds with those of non- Eerden & Munsterman 1986). individuals to be correctly assigned breeders. Almost exclusively, these to the level of subspecies. The measurements and observations are By observation from a combination angle of the gular pouch is a good restricted to dead birds but they can of body size, bill shape and measure to separate the subspecies be used to help to interpret dietary structure as well as the shape of (Newson et al. 2004). information obtained in the field. the forehead, it may be possible to estimate the sex of individual Measurements from Cormorant In terms of ‘size’, the careful birds in the field. Experienced and chicks also provide useful standardised measurement of various trained observers may thus estimate information (see 5.6 and Figure features of a Cormorant’s body such sex ratio in various flocks. Any 3.12). Thus, biometrics of nestlings as weight (‘mass’), wing-length, differences in the use of habitat can be used to estimate the age (in and bill shape and size can help to between males and females and days) by taking the wing length determine sex of birds. Females between adults and immature birds of chicks to reconstruct hatch are generally smaller than males is likely to have an ecological and egg laying dates based on as shown by discriminant analyses explanation, which will contribute the growth curve of the wings. discussed later on in this chapter. to our understanding of the Such information is important temporal and spatial use of different to understand the relationship Differences in structural size habitats across the species’ range between the timing of breeding and between males and females can for (van Eerden & Munsterman 1995). environmental conditions in the example affect individual hunting field, such as food abundance and performance (see Koffijberg & van Measurements of various features weather conditions for example. Eerden 1995). For example, a sex- of the Cormorant’s head can Additionally these measurements specific foraging skill could affect provide useful information on the can be used to determine the body foraging success and/or prey choice racial identity of the individual (see condition of chicks by calculating

Figure 5.2 Adult Cormorant (No.432), female, ventral side showing complete Figure 5.3 Adult Cormorant, dorsal black breast and belly, no breeding plumage; net weight 1,700g; ringed view, note the bronzy scapular 7.6.2002 in Sweden, shot 10.1.2006 in Austria (3 years, 7 months). feathers with black margins. Photo courtesy of R Parz Gollner. Photo courtesy of R Parz Gollner.

www.intercafeproject.net [55] the intercafe field manual

its relative mass (referring to age), which offers the opportunity of relating this to environmental conditions and survival.

5.2 Plumage colouration — age determination Only full-grown Cormorants appear to be completely ‘black’ but see Figures 5.2 and 5.3. Adult (pre-) breeding birds have an overall ‘black’ plumage, but this is glossed purplish or iridescent green with bronzy sheen on scapulars, tertials and upperwing coverts.

Further characteristics of the adult breeding plumage (in both sexes) are a ‘crest’ of long, white filoplumes extending down the back of the neck giving a variably ‘hoary’ appearance, and a large white patch of filamentous feathers on each flank. Birds have conspicuous yellow gular skin with a white border, a small orange patch below an emerald green eye with an Figure 5.4 (top) Immature Cormorant (No. 399), female, 1,680g, (shot in orbital ring of grey, yellow or black. Austria, winter 05/06). Photo courtesy of R Parz Gollner. The lores (the space between the eye and the bill in birds) is yellow Figure 5.5 (above) Immature Cormorant (No. 447), female, 2,034g, (shot in to orange, the bill is dark, becoming Austria, winter 05/06). Photo courtesy of R Parz Gollner. yellower at base of mandibles, the legs and feet are black. Mature plumage is acquired by year 3 or 4 and third-winter birds resemble post-breeding adults but still have a mottled belly (see Nelson 2005).

Unlike adult birds, immature individuals have highly variable breast plumage. This ranges from an almost completely white breast, through a white breast speckled with ‘black’, to an almost black breast speckled with white. There appears to be little, or no, relationship between the degree of ‘black’ on the breast and the Figure 5.6a–d Immature Cormorants, various extent and pattern of white age of these immature birds. The and dark on underparts. Photos courtesy of R Parz Gollner.

[56] the intercafe field manual

Figure 5.7 Wing of first-year Cormorant (top) and adult bird (above). Photo courtesy of T Bregnballe. variation of types within a number of immature individuals is shown in Figures 5.4–5.6.

Other than the breast, the remainder of the feathers of immature birds are brownish rather than the dark black of adults. An example of this difference in the wing feathers is shown in Figure 5.7.

During the breeding season (and well before when still present in Figure 5.8 Index of cover of filoplumes of head (0 = no plumes, 4 = fully covered) the wintering area) fine white head and of the thigh patch (again scored 0–4). Drawing courtesy of M R van Eerden. feathers (‘filoplumes’) appear on the adults and (later) on immature Cormorants. These become more in display behaviour, to attract and speckled, a small red dot prominent as the bird gets older. partners or impress neighbours. appears at the lateral base of the bill These diagnostic characteristics Observations of this index can help and the bare parts around the eye can help determine the proportion to describe breeding phenology get more colourful at the time of of sexually mature birds recorded and possible changes therein egg laying. All these characteristics in the field and at roosts. The (geographical range, seasonal and fade away one after the other, development of it is related to the annual variation). Related to the starting during the first half of the timing of breeding and can easily changing plumage, changes also incubation period. Of these the be scored in the field (see Figure occur in colour of the bare parts. thigh patch remains visible the 5.8). The breeding plumage is used The gular pouch becomes darker longest.

www.intercafeproject.net [57] the intercafe field manual

5.3 Body measurements — sex determination Body length Perhaps the most obvious measure- ment to take from a dead Cormorant is its length, total length being measured on the dorsal side from the tip of the bill to the end of the central tail feathers. This measurement is somewhat limited in accuracy, depending on how the dead bird was handled before being measured. If the bird was left hanging (see Figure 5.9) it will probably have been stretched. Care should be taken to avoid hanging or stretching the carcass.

In order to make measurements of Figure 5.10a Yellow lines indicating the starting position on the upper the same dimensions comparable, mandible for measuring bill length researchers have standardised the (L, right bar) and bill depth (D, left exact positions from and to which bar). For measuring nestlings with they take specific measurements. un-feathered forehead see also Provided that the same measuring Figure 5.17. points are always used, the most useful and accurate measurements can be taken from head, wing, tail, leg first, thebill (or ‘culmen’) length (‘tarsus’) or breast-bone (‘sternum’). is taken from the feathers in the centre of the forehead at the base of A good set of scales or an the bill, to its tip (see Figure 5.10a, automatic balance is needed to where the yellow line ‘L’ indicates weigh birds (nearest gram) and the position of the starting point on a pair of Vernier callipers and a Figure 5.9 Many hunters store dead the upper mandible). stopped rule to measure heads (legs Cormorants by hanging them from the neck: this stretches body length or the sternum) and wing (tail), The second is the bill depth (or and should be avoided. respectively. Measurements should height), a measurement taken Photo courtesy of T Bregnballe. be to the nearest 1 mm (for the vertically down on the base of head, tarsus, sternum, wing or tail). the bill from the upper to the oesophagus/stomach contents can lower mandible. Note that there Body Weight (‘Mass’) be more than the average difference is a difference in taking bill depth measurements in mass between males and females measurements when handling birds It is important to indicate whether (see section 4.4). It should also be of different age classes. Dealing the gross or the net body mass has noted whether the bird was wet (i.e. with nestlings, this measure is taken been taken. Gross weight includes taken from the water) or dry, when at the base of the nostrils (see Figure the oesophagus/stomach contents it was weighed. 5.17). In the case of fully-grown (food), whilst net weight can only birds, where no nostrils are visible be obtained by dissecting the bird. Bill measurements on the culmen, the corresponding This distinction should always There are two standard bill position to measure bill depth is be noted, because the body mass measurements in use, both of taken vertically down along the line recorded either with or without the which are taken with callipers. The of the edge of the front feathers on

[58] the intercafe field manual

Figure 5.10b Lateral and dorsal views of measuring bill length in adult Cormorants. Photo courtesy of S van Rijn.

5.4 Large Samples — sex and age discrimination Figure 5.10c Lateral view of an adult Here we discuss two studies that have Cormorant with the position for measuring bill depth indicated with red had access to relatively large numbers arrows. Photo courtesy of R Parz-Gollner. of dead Cormorants. Both studies explore differences between both adult and immature birds of each sex. the forehead as indicated with the They show the value of such samples line ‘D’ in Figure 5.10a. and the use to which the biometric measurements can be put. Wing measurements Standard wing length is measured Table 5.1 summarises four body with a stopped rule. The fold of the measurements taken from a large wing is placed against the stop and sample of ‘full grown’ Cormorants the primary feathers are straightened shot during control actions in Austria along the length of the rule. The (data collected between 1995/96– wing length is measured to the 2006 courtesy of Rosemarie Parz- extreme tip of the longest primary Gollner). Data from measurements feather. This measurement is shown of body mass (dry, net weight), bill in Figure 5.11 and is also referred to Figure 5.11 Measuring wing length. (culmen) length, bill depth, and wing as the ‘maximum wing chord’. Photo courtesy of T Bregnballe. length are tabulated separately for

www.intercafeproject.net [59] the intercafe field manual

male and female birds categorised as Table 5.1 Body measurements of fully-grown Cormorants shot during control either adult (AD) or immature (first actions in Austria. winter) (IMM). Mean (‘average’) Males Females measurements are given as well as the minimum and maximum AD IMM AD IMM measurements recorded, and the total culmen (bill) length number (N) of birds measured in the Mean 68.45 68.73 61.43 62.03 sample. Measures for bill and wing Range 62.8–77.8 60.8–77.3 55.1 –68.9 55.8–68.3 lengths were taken as indicated in N N = 96 N = 142 N = 57 N = 59 Figures 5.10a–c and 5.11.

bill depth Here, ‘net weight’ is obtained when the weight of food found in the Mean 21.77 21.74 19.36 19.33 stomach and oesophagus has been Range 19.7–24.7 17.3–26.6 17.4–21.2 17.6–22.3 substracted from the total (gross) N N = 95 N = 141 N = 57 N = 59 weight of dead birds and total weight was taken before dissecting wing length the carcasses. All birds in this Mean 359.12 359.59 338.23 338.36 sample have been dissected. Range 340–380 340–380 322–354 32–355 This comprehensive dataset shows N N = 95 N = 133 N = 52 N = 57 clearly that female Cormorants are considerably smaller (by each body mass (dry plumage, net-weight) of the four measurements taken Mean 2440,53 2378.77 1983.74 1991.36 here) than male birds. Furthermore, Range 1540–3440 1737–3230 1638–2580 1604–2623 although this difference in size is N N = 124 N = 175 N = 69 N = 75 also apparent for both adult and immature birds of each sex, there is very little size difference for either male or female birds between these Cormorants accidentally drowned Discriminant analysis was used to two age-classes (Figure 5.12). in fishing gear. A series of examine which combinations of measurements were taken from these measurements best predicted In these Figures (5.12a–c), the Box these birds which could be sexed on the sex of individual birds. The Plot graphics show lines at the top, the basis of external examination single measurement that best bottom, and through the middle and aged on the basis of their separated males and females of the boxes which correspond to plumage characteristics (see section was bill depth, according to the the 75th percentile (top quartile), 5.2). Measurements taken were as following equation: 25th percentile (bottom quartile), follows: D = 0.71 x Bill Depth — 14.35 and 50th percentile (the median). The horizontal ‘whiskers‘ on (i) Fresh body mass (dry where a resulting D value greater the very bottom extend from the plumage, nearest 1 g) than zero indicates a male bird and 10th percentile (bottom decile) (ii) Mass of fresh fish in the a D value less than zero indicates a and on the very top from the 90th oesophagus (nearest 1 g) female. percentile (top decile). A square (iii) Body length (length of symbol indicates the arithmetic stretched body from tip of bill When tested, this measurement mean for each sample. to tip of tail, nearest 1 cm) of bill depth correctly identified (iv) Wing length (nearest 1 mm) male and female birds in 85% of Similar information was collected (v) Sternum length (nearest 0.1 mm) cases. Even higher proportions by Koffijberg & van Eerden (vi) Bill length (nearest 0.1 mm) of individuals could be assigned (1995) from a sample of sinensis (vii) Bill depth (nearest 0.1 mm) correct sex when combinations of

[60] the intercafe field manual

    

        

         

                  

Figure 5.12a Sample size/box plot bill (culmen) length (left), bill depth (right); adult male — immature male, adult female — immature female. Measurements taken as indicated in Figure 5.10b and c.

  

   

   

         

                          

Figure 5.12b Sample size/box plot wing length; adult Figure 5.12c Sample size/box plot body mass (dry male — immature male, adult female — immature female. plumage, net weight); adult male — immature male, adult female — immature female.

www.intercafeproject.net [61] the intercafe field manual

Figure 5.13 Immature female (left) and immature male (right) in the 1st winter — note the difference in bill shape between the sexes (see text for details). Photos courtesy of R Parz-Gollner. body measurements were taken Cormorants did take significantly body size, bill shape and forehead together. Furthermore, overall larger individual fish (of some shape in the field with wild birds body measurements were better species) than females did. to determine different use of roosts for assigning individuals as male and sex ration in several parts of birds — which are bigger than Finally, these authors also the wintering range of the sinensis females and tend to vary in size less examined the general head/ race. How these differences would than the smaller females do (see bill shape of male and female apply to carbo birds is not known. Koffijberg & van Eerden 1995). As Cormorants (see their Figure 5) and is usually the case, determining the showed that females had a ‘saddle- sex of individuals is most difficult shaped and more slender bill’ than 5.5 Racial identity (carbo, in young, immature birds where males, also that males had ‘a more sinensis) size is more variable than it is in massive, rather straight-edged bill’. fully-grown birds of either sex. Similarly, the ‘forehead of females The use of biometric measurements often is more rounded compared to for sub-specific identification in These authors took their analysis males’. A series of photographs of Great Cormorants was originally further by exploring whether adult and immature Cormorants of investigated using skins from birds there were any differences in the both sexes is shown in Figures 5.13 of known sex and sub-species diet (using stomach contents, and 5.14. (Newson et al. 2004). This showed see section 4.4) of male or that the gular pouch (‘throat’) female birds. Although the Other authors (e.g. van Eerden angle was a useful character for species composition tended to & Munsterman 1995) have gone assigning individuals to subspecies. be similar for both sexes, male on to use the combination of Where further measurements were

Figure 5.14 Left - immature male, 1st winter (ring recovery from Finland, 8 months old); Right — immature male, 2nd winter (ring recovery from Finland, 1 year and 8 months both birds shot in Austria). Note the massive, straight-edged or ‘conical’ shape of the bill and difference in plumage coloration of immature birds of known age. Photos courtesy of R Parz-Gollner.

[62] the intercafe field manual

taken of bill depth and bill length, P. c. carbo sex-specific formulae allowed A) the majority of individuals to be correctly assigned to either carbo or sinensis races. This character needs careful examination when applied in the field; if viewing angle (horizontal) and position of head (lateral side fully exposed Gular pouch angle = 58º to observer) are favourable, this character can successfully being applied in the field.

Thus it is possible to identify carbo and sinensis individuals of known P. c. sinensis sex, as follows. Bill length (BL) is measured from the tip of the bill to the feathers at the centre of the bill, B) bill depth (BD) at the narrowest point in the middle of the bill (see section 5.3). Gular pouch angle (GPA) is measured to the nearest 1° using a rotatable protractor (see Figure 5.15). It is important to Gular pouch angle = 85º note that the baseline from which the angle is measured is not the bill line, but the line from the gape outwards, chosen because it gives greatest reproducibility between measurements. On live or freshly Figure 5.15 Gular pouch measurements for carbo (A) and sinensis (B) dead specimens, it is important Cormorants — After Newson et al. 2004. to note that the gular pouch can

Figure 5.16 Two individuals of nominate race P.c.carbo, shot winter 2001/02 in Austria. No. 219 (left) immature male, 2,921 g net weight, No. 226 (right) immature male, 3,140 g net weight. Photos courtesy of R Parz-Gollner.

www.intercafeproject.net [63] the intercafe field manual

be distorted to change its angle. With all measurements care should be taken to ensure that the pouch assumes a normal attitude by gently stroking it downwards before measurements are taken.

Males can be classified ascarbo if the result of the following calculation is greater than 4.66583.

(0.92133 x BD in mm) + (0.36504 x BL in mm) — (0.50198 x GPA in degrees) Figure 5.17 Nestling Cormorant, showing measurement of bill depth at base For example, a male Cormorant of nostrils (red line). This bird is likely to be a male, according to the un- with measurements (BD 12.2 mm, feathered forehead at an age of about 20 days (Estonian colony, June 2007). BL 68.4 mm and GPA 82 degrees), Photo courtesy of S van Rijn. gives a value of -4.9534, which is not greater than 4.66583 and so the individual is therefore categorised carcasses). The fold of the wing measure bill depth on fully-grown as belonging to the sinensis is placed against the stop and the birds the edge of the front feathers race. Similarly, females can be primary feathers are straightened is taken as a corresponding position classified ascarbo if the result of along the length of the rule. The (see Figure 5.10a). the following calculation is greater wing length is measured to the than 4.87236. extreme tip of the longest primary Measurements can be taken from feather. This measurement is shown hatching until fledging although (0.87159 x BD) + (0.56828 x BL) in Figure 5.11 and is also referred young birds tend to jump out of + (-0.61081 x GPA) to as the ‘maximum wing chord’. the nest between 23 and 41 days old and should be approached 5.6 Measurements from There are two standard bill with care, especially in ground- nestlings measurements used by researchers breeding colonies. Observers (see section 5.3 and Table 5.1 should not handle or stress the Similar body measurements to for carcasses), both of which are smallest (youngest) nestlings. those described earlier for adult taken with callipers. The first, the Time in colonies should be limited and immature Cormorants can bill (or ‘culmen’) length is taken as prolonged exposure to direct also be taken from younger birds from the feathers in the centre of sunlight can cause mortality of in the nest (see Figures 3.12). the forehead at the base of the bill, hatchlings. Nestling biometrics can be used to its tip (see also Figures 5.10a to determine age, growth rate, sex and 5.10b). The second is the bill Biometric data can be used to and condition of individual birds. depth (or height), a measurement determine the relative condition On each nest visit, body mass taken vertically down from the of individual nestlings, and these and other parameters — wing top edge of the upper mandible can be examined in relation to a length, bill length and bill depth to the bottom edge of the lower number of factors including brood of each nestling in a brood can mandible (Figures 5.10a and size and the age-rank of chicks (see be measured. Body mass can be 5.10c). Measuring nestlings with an Platteeuw et al. 1995). Wing-length determined to the nearest 5 g, unfeathered forehead, the bill depth is a strong predictor of nestling age using a Pesola spring balance. is taken at the base of the nostrils. (in days) when accurate growth data Wing length is measured with a Fully-grown birds do not show are available. With this the timing of stopped rule (see 5.3 the same for nostrils on the culmen therefore to egg laying can be back-calculated

[64] the intercafe field manual

and related to condition relative having whitish fronts, becoming Depending on their condition, to age and subsequent survival (in fully covered with down from day chicks can best be ringed between combination with colour-ringing). 26 on. This feature is a reliable 25–35 days old. Younger chicks can The bill shape (length related predictor to choose chicks which have larger tarsi when condition to depth) is a gross indicator of can be ringed, especially useful in is good and can bear 18 mm rings sex, which can be confirmed for tree nests. from 18–23 days on. In ground individually marked colour-ringed colonies an earlier ringing date birds later on, through observation Wing and tail feathers develop gives less disturbance and allows of sex-based behaviour. strongly after 28 days being the chicks to stay in the nest. From followed by scapulars and upper day 25 on chicks tend to walk away Age of young can be determined back from day 30–33. Fully more often and after day 35 young by stage of plumage development. developed wings appear normally tend to jump from lower bush nests Downy plumage appears after one not before day 40. At that time if approached. week, this is first blackish brown also belly, head and rump have no (10–20 days), then gradually more downy feathers. Wings tend becoming greyer. The feathering to increase in length after fledging of the forehead is a good indicator between 50–55 days. of age, chicks less than 20 days

www.intercafeproject.net [65] the intercafe field manual

6 MOVEMENTS T Bregnballe, J Y Paquet, S van Rijn and S Volponi

6.1 Introduction a few stopovers. Some Cormorants Ringing programs are extremely wintering in the southern and valuable tools for increasing our Cormorants move extensively central part of the Mediterranean knowledge about a number of around continental Europe and are known to have migrated up to topics relevant for understanding beyond (Figure 6.1). After the 1,100 km north in spring within 3 the ecology of Cormorants during breeding season young Cormorants days. and outside the breeding season first start to explore the environment and for developing knowledge- around the nesting site thereafter Cormorants from northern breeding based management strategies to they, as well as most adults, disperse areas may benefit from migrating resolve conflicts. For example, the in different directions, before south in a number of ways. For recoveries of ringed Cormorants initiating the real autumn migration. example, most fish move into deeper found dead and the readings of Not all Cormorants migrate south, water and become less available to codes on colour-rings attached indeed some remain within a Cormorants with declining water to live birds gives researchers few hundred kilometres from the temperature. Furthermore, some an opportunity to describe (a) breeding areas. The departure time regions occupied by Cormorants how birds belonging to different from the northern parts of Europe during the breeding season are so breeding populations are ranges from July to December but cold during the winter months that distributed outside the breeding most birds migrate south during standing fresh waters freeze for season (see Figures 6.2 and 6.3 September-October. Timing of weeks, or even months, preventing for example); (b) how migration later departure is partly related birds from feeding. However, routes vary among individuals and to drops in temperature. Some Cormorants wintering close to populations; (c) how individuals individuals appear to initiate their the colony are apparently at an time their migration; (d) how site- northerly spring migration from the advantage over long-range migrants faithful individuals are towards Mediterranean (including North because they are able to arrive early certain staging and wintering sites; Africa) as early as late January. For in the season and start breeding as and (e) how Cormorants respond example, colour-ringed birds seen soon as weather conditions become in Tunisia in January have been appropriate, leading to a breeding resighted in The Netherlands by success above average. Birds that mid-February before subsequently move south to warmer parts of reaching the breeding colonies in Europe (and North Africa and Denmark. In more northern areas the Middle East) will, once there, like Switzerland, parts of Germany conserve energy due to higher air and The Netherlands, departure for and water temperatures (van Eerden spring migration mainly occurs in & Munsterman 1986, 1995). The March. However, there is extensive southward migration also has costs variation in individual migration of course due to the energy needed itineraries. Some individuals make for migration and also because birds many small ‘jumps’ between that migrate south tend to return stopover sites whilst others tend and breed later in spring than those to migrate fast by making long- wintering close to the breeding distance flights separated by only areas.

[66] the intercafe field manual

Eerden 2007). Finally, the precision of estimates of survival for first- year birds and older ones can be improved by combining data from recoveries of birds found dead with resightings of colour-ringed birds in breeding colonies (e.g. Frederiksen & Bregnballe 2000, see also Figure 6.3).

6.2 Metal-rings Metal rings, generally include country information in the form of an address to be contacted in the event the ring is found on a dead bird or, less commonly, the information on the ring has been read from a live bird in the field. Metal rings are the most common form of individual bird recognition but are often used alongside colour rings. Metal rings are the usual means by which the general public provide data, particularly from birds that are trapped or shot.

Figure 6.2 Resighting locations of colour-ringed Cormorants from the Oostvaarderplassen colony (Netherlands) in 1983–2005. About 5,000 birds Cormorant chicks have been ringed have been colour-ringed and about 20,000 resightings recorded. with metal rings in Europe since at Map courtesy of van Rijn & van Eerden 2007. least the 1930s. The use of metal rings increased markedly after the mid-1970s. Chicks have been Figure 6.1 Flying Great Cormorants. to climatic variation between ringed with metal rings made of Photo courtesy of J Trauttmansdorff. seasons. aluminium or steel. Rings made of steel are now recommended Marking individuals in breeding because they have a higher colonies with colour-rings and the resistance to wear. The use of metal subsequent reading of these rings rings is entirely coordinated by in the field is an indispensable national ringing centres. tool for quantifying demographic parameters, such as emigration Since the late 1970s metal and immigration, survival, age rings have often been used in at first breeding, and age-related combination with colour-rings. reproductive performance. As Ringing with metal rings alone an example, Figure 6.2 shows (i.e. not in combination with the national and international colour-rings) is usually carried resightings of colour-ringed out because the only aim is to Cormorants from a single colony obtain information about birds in the Netherlands (van Rijn & van found dead or because it is too

www.intercafeproject.net [67] the intercafe field manual

costly (in terms of man power) to ring with colour-rings and handle the resightings afterwards. Furthermore, ringing activities might be limited to the use of metal rings only in order to minimise the duration of disturbance at the breeding colony.

