Studies on the Ecology & Behaviour of British Shrews

Studies on the ecology & behaviour of British shrews Churchfield, Jane Sara

The copyright of this thesis rests with the author and no quotation from it or information derived from it may be published without the prior written consent of the author

For additional information about this publication click this link. http://qmro.qmul.ac.uk/jspui/handle/123456789/1383

Information about this research object was correct at the time of download; we occasionally make corrections to records, please therefore check the published record when citing. For more information contact [email protected] -1—

STUDIES ON THE ECOLOGY & BEHAVIOUR OF BRITISH SHREWS

JA!IE SARA CHURCHFIEII), B.Sc. (LONDON)

Westfield College, University of London

A thesis submitted for the

Degree 'f Doctor of Philosophy

April 1979 ABSTRACT

All five species of British shrews (Sorex araneus, S. minut, Neomys fodiens, Crocidura suaveolens and C. russula were studied with the emphasis being placed on the commoner species.

The population dynamics and seasonal fluctuations in numbers of S. araneus and S. ininutus were investigated. A seasonal cycle of captures of S. araneus was demonstrated, with peaks of occurrence in summer, low numbers in winter and a re-emergence of high numbers in spring. Closer study indicated a great mortality of old adults and. juveniles in autumn which commenced before the onset of harsh weather conditions, but overwintering survival of remaining shrews was high. Home ranges and activity of S. araneus appeared to be reduced in winttc.

A study of food availability aid diet of S. araneus, S. ininutus and N. fodiens showed major prey items to include adult coleopterans, insect larvae, araneids, isopods and lumbrlcids which occurred In large numbers throughout the year; no decrease in numbers or biomass of prey was found to account for the decrease in body weight of shrews in autumn and winter aid their apparent decline in numbers.

Food consumption of shrews ranged from 4 of the body weight daily for C. suaveolens to i6 for S. minutus, but was not directly related to body weight within a species. Conaurnption by S. araneus was reduced at low temperatures.

Studies of fat storage by wild. shrews showed no great seasonal differences, although captive shrews ac.cumirulated fat in warm conditions.

Studies on the foraging and burrowing behaviour of S. araneu3 showed that they are generally poor burrowers but that they are able to recover insect pupae buried up to 120mm deep in soil.

It is sugges ted. that overwintering shrews adopt a more subterranean existence, spending longer periods in the nest to conserve body heat aid less time foraging. Mortality due to increased. predation in autumn, aid reduced activity on the ground surface probably account for low numbers of captures in winter. -3..

CONTENTS Page

Abstract 2

Table of Contents 3 Acknowledgements

Chapter I: Introduction 10

Chapter 2: Field Studies on the Population Ecology of Shrews 18 Section A: Introduction 19 Section B: Description of the Stx1y Areas 21 Part I : The Wilderness & Monks Wood 21 Part II: Alice Holt Forest 23 Part III: Water-cress beds 24

Section C Trapping Methods 26 General Methods 26

Part I : The Wilderness & Monks Wood. 3].

Part II: Alice Holt Forest :33

Part III: Water-cress beds 33

Section D: Results & Discussion 3.5

Part I : The Wilderness & Monks Wood 3.5

(1) Numbers of Shrews 35

a) The Wilderness 3.5 b) Monks Wood 38

(ii) Weight distribution 38 (iii) Input of Sorex araneus 40 (iv) Survival of S. araneus 41

a) Survival between trapping periods 41

b) Fate of monthly cohorts 43

(v) Life expectancy of S. araneus 43 (vi) Seasonal weight changes, survival & life expectancy of S. minutus & Necuys fodiens

(vii) Home range & movement of shrews 45 (viii) Population estimation 51

-4-

Part II: Alice Holt Forest • 55

(I) Numbers of shrews -- 5$ (ii) Sex ratio 56 (iii) Weight distribution 57 a) S. araneus .57 b) S. mi.nutus from Alice Holt Forest & The Wilderness 58 (iv) Onset of maturity 58

(v) Input, survival & life expectancy 60 Part Ills Water-cress beds 61 - Section Es Summy 63

Chapter 3: Factors affecting the Population Fluctuation of Shrews (5 Section A: Food. availability & diet of shrews 66

Part I: Diet 66

(i) Introduction 66

(ii) Method 68 (iii) Results & discussion 68 a) S. araneus 69

b) S. minutus 73

c) N. fodiens 73

Part II: Estimation of food availability 7L.

(i) Introduction 711.

(ii)Methods 7.5

(iii)Results & discussion 78

a) Vegetation/soil cores 79

b) Pitfall trapping 81

Part III:The relationship between diet & food availability - 83 Genera], discussion

Section B: Climatic Conditions 90

(k.) Introduction 90

(ii) Methods 90

(iii) Results & discussion 9]. Section C: Parasites of shrews 96

(I) Introduction 96 (ii) Methods 97 (iii) Results 97 a) Seasonl occurrence of parasites 99 b) Incidence of infestation in different age classes 100 of shrews

c) Infestation of different sexes of shrews 101 (iv) Discussion 102 Section D: Summary 105

Chapter kz Bloenergetics Studies on Shrewa 109 Section A: Food consumption & assimilation 110 Part I : Introduction 110 Part II; Methods 112 (I) Maintenance of shrews in captivity 112 (ii) Metabolic cages 113

Part III: Results 116 (i) Food consumption by shrews in staudard bins & in metabolic cages 116 (ii) Sexual differences in food consumpbion 1i6 (iii) Differences in food consumption & azslmilation according to species & body weight. 117 (iv) Food consumption & assimilation at different 119 Part IV: Discussion temperatures 119 Section B: Calorific & Water contents of shrews & their Diet 125

Part I : Introduction 325 Part II: Methods 125 Part III: Results 126

Part IV: Discussion 126

Section C: Production & seasonal weight changes 133

Part I : Introduction 133

Part II: Methods 134

Part III: Results & Discussion '34 -6-

(1) Production 134

a) Litter size & the size of pre-natal embryos 134

b) Post-natal growth & development of young shrews 136

c) Food consumption by nursing females 137

d) Re-location of nesting sites 138

e) Retrieval of the young, & fostering of Crocidura 138 suaveolens young, by S. araneus

f) Survival of young 139

(ii) Seasonal weight changes 140

Section D: Seasonal water & fat contents 1LI.$

Part I: Introduction 145

Part II: Methods 145

(1) Ether-shaking technique i46

(ii) Soxhlet method 147

Part III: Results 148

(1) Comparison of the two methods of fat extraction 148

(ii) Wild, shrews 149

(iii) Captive shrews 15].

Part IV: Discussion 152

Section E: Studies on the activity of S. araneus 156

Part I: Introduction i6

Part II: Method i6

Part III: Results 157

Part IV: Discussion 159

Section F: Summary 161

Char4er 5: Studies on the foragin& burrowing behaviour of S. araneus 165 Section A: Introduction 166 Section B: Methods 169

Part I: The experimental pens 169

Part II: Maintenance of shrews 169 -7-

Part liii The apparatus & experimental design 170 (1) Controls 171 (ii) Predation at different soil depths 171 (iii) Predation at diffeent prey densities 172 (iv) Varnished pupae 172 (v) Burrowing activity

Section C: Results l. Controls Part I: 17k Part II: Predation of buried pupae by shrews 174

(i) Predation at different soil depths 17k

(ii) Predation at different prey densities 176

(iii) The effects of weather, season & proximity of food points on predation levels

(iv) Varnished pupae 177 (v) Burrowing activity 178 Section D: Discussion

Section E: Summary 187

çpter 6: Final Discussion 188

Ref erences 199

Append Ic 214

a -U-

ACKNOWLE!)CEMENTS

My thanks are due to Dr. J. Griffith for his supervision of this research. I would also like to thank Professor J.E. Webb, in whose department the work was carried out, for his help ani encouragement. I am especially indebted to Dr. J. Gurnell for his assistance in collecting shrews and for much useful discussion in the preparation of this thesis.

I would like to convey my thanks and appreciation to all the members of the Zoology Department, Westfield College, too numerous to mention by name, for their unfailing assistance and forbearance over the last four years.

I wish to thank the officers of the Nature Conservancy Council and the staff of the Institute of Terrestrial Ecology Experimental

Station at Monks Wood, particularly Professor K. Nellanby, Mr. C. Shaw,

Dr. J. Mason and Mr. J. Woodward, for permission to work in Monks Wood

National Nature Reserve and neighbouring areas. Thanks are also due to the owners and workers of the cress beds on which I trapped.

Ms. Amelia Falk deserves many thanks for her patient and diligent typing of this thesis.

Finally, I would like to express special thanks to my parents who showed an unfailing interest in the work fr and offered assistance and encouragement when it was needed most.

The research for this thesis was carried out during the tenure of the 'Mary Scharliel Award', a postgraduate studentship administered by the University of London, and for whose patronage I am most grateful. -9-

A.E.C. & J.V.C. - 10 -

CHAPTER 1: INTRODUCTION - 11

Shrews are small insectivorous mammals commonly found in woodlands, grasslands and hedgerows. They are represented in Britain by five species: on the mainland and some of the offshore islands these species include Sorex araneus (the common shrew), S. mi.nutus (the pygmy shrew) and Neoinys fodiens (the water shrew).

S. araneus and S. minutus are morphologically very similar, both possessing a brown pelage, but may be distinguished by their size: S. ntinutus weighs up to only kg while S. araneus is considerably larger with a body weight of up to 1kg. N. fodiens is the largest

British species, weighing up to l8g and is characterised by its pelage which is black on the dorsa]. surface and usually white on the belly. It also possesses a cringe of stiff hairs on each foot and on the tail. It is the only aquatic species of shrew in Britain aM is usually found beside swift-running streams and rivers where it feeds on small freshwater invertebrates and vertebrates as well as terrestrial prey. The other two British shrews, Crocidura suaveolens and C. russula, are similar in size and colour to S. araneus but do not occur in mainland Britain: they are found in the Channel Islands, and C. suaveolens is found in the Scilly Isles. Both species are characterised by the teeth which lack the red tips of the other species, by long scattered hairs on the tail, and by their strong scent.

Shrews are amongst the commonest of British mammals but are not as numerous as small rodents, and this, cottp)ed with their solitary habits, causes them to be more elusive and less frequently caught. Additional problems have arisen in studyIng shrews through the difficulty of catching them alive and of maintaining them in the laboratod, where they have proved difficult to breed and keep for any length of time. Consequently, there have been comparatively few studies of shrews in the field or the laboratory, arid our - 12 - knowledge of the details of their biolor, particularly the seasonal aspeta, is comparatively Small.

One of the most interesting features of the ecolo' of shrews is their seasonal mortality. Amongst the first to report his observations was Adams (1910, 1913) who wrote about the habits of the common shrew, Sorex araneu, and the pygmy shrew, S. ininutus, in Britain,with particular reference to the rapid decline in their numbers in autumn and winter: the so called 'autumnal epidemic'.

He remarked upon the large number of dead animals he encountered. in autumn and it is noticeable that shrews can frequently be seen, or more often heard by their high pitched squeaking, during the summer months, but there is little sign of them in winter. The pioneering field studies by Crowcroft (195L a), Shillito,(1960),.

MichielBen (1966), aM Rudge (1965), aM the later study by

Pernetta (1973) on Sorex araneus and S • minutus substantiated. the observations by Adams on the seasonal population fluctuations of these shrews by providing numerical data based upon numbers of captures. They revealed high numbers of captures in summer, a drastic decline in autumn, very low numbers in winter, and a sudden increase in springb Although they hypothesised on the possible reasons for such fluctuations, little information was forthcoming on their true nature or cause • Workers in Poland aM. Czechoulovakia also provided information on the population fluctuations of shrews, and a wealth of laboratory studies contributed to our knowledge of seasonal morphological and physiological changes undergone by these animals. But despite this, the seasonal fluctuations in captures of shrews, with particular reference to their overwintering numbers, have yet to be explained.. This prompted the present study on the population dynamics of shrews and the possible factors, such as food availability, - 13 climatic conditions and parasitism, which might contribute to their seasonal fluctuations in numbers.

One of the difficulties of interpreting results of field studies on the population density of shrews pertains to the nature of sampling, and this is particularly so in the case of studies in Britain to date where mainly live-trapping has been carried out • There has been a tendency to equate captures with population density on the assumption that the numbers of captures are truly representative of the population, and this could be a major source of error contributing to misleading results. Live- trapping relies upon the behaviour of the animal to explore, enter and be captured in a trap. Not on'y does this involve the behaviour of an individual to a strange object within its home range but it also encompasses the whole pattern of behaviour of an individual within its habitat, such as its level of activity during trapping periods, and seasonal changes in activity or behaviour. Differences in activity and levels of 'curiosity' to 'new' objects may also vary between individuals. All these factors affect the trappability of small mammals (An&rze3ewoet al, 1967;

Andrzejelvsi(Jet al 1971; Gurnell, 1978) and it is unlikely that live trapping techniques represent truly random sampling irethods • Little is known of the factors affecting the trappability of shrews, particularly on a seasonal basis, and these features have rarely been considered in population studies to date. Part of the work presented in this thesis attempts to remedy this.

Another aspect of shrew biology which has given rise to great I interest is their metabolic rate. Shrews have often been quoted as having voracious appetites and. a metabolic rate far exceeding that of - 111w - other small mammals (for example, Pearson 1947, 1948; Harrison Matthews, 1972). This has given rise to numerous studies on food and oxygen consumption, but with equivocal findings. For example, Pearson (1947), Gebczynski (1965, 1969) and Gebczynska and Gebczynski (1965) maintain that shrews have a high metabolic rate while Morrison, Ryser and Dawe (1959), Hawkins, Jewell and Toalinson (1960) and Hawkins & Jewell (1962) found shrews to have a metabolic rate no higher than that of other small mam,iis. This prompted an investigation of the metabolic rate of shrews in the present study by estimating food, consumption and assimilation in accordance with the energy budget equation following (Petrusew-icz &

Macfadyen, 1970; Grodzinski, Klekowski & Duncan, 1975) s-

Assimilation Consumption - (Faeces ^ Urine)

A large number of British shrews of different species under controlled conditions were used in an attempt to clarify the anomaly of their metabolic rate.

The studies on the diet of shrews in the wild, and food consumption in the laboratory, lead. to an investigation of another important aspect of energy budgets, namely the calorific value of the food eaten and how this is equated with the calorific value of the body to be maintained. It is not known whether there is any correlation between diet and. calorific and. water contents of the prey eaten by snrews, that is, whether any selection for food. values occurs, or whether diet is purely a function of prey abundance and palatability.

A further parameter in the stu&y of energetics is production, shown in the following energy budget equation (Petrusewicz & Macfadyen, l97O Crodzinski et al, 1975):- _ 15 -

Assimilation - Respiration + Production

where production refers to the ener' channelled into growth of the

individual, and of new individuals through reproduction. Production of new individuals is difficult to assess in a wild population of live

shrews beyond an estimate of the numbers of juveniles which are successfully weaned and enter the trappable population. Estimates of litter size can be made from autopsies of pregnant females, but the

success of litters prior to weaning remains unknown, and there is little information available apart from the observations of the growth

of young S. araneus by Dehnel (1952) and Crowcroft (1954a). Yet another feature long associated, with the shrew and iminortalised. by

Shakespeare is its befligerant and quarrelsome nature, especially towaths other members of its own kind. How this behaviour is modified. during the rearing of young, and at what age agonistic behaviour first appears is unknown. As a result of such questions it was decided to undertake an investigation of the production and rearing of young, and associated behaviour.

Studies of the growth of wild shrews by such workers as Shillito

(1963b), Mezhzherin (196t,.) and. Michielsen (1966) revealed seasonal changes in body weights, with negative production in winter, manifested. by a marked winter decrease in weight. Despite much speculation, little is known of the conditions which produce these winter weight losses. This prompted a study of the seasonal, weight changes of captive shrews under various conditions, and. also of the way these weight changes are manifested in wild shrews. Seasonal changes in water content are quoted by Myrcha (1969) as being instrumental in the weight loss by shrews, but the significance of this seems unclear. More relevant could be changes in fat co'itent, and the storage of fat - 16 - could have considerable implications in the overwintering success of shrews. A study of seasonal fat and water contents was therefore carried out.

Associated with the overwintering of shrews and their trappability is seasonal activity. Shrews are not known to hibernate but Tudin (1962) does mention it as a possibility in the U.S.S.R.

Activity is also an important consideration in the study of energy budgets, since changes in activity affect metabolic rate in terms of food and oxygen consumption (Petrusewicz & Macfadyen, 1970; Schmidt—

Nielsen, 1972). Our knowledge of this aspect of shrew behaviour is very scanty, although the patterns of daily activity, both in the laboratory and in the wild, have been described. in some detail. An indication of activity and seasonal movements was forthcoming roa field trapping in the present study, but direct information had. to be obtained by laboratory studies on captive shrews,

The major part of a shrew's active period is probably spent searching for food amongst vegetation and soil, and so an important aspect of shrew ecology is their foraging and burrowing ability. There is confusion about the way in which shrews locate food, and there is no information about the effect of density or vertical distribution at prey upon foraging behaviour and ability. The fact that shrews exist with such voracious appetites is evidence of their successful foraging strategy, but just how effective this is remains unknown, and prompted the present investigation,

In conclusion, this thesis presents a detailed study of the population dynamics and seasonal fluctuations in numbers of British shrews, and by field and laboratory investigations attempts to explain the nature of the fluctuations and their causes, with particular - 17 - emphasis on the overwintering of shrews. Special attention is paid to the common shrew (Sorex araneu) but the other British species, including the pygmy shrew (S.mlriutus), the water shrew (Neomys fodiens) and the Scilly and Channel IslaM white-toothed shrews (Crocidura suaveolens and Crocidura russula) are considered. Each chapter introduces and discusses the literature relevant to the subject matter of that chapter in the light of results obtained. In addition to a population study this thesis also presents a study of the shrew as an individual and some aspects of its attendant biology, which is perhaps more relevant considering that it is essentially a solitary creature -18-.

CHAPTER 2

tELD STUDIES ON TEE POPDLkTION ECOLOGY OF SHBLWS - 19 -

SECTION A: IRODUCTION

There have been several field stud.ies to investigate the population ecology of shrews in Britain (Crowcroft,1954ShiUito, 1960; Badge, 1965; Pernetta l977a;Grainger & Fairley, 1978) but there is still comparatiyely little detailed knowledge of their population dynm1cB.

Seasonal population fluctuations of captures have been demonstrated but there has been little attempt to correlate these with actual population density of shrews, their input and survival rates, life-. expectancy, seasonal movements and activity, or with other factors which might affect such fluctuations. Im an ettempt to rectify this, the present study was undertaken.

Two major field projects were carried out to study the population dynamics of shrews and to investigate the factors affecting their seasonal fluctuations in numbers. The three different areas selected for study were (1) scru.b-grassland known as The Wilderness and (2) deciduous woodland in Monks Wood, both areas being located at the Monks Wood National Nature Reserve near Huntingdon, Cambridgeshire, and (3) mixed deciduous woodland. in Alice Holt Forest near Farnham, Hampshire. In areas (1) and (2) a programme of live-trapping was carried out in the two different habitat types whereas in area (3) shrews were studied by removal-trapping to provide specimens for autopay, which was carried out in conjunction 'with a study of the rodents by Dr. 3. Gurnell.

In The Wilderness and in Monks Wood, rex errneu and S. minutus were studied between January 1976 and. April 1978 by regular live-trapping, and. Capture-Mark-Recapture teclmicpies. Information was collected on seasonal numbers, weight distributions, input, survival, home range and - 20 - movement. Information on seasonal numbers and weiit distributions of shrews was collected in Alice Holt Forest from May 1975 to March

1978, and. all dead specimens were taken for autopsy to provide details of age-structure and. sex-ratio of the population, onset of sexual. maturity and. internal parasites.

The diet of shrews and the seasonal availability of food resources was studied. in The Wilderness; the levels of infestation by end.oparasites of S. araneus and S. minutus were recorded in Alice Holt Forest and. supplemented by any shrews which died in traps in The Wilderness and

Monks Wood; and climatologica.l data was collected from The Wilderness and Alice Holt Forest to investigate the possible factors affecting the seasonal fluctuations in numbers of shrews. These aspects of shrew ecology will be dealt with in greater detail in the following chapter.

In addition to these major field projects, trapping was carried. out on a short-term basis in two other areas in an effort to collect additional information on the water-shrew, Neornrs fodien. Although this species was very occasionally trapped. in The Wilderness, Monks Wood. and Alice Holt Forest study areas, captures were so few that little information on its ecology was forthcoming. Trapping expeditions were made to commercial water-cress beds in Buckinghamshire and Hampshire during 1975 and 1977 where this species was successfully located, and. information on its babitat preference, population numbers and. diet collected. - 21 -

SECTION B: DESCRIPTION OP THE STTIDY AREAS

PART I: THE WILDERNESS & MONKS WOOD

The main area selected for study was undisturbed. scrubland. of approximately 1.5 hectares in size situated beside the Institute of

Terrestrial Ecology Experimental Station at Monks Wood National Nature

Reserve, and known as The Wilderness. The reserve is located near

Runtingdon, Cambridgeshire, and is described by Hodge et al.