Interpretation of ring recoveries from birds found dead Information about recoveries of ringed Cormorants found dead is frequently used in research aiming to identify the geographical extent of the breeding areas from which staging and/or wintering Cormorants originate. However, the main interpretive problem here is that Cormorants have not been ringed simultaneously throughout their breeding range in proportion to the numbers actually breeding in the different countries. Thus for example, the lack of recoveries on Figure 6.3 Distribution of Cormorants ringed in the Danish Vorsø colony wintering sites in middle European (central Denmark) and found dead during winter (20 November–20 February) countries of Cormorants ringed in showing a total of 552 recoveries from 1947–2008. Belarus probably simply reflects Bregnballe & Gregersen unpublished, ring recovery map by T Bregnballe. an absence of ringing activities in Belarus and does not mean that Cormorants from Belarus would Similarly, there are also difficulties For example, the number of not migrate this far to the west if in using ring recoveries from dead Cormorants recovered drowned in appropriate. birds to describe the temporal use fishing nets might depend on the of staging and wintering areas and number of nets in use (Bregnballe Recoveries of rings are also also the timing of movements. For 1999), and the risk of dying in often used to define the areas example, partly because birds are Dutch, German and Swiss lakes is that Cormorants from a certain rarely recovered on the day they thus apparently higher in January breeding area use outside the die, recoveries of birds found dead than in autumn (Bregnballe et al. breeding season and to identify will tend to give a picture of later 1997). The time periods of culling which of these areas are of departure and later arrival than activity and/or Cormorant control greatest importance (see for was actually the case (Bregnballe actions in various countries also example Figure 5.2). The major et al. 1997). Another difficulty is have to be taken into account as problem here is that mortality that an increase in the number of these may also bias the timing and risks and recovery probabilities recoveries from a certain region locations of ring recoveries. will vary geographically, thus at a certain time of the year may affecting the conformity between be an effect of a local temporary 6.3 Colour-rings the geographic distribution of increase in mortality (and/or ring recoveries and the spatial in reporting rate of dead birds) The Great Cormorant is a good distribution of the population rather than an effect of presence example of the great interest in under investigation. of higher numbers of Cormorants. long-term, intensive individual bird

[68] the intercafe field manual

marking operations, yielding a wealth a distance through a telescope. perched with their legs exposed. of results on population dynamics, This type of ring is by far the Prior knowledge of the preferred migration, timing, reproduction, and most frequently used in Europe. resting places of Cormorants in behaviour as shown in numerous The colour ring is placed on a particular area is thus essential. papers using colour-ring data (e.g. one tarsus, while the traditional Usually, night roosts are good Fiedler 1999, Frederikson et al. metal ring is placed on the other places to find ringed birds because 2002, Sackl & Zuna-Kratky 2004, leg. A bar may separate the they are used by many individuals Bregnballe 2006, Bregnballe et al. digits or letters on the colour- throughout daylight hours as well 1997, 2006, Hénaux et al. 2007, van ring, thus increasing the number as at night. However, although day Rijn & van Eerden 2007). of combinations. However in roosts hold fewer birds than night such cases at a distance it may roosts do, they can be places where Since the early stages of the species’ seem that there are two rings on the birds’ legs are easier to observe. recovery, Great Cormorant chicks the leg instead of one. For instance day roosts often form have been intensively marked with ▪▪ Several smaller PVC rings on such sites as walls and trees colour-rings in several important of different colours, creating along rivers, dams, sluices, and high breeding populations throughout a multi-ring combination, tension poles. Colonies are also Europe. The different colour-ringing including the metal ring as one of good places to spot ringed birds, schemes have been well coordinated, the rings within the combination. although they are more sensitive providing motivating exchange There is no alpha-numeric sites to watch without causing between observers and program code to be read on this type of disturbance. Care must be taken in managers. This has ensured that no marking and so the exact colour- all situations, in the breeding colony miss-match occurred with different ring combination, including the as well as at wintering sites, because individual birds carrying the same metal ring has to be recorded to any unnecessary disturbance to mark. Colour-rings are relatively easy identify the individual. birds should always be avoided. to read on Great Cormorant legs, thanks to the perching behaviour of The coordination of all colour-ring When perched in trees, Cormorants this species (see Figure 6.4). schemes is undertaken in Europe are more prone to show their legs by the IUCN-SSC Wetlands if a light wind is blowing. In these Usually, colour-rings are placed International Cormorant Research conditions they have to stand up on 21–27 (but up to 35) day-old Group. It is very important that a little bit and tend to move more nestlings (see Figure 6.5). However, anyone wishing to colour-ring frequently to keep their balance, recently the colour-ringing of Great Cormorants first contacts the thus increasing the opportunities Cormorants captured as adults has European coordinator of this scheme, for their leg-rings to be visible to been initiated in Spain (see Figure before even planning, ordering, or an observer. In a night roost, during 6.8). Importantly, at all times (and making colour rings. The reference calm and cold weather, birds tend in all situations), Cormorant rings website for colour-ringing Great to stand/sit in a squat position should only be fitted by specially Cormorants is the Cormorant with their legs tucked well into trained and licensed ringers (see Research Group official website: their plumage and so their legs can Figure 6.6). One or several colour http://Cormorants.freehostia.com/ be extremely difficult to check. rings are placed on the tarsus, along index.htm and the current colour- In these circumstances, the best with a traditional metal ring of the ring coordinator for Europe can be opportunity to see the bird’s legs is national ringing centre. Basically, contacted via: [email protected] when a new bird comes back to the there are two types of colour-rings roost because the other birds stand used on Cormorants: up, ready to defend their branch or 6.4 Reading colour-rings on place in the roost, as the newcomer ▪▪ A single large PVC ring of Cormorants tries to land amongst the group. one colour, carrying an alpha- numeric code of generally 2 The best way of finding a ringed A good telescope is essential to or 3 (occasionally up to 4) Cormorant in the field is to scan read colour-rings, particularly characters that can be read from a group of birds while they are if fitted with a 20–60x zoom,

www.intercafeproject.net [69] the intercafe field manual

although a fixed wide-angle ocular (e.g. 30x) or a pair of binoculars are preferred by some observers because it is then easier to scan birds in flight and/or a larger number of perched birds (if they can be approached close enough without disturbance). Depending on weather conditions, colour rings are normally readable up to 300 m (sometimes 700 m). When using a telescope, a sturdy tripod ensuring maximum stability is essential.

The code on the ring should be read vertically e.g. parallel to the leg axis. Do not forget that the codes are repeated (normally 3 times) all around the ring, so if you see only a part of the letters/ numbers, you may be able to see the rest of them on the other side of the ring. Be careful to determine the reading direction Figure 6.4 An adult Cormorant with red colour ring on its left leg and a for the code on the ring (it can be metal ring on its right. Photo courtesy of J Helder. either way, up or down), to ensure that you correctly read ‘6’ and not ‘9’ by comparison with the 6.5 Reporting sightings of In every country, ringing operations other letters in the code sequence. ringed Cormorants are coordinated nationally by a Confusing codes are generally national ringing centre choosing avoided by ringers, or only used As well as the ring code, a from a list available here: http:// once (e.g. ‘666’ and ‘999’, and number of other essential, very www.euring.org/national_schemes/ letters like N, H, S). For the same important and useful pieces of contact_schemes.htm and thus reason not all the letters in the information should be recorded ensuring that a unique metal ring alphabet are used: in Italy, for if at all possible. An essential is associated with an individual example, only 15 letters are used. and supplementary check-list of bird, whatever the species, and Sometimes the ring code may recording details is given overleaf that the record of this ringed bird also include a vertical bar (see (Table 6.1). is kept in a centralised database. Figure 6.4) or far less commonly a No ringing operation should be horizontal one. In order to provide the most conducted without the agreement comprehensive and useful of the national ringing scheme. Needless to say, it can be really useful information possible to the Colour-ringing schemes are to photograph or sketch any ringed ringing scheme, it may be useful however not always coordinated by bird, (and to send them to the ringing to complete a simple colour- the official national ringing centres scheme manager), even if only low ring reporting sheet that can be (as they are for the traditional metal resolution pictures are possible. Such downloaded as an excel file at rings), although colour-ring users photographs or images can hold the following website: http:// sometimes do have an obligation to important details such as the bird’s Cormorants.freehostia.com/index. report their data to their respective plumage, size, or sex. htm ringing centre.

[70] the intercafe field manual

Figure 6.5 Photo left — Cormorant chick ringed with a metal ring on tarsus (Estonia, June 07), Photo courtesy of Stef van Rijn and Photo right — one marked with a plastic colour ring. Photo courtesy of M Marinov.

If a Cormorant with a metal ring is is being reported, the report should observed or found (see Figure 6.7 be sent to the corresponding scheme and also Figure 6.8), send as much manager. According to the code on relevant information as possible to the ring, the relevant ringing scheme the ringing centre in your country can be found using the following (see list on the Euring website: www. updated website: http://Cormorants. euring.org). If a colour-ring sighting freehostia.com/index.htm

Figure 6.7 Example for metal ring recovery: migrating bird shot during Cormorant control actions in Austria, an immature male ringed 17.06.03 in Finland and shot 31.01.04 in Austria. As well as indicating migration routes, graphical documentation of ring recoveries can be used to show plumage characteristics for birds of Figure 6.6 Field workers measuring and ringing Cormorant chicks. known age. Photo courtesy of M Marinov. Photo courtesy of R Parz-Gollner.

www.intercafeproject.net [71] the intercafe field manual

If the relevant scheme can not be Table 6.1 Details of information to be reported when providing a Great Cormorant found however, do not hesitate to resighting. send the data to any coordinator on What to report? Relative Comments the list who will then try their best importance to find the appropriate scheme for Ring code Essential Every detail is needed, including the position that particular ring. of the colour-ring (left or right leg) and the direction of the code-reading (top-down or down-top). Any uncertainty in the reading 6.6 Some ideas for using must be reported (e.g. for potentially colour-rings for new studies confusing letters such as ‘K’ or ‘X’). The position (or absence) of a metal ring (usually on the other leg) should be noted. Large-scale colour-ringing offers opportunities not only for the Place, location Essential Preferably with the geographic coordinates of the site. These can be easily found using various specific colour-ringing Google Earth®. Take care to communicate schemes themselves, but also for the format used to report the coordinates those interested in Cormorants (decimal degrees, degrees and minutes, or anywhere in Europe. There are otherwise). several examples of local studies, Date Essential If a colour-ring bird is repeatedly seen at the not themselves based on directly same location, all the observation dates are ringing the birds but on the useful, not only the date of first sighting. systematic recording of colour- State of the bird Essential Was it alive or dead? If dead, approximately ringed Cormorants in specific how long? locations (Reymond & Zuchuat Age Very important It is useful to check if the bird is adult or 1995a and 1995b, Yésou 1995, immature (this can also help sorting out any Retter 2000, Paquet et al. 2003, misleading readings). Galván 2005). Such systematic Name of the Very important If possible with e-mail address. recording of colour-ringed observer Cormorants can offer unique Habitat Useful Especially if no geographic coordinates are insights on the use of a specific given. Be as precise as possible e.g. bird area, but can also provide details on was along a river, in a harbour, in trees, in a behaviour, local movement, local colony, etc. site-fidelity, and the turn-over of Signs of breeding Very important For birds resighted inside breeding colonies, it individuals. Such local information activity is of great value to know whether or not the can be very useful in the planning bird was engaged in a breeding attempt. For of local Cormorant counts (see example, it is sitting on the rim of a nest. sections 3.3 and 3.4). Type of location Useful For example, is it a resting bird in a night roost? Or a bird seen on a day roost? Or is it on or close to some foraging ground? It is possible to organise regular Flock size Useful Record the number of other Cormorants surveys for ringed Cormorants present at the same site. at one or more roost sites using Time Not essential Can be reported if recorded. several observers who are able to organise themselves and exchange information. For example, if roost 6.7 More general literature and download material can counts are being organised (see information, selection of be found on the web, for example section 3.3), it is highly valuable useful links under the following links: to encourage observers to have a closer look at Cormorant legs and Further useful information about BTO — Bird Ringing in Britain and to record any sightings of metal or bird ringing, as well as related issues Ireland: www.bto.org/volunteer- colour rings that are observed. dealing with bird migration, relevant surveys/ringing/ringing-scheme

[72] the intercafe field manual

Figure 6.8 Example for metal and colour ring recovery showing site fidelity, movements and fly-ways between breeding and wintering sites: this Cormorant has been ringed as a chick in Sweden with a metal ring on the right tarsus. Wintering in Spain the bird was captured a second time as an adult on 03.03.2007 and a colour ring was fitted on the left tarsus. Two years later the bird was sighted and identified by reading the letters on the yellow plastic colour-ring in the field and the bird was reported on 22.08.2009 to be on the southern coast in Sweden again. Photo courtesy of Joan Aymerich, Grup d’Anellament Calldetenes Osona, Spain.

EURING (The European Union for Waterbird migration: Wetlands International (WI) Bird Ringing): www.euring.org; http://www.wetlands.org/ Cormorant Research Group (CRG): see download link to get a copy of Whatwedo/Savingwaterbirds/ http://Cormorants.freehostia.com/ the Euring Brochure (2007) Bird Flywaysforwaterbirds/tabid/772/ index.htm ringing in science and conservation. Default.aspx

European Colour Ring Birding: Wetlands International — follow www.cr-birding.org the bird: http://followthebird. wetlands.org/BirdsWeFollow/ Wetlands International: tabid/1539/language/en-US/ www.wetlands.org Default.aspx

www.intercafeproject.net [73] the intercafe field manual

7 INTRODUCTION Part Two — Working with Fishes D N Carss, R Haunschmid and I C Russell

This section of the INTERCAFE Cormorants (and just as Field Manual is devoted to a brief importantly, choosing the most overview of some of the field appropriate technique to do so) and methods available for assessing interpreting the resulting counts fish numbers in freshwater and are similar to those that have to coastal habitats. It aims to provide be considered when attempting the reader with a description of the to count fish. An understanding commonest techniques used by of the abundance of fish is a vital researchers and fishery managers for requirement for any informed assessing fish populations and of the evaluation of Cormorant-fishery problems associated with deriving interactions and for assessing reliable measures of fish abundance. the effect of measures aimed at Clearly, in the context of the Field reducing the birds’ impact on fish. Manual, interest in fish populations is due to the general concern over Across Europe, there is a broad the potential ‘impact’ of Cormorant diversity of fish species and a wide predation on fishes. However, in variety of habitats in which they live. contrast to Part One, there is less Unlike Cormorants, fish are seldom, focus here on the application of standard methodologies for fish assessment, because aquatic habitats and fish populations vary widely necessitating the use of a broad range of sampling and assessment techniques in different situations. Further, in many instances, well- established fishery assessment protocols are already followed. Nevertheless, we hope that readers will gain a better understanding of the work required to answer the question ‘How many fish are there in that lake, or stretch of river, or along that coast?’ and of the constraints on the resulting data.

Many of the ecological issues Figure 7.1 Typical river systems are dominated by ‘Trout Zone’ habitats in associated with both counting their headwaters. Photographs — Shutterstock.

[74] the intercafe field manual

here but to give the reader a feel for the complexity (in both time and space) of the ecology of fishes and their relationships with the diverse habitats in which they occur. This complexity has consequences for both the appropriateness of available methods to assess fish numbers and the likelihood of identifying and/or quantifying any Cormorant impact on fish at a specific location.

7.1 European fish diversity The published literature on fish diversity and distribution, behaviour, and ecology is vast and beyond the scope of this summary. However, a good introduction is provided by Wheeler (1978), Pitcher (1993) and Wootton (1990). Although the fish fauna of Europe is not particularly rich, the ‘Handbook of European Freshwater Fishes’ (Kottelat & Freyhof, 2007) lists 522 freshwater species plus 24 marine species that are also found in freshwater. There is an even wider diversity Figure 7.2 Typical river systems are dominated by ‘Bream Zone’ habitats in of marine fishes around Europe’s the lowlands. Photographs — Shutterstock. coasts (Hart & Reynolds, 2002). There is thus a huge diversity of fishes in European coastal and if ever, conspicuous and so counting or unit volume), or ‘catch per unit freshwaters, closely linked to them in an accurate, reliable, effort’ (CPUE: typically the number/ specific habitats and environmental and repeatable (i.e. ‘biologically weight of fish caught by a particular conditions. meaningful’) way is not easy fishing gear over a standardised and is commonly constrained time period). However, such sample By necessity, the techniques for by time, methodological and catches/counts are commonly also investigating fish populations financial limitations. Typically, fish extrapolated up to arrive at best are also numerous and diverse. population assessments are based on estimates of overall fish numbers in Knowledge of fish ecology and samples of the ‘population’. Such a wider area. behaviour also forces the researcher data are thus not absolute counts to consider the term ‘population’ but an index of the population It is only through knowledge of and this concept is discussed below. given in such terms as ‘biomass’ or both fish ecology and behaviour Although much of the following ‘standing crop’ (the mass of fish of that researchers can hope to assess discussion focuses on freshwater a particular species per unit area), populations. It is not the intention fishes, it will often be analogous to ‘density’ (numbers per unit area to review all the relevant literature coastal situations.

www.intercafeproject.net [75] the intercafe field manual

7.2 Fish distribution and in a defined area or habitat that movements interact (directly or indirectly) form a ‘community’ and, within Fish species are distributed a community, fishes that exploit throughout particular waters in the same resources in a similar both space and time (Wootton way are termed ‘guilds’. The 1990). A range of factors influence term ‘assemblage’ is often used the distribution and abundance to describe all fish species in a of fish populations, including defined area, regardless of whether seasonal factors and the life-history they interact or not. Often, several characteristics of the species different species may have broadly themselves (e.g. migration, habitat similar habitat requirements (at requirements of different life least at some time in their life- stages, etc.). For instance, many cycles) resulting in assemblages juvenile fish occupy a specific that are associated with particular habitat which changes as the fish broadly-defined habitat types. grow older - individuals of some species spend all their lives close In freshwater, broad-scale to where they were spawned distribution is chiefly controlled whilst other species may make by climatic, topographical and long migrations covering 100s or hydrological differences. In 1,000s of kilometres. Some fish rivers, for example, there is often also make vertical migrations, with a continuous increase in species Eel (Anguilla anguilla). Finally, individuals moving up and down in richness (i.e. total number of species) the river enters the ‘Bream Zone’, the water column at anything from with progression downstream. Thus, the true lowland zone, where the daily to seasonal frequency. Many a typical river starts with a zone gradient is very gentle and the water adult fishes also have particular (the ‘Trout Zone’) characterised by slow-moving. Although oxygen spawning habitat requirements. steep gradients, fast flowing water content is usually good, temperature Apart from specific movements for and cool temperatures and holds is more variable than in the other activities such as spawning, many Brown Trout (Salmo trutta), Atlantic zones and the substrate is often of the movements made by fishes Salmon (S. salar), Bullhead (Cottus silty and the water clarity low. Few are likely to be made in relation gobio) and Stone Loach (Barbatula upland species can survive here and to the need to forage and to avoid barbatula). The ‘Grayling Zone’ only a few upstream species (Roach, predators, and also in response to follows (and is slightly warmer) and Rudd, Perch, Pike) inhabit this zone changing environmental variables can hold all the above species with which is also characterised by Bream such as light levels and water the addition of Grayling (Thymallus (Abramis brama), Tench (Tinca temperature. thymallus), Minnow (Phoxinus tinca) and Carp (Cyprinis carpio). phoxinus), Chub (Leuciscus cephalus) and Dace (Leuciscus Overlaid on this, as one moves 7.3 Fish communities and leuciscus). This, in turn, leads to the westwards in Europe from the assemblages ‘Barbel Zone’, which has a gentle Danube, the freshwater fish gradient, moderate water flow and fauna gradually becomes more Individual fish live within complex temperature, with a good oxygen impoverished in terms of species webs of interactions (e.g. predator- content and mixed silts and gravels. numbers. At the far west (the UK, prey relationships, competition, It is characterised by the previously Ireland and Norway) the native reproduction) and processes (e.g. mentioned upstream species plus freshwater species are actually only the flow of energy and nutrients Barbel (Barbus barbus), Roach those that became trapped in rivers through the web) that can affect (Rutilus rutilus), Rudd (Scardinius as sea levels rose at the end of the (or be affected by) that individual erythrophthalmus), Perch (Perca last Ice Age and many of the current (Wootton, 1990). All those fish fluviatilis), Pike (Esox lucius) and fish fauna are present entirely as a

[76] the intercafe field manual

Figure 7.3 Many European river systems have been altered from ‘natural’ habitats (left) to ‘modified’ ones (right) — often to the detriment of the fish

communities within them. Photographs — Shutterstock. result of human activities (Wheeler changes and anthropogenic of vegetation clearance, draining 1978). However, at a more local activities. It is generally recognised and infilling, often related to scale prevalent habitat types and that the distribution of freshwater the expansion of agriculture or anthropogenic factors are likely to fishes has been altered more infrastructure, eutrophication, be more important in regulating the radically by human activity in the hydrological modification and distribution of different fish species. last few hundred years than was pollution, or for use in aquaculture, ever achieved by natural processes agriculture or industry. The In coastal areas, there is a similar over the preceding millennia. construction of dams and other variability in fish communities Inland water habitats and species structures along rivers has often associated with salinity, water are generally in a worse condition resulted in fragmentation and flow temperature, coastal structure, water than those in other ecosystems regulation in almost 60% of the currents, substrate type and algal (Hassan et al. 2005). These authors large rivers systems in the world. (seaweed) abundance and cover. conclude that more than 50% of For example, the first 1,000 km inland waters (excluding lakes and of the River Danube can be rivers) have been lost in parts of considered an almost uninterrupted 7.4 Ecosystem-level factors Europe during the last century. artificial waterway passing through affecting fishes 59 hydropower dams. In addition, Human activities have affected freshwater fish populations have The following subsection freshwater fish populations also been influenced directly is complied primarily from indirectly by changing aquatic by targeted action (e.g. fishery information taken from work habitats. Habitat degradation management initiatives, harvesting, undertaken from the Millennium is widespread and the species species introductions and transfers). Ecosystem Assessment (Hassan diversity of inland waters is et al. 2005). Ecosystem-level among the most threatened of all The FAO’s (Food and Agriculture factors affect fish in fresh and ecosystems. The direct drivers Organization of the United coastal waters through the fishes’ of degradation and loss are well- Nations) latest global assessment of relationships with aquatic habitats documented and include changes freshwater fisheries (1999) shows and sensitivities to environmental in land use and cover as a result that most capture fisheries that rely

www.intercafeproject.net [77] the intercafe field manual

on natural reproduction (perhaps occurring each year (Hickley 1994), meaning a ‘population’ of fish now around only one-third of all although many of these fish are that is exploited by a fishery and freshwater fishery production on simply relocated between natural which may be subject to some a global scale) are over-fished or sites rather than being farm-reared. type of management. Thus, are being fished at their biological For example, in England and Wales ‘stock’ is typically used for more limit, and that the most important 6,000 separate fish introductions commercially important species. factors threatening them are fish take place each year involving However, it is important to consider habitat loss and environmental some 2 million fish (Environment that predation by Cormorants, for degradation. Similarly, coastal Agency, undated). Similarly, at Loch instance, on a fish stock may not ecosystems are experiencing some Leven a freshwater lake in Scotland, have the same biological meaning of the most rapid environmental over 1,318,000 (7–15 cm long) as predation on a fish population. changes and, in all regions, coastal Brown Trout (Salmo trutta) were To further complicate matters, fisheries have depleted fish stocks. released between1983–1994, plus an many fisheries interests measure additional 40,000 (28–29 cm long) Cormorant impact in terms of the The human population of Europe Rainbow Trout (Oncorhynchus reduction in their fish catches. has almost doubled since 1900 mykiss) in 1993 and 30,000 in In general terms, larger stocks and agriculture and industry have 1994 (Carss et al. 1997). However, would be expected to result in developed massively. According to there is increasing concern over better catches, and this has been the UN Environmental Programme, the potential risks associated with demonstrated for many different almost 60% of Europe’s natural the stocking of fish, particularly in fish species. However, this does not wetlands have now been ‘destroyed’, relation to ecological imbalance and always apply and a range of factors leaving freshwater species declining changes in fish community structure, may influence the relationship at an increasing rate (e.g. see as well as the transfer of disease between catches and the underlying Kottelat & Freyhof, 2007). Recent and the loss of genetic diversity stock size. It has to be pointed out estimates suggest that nearly 40% (see Cowx 1999, Carvalho & Cross that the relationship between fish of European freshwater fish species 1998, McGinnity et al. 2003). catches, fish stocks and, ultimately, are threatened with extinction. fish populations is likely to be However, many new water bodies extremely complicated and very have also been created in recent 7.5 Defining fish populations difficult to quantify in many cases. decades as a result of human activity (e.g. sand and gravel extraction, new Given the complexity of fish As well as the complexity of fish recreational fisheries, etc.) and water distribution and movement in both distribution and movement in both quality is also improving in many time and space, it can be difficult time and space affecting what is places as discharge regulations have to define a ‘population’. A simple understood to be a ‘population’ of a been brought in, effluent treatment definition would be a group of particular fish species, the concept measures improved and with greater fish of the same species that are is further compounded when several general awareness of environmental alive in a defined area at a given species live together in the same issues. time. The area may be defined habitat as an assemblage. Moreover arbitrarily for the convenience of in almost all waters, given that Fish stocking is seen as an researchers/managers (e.g. as a several members of an assemblage important and commonly-used stretch of river or coast) or may may have direct interactions with tool in the management of fisheries be ecologically meaningful for each other, it becomes extremely for commercial, recreational and the population under study (e.g. a difficult to consider one fish species conservation purposes — with lake). These different approaches in isolation in an ecologically fish being stocked for mitigation, need to be borne in mind when meaningful way. This is particularly enhancement, restoration and the considering what effects predators important when considering creation of new fisheries. The scale such as Cormorants may have on potential Cormorant predation, of such stocking is immense with a ‘population’ of fish. Rather than as these birds are generalist many thousands of stocking events populations, many fishery managers predators which tend to take their involving many millions of fish use the term ‘stock’ — essentially prey opportunistically in relation

[78] the intercafe field manual

assessment (often over many years). This level of monitoring is thus usually limited to large, commercially-valuable stocks, and is unlikely to be available for many fish populations that are subject to Cormorant predation.

While many commercial fisheries in larger waters will be subject to regular assessment by the appropriate authorities, such assessments do not commonly extend to smaller water bodies, private fisheries, etc. Thus, the information available for fisheries subject to predation by Cormorants (e.g. stock size and composition, catches, etc.) will vary enormously. Figure 7.4 The definitions of fish populations and stocks, and their Although it cannot be assumed relationship with catches are all issues adding complexity to understanding that parallel fisheries data will the population ecology of fishes.Photograph — Shutterstock. be available alongside that of Cormorants, it is nevertheless possible that some studies on fish at to their general ‘availability’ in ▪▪ Estimates of population specific sites will be useful in this particular habitats. ‘Availability’ is abundance (fish numbers) for all context. a particularly difficult parameter stages of the life-cycle. for researchers to measure but is ▪▪ Measures of the rate of change The techniques described in probably influenced by several of abundance for these stages. the following chapter do not factors — a combination of the ▪▪ Mortality, growth and fecundity necessarily provide all the data relative abundance of various (potential reproductive capacity) required to quantify the population species and size-classes of fish, rates — which are all size- (age) dynamics of a fish species (or together with the ease with which related and which may also be the impact of Cormorants on a they can be located, caught, and functions of the density of the particular species or on members eaten (Marquiss et al. 1998). For population. of a fish community). However, further discussion on Cormorants ▪▪ The relationship between the they are the most common methods as generalist predators, see section abundance of the sexually used by researchers to answer the 10.6. mature portion of the population question ‘How many fish are in and recruitment to a defined age that water body?’ Furthermore, (or size) class — i.e. the stock- these methods are important to 7.6 Estimating fish abundance recruitment relationship. consider as they are usually the ▪▪ The amount of fish flesh ones that provide fish data for From the above, it is clear that generated by a cohort as the any investigation into Cormorant considerable research effort is fish grow and die — in order to impact at fisheries. The subject of required to rigorously quantify a assess production. scientific attempts to quantify such particular fish population/stock. In impact — through the integration order to understand the population Deriving such information is both of both Cormorant and fishery dynamics of a specific stock, the challenging and expensive, and data — is the subject of the final following information is ideally typically requires a continuous chapter of the INTERCAFE Field required: process of monitoring and Manual.

www.intercafeproject.net [79] the intercafe field manual

8 ESTIMATING ABUNDANCE OF FISH SPECIES S França, R Haunschmid, I C Russell and C Vinagre

As noted in the last chapter, there those mainly used in marine/ the seabed or through the water are numerous different techniques coastal fisheries, and those more column. Fish become trapped in the for assessing fish abundance (e.g. commonly used in freshwater, tail of the net — known as the ‘cod- see Nielson & Johnson 1983, although some apply in both. Many end’ (Hemingway & Elliot 2002). Gabriel et al., 2005). It is not of the limitations relating to the Fish of different sizes and species possible to review these in detail estimation of fish abundance apply have very different ‘catchability’, here, but the following provides a equally in both habitats. and this also varies with gear brief summary of the most common design (Jennings et al. 2001). There fish sampling methods that might are three main types of trawl nets, be applied in assessing fish stocks 8.1 Estimating abundance and these are distinguished by the and possibly used as part of an of marine and coastal fish way the mouth of the nets is kept investigation into Cormorant/ open, the habitat in which they are fishery conflicts and assessing species used, and the target fish species: Cormorant impact. As discussed in Chapter 6 of the INTERCAFE Field Manual, a Otter trawls — the net of an otter Fishing gears may be either ‘active’, wide range of factors influence the trawl is kept open by so-called where the gear is moved towards distribution of fish populations in ‘otter boards’ and is used for fish the fish and they are caught in it, the marine environment, although swimming in the water column near or ‘passive’, where the equipment climatic and oceanographic the seabed (‘demersal roundfish’ is static and fish swim into it influences (e.g. ocean currents) are and flatfish), fish swimming in and are captured (Hemingway & probably the main influences. Fish open water not close to the bottom Elliot 2002, Jennings et al. 2001). distribution is also influenced by (‘pelagic fish’), crustaceans Active gear tends to provide seasonal factors and by the life- and molluscs. The gear is most almost instantaneous samples from history characteristics of the species commonly used in estuarine and more or less defined areas, while themselves (e.g. migration, the coastal waters. passive gear usually integrates the habitat requirements of different life collection of fish over a longer stages). Thus, the use of appropriate, Pair trawls — are deployed period of time. Passive gear is targeted sampling methods is between two fishing vessels and therefore most appropriate for needed when estimating fish species are used to capture both demersal sampling areas that are poorly composition and abundance. A brief (generally bottom-living) and defined or unknown (Hemingway description of some of the main pelagic (generally living in the & Elliot 2002). The choice of sampling methods for marine and water column) fish. The gear is equipment also depends very much coastal areas is given below. typically used in shallow water. on habitat type and target fish species. Beam trawls — the net is held open 8.1.1 Trawl nets by a beam and has a heavily weighted For convenience, fishery sampling Trawl nets are funnel-shaped nets footrope on the bottom. It is designed techniques have been split between which can either be towed across to exploit demersal fish and shellfish

[80] the intercafe field manual

However, two or more sheets of netting can also be hung together to create specailsed gill nets known as trammel nets. For demersal species the nets are fixed to the seabed by a weighted ground rope and anchors. Gill nets for pelagic fish are suspended by floats at, or close to, the surface, and one end of the net may be attached to a boat (Hemingway & Elliot 2002).

Gill nets tend to be highly selective, small fishes swimming through the mesh while larger ones are not retained in it. However, the mesh sizes, lengths and depths at which Figure 8.1 Fish sampling with a trawl net in coastal habitat. the nets are positioned can be varied Photograph — Shutterstock. according to the target species. Standard survey gill nets for fresh and can be used in estuarine and be used in freshwater. Typically waters have been developed which coastal waters. It is also commonly these comprise single sheets of contain a range of mesh sizes so used for scientific sampling. netting (either fixed or drifting) that they will catch individuals which are hung vertically in the of lengths from 2 or 3 cm to in The aim of a trawl survey is usually water column to capture demersal excess of 1 m. A good reference to get indices of fish abundance or or pelagic roundfish (i.e. non- to the standard gill net used in density for a specific survey area flatfish species living in the water European fresh waters as required that are then extrapolated to give column either close to the bottom by the European Committee for values for a larger study area on or not), which become caught Standardisation (CEN) is given by the assumption that the necessarily in the nets by their gill covers. Appelberg (2000). limited sample is representative of this wider area. However, as trawls invariably select particular species (or sizes) of fish, they are very unlikely to provide information on the relative abundance of the fish assemblage(s) in a particular area. Similarly, trawling will only be possible in particular water bodies (or areas within them). For example, they may not be appropriate over rocky habitats or in those with substantial aquatic vegetation, and so may not be able to provide data for some Cormorant foraging habitats.

8.1.2 Gill nets

Gill nets are widely used in the Figure 8.2 Various freshwater fish species sampled with a gill net. marine environment, but can also Photograph — Shutterstock.

www.intercafeproject.net [81] the intercafe field manual

Figure 8.3 Setting (left) and extracting catches from (right) a gill net set in coastal habitat.

Photos courtesy of Marisa Batista.

Thus, as for trawls, gill net 8.1.3 Seine nets then gradually pulled ashore. Fish catches may not be completely landed by this technique are usually representative of the relative species Seine nets are usually simple sheets in excellent condition (Jennings et abundance (or size-range) of the fish of netting stretching from a lead line, al. 2001). Seine nets are commonly assemblage(s) in a particular area. designed to keep the bottom of the used in freshwater habitats and However, if used with care gill nets net adjacent to the substrate, and a estuaries as well as on the coast. can be operated efficiently in most floating headline. The nets may also habitats. One major disadvantage incorporate a cod-end in the centre. Seine nets may have quite fine of gill nets is that the fish caught in Seine nets are used in shallow water mesh, thus enabling them to capture them are usually recovered dead, and and are often operated from the all but the smallest fishes. However, they can also catch, and kill, non- shore, when the net is either set by because efficient capture depends target organisms, notably aquatic a boat or is walked out from the on the net being worked smoothly mammals and diving water birds. shore and set by hand. The net is through the water (particularly

Figure 8.4 Very fine mesh(left) allows all but the very smallest fishes to be sampled. Some seine netting operations require considerable manpower (right) and here sandy shores offer ideal sampling habitat for seine netting. Photographs — Shutterstock (left), Catarina Vinagre (right).

[82] the intercafe field manual

Figure 8.5 Fish sampling with a seine net in a rocky coastal habitat where Figure 8.6 Preparations for fish great care must be taken so that the net does not snag on the bottom and sampling with longlines. allow fish to escape.Photo courtesy of Ana Pego. Photograph — Shutterstock. when pulling it to the shore), or so of the water column, which line in some commercial (usually they cannot be used in habitats may be a problem for sampling marine) fisheries. However, the with dense aquatic vegetation or some species. Experience in technique is non-selective for rocky substrates as nets tend to marine fisheries has shown that fish species and sizes, and it can get snagged on these obstructions. comprehensive calibration of the result in considerable catches Similarly, depending on the depth equipment, coupled with accurate (and mortalities) of non-target of the seine net itself, and the water estimation of the acoustic target organisms, especially diving water it is being fished in, this technique strength of the species present birds. will not always sample the entire throughout their size range, is water column. crucial before reliable, quantitative 8.1.6 Catch data estimates can be generated. 8.1.4 Hydroacoustics Information on catches is widely 8.1.5 Longlines used for assessing fish stocks, Hydroacoustic techniques are particularly those in the marine used most extensively in marine Longline fishing uses large environment. Such data are of areas. However, the techniques numbers of baited fish hooks particular value where they are also have applications in larger attached at intervals to a main accompanied by a measure of effort freshwater bodies where other line (the longline) by means of so that data can be presented as methods can be difficult to apply branch lines called snoods. Each catch per unit effort (CPUE) indices. (Winfieldet al. 2009). This method snood is a short length of line Coordinated catch monitoring can provide an overall indication attached with a clip or swivel and programmes operate in most of fish abundance in an area and with a baited hook at the other commercially exploited fisheries. an approximate indication of end. Longlines can be set at the For species/fisheries subject to individual sizes, but it does not water’s surface or on the bottom such independent sampling, or provide reliable information on and be either fixed by means of an where a reliable licensing and catch species composition. Furthermore, anchor or left drifting. Hundreds reporting system operates, catch this technique as usually deployed or even thousands of baited data are likely to be particularly cannot detect fish in the top 2 m hooks can hang from a single reliable and useful.

www.intercafeproject.net [83] the intercafe field manual

8.2 Estimating abundance of Table 8.1 Summary of fish sampling methods used in the UK to evaluate the freshwater fish species status of fish stocks. Number in brackets indicate the minimum number of personnel required for the survey. There are various techniques that enable the status of freshwater Water Body Type Sampling Gear Used Sampling Strategy fish stocks to be assessed, and the Small streams up 1 hand-held electrode, DC, Depletion sampling between objective of any survey will be to 5 m wide. 50/100 Hz PDC electric fishing, stop-netted sections. central to assessing which of these generator supply, wading (3). may be appropriate in any situation. Small rivers 1 or 2 hand-held electrode(s), Depletion sampling between 5–15 m wide, DC, 50/100 Hz PDC electric stop-netted sections, if For example, various techniques pool-pool/riffle fishing, generator supply, possible. (e.g. electro-fishing, seine netting, topography. wading (3) or boat-based in or gill netting) might be used to deeper river sections (3–6). provide ‘snapshot’ estimates of a Small rivers Boat-based, more than two As above. fish population in a representative 5–15 m wide, hand-held electrode(s), DC, area/stretch of water and enable pool topography, 50/100 Hz PDC electric fishing, One-catch relative results to be extrapolated to the greater than 1m generator supply (3–6), multiple assessment, calibrated whole water. Other techniques deep. anode boom array (4). sampling. (e.g. observation by divers, hydro- Large rivers and Boat-based, 2 boats (7–8), more Depletion sampling. acoustics) may provide information canals, over 15 m than two hand-held electrode(s), Calibrated sampling. on fish stocks without the need to wide and greater 50/100 Hz PDC electric fishing, Relative assessment. handle the fish, while fish counters than 1 m deep. generator supply. Multiple Capture-recapture. can be used to record numbers of anode boom arrays (4), 4–7 Depletion or calibrated fish over an extended time period kVA generators. Seine netting sampling. (e.g. to assess the numbers of (‘wrap-around’ technique) (at Catch effort and trend adult Atlantic salmon Salmo salar least 6). Catch statistics/licence analysis. and migratory brown ‘sea’ trout returns. Calibrated biomass and Creel census. density estimates. S. trutta entering river systems). Hydroacoustics (2). However, the choice of method Large still waters. As for large rivers, electric As for large rivers. may be dictated as much by fishing and netting in margins considerations of cost effectiveness only. Gill nets. as by the accuracy of the results obtained. As an example, a summary of fish sampling methods fishing survey can be extrapolated mark/recapture experiments). used in the UK to evaluate the to the total area of the system Standardised assessment methods status of fish stocks is given in and tentative conclusions drawn have now been identified (European Table 8.1 (Cowx 1995). about the nature and extent of the Standards — EN) for use under the stocks of fish within it. Due to EU Water Framework Directive Applied correctly, many freshwater seasonal changes in the biomass for defining the ecological status sampling techniques can be used of a fish stock, species-specific of water bodies (EN 14011, CEN to generate quantitative or semi- migration periods, and the habitat 2002; EN 14962, CEN, 2004; quantitative estimates of freshwater requirements of specific life-stages, EN 14757; CEN 2005). In many fish biomass, but in some situations the timing and choice of sampling instances, these methods will also this may be impractical due to sites must be considered in detail. be appropriate for addressing river topography, current velocity, questions related to Cormorant- water depths, obstructions in the Statistical methods provide fishery interactions. Techniques water course, or because they alternative means of assessing fish for stock assessment in inland are potentially damaging to a populations in both freshwater fisheries are discussed at length in valuable fishery resource. Provided and marine environments (e.g. Cowx (1996) and some of the more enough sites are sampled within the use of catch and catch per common assessment methods are a catchment, then results from a unit effort [CPUE] data, or outlined in a little more detail below.

[84] the intercafe field manual

8.2.1 Netting which the fish is exposed, and affect gear efficiency (Reynolds larger fish will be influenced more 1983). Both seine nets and gill nets than smaller ones (Murphy & Willis (described above) are widely used 1996). Stunned fish are then usually In rivers, sampling is usually for sampling fish populations in removed using hand-held nets and carried out at fixed sites delimited fresh water, although trawl nets and placed in aerated water tubs to by seine or stop nets, man-made or fixed nets (e.g. fyke nets) are also recover. Susceptibility to electro- natural barriers, or along designated used in some instances. Relative fishing varies among species stretches of a watercourse. The changes in fish community structure because of innate differences in sampling runs may be repeated two and population size of those species anatomy and behaviour, and other or more times at each site to enable and sizes caught in the nets can parameters such as water turbidity, depletion estimates to be made (see be assessed directly by netting conductivity and temperature also review by Cowx, 1983). Where surveys where these are repeated on a regular basis. The data generated through netting surveys can also provide quantitative estimates of fish abundance in sampled areas, for example where repeat (depletion) or mark-recapture surveys can be carried out (discussed in more detail under electro-fishing below). However, netting surveys may often generate more qualitative information, for example in relation to the species and size composition of fish in the catches in a particular water, and data may be restricted to particular species and/or sizes of fish.

8.2.2 Electro-fishing Electro-fishing is often used for survey work in freshwater, principally in rivers and streams (in saltwater it is impossible to maintain electric fields of sufficient strength due to the high conductivity of the water). When direct electric current is passed through water, a fish within the electrical field will be attracted towards the anode (the positively charged electrode) becoming narcotised temporarily as it nears the anode. Fish exposed to alternating current are stunned by, but not attracted to, the electrodes. The intensity of this effect depends upon the potential gradient to Figure 8.7 Fyke nets hanging up to dry. Photograph — Shutterstock.

www.intercafeproject.net [85] the intercafe field manual

Figure 8.9 Electro-fishing from a boat in a wider section of river.

Photo courtesy of W Hauer.

Figure 8.8 Electro-fishing with backpack ‘shocker’ in a narrow, shallow section of river. Photo courtesy of W Hauer. sections are sampled using a single areas (less than 0.7 m depth), run, the resulting data will, at best, boat-mounted electric generators be semi-quantitative. Alternatively, or backpack ‘shockers’ can be used on such things as species mark-recapture methods can with varying numbers of anodes composition and fish growth rates. be used. Catch probability and according to the width of the stream catchability also need to be taken and the equipment being deployed. 8.2.3 Fish traps into account in assessing fish In non-wadeable areas, of up to 2 m densities using this approach, as depth, fishing has to take place from Many different types of fish these will vary with the species and a boat using equipment powered by traps can be used to sample fish size of fish and water conditions a generator (CEN 2002). populations. For example, in rivers, can influence the results. In general, fish traps can be used to provide fish are harder to catch effectively Large river sections can be sampled data for migratory species such as in faster-flowing shallow water using boom-mounted, multi-anode Atlantic Salmon (Salmo salar) and and in water deeper than about electric fishing equipment, and this sea trout (migratory Brown Trout 2 m. Catch probability describes can be effective at depths of up to S. trutta, both adults and smolts). the depletion rate over successive about 2 m. It may be possible to However, traps rarely operate with runs (based on certain assumptions) make some assessment of the fish 100% efficiency and it is usually (Hilborn & Walters, 1992), while evading capture, although this will necessary to estimate the trapping the sampler has to estimate be severely constrained in more efficiency in order to use resulting catchability. In shallow (wadeable) turbid water. Where sections are data quantitatively. This can be water bodies the depletion rate sampled using a single run, the achieved by the use of mark and should be more than 0.5 in order resulting data will, at best, be semi- recapture techniques, although even to fulfil the criteria for calculating quantitative. then the efficiency of the trap can fish density after Seber & Le Cren vary with environmental conditions (1967) or De Lury (1947). Netting and electro-fishing throughout the duration of the run surveys are more difficult to apply and the efficiency of tag or mark Different electro-fishing equipment quantitatively at still water sites, detection must also be taken into is available for different habitats particularly larger ones, and at such account. Estimates of Salmonid and sampling depths. In wadeable sites may only provide information populations obtained using these

[86] the intercafe field manual

Figure 8.10 A fish trap showing outstretched mesh ‘wings’ that guide fishes

in to a central holding compartment. Photo courtesy of W Hauer.

fish in freshwater, most commonly are capable of providing an accurate in large rivers (of more than 2 m estimate of numbers of adult methods may have wide confidence depth) and lakes to catch bottom- fish returning to a river system. limits attached to them. dwelling species. Longlines are However, sub-sampling of the unlikely to provide quantitative catch is often required to accurately Fish traps, and other static fishing data on fish stock size (although apportion the counts by species or gear such as fyke nets, can also may provide an index based between particular age groups. be used in slow-moving waters or on catch per unit effort — see still water situations to sample fish. 7.2.6) and may be most useful for 8.2.6 Catch records and However, such gear may be highly obtaining information on species CPUE data selective in the species and sizes of composition (CEN 2004). However, fish that are captured and is unlikely it is unlikely that all species (and/or Catch and Catch Per Unit Effort to provide quantitative data on fish sizes) of fish in most habitats will data may provide a useful index population or stock size, or reliable have an equal probability of being of abundance for freshwater fish information on species composition. caught. Measures should always stocks (see review by Cowx 1991), be taken to prevent accidental particularly where these exist as a It should be noted that fish traps catches (and death) of non-target reliable time series. The catch taken can capture non-target organisms, organisms. by fishermen/anglers also provides which may then drown within them. a direct measure of stakeholder In particular, in areas frequented by 8.2.5 Electronic fish counters satisfaction. In some large European otters (Lutra lutra), trap freshwater bodies, and for certain mouths may have to be fitted with Fish counters are mainly used particularly high value species such ‘otter guards’ to prevent the animals for monitoring runs of migratory as Atlantic Salmon, long-term catch entering the traps. adult Salmonids (trout and data are available. However, the salmon). Their efficiency depends ability to compare such data across 8.2.4 Longlines upon the reliability and design years requires some knowledge of the apparatus and appropriate of associated fishing effort. Even Longlines (described in section positioning of the facility. Properly assuming reasonably accurate catch 8.1.5) can also be used for catching set up and validated, such counters data, the issue of relating fish catch

www.intercafeproject.net [87] the intercafe field manual

to fish population or stock size may need to be considered. In general terms, a larger stock size would be expected to result in better catches, and such relationships have been demonstrated for many species. However, a consistent catch/stock relationship cannot be assumed, and exploitation rates can vary between fisheries, between species and over time.

Data from angling competitions (‘matches’) where good catch records are kept, and from questionnaire surveys of recreational anglers can also be successfully used to evaluate some fisheries, particularly larger bodies of still water and wide, deep rivers where most other techniques are ineffective. Such methods can provide a cost- effective method of monitoring long-term changes in a fishery, potentially over large areas or long stretches of rivers. The resulting data generally yield information on changes in fish community structure (or at least that part of it that is caught routinely) and relative abundance, rather than any quantitative measure of stock size per se. However, match catches Figure 8.11 Fish can be marked with tags (top) or with injected coloured dye can be employed in mark-recapture (above). Photos courtesy of W Hauer. exercises to yield quantitative estimates of fish densities. fish population (assuming certain population, they are randomly 8.2.7 Mark-recapture underlying criteria are satisfied). dispersed within it, the marking methods The method can be applied equally procedure does not affect the to migratory species (e.g. for behaviour or catchability of the Overall fish numbers in a specific assessing runs of adult Salmon or fish, and all fish in the population water can be estimated by giving smolts), or for non-migratory trout have an equal probability a sample of the fish population and other freshwater fish within a of recapture. Furthermore it within it a distinguishing mark defined area. is assumed that there is no (or tag) and releasing them back immigration or emigration, into the original population. The When arriving at population recruitment or mortality over the ratio of marked to unmarked fish estimates using this method it sampling period and that all marked subsequently recaptured can be is assumed that: marked fish or tagged fish caught are detected used to estimate the size of the are representative of the whole and reported.

[88] the intercafe field manual

8.2.8 Hydroacoustics still be used to identify trends and trawl surveys, more commonly alert fisheries managers to potential used in large, deep rivers and Hydroacoustic techniques are problems. There are a number of estuaries, may positively select for increasingly being used in inherent problems associated with bottom-dwelling species and mesh freshwater situations, particularly sampling fish stocks that need to be selectivity may also apply. on larger water bodies where other borne in mind and some of the most fish-sampling methods may be important are discussed below. In freshwater, electro-fishing more difficult to apply (Winfield becomes progressively less et al. 2009). Such methods need to 8.3.1 Fish distribution efficient as the area and depth of take account of variability in fish the sampled water body increases. distribution, since this is known Fish may not be evenly or This technique is also selective for to change both diurnally and uniformly distributed throughout larger individuals since the intensity seasonally for many species. In a water system, either in time of the effect depends upon the fresh water, higher fish densities or in space. For example, many potential gradient to which the fish tend to be recorded at night and in species aggregate in shoals, or is exposed. The recent development the summer or autumn months than may be associated with particular of boom-mounted multi-electrode at other times of day or season, habitat features, and other species electro-fishing equipment has presumably as a consequence of undergo seasonal migrations. Such improved the efficiency of sampling changes in fish behaviour (Winfield discontinuous patterns of distribution fish populations in larger rivers, et al. 2007). can make stock assessment particularly for smaller shoaling particularly difficult and mean that species. However, sampling 8.2.9 Fish production data stock estimates (e.g. numbers per stillwater systems in this way unit area or unit volume) vary both remains problematic. In such In the case of both intensive and temporally and spatially. situations, electro-fishing tends to extensive fish farms, managers will be more efficient at catching larger, typically have a reasonably robust 8.3.2 Gear Selectivity territorial species such as trout than knowledge both of fish numbers at for catching small, shoaling species. various stages of production and The efficiency (or selectivity) of the expected harvest levels. In of nets and almost all other fish 8.3.3 Gear avoidance addition, regular monitoring (e.g. sampling gear invariably gives of the quantity of food consumed rise to sampling errors. This can behaviour by the fish) may provide a fish be the result of the positioning Some species of fish are more adept farmer with an indirect estimate of of gear (e.g. on the bottom or than others at avoiding capture, fluctuations in a fish ‘population’ at the surface), whether it is or are not adequately sampled by (based on experience of the moving or static, and also a certain methods. The assumptions expected consumption rates for consequence of features of the underlying electro-fishing, seine particular species and sizes of fish gear itself (e.g. net mesh-size netting, depletion, and mark- under particular conditions). and hanging arrangement). Nets recapture estimates may be violated usually only catch a certain group by initial (or acquired) gear of species, or a particular size- avoidance behaviour by fish. 8.3 Problems in estimating range of fish, and smaller fish the abundance of fish (individuals and/or species) are 8.3.4 Stocking commonly under-represented in Commonly, the results obtained fish biomass estimates. This may Assessing fish stock sizes, in the from fish stock assessment have implications for researchers context of evaluating the impact of techniques do not provide an and fisheries managers, as fish Cormorant predation for instance, estimate of absolute abundance recruitment is very variable and also necessitates consideration of of the fish stocks present, but information on year-class strengths the role that stocking might play they may give an index of stock may therefore be missed in many in influencing fish population abundance. Nevertheless, they can sampling programmes. In addition, structure and size. Stocking is

www.intercafeproject.net [89] the intercafe field manual

carried out routinely at many techniques described above) can be in comparison with the mean value freshwater sites in order to enhance substantial over time. For example, derived from the full six years of stock levels, commonly for angling Figure 8.12 illustrates variation in sampling (Figure 8.13). purposes. However, as discussed Brown Trout density at the same in chapter 6, there are concerns small stream site over several Ideally, therefore, data sets should over such stocking in relation consecutive years. This dataset be large and cover extended periods to ecological imbalance and the clearly indicates differences in fish so that natural variability in the possible effects on fish community abundance (numbers per hectare) fish population estimates can be structure. There is also the both among seasons and between quantified from them. Only then possibility that stocked fish may years (Haunschmid 2004). The is it possible to explore what be more vulnerable to predation stream section was easy to survey other factors may be responsible than wild conspecifics, since and had a high catch probability, for unexpected changes in fish although anti-predator behaviour and the resulting estimates are abundance as measured through in fish is in part based on inherent considered to reflect the natural standard sampling techniques. predisposition, this is also acquired variability in fish numbers at this through learning during a fish’s particular site. 8.3.6 Reliability of data lifetime (Magurran 1990a and b, Kieffer & Colgan 1992). Statistical sub sampling from Estimates of fish stock abundance the data presented above (Figure are frequently influenced by a 8.3.5 Variability over time 8.12) indicates that calculating fish range of potential confounding abundance from any one (randomly factors and fish density estimates Natural variability in estimated chosen) sample out of six leads to a often have large levels of fish numbers (from any of the mean difference of more than 20% associated statistical error. This

Figure 8.12 Density of Brown Trout in a small stream estimated from repeated standardised sampling over a 6-year period.

[90] the intercafe field manual

Figure 8.13 Differences in the estimated density of Brown Trout in relation to sample size.

needs to be recognised when be obtained from the fish sampled specifically related to particular interpreting results and using for stock abundance estimates is predators and may provide further these in estimating levels of the proportion of individuals that information on the potential impact Cormorant impact for example (see are wounded or damaged by bird of predators on their prey. chapter 10). In this context, one predators such as Cormorants (see additional piece of data that can chapter 9). Such wounds can be

www.intercafeproject.net [91] the intercafe field manual

9 INDICATORS OF CORMORANT DAMAGE H Engström, P Musil and I C Russell

The wounding of fish by and grasped by the of the bird, 9.1.2 Heron Cormorants is widely reported in and these marks are characteristic the scientific literature and there of specific bird predators. Fish injured by herons (e.g. Grey is evidence that wounding rates Heron Ardea cinerea) tend to have and bird densities are correlated 9.1.1 Cormorant characteristic parallel wounds (Klee, cited in Russell et al. 1996). caused by the edges of the bird’s Although Great Cormorants are The Great Cormorant has a large bill. Such fish are often damaged known to be vectors for various powerful beak with a strong hook, on both sides often around the fish parasites and possibly also which produces a characteristic fish diseases (Beveridge 1988), the deep triangular wound or slash extent to which the birds might on one side of the fish (often influence parasite infestation rates puncturing the body wall), and or disease incidence in fish (and an area on the other side where ultimately fish population size) scales have been scraped off by the is unclear. Behavioural changes lower mandible (van Dobben 1952, associated with Great Cormorant Ranson & Beveridge 1983, Carss & presence, and reduced catchability Marquiss 1991, see also Figure 9.1). of fish, are widely reported by In other cases, fish may be gripped fishery managers (Callaghanet tightly on the head (Carss 1990). al. 1994, Carss & Marquiss 1995) but once again it has not been possible to quantify the impact of behavioural or stress responses at the fish population level.

9.1 Wounding of fish In addition to consuming fish, predators such as cormorants are commonly reported to injure or damage fish, as a consequence of capturing some individuals that subsequently escape. Such wounded fish have been reported from both fish farm operations and the wild. Damaged fish bear marks Figure 9.1 Typical wounds caused by Great Cormorant. consistent with having been hooked Photo courtesy of W Hauer. Photo inset courtesy of J Kortan.

[92] the intercafe field manual

epidermis and scales makes fish more susceptible to secondary infections, such as bacteria and fungi, and this can lead to the breakdown of body tissues (Carss & Marquiss 1991). Some injured fish may die from the infections or wounds, but the extent of such mortality is difficult to quantify.

9.2 Damage at fish farms At fish farms, scarred individuals may be unmarketable and thus represent a direct economic loss. Surveys at Salmonid cage farms in Scottish lochs, where Great Cormorants appear not to enter cages but attack fish through the netting (Carss 1993), have indicated relatively low levels of fish with Cormorant damage (<1% of Rainbow Trout Oncorhynchus Figure 9.2 Typical wounds caused by heron. Photo courtesy of J Kortan. mykiss, mostly 5–9 cm in length). However, the frequency of operculum (gill cover) region, as although the beak tip occasionally wounding depends on fish age, the fish are generally attacked from punctures the skin, the resulting size and species, and on whether above. The whole fish body may wounds are usually less severe than fish are held in net cages, ponds be covered in such marks if the those caused by Cormorants (Carss or lakes (Beveridge 1988). Studies heron has manipulated the fish for & Marquiss 1991). in the Netherlands have recorded some time prior to it being dropped the occurrence of ‘many’ wounded (Carss & Marquiss 1991, see also 9.1.4 Gulls fish at a large Carp Cyprinus Figure 9.2). In some instances, carpio culture site where up to larger fish may have wounds Based on observations at some 500 Cormorants fed (Moerbeek et consistent with having been stabbed fisheries in Northern Sweden, al. 1987). In Denmark, 32-45 cm by the heron’s bill. gulls (Larus spp.) may also inflict Rainbow Trout placed into pound wounds on fish held in cages (H. nets (fish traps) suffered a 3.5% 9.1.3 Shag Engström, unpublished). Marks damage rate in the first 8 hours caused by gulls seem to be difficult (Dieperink 1993). The Shag (Phalacrocorax to tell apart from damage caused aristotelis) is similar to the Great by Cormorants, although gulls lack Data documenting the relationship Cormorant in physical appearance, the hooked upper mandible and between the body size of but is a smaller bird and has a are unlikely to puncture fish in the Carp and the relative levels of correspondingly smaller beak same way. No detailed studies have damage caused by Cormorants which lacks the strong hook of the yet been carried out on the damage at Carp ponds are available Cormorant. The mandibles are finer to fish caused by gulls. from Oberlausitzer Fischteiche and their edges more parallel. The (Saxony, Germany). The highest resulting wounds are smaller than The wounds suffered by fish vary levels of damage were recorded in those made by the Cormorant, and in their severity. Damage to the Carp with a length of 26–35 cm

www.intercafeproject.net [93] the intercafe field manual

Table 9.1 The number of Carp injured by Cormorants in relation to fish body length. Samples of 200 Carp were taken from 20 different stocks in Saxony (Oberlausitz). Injuries are presented as percentage of all samples (Seiche & Wünsche 1996).