The Wilderness lies on heavy Oxford clay arid chalky boulder cla7 and the site is level. Consequently, water—logging is not uncommon,. particularly in winter months. It is bounded on one side by Monks Wood.. itself, a large area of mixed deciduous woodland, another side is bounded by a road, the third side by a regularly mown field with grass less than 150 mm high, and the fourth side by the Institute of Terrestrial Ecology Research Station buildings and experimental plots.

A map of the study area with the main vegetational areas ma&ed in is given in Fig. 2-1, and a photograph in Fig. 2-2..

A list of the major species of plants identified in The Wilderness is given in Table 2-1 • Once a pasture, the area had been left un—managed. for many years, and had developed a dense covering of vegetation, dominated by grasses (Deschampsia flex.osa and Calama-ostis eie.1oa), bramble (Iubus fruticosus), wild roses (Rosa canina) and willow—herbs

(Epilob iurn sp.) interspersed with young oak trees (Quercus robur) and thickets of hawthorn (Crataegus monogyna), blackthorn (Prinus spinosa) and dogwood (Swida sang nnea). Abundant and. varied ground cover was available at all times of year and. especially in summer. The availability of cover was important not only for the small mammals, but .0 : . .02 -• - .0 i - bE - a o 0 z Ui C., — o g o U.' C - - - 0 -J oE - - a o - / 2,g 0 '-' in

00 . I

S 0 0 U. cJ. U-

0 • II

S 0 I

/ '7/

•• ,/:: ...y

Pt..._ 4 •: ,..i ..

1. .

4 j ; T;l1:. , - 1? : 4•• - 9 . .". ;.. •. .' ,' .. :' /''

I . '. :t.4 -\4;; :: I •'-• 71 1,4 : r'I - 1 . 1.4 1;i 6.'• .

\ 4, L

Fig. 2 - 2: The Wilderness Study area Trees & Shrubs

Betula pendula J. inflexus Crataegus monogyna Lotus corniculata Fraxinus excelsior Potentilla anserina Prunus spinosa P. sterilis uercus robur Prunella vulgaris Rosa canina Pulicaria dys enterica Rubus fru. ticosus Ranunculus acris Salix cinerea Runiex acetosella. Swida aangtzinea B. crispus Senecio 3acobaea Grasses S. vu.igaris Sinapis arvensis Agrostis stolonifera Trifolium pratense Calamagrostis epigejos T. repene Deschanipsia flexuosa Urtica d.ioica Holcus lanatus Poa pratensis

Other Ground Flora

.Axiagaflis aruensis Anthri.scus sylvestris Arcaea lutetiana - C entaurium erythraea C. puichellurn Circium app. Dipsacus pilosus Epilobium parviflorum E. roseum Eu.phrasia nemorosa Filipendala ulmaria Fragaria vesca Juricus effuss

Table 2-1: Plant species identified from The Wilderness

t - 22 - also for a large variety of invertebrates. A list of the small mammals trapped in the study area is included in Table 2-3. In addition to the small mammals trapped, other larger mrnm1s were known to be present, and were occasionally sighted. These included rabbits (Oryctolagus cwuculus), foxes (Vulpes vul pes), muntjac deer (Muntiacus reevesi) and. iqoles (Talpa europaea). Avian predators included. tawny owls (Strix aluco) and kestrels (Falco tinnunculus)..

The Wilderness study area was extended to include the area of Monks

Wood adjacent to it, and. a small trapping grid was established there

(see Fig. 2-1). This area was a long narrow strip of approximately 0.5

hectares in sIze and approximately 40m wide and. 20Cm long. It consists

of mature deciduous woodland with oak (Quercus robur) predominating, and

other trees including ash (Fraxinus excelsior) and. elm (ulnus carpinifolia). Hawthorn (Crataegus monoma), blackthorn (Prunus soinosa),

hazel (Corylus avellna) and. honeysuckle (Lonicera periclymenum) are common shrubs. Ground vegetation is dominated by dog's mercury (MercuriaJ is perennis), bramble (Rubus fruticosus), bluebell (Endymion

non-scriptus), wood anemone (Anemone nemorose) and. lesser celand.ine (Ranimculua ficaria). The amount of cover varied throughout the year.

In Spring and Summer abundant ground. cover was evident, when dog' a mercury and. bluebell reached their peak of growth. Once these plants

had died. off, ground cover was sparse and consisted. mostly of moss and

leaf-litter.

Fewer species of small mammal were encountered. in Monks Wood. than

in The Wilderness (see Table 2-3). The larger mammals and predatory birds found in The Wilderness were also evident in Monks Wood.. - 23 -

PART II: ALICE HOLT F0RES!r

The study area in Alice Holt Forest consisted of approximately 0.72 hectares, surrounded by a border strip 5m wide, of mixed deciduous woodland situated within the Forestry Commission Compartment 32, near

Alice Holt Research Station, near Farnham, Hampshire. The study area U lies upon Gault Clay at approrimately lOOm above sea level, andjwefl-

drained. The study area is bounded on all sides by deciduous woodland.

A list of the plant species identified from Compartment 32, im which the study area was located, is given in Table 2-2. The domirnt species of trees were oak, hazel, sweet chestnut (Castanea sativa), beech (agus sylvatica), yew (Taxus baccata) and holly (flex aciuifolium). The ground cover was dominated by dog's mercury, bramble, honeysuckle, bracken (Pterid.ium acrnilinum), ivy (Hed.era helix), nettle (Urtica di o ica), and wood sorrel (Oxalis acetosella), and during spring and early summer only, bluebell and. primrose (Primula vulgri). During autumn and winter when the main cover died off, grass tussocks provided the most cover with species such as Aostis app., Pestuca ovina, Descham psia app., Dactyla g].omerata. and Brachy-po&ium sylvaticum. More ground cover was available in this study area than in the Monks Wood. study area where grasses were not common, and. the ground cover in winter extremely sparse.

Small mammals cauit in the area are included in Table 2-3, of which the commonest were the wood Inouse (Apodemusylvaticus) and. the bank vole

(Clethrionomys gloreolus). Of the larger mammals, roe deer (Capreolus capreolus), were common and predators included weasels (Mustela rivalis), foxes and. tawny owls. Trees & Shrubs Other Ground Flora

A. caxnpestre Ajuga reptans A. pseudoplatanus Anemone nemorosa Betula pendu].a Arum niaculatum Castanea sativa Bellis perermis Corylus aveUana Blechnum spicant Crataegus monogyna Chamaenerion angus tifohium Pagus sylvatica Digitalis purpurea Fra.xinus excelsior Dryopteris filix-mas hex aquifolium Eidynion non.-scriptuø Lonicera perio].ymenum Euphorbia amygdaloides Pru.uus spinosa Filaria verua uercus robux Glechoma hederacea Rosa canina Hedera helix Rubus fra ticosus Juncus effusus Salix caprea Luzu.la pilosa Sorbus aucuparia L. ey]xatioa. Taxus baccata Mercurialis perennis Grasses Oxalis acetosella Plantago lanceolata Agrostis etolonifera P. major A. tenu.ta Primula viilgaris Brachypod.ium sylvaticum Prunella vulgaris Bromus ep. Pterth.um q.u1uinum Dactylis glomerata Ruinex acetosa Deschaznpsia cespitosa It. acetosella D. flexuosa Sonchus arvensis Festuca oviria Urtica diolca Holcus lanatus Viola riviniana if. mollis Lolium pereime Poa annua P. pratense P. trivialis

Table 2-2: Plant species identified. from Compartment 32, Alice Holt Forest SPECIES STUDY AREA

The Wilderness Monks Wood. Alice Holt Forest Water—Cress Be

Sorex araneus x x x x

5. minutus x x x x

Neomys fodiens x x x x

Apodemus sylvat 1 CUS x x x x

A. flavicollis x x

Micromys minu.tus x

Cl ethri onomys glareolus x x x x

Microtus agrestis x x

Mustela nivalis x x

Table 2-3: Small Mazmals caught in the different study areas

(A cross indicates presence). - lli -

PART III: WATER-CRESS BEDS

Four different commercial water-cress beds were selected for study. These included one bed. on the River Chess at Chenies, Buckinghamshire and three sites on the River Aire in Eampshire, between Bishop Sutton,

Bighton and Airesford, approximately 11 miles north-east of Winchester.

Each set of cress beds has a total area of 0.5 - 2.0 hectares, which is divided up into small rectangular sections approximately 20m wide aM up to lOOm long by low concrete walls about 300 mm high. Within each of these sections, or beds, the cress is cultivated. on a substrate of gravel and silt with fresh water flowing through it. The cress grows throughout the year, reaching a height of appro'rlmately 250mm before it is cut, the peak harvesting time occurring in summer, and continuing until December to a lesser extent. A typical cress bed. is illustrated. in Fig 2-3, a photograph of the beds taken in summer.

Between groups of 2 - 4 beds lie grass-covereJ. banks along which wheelbarrows or small tractors can be driven. These banks reach a height of approximately 0.5 m above the water surface, and generally elope steeply towards the water. They are almost completely surrounded by water, and form the main habitat of the water shrews. The banks provide some cover in the form of grasses (mostly Pemtuca oviria and. Dactylis glomerata) and buttercup (Ranunculus rep ens) but thi a is regularly mown and. so seldom reaches a height of more than 250mm. Additional cover is provided by old cress uprooted from the beds and left on the banks to rot.

The area surrounding the whole set of cress beds usually comprises a border up to 3m wide of grasses and. buttercups, which is also kept mown. The cress beds are constructed on a river-system, usually in calcQreous or limestone areas. The river generally flows down one side of the beds, bordered by grassy banks, and from it water is diverted and/or pumped

-;;• t1.

.'..-!\--

Fig. 2 - 3: Watercress beds

14 ,__ ___

.:: _____

Fig. 2 - 4: Watercress beds with traps in situ (4) - 25 -

into the cresa beda.

Compared. with the scrb1and. and woodland. study areas described. previously, the water—cress beds offer very little cover. However, small mammals are numerous and. the species trapped. in the cress bed. study areas are listed in Table 2-3.

?4ajor predators of small mammals in and. around the cress beds - include domeotic oats, and. probably foxes, owls and kestrels. - 26 -

SECTION C: MPHODS

GENERAL METhODS

Longworth live-traps, described by Chitty & Kempson (1949) were used to catch shrews, two traps being placed at each trap point. AU the traps were provided with hay bedding and. whole oats were placed in the nest box for any rodents accidentally caught. No bait was used. for shrews for several reasons:-

1. During preliminary trapping excursions shrews were not found. to be attracted. to baits and numbers of captures were not found. to increase with the use of baits such as minced beef, liver or blowfly pupae.

Shrews in laboratory studies appeared. to enter 'unbaited. traps 'out of curiosity: 2. The survival time of shrews in traps where bait was provided. was not found to be greater than those in unbaited traps. It appears that the stress of confinement contributes to the death of shrews in traps rather than the lack of food. Shrews which had died. in traps and were subsequently autopsied. were frequently found to have haemorrhages in the stomach wali, which in man is a sign of stress.

3. The use of baits was tune-consuming and inconvenient. Fresh meat rapidly decomposed in wazn conditions, fouling the traps and possibly discouraging animals from entering them. In cold weather it froze solid, making it inedible.

4. The use of bait complicated. the analysis of fascel pellets collected. from trapped. shrews.

Following each sampling period the traps were thorou&fly cleaned. and fresh bedding provided. The traps were checked for signs of damage and. repaired where necessary. Treadle settings were also checked, the traps - 27 -

being adjusted to catch animals of 2g weight or more in The Wilderness and. Monks Wood in order to catch pyy shrews, and 6g weight or more in Alice Holt Forest, where rodents were being studied. in addition to shrews.

All small mammals caught in the study areas were recorded. and the

date, time of capture and. reference number of the trap point noted.

Special attention was given to the shrews. All shrews were weighed, the colour of the pelage and reproductive condition recorded.

Weighing in the field was carried out by placing the animal in a small polythene bag attached to a Pesola spring balance weighing from o - 30g.

In the field, weighings were made to the nearest 4g. In the laboratory, live and. dead shrews were weighed to two decimal places using an OertLing balance.

The condition of the pelage of common shrews was recorded as being brown or dark-brown, for this is relevant in age-determinations of shrews

in the field. As previous studies have shown (Shillito, 1960, l963., Michielson, 1966) and results of the present study confirm, juvenile S. araneus are brown in colour during the first few months of life, bua

in September/October they moult arid prod.u.ce a thicker winter coat of a dark-brown colour. Moulting occurs again in the spring, but the dark

colour of the pelage is maintained in mature animals. In the field,

three age-classes can be recognised by a combination of pelage colour,

body wsight and. date of capture. Table 2-4Asummarises these conditions.

Common shrew juveniles are not weaned and. do not leave the nest until reaching a body weight of approximately 6g, but the sub-adults may lose

weight in winter and weigh as little as 5.5g, which accounts for the weight classes in the table, derived. from results of the present study. - 28

Month of Capture Pelage Colour Body Weit

Juveniles May - September Brown 6.0 - 7.5

Sub-adults October - March Dark-Brown 5.5 - 8.0

Adults April - September Dark-Brown 8.0 - 13.0

Table 2-4* • Age Classes of S. araneus

The sexing of live shrews in the field, based. on external characters alone, is difficult although it is sometimes possible to identify mature artn 1 s. The testes of the males are abdominal but increase to such a size as to be visible when mature, as large swellings under the skin anterior to the anus. Mature females can be distinguished. by their nipples and. abdominal swelling when preiant or lactating. The body weight of shrews is a guide to their maturity. Tmrnature shrews (both juveniles and. sub-adults) cannot be sexed accurately because in males the testes and. penis are too small to be visible, and in females the nipples are small and hidden amongst dense fur. Palpation of the abdomen sometimes causes the penis of males to evert, but this is most easily done in mature mimals. There are no other external characteristics to distinguish the sexes.

Each live shrew caught in The Wilderness and. in Monks Wood was marked before being released at the point of capture. Marking was carried. out by- toe-clipping and. different combinations of toes clipped. indicated. particular individuals • The termiiii digit was renoved. from toes of the fore and bind feet with a pair of sharp scissors. Shrews, unlike rodents, do not have well-developed. fleshy pads on. the toes, and. clipping of the terminal digits seldom resulted. in bleeding, Toe-clipping did. not appear to adversely affect the animals: shrews marked. in this way were frequently recaptured, and. the wound was found. to have healed satisfactorily. In - 29 -

addition, several an.imals were observed to forage and feed easily following release after toe—clipping. Shrews are pentadacty]. and a total of twenty toes are available for clipping on each shrew: The minimum number of toes necessary to mark the were clipped in each. case, and never exceeded two toes cut on any individual.

Shrews which had died. in traps, mostly in Alice Holt Forest where removal trapping was carried out, were kept and. frozen to await autopsy.

The following features were observed during the autopsy operations:-

1. Species of shrew.

2. Condition of shrew, healthy or otherwise. Colour of pelage.

3. Body length and tail length.

4. External parasites were identified and counted. Ectoparasites leave the host as soon as it s body cools down following death so that few parasites were found..

5. Sexu.a]. condition: the body was opened. and the sex of the animal determined together with its condition of maturity. The testes of e'z'& male were measured, the maximum length and width of each being determined and the mean of the two testes recorded. The testes were also examined for the presence of spermatozoa. In the females, the total uterus length was recorded. as the sum of the lengths of each uterine horn (see Brambell, l935, and. the numbers and sizes of any embryos present were recorded.

6. Food remains and gut parasites: the whole gut was removed and. cut into sections inclvding the stomach and. oesophague, the duodenum and ileum, and the colon and rectum and examined separately for food franents and. parastes. Food remains were rarely found. in the gut of trapped animals, the gut being evacuated prior to death. Any parasites found were identified. as far as their state of preservation would allow, and counted. - 30 -

7. Other internal parasites: the body was searched internally for

signs of damage and. parasites. Particular attention was paid. to the connective tissue surrounding muscles of the back, neck, sternum and. - limbs where encysted nematodes occurred..

8. Pat and. water content: the remainder of the body was cut into

pieces, dried in an oven at 6000 until constant weiat was obtained.

(which required approximately 3 days), and fat extracted. The method. of

fat extraction will be described in Chapter 4, where laboratory experiments are discussed.. - 31 -

PART I: THE WILDERNESS AIfl) MONKS WOOD STUDY AREAS

In the Wilderness, a study area of ].50m by 10Cm was measured out and 38 trap points established within tius area at approximately 15m intervals. A map of the study area showing the main vegetational areas and the trap points is given in Fig.2-1. The study area covered the whole of the Wilderness including the tall hedges at the edge, with the exception of a small area adjacent to the Research Statipn buildings which tended to be heavily grazed by rabbits.

The Monks Wood study area was adjacent to The Wilderness and. fozned. an extension of The Wilderness trapping grid. Two rows of 10 trap points in each row were established with an interval of 15m between trap points. This study area is also depicted in Fig. 2-1

Traps were set in used. tunnels where possible, amongst grass, willow-herb, dog' s mercury and. brambles. They were mostly hidden from view by the surrounding vegetation and. so were rarely covered with grass or moss to camouflage them. Shrews were found. to enter traps irrespective

of any camouflage.

Each sampling period. consisted of three days, during which trapping

in The Wilderness was carried. out during daylight hours only. Shrews

are active by day and by night, and the difficulties of trapping them

alive at night in rough, almost inpenetrable vegetation, when traps have

to be checked. at frequent intervals to prevent captives from dying,were dispensed with in favour of trapping during daytim€ only. In Summer, traps were opened at 8 a.m. and. closed at 7 p.m., and in winter they were opened at 8a.m. and clöed at dusk. As very few shrews were caught in winter months the difference in number of trap hours between summer and.

winter sampling periods was not deemed to make results incomparable - 3Z

(as will be discussed later). During the night traps were left closed to prevent mice and voles from entering and. fouling them.

The difficulty of catching shrews alive and. the need. to obtain live

animals for the study resulted. in traps being visited. regularly every

2 hours during the trapping period in The Wilderness. This seemed to

be the optimum period for leaving traps unattended: left for longer and.

mortality of shrews increased, particularly of pygmy shrews and. those caught several tines in succession. More frequent visits to the traps

would. result in increased d.isturbance of the habitat and. the creation of mu.d.dy pathways between the trap points which could affect the ranging behaviour of shrews. Very few deaths of shrews were incurred as a result

of this trapping strategy: the mortality rate over the whole trapping programme was only compared. with approximately 70% at Alice Holt Forest where traps were visited only once or twice per 24 hours..

In Monks Wood, captures of shrews were infrequent and. so traps were

left set open for 24 hours each day and examined morning and evening.

A list of the trapping dates for The Wilderness and Monks Wood. is

given in Table 2-4. The trapping programme was commenced. in January 1976 in The Wilderness and extended to include Monks Wood in March 1976. Both study areas continued to be trapped until April 1978. Sampling frequency was increased after the first eight months of the study to

investigate more fully the trends apparent in population fluctuations, and thereafter intersampling periods were relatively constant. However,

when shrews were most numerous, namely in the surnner months, the study

areas were trapped more frequently than in winter when shrews were elusive and rarely entered traps.

TRAPPING DATES (inc1usii) INTERSAMPLING PERIOD (Days)

January 9 - 11 1976 74 March 24 - 26 1976 54 Nay 17-19 1976 71 July 28 - 30 1976 55 September 21 - 23 1976 23 October 14 - 16 1976 28 November 11 - 13 1976 27 December 8 - 10 1976 49 January 26 - 28 1977 36 March 3-5 1977 21 March 24 - 26 1977 42 May 5-7 1977 28 June 2-4 1977 29 July 1-3 1977 28 July 29 - 31 1977 34 September 1 - 3 1977 33 October 4 - 6 1977 36 November 9 •- 11 1977 31 December 10 - 12 1977 74 February 22 - 24 1978 49 April 12 - 14 1978

Table 2-4 List of Trapping Dates in The Wild.eriiess and. Monks Wood. - 33 -

PART II: ALICE HOLT FOREST

A trapping grid was measured out covering an area of 8Cm by 9Cm within 'which trap points were established at 1Cm intervals on the grid., giving a total of 72 trap points at which two traps were placed per point.

Traps were left open for four compiete days and. nights on each sampling occasion and were e y m1 ned once daily. Trapping was carried. out for three years at approximately four-weekly intervals. A list of the trapping dates during the sampling programme is given in Table 2-5.

PART III: WATER-CRESS BEDS

Water-cress beds were sampled on four different occasions for periods of between three and seven days each tine with the prime aim of obtaining live water-shrews for laboratory experiments. Table 2-6 givea a summary of the cress beds sampled, the trapping dates and numbers of

traps employed on each occasion.