Carp length Small injuries Moderate Large injuries Total injuries (cm) injuries 5–10 3.2 - - 3 11–15 12.0 - - 12 16–20 6.1 0.8 0.2 7 21–25 10.6 4.1 0.5 15 26–30 14.9 1.9 1.2 18 31–35 16.7 2.8 - 20 36–40 12.1 0.4 - 13 41–45 4.4 1.5 - 6

(Seiche & Wünsche 1996). This fish ponds (Czech Republic), probably reflects the fact that where ponds contain various smaller fish (<25 cm) are more age cohorts of Carp, indicates likely to be manipulated and that Cormorants selectively use swallowed successfully and thus fish ponds containing one-year less likely to escape, while larger old Carp (10–20 cm in length). by Cormorants and the body length fish (>40 cm) are mainly be too Similarly, fish of this size were of the fish (Table 9.1). Most (50%) large for Cormorants to attack. also recorded more frequently in injured Carp are in the length However, this finding probably to their diet (Musil et al. 1995, Musil range 26–40 cm. It is thought that some degree also reflects prey-size & Janda 1997). smaller fish are swallowed with few selectivity — with smaller fish escaping, whilst larger fish are less probably being the optimal size for Researchers in Saxony have found commonly hunted by the birds. foraging Cormorants. For example, a strong relationship between the evidence from South Bohemian proportion of Carp that are injured 9.3 Damage to wild stocks Levels of fish-eating bird damage in different wild stocks are summarized in Russell et al. (1996). For example, various studies on lakes and rivers in Germany and Switzerland have noted wounding rates in the spring (after winter Cormorant flocks depart) of 10–25% (sometimes 50–60%) among species such as Grayling Thymallus thymallus (see Figure 9.4), Brown Trout Salmo trutta, Whitefish Coregonus spp., Chub Leuciscus (Squalius) Figure 9.3 Typical imprint of Cormorant´s beak on a two-year-old (‘mirror’) cephalus and Pike Esox lucius Common Carp. Photo courtesy of J Kortan. (Staub et al. 1992, Staub & Ball

[94] the intercafe field manual

between fish species (Engström 1998). Habitat use and the conspicuousness of different fish species may also affect encounter and wounding rates, with ‘non- cryptic’ species likely to be more frequently attacked than ‘cryptic’ species. The toughness of fish skin may also affect the severity of wounding. For example, some fish species like Eel Anguilla anguilla, which are common prey of Cormorants at certain times, rarely appear to carry open wounds. Wounding rates are also likely to vary with the foraging and prey manipulation experience of the bird, with higher wounding rates likely to occur with inexperienced immature birds.

Figure 9.4 Whitefish Coregonus( spp.) injured by Cormorant. Ultimately, fish damaged by Photo courtesy of W Hauer. predators such as Cormorants are often conspicuous in catches and levels of damage can be relatively 1994). Other studies from this were significantly higher for high for particular times and places. area reported that some Grayling Chub than for Dace. In Denmark, It is also possible that wounded showed healed wounds from a experiments in coastal pound nets fish are more likely to be caught previous winter’s attack, and (where fish were constrained within in fish samples and thus their true that levels of wounding were not fish traps) indicated that 13–35% abundance in the population is considered to adversely affect of trapped Cod Gadus morhua, overestimated. The proportion the species’ population dynamics Herring Clupea harengus and of damaged fish which survive (Suter 1995). Flounder Platichthys flesus were their injuries is not known, but wounded by Cormorants during clearly the most severe wounds On a river in North West England, each tidal cycle (Cornelisse & are likely to be fatal and even less 2–18% of Dace Leuciscus Christensen 1993). serious wounds can cause direct leuciscus, Roach Rutilus rutilus economic losses as such individuals and Chub had marks resulting from Other authors have also noted are often unmarketable. At the Cormorant attack (Davies, cited higher wounding rates among population level however, the in Russell et al. 1996). Damaged some species than others, for effects of such damage (or those of fish were found to be significantly example Whitefish caught in secondary effects through increased larger than undamaged ones, and pound nets. This finding possibly susceptibility to disease) are the proportions of stock damaged reflects behavioural differences currently poorly understood.

www.intercafeproject.net [95] the intercafe field manual

10 ATTEMPTING TO INTEGRATE CORMORANT AND FISH DATA D N Carss, R Haunschmid, R Parz-Gollner, I C Russell and J Trauttmansdorff

10.1 Introduction others (e.g. when the understanding qualitative approaches are to be of impact relies more on human taken seriously and, indeed, a This final Chapter of the attitudes and values (see for qualitative approach may yield INTERCAFE Field Manual is example chapter 10 of Marzano & in some cases a more accurate devoted to the subject of scientific Carss 2012), it will be necessary assessment of impact on social attempts to quantify Cormorant to assess impact qualitatively. grounds as experienced by fisheries impact on fisheries — through the It is important to realise, in this stakeholders undertaking their integration of both Cormorant and case, that both quantitative and commercial and/or recreational fishery data obtained by researchers often using the methods outlined and discussed previously in Part In managing fishery resources, managers need to consider the state of One and the first section of Part the resource and the provision of fishing opportunities. Satisfactory Two, respectively. Whilst this fishing depends on the presence of healthy fish in appropriate numbers; chapter focusses on the viewpoint regulation of fisheries cannot be considered in isolation from the of the ecological researcher, and ecosystem of which fish are a part. thus may of be of limited practical value to most fisheries stakeholders The word ‘fishery’ has a range of meanings, which are important in or managers, it is hoped that it understanding the Cormorant-fishery conflict and the potential ‘impact’ will also shed some light on the of the predators on their prey. ‘Fishery’ can mean: issues facing biologists addressing ▪▪ The ecosystem including its fish, whether or not these are fished for Cormorant predation issues. (to ecologists).

Before discussing the integration ▪▪ The enterprise managed (to fisheries managers). of Cormorant and fishery data ▪▪ The place where they fish (to fishermen/anglers). and predator-prey relationships more generally, it is important ▪▪ The sphere of activity covered by the relevant law (to policy to take some time to consider makers). how Cormorant ‘impact’ can Thus, conflicts (and ‘impact’) might therefore be viewed at: be understood or interpreted by ▪▪ The ecosystem level – as changes on fish stock size or species different people, depending on composition (which may operate over the long-term), and their often diverse values, attitudes and expectations. Impact clearly ▪▪ The ‘enterprise’ level – as reduction in fishery income, and/or means different things to different ▪▪ The resource/individual level - as reduction in amenity value (both people. In some circumstances of which may be felt directly in the short-term). (e.g. biological or economic), it may be possible to attempt to Text Box 10.1 An example of spatial scale when considering Cormorant impact on measure impact quantitatively. In fisheries (modified from Russell 2006).

[96] the intercafe field manual

be easy. Smaller water bodies (particularly ‘closed’ ones within well-defined spatial limits) and the fish in them often react more quickly to changes in environmental factors (both biotic and abiotic) than do larger ones that can usually buffer any effect to some degree at least. It could be similarly argued that Cormorant predation is more likely to have an impact on a particular fishery or fish species operating within a smaller water body than within a larger one (or a series of similar habitat patches spread throughout the landscape).

Figure 10.1 Harvesting fish from a Carp pond – a relatively small ‘closed’ Furthermore, within this concept system. Photo courtesy of INTERCAFE. of ‘frames’, careful consideration must be given to the criteria upon which any impact will be judged. activities. It is also necessary to catches even if a popluation Obviously, if there is an impact, think about the ‘frame’ within bounces back in the longer term. it should be measurable — but which knowledge of Cormorant From a biological perspective, what should be measured? How? impact and associated bird and long-term population resilience to By whom? Over what period of fishery data are to be interpreted. predation might be equated with time? What information would be Clearly there are both spatial and no serious damage, but this view required to demonstrate impact? temporal elements to this but there is unlikely to be shared by fishery These are all pertinent questions, are also a variety of scales over managers or owners, recreational though some may feel they are just which impact at a fishery (a term anglers or fish farmers whose adding unnecessary complexity to with a range of meanings) can be catches (income or amenity values) something that is obvious — that considered (see Text Box 10.1). are depressed in the short-term. Cormorants do have negative impacts on fisheries and thus The issue of temporal scale is In most circumstances, the should be controlled or managed to important because much of what term ‘impact’ is implicitly used reduce this. Nevertheless it should is discussed in relation to fish to describe a situation where be acknowledged that there is often populations (e.g. compensation, Cormorants are consuming serious disagreement over whether lack of population effects) occur sufficient fish in a system so as or not Cormorants cause an impact over the long-term, probably on to affect it negatively from a at fisheries precisely because the a generational scale. However, human perspective. This almost term itself means different things to many conflicts with fisheries invariably equates to a reduction different people who use different occur over short timescales (i.e. to fish catches, fish value, or to a knowledge bases to quantify it. weeks/months). For example, specific portion of that catch (e.g. there is widespread agreement a particular species or age/length that Cormorants often forage very class of individuals). However, 10.2 Interpreting efficiently at a site and then, once choosing the ‘correct’ spatial understandings of Cormorant prey becomes relatively scarce, scale within which impact is ‘impact’ at fisheries move on to an alternative one. Thus, both biologically meaningful and one might well expect to see short- meaningful to specific fisheries One very practical way of exploring term impacts such as depressed stakeholders may not always the complex potential impacts

www.intercafeproject.net [97] the intercafe field manual

As with all European Directives, fisheries), one very important Member States are given certain outcome of this approach is the flexibility about the way in which diversity of interpretations across the Birds Directive’s binding Europe. This is particularly true in obligations are achieved. Put the case of derogations whereby very simply, Member States are Member States request permission signed-up to the obligations of to manage the species. They need to the Directive but the precise route consider carefully what constitutes for getting there, constrained to ‘serious damage to fisheries’ and some degree, is left partly to the ‘no other satisfactory solution’ for discretion of individual Member instance, and there appear to be States. In terms of the management differing interpretations. of Cormorants across Europe (and, by implication, the rationale Prior to the REDCAFE and behind the necessity to undertake INTERCAFE projects, attempts management in the first place to resolve these problems (i.e. their impact or damage to relied largely on the work of

Figure 10.2 Like almost all European Article 1 — the Directive relates to the conservation of all wild bird wild birds, the Great Cormorant is species. protected under the Birds Directive. Photograph — Shutterstock. Article 2 — Member States take requisite measures to maintain populations at levels that correspond to ecological, scientific and cultural requirements… taking account of economic and recreational of Cormorants at fisheries is to requirements. consider it from the perspective of legislation. This is a useful Article 3 — the general duty of habitat/biotope conservation. approach because the derogation facility within the legislation, Article 4 — specific habitat conservation measures, especially requires consideration of the Special Protection Areas for Annex I (i.e. considered endangered and ‘damage’ that Cormorants can requiring special conservation measures — does not apply to the Great cause and, as such, allows us to Cormorant) and migratory species (e.g. the Great Cormorant), especially think about ‘impact’. Indeed to regarding wetlands. many people the terms ‘serious damage’ and ‘impact’ have become Article 5 — establish a general system of protection, prohibiting in synonymous in their discussions of particular deliberate capture and killing, disturbance, destruction of the Cormorant-fisheries conflicts. nest or the taking of eggs.

At its heart, the general pan- Article 8 — prohibited methods of killing and capture. European scheme of protection for Cormorants (like almost all other Article 9 — derogation scheme of Directive, where no satisfactory wild birds on the continent, and in solution is found and for specific reasons including ‘to prevent serious Great Britain and Ireland) is based damage to fisheries and water’ and ‘for the protection of fauna and on the Birds Directive (79/409/EC). flora’. Key provisions of the Directive in relation to Cormorant conservation Article 10 — encouragement of research. and management are shown in Text Box 10.2 (see also chapter 8 of Text Box 10.2 Key provisions of the EC Birds Directive in relation to Cormorant Marzano & Carss 2012). conservation and management.

[98] the intercafe field manual

Figure 10.3 Fish, and the fisheries they support, usually have very strong economic elements — be they commerical or recreational. Photographs — Shutterstock. biologists — because biological practical difficulties in establishing problem. Thus the quantification understanding has fundamental any effect incontrovertibly. of impact from a biological importance (Carss et al. 2009). Yet Indeed, in the light of seeing large perspective has to be site-specific. within the biological community numbers of birds at their fisheries there has long been debate consuming large numbers of From a purely biological surrounding the question of valuable fish, the onus placed on perspective, there are well- whether fish-eating birds seriously fisheries managers and anglers documented and acknowledged damage fisheries (e.g. Marquiss to demonstrate or prove ‘serious problems of quantifying Cormorant & Carss 1997). Several major damage’ from a biological impacts generally, and ‘serious reviews of the scientific literature perspective before being granted damage’ specifically (see also have concluded that, despite many a derogation to kill birds has Russell 2006). However, the studies, there are few instances been criticised for putting these derogation procedure is also where scientists have demonstrated stakeholders at a disadvantage ‘intended to prevent damage (i.e. long-term, population level (Marquiss & Carss 1997). Case it is not a response to already effects of Cormorant impact on studies in the UK (e.g. Feltham et proven damage)’ and can require fish populations, other than at al. 1999) indicated that Cormorant a ‘reasonable basis for concluding relatively ‘closed’ simple systems predation at some sites may be that damage will be serious in the such as fish farms, ponds, or high enough to cause a decline absence of action’. Thus while within some heavily modified in the fishery, but not at others offering the opportunity to manage running waters. However, this where other factors may be more Cormorants without prior proof lack of biological evidence is significant. This study concluded of damage, the issue of what recognised not to be necessarily that the impact of Cormorants on constitutes ‘serious damage’ is still because birds rarely affect inland fisheries was a problem at unclear, but is expected to involve fisheries, rather it reflects the specific sites rather than a general some subjective evaluation.

www.intercafeproject.net [99] the intercafe field manual

Furthermore, the derogation also ‘infers an economic interest’ and impact is very often thought of in economic terms by those affected by Cormorants. Unfortunately, while the economic issue may appear easier to quantify than a biological impact, relatively few clear data appear to be available. Indeed, the REDCAFE project specifically explored the reasons for the non- disclosure of relevant financial information within the fisheries sector. This may well be, in part at least, because there are no clearly- defined criteria for measuring impact and so there are no comparable economic data. However, the factors affecting fishery economics are also complex and Cormorant- fishery conflicts rarely, if ever, occur in isolation from other factors. Similarly, INTERCAFE had little or no access to financial information that would irrefutably demonstrate that Cormorants were the direct cause of financial losses to a fishery, other than at fish farm sites and small, enclosed recreational fisheries. Again, this does not mean that Cormorants are not capable of causing a financial loss, but merely that the economic data relating to fish losses caused directly by Cormorants have not been collected or collated, or are not readily available, and other possible related losses (e.g. reduction in amenity Figure 10.4 Teasing apart the influence of predation from other sources of fish mortality is not an easy task: predation, pollution, habitat modification. value or licence sales) are equally Photographs — Shutterstock. difficult to quantify.

There is a growing body of stakeholders stated very strongly to biological, economic, social and evidence (see Carss et al. 2009) that the birds undoubtedly had a cultural perspectives in relation that, regardless of the availability serious impact on their fishery, to Europe’s Cormorant-fisheries of biological and/or economic and numerous important social conflicts (see also Marzano & Carss data, the impact of Cormorants and cultural issues have been 2012), it is impossible to argue on fisheries is believed to be great discussed in relation to this (see against the validity of many of these in some situations. Indeed the for example Carss et al. 2009). socially- and culturally-derived INTERCAFE project documented Given INTERCAFE’s generally claims of Cormorant impacts and numerous occasions where fisheries integrative (‘relativist’) approach these claims must be taken seriously.

[100] the intercafe field manual

Clearly, a purely biological inadequate and so consideration of cultural requirements as well as approach is usually inadequate to broader environmental ‘values’ is economic and recreational ones in the task of trying to address the necessary. For instance as habitat some justifiable and fair manner. Cormorant-fisheries conflict in restoration projects continue there This multidisciplinary perspective full. Similarly, although economic may be conflicts of interest over should also be borne in mind when interests are frequently cited in particular species. When these considering Cormorant ‘impact’ relation to the damage and impacts species are scarce there is little as explored primarily from an that birds have on fisheries, clear problem but when they become ecological perspective in the economic data per se are generally abundant (and perhaps prey on present chapter. inadequate too. other species of conservation value) how is conservation (or Although notoriously hard to management) to be prioritised? 10.3 Introduction to the define, terms such as ‘serious ecological perspective damage’ are clearly opaque and Thus, in keeping with the wording there seems a genuine need across of Article 2 of the Birds Directive, As a synthesis, this final chapter Member States to clarify them a more holistic approach to the explores the inherent difficulties as much as possible. The purely issue of quantifying Cormorant in integrating both Cormorant biological or economic approaches impact should almost certainly and fishery data obtained by to determine impact are often include ecological, scientific and researchers. It also aims to discuss

Methodological requirements for data collection and parameters affecting the ecological behaviour and population dynamics of Cormorants and fish. This theoretical sketch diagram illustrates the complexity and breadth of Cormorant- fish interactions.

www.intercafeproject.net [101] the intercafe field manual

Researchers recognise (at least) twelve levels of biological integration, each with its own separate and distinct series of attributes and methodological problems. The study of ecology is concerned with the top five (upper case) levels. As shown on the right, as we ‘progress’ from individual organisms, through each of the ecological levels, our scientific understanding decreases. The Increasing levels of reasons for this are not hard Biological Integration to understand — they include the increasing complexity of these ecological levels and the difficulties involved in level for either Cormorants or Though far from complete, this collecting and analysing rigorous fish species (see, for example, sketch conveys the breadth and quantitative data on them. the methodological requirements scales required of scientific boxes in Figure 10.1). Similar information in order to define and Text Box 10.3 Levels of biological care is needed when interpreting begin to quantify any ecological integration (based on Krebs 2001, p10–12). such datasets, particularly when relationship. Many terms used in Cormorant and fish data are the diagram represent a network of combined to infer something different factors, often interacting some of the main interdependencies about the impact of the predator with each other. For instance, fish between these data sets. on the prey. For example, the abundance can be influenced to When discussing two different diagram on the previous page a great extent by fish mortality. populations and their associated, shows that key biotic and abiotic This mortality is the sum of all inter-related systems, there are factors can affect both Cormorants deaths in the fish population many more environmental factors and fish and suggests that the regardless of their cause, and so than just the Cormorants and ‘system’ has the potential to be includes things like harvest of their prey. These factors may be highly dynamic. This dynamic adult fish by humans, starvation of ‘biotic’ (i.e. biological) or ‘abiotic’ system deals with two very fry as they compete for space in (non-living, usually physical and different organisms that have a habitat, predation, those dying chemical) and fit together in a different demands on habitats, through diseases or parasitism, complex, dynamic way, influencing exhibit different behaviour, and and deaths caused by one or more both the results of sampling efforts have very different distributions abiotic factors. Clearly, it is a and interpretations of these results. in both space and time. The major challenge for researchers to It seems very rare that the cause methodological requirements quantify the relative proportions of of any impact on a fish population and key parameters affecting the each of these causes of mortality, is easily identified and that it is populations of Cormorants and to tease apart the mortality caused caused by a single factor (though a fish highlighted in red reflect both by predators, the proportion of pollution incident in a river or lake similar methodological issues to be this that is caused by Cormorants is an example of this). considered by researchers and also and the ultimate effects of this those environmental factors that portion of overall mortality on the As discussed earlier, great care influence both predators and their fish population (stock) or resulting is needed to collect rigorous prey (but often in very different catches. A key scientific issue ecological data at the population ways). here is whether, in the system

[102] the intercafe field manual

under consideration, the different prey. Closely associated with and to the suitability of particular mortality factors are additive or this are the methodological foraging habitats in relation to fish whether compensatory factors considerations which form the bulk availability (see chapter 8 of van operate. In other words, whether of this Field Manual. However, Eerden et al. 2012). different mortality factors are in order to interpret the resulting cumulative, or whether when one biological data, in an attempt to Ecological behaviour is influenced factor increases another decreases. quantify the impacts of Cormorant by a wide range of parameters. In the latter instance, mortality by predation on fish for instance, For instance, fish stocks are predators may be compensated for we need to consider ecological affected and regulated by abiotic if fewer individuals die later on behaviour, particularly of the prey. and biotic parameters and so the due to starvation or competition, data obtained by researchers from or if the growth rate of surviving It is also important to acknowledge sampling must be considered as individuals increases. the close inter-relationships a ‘snap-shot’ of the population at between Cormorants and their prey. a specific time (and place). Fish The type of the water body also For example, Cormorant behaviour populations may be influenced influences the complexity of the is most likely to be influenced by extreme events such as heavy system. As discussed elsewhere by the ‘availability’ of fish (see flood or drought and these can in the Field Manual, the methods section 7.5 for a discussion on be generally inferred from fish available to researchers to study ‘availability’), which itself is samples (for instance through both Cormorants and fishes heavily influenced by the behaviour sudden changes in mortality rate become less accurate as the size of the fish themselves. At spatial or loss). Unfortunately, the effects of the system under investigation and temporal levels, changes in fish of many less extreme events are increases. Similarly, as these availability are almost certainly the much more difficult to record in, or systems become increasingly reason for Cormorants’ north-south infer from, data from fish samples. complex with increasing winter migrations and also both Furthermore, the influences of such geographical area, so our scientific their flock size whilst foraging at events on fish populations are likely understanding decreases (see a particular site and their choice to be related to the area over which Text Box 10.3). This increased of that site, or ‘willingness’ to they occur. Similarly, the area complexity at the higher switch to an alternative one. from which fish can be sampled (ecological) levels has serious Clearly, fish availability actually (e.g. point, stretch, whole river, implications for researchers as it becomes zero in those waters that catchment), and hence over which means that issues like Cormorant- become ice-covered in winter and the population can be ‘understood’, fisheries interactions, associated this must be one of the strongest is likely to influence our level predator-prey relationships, and drivers for Cormorants’ north- of understanding. Generally, the potential or actual ‘impacts’ of east to south-west seasonal smaller the area sampled, the the birds on the fish are subject to movements (see chapter 6 of more accurate the population decreasing scientific understanding. van Eerden et al. 2012). At the estimate but the less likely it will population level, the availability be to appreciate the influence of of fish, driven ultimately by abiotic and/or biotic factors at the 10.4 Review of some key seasonal changes in temperatures population level. considerations for Cormorant- across northern Europe, will fisheries interactions ultimately influence both timing Interpreting Cormorant numbers and the number of both adult birds is similarly complex. For instance, The spatial and temporal reaching breeding condition and researchers may place different complexity of prey-predator of nestlings that fledge and leave interpretations on day, night roost, systems, including those the nest successfully (see chapter and breeding counts. Furthermore, involving Cormorants and fishes 7 of van Eerden et al. 2012). At the ‘accuracy’ of the count is likely (see also section 10.1), requires the population level, Cormorant to fall as the area covered by the consideration of the ecological distribution in winter appears to be count increases. For example, behaviour of both predators and determined by temperature limits counting the exact number of birds

www.intercafeproject.net [103] the intercafe field manual

using a day roost is likely to be ▪▪ Bird daily food intake — upper considerably easier than counting limit derived from literature birds foraging along a winding ▪▪ Salmon standing crop — not section of river. Thus researchers measured directly. Based on are constantly interpreting the data data from four other rivers they collect within a given ‘frame’ Likely to be underestimated that includes both the space and because of methodological time-period under consideration limitations and does not take as well as the specific research Figure 10.5 Juvenile Atlantic into account the seasonal question for which the data were Salmon Salmo salar and Three- movement of fish within the collected (this is not necessarily the spined Stickleback Gasterosteus catchment same as the ultimate question one aculeatus in a Cormorant’s stomach may want to ask of the available contents. Photo courtesy of D N Carss. Marquiss et al. (1998) stated that information — a question which at certain times and places, the might require different data). fish consumption by the birds as apply over full generation cycles it estimated using these calculations may be very difficult to determine suggested that the birds would their extent. It should further be remove all fish from the river 10.5 The effect of Cormorant noted that if a stock is reduced within two months. That this did predation on fish — some by predation, catches may be not happen implied that some of the significantly impacted even though words on data requirements study’s assumptions were invalid. this may not prevent the fish Either estimates of consumption Where the fish removed by population from sustaining itself at were too high and/or the estimates Cormorants are of an exploitable the lower size. size and species, the losses will of Salmon standing stock were too low. Alternatively, the figures and tend to have a direct effect on the As an example of an attempt to calculations could be realistic but numbers potentially available to a assess fish-eating bird impact on the fish population might somehow fishery. This will apply particularly juvenile Atlantic Salmon Salmo be able to compensate for the where the fish have been stocked salar from the sort of bird and fish recorded levels of predation (see and are not expected to contribute to data obtained by methods discussed Sections 10.7 and 10.8). the natural production of the waters. in this Field Manual, Marquiss et The same will be true where there is al. (1998) compared consumption no population compensation, either rates of birds (Goosander Mergus Clearly this research shows in stocks artificially maintained merganser, Red-breasted Merganser the difficulties in moving from at levels above the carrying M serrator, and Great Cormorant estimates of fish consumption by capacity or, for example, where the Phalacrocorax carbo) with the best predators to their impacts on prey removal of Salmonid smolts by a available scientific estimates of fish populations. The inference from predator will be expected to have a abundance (estimated as numbers of this and other studies seems to be proportional effect on the numbers fish per 100 m2 in relation to river that the fish losses to birds can be of returning adult fish. Where such width). These authors discussed substantial, albeit impossible to losses occur, catches are likely to be a number of potential sources of quantify in population terms using reduced, since larger stocks would error in their calculations. They are commonly collected data. be expected, in general, to provide listed here to highlight difficulties better catches. in estimating impacts even in a Marquiss et al. (1998) also relatively simple system. highlighted that a major problem These effects may be reduced, but in calculating the proportion of are unlikely to be eliminated, if ▪▪ Bird density — data from the fish standing crop consumed by predation occurs at a stage where study river birds was collecting accurate data compensation occurs. However, ▪▪ Bird diet — data from the study for all the necessary parameters at as compensation effects tend to river a specific site. The main priorities