Location of Cress Bed Trapping Dates no. Traps

Chen.ies, Bucks. Jan. 7 - 13 1975 40

U II May 2]. - 23 1975 60 Aires ford, Rants. 75 Sept. 30 - Oct. 2 1975 Bishop's Sutton, Rants. 75 Aires ford 50 Bishop's Sutton Aug. 22 - 25 1977 50 Bighton 50

Table 2-6 Trappir on Water-Cress Beds DATES OF TRAPPING PERIOD (Inclusive)

May 2-5 1975 September 8 - 11 1976 May 30 - June 2 1975 October 23 - 26 1976 June 19 - 22 1975 November 20 - 23 1976 July 26 - 29 1975 December 18 - 19 1976 August 16 - 19 1975 January 19 - 22 1971 September 14 - 17 1975 March 4. - 7 1977 September 30 - October 3 1975 April 4 - 7 1977 October 17 - 20 1975 May 12 - 15 1977 October 31 - November 3 1975 June 16 - 19 1977 December 14 - 17 1975 August 1 - 4 1977 January 12 - 15 1976 August 24 - 27 1977 February 27 - March 2 1976 September 29 - October 1 1977 March 18 - 21 1976 October 25 - 29 1977

May 3- 61976 December 1 - 4 1977 June 11 - 14 1976 January 27 - 29 1978 August 5 - 8 1976 March 2 - 5 1978

Table 2-5' List of Trapping Dates at Alice Halt Forest -

Longworth traps were placed singly at ]. - 2m intervals along stretches of grassy banks between and beside the cress beds where openings to the shrews' burrows could be seen. A few traps were placed amongst the cress in the beds. Fig 2-4 is a photograph of a typical. study area and some traps in situ. The traps were left open for 24 hours each day and examined at intervals during the day and. night. AU captures were recorded. and live animals, with the exception of water'-. shrews, were released. The diet of water-shrews was also studied, and is described. in Appendix 3-2. - 35 -

SECTION D: RESULTS & DISCUSSION

PART I: THE WILDERNESS & MONKS WOOD

Population Dynamics. i) Numbers of Shrews a) The Wilderness.

The numbers of juvenile, sub-adult, adult and the percentage of adult S. araneus and S. minutus caught in The Wilderness are given in

Table 2-7 and Fig. 2-5. The total numbers of individual S. araneus and. S. minutus of all age-classes are given in Fig. 2-6. These numbers represent captures in 1.5 hectares, the area of The Wilderness. A single N. fodiens was caught here in May 1976.

Between 40% and lOCP/o of all the S. araneus caught were found in the part of The Wllderrbess biown as Field Edge. This s 'b-area occupied some

30m by 80m (0.24 hectares) and was located adjacent to a mown field (see plan of study area, Fig. 2-1). Between Field Edge and the main part of

The Wilderness was a narrow strip of grass, approxiiiiately 20m wide and. heavily grazed by rabbits, in which no shrews were found.. The numbers of shrews caught in the two sub-areas of The Wilderness are given in

Table 2-8. In all, but four trap periods more shrews were caught in Field Edge than Li the remainder of The Wilderness • The maximum number of S. araneus caught in Field Edge during one trapping period was 20, in

April 1978, whereas only 7 were caught in the remainder of The Wilderness in the same month, but the reason for this is not clear. Vegetation cover over the wnole of The Wilderness was dense throughout the year. Field Edge comprised mainly tussock grass while the remainder of The

Wilderness consisted primarily of willow-herb and bramble. However, another stand of tusoock grass in one corner of The Wilderness did not

S. araneus S. minutus

Total Total % Sub- Sub- adult Adult mdi- Capt- Adults Adult Total Adults adult Capt.- Breedi. vids. ures :Breed- ures ______1-rig JAN 1976 1 1 1 0 MAR 4 6 10 13 60 22MAY 22 38 100 1 1 100 July 7 4 11 15 36 5 1 6 20_ ST 15 3 18 52 17 5 5 0 OCT 7 7 15 0 2NOV 2 3 0

2DEC 2 3 0 JAN 1977 3 3 3 0 1 1 0 MAR 0 0 0 2APR 4 6 10 67 MAY 6 6 10 100 1 1 100

JUN 10 10 16 100 1 1 100

Jul 8 4 12 18 33 2 2 0 AUG 23 3 26 43 12 10 1 11 10

SEPT 29 1 30 70 3 3 3 0 OCT 24 24 38 0 8 8 0 NOV 8 8 2]. 0 1 1 0 DEC 4 4 6 0 PEIB 1978 6 6 8 0 1 1 0 2APR 25 27 52 93 3 3 100

Table 2- 7: Numbers of S. araneus and. S. minutus caught in The Wilderness. 4 0. uz-° E Ow uJ >-0-rI 0

a C,

t) 0 't) 0 N 0 p.. N C, N o SIOflPAIPU 0- z o LU

I 4 i I

cq.

SIOflPJMPU

THE WILDERNESS FIELD EDGE TOTAL % INHABITING MAIN AREA FIELD EDGE

JMUARY 1976 1 - MARCH 10 - MAY 22 - JULY 11 -

SEPTThIBER 10 8 18 44.4

OCTOBER 0 7 7 100 • 0

NOVEMBER 1 1 2 50.0

0 2DECBER 2 100.0

JMIJARY 1977 1 2 3 66.7

0MARCH 0 0

MARCH/APRIL 1 5 6 83.3

3MAY 3 6 50.0

10JT 2 8 80.0

JULY 4 8 12 66.7

JULY/AUGUST - 14 12 26 46.2

30SEPTFBER 13 17 56.7

OCTOBER 12 12 24 50.0

NOVEd1BER 2 6 8 75.0

DECEMBER 1 3 4 75.0

FEBRUARY 1978 3 3 6 50.0

APRIL 7 20 27 74.1

Table 2-8. Numbers of Shrews caught in the different areas of The Wilderness

t - 36 maintain a high number of shrews. Tk dense population of shrews in Field Edge may be attributed to it being a relatively isolated area of dense cover, with a mown field, a grazed strip, a road. and buildings occupying each side.

From Fig. 2-6 it is apparent that there is a seasonal cycle in captures of S. araneus, with peaks of abundance in summer months and a marked decrease in winter, as has already been observed by many workers, for example Adams (1910 , 1913), Mezhzheiun (1960), Shullito (1960),

Bucicner (1969), Randdlph (1975o) and Pernetta (l97l). Numbers of captures increased rapidly during spring to a maximum between May and.

September or October of 20 to 30 m,-tnals per trapping period. In autumn captures declined sharply towards winter, when only small numbers of shrews were trapped per trapping period, suggesting densities of

0 to 8 individuals in 1.5 hectares. These low numbers persisted until the following spring.

Young shrews, characterised by their low body weight and light-. coloured pelage, appeared between May and September. During summer the population structure changed dramatically in favour of these young animals, which, by October, constituted 10(/ of the captures, adults having disappeared, probably as a result of death by senescence or predation. Young snrews overn.ntered as sub-adults after moulting in September/October and replacing the light summer pelage with a thicker, darker-coloured one. Following low captures of individuals during winter, a sudden re-emergence of larger numbers of shrews occurred in spring, many of which had been marked in the previouc aunner or autumn, but were not trapped during the intervening months. By March or April shrews were approaching breeding condition and by May l0C of the population consisted of mature adults who were active on the ground surface - 37 -

searching for mates. By July, the first juveniles had appeared in the

population, following a gestation and. w i"ig period, totalling six weeks.

The decline in numbers of captures in winter may be attributed. to one or more of several factors:-.

1. Death due to starvation, pred.ationor adverse weather cond.itiona.

2. Emigration from the study area. -

3. Reduced trappability caused by a change 'to a more subterranean existence and. a decline in activity on the ground surface during

adverse weather conditions.

Only one specimen of N. fodiens was trapped in The Wilderness

study area during the sampling programme, although four were caught in

Monks Wood, adjacent to The Wilderness, and. information from these will. be included in the diet studies. Very few S. minutus were caught and. so information concerning their population dynamics and seasonal numbers

is sparse. Longworth traps were set to their finest adjustment in order

to catch these shrews but whether the low numbers of captures reflects their actual abundance or their behaviour towards the traps is not known. Captures of S • minttu showed a seasonal cycle .in numbers similar to that

of S. araneus (Fig. 2-6) with peak numbers in July-August. This is also reported by Grainger & Fairley (1978). It appears, superficially at least, that,like aneus, they overwiriter at very low densities

(two or less per three heotares).

The fluctuations of both S. araneus arid S. minutas showed. differences between the two years sampled (see Fig. 2-6). Both species were more numerous in 1977 than in 1976, and peaks of abundance occurred considerably later in 1977, with maximum numbers of S. araneus for 1977 - 38 -

in August-September and October, but in May and September in 1976.

This was noted too by Borowski. and Dehnel (1952) and Aulak (1970).

b) Monks Wood.

The numbers of S. araneus caught in the Monks Wood study area are

given in Pig. 2-7A. A total of only 36 S. araneus were caught between

September 1976 and. April 1978. This may reflect a habitat preference for areas of denser vegetation, provid.ing greater cover, such as was

found in The Wilderness scrubland, but was present only in summer months in Monks Wood, in the foiin of dog's mercury.

Despite the paucity of S. araneus captures, a seasonal. cycle occurred, with low numbers in winter and. a peak in summer comprising mostly juveniles.

No S. minutus were found in Monks Wood, but a total of four

N. fodiens were captured between September and. December 1971.

ii) Weiit Distributions

The weight distributions of S. araneu caught in The Wilderness

during each trapping period, are given in Fig. 2-8 where ¶ weight

classes are shown from 5.0 - 13+ g inclusive. Results are not divided into males and females owing to the difficulty of seng shrews reliably in the field, especially immature m,fmals.

In spring the population comprised shrews approaching breeding

condition and. from May onwards the peak weights were found., with adults

reaching 9g and over. Between May and September the weight distribution was large, as both adults and juveniles were found in the population.

The time of greatest biomass was found between July and September when population numbers were high and comprised mostly young, immature shrews weighing 6.5 - 8.0g. In October and November the mean weight of shrews started to decrease, and the overwintering DuDu1at1on consisted en±i1v a I21

U.,

'no

.-c4

p. .t ;

'0 W N

dq, SIOflPIAIPUI Z (6) SSD) 446'M a • r - - I uj8 ______'0 . .< 4d

-I'.- 0,6o u- I--

'00 - I... I'%O z - p.. 0 0 P. 0 (') r% L U, N .1

- 0'. N '-.4

.o2

U-

V) 0

z 0

.— . (• N Th I 0.. -' I— 2 -Lr

QPs-I' 0q'. U- -o •1

0 0' I's •0 I I I I I I 0 0 '0 0 0 0 0 CM P.. 2 - 0 0. '0 - (6) sso 4t16IM - - of immature animals with a low body weight with a mean of 6.lg in December - February.

Weight distributions of S. araneus in Monks Wood showed similar trends, and are presented in Fig. 2-7B. These trends are also suggested for S. minutus caught in The Wilderness, but the results were not conclusive owing to the small number of captures. The maximum weight recorded for this species was for an adult weighing 5.3g in May 1976, and the minimum was 2.5g in January 1977 for a sub-adult. Weight distributions of pygay shrews are ahown in Pig. 2-9. No data was obtained for N. fodiens.

That shrews lose weight during winter can be seen by the weight of sub-adults caught repeatedly from the summer of their birth to the following spring, and these are presented as examples in Pig. 2-10 A & B.

A notable example is S.a. D1O which weighed 6.og when first caught as a juvenile in July 1976, had increased in weight to 8.5g by October, but when recaptured in January 1977 weighed only 5.75g. It had. gained weight to 7.5g in March and by July it was weighing ll.5g and in breeding condition. Other examples of weight loss over winter are E6 which decreased from 8.5g in November to 5.5g in December, and. ff5 which weighed 7.5g in September and only 5.Og in Februar r fcllowi.ng.

The p'nenomenon of weight loss coupled wth a decrease in body dimensions of overwintering sub-adults has been observed by other workers

(Adams, 1913; Brambell, 1935; and. Shillito, 1963hin rita.in 1 Michielsen,

1966 in the Nebherlands, and Kubik, 1951 ; Borowaki and. Dehnel, 1952; and. Mezhzherin, 1964 in Poland). Niethammer (1956) attributed the loss in weight to a shortage of food combined with the physiological strain of moulting, and a drop in external temperatures; increase in food supply plus a rise in temperature permitted. an increase in weight in spring, *

r)o HI __

—o

(b có

—o

II. Er

'to

4 N°

-4

—o a 2

d. (6) scop 446'M — < < .__ 4

z I 0

— -s ,-

U- I -I 0 t z 0 'I U)

o 0 0 C') (S4 (6) 4qBlaM Apog I

O G. x o

'Job

'I)

I-.

(6) 411619M Apo despite the 'strain' of the spring moult. This is contrary to the view of Debnel (1949), who could find no increase in the spring food supply to account for the weight increase of shrews. Pucek (1970) explained the seasonal changes in body weight, particularly of internal organs, by tissue dehydration in autumn and. winter, behind which were 'morpho-physiologicaJ. changes' in endocrine glands and. variations in metabolism. The influence of food supply and weather conditions on the seasonal weight changes of shrews will be discussed later.

iii) Input of S. araneus This refers to the number of new pniTni5 in the population and is indicated by the numbers of previously unmarked captures during each trapping period. It comprises n,inals born within the study area and those which moved into the area between sampling periods, or resident

riimals previously uncaptured. Input figures have been expressed in

Table 2-9 and in Pig. 2-11.

There was a large input of individuals in summer during the breeding season. Jiveniles continued to appear in the population until October, but thereafter input was reduced, although a few unmarked young animals still appeared during the winter of 1976, after breeding had fiiusiied. This was probably due to the prolonged surface activity of young shrews not yet settled in definite home ranges, but which had. not been caught betore, or may represent some immigration. The winter population was small and input was low, most captures having been previously marked and appearing to be resident in the area. In March

April, when shrews were approaching maturity, a high input was indicated: 8 of the population in April 1978 were unmarked adult shrews, mostly mature males. This may be due to two different reasons or a combination of both: i) a number of resident animals which were not captured as young

pap, Qc+cI (D (

(D(D F-' c . I_J '0 \0 - -3 -4 -3

I.-' 0 0 0 0 I-' -3 C 0 0 0 0 F- 0 I- J1 - 0 -

- 02

£' 4 I-" ,_& F-' - Vi - 0 Vi 0 0 0 0000ui00000'M000000 ,' I-.o • S S S . . S t- c+ O O'VO o

N) 1-' 000OOOF-' O'sjiN)000000l0V1 0 02

'd c1- lb (D cl CD r.) O\ Vi Vi - 1-3 0000 00%Ji00 000000 OVi I i 02 • • • . . • • • . c1-P) 0 %iJ 0 0 0 0 0 - 'i i

C-.-

N) F-' F-' I-' F-' F-' I-' 0 0 0 0 l- -4 \0 '.0 O\ N) 0 I-' 0 F-' I-' Vi 0 '.0 '.0 I-' 1-3

-1 . Vi - -4 0\ Vi Vi I-' CD '.0 CD '.0 - Vi Os '.N Vi 0 0 '.. 0 VI 0 0 - '.,.i 0 Os 0 • S S S S • S S S • S • • S S I-' 0 0 0 CD - I-' 0 0 0 '.i 0 '.i VI '.0 . 0

F-a 0 P I I CD - -3 CD F-' I-' '.0 VI -4 CD CD I-' -3 • S S S S S S S • 5 • • • • • - N) - OS - '.) VI CD Vi N) - Vi Vi

0 P I I 0 0 I-' N) )- 0 I-' I 0 I I 0 0 N) 0 0 .___ S rJ) • S S S • • S S • S • S CD -.4 Vi Vi - F-' - 0 Vi - '.0 CD '.0 CD CD CT'. CD Os O5 Cs \_,4 VI CD -4 S '—. .__ %_• .__ %__ , .— - .- —_ -

- '+1. - individuals in autumn, but survived winter, were caught in spring for the first time while they were seaching for mates (ii) immigration from a neighbouring area, probably Monks Wood, or fringe areas of The Wilderness where traps were not laid. Mich.ielsen (1966), too, recorded numbers of previously unmarked S. araneus in spring and.

suggested that these individuals, which were mostly males, had.

'territories' outside the trapping area but had extended their home ranges into the study area in. the seach for mates. In The Wilderness, - new adults continued to enter the population until July or August,

after which the input comprised juveniles only. All the input nmala were in good condition and had body weights 52.m.11ar to those of resident

that is, previously marked, shrews.

iv) Survival of S. araneus. a) Survival between trapping periods The survival of S. araneus juveniles/sub-adult3 and adults froni month to month is shown in Table 2-10 and Fig. 2-12 A & B. This takes into account animals marked in previous trapping periods which were not captured during a following trapping period, but were known to be alive owing to subsequent recaptures at later dates. iuinals which were not recaptured at all may not have died, but simply moved out of the stu.dy area.

Survival tended to be lower in summer than in wintor for all

shrews. Result2 showed a high mortality of adults when breeding had. finished, probably due to old age and relaced catses. During summer and.

autumn, when conditions were warn, shrews were aotive on the ground

surface and the reduced survival of juveniles and sub-adults may represent (1) mortality by increased preLatiori or (2) emigration to a less densely populated area. Survival was at its lowest in. July 1976 H H H '.0 -3 - -3 CD --4 a'. I-$fl CDC 0 -i,1 o ctI I Ct

CD H ro ro CD '.11 O' 0 0 ',i -3 CD '. '.o '. 0 r q ci- R 0

CD &ca

0 0 0 0 0 ro ro . - -. 0 0 0 0 0 H r H 0 0 O E-c CD ci- 0

p & H H H H I.-' OR0 CD H 10 ro CD -4 CD -P. -P. -3 -3 CD '.0 '.0 '0 -P. 10 ,i 0 I-4 SO p -OH ci-':9 0 CD D2 C I-' H I-' to '.1 10 I-' ci- H 10 '1 - 0 ¼11 CD 0 0 - CD '.0 '.0 0 0'. -3 0 - H 0

(DtD '

CD P.

0 0 0 0 ro '.Ji -3 '...n 0 0 : 0 '. '..,i '. 0

- '.0 '.0 '.J1 0'. 0'. -.3 -3 CD 0) 0 '.0 '.J1 P. '.J1 1') H 1'.i 0 0 '.51 '.Jl 0 0 '.51 -4 0) 0 0 0'. 10 0 0 0 .

0' - 1 Jl CD 10 H 00000 O'.0 O'.-3 000000 0'.0 0'.O 4OHO OH-3 I-' . CD

H -3 '.0 '.0 'J '.51 0'. '.4 '.4 '.51 -4 CD 0) 0 '.0 -P. \-N \.$ 1- I'.) 0 0'. '.51 0 0'. -3 -4 -3 CD 0 0 -3 '., '.0 0 0 p o 0'.CD'0000

00 H1-''- J'.j 10 HI-' H I-' H ro H I-' t-. '.1 - 1) 0'. \$ I-' - '.0 CD '.0 '.0 0 '.0 10 1..) 0 I-' CD 4.. D

U) Zcd - UJ vi Q. lu -= vi E -I C- 0 vi >. -0 0

4. ii

4. o 0 o o 0 0 ( - o co .,o • N co N I- o I0.hhmflS % - - and in September and October 1977. Once winter approached $ survival of individuals increased to 90 - 1OC even though the total population represented by- numbers of captures, was small. It decreased slightly in March/April probably as a result of predation and/or stress of shrews active on the ground surface once again with the advance of the breeding season.

Nature adults were only present in the population for a very limited period. By May all were mature and from then onwards their survival - decreased. and. by October all adults had d.isa;peared from the population..

The figure for August 1977 is slightly misleading: 3 adults were known to be alive of which 2 survived until September, givIng a high percentage survival figure. However,none of the September cohort of adults in either year survived to the following month.

To conclude, the percentage survival of shrews (which is not a reflection of population Size) was higher during winter than in summer, most of the reduced population of sub—adults surviving until the breeding season the following year. The survival of shrews through the winter is further emphasised by Fig. 2-13. Here, numbers of marked shrews known to be alive from their subsequent recapture at a later date, but which were not caught during intervening months, are presented. During summer months, most marked pnima1s were recaptured du.rIng each successive sampling period, but large numbers of shrews which were known to be alive failed to be caught during winter months. Either these individuals moved out of the study area temporarily between captures, or they were untrappable due to reduced or modified activity on the ground surface.

t E 0 z *

I9I

be .- r14

o 0 (N 0 I-

pQddoij ou nq 4uacJd SIOnPIAIPUI - 43 -

b) Fate of Monthly Cohorts

The survival of S • araneus throughout the year has also been

expressed in teins of the fate of monthly cohorts where a cohort is the number of marked animals known to be alive at each trapping period.

The total numbers of survivors from month to month, that is, the fate of each monthly cohort, is shown graphically in Fig. 2-14 where the

upper limits of the vertical bars represent the minimum number of iimals known to be alive each month. Their fate is shown by their gradual decrease in numbers in successive months.

Mature shrews which overwintered as sub-adults disappeared

(presumably they died) from the population in July, August and September

following breeding. Juveniles decreased in number in autumn but the

survivors remained throughout the winter to maturity the following spring.

The percentage survival of monthly cohorts, given in Table 2-il, shows that shrews had a very limited life as mature individuals, the May and June cohorts surviving for only 3 months. The maximum life-span of S. araneus is shown to be only approximately 13 months, indicating a very rapid turnover of individuals in the population. v) Life-expectancy of S. ararieus1 The life-expectancy, from the time of first capture, of S. araneus in The Wilderness is shown in Table 2-12 where the number of captures from each monthly cohort recaptured in successive months is calculated and expressed as the percentage of the original capture at time zero

(the first capture time). Again, it can be seen that shrews do not survive more than 13 months.

t The mean percentage survival values for each successive month of capture, derived from Table 2-]. 1, are depicted in Fig. 2-15. During the first two months of life survival was low, agreeing with the findings I

cs1 0 CO 0 C') ('1 N- N _ pouid 6uiddoi qDoo GAPID aq 04 UMOUI OU UIW H ¼0 -a - - H c o '.3

H CD Fs3

H 8o

Q\000 -a I-I 8 OO

0 H3 H 0 Q 800° '.3 I-i OON)0000 0 '-.3 H I-' rn o o o 0 N) 0 0 0 0 H 000-a000°0 CD cn

H o CD - - - - OOCDcDcoON)0000

CD I-' 0-a -a as I-a. 0 CD 0 \s -a -a 0 as 0 0 0 H pa CD

CD 80000 to I-' 0 N) IS) N) N) N) IS) H ci .,o 0 as N) IS) 0 .J1 IS) IS) 0 H 0 0 0 '-' -' -a '.4-4 H o-a'.000000000000°0 as

g-a-000000000000°-a as

H 0 OSji'J100000000000000%J1J1'JI 0 '.0

H o0 IS.)Y1¼M rs)..M H-s00000000000O 0 I-a J) SJI N N.) - 0 r. N.) H -3 ' 0 0 0 0 0 0 0 0 0 0 0 0 0

I-' 0\O O)- N) N)¼11 0 N)U1 CD CDO00000000000 0 '.0 H 0 -i as N) IS) '..N 0 -i H '..n N) '..,i 0 0 0 0 0 0 0 0 0 0 0 0 0

—J —I H H I.-' I-.' H H I-' H I-' I-' H H H I-' H H H I-' (DO 00000 C) 00000000000000 0 X 0 00000000000000000000 H a I I \O Vi Vi — O\ - —3 I CD 0 O . I I I I H CD '.0 F) rs) ¼$ - —3 C' - CD \O 0 0 —3 0 Pa s—' Vi - %.O Vi ,1 Vi 0'. I3 -1 .. 0) '..O Vi 0 M I r H S-fl — '0 r - 0 I 0 0 rs 0 .O 0 0 ti

CD S-fl I CDL S-ltO'J100r5)CD0 I --P- I I I - Vi H —3 0 r CD 0 I'.) 0 CD

- s I - I I'.) ¼ 0 0 0 — - 0 I - n 0 0 a'. — CD — ¼ Vi CD —3 0 & S-4 I 15) I V4 0 0 0 S-fl 0'. -3 0 l - 0 —3 H a) '-'a 0-3 CD 0 0

_ '-, I IS) I 0 0 0 IV - 0'. —3 0 - 0 0 0 Vi S-N Vi30 0

"I 15)1 '.JI O00IS)0\S-,l $001 -3 a 0\ IS.) S-N I'.) ID & r') ') I 01 000 IVIV00O 0 0) a) '.N I 01 000IVH I 00 \0 —3 0 0'0 0 '5) 0 I 0 0 0 0 0 H S-N 0 0 0 H H HO 0 CI OOO000 I 0 H 0 H '5) OI 000015)000 H 0 P.) 0 ii 0 01 0000001 H C'. 0 ,- 01 000000 H

H 0100000 H lCD Vi I H ID 01 0000 H H CDIV 0'. 0 O0O0D

B 1000 1-' 0 a) i

000 H CD Pa '-0 ml H 00 ts (DId 0 '5.) co 0 IV I', 15) (0 '5) - I-

ci- CD IL C') I-

a.. •

4- a.

0 0 0 E 4-

C') U)

o 0 0 0 0 0 o '0 N jOAIAJAS % - L44

of Pernetta (1977t, with only 5C% of shrews surviving this period. Thereafter, survival was steady and. between 4/o and 5 survived to

be six months of age, which took them through the major part of the winter to February or March. With the coming of spring survival decreased steeply until by August most adults had. died.

To summarise, 5C9 of shrews died within the first two months of life,

40 - 5 survived to become six months of age, and 20 - 3C$ of the original number survived to breed. These results are similar to those of

Shillito (1960) who found 6cP/ of S. araneus died before the autumn and

3C of the original population survived the winter. Michielsen (1966) found. up to 6 of the population survived between July and January.

No shrews in the present study survived longer than 13 months. vi) Seasonal weight changes, survival & life-expectancy of S. nanutus arid N. fodiens Data on the seasonal weight changes, survi.val and. life-expectancy of both S. minutus and N. fodiens is not available from this study owing to the low numbers of captu.res. No rece.ptures of N. fdiens were made, and only two S. minutus were recaptured. S.m. A]. was first captured in

August 1977 weighing 4.Og as a juvenile, and was subsequently recaptured the following April prior to breeding and. still weighing 4.0g. S.w. BlO, captured in Octohcr 1977 weighing 3.Og was also recaptured in April 1978, arid had increased. to 4.Og in weight. This represents a period of 7 - 9 months before these shrews commenced bree&uig, and. so S. minutus probab].r has a life-span similar to that of S • araneus. No. S. minutus were recaptured during winter and so no inforrnaticri on i..eight loss was forthcoming but Grainger & Fairley (1978) observed a decrease in. weight, also aocompnied. by skull regression, in overwintering S. minutus in

Ireland. Pernetta (1977a) found peak mortality of S. minutus occurred between two and four months of age, rather later than in S. araneus. - "'.7 -

He suggested this was caused either by death of resident individuals or to dispersal occurring later in life in S. minutus.

vii) Home Range & Movement of Shrews

Burt (1943) defined home range as the area traversed by the individual in its normal activities of food gathering, mating and cari.ng for its young. Occasional aounieys out of this area were excluded. More recently, Jewell (1966) has proposed the term 'life- time range' which refers to the total area with which an animal has become familiar, including seasonal home ranges, excursions for mating

and. routes of movement. The problems associated. with the determination of home ranges from trapping data has been reviewed by Kikkawa (1964).. The analysis of home range depends for its accuracy upon the number of

rocaptures made: according to most authorities a minimum of ten recaptures of an individual are necessary before a home range can.

reasonably be calculated (Hayne, 1950; Stickel 1954; Kikkawa, 1964;

F].owerdew, 1976; Hawes, 1977). The home ranges estimated in. this atudy are ba9ed. upon movement recorded by Longworth trapping which

relies for its success upon the behaviour of the nfma1. This iiay not

provide a true indication of movement since the mechanism of trapping,

the spacing of the traps arid the reaction of individuals to the traps may greatly influence patterns of movement. The presence of a trap

at a particular site may encourage an animal to enter it and thereafter

return to it, as was the case when some shrews were oauat repeatedly in the same trap each day. Consequently, home range arid pabterns of movement revealed by trapping data should be viewed with caution. The numbers of recaptures made in the present study were not

sufficient in most cases to provide reliable information on the size

of home ranges of shrews, but this has already been presented by

Shillito (1963o),Yiichiels en (1966), Bukn.er (1969) and Pernetta (1977a) for Sorex araneus and. S. nnnut, and by Platt (1976) and Hawes (1977)

for American specios. However, the ranging of shrews in The Wilderness - -

has been investigated by an analysis of movement revealed by Capture-Mark-Recapture data. Table 2-13 provides a summary of trap-revealed movement of juvenile, sub-adult and adult S. araneus

dur3.ng each trapping period. As trapping was carried out according to a standard method each time, the results are comparable fron

month to month. Fig. 2-16A & B present movement of the three age

classes of shrews in terms of mean total distance travelled and.

mean distance travelled between captures wi±hin each trapping period.

Movement was greatest in spring and. summer when adult shrews

were breeding, and. most captures were males, particularly in April

and May when they were presumably searching for mates. In his stud7

of Sorex vagraris and S. obscurtis in British Columbia, Hawes (1977)

found home ranges, and hence movement, increased with the onset of the breeding season, especially in males. In the present study, juveniles were numerous and active on the ground surface .Ln summer but movements between trap points were limited to short distances and. were not as extensive as the forays made by breeding adults. The mean distance

moved between captures was 7m by juveniles and 34in by adulta. The mean total distance moved during a trapping period by juveniles was

greatest in September 1976. This was due primarily to one individual

which travelled considerable distances between captures. If the

figure for this individual is excluded, the mean total distance moved is

reduced from 5Gm to 41m.

This accords with the results of Shillito (l963c who found that some juveniles travelled large distances in the first two months of life

but most of them remained in a small area.

In winter, movement decreased to a minimum in November, December

and February. In February 1978, six recaptures of sub-adults were made and the greatest movement both 'zithin the trapping period and

0 Cl)

H 'J.II H1 -3 -3 £0 0\

N) J1 N) N) ¼ê C.1 C) c1 - 0 0 M .

H H £0 N) N) .jl N)

H Is) 00 N) H H H -4 0\ . • 40 CD CD H Is) H • •H • —.) • CD as II Cl) j H Is) H N) H I-' s.N H • S • —1 N) 5.ft

H H H C H - N) • S • J1 CD Cx) CD

0 H H H • S C_I •o . • 00 CD 0 • cx) '.Jl 0 CD 'do H H 0 Cl) H 0 N) 0 . S • I • 0 -4 0 So 'J \, O\ o • 4J. CD 0 0 0 0 ¼$ 0 o S 5 . • 0 • p. '0 'Ji 0 • S

H co CI (DO H • o — 0 I.i

H H H o o I) . d- o co . - • I-I. p.

H H Is) N) .SN 0 S)1N) - — o • —so -.5-. CD pa S-n N) H CI 00 S.N ..N 0 a) • CD H • 'do Is) 1-' CDci 0 CD SN —1 o •

N) -j - s_n • t SN S-il 0 - I-' N) -4 HP' Jm

Table 2-13: Intra-sampling period movement of S. araneu 5 Juveniles, Sub-Adults & Adults

. - 0. =. E LU C 0,.. 0 0 LU > _i 3 -D 0 EzD *

z 0 U,

0 z 0 U)

, _, ' , ._J , ._J 0 0 '1) c o R Cl C1 - (w) OAOW QDUO4SIG t.7 -

since the previous trapping period, was 15m (inter-sampling period movement is discussed later). Reduced activity in winter in terms of movement on the ground surface is shown in Table 2-14. Eere, both the total distance moved, and the mean distance moved between captu.res, within a trapping period are approximately twice as great in summer as in winter. This indicates that home range may also be smaller in winter than in summer.

These findings agree with those of Shiflito (1963cwho found a mean home range length of 37m in juveniles in summer but ouly 28.4a in overwintering sub-adults. Michielsen (1966) found home range to be larger in sub-adults than in juveniles but home ranges of adults were extended to greater distances than the other age classes. Buckner

(1969), however, found no difference in home ranges from month to month. Total Distance moved within sampling period Mean Distance moved. (m) (in)

March - September 34.7 14.7 October - February 16.9 8.7

Table 2-14

The mean distance travelled 'by juveniles, sub-adults and adults between sampling periods is shown in Table 2-15, where movement has been estimated on the basis of a comparison of centres of activity between successive trap periods. Again, adults tended to move further than other age classes, and. move nnt was greatest in spring and. summer. Sub-adults moved very little between sampling periods during winter.

Shrews generally failed to move far from their original sites of capture, a Table 2-16 shows, indicating that home range remains in approximately the same location throughout life. Most shrews were repeatedly captured at the same trap point month after month, and. many moved between only two or three neighbouring trap points during their life. Hawes (1977), too, found that home range location

tended to remain stationary throughout life. Platt (1976) found.

that the size and location of the home range of Blarina brevicaud.a

in the U.S.A. remained stationary provided, prey density was high, but that it increased in area and often shifted at low prey densities.

That a shrew maintains a home range in the same area throughout its life suggests that conditions for residence there are favourable, but 'nomadic behaviour' becomes more apparent in the breeding season.

Mean distance moved between sampling periods m, Total Juv Sub-Adult Adult Individuals MAY - SEPTEMBER 9.4 (14) 15.0(1) 10

SEPTEMBER-FEBRUARY 4.8(16) 11

FEIBRUA.RY-NAY 13(3) 3

Table 2-15

(Brackets denote number of moves observed)..

Mean Overall intersampling period. No. Moves No. Individua movement (i) Observed.

11 53 16

Table 2-16

Some examples of trap-revealed home ranges are shown in Fig. 2-17

and Pig. 2-18 where they have been mapped by drawing lines connecting the outermost points of capture. Only shrews caught ten or more times

are included. In Fig. 2-17, home ranges cf individuals are shown in terms of the summer juvenile stage, the overwintering sub-adult stage, and. as mature adults in spring and summer. In most cases, home range tended to be larger in juveniles and. adults than in the overwintering sub-adults. Juveniles tended to move between several trap points before a regular, and. usually smaller, home range was

e • • S araneuj 010

• • • p • July 1976 July 1977

• S • S •

. S • • S $arafl_eti E6 Sept 1976Julyl977 ___:__:1:: Lob :

. S S S S

•i ' 5rone 87 • S S S July 1977-Dec. 1977

e [La., I • • S • S S

• . (' _____S.araneus A3 • July 1977-Apr 1978 • • • • •

9- ILab 1 • . • • • J

S orofley A2 July 1977- Apr 1978

S S S • S •

•ci -e-_ • . • S.araneu B2 \ . - . - -- • _L' July1977-Apr 1978

L -. - - - - - LEGEND • Trap point - Juvenile • Sub-adult - Adult Scale 9 1 Fig. 2 -17 Some trap -revealed home ranges of individual S araneus at different metres ages in 'ihe Wildernebs

- . .

C . C C A S S • Soraneus B46 S S S Mar —Sept 1976 --•- -- • . S S - - - a - • ------usual home rang.. • ------distance covered m one day outside usual home rc*ngi

: \ \ ri1 :iiii______

B 4 shrews caught between July 1976 &July 1977.

C C S . . ./1c . 6 shrews caught between C July 1977&Apr 1978 •.. •

•''.::.:.::,, S S LEGEND 1 •Irap point I Lab Scale I ___ J__• 9_1 I I metres

Fig 2-18 Some trap -revealed home ranges of S araneus caught at least 10 times in The Wldernesg. - 1+9 -

established, which was maintained throughout the winter. Once the breeding season approached, movement between trap points increased

once more,

Both Fig. 2-17 and. Fig. 2-18 show that shrews tended. to keep to a discreet area, although it varied in size between individuals and.

occasional sorties were made outside it. For example, in Fig. 2-18 A

Sa. .B46 travelled at least 135m away from its noma]. home range in -

September 1976 before returning to it the sane day, but was otherwise

not caught outside its own area. The size of the home range varied.

from 20 - 75m in length, similar to that found. by Shullito (1963 a.)

which varied from 8-54m in length, with an average of 28-39zn, and Michielsen (1966) of 370 - 630mg

Fig. 2-18B & C show how the home ranges of shrews overlapped..

This was also observed by Shil].ito (1963) and. Buckner (19 69), but

Nich.telsen (1966) found home range to overap in the breed.ing season only. Michielsen suggests that a home range is held. by an individual

throughout life, but that in the breeding season it is extended,

especially by males, to overlap with others. .thickner (1969) found t hat the well-defined home ranges of the autumn and. winter were mostly vacated at the onset of the breeding season, and that males became

nomadic in spring, but this may mean an extension of the home range

rather than its vacation,

No differences in the home range betweec. the sexes was apparent

in this or other studies, except for the .z.ncrease an movement of

males in spring and summer (Michielsen, 1966; Buckner, 1969; Hawes,

1977). .50..

No recaptures of N. fodiens were made, and only two S. minutus

were recaptured in The Wilderness. S.m, BlO moved 6Oxn between

October 1977 and. its recapture in April 1978, and S.m. Al moved only

30 metres between July 1977 and April 1978 during which tine it was captured only twice.

To conclude, the results of the present study suggest that a

discreet home range is maintained by most shrews throughout their - life, which tends to remain stationary in its location but varies in

size according to age and time of year. Movement was reduced in winter, but reached a peak in the breeding season when large numbers

of nomadic males were encountered. Hawes (1977) suggests that individual success in overwintering, measured by the attainment of sexual maturity the following year, is closely correlated with

territorial success. Michielsen (1966) concludes that shrews are- territoriai on the basis of the pattern of distribution of the trap— revealed home ranges which remained stationary, and. in most cases

showed complete exclusion combined with complete adjacency. Croworoft

(1955) suspected that shrews were territorial because captive S. araneus would drive away others from certain areas. Michielsen (1966), therefore, assumed that boundaries of home ranges in the wild are fixed primarily by fighting and so refers to territories as well as home ranges. However, there is still no proof of territoriality in wild. shrews, and this, coupled with the overlapping patterns of movement of S. araneus revealed, in the present study, leads me to refer to home range in preference to territory. The maintenance of a territori, which may be synonymous with home range in the present study, would be of particular importance for females in summer while they are nursing young and requiring an abundant food. supply, and. for all shrews in winter to reduce competition for food and. nesting sites during unfavourable weather conditions, and. so aid individual survival. - 51 -

More work is required to discover the role of the home range of shrews and whether it can be referred to accurately as a territory. viii) Population estimation.

The problems associated with population estimation have been reviewed by Kikkawa (1964), Southwood. (1966) and. Seber (1973). One of the assumptions common to most methods of estimating population density from trapping data is that all individuals in the population have an equal probability of being captured on any sampling occasion. This may be influenced by the trapping procedure itself, or by certai'i characteristics of the population, which affect random sampling, and leads to error in the calculation of population densities.

In this study- no recaptures of N. fodiens were made and. recaptures of S. minutus were not sufficient to use in population estimates using

Capture-Mark-Recapture methods. All population estimates using C.LB. methods are based on S • araneus caught in The Wilderness.

Table 2-17 and Fig. 2-19 give three different estimates of population density throughout the trapping programme. These include (i) estimates calculated by the simple Lincol!1 Index method (Southwood. 1966), (2) numbers of individuals captured. and. (3) the minimum number iown to be alive, or the calendar of captures method (Petrusewicz and.

.And.rezejewski, 1962).

The problem associated with the assurnpion that all individuals are equally catchable is well demonstrated inthe winter estimates of populationdensity using the simple Lincoln Index method, where numbers of individuals caught during each sampling period were used and compared. with numbers caught in the following sampling period. DENSITY PER HECTARE MINIMUM NUMBER NUMBERS OF DATE LINCOLN INDEX KITOUN TO BE INDIVIDUAL S PRESENT JANUARY 1976 0 0.7 0.7

MARCH 36.7 6.7 6.7 MAY 132.0 14.7 14.7 JULY 28.7 8e0 7.3 SEPTEMBER 14.7 12.7 12.0 OCTOBER 4.7 6.7 4.7 NOVEIBER 1.3 6.0 1.3 DECEMBER 1.3 6.0 1.3 JANIJARY 1977 0 5.3 2.0 MARCH 0 5.3 0 APRIL 4.0 6.0 4.0 MAY 4.7 4.7 4.0 JUNE 24.0 7.3 6.7 JULY 28.0 8.7 8.0 AUGUST 28.7 17.3 17.3 SEPTEMBER 28.7 21.3 20.0 OCTOBER 4.0 16.0 16.0 NOVBER 5.3 8.7 5.3 DECEMBER 2.7 8.0 2.7 FEBRUARY 1978 8.7 7.3 4.0 APRIL 18.0

Table 2-17: Pcpu1aton Estimates of S. aianeus in The Wilderness

Based. on Lincoln Index calculations Of C.M.R. data, minimum numbers known to be alive, and numbers of ind.ivid.iials caight per trapping period.. , .-6 E a V C K- V 0 C z_-a woc 0 Z 10

V V

0 - 'I, - - - - - b 0' - - -I .-

IL-

p-I

CM o 0, CM 0 CV) CV) CV) CM I-. jo;,aq id °N -52-

The nature of the sampling programme, with frequent visits to traps each day to release shrews, complicated population estimates. This, coupled with the fact that few individuals were caught during sampling periods at certain times of year (particularly in winter months) lead. to population estimation based on. numbers of captures within sampling periods being abardoned in favour of inter-sampling period estimation.

This presents an additional problem. in providing an accurate time-base for the population density estimated by the Lincoln Index method. Estimates were based on the ratio of the numbers of shrews marked and. released in one trapping period (first sample) and. the total number caught in the following trapping period (second sample) to the number of recaptures appearing in the second sample. Estimates were assigned to the latter or second sample, but to which of the two sampling periods, or to which intervening month, the estimates really pertain remains unknown.

A marked decline in population density, as estimated by the Lincoln Index method, was found in winter and which in January and

March 1977 reached zero. This is incorrect, for it could not permit the sudden increa.3e in numbers the following spring, particularly in view of the other assumption made in C.LR. estimates that there is no change in the pomlation through :mmigration, emigration, natality or mortality between sampling periods. If the Lincoln Index figures are compared witn the n'imbers of known sirvivors over winter, it can be seen that shrews did overwinter in the study area, and. hence the latter estimate represents a more accurate record of the population size, although it, too, is probably underestimated because it assumes that all individuals have been caught.

This anomaly is not so apparent in the summer estimates of population density, when shrews are active on the ground surface - 53 -

and large numbers are encountered.. Here the Lincoln Index estimate probably approaches the true value, for it shows a summer peak in excess of both the captures of individuals and. the number of known survivors, both of which tend to under-. estimate the population size by ssllm' ng that all individuals have been cauit and marked. The large increases in population density in March and May 1976 may be attributed to the low numbers of recaptures sustained. at the begirming of the sampiing - programme, leading to an inflated Lincoln Index estimate.

The problem of population estimation is highligited again, however, between May and. June 1977 when a rapid. increase in estimated numbers is apparent, prior to the emergence of juveniles into the population. Shrews are in breeding condition at this time of year, and. most captures were mature males, including many previotisly unmarked. ones, active on the ground surface searching for mates. An equal proportion of females are present, as the sex ratio is approximately equal, but many of them will be spending most of the time in the nest nursing their young, and so appear infrequently in traps. The resulting low numbers of recaptures greatly increased the Lincoln Index estimates for June 1977, a1thoui juveniles had not yet appeared. in the trap:ped population.