[104] the intercafe field manual

These authors consider that attempts to quantify (B) Knowledge of fish populations and factors that Cormorant impact at fisheries would benefit greatly affect them from information relating to periods both before and after Cormorant presence. They also argue Definition and scale – helping to develop criteria that better understanding between the variability for what data are required within Cormorant and fish populations is needed, ▪▪ A definition of ‘impact’ is essential — impact on alongside recognition of the nature and scale of such populations does not necessarily equal impact on variation. Requirements (many of which are described catches. elsewhere in the Field Manual) are presented for both ▪▪ The level of impact that triggers management Cormorants and fish. action must be clearly defined and stated. ▪▪ One approach could be to define impact in (A) Knowledge of Cormorant diet and biology economic terms [see 10.1] where losses can be attributed specifically to Cormorants — financial Diet loss (impact) would not have occurred in the absence of Cormorants. ▪▪ Confidence in diet assessment methods appropriate ▪▪ Spatially, are impacts considered local (or site- to specific situation. specific) or regional? Such spatial differences will ▪▪ Determination of prey species, size-range, prey affect the likelihood of both an impact and our selection, and associated temporal and spatial ability to quantify it. changes. Simply identifying prey fishes, either ▪▪ Cormorant impacts have to be considered in relation absolutely or relatively, is inadequate. to other factors affecting fish populations and ▪▪ Consumption rates for all prey species, not just catches — studies of how other fish-eating predators those of commercial interest. (including fishes), angler/commercial harvest, ▪▪ Estimates of Daily Food Intake (DFI) — varying pollution, and competition within the fish community with season, energetic demands, calorific content affect fish populations will help understand the of prey. relative magnitude of a Cormorant effect. ▪▪ Quantification of ‘impact’ requires that all the above are combined and may be compared to the Space and time situation without Cormorants. ▪▪ Knowledge is required on the abundance of fishes Biology and their distribution — including their ranges and movements — both spatially and temporally. ▪ Cormorant demography — ratio of breeders to ▪ ▪▪ Population indices for fishes are also required — non breeders, sex and age ratio in relation to their ideally from regular species-specific fish surveys. associated diet and DFI. ▪▪ The most useful data sets will be in the form of ▪▪ Seasonal foraging range — helps to understand long-term trends. local versus regional ‘impacts’ to fish populations. ▪▪ The timing, magnitude, and frequency of Population dynamics and habitat characteristics Cormorant foraging activities, in relation to the foraging site under consideration. ▪▪ Essential information includes information on ▪▪ Seasonal changes in these parameters. dynamics of fish populations (not necessarily just those of commercial interest), including age-structure, mortality (including changes in the relative importance of different sources of it), survival, growth rates, and recruitment. ▪▪ Impact studies require a better understanding of Text Box 10.4 Requirements for quantifying compensatory effects on natural fish mortality. Cormorant impact to fisheries (based on Wireset al. ▪▪ The most useful datasets will be in the form of 2003: 391–2). long-term trends.

www.intercafeproject.net [105] the intercafe field manual

identified to allow more accurate calculations were to produce matched datasets (birds and fish) for specific foraging sites that provided:

▪▪ More accurate estimates of daily food intake of birds ▪▪ Knowledge about prey switching by birds and the role of alternative habitat ▪▪ Estimates of bird densities at foraging sites where there are dietary data ▪▪ Direct estimates of fish standing crop (abundance) at bird foraging sites and of fish Figure 10.6 How many Cormorants? Do large numbers of fish support large movement patterns numbers of birds? Are fish affected in relation to Cormorant numbers? Photo courtesy of Janis Kuze. In this context, an in-depth review of the types of high quality data (including those mentioned with the methodological difficulties are ecologically very complex. above) which are ideally required discussed elsewhere in the Field The next section discusses for scientifically quantifying Manual. These problems are often this complexity, particularly in Cormorant impact at fisheries is further compounded by the common relation to the careful and accurate summarised in Text Box 10.4. mismatch (in time and/or space) interpretation of data. Many of these issues are then between the available Cormorant discussed in more detail in later and fish data. Generally, assessment sections. becomes increasingly difficult 10.6 Predator-prey as one moves from small highly relationships — some concepts Whilst the information in Text managed water bodies to larger Box 10.4 is fine as a research wish more natural ones (see also sections This brief review of some of the list, it is unrealistic as a set of 7.5, 7.6 and chapter 8). Moreover, key ideas and concepts associated recommendations for what might there is a general consensus that the with predator-prey relationships be accomplished in practice. There relevant bird data are more readily is drawn from general texts (e.g. is thus a need to recognise the available, and perhaps more robust, Cherrett 1989, Crawley 1992, need for a pragmatic approach to than many of the available datasets Begon et al. 1996) as well as quantifying Cormorant predation on fish, their population dynamics specific works on fishes (Pitcher at fisheries (see also sections and status and distribution. Ideally, 1986, Wooton 1990). It follows 10.1, 10.2 and 10.13). Indeed the fish data should include some very closely the main points fisheries themselves are often knowledge of the community discussed by Marquiss et al. (1998, managed very successfully on far structure and productivity of pp 62-64). less information than that listed indigenous stocks and the effects of here. possible compensatory mechanisms, In practice, a number of density and also of the effects of any dependent and density independent It is clear that obtaining relevant stocked fish. processes can act on a population data that allow us to quantify at any one time (see section populations of either Cormorants or Finally, far from being a simple 10.7). Their relative importance of fish species is difficult and many predator-prey relationship, those usually varies in space and time, of these problems are associated between Cormorants and fish thus altering the position of the

[106] the intercafe field manual

equilibrium density within and researchers to predict that several are foraging in modified habitats between populations. Furthermore, factors besides the densities of and/or sites where only one (or these processes are likely to affect predators and prey are important. a few) fish species occur or are populations differently depending ‘Specialist’ predators consume being raised (e.g. fish farms). on when in the life-cycle they occur. a single, or small number of, Thus understanding fish stocks, in To assess any effects of predation prey types (or species) whilst relation to their associated habitats on prey populations, it is vital to ‘generalist’ predators feed on a or fishery types is important. know how predators respond to wide variety of prey. In theory, the changes in the density of their prey. effects of specialist and generalist There are essentially two types predators on their prey are likely 10.7 Using fish stock of response. First, a numerical to be very different. Generalists assessments as a basis for response whereby higher numbers are likely to have a bigger impact assessing Cormorant impact of prey are likely to support higher on their ‘preferred’ prey through numbers of predators. Second, a their ability to switch to other Commonly, the results obtained functional response whereby the prey types (species). Studies of from fish stock assessment number of prey eaten by each predation should not therefore techniques (see chapter 8) do not predator is affected by variation in merely consider the ‘preferred’ provide an estimate of absolute the density of that prey. prey in isolation but also consider abundance of the fish stocks present other accessible prey and the in a water body, but they may give Such functional responses are likely proximity of alternative foraging an ‘index’ of stock abundance. This to be affected by a number of factors. sites. In the case of Cormorants (a can nevertheless still be used to For example, not all prey are equally generalist predator), this is likely identify trends in fish abundance and vulnerable to predation. Predators to mean that most, if not all, of the alert fisheries’ managers to changes may select old, young, or sick prey fish assemblage(s) in habitats used and potential problems. There are that are easier to catch. Where this by the birds for foraging might a number of inherent difficulties is the case, predation is less likely be considered and quantified to associated with sampling freshwater to affect the overall population some degree. In some situations fish stocks that need to be borne in dynamics of the prey. Conversely, this is actually less of a problem mind in relation to the integration predators may select highly in practice because Cormorants of fish stock assessment data and conspicuous prey that are feeding actively or breeding and which (if not eaten) would have made a large reproductive contribution to the population. Similarly, not all individual prey will have equal access to refuges from predator attack — and in many cases those with the best access to cover might be the ‘fittest’ individuals. Thus, in some cases, those prey making the biggest reproductive contribution could be least vulnerable to predation. Conversely, if a prey species relies on being cryptic to avoid predation, prominent territory holders may be most vulnerable.

Considering predator-prey interactions within the functional Figure 10.7 Fish catches from a Greek coastal fishery — a very large ‘open’ response framework allows system. Photo courtesy of D N Carss.

www.intercafeproject.net [107] the intercafe field manual

parallel Cormorant data when lakes and the marine environment attempting to quantify the impact of (i.e. spatially ‘open’ systems often the birds at a specific fishery. with reduced management and/ or fish stocking) is much more Information on the distribution complex. Detailed long-term of fish species and the size of monitoring of such sites, in the fish populations is only one of absence of pressures such as the requirements for trying to Cormorants, may allow the natural understand the interaction between variability in fish populations Cormorants and fisheries. This (standing crop, recruitment, size is because predator-prey ecology structure of the population, cohort involves the complex interaction mortality) to be described and to Figure 10.8 Stocking waters with between two dynamic systems, and serve as a basis for comparison fish is a common practice in many the various factors affecting them. (e.g. in terms of fish density, freshwater fisheries but it can often While predation at an individual length-frequency distribution, etc.) have large effects on the ecology level (e.g. a fish consumed by a with similar water bodies affected of the system. Photo courtesy of Cormorant) requires simply that by Cormorants. Information on INTERCAFE. a foraging predator and a suitable the number of Cormorants present prey are at the same place at the (e.g. Cormorant-days in a certain same time, this does not explain the period or area, see section 3.4.1) population regulation mechanisms possible implications at the long- and their diet and daily food affecting freshwater fish stocks and term, population level, nor does it intake (see chapter 4) will also be their possible effects on production provide any measure of whether important in such studies. However, and catches. ‘serious damage’ has resulted (from the costs and practicalities of such a biological perspective, but there monitoring mean that such rigorous are others, see Text Box 10.2 and data sets on both birds and fishes 10.8 Natural regulation in sections 10.1, 10.2 and 10.13). will inevitably be confined to fish populations relatively few sites. In relatively simple, man-made The species composition of structures like fish ponds or fish There are certainly some cases fisheries, and the size of fish stocks farms (i.e. spatially ‘closed’ where such data on fish and within them, can vary enormously, systems), where controlled Cormorants are available that allow under the influence of a wide range conditions apply, estimating the us to consider the estimated numbers of factors. For many species, the mortality of a fish stock due to of fish eaten by the birds in relation population size is determined as predators, and hence the level to some measure of the standing a balance between gains (births, of impact, can be relatively stock of the prey. Nevertheless, they immigration) and losses (death, straightforward. For example, do not necessarily allow us to assess emigration). If the rates of either yearly losses in the absence of the impacts of predation on fish gains or losses (as a proportion of Cormorants could provide a basis populations because the predator- the population) are unrelated to for comparison with sites visited prey interactions are complicated population density, processes are by the birds. Although other (Russell et al. 1996, see also said to be ‘density independent’. explanations for the losses (see sections 10.10 and 10.11). Natural If, as population density increases, section 10.12) would need to be variation in fish populations and the the proportional rate of gain ruled out before a Cormorant effect effect of this on both fish production decreases (or of loss increases), could be established unequivocally. and catches makes it additionally then processes are considered to difficult to determine the effect of be ‘density dependent’. Generally, Inevitably, understanding predator- a single factor, such as predation, in natural fisheries, stock levels prey interactions and assessing the on the numbers of fish present. The are regulated by various factors status of fish populations in more following sections provide some and they serve to provide an upper natural systems like running waters, background information on the limit or ‘carrying capacity’ for

[108] the intercafe field manual

the fish stock which involves both environment ceases to impose will be influenced by the timing density independent and dependent a limit on the carrying capacity. and duration of any impact from factors In effect, carrying capacity For example, in Atlantic Salmon predators, other biotic and abiotic represents the stock size that, Salmo salar and Sea Trout limiting factors and by year-to-year on average, the habitat is able to (migratory Brown Trout S. Trutta) changes in the productivity of the support. Where gains equal losses, there is thought to be virtually no stock. Moreover, compensation also the population is at equilibrium compensation after the fish reach only works within certain limits. but if there is a switch (in either the smolt stage and emigrate to For example, many fish populations direction) only density dependent sea. Furthermore, for all fish stocks are subject to substantial variation processes can act to re-establish maintained at high levels (i.e. above as a result of fluctuations in the equilibrium in a process known as carrying capacity) through stocking, year-class strength of different ‘regulation’. However, in managed there will be no natural compensation fish species, with ambient water or intensive fisheries, stock for fish losses to predation. temperatures in the first year of life densities that significantly exceed being a key determining factor. The the natural carrying capacity can The compensatory processes also sustainability of such fish stocks be achieved through additional mean that natural mortality will often depends upon occasional stocking and other management be reduced when stocks fall below large year-classes, which remain practices. This is typically confined their carrying capacity due to identifiable as the fish grow over to smaller, spatially closed sites. reduced competition for resources. a period of years. In addition, As a result, the productivity of the in larger rivers and lakes, fish In healthy natural populations, population (e.g. fish growth and populations will typically be larger the number of fish produced will number of surviving offspring per in overall terms but occur at lower generally exceed the numbers which adult) will increase as the stock densities and be more spatially can be sustained by the available size falls. This means that if stocks separated than those in smaller resources. As a result, natural are depleted by increased levels of habitats, and this may confer mortality will tend to result in the predation (or human exploitation), greater resilience against predation population declining towards the they may still be self-sustaining, over the longer term. carrying capacity level. Under these albeit at a reduced average level. circumstances, some or all of the Thus fish populations may be There are, of course, many other fish lost to predation may simply be self-sustaining at a wide range factors (such as human exploitation, replaced, in the long term at least, of population sizes below their disease, other bird and mammal by fish that would otherwise have maximum carrying capacity. Only predators, and piscivorous fish) died from other causes, a process if the rate of removals (by whatever that also impact on fish population known as ‘compensatory mortality’. cause[s]) exceeds the maximum size and structure. However, the This mechanism depends on there productive capacity of the stock assumption that catches of fish being continual competition for will the population collapse. A will be larger where there are available resources, such as space clear implication of this is that larger stocks of fish in a water or food. It therefore tends to have where populations are reduced, body (i.e. that catches and stocks stronger effects on the relatively even if they are self-sustaining, are positively related) is widely large numbers of very young fish catches will tend to be depressed accepted by anglers and fishery in a population and less on the and impacts may still be occurring, managers and supported by a smaller numbers of large adults. It albeit difficult to quantify. substantial body of scientific should be noted that the potential evidence. Consequently, fisheries for compensatory mortality is managers commonly resort to greatly reduced in stocks below 10.9 The effect of natural stocking to increase fish densities. carrying capacity (the situation variation on production and However, the introduction of for the majority of Salmon stocks catches stocked fish may make the in different parts of Europe for quantification of Cormorant impact example). Compensatory effects The scope for compensation to even more complex because can also be affected where the operate within a fish population the population dynamics of the

www.intercafeproject.net [109] the intercafe field manual

prey — perhaps in this case, stocked fish — are not constrained by natural gains and losses (see below). Moreover, fish introduced through stocking may be more vulnerable to predation than those occurring naturally, particularly during the first days of release. Cormorant predator-prey relationships are clearly complex and so can be very difficult to understand in terms of biological impact, an issue discussed in the following section.

10.10 Why do biologists find it difficult to measure a Cormorant ‘impact’ at Figure 10.9 Cormorants can sometimes compete directly with fishermen for fisheries? the same species and sizes of fish.Photograph taken from Fishing Club Bohinj presentation to INTERCAFE. As discussed earlier (section 9.1), the word ‘fishery’ has different meanings to different concerns, as scientists have positive relationship between catch people, similarly there are many very seldom been able to collect and stock size. different ‘frames’ through which irrefutable evidence of Cormorant Cormorant impact at fisheries impact, particularly at larger fishery There may also be a difference may be interpreted. The biological sites. This section, deliberately between the catches made by perspective is only one of several, focussing on the biological (natural humans and by other predators often inter-woven, perspectives science) perspective, further such as Cormorants or predatory that can be applied to the issue explores why this might be so. fish, and this is more likely of Cormorant impact on fisheries in ‘well-balanced’ aquatic (see also sections 10.1, 10.2 and Usually, ‘Cormorant impact’ is ecosystems or those under more 10.13 and chapter 12 of Marzano taken to mean a situation where natural conditions, where the & Carss 2012). INTERCAFE thus birds are eating sufficient fish in a level of competition for the same recognises (see also Marzano & system so as to affect it negatively individuals of the same species Carss 2012) the management needs from a human perspective. In and sizes can sometimes — though of many involved in commercial practice, this almost always means a certainly not always — be and recreational fisheries where reduction to fish catches, fish value, relatively small between humans the evidence base has to weighed or to a specific portion of that catch and Cormorants. There is, however, against many factors other than (e.g. a particular species or age/ even greater potential for direct natural science, including social, length class of individuals). The competition in some highly cultural, economics, and legislative relationship between a fish stock managed fisheries, for example ones. Nevertheless, biologists have size and the catch can be affected by at intensive fish farms, or angling long been charged with addressing a range of factors. Typically, though lakes stocked regularly where fish and quantifying impacts (see not always, larger stocks will result community structures may be quite reviews in Marquiss & Carss 1994, in better catches (this is, after all, simple or even monocultures. Russell et al. 1996, and also Carss why fish are stocked at some sites) et al. 2009). This has invariably and there are published accounts Besides direct losses, there is also led to frustration amongst fisheries for many fish species indicating a a range of ‘indirect’ Cormorant

[110] the intercafe field manual

impacts that may need to be taken of Cormorants on fisheries. inferences just mentioned are too into account. These include: In relation to the ecological simplistic to be realistic in nature. processes crucial to predator-prey ▪▪ the physical damage from the interactions, ecologists considering It is beyond the scope of the Field birds to those fish that they are natural enemies and their prey Manual to go into each of these unable to swallow or which need to address several interrelated questions in detail but, nevertheless, escape having being captured; questions (Crawley 1992): it is clear that a huge amount of ▪▪ the risks of birds spreading scientific data would certainly diseases and/or parasites ▪▪ How do predators affect the be required to answer each of between fish stocks; abundance of their prey? these questions categorically for ▪▪ the predation on smaller ▪▪ Can predators regulate prey any specific Cormorant predator- individual fish which, had they population density? prey situation. Focussing on not been eaten and survived ▪▪ What determines the abundance the fourth question above, in to grow larger, would have of predators? relation to Cormorants and been legitimate catches for ▪▪ What factors influence the fisheries for example, researchers commercial fishermen, quarry pattern of dynamics exhibited need to consider a number of for recreational anglers, or part by a particular predator-prey important issues. Moreover, they of the harvest at fish farms; relationship? can not address these questions ▪▪ predation on smaller fish species simultaneously, by experiment or in of no commercial value that are These appear to be simple theory, as these relationships turn prey for larger fishes that are questions — if a predator kills out to be extremely complex and themselves the target for human prey, the prey must become less require, in theory at least, a large fisheries; and abundant and, if prey numbers amount of necessary information ▪▪ the ‘stress’ caused by foraging decline, predators must switch to (see Text Box 10.5). birds to fish, which causes another prey, or foraging site and/or them to move into different become less abundant themselves. In reality, the requirements detailed habitats, change their feeding However, the consensus amongst in Text Box 10.5 are essentially a habits, lose condition, become researchers is that these are far wish list and is wholly unrealistic more susceptible to diseases or from trivial questions and the in the context of most Cormorant- parasites, grow more slowly, and generally be less ‘available’ (and valuable) to human fisheries.

In most cases other than in relation to physical damage to fish (see chapter 9) and parasite infections (Cormorants and other fish-eating birds are known to be the definitive hosts for a number of parasites which also affect fish), definitive biological data may be difficult to collect. Clearly further research is necessary to further understand and quantify these issues.

Leaving the indirect biologically- based impacts aside, the remainder of this section focuses on the Figure 10.10 Sub-Alpine river in Slovenia — when lowland standing scientific requirements for a waters have ice-cover in winter, such running waters can be ‘hotspots’ for quantification of the direct impacts Cormorants. Photograph — Shutterstock.

www.intercafeproject.net [111] the intercafe field manual

1. The intrinsic rate of increase of the prey — the where the rate of growth of both predator and rate at which the prey fish population increases prey is zero. In this manner it is possible to relate in size — the change in fish population size per changes (increases or decreases — defined by the individual fish per unit time. slope of the isocline) in the Cormorant to those in the fish andvice versa. It suggests strongly at what 2. The functional response of the predator — the densities Cormorants might be limited by the fish relationship between the Cormorant’s consumption (or by other things) and when the fish might be rate of fish and the density of those fish. limited by Cormorants (or by other things).

3. The predator’s spatial foraging behaviour — the 7. The size of the prey refuge — this can vary foraging behaviour of Cormorants in relation to the considerably and encompasses every situation space they are currently using as a foraging range. where fish are ‘free’ of Cormorant predation. Prey refuges could thus include a hiding place 4. The nature of prey density dependence — the for an individual fish, the avoidance behaviour way in which the death rate in the fish population or crypsis of fish, habitats that are inaccessible to increases, or the birth or growth rate of it, Cormorants, habitats that do not overlap with the decreases as the density of the fish population geographical range of Cormorants, the ‘choice’ increases. Alternatively, — the way in which the of the predator to select an alternate prey, and death rate in the fish population decreases, or the the presence of fish in an area of low Cormorant birth or growth rate of it, increases as the density density. of the fish population decreases. 8. The rate of predator immigration — the movements 5. The nature of predator density dependence — the of Cormorants into the population from elsewhere. way in which the death rate in the Cormorant population increases, or the birth or growth rate 9. The rate of between-patch prey dispersal — fish are of it, decreases as the density of the Cormorant not uniformly distributed across habitats or within population increases. Alternatively, — the them but tend to be concentrated in ‘patches’ (e.g. way in which the death rate in the Cormorant shoals of fish in a lake, higher densities of fish in population decreases, or the birth or growth rate particular stretches of river — or at particular times of it, increases as the density of the Cormorant of year). The dispersal rate of interest here is that at population decreases. which individual fish spread away from each other between these patches. 6. The relative slopes of the isoclines at the intersection — isoclines are lines drawn on a It is important to note that the above refers to the two-dimensional plot of predator density against simplest case, that of a ‘specialist’ predator — one prey density. They link points that give rise to the taking a single (or few) prey type(s). As ‘generalist’ same rates of population increase for the species predators, Cormorants consume many different types being considered. Thus they divide the area of of prey fishes and so trying to explain the dynamics the graph into a zone of prey increase and a zone of this more complex predator-prey interaction could of prey decrease. Similarly, the predator isocline ‘exhibit chaotic dynamics of almost unimaginable divides the graph into an area of predator increase intricacy’ (Crawley 1992, p89). and one of predator decrease. Isoclines intersect

Text Box 10.5 The minimum amount of necessary information to understand the dynamics of one specialist predator and its prey (based on Crawley 1992, p89). Clearly, this list is too simplistic for Cormorants foraging anywhere but in a single-species monoculture fish pond (see text).

[112] the intercafe field manual

fishery conflicts. Moreover, the large proportion of the fish available complex ecological processes list is too simplistic for all but the in a relatively small local area for so-called ‘trophic cascades’ simplest predator-prey relationship relatively quickly after ‘discovering’ in lakes. Their work focused on where Cormorants may be foraging the site. It seems that such localised predator-prey relationships in these solely on a monoculture of fish. predation ‘hotspots’, where fish habitats but had fish as their top This highlights a major problem appear particularly vulnerable to predators. Here, variability at the facing researchers — that even predation and where losses to them top of the food web (acting through complete understanding the may be considerable, are the main the selective predation by the fish dynamics of the simplest predator- focus of the greatest concerns of community) was hypothesised to prey relationship is likely to fisheries managers, fishermen and ‘cascade’ through the zooplankton be unrealistic — and leads to a anglers. However, although this is a and phytoplankton to influence recognition that researchers will widely held view of the sequence of ecosystem processes (e.g. primary never be able to provide all the events, the ‘hard’ scientific evidence production). These researchers answers. This again highlights that may not be particularly robust for set out to understand this process, the need for a pragmatic approach the reasons already cited. beginning with an exploration of is needed both in assessing how best to test their ideas in the Cormorant impact at fisheries and In conclusion, each of the field. Thus, although this work in managing Cormorant-fisheries components of a specific predator- will not help us directly with our conflicts (see section 10.13). prey relationship is based on a question of whether Cormorants simplified or abstracted description impact upon their fish prey (it Scale is also a vitally important of the real world. Intuitively, people does not include fish-eating birds consideration in predator-prey may think they know the answers as top predators), it may suggest relationships. For example at the to some (if not all of them) and, in how best to look for, and measure smallest spatial scale, every death some cases, there are scientific data in biological terms, Cormorant caused by a predator could be to help understand at least some of impacts in real world situations. considered as a local extinction these components. However, when (Crawley 1992). At a large spatial researchers attempt to put them Kitchell and Carpenter (1993) scale, the interaction might be together in order to understand the considered five essentially stable - there may be little overall dynamics of a specific predator-prey different ways in which effect of predation because of local relationship in nature, immensely researchers can test their ideas instabilities or as a result of local complex systems are formed. Given about ecosystem processes. regulating mechanisms. Similarly, this biological complexity and the These are comparative studies, temporal scale is also important recognition that scientific evidence long-term studies, simulation as there may well be potential for the numerous permutations of analyses (modelling), mesocosms for short-term impacts (on fish Cormorant-fisheries interactions (laboratory or field enclosure catches or income, say) but perhaps is likely to be always incomplete, experiments), and experimentation compensation and/or recovery (at where can researchers and other at the ecosystem scale. Whilst the population level) in the fishery stakeholders turn to better understand each has its benefits and associated over longer time scales. the potential impact of the birds on limitations, they considered fish from a scientific perspective? experimentation at the ecosystem For species that live in fragmented scale to be most appropriate, landscapes as ‘sub-populations’ complementing this where separated in space, local extinctions 10.11 Manipulative appropriate with other methods. of prey may occur following the There is a large body of literature discovery of a patch by predators. field experiments and on the problems and pitfalls of In theory, this may be the case at understanding complex experimental design in ecology some predation ‘hotspots’ (e.g. Carp ecological processes (see for example Krebs 1989, Cyprinus carpio aquaculture ponds, Raffaelli and Hawkins 1999) and stretches of sub-Alpine rivers) Kitchell and Carpenter (1993) whilst it is not within the scope where Cormorants may consume a explored how best to understand of this Field Manual to discuss all

www.intercafeproject.net [113] the intercafe field manual

Given the time required and the geographic area to be covered, it is perhaps not surprising that relatively few experiments have attempted to assess the effect of fish-eating birds on fish stocks at a catchment scale. Moreover, those experiments that have been conducted have involved sawbill (Mergus spp.) ducks not Cormorants. Experiments have also focussed on reducing bird numbers, thus reducing their predation pressure on the fish, and have mostly been conducted in Canada in an attempt to increase salmon harvest. These experiments have been extensively reviewed (Marquiss & Carss 1994, Russell et al. 1996, Marquiss et al. 1998) and none of them unequivocally showed that killing sawbill ducks (reducing the numbers of predators) increased fish (salmon) harvest, mainly Figure 10.11 Do fish populations (stocks, or catches) change as a result of because of poor experimental Cormorant presence? Photographs — Shutterstock. design or implementation.

differences will be more convincing of fish-eating birds and fisheries, Text Box 10.6 Experiments to reduce such manipulative experiments the impact of fish-eating birds on fish and could be statistically significant stocks. if they are shown consistently have usually involved the large- in more than one treatment area scale removal of the predators compared with more than one (or a significant reduction in their these, some consideration of field control area (each of these pairs numbers) with the aim of measuring experimentation is necessary here. being a so-called ‘replicate’). This a subsequent ‘improvement’ in replication is important because some measure of the fishery. Scientifically, an ‘experiment’ otherwise it could be argued that the However, the lack of replication is a carefully-designed process differences are nothing to do with often invalidates such ecological undertaken by researchers the experimental treatment but are field experiments (see Text Box to test a specific hypothesis due to other factors. Thus, if done 10.6) and so it is not possible to about nature — for example, a correctly, statistical analysis can relate any measured effect (if, ‘treatment’ is applied to one area then resolve a particular cause and indeed, there is one) to the specific (or ‘experimental unit’) and not to its specific effect(s). act of reducing bird numbers. another. Any differences between the treatment and the so-called However, as noted above (section It therefore seems that unequivocal ‘control’ area can then be compared. 10.10), real-life Cormorant-fishery scientific evidence of Cormorant For the results of an experiment interactions are generally so (or other fish-eating bird) impact to be convincing, there must be complex, and scales so large, that on fisheries is extremely difficult to clear-cut differences between the standard scientific experimentation obtain. This is true whether impact treatment and the control. These is not possible. Thus, in the case assessments are direct comparisons

[114] the intercafe field manual

These factors need to be considered carefully as possible explanations, before claims that a reduction in fish abundance has been caused solely by Cormorant predation.