Seasonal changes in the behaviour of shrews, influencing their activity on the ground surface and. their tendency to enter traps, therefore adversely affects any estimation of population density, invalidating it at certain times of year. Consequently, population estimation will not be considered in further detail.

Population estimates from other sources are compared with those of the present study, detemnined. by the criteria mentioned above, in Table 2-18. Great variation in population densities is apparent,

according to habitat, location, time of year and. trapping technique. Population estimates from the present study are lower, particularly in summer, than many of the other studies shown, but most closely

resemble those of Michielsen (1966) and Buclaier (19e9). Densities of S. ininutus, indicated by numbers of captures (0.7 per ha in

winter and 7.3 per ha in summer), also approach those of Mich.telsen (1966) of 4 per ha in winter and. 10 per ha in summer.

Despite the differences between population estimates in the studies mentioned above, a trend of high summer numbers and low overwintering numbers is still apparent. The variation between population estimates, particularly for summer values, do not appear to reflect habitat differences. There is no evidence of a preference for any particular biotope, although it is probable that localised. habitat differences and attributes contribute to the selection by shrews of certain areas for colonisation and. govern the numbers which can be supported there. Local differences were encountered in the present study where certain parts of the study area supported larger numbers of shrews than others, as ha3 already teen mentioned.. If such differences exist within a mere 1.5 hectares of study area and populations differ in size from year to year, as was also found, it is only to be expected that population estimates produce considerable variation between different biotopes, locations and. years of study.

No. of S. araneus per hectare

SUMMER WINTER HABITAT/LOCATION AUThORITY

49 - 62 - Deciduous Woodland., England Crowcroft (1954a)

69 17 - 27 " " " Shillito (1960)

- 2 - 7 " " " Buckner (1969)

6 - 52 Various forest biotopes, - Aulak (1967) Poland.

18 12 - 13 Dune Scrub, Netherlands Michielsen (1966)

67 Gras si and, France Yalden (1974)

42 Gras si and, England. I,

8 5 Grassland, England. Pernetta (l976

7 - 21* 5 - 9* Scrub-grassland, England Present Study

24 - 29' I, if

Table 2-18 Population Estimates of S. araneus

* minimum numbers known to be alive

' Lincoln index - 55 -

PART II: ALICE HOLT 1DREST

1) Numbers of Shrews.

Captures of S. ararieus arid. S. rninutus in Alice Holt Forest between April/May 1975 and April 1978 are shown in Fig. 2-20. No N. fodlens were caught during the trapping programme. Apparent from Fig. 2-20 is the marked seasonal cycle in numbers

of captures of S. araneus, with peak numbers in July, August and. - September each year, and low numbers during winter months. Also noticeable is the magriitude of the seasonal cycle in shrew numbers

which differed from year to year. Numbers of S. araneus in 1977 were very low compared. with the two previous years • Both the

seasonal cycle in numbers and. its difference in maguitade between

years show similar trends to those already described for The Wilderness and. Monks Wood sites. These differences may be

attributed to (a) food availability or (b) environmental conda.tion

which will be discussed later.

Captures of S. minutus were very few at Alice Bolt Forest,

a total of 20 being caught for the whole sampling programme. A seasonal cycle in numbers, as has been shown in The Wilderness and Monks Wood, is not apparent, although winter captures tended to be

extremely low. The low numbers of captures may be a reflection of the habitat preference shown by S. minutus for grasslands rather

than woodlands (Pernetta, 1973) and. it has already been remarked

that this species was apparently absent from Monks Wood although it was caught in moderate numbers in The Wilderness, adjacent to it.

Of the shrews caught at Alice Bolt Forest, most died in the

traps and were subsequently subjeoted. to an autopsy eymination, I I

0 Q 0 0 0 0 C, (4 IOnPMPU1 -56-. which provided data on body weight, sex, reproductive condition and parasites. One hundred and seventy-one S. araneus were autopsied, and a summary of the information obtained is presented. in Table 2-19, which shows the numbers of males and females, juvenile, sub-adult and adult, caught during each trap period. Total numbers of juveniles, sub-adults and. adults are shown graphically in Pig. 2-211 together with the percentage found. to be in breeding condition (Fig. 2-21B). A similar trend in population dynamics to that found. in The Wilderness is demonstrated, with large numbers of juveniles present in the summer, which overwintered as sub-adults. Maturity was reached between March and Nay whereupon breeding commenced, and. the major component of the population then comprised adults, the first juveniles appearing in Nay. As summer progressed the numbers of adults innni shed until by September/

October the total population consisted of sub-adults born during the summer,

The low numbers of captures of S. minutus provided little material for autopsy. Consequently, autopsy data obtained from

Alice Holt Forest has been combined with that for The Wilderness, where a number of shrews died. in the traps. The numbers of juveniles, sub-adults and adults ey nnned are given in Table 2-20.

A total of 24 5. minutus were autopsied., which was not sufficient to give conclusive results of a seasonal nature. Only one adult male was found, arid no pregnant females, the rema..uider of the sample comprising juveniles or sub-adults. ii) Sex Ratio.

The numbers of male and. female S. a:raneus caught and autopsied. are shown in Fig. 2-22, which shows the proportions of the two sexes to be approximately equal and agrees with Slcaren (1973) p Rather

MILES LES

MONTH Total sub.- mature sub- mature TR.A.PPED autopsied juv adults adults adults adults

10J1JLY 1975 34 0 7 15 0 2 AUGUST 9 8 0 1 0 0 0 SEPTEMBER 13 6 0 4 3 0 0 0OCTOBER 12 8 1 0 2 1 0DECEMBER 7 0 3 0 4 0 MAY 1976 5 o 0 4 1 0 0 AUGUST 18 6 0 4 5 0 3 12SEPTEMBER 19 0 0 7 0 0 0OCTOBER 7 0 4 0 3 0 NOVEMBER 3 o 3 0 0 0 0 DEMBER 19 o 8 0 0 11 0 JAMIJARY 1977 6 0 4 0 0 2 0 0MAY 3 0 2 0 0 1 AUGUST 4 2 0 0 1 0 1 SEPTEMBER -5 2 0 0 2 0 1 OCTOBER I 0 0 0 1 0 0

MARCH 1978 1 0 0 0 0 0 1

APRIL 5 0 0 5 0 0 0

Table 2-19: Numbers of S. araneus autopsied from Alice Holt Forest

t VI

0 4 0

—.

N U

0 •1

0 U VI U f 0 f 0

CM CM

S)OflpIAIpU MALES F24ALF

!1ONTH Total sub- mature sub- mature j'v' TRAPPED Autopsied adults adults adu.lta adults

JTJNE].975 1 1

JULY 4 1 1. 2

AUGUST 1 1

OCBER 3 2 1

MARCH 1976 1 1.

MAY 1 1

JtUJY 2 1 1

SEPTEIVtBER 1 1

NOVEXBER 2 1 1

DECEMBER 1 1

JA1UARY 1977 1 1

JULY 1 1

AUGUST 3 1 2

SEPTEMBER 2 1 1

Table 2-20: Numbers of S. minutus autopsied. from Alice Bolt Forest and. The Wilderness r's

0 0 N SOflPIAPUI -57-

more males than females tended to be caught, which is probably

a reflection of their activity on the ground surface rather than

their unequal ratio in the population or their behaviour towards

the traps. An approximately equal sex ration is also reported by Middleton (1931) and. Pucek (1959). During spring and summer males are active on the ground surface searching for mates and.

defending territories, and. so nu.ght be erpected to be caught more frequently than nursing females. The datd. on S. rninutus

were insufficient to provide any conc]usions about the sex ratio of this species.

iii) Weight Distribution. a) S. araneus• The weight distributions of S. araneus autopsied from Alice Holt Forest are shown in Pig. 2-23, where males and females are

combined into different weight classes. Ta 1 le 2-21 gives the mean weights for the different age classes of shrews.

MALE ALE Juv. S.A 1 Adult Juv. S.A. Adult (Pregnant)

Mean Weight g. 6.48 6.11 9.05 6.43 5.91 8.60 10.10

S .D. 0.54 0.58 0.68 0.55 0.57 1.67 1.69 Table 2-21

There was no significant difference between the weigrits of male

and female juvenile and. sub-adult shrews, but mature males were

found to be heavier than mature females, except when the latter were pregnant. The loss in weight of overwantring shrews was

found to be a feature of the Alice Holt Forest population as well as that of the Wilderness, for sub-adults, (both males and. females), were lighter in weight than the juveniles caught in the summer soon after their weaning, - I,)O ,._ 0

—o

OO '0 fl -p%;

0 0. r.. N

0 ,., 0

-'06 .- p._ 0.

0. r - -i .- - - g 0;. I E-- '0 c og4 2 •0 oI (6) scop 4'I6M - 58-•

As with The Wilderness shrews, the widest weight distribution was found in summer, when the population comprised both adults and juveniles.

Shrews reached a peak weight between May and September when breeding occurred. Thereafter, weights decreased, a mean weight of 5.7kg being recorded for January 1977, until by April/May the maturing shrews had. gained weight. b) S.minutus from Alice Holt Forest and The Wilderness.

Because of the low ntunbers of captures of S. nunutus in Alice Holt

Forest, the results of weight distribution have been combined with those of S minutus from The Wilderness, although the resu.fts are not strictly comparable because the accuracy of weighing differed in each study area

(see p.27), the results are presented in Fig.2-2 I . The body weights of

S minutus ranged from 2.36 to 3.50g. Mature, but not pregnant, females weighed 2.59g, and the single mature male captured weighed 3.5g. Few shrews were caught during winter, and so a decrease in weight in winter was not conclusive, unlike those found in The Wilderness. Apart from one immature animal found weighing 3.5g in January 1977, peak weight was achieved at maturity in May. 1then the autopsy results from Alice Holt

Forest and The Wilderness are combined into one year (Fig. 2-251, weight is found to decrease in November and December, (confirming the findings of

Grainger & Fairley, 1978), to increase in April and May, and have its widest distribution in July and August, when adults and juveniles were present.

iv) Onset of Maturity

Following a decrease in body weight of shrews in winter, when no growth in body size or reproductive organs was recorded, there was a period of very rapid growth to maturity of males and females simultaneously

in Spring (April/May). Fig.2-26 shows the growth of the testes (in terms of mean length) and the uterus (total length) of a cohort of S. araneus born during Summer at Alice Holt Forest. No results were obtained for the

transition period •1 o Iv, VI - oV) .1. C4

-.,

—c'.'0 1

U-

0 —o II -o

C4..'3 z 0 0 g U)

'0 (.4 9

U- U 0 cv z

C., 0

_.0 ('.1°

ir.s! N0 —' a — 'V 'V - '.4 (:4 I I I I

z (6) ssop 4q6'M . 7.- oIfl

0 U-

OPs

— N N NO IL 0—n

p.' (I) no

— -no

— n o

NO -)

- 0- I I I I I U', Ul Q o N 0o 7 (6) SSD 4161aM (ww)J4u1 s n i ci4fl UW o o 0 0 o o o 0 0 o U, 9 ol '4

ZELU '4

0 t 'e- ' f9 46 :1

S.. I1

LL 0. .i.erl

—) ___ 0

U S4 o

b rZ o 0 0 o 0 0 o '0 (54

(ww) 6u;-i SI 4.L UDW - 59

(February/March) between immaturity and breeding condition, but

Brambell (1935) found males to mature in March at the earliest, and to mature before females. Shrews remained immature from the

time of birth, throughout the winter, until approximately March,

during which time the testes remained at about 1.5mm in length and. the uterus at 4-7mm in length. By April/May, when maturity

was complete, testes reached a length of 6.5 - 8.1 mm and. the uterus 17mm (or, over 40 mm when bearing embryos). Testes

remained large until the males disappeared from the population.

(ie. until death), but the uterus shrank to approximately 15mm

in August following breeding. There is no evidence to prove that

females survive longer 'than males, although in the August and.

September trapping periods two mature females but no mature males

were caught. It (s possi bk. that males die soon after mating, while females remained until the end of the summer or beginning of autumn to finish nursing the young. No mature shrews were caught in October, and, according with the findings of Middleton

(1931) and others, none were caught during the winter.

The relationship between testes and. uterus length and. body weight is shown in Pig. 2-27 A & B. Males reaciied sexual maturity at about 8.Og when the testes measured 7.0 - 8.5mm and. contained

sperm. Females were in breeding condition at 7.5g. Referring to the weight distribution of S. araneus in Alice Eolt Forest, and.

also to The Wilderness to supplement the months where data was not

obtained from the former study area, it can be seen that body weight did. not reach 7.5g until April in 1978 at Alice Rolt Forest; but had done so by February in The Wilderness • The bulk of the population, however, did. not increase to maturity weight until April/

May in both study areas • Mature males, caught alive in The Wilderness were not recorded until Apri],when the testes were first observed. A >180

150

E - 4- a) - 120

39 Body weight (g) - - -

100 B

• . . 80 • . • . -c ••. • . .

,5 40 C 0a, 20-. ... . S . .. . . 0 I 50 60 70 80 90 100 11•0 120 Body weight (9)

Fig. 2-27 The relationship between body weight and reproductive condition as indicated by the size of the reproductive organs in S arane's from Alice Hoit Forest. - Cu -

as swellings around the cloacal region. The growth to sexual maturity was, therefore, very rapid in males, and probably in

females too, and. occurred within approximately one month (between

March and April for the majority of the population).

The contribution of the reproductive organs to body weight at maturity is shown in Table 2-22. MALZ MAIS Nature Immature Uterus & i Testes Testes 7 Embryos Mean weight g. 0.2030 0.0024 2.1271 0.1003

% of total body weight 2.24 0.04 21.10 0.99 Table 2-22

Although the testes at maturity measure 7.0 - 8.5mm each, this represents only approximately 2% of the total body weight, whereas a mature female may have 21% of her body weight incorporated in her reproductive organs when pregnant.

Most captures of S. niinutus were Juveniles or sub-adults with

testes measuring only 0,8 - 1 • 2mm and uteri 4.0 - 6.0mm • The male in breeding condition had testes 6.0mm long, and. the maximum length

of the uterus in a mature, but not pregant, female va.s 8.0mm. As

with S. araneus this species overwintered in an immature state and. reached breeding condition in May.

v) Input, Survival & Life-Expectancy,

The sampling programme carried out at Alice Kolt Forest was essentially a removal trapping. Consequently, the results could be considerec. as a fonu of input, the frequent and regular trapping periods tending to sample the invading populations rather than the resident, established ones. As a result, no data is available on the survival or life-expectancy of shrc's in this study area.

- oJ. -

PART III: WATER-CRESS BEDS

The numbers of shrews caught on conimercia]. water-cress beds

are shown in Table 2-23. Al]. trapping periods lasted. for three days, with the exception of the first one at Chenies in Janua.ry

1975, which was continued for seven days, although the single water-shrew capture was sustained, on the first day of trapping,

and the common shrews on the third. day. An additional three-day

trapping period at Cheiues in May 1975 produced no captures of shrews.

Number Caught Sorer LOCATION DATE Neomys Sorer •fodi. ______minutu Chenies, Bucks 3'ANtrARY 1975 1 1 0 Airesford, Rants STEER 1975 10 7 1 8Bishop's Sutton, I, 3 0 Rants

Airesford., Rants AUGUST 1977 3 4 2

U Bishop's Sutton tt 2 5 0 Rants -

14Bighton, Rants I, t 6 2

TOTAl. 30 33 5

Table 2-23

The numbers of N. fodiens caught in 'the brief trapping periods on the water-cress beds, par'Cicularly those in Hampshire, far exceeded numbers of these shrews caught in other study areas during

the whole trapping programme.

All captures were made on the grassy banks surrounding the

cress-beds and. although some shrews, including S. araneus, were

caught very near to the water's edge, none were actually caught

within the cress-beds. N. focliens was particularly numerous on the

long, narrow grassy banks between groups of cress-beds, or between cress-beds and the river, where they were cut off by water except

for a small strip at one end. Here they appeared to inhabit small rounded burrows, as shown in Fig. 2-28. In these places the on].y

other small mammal caught was Clethrionomys glareolus, other shrews being caught only on grass banks Cuxrounding the edge of the cress-

beds. N. fodiens seemed to prefer living in close proximity to the

cress-beds, for none were caught on river-banks any distance from the beds.

Few shrews were caught during daylight hours when the cress- beds were being worked upon. Cress-workers reported that they never,

or very rarely, sighted water-shrews wule they were working. Most shrews, includ.ing N. fod.iens, were caught soon after dark and. during the night. N. fodi ens was fouth to survive in the traps for no more than four hours.

A study of the diet of N. fodiens in water-cress beds and. deciduous woodland is given in Churchfield (l97, in press) in Appendix 3-2, where the relative importance of acpiatic and. terrestrial prey items is discussed. •1 '

.:: a Ø14• , -%

- I'

,.,' A' :4'' 'S

vs - 411Al . - -- ., ' & : Ad"

. II, I 1ie

,'d. 4' .19' _IIt

• ''

Fig. 2 - 28: Burrow entrance of N. fodiens -

SECTION E: SUNMMiY

S. araneus

1. Using live-trapping method2 in BCZfD grassland. and. deciduous woodland habitats a seasonal cycle in numbers of shrews captured.

has been demonstrated: numbers were high in summer and low in. winter. Trap-revealed population densities aJ.o differed between

years.

2. Shrews entered. breeding condition in March/April, and by May - al]. of the trapped population were mature adults • Adults disappeared from the population following breeding.

3. The recruitment of young occurred between May and October, and by late October the whole trapped population comprised young animals • Shrews overwintered. as immature sub-adults..

4. The sex ratio of shrews was found. to be approximately equal.

5. A seasonal cycle in body weight of shrews has been demonstrated., with a decrease in. weight of sub-adults in. winter.

6. Input of new ina.ividuals was high in summer, when juveniles were numerous, but low in winter. High input in spring is

attributed to (1) increased activity of residen.t but previously

untrapped shrews and. (2) tmmigration as the breeding season approached..

7. Mortality of shrews appeared to be higher in summer than in.

winter. 5C of shrews survived to two months of age. 40 - 5(/

survived through most of the winter to six months of age, but

only 20 - 30% cf all shrews survived. to breed. 8. The maximum life-span of shrews was shown to be approximately 13 months.

9. The rapid decrease in numbers of s1rews caught in late summer and. autumn is attributed to death of old adults by senescence following breeding, and. of uveru.les and. ub-adult by predation

and. emigration. -

10. Trap-revea.led movement of shrews was greatest in spring and. summer. In winter, movement on the ground surface decreased.. 1].. The home range remained in the same location throughout life,

but tended to be larger in adults and. juveniles than in overwintering sub-adults. 12. The problems of population estimation have been emphasised.

Lincoln Index estimates have been shown to underestimate the population density in winter when a more reliable estimate was - provided by the 'calendar of captu.res' method.

13. Population densities of 15-29 shrews per hectare were found. in summer and 5-9 per hec tare in winter. 14. Overwintering numbers of shrews were found to be higher than numbers of captures indicate. This is attributed to changes in trappability resulting from reduced. activity of shrews on the ground surface in winter.

Other Species

1. S. minutus exhibited a seasonal cycle in numbers of captures and. body weit changes similar to S. araneus, but owing to the smi1 numbers of captuies, details of the population dynamics of this species were not coiclusive. 2. An additional trapping study was carried out around commercial watex-.oress beds where N. fod.iens was caught. These shrews were found. to live in burrows in grass banks bet.ieen cress-beds which were almost completely surrounded by water. -

CHAPTER 3

FACTORS APFECTING THE SEASONAL POPULA.TION UCTUATIONS OF SHREWS - -

SECTION A: FOOD AVAILABILITY & THE DIET OF SBBE1S

PART I: DIET

i) Introduction The food of the Soricidae has been studied by several, workers in different ways. Many laboratory studies of the food preferences

of captive shrews have been undertaken (Tupikova, 1949: Boroweki &

Dehnel, 1952; Crowcroft, 1954a; Hawkins and Jewell, 1962; and Rudge, 1965) 50 that our knowledge of what different species will eat is considerable. Data on the food items taken in the wild are also

accumulating from the extensive studies of stomach and gut contents

of various species from different countries, for example

Czechoslovakia and. U.S.S.R. (Pelikan, 1955; Nezhzherin, 1958), U.S.A. (Hamilton, 1930; Buckner, 1964) and Britain (Ru.dge, 1968; J Pernetta, 1976b; Grainger arid Fairley, 1978). Diet has even been studied by examining the gut parasites of shrews, the occurrence of various parasites indicating that the diet corrained certain prey

items which acted as intenediate hosts leading to the infection of the shrews (Kisielewska, 1963).

The disadvantage of the methods mentioned above is the necessity

to kill shrews in order to obtain samples. Consequently, there are no studies in which food consumption and food availability have been investigated on a simultaneous and continuous basis, and. little is known of the relationship between food availability and. population

density of shrews. Rudgea (1965) studies in woodland did not

produce continuous records of food. availability and diet in a

single area with an experimental population, although he went part of the way in this. Pernetta (1976b) showed a relationship between

food. availability and diet in S. minutus, but his method. of - 67 invertebrate sampling was not suitable for the study of prey items of S. araneus.

The object of this study was to monitor food availability and. diet simultaneously for a population of S. araneus. Por this, faeca]. pellet analysis was undertaken. The value of faecal pellet

analysis haa so far been overlooked.. Pemetta (1973) mentions it as a possible means of studying diet, but dismisses it on the basie -

that faeoal pellets collected from a threw on a single occasion

represent only 5% of the daily faecal output and. so, in his view, its value in quantitative studies of the diet of individuals is negligible. However, faeoaJ. analysis has the advantage of producing

numerous samples from a population of shrews without adversely affecting the animals and thereby giving a continuous record of diet. Provided faeca3. pellets are collected from large numbers of shrews the resulting samples must be random and so ean provide quantitative estimates of the diet in a population of shrews. Ru.dge (1968), found that the passage of food, through the gut of shrews is vcry rapid, the interval between ingestion and egestion

being as little as 30 minutes, and. this is to the advantage of

faecal analysis becavse it facilitates collection. He found that

the rapid passage of some items such as isopods, and the swift

breakdown of others such as lumbricids, left a disproportionately

large amount of molluso and geophilid. remains in the stomach.

Stomach contents were, therefore, not a true reflection of the

original contribution of the food items, This difficulty is not

encountered in faecal analysis, for remains of all invertebrates

can be found. not only are chitinous components not digested, but

also a proportion of the softer parts of invertebrates such as caterpillars arid dipteran larvae, which is irrespective of the

throughput rate. Moreover, 9% of samples contain identifiable - 68 -

remains, unlike the frequency of empty stomachs encountered in stomach analyses. Consequently, a wealth of material is available for study.

ii) :h Thiring each trapping period between July 1976 and April 1978 the faecal pellets were collected from Longworth traps in which shrews had been caught. The faecal pellets of all shrew species cauit were found mainly in the tunnel section of traps, from which they were easily removed with forceps and placed in specimen tubes containing 7C alcohol. As many faecal pellets as possible were collected from each trapped Individual, and these were taken as constituting a single sample. Numerous samples were collected during summer months when the shrew population was at its peak. Fewer were obtained in winter 'then the captures decreased, but it was assumed that these were still representative of the diet.

Faecal pellet samples were sorted under a binocular microscope

and the presence of any identifiable food. item was noted. Items were

recognisable as lumbricid chaetae, gastropod. shells, arthropod. limbs, wings, head—capsules, mouthparts and. body chitin, and even softer

parts of the body wall inclu&ing spiracles. The identification of

these remains was facilitated by the construction of a reference collection of potential prey items taken from the study area and. mounted on slides in Hoyer's solution. .axititative estimates were made impracticable by the fragmentary n.ture of the material, and it was rare that complete items in a sample could. be counted.

iii) Results & Discussion

Two hundred and. forty— Six samples, collected from The Wilderness

were examined. These included 220 samples from S. araneus, 22 from - 69 -

S. nanutus and four from N fodiens. Six of these samples

contained only unidentifiable components.

a) S. araneus: a complete list of the prey items identified from S. araneus is presented in Appendix 3-1. Table 3-1 shows

the percentage frequency of occurrence of these prey items during

each month of sampling. A wide variety of invertebrates appear

in the diet of S. araneus; this confixms the findings of Rudge -

(1968) and Pernetta (1976b) with three notable additions: two occurrences of millipedes, four of ants, and. one of down feathers which may- indicate that carrion had been eaten. Shrews have not been found to feed on any of these items in food. preference tests.

These items, together with coflenibolans, free-living mitea, pseudoscorpions and. earwigs occurred. only in small numbers and. constituted a minor proportion of the biomass of the diet. Rudge

(1965) considered invertebrates less than 3mm in length of negligible importance as prey and the possibility of accidental ingestion camiot be excluded. This view is also held In the present study and. consequently these items, thoui recorded in Appendix 3-1 and Table 3-1, were not included in further treatment of the data.

One constituent of the faecal pellets which was wholly discounted was vegetable material, although it wis frequently found. Lavrov (1943), Kangur (1954) and Zablotskaya (1957) suggested that plant material was eaten in greatGr quantities in winter to supplement the diet, but Rudge (1965) considered it to be an accidental introduction and reports only a single example where plant material increased in occurrence in the gut to suggest that it might compensate for the drop in the invertebrate food. supply in winter. Buchalozyk (1961) found in his studies of salivary glands

0t- ø t-4"xjb o -0t-4C) s00C) 0l CDP,00dI.a. I.1. m tflO(D c-1- 0 c+ CD 0 C) (-$- (D c-i- 0 e+- I P3 OPjjd P3 H 0 1 CD r c) : rIc, Ii I-' i-I i I-' rrj :j CD QCD'-•1CD (Dc) (t! P. CD H CD 0 0 0 I-'. P3 r P.O 0 'd 0 H Fd CD(D0 OQ.OpIa (fl-'c4- p. jao7ocD (flOP. CDOCDO QOO 0 CD d CD CD c- CD b H c- CD o s-' p. p. 'd-'P' CD1c+(DP I-IH P P3 ooq. P3I- : sP3P.ji-i '-3 ij mi-P'cP1P ci OCD '

N N a'. ,.-a -.a N N ¼..) 0' H - H N ...... 0000000 N 000 1%) 000 0'..) 0 0) 00 H 00 - 0 I-' 00 H 1300 I-I H H H . ) H H 0000P000P0000003oo?'" 0 0• Hf-LItLn0 • . • • 0 c CD -- a'. Lit H N I-' N N H N H N '.0 N ¼..) 0000 -] 0 - CD CD - CD sj . H N H 0 . S • S 00 . • 0. . • . 0 S 00 . 0. • 00 . 0 . • • 0 • HI- H 0 N 00000000

00000000000 00000000000000 00000000 H H I-' 0¼..) ¼...) 0 ¼..) 0 0 ¼..) ¼..) 0' c-I .D ) 0' ...) 0 0 0¼..) ¼..) 0'. 0 0 O ... 00 . 00.0 . 00 . 0000 0000 00)00030

N N H N N H - H 1') H 00000000000000000000000000 0000; f.J S • S S • 0 . N 00 0 00 0 0 00 000000000 0'. '8 • . . . 0000000. 00000. 00000 . 00 0 0 0 0 00 • 0 0 0 0 H H N I-' C' ¼...) N CD'. '4 LIt '4 LIt-'.) - 0 0¼..) Lit 0 '.0 0 CD 0 " ¼.,.) ¼..) 0 0 • 0 0 0 0 H 0 0 • 0 • 0000 . • 0. H • S • S • • S -3 • •-'.,] -'.] ._-4 -4 LI) N ¼j cOO'.4 0'. . H 00000-4 . ¼..iO . 0¼..)000CDO0%0-..]0000NL,t H • _.3. t..) . • . . • • S • • 000 0'. CD CD Li ¼..) jtLIfl.n H HHH 0 0N.NNN • S • S S • • S • • • • • . 5 0. 0' 0 • 0 . • . . . S 0¼..) Lj CD N N CD.) ¼.jt ¼..) 1'.) N Lii '4 N 4 LIt _4 N'.tj*Ljt t) - . N H H H C) H N H H N 4 CDLI) .¼..) C7'. 000 - 0 ¼..) 0 H 0 LI3 -.) - ¼.) 00 - •H S 0'405 LI)5 00 - 5CD • • H 00 '• S S 5 5 5 S • • S S S - '4 0) LIt s) 0 0 H - 0 -4 0

H I-' I-' I'.) 1-' I-LI) N 0 - . . '0 -.) 0 CD N 0)0% 00'A -i- N00 . CD 00 CD N - 0 Ut 0 CD 0 CY'. 1'.) - ¼/i • U • • • • • • . . S • S • • S • U • • S 5 5 5 t) N .3 N 401'.) Ut LI) Ut LI) 4 N 0)19 LI) LI) 1'.) 0 LI) '4 Ut ¼jt 0 H ..) LI) H LI) H ).J H N H N H - 00 Ut. LitS S0 '0• •Ut U I-' SUt •U'. S Ut 5¼'. • 000 UtS Ut• • 0'. S S0000 UtS 0 U'.• 0 H• 0 Ut• 0 SH -0 S 4 5Ut %H C ¼.a)LI)Ut 0¼..) 0%LI)¼..)U'. ¼.)LI) 0)¼.jtLit CD LI) H CD HUt H ti Cl'. H H 0 LI) -J LI) H LI) '-' Cl'. H H H H Lit Ut LI) 000'. 0' 0' 0 LI) 0' LI) 0'. LI) 0' Cl'. 000'. 00' 0 000'. 00'. 00 LI) C'. S I • S 0. 000. 0. 0. . • S S • S S S S S 4 0 LI) LI) '4 LI) '4 4 '400¼..)

000Q0000 0000 000 CD 000 LIt 0 0 Ut 0 LI'. 00 Ut _ - H 1'.) 1'.) H F.) N H '0 H F) 1.) 1'.) 000% 00 LI) Cl' 0' 00¼..) 00'. 000% O'.LI) LI) LI) 0 0'.. 0'. 0' 0 Cl'. H S S • S S I S S S S S • 00 0 0 00. 5 • 05 • i_J '4 OLI3'3'4 0 LI) '-4 '.J -4¼.) ¼.) LI) 0 .-9 -.J '4 0-4

Table 3-1: Analysis of faecal pellets of S. araneu from The Wilc1ernss % Frequency of occurrence ot prey items - 70 -

that shrews were almost completely carnivorous, even in winter.

Recently—captured shrews were frequently observed, in the present

study, to bite grass bedding material in their cages, but not to eat it, and it is suspected that this is a form of displacement activity which could also occur in Longworth traps. Consequently, vegetable matter was not included in the dietary recordn.

The mean number of prey items identified per sample are given in Pig. 3-1 where between three and. six different itens occur per sample. The maximum number identified from any sample was 12. Winter samples (December and January) show a greater number of different prey items taken which contrasts with Pernetta' s findings (1976b) of increased variation in the diet in summer, The largest number of prey types occurring in winter diets does not seem to be a function of increasing numbers of invertebrate types available as will be revealed later.

Prey items were combined into nine major groups which most frequently appeared in the diet and. also in the invertebrate samples collected (see later). These groups include adult coleopterans, hemipterans, adult dipterans, insect larvae, chulopods, araneuds and opilionids, uopods, lumbricids and. gaatropod. The percentage frequency of occurrence of these major groups in the faecal pellet samples of S. ararieus is shown in Fig. 3-2. Adult coleopterans, insect larvae, araneids and. opilionids, and isopods were the major items in the diet in all seasons. Of intermediate importance were luxnbricids and adult dipterans. Other groups, including hetnipteran, chilopods and snails were more sporadic and less frequent in occurrence. Coleopterans were a dominant item in the diet of shreva found by Mezhzherin (1958), Rud.ge (1968) and. 0. E a (A 0 z *

4. 4. H

a1dwos iod °N

- 7]. -

Pernetta (1976b) in different localities, but considerable

variation in occurrence is apparent between other groups of invertebrates. Rudge found lumbricids were another important

item inthe diet while lepid.opteran larvae, isopods, chilopods and molluscs were intermediate in frequency. Araneids, opilionids and dipteran larvae occurred infrequently. Pernetta

(l976b), too, found. lumbricids to be a major item, but 50 were opilionids, while insect larvae and. spiders were of secondary importance and henapterans were absent. The differences between studies is emphasised by Fig. 3-3 where results from the present study are compared with results of stomach analyses performed by

Pernetta (1976b) on a monthly basis. The percentage composition in

the diet of four major food types are given, namely adult

coleopterans, insect larvae, isopods and lumbricids.

The variation in importance of different food types between studies, taken on a yearly basis, can be explained by the nature ot the localities, their floras arid weather conditions governing the

abundance of certain prey items. There is also great variation in the diet between years which complicates seasonal comparisons. In. this study, for example, lumbricids were found. frcquently in the

diet during 1977 but rarely in 1976, but this may be explained by the summer drOUit of that year when earthworms retreated. away

from the ground surface. On the basis of 1977 results, lumbricids could. be considered. a dominant food item and. thereby agree with the results of fludge and Pernetta. The variation in dominant components in the diet between the two years sampled can be seen in Fig. 3-4 showing the composition of the diet in spring (between March and May), summer (between June and August), autumn

(September and. October) and winter (between November and. February) of 1976/77 and 1977/78. Adult coleopterans, insect larvae, -2C.

Uiz.

Iii U, z U, 4 - 4 V U, -a I- 0 U 0 U,'U —a U, z 5

- k

L 'I) % I '.4

ci - - I C 4'

b a' -) 56

// 'I 6 ___ I I J L_ 1 _J I P - I In 00 2 00 in 00 0 0 0 C', C') 0

-72- araneid.s and opilionids and. isopods are the main components while other groups make a variable contribution. Adult coleopterans remain relatively constant in their contribution

during both years with autumn and. winter increases contrasting with Pernetta's summer peaks. Insect larvae show great variation

especially in the summers of 1976 and 1977, but demonstrate spring and. summer peaks in both years. Pernetta (1976b) also found summer peaks of occurrence in insect larvae, but Rudge found d.ipteran -

larvae were most frequent in the diet in autumn and winter

Variation between years can be seen in Perrietta's studies: he states that lumbricids were more important in winter than summer diets, but only in one year sampled were his results significantly higher for winter, and the present study shows both winter and. summer peaks in 1977.

The si'ze of invertebrates in the diet of S • araneus ranges

from large lumbricids, tipulid and lepidopteran. larvae greater than

20mm long, down to small etaphylinids, hemipterans, collembolana

and mites of less than 5mm. The sizes of prey items were divided into two categories where possible: large (p5mm) and. small (5mm). The sizes of araneldd, opilionids and snails were difficult to

assess because of the franentary nature of the remains in the

faecal pellets, and. so were omitted from estimates of the size composition of invertebrates in. the diet. A high proportion of invertebrates of less than or equal to 5mm were recorded in the diet, with a mean of 40. C9 for the whole sampling programme (range 17.4- 75.C1)). These consisted primarily of small staphylinids, hemipterans adu].t dipterans and. isopods, with occasional occurrences of other coleopterans, mites, collembolans, pseudo—

scorpions and. ants. High percentages of small invertebrates in the diet tended to accompany large numbers of small staphylinids in the -73- invertebrate samples, with peaks in autumn and early winter. b) S. minutus: a complete list of prey items identified. in faeca]. pellets from this species is given in Table 3-2. Table 3-3 shows the frequency of occurrence and. percentage composition of major prey items in the diet of S. minutus in The Wilderness..

Notable by their absence are lumbricids, which agrees with the findings of Pernetta (l976b) and Grainger & Fairley (1978). The - major items recorded in the diet for the complete sampling prograe are shown in Fig. 3-5. Like S • araneus, the maaor food items comprised adult coleopterans, araneids and. opiliomids and insect larvae, and accords with the results of Mehzherin (1958),

Pernetta (1976b) and Grainger & Fairley (1978). Remipterans and. adult dipterans were taken in higher proportions than by S. araneus but iopods and chulopods occurred infrequently, althoui isopods were a major dietary item for S. nanutus in Ireland (Grainger & Fairley, 1978). 57.4% of prey items taken by S. minutu.s were less than or equal to 5mm in body length, which is a slightly higher proportion than that taken by S • araneus.

c) fodiens: All four samples of faecal pellets e,rim,ned were obtained from the Monks Wood. study area and not from The Wilderness.

The results of the analysis are given in Table 3-2 with those of S. minutus. A fuller account of the diet of N. fodiens in the wild is given by Churchfield (197q , in press) and. presented in Appendix 3-2. Results from the limited data available from Monks Wood showed that equal proportions of staphylinids, adult dipterans, isopods, opi1ionids chi].opods and lumbricids were taken.

0 0

o H 1'

W U— r4 0 H 0• a) H •rI

Es ('J 1. c\J

H H H H co N- 0 H H H H

H H H H H H 0

a) Es a) t C) v-I H H H H C.J (SJ (.J 0 H vi

a) Es H —I H H H H OJ H Es

4-' O co ) CJ r-i H H C\I H

E 1-I -1

H H

H H H

F-s H ,-4 H H H a) E E C) Ca H L(\ a) P4 H C) - .4. -'-- C) v-I C) Ca a) Ca C) C) -s Ci ,r Ci Ca Ci • E -i-i -.1--I U) v-I -s o H a) If\ •.-I • C) C) al .s + U) C) s-I s-I P4 E s-I H C) i-I -P (i U) Ca 0 es-s .4-' i v-I Ca U) P4, vi V-4I c3 -s C) E ci v-i P4 C) ) -q -P v-I I c C) C) P4 C) 0 v-I -P • 01W (1)10 U)-4 4 P4. - ti T3 P4 -v-s P1 I v--i P4 0 v-I v-I 0 0 rlv-4 0 -4 H i ai .I a) Ola) a) 0) b C) ol.c a) .CLLH Ci ..c CI) P. P. (15 (13 I -P 0J -I-' 0) 4-' 0 (1) 1-' 0 v-I C) C)E0 0

'-I

I—.

C.J i—I i-I r-4 C) P

'-4

H 43 C) H U] E 0) 4.3 (J 1-4 i-I

l r4CJ .o ,-

'-4

'-4 '-4

ci) cI_ 1-4 '-4

4-itO

U] U] U) P1 P1 tOw H Lf\L(\ .4 .4 1 (U 00 tO U) 0 U] E 0) 'Ti Ti r-1 0 -I E CO H H 0 U) 0 v 'Ti 0 0 'Ti rJ 43 H CO ,—I H 0 'Ti 0 ,0 C) CO P1*) d ,0 P4 HO H U] 0 H 0 0 0 0) H ,-1 0) H 'Ti 0) B P Pi. ,-4 $ 3,0 H r1 ., 0 043 P4 HE (' ,—1 ci C) 4) 0 CC 4) .4 .-4 -4 P1 0 C) U]

p W '1 CD P 0 rn I- C fr i- CD I-' c+ 0 CD 0 C) c$ CD CD P I- c+ c4. CD 0 g w P.. 0 CD C) b- -I o 'd P.tl CD 3 CD i 'lCD CD CD rn 0 CD '1 0 3 'd CD c4 -, I-" CD CD l.a. 0 CD I... I-a Pa CI w Cl) l-•' I.-' l.- l- l- - 0000000000 0000 I-

C4 - • I-a0 0 I-' I.ad 0 0 0 0 0 0 0 0 0 CD c+. I-I I-' P-a C.1 0 00 I-' 'I 0 000000000 CD CD 0 I-' 000000000 C I-' 1a I-' I-a N) N) - -. N) N) '.fl '.0 0 0 0 n • ...n • ¼.n • 0 Co C —) J1 Lfl C 0 CD 0 c\ (1J(jJ 0 0 0 a'. 0 a'. 0 0 ti-' 0 I-' —• —. • • 0 Ii 0 CD -a 1'.) '.Jl .7I 0 -N CD '- 0 00 00¼J1¼J1 000 CD

IJ. I-a ,-' I- I-1 l-J 0 0 000 I-' 0 0.. 000000000 .4 CD I-a00000 I-' CD 00000 0 0 I- N) J1¼)1 - CD 0 0 0 0 -N ' 00000 0 N) N) I'.) N) N) ca I' 000000000 I- a'. 000000000 I-.a C.

00 0 000 I- • . S IjJ t,J I.1

I-.' 0 P-a t4 - 000000030 I-' C) '.0 0 IjJ I- I-' N) C -4 E 00 •0¼1 • . •J1 • OLfl• • I-' Co Q-4 'd0 I- U) i u.a a'. co '.ji u., I-'.

N) I'.) I-' I-' IjJ 0 0 N) 0 N) 0 1á I-' UJ 0 • • • N) N) I-a 0 - N) I- - N) C) o o . 0 I-' - • S S • S • - I- (jJ I) '.

1a t.%) N) N) N) .4 000000000

0000.fl00fl0 Ia 0 0 '.0

0 0 0 0 0'. 0 '.0 0'. Co - d Co • • S S -N N)

FIGURE 2 CHANGES IN TOTAL PRO PORTION FRAGMENTED IN RELATION TO CHANGING FARM SIZE

NI LU U) LP) E LU (0 Li

ru 4- 2 C

C ro Li C) C - 0% 4- C C) Li Li.) C- 0 LU

Li Li.)

-50% Percentage fragmentation 1958-78

Source fetdwork - 7k -

PART II: ESTIMATION OF )OD AVAILABILITY i) Introduction

Shrews have a catholic diet (Crowcroft, 1954aShi1lito, 1960) so that any assessment of food availability recpures a wide spectrum of invertebrate types to be taken into account. This poses difficulties not only in taking samples of invertebrates from Boil, leaf litter and aerial vegetation, but also in their subsequent - extraction. Macfadyen (1955) states that there is no single perfect technique for the quantitative extraction of all soil ntma1s, and equally this could. be said of their sampling. Most sampling and extraction methods are desiied for particular groups of invertebrates on the basis of their activity and behaviour, so making the employment of available proved techniques unsuitable in this study.