▪▪ Past events and effects: sudden severe impacts such as toxic agents, waste water, extreme flooding. ▪▪ Climate: long-term effects. of bird and fish data or whether ▪▪ Chemistry: water parameters such as ammonia, nitrite, nutrient they are indirect, through the real levels, pH, ‘hardness’. world experimental manipulation ▪▪ Water-physics: water parameters like temperature, oxygen, of predator numbers. However, this conductivity. lack of evidence does not mean that ▪▪ Hydrology: amount of discharge, flow regimes and water levels. birds like Cormorants do not have ▪▪ Morphology: river bed, river bank characteristics, provision of an impact on fish stocks — it merely cover. emphasises that it is very difficult to ▪▪ Food: abundance and availability of food. demonstrate this scientifically. ▪▪ Diseases: parasites, bacteria, viruses. ▪▪ Growth: leading to an increase in biomass. ▪▪ Fisheries management: human interference to wild fish stock, 10.12 How best to make including exploitation and stocking, also the presence of alien progress? Data requirements species. ▪▪ Predation by other species of fish-eating birds, mammals (e.g. and methodology American Mink Mustela vison, Otter Lutra lutra, or seals Phoca spp.), and/or predatory fishes such as PikeEsox lucius, Perch For the researcher, perhaps the Perca fluviatilis and Pikeperch Sander leucoperca in freshwater most relevant development in and Mackerel Scomber scombrus, Cod Gadus morhua and Bass the consideration of Cormorant- Dicentrarchus labrax in marine ecosystems. fisheries relationships is the ▪▪ Density dependent mortality and natural population variability: application of statistical techniques, mortality changes due to changing fish densities inducing novel to ecosystem ecology, that fluctuations in measured fish density/biomass between years. are appropriate for real world experiments (Carpenter 1993). Like Text Box 10.7 A checklist of key factors that could significantly influence fish the Cormorant-fisheries situation, abundance.

www.intercafeproject.net [115] the intercafe field manual

Figure 10.12 Geographic scale is a key ‘frame’ through which to view the likelihood of Cormorant predation (and of quantifying it) at a fishery; small stream flowing into a river main stem; large lake. Photographs — Shutterstock.

reliable statistical inference we need spatial controls for all ecological experiments.’ Nevertheless, careful data collection and interpretation, and a good understanding of the dynamics of the fishery under investigation in terms of natural Carpenter (1993) argues that to the replicated manipulation regulation (see section 10.8) and interpreting what happens in his discussed above is to make before- the likely effects (and scale) of study lakes during experimentation after comparisons, where the other factors (besides predation) can be reduced to two simple experimental unit can act as its on the fish population (see Figure questions: Did the system change? own control. Fisheries and wildlife 10.1, also text below), the ‘before’ If so, did the manipulation cause management often rely on time- situation might be able to act as a the change? Analogous questions in series data in which a management legitimate control. the context of Cormorant-fisheries manipulation occurs at a given time. studies would be: However, the problem here is that To answer the second question without an adequate control, all above, one must show that the ▪▪ Did fish populations (stocks or before-after comparisons assume manipulation (or the presence of catches) change? that natural systems stay essentially Cormorants) was the most plausible ▪▪ If so, did the presence of the same over time and as Krebs reason for the change. This can Cormorants cause the change? (1989: 270–71) points out this is ‘a either involve looking very carefully dubious balance-of-nature model at the before-after situation in Carpenter (1993) states that, that has been found invalid time the same system (see above) or without replication, it is only and time again in ecological work. looking at similar systems (with possible to answer the first question Populations and communities and without Cormorants), well- through experimentation. An change over time in a way we only argued ecological interpretation and alternative experimental approach dimly understand, and to achieve an acceptable level of significance

[116] the intercafe field manual

assign any effect specifically to Figure 10.13 Double-crested Cormorant predation as there are a Cormorant Phalacrocorax auritus. Photograph — Shutterstock. number of factors that could also be affecting the fish population of interest (see Text Box 10.7). Each one could have a significant, measurable, effect on the results of fish sampling, potentially as big (or bigger) than the likely effect of predators such as Cormorants.

Typically, the same general suite of factors can affect a variety of water bodies and wetland habitats frequented by Cormorants and where fish samples may be collected. These include ponds, rivers, lakes, and coastal areas and it is not uncommon for researchers between any recorded changes in and/or after Cormorants will make to obtain fish data from any of these fish in waters with and without it difficult to detect any effect habitats in the form of biomass Cormorants. Careful use of time amongst the natural variation in the records for fish species. Regardless series statistical techniques (looking data. The worse case scenario (and of the form of the data (i.e. long- for non-random changes in the one commonly facing researchers) term time trends or before-after system under study, see Carpenter is thus that there are few (or only comparisons), the key issue is 1993) could thus help us search for single) data points either before often one of interpretation — is and quantify a Cormorant ‘impact’ and/or after Cormorants. In such Cormorant predation the most at a fishery. It is also important to situations, it is very difficult to likely cause of any observed remember that losses to predators such as Cormorants (which can be estimated with reasonable accuracy, see chapter 3) should be viewed in the context of the other environmental and anthropogenic factors which might affect fish stock (and ultimately catch) size (Russell et al. 1996, and also see Text Box 10.7).

One further important issue to consider here is the amount of data available to researchers. Clearly the longer the run of data both before and after Cormorants, the more likely it will be to detect any differences between the two situations and begin to explore whether it is most likely to have been caused by the birds. A small Figure 10.14 Typical shallow lake Alewife Alosa pseudoharengus spawning number of data points either before habitat in Connecticut, USA. Photograph — Shutterstock.

www.intercafeproject.net [117] the intercafe field manual

changes in fish biomass, numbers scale, interpretation of the results or some index of these? becomes more complex. Smaller waters and their associated fish However, whilst this holistic communities often react faster approach (Text Box 10.7) is useful to environmental factors while to assess the various factors likely to larger ones may be buffered to have influenced fish abundance from some extent. Similarly, Cormorant a purely biological perspective, it predation is more likely to have an is probably unrealistic to think of impact on a fishery or particular being able to eliminate some of fish species in a smaller water body them (e.g. climate effects) as likely than in a larger one. causes of change. Similarly, for many Cormorant-fisheries conflict The increasing complexity of situations, assessing these factors sampling Cormorants and fish will be beyond the scope of many rigorously in larger water systems, stakeholders, including researchers. the need for more intensive There is another danger in this research effort, accompanied approach in that it could be seen as by a general reduction in the requiring an unreasonable burden of accuracy of the results, could proof that Cormorants are the sole lead to questioning whether cause of fish declines. However, scientific understanding of real-life Figure 10.15 Double-crested on the other hand, an awareness of Cormorant-fisheries interactions Cormorant after a successful foraging it may guard against the intuitive is possible at all. Such a question bout. Photograph — Shutterstock. reaction that if fish abundance has should not be seen as a means of declined then Cormorants must be deflecting the issue of predation solely to blame. Commonly such into ever more complex ecological results comparable with other assessments are more likely to boil processes, but a recognition that investigations through the down to a number of factors and measuring predation, in a scientific standardisation of methods. hard practical choices. If fish are in context, is also extremely difficult. Furthermore, there are strong decline and management is needed, Nevertheless, there are numerous pointers as to how best develop how do fisheries owners and claims of Cormorant impact study designs that will give managers prioritse addressing the at fisheries and factors besides maximum insight into the various limiting factors, taking into ‘scientific proof’ could quite biological reality of the system, account such things as the likely legitimately be used to define such through the collection of rigorous scale of any effect, practicalities, impact (see also sections 10.1, 10.2 field data. In addition, a broader costs, and short- versus long-term and 10.13). Ideally, with current understanding of the diverse considerations? political demands for evidence- factors affecting both Cormorants based policies, researchers should and fish can only help researchers This chapter has discussed the try to quantify the effect(s) of tease apart any Cormorant impact general picture of a theoretical Cormorants on fisheries of interest on fisheries, but will also be of Cormorant-fish ‘system’, showing to humans, but this will inevitably value for day-to-day management that both the complexity and the be constrained by practicalities practices. ecological knowledge required to and there will be a need for a understand it, increases with the level of pragmatism in assessing Whilst most available estimates size of the system. It also discusses Cormorant-fishery conflicts. of the impact of bird predation how the accuracy of available on fish catches have relied on scientific sampling methods As this Field Manual demonstrates, either experiments or theoretical decreases with the increasing size there are robust methods with modelling, the findings discussed of the study system or water body. which researchers could address here suggest that when used in Similarly, with this increase in many issues and make their isolation, neither approach is likely

[118] the intercafe field manual

effect of Cormorant predation, numbers and distribution mirror amongst other factors, on fish. both the environmental status of Indeed such an approach has been aquatic environments and their adopted by researchers in the USA condition. The decisive factors investigating the impact of Double- for Cormorants on a site-by-site crested Cormorant (P. auritus) basis are ultimately the availability predation on fish including the of, and accessibility to, water and Alewife (Alosa pseudoharengus) in food resources, while climate and a shallow lake and associated water prevailing weather conditions are bodies. Counts of Cormorants the most relevant driving forces for and assessments of their diet were large-scale movements, distribution combined with a bioenergetics and length of stay. It is within this model of their food consumption dynamic framework that attempts and with estimates of fish numbers to measure Cormorant impact at and mortality. In this case, the specific fisheries are made. authors concluded that although the Cormorants were important Importantly, it is also clear that in predators of spawning Alewifes many places Cormorant predation (consuming 30% of the spawning (and any resulting impact) on stock in 2005 and 18% in 2006 and fisheries is occurring against a representing 48% of the overall backdrop of increasingly rapid mortality of spawning fish in 2005), environmental (and social, cultural to prove entirely satisfactory (see they did not have notable impact and economic) changes (not always also Marquiss et al. 1998: 80). on fish mortality or population negative) across many European Clearly, large-scale catchment size, concluding that under current wetlands. Although many who experiments are likely to be circumstances, they did not pose make a living or spend their cost-prohibitive and of limited an immediate threat to the recovery recreational time growing and/or application beyond the time of regional fish stocks (Daltonet catching fish share concerns over and place in which they were al. 2009). The authors recognised, these wide social, economic and conducted. The alternative, of however, that Alewifes would not environmental issues, they may modelling Cormorant predation, be able to produce sustainable year feel powerless to influence them is theoretical and too dependent classes below a critical spawner and so it is sometimes easier for on untested assumptions to density and that Cormorants had them to blame Cormorants for provide a reliable tool for fisheries the potential to cause such impacts reduced fish catches or changes to management (for example see in years of low Alewife escapement fish community structure (Carss Marquiss et al. 1998: 53). Thus an with detrimental consequences for et al. 2009) even when this is not iterative ‘model-field experiment- regional alewife populations. necessarily proven. remodel’ approach may be the most efficient way forward, where Given this situation, and the the validity of model predictions 10.13 Concluding ecological complexities involved are tested in the field through remarks — integrating (see section 10.3 and Text Box 10.4 observation and (preferably) perspectives and also van Eerden et al. 2012), experiment. Results can then be the impact of Cormorants in the used to refine the original models, As described and discussed vast majority of fisheries-conflict which can be tested again. in detail in INTERCAFE’s situations has never been studied publication ‘Cormorants and the scientifically. Similarly, where it This approach would lead to European environment: exploring has, scientific data have not always increasing levels of understanding, Cormorant status and distribution proven an impact. Assessing the and a higher likelihood of on a continental scale’ (van impact of Cormorants is evidently researchers understanding the Eerden et al. 2012), Cormorant more than a purely ‘scientific’

www.intercafeproject.net [119] the intercafe field manual

issue. There may well be strong of the data collected on both bird within these ‘frames’, careful evidence of Cormorant impact and fishes (e.g. see section 10.11). consideration must be given to the from other knowledge sources Detailed research work of the type criteria upon which any impact and, indeed, many stakeholder described in the Field Manual will be judged. Given the lack groups consider Cormorant impact will help our understanding of of agreement over whether there to be proven based on their own the ecological webs within which are impacts or not (partly, at least information, knowledge and local Cormorants and fish exist in a because the term means different experience. standardised way and will, to some things to different people) it is degree at least, contribute to the clear that this is an important area A sceptical reader may conclude mitigation of Cormorant-fishery requiring further study. However, a that investing in further biological conflicts and help to guide future qualitative approach (on social and research into Cormorant-fishery research on what will undoubtedly economic grounds, say) may well interactions is neither likely to remain a fascinating, if contentious, give a more accurate assessment of be cost-effective nor particularly issue across Europe. Cormorant impact as experienced useful. However, this would be by fisheries stakeholders than the wrong. Cormorant ecology, from Besides a more realistic assessment traditional biological approach. the local to the continental scale, is of the biological factors concerning ultimately governed by the species’ Cormorant impact there is also The discussions above clearly relationship with its environment a more general need to consider recognise that scientific data alone and associated fish prey (see other important elements (see also will not provide all the answers to van Eerden et al. 2012). Here, section 10.1 and 10.2). The word Cormorant-fishery conflicts and that environmental conditions will be ‘impact’ means different things to a pragmatic approach is needed to paramount in that they affect such different people (as does the term managing them. Thus, in keeping things as the presence or absence, ‘fisheries’) and both quantitative with the wording of Article 2 of productivity, abundance, quality and and qualitative approaches to the Birds Directive, a more holistic availability of fish species. There is assessing it are essential. The approach to the issue of quantifying thus clearly a lot more to be learned ‘frame’ within which knowledge of Cormorant impact should almost about these ecological relationships. Cormorant impact and associated certainly include ecological, Only with sound biological data are to be interpreted also scientific and cultural requirements understanding of Cormorant- requires consideration. There are as well as economic and recreational and fish ecology are we able to important spatial and temporal ones in some justifiable and fair manage the species in the long- elements to this and a variety of manner. Whilst this publication term (Wires et al. 2003) should scales over which impact can be focuses primarily on ecological this be necessary. There is still examined. Choosing the ‘correct’ and scientific perspectives, much much to learn about the ecological spatial and temporal scales within of INTERCAFE’s work (see van relationships between Cormorants which impact is both biologically Eerden et al. 2012, Russell et al. and fishes. Such knowledge will meaningful and meaningful to 2012, and Marzano & Carss 2012) require dedicated, skilled fieldwork specific fisheries stakeholders has focussed on integrating them and carefully considered integration may not always be easy. Similarly, with others.

[120] the intercafe field manual

REFERENCES

Preface Overviews. Report to the European thriving population of migratory Commission, contract number Cormorants Phalacrocorax carbo Carss, D N, Bevan, R M, Bonetti, Q5CA-2000–31387, pp.374. in man-dominated wetlands. Ardea A, Cherubini, G, Davies, J, 83: 1–9. Doherty, D, EL Hili, A, Feltham, Cramp, S, and Simmons, K E L. M J, Grade, N, Granadeiro, J 1980. Handbook of the Birds of van Eerden, M V, van Rijn, S, P, Grémillet, D, Gromadska, J, Europe, the Middle East and North Volponi, S, Paquet, J-Y, and Carss, Harari, Y N R A, Holden, T, Keller, Africa. The Birds of the Western D N. 2012. Cormorants and the T, Lariccia, G, Mantovani, R, Palearctic, Volume I Ostrich to European Environment: exploring McCarthy, T K, Mellin, M, Menke, Ducks. Oxford University Press. Cormorant ecology on a continental T, Mirowska-Ibron, I, Müller, W, scale. COST Action 635 Final Report Musil, P, Nazirides, T, Suter, W, Gabriel, O, Lange, K, Dahm, E, I. ISBN 978-1-906689-07-02. Trauttmansdorff, J, Volponi, S, and and Wendt, T (eds.). 2005. von Wilson, B. 1997. Techniques Brandt’s Fish Catching Methods of for assessing Cormorant diet and Counting Cormorants the World. 4th Edition, Blackwell food intake: towards a consensus Publishing, Oxford. Blackburn, J. 1999. Tree-nesting/ view. Supplemento alle Ricerche di inland Cormorant colonies. Ringers Biologia della Selvaggina XXVI: Gilbert, G, Gibbons, D W, and Bulletin 10:12. 197–230. Evans, J. 1998. Bird Monitoring Methods, Royal Society for the Bregnballe, T. 1996. Reproductive van Eerden, M R, Keller, T M, and Protection of Birds, Sandy, UK. Performance in Great Cormorants Carss, D N. 2003. Preface. In: during Colony Expansion and Keller, T M, Carss, D N, Helbig, Hoffmann, J. 1858. Der Stagnation. PhD thesis, 102 pp. A J, and Flade, M. Cormorants: Kormoran und dessen Verwendung ecology and management. zum Fischfang, Tafel 22. In: C. Bregnballe, T, and Gregersen, J. Proceedings of the 5th International Hoffmann (ed.) Das Buch der Welt. 2003. Breeding success of Great Conference on Cormorants. Hoffmannsche Verlagshandlung, Cormorants Phalacrocorax carbo Vogelwelt 124, Suppl. Stuttgart. sinensis in the Vorsø colony: variation among colony sections. Introduction Johnsgard, P A. 1993. Cormorants, Vogelwelt 124, Suppl. 115–122. Carss, D N (ed.). 2003. Darters and Pelicans of the World. REDCAFE: Reducing the Conflict Smithonian Institution Press, del Hoyo, J, Elliott, A, and Sargatal, between Cormorants and Fisheries Washington. J (eds.). 1992. Handbook of the on a pan-European Scale — pan- Birds of the World, Vol. I ICBP, European Overview. Report to the Nelson, J B. 2005. Pelicans, Lynx Edicions, Barcelona. European Commission, contract Cormorants and their Relatives. number Q5CA-2000–31387, pp.169. The Pelicaniformes. Oxford Frederiksen, M, Bregnballe, University Press. New York. T, and Reymond, A. 2003. Carss, D N, and Marzano, M (eds.). Estimating turnover at a staging 2005. REDCAFE: Reducing van Eerden, M V, Koffijberg, K, site: how many Great Cormorants the Conflict between Cormorants and Platteeuw, M J. 2003. Phalacrocorax carbo sinensis used and Fisheries on a pan-European Riding on the crest of the wave: the Lake Geneva roost in autumn Scale — Summary and National possibilities and limitations for a 1987. Vogelwelt 124, 123–125.

www.intercafeproject.net [121] the intercafe field manual

Harris, M P, and Forbes, R. 1987. of Raptor Populations in Eastern Olver, M D, and Kuyper, M A. The effect of date on counts of Fennoscandia. Proceedings of the 1978. Breeding biology of the nests of Shags Phalacrocorax workshop, Kostomusksha, Karelia, White-breasted Cormorant in Natal. aristotelis. Bird Study 34: 187 190. Russia, November 8–10 2005. Ostrich 49, 25–30.

Hughes, B, Sellers, R M, and Ekins, Krag, S M J. 2003. Variation Pihl, S, and Frikke, J. 1992. G R. 1998. Bird monitoring in nest survival within a Great Counting birds from aeroplane. In: methods — Cormorant. In Gilbert, Cormorant Phalacrocorax carbo Komdeur, J, Bertelsen, J, Cracknell, G, Gibbons, D W, and J Evans sinensis colony. Vogelwelt 124, G. (eds.) Manual for Aeroplane (eds.): Bird Monitoring Methods: Supplement: 131–137. and Ship Surveys of Waterfowl a manual of techniques for key and Seabirds. IWRB Special UK species. Royal Society for the Laursen, K, Frikke, J, and Kahlert, J. Publication 19, 8–23. Protection of Birds, Sandy, UK, 2008. Accuracy of ‘total counts’ pp.71–77. of waterbirds from aircraft in Quintana, F, and Yorio, P. 1998. coastal waters. Wildlife Biology Kelp Gull Larus dominicanus Im, B H, and H Hafner. 1984. 14/2, 165–175. predation on an Imperial Impact des oiseaux piscivores et Cormorant Phalacrocorax atriceps plus particulierement du Grand Lehikoinen, A. 2006. Cormorant colony in Patagonia. Marine Cormoran (Phalacrocorax carbo in the Finish archipelago. Ornis Ornithology 26: 84–85. sinensis) sur les exploitations Fennica 83: 34–66. Rotella, J, Taper, M, Stephens, S, piscicoles en Camargue, France. and Lindberg, M. 2007. Station Biologique, La Tour du Mayfield, H. 1961. Nesting Extending methods for modelling Valat. success calculated from exposure. heterogeneity in nest survival data Wilson Bull. 73: 255–261. using generalized mixed models. Im, B H, and H Hafner. 1985. Studies in Avian Biology 34, 32–44. Impact des oiseaux piscivores et Mayfield, H. 1975. Suggestions plus particulierement du Grand for calculating nest success. Wilson Siegel-Causey, D, and Hunt, G L. Cormoran (Phalacrocorax carbo Bull. 87, 456–466. 1981. Colonial defense behaviour sinensis) sur les exploitations in Double-crested and Pelagic piscicoles en Camargue. Bull. Nelson, J B. 2005. Pelicans, Cormorants. Ardea 98: 522–581. Mens. Office Nat. Chasse no. 94, Cormorants and their Relatives. 30–36; no. 95, 35–42. The Pelicaniformes. Oxford van Eerden, M R, Zijlstra, M, and University Press. New York. Munsterman, M J. 1991. Factors Kopciewicz, P, Stempniewicz, L, determining breeding success in Nitecki, C, Bzoma, S, and Newson, S E, and Bregnballe, T. Cormorants Phalacrocorax carbo Olszewska, A. 2003. Changes 2003. Estimating reproductive sinensis. In: van Eerden, M R, and M, in the breeding success of Great success of Great Cormorants Zijlstra (eds.): Proceedings workshop Cormorants Phalacrocorax carbo Phalacrocorax carbo: reliability 1989 on Cormorants Phalacrocorax sinensis in the expanding colony at and limitations of current carbo. Rijkswaterstaat Directorate Katy Rybackie (N Poland): effect methodology. Vogelwelt 124, Flevoland, Lelystad, The of phenology and age of subcolony. Supplement, 143–148. Netherlands, pp.67–73. Vogelwelt 124, Suppl.: 127–130. Newson, S E, Hughes, B, Hearn, R, van Rijn, S, van Eerden, M R, Korykakin, A, and Boyko, N S. and Bregnballe, T. 2005. Breeding and ZijlstraI, M. 2003. Body 2005. The White-tailed Sea performance and timing of breeding condition of young Great Eagle Haliaeetus albicilla and of inland and coastal breeding Cormorants Phaalcrocorax carbo the Common Eider Somateria Cormorants Phalacrocorax carbo sinensis as an indicator of colony mollissima in the Gulf of in England and Wales. Bird Study performance. Vogelwelt 124, Kandalaksha, White Sea. Status 52, 10–17. Suppl., 167–171.