In his study of the food available to shrew3 Pernetta (1973) based his sampling technique on his theory of tactile detection of prey active on the ground surface and employed a simple and. economic system of pitfall trapping. Eis method had two major drawbacks, firstly, that pitfall trapping selects for certain types of invertebrates highly active on the ground surface at the time of sampling, cuch as adult coleopterans and. araneid.s • Invertebrates rarely encountered in pitfall traps, such as lunibricids and. geophilomorphs (and certainly not in the numbers proportional to their population sizes in soil) nevertheless do occur in the diet of shrews (Shillito, 1960: Rud.ge, 1965). Secondly, invertebrates available as prey for shrews occupy a vertical as well as a horizontal profile, whether it be lumbricids in the soil, chrysomelids on aerial vegetation or flying dipterans, and this is not taken into account by pitfall trapping. Likewise, shrews - 75 - are active throuiout much of the soil/ground, vegetation profile being burrowers and inhabitants of subterranean tunnels, as well as good climbers.

Rudge (1965) realised the importance of vertical sampling of invertebrate prey in his woodland study of shrews, and took cores of leaf-litter and surface soil. He extracted the invertebrates from the samples using a Kempson extractor, but this method again. - tends to be selective for the most active invertebrates such as Coleopterans, hemipterans and collembolans, while those which are particularly resistant to the effects of heat and. desiccation such as geophilid centipedes, or are donnant such as d.ipteran pupae, are not extracted (Macfadyen, 1955; personal observation).

It was the aim f this study to assess the relative abundance of invertebrate prey types on a seasonal basis and methods were employed which would sample the fauna in a vertical as well as a horizontal profile. ii) Methods

The present st.tdy took place in mixed scrublaiid where surface vegetation was thick and abundant at all times of year, which further complicated the taking of samples and. made the subseqaent extraction of invertebrates difficult. A large metal cylinder measuring 345mm lu and 300mm in diameter and having an area at one end of 7069 sq.mm was employed to take cores of vegetation and. soil. One end of the sampler was sharpened to cut out a core of soil. The sampler, with sharpened end downwards, was rammed hard into the ground to a depth of 80mm with a large heavy-duty polythene bag fitted over the top to contain aerial vegetation and any flying or jumping insects. A flat spade was dug into the soil - 76 - underneath the sampler whereupon the whole cylinder and the contained sample was upturned into the polythene bag, the neck of which was promptly secured with string to prevent the escape of any invertebrates. In this way a large sample of aerial and. ground vegetation, comprising m-any grass, willow-herb, bramble and moss, together with the surface soil, where the maority of soil fauna was found, could be collected. This proved satisfactory for all groups of invertebrates with the exception of fast flying - insects such as large dipterans and syrphids whose numbers are consequently considerably underestimated in the results.

Ten vegetation/soil samples were taken from The Wilderness study area during each trapping period between January 1976 and.

December 1976 inclusive, and in June and July 1977. This involved considerable time and effort in collection and extraction as each. sample was very large, so for the remainder of the sampling programme

(up to April 1978) the number of samples was reduced to 5 per trapping period. It was not the aim of this study to investigate in great detail the population dynamics of the invertebrate fauna in the area, but merely to obtain an indication of the relative abundance of potential prey items at different times of year. For this reason it was deemed acceptable to reduce the number of samples taken.

The vegetation/soil samples were taken from 10 points on a stratified grid. Sites were selected away from the trap points themselves to avoid disturbance, and the same sites, occupying an area of some 20 sq.m were used each time in order to stand.ard.ise the sampling programme, while encompassing the whole range of vegetational types available in the study area. -77..

The activity, and therefore distribution, of invertebrates

amongst vegetation depends greatly upon the time of day and

prevailing weather conditions (Southwood., 1966). In order to

reduce variation due to these causes to a minimum, all samples

in each batch were taken on the same day and. collected as near as possible to midday. As shrews are active, and. need. to feed,

throughout the day and. night, whatever the weather, the relative activity or non—activity of prey items at certain times of day

was disregarded for the purposes of this stLldy.

After collection, the samples were taken to the laboratory where they were sorted as promptly as possible, usually on the same

day. Sorting was carried out by hand. for, although tedious, it was

found to be the most rapid and. efficient method to employ. Initially, Kempson. extractors were used. but a comparison of this method. and. hand. sorting proved the fomer to be impractical as

well as inaccurate. The samples were vary large and dense, with

heavy clay soil that tended to bake hard in the apparatus,

entombing the less active invertebrates (such as geophilids and.

insect larvae) before they could escape, zaid thereby selecting

for certain faunal types only. Rudge (1965) repo'ts that litter

inhabitants of less than 3mm body length were of negligible importance as prey of shrews and, in. accordance with this,

invertebrates of this size were disregarded.

Larger invertebrates were easily seen in f'eFh samples where

they were alive, and could be reliably removed by hand. Care was taken to remove all flying insects first by introducing a 'pooter'

throgh the neck of the bag containing the sample. The sample was

then tipped into a large steepsided tray for thorough sorting, the invertebrates being placed. in 7( alcohol to await - identification and counting.

In ad.d.ition to the 3—dimensional sampling of vegetation and soil, 10 pitfall traps were dug in at 10 points on a stratified grid, to investigate the activity of invertebrates on the ground surface during each trapping period. These were in operation from November 1976 to April 1978. Between trapping periods the pitfalls became flooded and were pushed. ou.t of the ground by - rainwater so they were employed for the three—day duration of each trapping period only. Plastic cups containing some 7/o alcohol with glycerine were used, and. each one was sunk into the ground and its lip obscured by moss or earth. The sane sites were used on each sampling occasion. Invertebrates caught were removed after the three—day/night period, identified and. counted, the total number of different prey types caught being recorded..

All invertebrates were identified and. categorised. according to type and size (3-5, 6-10, ll-20,>20=). For each set of vegetation/soil cores the mean number of all invertebrates greater than 3mm in length per sample was calculated and. this was subsequently used in calculations of numbers per m 2. In additions the invertebrates taken from the vegetation/soil samples, excluding such items as ants and diplopod.s which initially were not found in the diet of shrews (although these two examples were subsequently identified in two faecaJ. pellet samples), were dried at 60°c to constant weight to obtain a biomass estimate in tenna of g dry weight per m 2of potential food. available.

iii) Results & Discussion

Appendices 3-3, 3-4 and. 3-5 comprise the invertebrates

collected month by month in the vegetation/soil cores and in - 79 -

pitfall traps respectively. As it was not possible to identify many of the prey items in the faecal pellet samples to species level, to facilitate comparison the catches of invertebrates

were divided into the same nine main food types as follows:—

Adult coleopterans Hemipterans

Adult dipterans Insect larvae Ararieids & opilionids Chilopods Isopoda Molluscs

Lumbricida

There are two notable omissions, namely ants and. diplopoda, which rarely entered the diet of shrews during this study, and. so have not been included in further treatment of the results. Slugs

are eaten by shrews (Budge, 1968; Pernetta, 1976b) but were not identified in the diets of shrews in The Wilderness. They are included with gastropods as a potential. food. item, but are not discussed. in detail. Large numbers of collenibolans and. mosquito

adults were collected in pitfall traps, but occurred very- rarely

in the diet of shrews, and so are not included in further treatment of the data either.

a) Vegetation/soil cores The seasonal number of major potential food types collected in vegetation/soil cores are hown in Fig. 3-6. The most numerous potential prey invertebrates in the vegetation/soil cores (down to a depth of 80mm) were adult coleopterans arid lumbricids,

E- °c O

Ui -z

I, 41 ui auaiino o Xuenba % 0 ______o 0 I I I < 1 1. 1 S. - S S I - S I - - \ - I - I . / . . 'I 4 - 4 - - If I - , S _ • - - / ___

i - I k- - ° )s. ç_ SI)

4 4 ç I-

4JI -'

- 'ks) ^ -' - . - - I -

- -

4 I. 4 I. I - I - $ I 5 I . S I $ I- I ,-i - I • - I I , I g • I I I S ' -

4 . -I ------0 4 - .z / •: • 4 - I 4 p. r - j .° 4 4 SI) 4 S - - $• - • b I• i s a8 I S p • • - - . - I. I .-$

I) -1 — .0 I I . • z . - •0 -I, - • -z I- II.' I) 0. - - .0 . - 0 4< - •0 .0- .• U - 0 0 U -J- .1- 0. -J z - - - - I--I LI - < lx I I J !_ I I I I I I J I I I I I - _i•••__' 00 0 0 00 0 00 00 0 0 0 00 0 0 0 0 0 00 0 0 * N 0 0 0 0 0 N 0 0 0 0 0 - N - N N - N iad ioqwn - 80 -

reaching densities of up to nearly 900 individuals per in2 each,

and. araneids and opilionids with densities up to 200 per in2.

I'Iost apparent from Appendix 3-4 is the predominance of small invertebrates with body lengths of 3-5mxn. These included araneida, hemipterans and coleopterans, particularly staphylinida belonging to the genus Tachyporus.

The variation in numbers of invertebrates between the two years sampled tends to override the seasonal change in numbers, for example, hemipterans and adult coleopterans were more numerous in 1976 than in 1977, and. vice-versa for lumbricida, This can be accounted for by the weather conditions for the two years of

sampling: 1976 being dry and hot while 1977 being predominantly wet and cold (see later).

Lumbricids showed peak densities in the upper 80mm of the soil during the months of high raiznfall. They showed. a marked increase in numbers from a scarcity in the dry summer of 1976 to a peak during the following wet winter. Plying insects and invertebrates active amongst aerial vegetation such as coleopterans, hemipterans, d.ipterans and araneids, were numerous during the summer months of

1976 and decreased towards winter. In the following year, with its cold, wet summer, there was a much later increase in numbers of these groups towards autumn.

Some groups, namely isopods, molluscs and insect larvae, were common throughout the sampling programme, the first ti,ro inhabiting surface so.l and. tussock vegetation, and the latter, comprising lepidopteran and tipulid larvae, being quiescent in the soil over the winter. - 81 -

The seasonal composition of the invertebrate fauna is

shown in Fig. 3-7 as pie-charts, based on the numerical proportions of each dominant group collected during each of the

four seasons, spring, summer, autumn and. winter of the two years

sampled. Again, considerable variation it3 demonstrated. between one year and the next, but adult coleopterans, 1umbricid,

araneids and opilionids predominate. All other groups, including insect larvae, are variable in their contribution.

Estimates of biomass and total numbers of potential prey invertebrates per are depicted in Pig. 3-8. These emphasise

the variation between the two sampled. years and. the relative abundance of prey in winter months. The major contributor to the biornass was the Lumbricid.ae and. the results reflect the numbers of earthworms present in the samples. This is particularly marked in the period following the dry summer of 1976, with a peak of 16.5g dry weight/rn2 in March 1977 when large .Lurnbricids 20 mm) were abundant. Insect larvae, especially tipulids and. large lepid.opterans

(mostly Hepialus lupulina) also formed an important proportion of the biornass. Coleoptera, although numerous, contributed comparatively little to the biornass becanse of their generally smaller size. b) Pitfall Trapping

Considerable differences in the relative numbers of invertebrates are found in the two methods of sampling employed.

Fig. 3-9 shows the seasonal numbers of invertebrates obtained in vegetation/soil cores and. pitfall traps. The captures obviously are not comparable on the basis of numbers alone, for the vegetation/ soil cores show mean numbers of invertebrates per sample while the pitfall samples represent the total numbers of captures over the three— day trapping period, However, i.t is interesting to compare II, 1.-s I II I liii I I 0 0. ,1iiIlij s-s. I I III uJ z I -o oa V o •o °- E oo II 1.-s

0% I- II z fIYII t I I -C' (. z II, \--'•\ , UI \ -o ll I-. C 0 4-I .0ri a. O

z In -D LU 2-u 0

0 U O LU U) o - —J UI - r-Ifl -

U, * 4, 4- U, 0 1 -Q 4) 4-. I- 4) co > C

0 U- z -*

-1 :1 .

"4. 14

u-* I -) O z o

-) —*I

-.5 9 0 00 0 9 0 0 I- '0 I.-

w .sed 4M XJG w sad °N 8

..a 79 U, .1,, U, U z U, 0- 4 -a ILl 0-a - , <0 J

'II-, 9

0. IL

'U-

o 0 0 00-

SIOOPIAIPUI - 82 -

the cycle of numbers. Pitfall samples show marked seasonal peaks which are more evident than those of the vegetation/soil cores. This is emphasised by the adult dipterans, coleopterans and araneids in summer, and molluscs in autumn. Some groups, such as lunibricids and. chilopods,were rarely encountered in pitfall traps, whereas fast flying insects, especially Diptera, which were infrequent occurrences in vegetation/soil samples, fonn a major

component of the surface active fauna in all four seasons depicted. - in the pie chart of Fig. 3-10. Other groups, such as heinipterans, isopod.s and araneids were found. in si m4 1 ax proportions in both vegetation/soil cores and. pitfall traps.

The relative abundance of certain prey items in the vegetation /aoil cores and pitfalls are directly comparable in Fig. 3-11 where the percenta€e frequency of occurrence of selected invertebrate types throughout the sampling programme has been calculated. The levels of abundance of isopods, araneids and opilionids, and insect larvae in vegetation/so!1 cores were similar throughout the year while coleopterans occurred in proportionately larger numbers in winter samples. However, occurrence in pitfalls showed. marked peaks (in summer for isopods, araneids and opilionids, and. coleopterans, and in winter for insect larvae). Therefore, although abundance may remain relatively constant throughout the year in the soil/ground vegetation profile, activity of invertebrates on the ground surface, reflected by numbers in pitfall traps, varies from season to season. This emphasises the drawback of sampling the invertebrate fauna by one particular teonuque and. the difficulty of incorporating the whole spectrum of invertebrate types in a stu.dy of this kind..

- 83 -

PART III; THE RELATIONSHIP BETWF DIET ID FOOD AVAILABILITY: GENERAL DISCUSSION

A major criticism of the method of invertebrate sampling and collection of faecal pellets concerns the time at wiach samples were taken. Pitfall traps caught invertebrates throughout the day and night, but in order to standard.ise the method, vegetation/soil

cores were always taken around midday, thereby disregarding any

daily differences in activity and. behaviour of potential prey items, - which may affect their distribution amongst the soil and. ground vegetation. Likewise, faecal pellet samples were taken only

during daylight hours. It is possible that there is a correlation between circadian activity and distribution of prey items and their

occurrence in the diet of shrews, which has been overlooked in this study. Information on the diet of shrews has been based upon their

feeding habits during daylight hours, and it is possible that the constituents of the diet change at night. An example is earthworms, which tend to be more active on the ground surface at night when they may be eaten more frequently by shrews than was suggested in the results presented here. Despite the limitations of the methods. employed, the prime aim of this study has been satisfied, for the results are representative of seasonal trends in diet and. food. availability, and on this basis they will be discussed in the light of results obtained by other workers.

Rudge (1965) found a close relationship between abundance of certain potential prey items in the leaf litter and their contribution in the diet of shrews. Major groups, such as lumbricids, isopods and snails, with a high frequency of occurrence in guts and a high individual biomass tended to be prominent in the diet of

S. araneus in all months, despite changes in numbers in the litter. -8k-.

The frequency of occurrence of secondary food groups, such as slugs, lithobiomorphs, opilionids, araxieids, d.ipterans and

insect larvae, was more closely dependent upon their current status in the litter.

Pernetta (1976b) found a close correlation between the diet of S. minutus and the availability of surface active fauna for

certain dominant prey items (araneids, oilioru.ds, adult

coleopteraris and insect larvae) but was unable to demonstrate

this for S. ararieus owing to his unsuitable method of

invertebrate sampling. The seasonal composition of the diet of

S minutus compared with the availability of invertebrates estimated

by both vegetation/soil cores and. pitfall trapping in this study is shown in Fig. 3-12. Little correlation is apparent between diet and. food availability except for the latter part of the sampling period. when the numbers of faeca]. samples eTimi?,ed increased. Here, the dietary composition in the fonn of adult coleopterans, adult dipterans and insect larvae reflects the occurrence of these items in pitfall traps and. hence their activity on the ground surface. When diet of S. araneus and food availability are compared for the present study little correlation is apparent. Pig. 3-13 A & B shows the percentage composition of the diet of S. araneus in

The Wilderness together with the percentage composition of the

invertebrate fauna sampled by vegetation/soil cores and. pitfall

trapping, for the whole sampling period. Where pitfall trap

captures were insi.iificant in. numbers (as in the case of

myriapoda) they are omitted from the graphs.

t The diet of S. araneus shows no corrol.ation with the relative

activity of prey items on the ground surface. Adult coleoriterans were common throughout the sampling period in both invertebrate

LEGENO £6 Out •.—* Pitfall, 0—C Veg.,6ouI co-.

20r ..11ILU)Ua A 01 t-r°rO-r-_n_1. I - ARANEIDS & OPILIONIDS

INSECT LARVAE A 60 I' I 50 , ' I I %40 p I p I - -- ' - - - - 'N ADULT DtPTERANS

HEMIPTERANS

--- ..,- _____&

100 ?\IJULI L*,JLC.JF Pr4

80 a'

60 :

0 I _t I I I I i I I J A S 0 N D J F MA IA J J A S 0 N D J F MA 1976 1977 1978

Fig 3-1' The percenbo composition of mior invertebrate prey tpes In the diet of S mImdu and in pttfulitcgctitIon/Soil oru. smp1e from The Wilderness LEGEND a--. Diet —u Pitfalls D—Q Veg./ioul cores A

P.JSFCT I ARVAF

ADULT DIPTERANS

%2o - - _A. - - - i_-- ;- - 0 30

0 80 ADULT COLEOPTERANS

40 - 20 .- - -

0 JASON DJ FMAMJ JASONDJ FMA 1976 1977 1978

FIg. 3-13 A The percentage composition of the major invertebrate prey types in the diet of S araneus and in pitfall and vegetition/soll core samples from The Wilderness B WMBRICID$

- ---a A... - -a---'- 0 - I SNAfl.S

A------IA - I 0 I SOPODS

20r CHILOPODS

, % / sa_ ------.J1 0 80 ARANEIDS &OPILIONIDS

- - i - i I t I I I I i I M J J A S 0 N F M A 1976 1977 DI.) 1978

Fig 3-13 B The percentage composition of the major invertebrate prey types in the diet of S araneus and In pitfall and vegetation/soil core samples from The Wilderness. - 85 - samples and the diet, and. their occurrence in the diet reflects their abundance in the vegetation/soil samples rather than in pitfall samples. The major component of the coleopterans in the dJ.et and. their greatest numbers in samples comprised small staphyliiuds commonly found. buried. in vegetation and surface soil.

Isopods were another major component of the diet at all times but their peak percentage in invertebrate vegetation/soil cores and in the diet do not correspond, except during May and. June when they occur in large numbers both in core and. pitfall . Isopode were common in sazples of invertebrates throuiout the year and. their occurrence inthe diet may depend. upon the relative abundance of other food items more palatable to shrews.

Araneids and opilionid.s showed. a large increase in numbers in pitfall samples in May and. June 1977 which was not reflected. in the diet, even thoui araneids were a common constituent of the diet. The apparent discrepancy between availability and. their incidence in the diet in May and June 1977 may be explained by Pernetta (1976b) who found large n'unbers of spiders of the family Lixiyphiidae in invertebrate samples whereas gut samples contained mostly lycosid3 and olubiomids. Thus the predation of certain invertebrates by S. araneus may reflect taste rather than abundance. Consequently, a surfeit of certain food. items may not be exploited. while less common but more palatable ones are, a varied. diet being preferred. Insect larvae, for instance, seem to be selected. for:they were taken by shrecis in much larger proportions than they occurred in vegetation/soil cores and pitfall samples.

This may also be seen when examining the graph for lumbricids, comparing diet and availability. When lunibricids were very - 86 -

numerous they did riot appear correspondingly as frequently in

the diet, whereas when numbers decreased, their incidence in

the diet simultaneously increased. Although they contain a

high proportion of water, lumbricids are rich in energy and.

possess little indigestible material apart from chaetae, so that

shrews might be expected to feed. on them in large numbers when they are abundant rather than on highly chitirious items; however this was not found to be the case.

Chilopods, gastropods and hemipterans occurred in small numbers in invertebrate samples and in similar proportions in the

diet, although fluctuations in numbers of these prey do not show corresponding increases in the diet, except to a small extent with hemipterans which are more seasonal in occurrence than the

other groups, with peaks in the summer..

Comparing Fig. 3-4 with Figs. 3-7 arid 3-10 (pie charts) showing the composition of the diet and invertebrate fauna in

different seasons, no correlation is apparent except for adult

coleopterans and the less numerous groups such as molluscs, hemipterans, and. ch.ilopods. Lumbricide were the largest fraction

of the invertebrate fauna in spring 1977 but occurred rarely in the diet. Insect larvae, which were the major component of the diet at this time, appeared in smaller proportions in the invertebrate samples, and. similarly in summer 1976.

This seems to suggest that prey items which are the most numerous in the habitat, such as coleopterans and lumbricids, are only eaten in certain numbers and so their occurrences in the diet and in invertebrate samples are not closely correlated.

Exceptions are apparent where extreme fluctuations in the - 87 -

populations of prey items occur and then their incidence in the diet and in field samples may follow similar trends. On the other hand, prey items appearing frequently in the diet, such as insect larvae, are taken in larger numbers than their proportion in the soil. This suggests that they are highly preferred and. actively selected for.