[122] the intercafe field manual

Walsh, P M, Halley, D J, Harris, Carss, D N (ed.). 2003. Reducing Cormorants Phalacrocorax carbo L. M P, del Nevo, A, Sim, I M W, the Conflict Between Cormorants in relation to angling catches on a and Tasker, M L. 1995. Seabird and Fisheries on a pan-European coarse river fishery in north-west Monitoring Handbook for Britain Scale. REDCAFE Overview. Report England. In: Greenstreet, S P, and and Ireland. JNCC/RSPB/ITE/ to European Commission, contract Tasker, M L (eds), pp.106–110, Seabird Group, Peterborough. number Q5CA-2000–31387, pp.169. Aquatic Predators and their Prey, Blackwell Scientific Publications, Worden, J, Hall, C, and Cranswick, P. Carss, D N, Bevan, R M, Bonetti, Oxfords. 2004. Cormorant Phalacrocorax A, Cherubini, G, Davies, J, carbo in Great Britain: results of Doherty, D, EL Hili, A, Feltham, Dirksen, S, Boudewijn, T J, Noor- the January 2003 roost survey. M J, Grade, N, Granadeiro, J huis, R, and Marteijn, E C L. 1995. The Wildfowl & Wetlands Trust, P, Grémillet, D, Gromadska, J, Cormorants Phalacrocorax carbo Slimbridge. Harari, Y N R A, Holden, T, Keller, sinensis in shallow eutrophic T, Lariccia, G, Mantovani, R, freshwater lakes: prey choice Wright, G A. 2003. Turnover in McCarthy, T K, Mellin, M, Menke, and fish composition in the non- a wintering Cormorant population: T, Mirowska-Ibron, I, Müller, W, breeding period and effects of large- implications for management. In: Musil, P, Nazirides, T, Suter, W, scale fish removal. Ardea 83(1): Cowx, I G (ed.) Interactions Between Trauttmansdorff, J, Volponi, S, and 167–184. Fish and Birds: implications for Wilson, B. 1997. Techniques management, pp.345–353. Fishing for assessing Cormorant diet and Enstipp, M R, Grémillet, D, News Books, Oxford. food intake: towards a consensus and Lorentsen, S-H. 2005. view. Supplemento alle Ricerche di Energetic costs of diving and Cormorant diet, food and energy Biologia della Selvaggina XXVI: thermal status in European shags intake 197–230. (Phalacrocorax aristotelis). Journal of Experimental Biology 208: AFORO online otolith database: Carss, D N, and Ekins, G R. 2002. 3451–3461. www.cmima.csic.es/aforo/index.jsp Further European integration: mixed sub-species colonies of Great Enstipp, M R, Grémillet, D, and Black, J A. 1964. A further look Cormorants Phalacrocorax carbo Jones, D R. 2006a. The effects at Cormorants. Salmon and Trout in Britain — colony establishment, of depth, temperature and food on magazine 172: 166–180. diet, and implications for fisheries the diving energetics of Double- management. Ardea 90 (1): 23–41. crested Cormorants (Phalacrocorax Brugger, K E. 1993. Digestibility auritus). Journal of Experimental of three fish species by Double- Carss, D N, and Marquiss, Biology 209: 845–859. crested Cormorants. Condor 95: M. 1997. The diet of Cormorants 25–32. Phalacrocorax carbo in Scottish Enstipp, M R, Daunt, F, Wanless, S, freshwaters in relation to feeding Humphreys, E M, Hamer, K C, Butler, P J. 1993. To what habitats and fisheries. Ekologia Benvenuti, S, and Grémillet, D. extent can heart rate be used as Polska 45 (1): 207–222. 2006b. Foraging energetics an indicator of metabolic rate in of North Sea birds confronted free-living marine mammals? Carss, D N, and Marzano, M (eds.). with fluctuating prey availability. Symposia of the Zoological Society 2003. Reducing the conflict between In: Boyd, I L, Wanless, S, and of London 66: 317–332. Cormorants and fisheries on a pan- Camphuysen, C J (eds.) Top European scale. REDCAFE Summary predators in Marine Ecosystems Butler, P J, Green, J A, Boyd, I L, & National Overviews. Report to (Symposium of the Zoological and Speakman, J R. 2004. European Commission, contract Society London). Cambridge: Measuring metabolic rate in the number Q5CA-2000–31387, pp.374. Cambridge University Press. field: the pros and cons of the doubly labelled water and heart rate methods. Davies, J M, and Feltham, M J. Enstipp, M R, Grémillet, D, and Functional Ecology 18: 168–183. 1996. The diet of wintering Jones, D R. 2007. Investigating

www.intercafeproject.net [123] the intercafe field manual

the functional link between prey Grémillet, D, Storch, S, and energy content of the lesser sandeel abundance and seabird predatory Peters, G. 2000. Determining (Ammodytes marinus) and other fish performance. Marine Ecology food requirements in marine top preyed on by seabirds. Journal of Progress Series 331: 267–279. predators: a comparison of three Zoology of London 224: 501–517. independent techniques in great Enstipp, M R, Grémillet, D, Cormorants, Phalacrocorax Horoszewicz, L. 1960. The value and Jones, D R. 2008. Heat carbo carbo. Canadian Journal of of lower pharyngeal arches as increment of feeding in double- Zoology 78: 1567–1579. species criteria for defining fish of crested Cormorants (Phalacrocorax the Cyprinid family. Roczniki Nauk auritus) and its potential for Grémillet, D, Wright, G, Lauder, A, Rolniczych 75 b–2: 237–258. thermal substitution. Journal of Carss, D N, and Wanless, S. 2003. Experimental Biology 211: 49–57. Modelling the daily food require- Keller, T. 1995. Food of ments of wintering great Cormorants: Cormorants Phalacrocorax carbo Feltham, M J. 1990. The diet of a bioenergetics tool for wildlife sinensis wintering in Bavaria, red-breasted mergansers Mergus management. Journal of Applied southern Germany. Ardea 83 (1): serrator during the smolt run in Ecology 40: 266–277. 185–192. northeast Scotland: the importance of salmon Salmo salar smolts and Grémillet, D, Chauvin, C, Wilson, Keller, T M, and Visser, G H. 1999. parr. Journal of Zoology of London, R P, Le Maho, Y, and Wanless, S. Daily energy expenditure of Great 222: 285–292. 2005a. Unusual feather structure Cormorants Phalacrocorax carbo allows partial plumage wettability sinensis wintering at Lake Chiemsee, Goldstein, D L. 1988. Estimates in diving great Cormorants Southern Germany. Ardea 87: of daily energy expenditure in (Phalacrocorax carbo). Journal of 61–69. birds: the time-energy budget as Avian Biology 36: 1–7. an indicator of laboratory and field Lifson, N, Gordon, J B, and studies. American Zoologist. 28: Grémillet, D, Kuntz, G, Woakes, McClintock, R. 1955. Measure- 829–844. A J, Gilbert, C, Robin, J-P, Le Maho, ment of total carbon dioxide 18 Y, and Butler, P J. 2005b. Year- production by means of D2 O. Govedicˇ, M, Janzekovic, F, and round recordings of behavioural and Journal of Applied Physiology 7: Kos, I. 2002. The diet of Great physiological parameters reveal the 704–710. Cormorants Phalacrocorax survival strategy of a poorly insulated carbo on the Sava river between diving endotherm during the Arctic Libois, R M, and Hallet-Libois, C. Ljublijana and Zagorje (Slovenia). winter. Journal of Experimental 1988. Éléments pour l’identifi- Acrocephalus 23 (110–111): 5–20. Biology 208: 4231–4241. cation des restes crâniens des poissons dulçaquicoles de Grémillet, D, and Schmid, D. 1993. Hald-Mortensen, P. 1997. Does Belgique et du nord de France. 2: Zum Nahrungsbedarf des Cormorant food tell more about fish Cypriniformes. Fiches d’ostéologie Kormorans Phalacrocorax carbo than Cormorants? Supplemento alle animale pour l’archéologie. Serie sinensis. Gutachten im Auftrag des Ricerche Biologia della Selvaggina A: Poissons, n°4:24. Ministeriums für Natur, Umwelt XXVI: 173–180. und Landesentwicklung des Landes Libois, R M, Hallet-Libois, C, and Schleswig-Holstein. Universität Hennemann, W. 1985. Rosoux, R. 1987. Éléments pour Kiel. 55pp. Energetics, behavior and the l’identification des restes crâniens zoogeography of and des poissons dulçaquicoles de Grémillet, D, Schmid, D, and Double-crested Cormorants. Ornis Belgique et du nord de France. I: Culik, B M. 1995. Energy Scand. 16: 319–323. Anguilliformes, Gastérostéiformes, requirements of breeding great Cyprinidontiformes et Perciformes. Cormorants Phalacrocorax carbo Hislop, J R G, Harris, M P, and Fiches d’ostéologie animale pour sinensis. Marine Ecology Progress Smith, J G M. 1991. Variation l’archéologie. Serie A: Poissons, Series 121: 1–9. in the calorific value and total n°3:15.

[124] the intercafe field manual

Leopold, M F, Damme van, C J G, carbo at some different fishery types prey size in otters and mink. Philippart, C J M, and Winter, C J in England and Wales. Vogelwelt Journal of Zoology of London 192: N. 2001. Otoliths of North Sea 124, Supplement 255–259. 25–31. Fish. ETI & UNESCO. The World Biodiversity Database. CD-ROM. Rutte, E. 1962. Schlundzähne Veldkamp, R. 1995. Diet of von Süßwasserfischen. Paläont- Cormorants Phalacrocorax carbo Lovvorn, J R, and Jones, D R. ographica, Abt. 120: 165–212. sinensis at Wanneperveen, The 1991. Body mass, volume, and Netherlands, with special reference buoyancy of some aquatic birds, Schulz-Mirbach, T, and Reichen- to Bream Abramis brama. Ardea 83 and their relation to locomotor bacher, B. 2006. Reconstruction (1): 143–156. strategies. Canadian Journal of of Oligocene and Neogene fresh Zoology 69: 2888–2892. water fish faunas — an actualistic Zijlstra, M, and van Eerden, M R. study on cypriniform otoliths. Acta 1995. Pellet production and the März, R. 1987. Gewöll- und Palaeontologica Polonica 51 (2): use of otoliths in determining the Rupfungskunde. Akademie Verlag. 282–304. diet of Cormorants Phalacrocorax Berlin. carbo sinensis: trials with captive Speakman, J R. 1997. Doubly birds. Ardea 83: 123–131. Mehner, T. 1992. Bestimmung Labelled Water: theory and practice. von Fischresten aus Fraßresten der Chapman & Hall, London. Sexing, ageing and measuring Vögel. Der Falke 92/39: 56–59. Cormorants Storch, S, Grémillet, D, and Koffijberg, K, and M R van Eerden. Pennycuick, C J. 1989. Bird Culik, B M. 1999. The telltale 1995. Sexual dimorphism in the flight performance: a practical heart: a non-invasive method to Cormorant Phalacrocorax carbo calculation manual. Oxford determine the energy expenditure sinensis: possible implications for University Press, Oxford. of incubating great Cormorants differences in structural size. Ardea Phalacrocorax carbo carbo. Ardea 83, 37–46. Petrova, G, and Zivkov, M. 1989. 87: 207–215. A contribution to the methodology Millington, R. 2005). Identification of the study of predator fishes food. Strod, T, Izhaki, I, Arad, Z: Weihs, D, of North Atlantic and Continental Vets.cs. Spolec. zool. 53: 280–290. and Kazir, G. 2003. Cormorants Cormorants. Birding World 18, 3, Phalacrocorax carbo swallow 112–123. Przyblski, M. 1996. Variation in fish under water. Vogelwelt 124, fish growth characteristics along Supplement: 270. Nelson, J B. 2005. Pelicans, river course. Hydrobiologia 325: Cormorants, and their Relatives. 39–46. Trauttmansdorff, J. 2003. The Pelicaniformes. Oxford Analysis of Great Cormorant University Press. New York. Ricker, W E. 1979. Growth rates Phalacrocorax carbo sinensis and models. In: Fish Physiology stomach contents from different Newson, S E, B, Hughes, I C, Vol. VIII. 677–743. Academic Press. areas of Austria and Liechtenstein. Russell, G R, Ekins, and R M, Vogelwelt 124, Supplement 271–276. Sellers. 2004. Sub-specific Ridgway, M S. 2010. A review of differentiation and distribution of estimates of daily energy expenditure Wilson, R P, and Culik, B M. great Cormorants Phalacrocorax and food intake in Cormorants 1993. Activity-specific metabolic carbo in Europe. Ardea 93: 3–10. (Phalacrocorax spp.) Journal of rates from double-labelled Great Lakes Research 36: 93–99. water studies: are activity costs Platteeuw, M, Koffijberg, K, and underestimated? Ecology 74: Dubbeldam, W. 1995. Growth of Russell, I C, Cook, A C, Kinsman, 1285–1287. Cormorant Phalacrocorax carbo D A, Ives, M J, and Lower, N J. sinensis chicks in relation to brood 2003. Stomach content analysis Wise, M H. 1980. The use of fish size, age ranking and parental of Great Cormorants Phalacrocorax vertebrae in scats for estimating fishing effort. Ardea 83, 235–245.

www.intercafeproject.net [125] the intercafe field manual

van Eerden, M R, and Munsterman, Süddeutschland und Österreich in Cormorants Phalacrocorax carbo M J. 1986. Importance of the — eine Ringfundanalyse 19986– passing through or wintering in Mediterranean for wintering 1999. Ornithologischer Beobachter Switzerland. Ardea, 83: 275–280. Cormorants Phalacrocorax 96, 183–192. carbo sinensis. In: Medmarvai & Reymond, A, and Zuchuat, O. 1995b. Monbailliu (eds.) Mediterranean Frederiksen, M, and Bregnballe, T. Perch fidelity of Cormorants Marine Avifauna, NATO Asi Series, 2000. Evidence for density- Phalacrocorax carbo outside the Vol. G12:123–141. Springer Verlag, dependent survival in adult breeding season. Ardea, 83: 281–284. Berlin. Cormorants from a combined analysis of recoveries and van Eerden, M R, and Munsterman, van Eerden, M R, and Munsterman, resightings. Journal of Animal M J. 1986. Importance of the M. 1995. Sex and age dependent Ecology 69: 737–752. Mediterranean for wintering distribution in wintering Cormorants Cormorants Phalacrocorax carbo Phalacrocorax carbo sinensis in Frederiksen, M, Bregnballe, T, sinensis. In: Medmaravis & Western Europe. Ardea 83, 285–297. van Eerden, M, van Rijn, S, and Monbailliu (eds.) Mediterranean Lebreton, J D. 2002. Site fidelity of Marine Avifauna, NATO Asi Series, Movements wintering Cormorants Phalacrocorax Vol. G12:123–141. Springer Verlag, carbo sinensis in Europe. Wildlife Berlin. Bregnballe, T. 1999. Seasonal Biology 8/4, 241–250. and geographical variation in van Eerden, M R, and Munsterman, net-entrapment of Danish Great Galván, I. 2005. Migration M J. 1995. Sex and age dependent Cormorants Phalacrocorax carbo strategies of the Great Cormorant ditribution in wintering Cormorants sinensis. Dansk Ornitologisk wintering inland in Spain. Phalacrocorax carbo sinensis in Forenings Tidsskrift 93: 247–254. Waterbirds 28 (3): 301–307. Western Europe. Ardea 83:285–297.

Bregnballe, T. 2006. Age-related Hénaux, V, Bregnballe, T, and van Rijn, S, and van Eerden, M. fledgling production in Great Lebreton, J B. 2007. Dispersal 2007. Recent population changes Cormorants Phalacrocorax carbo: and recruitment during population of Cormorants Phalacrocorax influence of individual competence growth in a colonial bird, the great carbo in the IJsselmeer area, the and disappearance of phenotypes. Cormorant Phalacrocorax carbo Netherlands. Aves, 44 (3): 151–154. Journal of Avian Biology 37(2): sinensis. J. Avian Biol. 38: 44–57. 149–157. Sackl, P, and Zuna-Kratky, T. 2004. Paquet, J Y, Dermien, F, Lacroix, P, Herkunftsgebiete, Wanderrouten Bregnballe, T, Frederiksen, M, and Lucas, P, and Pourignaux, F. 2003. und Zugablauf in Österreich Gregersen, J. 1997. Seasonal Year-to-year site-fidelity of wintering rastender und überwinternder distribution and timing of migration and migrating Great Cormorants Kormorane (Phalacrocorax carbo of Cormorants Phalacrocorax Phalacrocorax carbo in Belgian sinensis) — eine Ringfundanalyse. carbo sinensis breeding in Meuse valley. Vogelwelt 124, Suppl.: Egretta 47: 39–65. Denmark. Bird Study 44: 257–276. 53–60. Yésou, P. 1995. Individual Bregnballe, T, Frederiksen, M, and Retter, M. 2000. Origin and migration strategies in Cormorants Gregersen, J. 2006. Effects of migration pattern of Cormorants Phalacrocorax carbo passing distance to wintering area on arrival (Phalacrocorax carbo sinensis) in through or wintering in Western date and breeding performance in the Upper Rhine area — an analysis France. Ardea, 83: 267–274. Great Cormorants Phalacrocorax of colour-ring sightings of wintering carbo. Ardea 94(3): 619–630. and staging birds. Naturschutz südl. Other relevant literature Oberrhein, 3: 1–11. Fiedler, W. 1999. Kormorane Jenard, P. 2007. Suivi détaillé de Phalacrocorax carbo als Reymond, A, and Zuchuat, la reproduction du Grand Cormoran Durchzügler und Wintergäste in O. 1995a. Axial migration routes (Phalacrocorax carbo) dans les

[126] the intercafe field manual

colonies d’Hensies en 2005 et Available at: http://www. Pitcher, T J (ed.). 1993. Behaviour comparaison avec les résultats de environment-agency.gov.uk/static/ of Teleost Fishes. Second Edition, 1999. Aves 44 (2), 79–90. documents/Business/stocking__ Chapman & Hall, London. eng_172017.pdf Paquet, J-Y, Malengreau, A, and Wheeler, A. 1978. Key to Simar, J. 2007. Dispersion des Hart, P J, and Reynolds, J D. 2002. the Fishes of Northern Europe. Grands Cormorans (Phalacrocorax Handbook of Fish Biology and Frederick Warne & Co. Ltd. carbo) de la colonie d’Hensies Fisheries (Volume 1), Blackwell (Z.h.i.b. des marais d’Harchies): Scientific Publications, Oxford, Wootton, R J (ed.). 1990. premiers résultats du baquage 432pp. Ecology of Teleost Fishes. couleur. Aves 44 (3), 145–148. Chapman & Hall, London. Hassan, R, Scoles, R, and Ash, N. Sandor, D A, Kiss, J B, and 2005. The Millennium Ecosystem Estimating abundance of fish Domsa, C. 2009. The importance Assessment Series; v.1 Current species of Northern Dobrogea in the State and Trends. Island Press, Appelberg, M. 2000. Swedish migration of Great Cormorant Washington, 917pp. (Phalacrocorax carbo). Scientific standard methods for sampling Annals of the Danube Delta Inst., freshwater fish with multi-mesh Hickley, P. 1994. Stocking and gillnets. Fiskeriverket Inf. 2000, 1. Tulcea, 15: 41–46. introduction of fish — a synthesis. In: I G, Cowx (ed.) Rehabilitation of European Committee for Working with Fishes Freshwater Fisheries, pp. 247–254. Standardisation (CEN) (2002) Carss, D N, Marquiss, M, and Fishing News Books, Blackwell European Standard (EN) 14011: Lauder, A W. 1997. Cormorant Scientific Publications, Oxford. Water quality — sampling of fish Phalacrocorax carbo carbo with electricity. Ref. No. prEN predation at a major trout fishery Kottelat, M, and Freyhof, J. 2007. 14011:2002 E. CEN. 16 pages. in Scotland. Supplemento alle Handbook of European Freshwater Ricerche di Biologia della Fishes. Publications Kottelat, European Committee for Selvaggina Volume XXVI (No. 1): Switzerland, 646pp. Standardisation (CEN) (2004) 281–294. European Standard (EN) 14962: Marquiss, M, Carss, D N, Armstrong, Water quality — Guidance on Cowx, I G. 1999. An appraisal J D, and Gardiner, R. 1998. Fish- the scope and selection of fish of stocking strategies in the light eating Birds and Salmonids in sampling methods. Ref.No. prEN of developing country constraints. Scotland. Report to The Scottish 14962: 2004:E. CEN. 25 pages. Fisheries Management and Office Agriculture, Environment Ecology, 6: 21–34. and Fisheries Department. Scottish European Committee for Office, Edinburgh, 156pp. Standardisation (CEN) (2005) Carvalho, G R, and Cross, T. 1998. European Standard (EN) 14757: Enhancing fish production through McGinnity, P, Prodohl, P, Ferguson, Water quality — Sampling of fish introductions and stocking: genetic A, Hynes, R, O’Maoileidigh, N, with multi-mesh gillnets. Ref. No. perspectives. In: I G, Cowx (ed.) Baker, N, Cotter, D, O’Hea, B, prEN 14757:2005 E. CEN. 27 pages. Stocking and Reintroduction of Fish. Cooke, D, Rogan, G, Taggart, J, Fishing News Books, Blackwell and Cross, T. 2003. Fitness Cowx, I G. 1983. Review Scientific Publications, Oxford, reduction and potential extinction of the methods for estimating pp.329–337. of wild populations of Atlantic fish population size from salmon, Salmo salar, as a result of survey removal data. Fisheries Environment Agency (undated) interactions with escaped farmed Management 14:67–82. Stocking Fish: a guide for fishery salmon. Proceedings of the Royal owners and anglers. Environment Society of London, Series B: Cowx, I G. 1991. Catch effort Agency, Kidderminster, UK. Biological Sciences, 270: 2443–2450. sampling strategies: conclusions

www.intercafeproject.net [127] the intercafe field manual

and recommendations for Kieffer, J D, and Colgan, P W. still waters using hydroacoustics management. In: Catch Effort 1992. The role of learning in fish and survey gill netting: experiences Sampling Strategies (Ed. I.G. behaviour. Reviews in Fish Biology with Arctic charr (Salvelinus Cowx). Fishing News Books, and Fisheries 2, 125–143. alpinus) in the UK. Fisheries Blackwell Scientific Publications, Research 96: 30–38. Oxford. Magurran, A E. 1990a. The adaptive significance of schooling as Indicators of Cormorant damage Cowx, I G. 1995. Fish stock anti-predator defence in fish. Annales assessment — biological basis for Zoologici Fennici 27, 51–66. Beveridge, M E M. 1988. sound ecological management. Problems caused by birds at inland In. Harper, D, and Ferguson, A Magurran, A E. 1990b. The waters and freshwater fish farms. (eds.): Biological Basis for River inheritance and development of EIFAC Tech. Rep. 51:34–73. Management. Wiley and Son, minnow anti-predator behaviour. London: 375–388. Animal Behaviour 39, 834–842. Callaghan, D A, Kirby, J S, and Bell, M C. 1994. An assessment Cowx, I G. 1996. Stock Murphy, B R, and Willis, D W of Cormorant Phalacrocorax carbo Assessment in Inland Fisheries. (eds.). 1996. Fisheries Techniques. occupancy and impact at stillwater Fishing News Books, Blackwell Bethesda: American Fisheries game fisheries in England and Scientific Publications,Oxford, 513pp. Society. Wales. Report to the Association of Stillwater Game Fishery Managers, De Lury, D. 1947. On the Nielsen, L A, and Johnson, D L WWT, Slimbridge. estimation of biological populations. (eds.). 1983. Fisheries Techniques. Biometrics 3: 145–167. The American Fisheries Society, Carss, D N. 1990. Beak-prints Southern Printing Company, Virginia, help in war against aerial invaders. Gabriel, O, Lange, K, Dahm, E, USA. Fish Farmer, November-December and Wendt, T (eds.). 2005. von 1990: 46–48. Brandt’s Fish Catching Methods of Reynolds, J B. 1983. Electro- the World. 4th Edition. Blackwell fishing. In: Fisheries Techniques Carss, D N. 1993. Cormorants Publishing, Oxford. (eds: Nielsen, L A, and Johnson, Phalacrocorax carbo at cage fish D L) American Fisheries Society farms in Argyll, western Scotland. Haunschmid, R. 2004. Dynamik Bethesda, Maryland. 468pp. Seabird 15: 38–44. des Bachforellenbestandes an drei Untersuchungsstrecken der Kleinen Seber, G A F, and Le Cren, E Carss, D N, and Marquiss, M. Mühl (OÖ). Dissertation an der D. 1967. Estimating population 1991. Avian predation at farmed Universität Salzburg. 131pp. parameters from the catches large and natural fisheries. Interactions relative to the population. Journal between fisheries and the Hemingway, K L, and Elliot, M. of Animal Ecology 36, 631–643. environment. Proceedings of the 2002. Fishes in Estuaries. Institute of Fisheries Management Blackwell Science. Oxford. Winfield, I J, Fletcher, J M, and 22nd annual study course. James, J B. 2007. Seasonal Hilborn, R, and Walters, C J. variability in the abundance of Carss, D N, and Marquiss, M. 1992. Quantitative Fisheries Arctic charr (Salvelinus alpinus (L.)) 1995. Fish-eating birds: Stock Assessment — choice and recorded using hydroacoustics in perceptions and realities. British dynamics. Chapman and Hall, N.Y. Windermere, UK and its implications Trout Farming Conference, London. 570pp. for survey design. Ecology of Sparsholt, September 1995. Freshwater Fish 16: 64–69. Jennings, S, Kaiser, M J, and Cornelisse, K J, and Christensen, Reynolds, J D. 2001. Marine Winfield, I J, Fletcher, J M, James, K D. 1993. Investigation of Fisheries Ecology. Blackwell J B, and Bean, C W. 2009. a cover net designed to reduce Science. Oxford. Assessment of fish populations in southern cormorant (Phalacrocorax

[128] the intercafe field manual

carbo sinensis) fisheries depredation Seiche, K, and Wünsche, A. 1996. Carss, D N, Bell, S, and Marzano, M. in a pound net. ICES J. Mar. Sci., Kormoran (Phalacrocorax carbo L.) 2009. Competing and coexisting 50:279–284. und Graureiher (Ardea cinerea L.) with Cormorants: ambiguity and im Freistaat Sachsen. Freistaat change in European wetlands. In: Dieperink, C. 1993. Fisheries Sachsen, Staatsministerium für Heckler, S. (ed.) Landscape, Process depredation caused by cormorant Umwelt and Landesentwicklung, and Power: re-evaluating traditional (Phalacrocorax carbo sinensis Dresden. 102 pp. environmental knowledge, pp.99– Shaw). EIFAC Workshop, Starnberg, 121. Berghan Books, Oxford. Germany, July 1993 (abstract only). Staub, E, Krämer, A, Müller, R, Ruhlé, Ch, and Walter, J. 1992. Cherrett, J M. 1989. Ecological Engström, H. 1998. Conflicts Einfluss des Kormorans (Phala- concepts: the contribution of between cormorants Ph. c. sinensis crocorax carbo) auf Fischbestände ecology to an understanding of and fishery in Sweden. Nordic und Fangerträge in der Schweiz. the natural world. Volume 29 Journal of Freshwater Research Schriftenreihe Fischerei, No. of Symposium of the British 74:148–155. 50:1–131. Ecological Society, Blackwell Scientific Publications. Moerbeek, D J, van Dobben, W Staub, E, and Ball, R. 1994. Crawley, M J. 1992. Population H, Osiek, E R, Boere, G C, and Effects of cormorant predation on dynamics of natural enemies and Bungenberg De Jong, C M. 1987. fish populations of inland waters. their prey. In: Crawley, M J. (ed.) Cormorant damage prevention Working document for 18th Session Natural Enemies: the population at a fish farm in the Netherlands. of EIFAC, and Report of the biology of predators, parasites Biological Conservation 39: 23–38. EIFAC Working Party, Starnberg, Germany, July 1993, EIFAC/ and diseases, pp.40–89. Blackwell XVIII/94/Inf.8 Rev. May 1994. Scientific Publications. Musil, P, Janda, J, and De Nie, W. 1995. Changes in abundance Suter, W. 1995. Are cormorants Dalton, C M, Ellis, D, and Post, D and foraging habitat selection in Phalacrocorax carbo wintering in M. 2009. The impact of double- Cormorant (Phalacrocorax carbo) Switzerland approaching carrying crested Cormorant (Phalacrocorax in the south Bohemia. Ardea 83: capacity? An analysis of increase auritus) predation on anadromous 247–254. patterns and habitat choice. Ardea, alewife (Alosa pseudoharengus) in 83:255-266. south-central Connecticut, USA. Musil, P, and Janda, J. 1997. Canadian Journal of Fisheries and Habitat selection by Cormorant van Dobben, W.H. 1952. The Aquatic Science 66: 177–186. (Phalacrocorax carbo) on South food of the cormorant in The Bohemian fishponds. Ekologia Feltham, J M, Davies, J M, Wilson, Netherlands. Ardea 40:1-63. Polska 45: 173–180. B R, Holden, T, Cowx, I G, Harvey, J P, and Britton, J R. 1999. Case Attempting to integrate Ranson, K, and Beveridge, M C M. studies of the impact of fish-eating Cormorant and fish data 1983. Raiders from the skies. birds on inland fisheries in England Bird predation at a freshwater cage Begon, M, Harper, J L, and and Wales. Report to the Ministry farm. Fish Farmer 6: 22–23. Townsend, C R. 1996. Ecology. of Agriculture, Fisheries and Food, 3rd edition, Blackwell Science. 144pp. Russell, I C, Dare, P J, Eaton, D R, and Armstrong, J A. 1996. Carpenter, S R. 1993. Statistical Kitchell, J F, and Carpenter, S R. Assessment of the problem of fish- analysis of the ecosystem 1993. Cascading trophic eating birds in inland fisheries in experiments. In: Carpenter, S interactions. In: Carpenter, S R, England and Wales. Report to the R, and Kitchell, J F. (eds.) The and Kitchell, J F. (eds.) The Ministry of Agriculture, Fisheries Trophic Cascade in Lakes, pp.26– Trophic Cascade in Lakes, pp.1–14. and Food (MAFF Project VC0104), 42. Cambridge Studies in Ecology, Cambridge Studies in Ecology, 130pp. Cambridge University Press. Cambridge University Press.

www.intercafeproject.net [129] the intercafe field manual

Krebs, C J. 1989. Ecological Pitcher, T J (ed.). 1993. Wires, L R, Carss, D N, Cuthbert, Methodology. Harper Collins. Behaviour of Teleost Fishes. 2nd F J, and Hatch, J J. 2003. Trans- Edition, Chapman & Hall, London. continental connections in relation Krebs, C J. 2001. Ecology: to Cormorant-fisheries conflicts: the experimental analysis of Raffaelli, D, and Hawkins, S. perceptions and realities of a ‘bête distribution and abundance. 5th 1999. Intertidal Ecology. Kluwer noire’ (black beast) on both sides Edition, Benjamin Cummings, San Academic Publishers. of the Atlantic. Vogelwelt 124, Francisco. Supplement: 398–400. Russell, I C. 2006. Definitions. Marquiss, M, and Carss, D N. 1994. In: Carss, D N, and Marzano, Wootton, R J (ed.). 1990. Avian Piscivores: basis for policy. M. (eds.) Commercial Carp Ecology of Teleost Fishes. National Rivers Authority, Bristol. Aquaculture. INTERCAFE Chapman & Hall, London. meeting report, Saxony, Germany, Marquiss, M, and Carss, D N. September 2005: 10–14. APPENDIX ONE: RELATED 1997. Fish-eating birds and CORMORANT SPECIES IN fisheries. BTO News 210/11: 6–7. Russell, I C, Dare, P J, Eaton, D R, EUROPE British Trust for Ornithology. and Armstrong, J D. 1996. Assessment of the problem of fish- Cramp, S, and Simmons, K E L. Marquiss, M, Carss, D N, Armstrong, eating birds in inland fisheries in 1977. Handbook of the birds of J D, and Gardiner, R. 1998. Fish- England and Wales. Report to the Europe, the Middle East and North eating birds and Salmonids in Ministry of Agriculture, Fisheries Africa. The birds of the Western Scotland. Report to The Scottish and Food. Directorate of Fisheries Palaearctic. Volume 1:216–219. Office Agriculture, Environment Research, Lowestoft, 130pp. Oxford University Press. and Fisheries Department. Scottish Office, Edinburgh. van Eerden, M, van Rijn, S, Johnsgard, P A. 1993. Volponi, S, Paquet, J-Y and Cormorants, Darters and Pelicans Marzano, M, and Carss, D N Carss, D. 2012. Cormorants of the World. Smithsonian (eds). 2012. Essential social, and the European Environment: Institution Press, Washington. cultural and legal perspectives exploring Cormorant status and on Cormorant-fisheries conflicts. distribution on a continental scale. Nelson, B J. 2005. Pelicans, INTERCAFE COST Action 635 INTERCAFE COST Action 635 Cormorants and their Final Report IV. ISBN 978-1- Final Report I. Relatives — the Pelecaniformes. 906698-11-9. ISBN 978-1-906698-07-2 Oxford University Press, Oxford.

[130] the intercafe field manual

11 APPENDIX ONE: RELATED CORMORANT SPECIES IN EUR0PE

The most abundant cormorant light colouration on their underside eyes, black scapulars and upper species in Europe is the Great and from Great Cormorants by wing-coverts with a grey tinge Cormorant Phalacrocorax carbo, their smaller size. Overall, Pygmy and darker edges giving scaled which has two subspecies or races Cormorants are around half the size appearance. Otherwise, plumage (see chapter 2). However, two other of Great Cormorants. is black with a green sheen above cormorant species occur throughout and light brown below. The bill is Europe, the Pygmy Cormorant 11.1.1 Sex & age black-brown, the bare facial skin (Phalacrocorax pygmeus) and the is tinged pink, and the feet are European Shag (Phalacrocorax Adults: in autumn and early winter, blackish. In summer (moult starts aristoteles) with two subspecies the head, upper half of neck, in December), the head and upper (Ph. a. aristotelis and Ph. a. and the breast are dark brown, neck are strongly tinged red-brown, desmarestii). The geographical sometimes looking black with a becoming nearly black before ranges of these three cormorants brown tinge. There is some black breeding, with short crest and overlaps in Europe and there is on the forepart of the head, variable scattering of white filoplumes over thus potential for some confusion white on the throat and around the head, neck. The underbody and between them. This Appendix upper tail-coverts are black and the provides a summary of information scapulars and upper wing-coverts about these two other European greyer with a more marked scaled cormorant species (and mentions appearance. Females are slightly differences between them and P. duller than males. carbo) and is taken from three publications (Cramp & Simmons Juveniles: have a dark brown 1977, Johnsgard 1993, and Nelson crown and neck, whitish chin, grey- 2005). brown fore neck and breast, and brownish-white belly blotched with darker brown and orange-brown. 11.1 Pygmy Cormorant They have blackish flanks and (Phalacrocorax pygmeus) the under tail-coverts are black- brown, the back has lighter feather This is the smallest of the three edges, and grey-tinged scapulars species of European cormorants and upper wing-coverts with dark and has a distinctly thinner and edges. The bill is yellowish. shorter bill. In breeding plumage the neck and head plumage is a Size: body length = 45–55 cm, rich brownish black, the body wing span = 80–90 cm, wing length black with greenish iridescence (males) = 195–217 cm, (females and some white spots not present = 193–208 cm, weight (males) = in other seasons. Immatures are 650–870 g, (females) = 565–640 g, distinguishable from immature Figure 11.1 Adult Pygmy bill length (males) = 30.5 mm, European Shags by more extensive Cormorant. Photo courtesy of R Sauli. (females) = 29.2 mm.

www.intercafeproject.net [131] the intercafe field manual

11.1.3 Roosts The roosting sites of Pygmy Cormorants are usually located on very inaccessible islands or flooded trees over the water. Unlike Great Cormorants, Pygmy Cormorants often arrive at roosts individually (especially at the beginning of a roost’s occupation). Flocks of up to 120 or more birds arrive during the peak hours, but normally arriving groups number 2–40 individuals. Pygmy Cormorants do not wander around roosts, trying to see if it is safe enough, but land directly. However

Figure 11.2 Pygmy Cormorant nestlings. Photo courtesy of S Volponi. they do use the highest parts of neighbouring trees and bushes to check the roost sometimes. As 11.1.2 Diet on trees branches) can generally dusk approaches, birds occupy the be collected and used for further main roosting trees and bushes, Primarily fish, but data are sparse processing and investigation. especially their lowest branches with other prey only occasionally Smaller fish species appear to (unlike Great Cormorants), where recorded, including young water- be completely digested judging they concentrate at up to 6–8 voles, frog larvae, crustaceans (e.g. by their absence in pellets. Thus birds per linear meter. Although shrimps), aquatic invertebrates and pellets from Pygmy Cormorants are foraging birds can be present at a leeches. Remains of vegetarian origin not useful for estimating either fish roosting site throughout the day, may also occur in diet. Normally species composition or the daily the first birds usually arrive about feeds during daytime either singly or food intake of the birds — they 3 hours before dusk, and the peak in pairs. It is less often seen in groups merely give a very general picture arrival time is 60–90 min before (during winter mostly). Average of the prey taken. dusk. The roost may be considered weight of fish needed per day is estimated to be 115 g.

The pellets of the Pygmy Cormorants (unlike those of Great Cormorants) appear to have a smaller but thicker mucus covering, containing very few remains from the fish eaten (otoliths in general). Empty pellets are not unusual. The pellets can be opaque white, yellow or brown coloured. As roosting sites are, in most cases, located over water, most pellets just fall into the water or become useless for analysis as any bones fall from the waterlogged mucus. Thus only those pellets that fall onto dry ground (or that become stuck Figure 11.3 Pellets of Pygmy Cormorants. Photo courtesy of I Nikolov.

[132] the intercafe field manual

occupied some 20–30 min before of the roost, many birds may arrive shores and cliffs. It is smaller than dusk (contrasting with Great individually at the same time from the Great Cormorant and lacks Cormorants, where some birds many directions and even if the that species’ white chin and thigh arrive after it is dark). whole roost could be overlooked by spot in the breeding period. During one person, it is not always possible this time the European Shag has a Flocks of Pygmy Cormorant appear to follow and count all the birds. pronounced head crest. The wing- to have no special order. They are Many birds approach the roosting beats of the European Shag are ‘chaotic’ and much more difficult to site flying very low over the water considerably faster than those of the count than the Great Cormorant’s (3–10 m) and if the background is Great Cormorant. The overall green V-shaped flocks. Additional dark (or the roosting site can not gloss of the Shag’s plumage and its difficulty may be caused by the way be overlooked from an observation yellow bill base and gape separate it the birds fly — sometimes Pygmy point) they can be easily missed if from the other European cormorant Cormorants make loops within the only one person counts. Sometimes species. At longer distances it is flock just before landing at the roost, during the peak hours of roost sometimes difficult to distinguish thus disarranging the formation even occupation, all the birds take off for European Shags and Great more. This behaviour means that to no apparent reason and fly over the Cormorants but at all times the count a returning flock as accurately roost whilst other birds continue slimmer build, shorter neck, smaller as possible, it should be observed at to arrive at the same time — an head and faster wing-beats of the least 200–300 m before reaching the impossible situation to handle for Shag help identification. There are roost site. just one observer. two subspecies of European Shags in Europe: P. a. aristotelis (north The best time for counting larger 11.1.4 Breeding and west) and P. a. desmarestii roosts is before dusk as birds/ (Mediterranean and Black Sea). flocks arrive one by one. There are Pygmy Cormorants produce one two main reasons for not counting brood, the clutch size = 4–6 eggs 11.2.1 Sex & age Pygmy Cormorants in the morning: (usually 3–7). Egg length = 40–52 x most of the birds take off at the 28–33 mm and egg weight = 23 g. Adults: The breeding plumage of same time (not in small groups) and Incubation lasts 27–30 days. the European Shag is mainly oily accurate counting is not possible at black glossed dark green, but the all; on the other hand as the birds Colonies are often polyspecific head and the neck are glossed dark are highly concentrated on the tree (mixed with herons, egrets, blue-green. The mantle, scapulars branches (and many of them remain spoonbills, ibises or other and wing-coverts are purplish, invisible in the dense group) it is also cormorant species). Nests are latter feathers are bordered with not possible to count them accurately located on trees or in dense velvet-black to give a scaled effect while they are standing in the roost. reedbeds. While counting or ringing across the mantle and wings. The birds within colonies the fact that forward-curved crest is usually When counting bigger roosts of many other species may be present conspicuous during the breeding Pygmy Cormorants a few things too (with different breeding periods period. They have a black bill with should be considered. Counts and behaviour) must be considered a yellow base to the lower mandible generally require the presence of carefully as such activities within and an orange-yellow gape. The a second observer, as birds may the colony may cause significant eyes are bright green. The legs arrive from any direction and if the and undue disturbance to the whole and the feet are black. The non- roosting territory is big enough (and mixed colony. breeding plumage (moult starts has many trees), lots of birds can be June–August) is duller and browner missed. Thus the two counters may without the crest, whilst the chin be forced to count separately from 11.2 European Shag is white to brownish-white and the the opposite sides of the roosting (Phalacrocorax aristotelis) throat is brown. site with a preliminary stipulation about the method of counting. The European Shag is found almost Juveniles: The juvenile plumage is Again, depending on the location exclusively in salt water off rocky medium to dark brown above with

www.intercafeproject.net [133] the intercafe field manual

= 90–105 cm, wing length (males) remain solitary in winter and when = 261–278 mm, (females) = 251– away from the nesting colony. 269 mm, weight (males) = 1,930 g Occasionally they flock at fish (breeding time), bill length (males) shoals but with minimal social = 55.0 mm, (females) = 55.3 mm. interaction. This essentially marine species does not usually range 11.2.2 Diet far from the coast, preferring rocky coastlines, where they roost Chiefly and often entirely fish, on stack rocks or cliffs. Large most times caught under water and numbers may use the same rock, brought to the surface. Unlike the but some roost alone on narrow Great Cormorant, which (at sea) ledges. mainly hunts on bottom-living fish species, European Shags prefer 11.2.4 Breeding species of the open like Clupeidae (Herring fishes) andGadidae (Cod European Shags produce one fishes). As with other cormorants, brood, the clutch size = 1–6 eggs European Shags produce pellets (usually 3), egg length = 52–72 which contain the undigested x 35–41 mm, egg weight = 49 g, remains of their prey. incubation lasts 30–31 days.

Figure 11.4 Adult European Shag 11.2.3 Roosts Shags usually breed in small, (P. a. aristotelis). loose colonies, though sometimes Photo courtesy of R Bardgett. Shags are less gregarious than densely. They defend only the nest- Great Cormorants. They often site territory. inconspicuous green gloss.There is a scaly effect on the mantle and scapulars but pale tips to the wing- coverts and colouration is paler on the sides of the head. Plumage is generally paler brown below, with small but variable areas of brownish- white on the chin and in the centre of throat, on the breast and belly. The P. desmarestii race is wholly whitish to brownish-white below.

Adult shags are unmistakable and immatures differ from young Great Cormorants in their smaller size, slimmer build, and much more slender bill. Also, (except for demarestii) there is much less white on the brown breast, though sometimes there is a white spot on the chin. Figure 11.5 Adult European Shag P. a. aristotelis taking flight – identifiable Size: body length = 65–80 cm by greenish plumage, head crest, yellow base to lower mandible, and (males on average larger), wing span indication of fast (frequent) wing beats. Photo courtesy of Scott Jones.

[134] the intercafe field manual

12 APPENDIX TWO: NATIONAL AND INTERNATIONAL CORMORANT COUNTS, EXAMPLES OF DATA FORMS

 (1) Belgium 5HFHQVHPHQWVKLYHUQDX[GHVGRUWRLUVGH*UDQG&RUPRUDQ  &HIRUPXODLUHSHXWVLPSOHPHQWrWUHFRSLpFROOpGDQVOHFRUSVGHPHVVDJHG·XQPDLODXFRRUGLQDWHXU6L ODGRQQpH D pWp HQFRGpH SDU YRV VRLQV VXU ZZZREVHUYDWLRQVEH PHUFL GHFRSLHUFROOHULFLO·85/GH O·REVHUYDWLRQ««««««««««««««««««  6LWH ORFDOLWpOLHXGLWVLQRXYHDXVLWHFRRUGRQQpHV/DPEHUWRXDXWUHORFDOLVDWLRQSUpFLVH  ««««««««««««««««««««««««««««««««««««« ««««««««««««««««««««««««««««««««««««« 'DWH«««««««« +HXUHVGH«««««j««««« 2EVHUYDWHXU V «««««««««««««««««««««««««««««« 0pWpRORFDOH SUpFLVHUVLSUREOqPHSRXUOHGpQRPEUHPHQW ««««««««««««« ««««««««««««««««««««««««««««««««««««« 5HPDUTXHFRQFHUQDQWOHGpQRPEUHPHQW«««««««««««««««««««« ««««««««««««««««««««««««««««««««««««« $XWUHVHVSqFHVSUpVHQWHVGDQVOHGRUWRLU *UDQGH$LJUHWWH+pURQFHQGUpHWF  «««««««««««««««««««««««««««««««««««««  1RPEUHWRWDO ILQDO GH*UDQGV&RUPRUDQV«««««««««««««««««««««« $JHUDWLR 1RPEUHG DGXOWHV«««««««««««««««««««««««««« 1RPEUHG LPPDWXUHV««««««««««««« 1RPEUHG LQGpWHUPLQpV«««««««««««««  2ULJLQHRXGpSDUWGH&RUPRUDQVSHQGDQWOH GpQRPEUHPHQW IDFXOWDWLISHUPHWVXUWRXWGHQRWHUOHV pYHQWXHOVIXLWHVG¶RLVHDX[RXDUULYpHVPDVVLYHVSURYHQDQWG¶XQGpUDQJHPHQW   1RPEUH+HXUH 9HQDQWGH 3DUWDQWYHUV BBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBBB ««« «««« «««««««««««««««««««««««« ««« «««« «««««««««««««««««««««««« ««« «««« «««««««««««««««««««««««« ««« «««« «««««««««««««««««««««««« ««« «««« ««««««««««««««««««««««««  eYHQWXHOOHVREVHUYDWLRQVG RLVHDX[PDUTXpV ««««««««««««««««««««««««««««««««««««« «««««««««««««««««««««««««««««««««««««   8QJUDQGPHUFLSRXUYRWUHFROODERUDWLRQ %HVRLQG XQFRPSOpPHQWG LQIR"-HDQ

   

www.intercafeproject.net [135] the intercafe field manual

(2) France

Ministère de l'Ecologie et du Développement Durable RECENSEMENT DES DORTOIRS DE GRANDS CORMORANS HIVERNANTS HIVER 2002-2003

Fiche à retourner remplie à : Département : N° Loïc Marion, coordinateur national MNHN-Université de Rennes Laboratoire d'Evolution des Systèmes Commune : Naturels & Modifiés, Campus Beaulieu Code postal : 35042 RENNES cedex tel. 02 23 23 61 44 fax 02 23 23 51 38 Nom du dortoir, lieu-dit :

Coordonnées géographiques (IGN) Organisme recenseur :

Date du comptage : Nom du coordinateur départemental ou régional : Heures du comptage :

Nom de l'observateur :

adresse de l'observateur :

RECENSEMENT DU Nombre d'oiseaux présents au dortoir au mois de : DORTOIR Octobre Novembre Décembre Janvier Février Mars Recensement de 2000-01 (rappel) Evolution intermédiaire (si connue) 2001-02

Recensement de 2002-03

Biotope et supports du dortoir :

Autres renseignements éventuels :

RAPPEL : seuls les dortoirs nocturnes doivent être recensés, à l'aube ou au crépuscule. Le coordinateur départemental ou régional doit envoyer chaque fiche (ou sa copie pour le réseau CSP) au coordinateur national (L. Marion), avec une carte de synthèse (photocopie d'une carte IGN) localisant tous les dortoirs recensés

[136] the intercafe field manual

(3) Italy(3) Italy

CENSIMENTO NAZIONALE CORMORANI 2000 - 2001

INFORMAZIONI SUL DORMITORIO Regione: Zona umida (1): Codice INFS: Provincia - Comune: Località dormitorio (2): Dormitorio censito anche in passato? si no Tipologia dormitorio (3): Data del rilevamento (4):

TIPO DI CENSIMENTO (5) Censimento eseguito dopo il tramonto al dormitorio ore: Censimento eseguito contando uccelli in volo verso il dormitorio dalle ore: alle ore: Censimento eseguito contando uccelli in volo dal dormitorio dalle ore: alle ore: Censimento diurno in zone di alimentazione ore: Altro (specificare) ore: Punto di rilevamento: Distanza rilevatori-dormitorio: Numero totale individui: Campione classi d'età (6): Numero ind. controllati: Adulti: Non Adulti: Nome dei rilevatori: Indirizzo di almeno un rilevatore:

E-mail: Annotazioni:

(1) Nome della zona umida visitata (possibilmente come indicato nel "Elenco delle zone umide italiane e loro suddivisione in unità di rilevamento dell'avifauna acquatica. "; INFS, Documenti Tecnici 17). (2) Toponimo conosciuto per l'area interessata dal dormitorio. (3) Alberi=1; Scogli=2; Pali e strutture per la mitilicoltura=3; Altro=4. (4) I tre periodi utili per i censimenti sono tra il 1 e il 7 dicembre, tra il 10 e il 20 gennaio e tra il 1 e il 7 marzo. (5) A parte casi particolari, dovrà essere utilizzata la tecnica del conteggio diretto del dormitorio dopo il tramonto. Può essere utile (talvolta necessario) contare i cormorani in arrivo al dormitorio nelle ore (almeno 2) precedenti il tramonto oppure quelli in partenza nelle prime ore del mattino. Nel primo caso sarebbe importante avere una stima precisa del numero di cormorani presenti al dormitorio prima dell'inizio del conteggio. E' importante che tutti i dormitori di una stessa zona vengano censiti contemporaneamente, compilando una scheda per dormitorio. (6) Per quanto riguarda il rapporto giovani/adulti è necessario controllare un buon numero di individui (>100) scegliendo dei campioni di cormorani in punti diversi del dormitorio. E' possibile, infatti, che giovani e immaturi si raggruppino in zone particolari, ad esempio all'esterno del roost (adulti = petto nero; non adulti = petto bianco, biancastro o bruno).

www.intercafeproject.net [137] the intercafe field manual

(4) European level — WI- Cormorant Research Group 2003 pan-European winter count.

:HWODQGV,QWHUQDWLRQDO&RUPRUDQW5HVHDUFK*URXS

1DWLRQDOFRRUGLQDWRUVSOHDVHFROOHFWIRUPVLQ\RXUFRXQWU\DQG VHQGQDWLRQDOUHVXOWVDQGPDSLQGLFDWLQJWKHORFDWLRQRIURRVW VLWHVWR(XURSHDQFRRUGLQDWRUV SRVWDODGGUHVVVHH LQIRUPDWLRQOHDIOHW  1RUWK:HVW(8/RLF0DULRQ/RLF0DULRQ#XQLYUHQQHVIU &HQWUDO(85RVHPDULH3DU]*ROOQHUSDU]#PDLOERNXDFDW

3DQ(XURSHDQ&RUPRUDQW0LGZLQWHU&HQVXV -DQXDU\

3DUW$$IILOLDWLRQDQGORFDWLRQ &28175< QDPHRISURYLQFH 'DWHRIFRXQWLQJ

1DPHRIUHVSRQGHQWORFDOFRXQWHU FRXQWLQJWLPHIURPXQWLO

/RFDWLRQQDPHRIURRVWVLWHDGGDOVRQDPHRIZDWHUERG\ULYHUHWF

JHRJUDSKLFDOFRRUGLQDWHVRIURRVWVLWHRUQHDUHVWWRZQYLOODJH

W\SHRIZDWHUERG\ SOHDVHVSHFLI\PDULQHFRDVWULYHUODNHSRQG 

W\SHRIURRVWVLWH SOHDVHVSHFLI\WUHHVEDUHJURXQGSROHLQWKHZDWHULVODQGV SODWIRUPVDUWLILFLDOVWUXFWXUHVRWKHUV

ORFDOFOLPDWLFFRQGLWLRQV ZLQGLFHFRYHUUDLQVQRZIDOOIRJJRRGYLHZ

3DUW%FRXQWLQJGDWD WLPH QXPEHUDFWLYLW\RIELUGVUHPDUNV

7RWDO ILQDO QXPEHU XVHEDFNVLGHIRUIXUWKHUQRWLFHV RIFRUPRUDQWVSUHVHQW DGGLWLRQDOFRXQWLQJUHVXOWVDQG REVHUYDWLRQV

$JHUDWLR DGXOW MXYHQLOLPPDWXUH QXPEHU 1 HVWLPDWH 

DFFXUDF\RIFRXQWLQJGDWDUHFRUG   

:HWODQGV,QWHUQDWLRQDO&RUPRUDQW5HVHDUFK*URXS

(8URRVWBIRUPSDJH [138] the intercafe field manual

3DUW%DGGLWLRQDOFRXQWLQJGDWD WLPH QXPEHU DFWLYLW\RIELUGVUHPDUNV

7RWDO ILQDO QXPEHURI FRUPRUDQWVSUHVHQW HQGRIREVHUYDWLRQWLPH

$GGLWLRQDOFRPPHQWV

3DUW&IRUPUHWXUQ 3OHDVHUHWXUQIRUPWR 1DPHRI1DWLRQDO&RRUGLQDWRU

$GGUHVVRI1DWLRQDO&RRUGLQDWRU

7+$1.<289(5<08&+)25<285+(/3$1'&223(5$7,21

(8URRVWBIRUPSDJH

www.intercafeproject.net [139] the intercafe field manual

13 APPENDIX THREE: WORK GROUP 1 MEMBERSHIP

The INTERCAFE Work Group of INTERCAFE, the participants other Work Groups also made 1 met and undertook work at listed below attended some or presentations and contributions to each of the stakeholder meetings all of the Group’s meetings and Work Group 1 meetings, but are not and during the between-meeting contributed greatly to them. named individually here. periods. Over the four-year span INTERCAFE participants from

Name Affiliation and country 1 Mennobart van Eerden (WG1 Institute for Inland Water Management and Waste Water Treatment (RIZA), Netherlands Co-chair) 2 Stef van Rijn Institute for Inland Water Management and Waste Water Treatment (RIZA), (WG1 Co-chair) Netherlands 3 Stefano Volponi Instituto Nazionale Fauna Selvatica, Italy (WG1 Co-chair) 4 Zeef Arad Institute of Technology – Technion, Israel 5 Dariborka Barjaktarov Natural History Museum, Belgrade, Serbia 6 Janis Baumanis† Institute of Biology, Latvia 7 Thomas Bregnballe National Environment Research Institute, Denmark 8 Szymon Bzoma Sea Fisheries Institute, Gdynia, Poland 9 Henri Engström University of Uppsala, Sweden 10 Manfred Enstipp Centre for Ecological and Physiological Energetics, Strasbourg, France 11 Marijan Govedic Centre for Cartography of Fauna and Flora, Ljubljana, Slovenia 12 Reinhard Haunschmid Federal Agency for Water Management, Institute for Water Ecology Fisheries and Lake Research, Austria 13 Mikael Kilpi ARONIA Environment, Åbo Akademi University & Sydväst Polytech, Finland 14 Emmanuil Koutrakis Fisheries Research Institute, 15 Vilju Lillileht Estonian Agricultural University, Tartu, Estonia 16 Svein-Håkon Lorentsen Norwegian Institute for Nature Research (NINA), Norway 17 Loïc Marion University of Rennes, France 18 Karlis Millers Institute of Biology, Latvia 19 Ivailo Nikolov BALKANI Wildlife Society, Sofia 20 Jean-Yves Paquet Central Ornithologique Aves, Belgium 21 Josef Ridzon Society for the Protection of Birds in Slovakia, Bratislava, Slovakia 22 Josef Trauttmansdorff Otto Koenig Institute, Stockerau, Austria 23 Catarina Vinagre University of Lisbon, Portugal 24 Ian Winfield Centre for Ecology & Hydrology, Lancaster, UK

[140] the intercafe field manual

COST–the acronym for European Cooperation in implementation and light management of the research Science and Technology- is the oldest and widest initiatives) are the main characteristics of COST. European intergovernmental network for cooperation in research. Established by the Ministerial Conference in As precursor of advanced multidisciplinary research November 1971, COST is presently used by the scientific COST has a very important role for the realisation communities of 35 European countries to cooperate in of the European Research Area (ERA) anticipating common research projects supported by national funds. and complementing the activities of the Framework Programmes, constituting a ‘bridge’ towards the The funds provided by COST — less than 1% of scientific communities of emerging countries, the total value of the projects–support the COST increasing the mobility of researchers across Europe cooperation networks (COST Actions) through which, and fostering the establishment of ‘Networks of with EUR 30 million per year, more than 30,000 Excellence’ in many key scientific domains such as: European scientists are involved in research having a Biomedicine and Molecular Biosciences; Food and total value which exceeds EUR 2 billion per year. This Agriculture; Forests, their Products and Services; is the financial worth of the European added value Materials, Physical and Nanosciences; Chemistry and which COST achieves. Molecular Sciences and Technologies; Earth System Science and Environmental Management; Information A ‘bottom up approach’ (the initiative of launching and Communication Technologies; Transport and a COST Action comes from the European scientists Urban Development; Individuals, Societies, Cultures themselves), ‘à la carte participation’ (only countries and Health. It covers basic and more applied research interested in the Action participate), ‘equality of and also addresses issues of pre-normative nature or of access’ (participation is open also to the scientific societal importance. communities of countries not belonging to the European Union) and ‘flexible structure’ (easy Web: http://www.cost.eu

www.intercafeproject.net [141] the intercafe field manual

© COST Office, 2012 Legal notice by COST Office No permission to reproduce or utilise the contents of this book Neither the COST Office nor by any means is necessary, any person acting on its behalf other than in the case of images, is responsible for the use which diagrams or other material from might be made of the information other copyright holders. In such contained in this publication. The cases, permission of the copyright COST Office is not responsible for holders is required. This book the external websites referred to in may be cited as Carss, D N, Parz- this publication. Gollner, R and Trauttmansdorff, J. The INTERCAFE Field Manual: research methods for Cormorants, fishes and the interactions between them. COST Action 635 Final Report II. ISBN 978-1-906698-08-9

ESF provides the COST Office through an EC contract

COST is supported by the EU RTD Framework programme

Published by: NERC Centre for Ecology & Hydrology on behalf of COST

Short Title: COST Action 635 Final Report II

Year of Publication: 2012

ISBN 978-1-906698-08-9

[142] www.intercafeproject.net