A plot of the density of different invertebrate groups - against their frequency of coccuirence in the diet (Pig. 3-14 A&B) further emphasises that the proportion of different prey types in the diet does not generally depend upon their abundance. (Adsit dipterans have been excluded because of their underestimation in. samples). Although some prey items may be present in considerable numbers in the habitat they may not appear in the diet at all, for example, ].umbricids. Only two cases (of snails ard. hemipterans) were found where no prey items occurred In either the diet or in the field samples. However, reference to Fig. 3-14B shows that there is a relationship between percentage frequency of occurrence in the diet and. abundance of ne prey type, namely adult coleopterans, and the calculated regression line has been included to illustrate this. As has already been mentioned, adult coleopterans constituted. a major item in the diet of

S. araneus in all seasons and Fig. 3-14B indicates that at times of peak abundance they are greatly exploited by many shrews.

Overall, there is no sigoificant correlation between diet and seasonal food availability,which suggests that food is always abundant and that shrews are expert at locating it.

This is further borne out by Hezhzherin (1958) who states that the main prey items (in his case, coleopterans) do not undergo great quantitative changes in abundance but their species a

C3 v, —a N

- z lu 0 N 'U Ø _a. o6 .-l4 o -u N -ø

.E< N N

çC4 NE I.. 1:: 4;

I- 4) I- 0. '4- 0 '06 —c 0) 0

0 6

o 0 0 0 0 oo o 0 N up u; uaiino jo Xuanbj o.. 0

cp N

-z N Oc o — a-

- . N

N N

ON NE

4) I- 0.

r o C 0) U '-4 0 "4 _I U 0 o. N -S

9 I.

0 0 0 0 0 06 0 .0 N 1P uu uajjno ;o XDuanboJj % -88- composition alters, Food items of minor importance undergo considerable quantitative changes. Results of invertebrate sampling and a knowledge of the diet of shrews obtained in the present study show that prey items are indeed always abundant, but in different proportions at different times of year. An examination of the number of available prey items taken coflectively and biomass measurementof between l-16.5g dry weight/rn2 shows that there is no dearth of food resources for shrews in the critical - periods, such as autumn when the shrew population decreases, in winter when environmental conditions are harsh, or in spring when shrews become active on the ground surface again and breeding commences, There is, however, considerable variation in biomass between years which could partly explain differences in population levels of shrews in successive years. Rud.ge (1968) reports that S • araneus is an opportunist feeder whose diet is governed more by changing availability of prey than by 'taste' or 'preference'. Results of the present study suggest that it is indeed an opportunist feeder, exploiting the most commonly occurring invertebrate groups and. therefore the most easily located ones, such as adult ooleopterans, araneids and. opilion.ids, isopods and. lumbricids, throughcut the year, and. supplementing it with more seasonal groups,for example,hexn.ipterans and adult dipterans.

However, an element of taste or preference is shown for some groups of minor importance, such as ohilopods and. gas tropod.s, which are ever present in relatively small numbers, and. others, such as insect larvae, which are not active on the ground surface and, therefore, require more diligent searching.

Pernetta (].976b) suggests that competition between S. araneus and S. minutus is reduced by a considerable dietary separation of - 89 - these two species by virtue of the surface activity of prey- and its size. The limited, data presented. here show that S.. minutu does indeed tend to feed. more on surface—active fauna with the exclusion of certain burrowing foins such as luxnbrioids, and. a greater proportion of the prey tends to be of a smaller size. Tbts is further shown by News tead. (1947) who found incidents of S. minutus feeding almost exclusively on greenfly of the genus Siphonophora. Such dietary separation has not been demonstrated. by - the larger species, N. fodieris, except when found. in aquatic habitats (Churchfield, 197 in press) and. this, together with the 1xow1edge that prey items are always abundant, tends to reduce the aiiificance of dietary- separation in the competition between shrew species. -90-

STION B: CLIMATIC CONDITIONS

1) Introduction Great differences of opinion exist as to the effects of adverse weather conditions on the mortality of shrews, owing to the lack of conclusive results on this subaect, Nezhzherin (1960) concluded from his studies that climatic factors were not instrumental in the regulation of population densities of shrews. -

However, Yudi.n (1962) mntained that the autumn reduction in numbers is caused by death under unfavourable weather conditions, and this view was also held by Shlflito (1960) who thought that mortality was caused by a decrease in temperature. Nevertheless,

Adams (1910) reported that shrews were frequently caught in frosty weather but rain caused a decrease in the numbers of captures. It is possible that weather affects activity and hence the trappability of shrews rather than their mortality. During the trapping programmes in The Wilderness and. Alice Holt Forest, meteorological data were recorded in an attempt to investigate the eifects cf climatic conditions upon seasonal captures of shrowa. ii) Methods A meteorological station in close proximity to the study areas was available for collection of data for both The Wilderness and.

Alice Holt Forest. Mean monthly maximum, m'nimum, grass minimum and. soil minimum (20cm. depth) temperatures were recorded, together with total monthly rainfall. The same data were recorded for each day of the trapping periods. In addition, maximum and minimum temperatures were recorded from the study areas themselves. All records were made at 0900 hours. - 91 -

iii) Results & Discussion

The maximum and minimum temperatures recorded from the weteoro1ogica.l stations and. within the study areas are compared in Figs. 3-15 and. 3-16 for The Wilderness and Alice Holt Forest respectively. The mean maximum and. minimum temperatures for the duration of each trapping period, are shown. Although winter minimum temperatures were slightly higher within the wood. at

Alice Holt Forest than at the 'neteorological station, the -

differences between data from the meteorological stations and the

study areas are small: from 2.10 to 5.1°C for mean minimum and

maximum temperatures respectively. In view of this, the results

obtained from the meteorological stations will be discussed. in

further detail, because they were the most comprehensive records

made.

The mean monthly temperatures and. total rainfall taken from The Wilderness and Alice Holt Forest meteorological stations are

compared with seasonal captures of S. araneus in Pigs. 3-17, 3-18

and 3-19, from which a relationship between temperature and numbers of captures is apparent. An annual cycle in temperatures is found. which corresponds to the seasonal cycle in shrew captures, with large numbers of shrews being caught in summer when temperatures

were at their highest and. vice-versa in winter when temperatures were low. However, few shrews were caught at Alice Holt Forest

during the summer of 1977, despite temperatures being similar to previous years. Numbers of captures the preceding winter were

larger than the previous years, a marked decline in numbers not

being apparent until the temperatures were rising in spring and.

summer. It seems doubtful, then, that temperature directly

affects population fluctuations of shrews.

4- -- a ,. -u v-u c 2 z u uJ

Lu '9

-) a

4) z a) 0

dv,,' 6• 6 •ø . 4) 0. Ii -),- a)

CE4

—

U- H

r..

I, ii 0. II 'I

I I I I I N—. 03 0 0 O C4 03 ' 0 I N (N N '-

Ddw1 wnwIxow uoew D dwj wnw,xo uD3W 00 00 00 - 0 z a 'S UJ 'S 0. 'S - z 'S 'S 9 6

- - -

• -

'S /

z I 'S 0 'S "S U) ('I, S. d S. U 2°

S.. .5 II IL I-

'5_ ____

4 <( -) j I ' I '0 0 CO ' 0 '0 (N CO "1 0 (N (N -

-% dwj wnwiuiw uoew dwej wnwixo uoaw -0 sneuoio g o sain4do) 0 N N N 0 I I I I I- S — UE - o g c.1 - • E o E E_ Z S LU DEE o - - -J 0 z 0 U)

k 0

/ U-

—)

0 C,

'1) z I.' (i 0 U) .(

— C.. 'I

U-

c 'i)

j 1z I I I I I I I N N N0 N 0 4 D0 QJfl40JadWj uOGW

snauoio o soJn4doD

Co o N c) N Co 0 E V- o c -u o E

0 LU -I Un'

------.------Or..

C;: I :

Co 0 N N N D0 ein4oiadwej uoe\y

snuoJo g ;o s3JnIdo) o o 0 C ) - 0

- -D 1 IL 12 a z- w - C o 2 uJ / -j - - - ::z , p. - Ij

0 I 'U. I. .4

0 0 4) c I-

I- -i dO 02 0 0 U)

a_cç_ -., 'in L L4

U-

o_ 0 00 0 0 ' 0 CN c1 - (Ww) J0UI0 10401 - 92 -

Worthy of note is the fact that soil temperatures, recorded at 20cm depth, were siu.ficant1y 1uier than mean minimum or

grass minimum temperatures at all times of year, although they

showed similar seasonal fluctuations. Voles and. woodmice were found to burrow to at least this depth in artificial enclosures in laboratory studies, and it is likely that shrews, which are not

good burrowers, inhabit tunnel systems of other small mmmi s.

Population fluctuations, reflected by numbers of captures, may

be correlated with rainfall for, at Alice Holt Forest, shrew captures

decreased. when rainfall was at its greatest. For example, capturea

were low in the wet summer of 1977 compared. with the hot dry sunnner of 1976, when shrews were numerous. A similar situation is not apparent in The Wilderness, however. Overwintering numbers were

low in 1976 and 1977 when rainfall was high, but the wet summer of 1977 still produced. a population of shrews larger than the previoua year, and. the subseqtient cold, wet winter also proauced larger catches than the previous year.

Numbers of captures of shrews may be related more closely to the weather conditions prevailing during the trapping periods than to mean monthly estimates. Excessively cold or wet conditions during

hours of trapping could reduce the activity of shrews on the ground

surface leading to fewer captures rather than increased mortality..

Pigs. 3-20 A & B show the relationship between the numbers of

S. araneus individuals caught in The W1derness and mean grass

minimum temperature and total rainfall respectively d.uring each

trapping period. Summer and. winter trapping periods are

distinguished. Again, captures were high when temperatures were

at their greatest and. rainfall at its lowest, especially in summer,

but declined with a reduction in temperature and. increase in rainfall U0. 9< > o. -J

E E F 0) 0 - w

446n0) s1onpuAIpu 801

0 .4 ft

:11

00 Cl t1

4 LIBflO, SIWnPtMPUI 6oi - 93 - in winter. Beat loss is accelerated when the fur is wet and. so it is conceivable that shrews would reduce their surface activity during rain, especially when temperatures are low, and remdn in subterranean burrows. Exceptions are evident on the graphs, for example, in January 1976 and. March 1977 weather conditions were exceptionally mild, and. there was no rain in the March sampling period, but no shrews were caught, whereas in September 1977 rainfall was high but activity by shrews was great and. captures numerous. In April 1978, despite low temperatures and high rainfall, large numbers of shrews were captured: at this time maturity is reached and shrews are active on the ground surface searching for mates and. this occurs regardless of weather conditions.

Very little is known about the effects oi' daylength on the seasonal activity and habits of small mammals. Fig. 3-21. shows the relationship between daylength and numbers of captures of S. araneus in Alice H0lt Forest and The Wilderness combined into a single 12— month period, from which a correlation Is apparent, mainly in autumn and winter. However, there is a much closer correlation between daylength and temperature evident from 'ig. 3-21.. It seema doubtful, therefore, that d.aylength has much effe't on the behaviour of enrews, since they are active both by clay and bT night, unlike most birds whose seasonal habits are closely linked via their hormone systems, to the photoperiod. (Lofts & Murton, 1968; Murton & Kear, 1978).

It is apparent from the results that seasoua]. changes in temperature and. rainfall are accompanied. by similar fluctuations in numbers of shrews, but that climatic conditions cause these fluctuations seems unlikely. Daily captures of shrews tended to be affected more by high rainfall than by temperature, captures in (sino4) 446uaIXoG

•. E V I- - a_c 0

?1 ? "a

U) co c L- C) U) —

) U)

a I,

o. a

dwai/ saindø °N most cases being reduced. during rain This is contrary- to the

findings of Boroweki and Debnel (1952) who observed. an increase in numbers of captures, arid, hence activity, of S. araneus during rain, which they explained as an increasing need. for shrews to forage for

insects which the ram renders inactive and. so more d.ifficult to

locate. However, Nystkowska and Sidorowicz (1961) found no connection between the numbers of captures and. the activity of insects, although they also found that the numbers of captures

were almost doubled by rainfafl..,

The correlation between climatic conditions and. numbers of

captures may be interpreted, in two ways:(a) reduction of captu.res

in winter may be equated with mortality cauned by adverse conditions

or (b) fewer captures may be an indication of reduced. activity on.

the ground surface when conditions are cold. and wet. Numbers of shrews caught in winter months were consistently low, together with

temperatures, and. rainfall tended. to be high, therefore it is

difficult to ascertain what effects changes in weather conditions have on the behavi our of shrews in winter.

Studies on the effects of climatic conditions on other small mammals indicate that they do influence the activity of these

animals. For example, the work of Evans (1942), Brown (1954) and.

Tanton (1965) on Ar,odemus sylvaticus showed that cold. conditions had. no detrimental effect on numbers of captures, but that numbers

increased if there was a sudden rise in temperature. Brown (1956) also found. that the activity of Apodemus, Clethrionomys and.

Microtus above ground was more influenced. by rainfall than by

temperature. Gurnell (1975 & 1976) found that cold conditionø, particularly a combination of cold and wet conditions, reduced the numbers of captures of A. sylvaticus, and. their activity on the ground surface was reduced. - 95 -

It is probable that climatic conditions do affect catches of shrews, particularly in winter, but a direct correlation between weather and. mortality cannot be ascertained. Laboratory studies on captive shrews, to be discussed later, show that they are hardy and. not susceptible to death from cold. However, it is known that one of the most important factors stimulating the activity of small marmals is temperature: on the basis of numbers of captures, Mystkowska & Sid.orowicz (1961) found that shrews were most active between the minimum temperatures 13. 1-16 .0°C and at over 27°C for maximum temperatures. These accord with summer temperatures in Britain, when shrews are breeding arid population numbers are higt. The effects of adverse climatic conditions, therefore, majbe to modify behaviour, particularly activity on the ground surface, rather than increase mortality. This will great1 influence numbers of captures, particularly in winter, and. hence the decline in population density of shrews, based on numbers of captures, may be grossly exaggerated.

t - 96-.

SECTION C: PARASITES OF J1BEWS

i) Introduction In his study of the population ecology of S. araneus in

England, Buclmer (1969) suggested that the incidence of infestation by parasites may have a significant bearing upon the density of shrews. Be found that the level of infestaton by Porrocaecum s. a splrurLd-type nematode found encysted as larvae in the gut - mesentary and connective tissue beneath the skin, could be correlated with the decline in the population numbers of shrews in winter. The percentage of individuals infested with Porrooaecum ep. was greatest between November and January whereas between February and August the parasite was not recorded at all. Re postulated that the parasite may secrete toxic substances and. may cause irritability or unrest in its host, mkirig it more subject to predation, and thereby contributing to a decrease in population numbers. Tudin (1962) in U.S.S.R. also found higier levels of infestation by Porrocaecum sp. together 4th helminth parasites, in overwintering shrews than in juvenile shrews. However, he could find no adverse pathological or behaviours]. effects caused by these parasites. Borowski & Dehnel (1952) maintained that shrews do die in autumn as a result of infestation by parasites, and that this was the factor regulating numbers of young shrews.

It was dectded to undertake a study of the degree of infestation of shrews by endoparasites at different times of year to discover more of the possible significance of parasitism on the seasonal fluctuations of shrews. - 97 -

Method miring the course of the autopsy work carried out on shrewa obtained from Alice Holt Forest, all parasites were recorded and counted. Results were expressed in terms of mean numbers per host individual (intensity of infestation)and the percentage of the shrew population infested (incidence of infestation). These data were supplemented by autopsies performed on shrews which died. in traps in the other study areas, including The Wilderness and. Monks Wood, and. the Airesford cress beds.

Ectoparasites leave the host as soon as the body cools after death. Consquently, few were found. on the shrews and. quantitative assessment proved impracticable, and. so the present study was limited. to an investigation of the endoparasites only. The seasonal dynamics of the tick Ixodes triangulices found on small mmnials including shrews, has been dealt with by Randolph (1975a, l975b).

Identification of the parasites collected, from the git, body cavity and beneath the akin was limited by the state of preservation of the host and the parasites within it at the time of autopsy.

However, it was the purpose of this study to investigate levels of infestation by particular groups of parasites rather than the biology of cpeciflc examples. iii) Results

A list of tbe parasites identified, the species of host and. the mean and range of uxnbers found per host individual is shown in Table 3'-4. -98..

S. araneus S ininutus N. fodiens No. Shrews examined. 157 23 10 Presence Mean Range P. Mean Range P. Mean Ran

MATODES

Staminerinema soricis X 15 1-60 X 10 1-15 X 3 1-5 Porrocaecum sp. X 3 i-ia DIGEMEANS

Brachylaimus oesophagei X 3 1-16 X 1 1 CESDES

Choanotaenia Crassiscolex X x 21 1-275 13 1-93 Hymenolepis Lingularis X x x

Soricinia sp. x Vaxnpirolepis sp. X 17 4-37 Triodontolepis bifurca I

Coronacantha ep. x

Table 3-4: End.oparasites identified from shrews

All parasites, with the exception of Porrocaecum sp. were located in the gut of shrews. Porrocaecum ep. was found in the connective tissue of the body wall, particularly of the dorsal and shoulder region of S • araneus where it was common. This nematode was not recorded in either S. nariutus or N. fod.iens. Yudin (1962) also found. no Porrocaecum ep. in S. minutus but Lewis (1968) did record it in shrews in Vales, and. so did Grainger & Fairley (1978) in Ireland. No digeneans were found in S. minutus in the present study. Five different genera of cestodes were identified from the gut of N. fodiens but fewer were identified from S. araneus and. S. minutus, possibly owing to the poor state of preservation of the specimens found. -99-

a) Seasonal occurrence of parasites The intensity and incidence of infestation by nematodes,

digeneans and cestodea in S. ararieus cauit in Alice Holt Forest

at different times of year a.re shown in Table 3-5 and. Fig. 3-22.

Digeneans and. gut nematodes tended. to have peak occuirences

in summer months. Cestod.es were more numerous in the summers of

1975 nd 1976 than the winters but, with the exception of October -

1975, the incidence of infestation was high at all times.. Porrocaecum ap. reached peaks in January and. May but had. dimfriished in both intensity and incidence of infestation by August, and remained low until after December in the years sampled. The greatest degree of infestation of all parasites corresponded with the rise in population size of shrews, with the exception of January figures for Porrocaecum ap. When shrew captures were low, as in winter, parasites also tended to be low in intensity and incidence of infestation.

The results for autopsied S. araneus from The Wilderness and.

Monks Wood study areas (collectively) are presented in Table 3-6 and.

Fig. 3-23. The major difference between these shrews and. the population from Alice Holt forest was the incidence of infection by

Porrocaecum sp. and Staimnerinema sp. which was very low in the former study area where these nematodes were only recorded for two and. three months respectively. Only forty one shrewa were e y m1nec1 from

The Wilderness and Monks Wood. whereas 116 were aut.psied from Alice Holt Forest, but this should not have affected the rasu.lts unduly, for other parasites were recorded in shrews from The Wilderness and

Monks Wood. Levels of infestation by d.igeneans were similar to those at Alice Holt Forest, but cestodes had a lower intensity and incidence in The Wilderness and Monks Wood. Cestod.es showed summer C4 b 0 C/) C4 t: 0 i ti 0 C) t'i xj ç) 0 '-4 C) -31 0 I-i '-3

(iJ -4 -4 Vi

I-' 0 0 0 0 0 I I • • • S • S S S S • -J (J t' -J -J (J N) - Vi N) Fl I j -1 10 N) L.C/)0 10 I C) Vi 1') L'J Vi N) N) I-J 0 0' 0 Fl Li-' - N) 0 0 8 0 0

0 Lii H) (11 Fl I-' N) N) a Fl 0 0 O\ N) N) 03 -3 H) • • S S S • S S • (jJ -3 I 03 0 d J 0 - - CCJ)0 CD Vi 1-3

Fl Fl Fl -3 0 N) O 0 O\ - 0 tjJ '0 0 Vi 0 Vi Vi 0 0 0 0 • S • • (1) H) 0 0 0 jJ - 0 -4 0 ' I-.'

Fl N) 0 0 Fl Fl Fl N) Vi • S I I S S S S S I I • CD 0 t' - 03 -3 -4 Fl 03 N) Fl ccao

I-I I- r 0 N) LJ O - Vi 0\ 0 - Lii 0 Vi I- 0'. -4 0 0 0 0 0 0 0 Lii 1(13 • S S • S • • S S S (1) 0 0 0 0 - '0 0 0 0

CD I-I Fl Fl N) - ('.3 0'. Fl frJ 0'. 03 - - Fl I-' Vi 0'. N) -4 1-1 0 • S S • • S • S S • S 0 (.J tJ Vi 0'. - - Vi 0 0'. 0 0 C!)

0 I- I-' Fl F-' - I-' Fl 0 0'. -3 03 0 0 - 03 0 0 0 0 0'. Vi 0 -3 0'. 0 0 0 0 4 .0 S S '.0I •0 S • • • S S S 0 -3 0 Lii 0 0 03 - 0 0 0 Fl

Fl I-' I--' . I.•J - '.0 -4 CD Vi Lii N) Fl LJ - (1)

pJ F F-1 -4 ) Fl Fl I-' Fl - 0 0 0 0 0 0 -3 '.0 0 0 0 (J 0 "j'-3 o o 0 0 0 0 -3 0 0 0 - 0 ti0 • . S S S • S S S • S 5 5 (/)1-3 0 0 0 0 0 0 0) LJ 0 0 0 (JJ 0

(o- -4 G4S33UI SMa14S % 0 0 0 V ç o2 ri 9 I ,1 S ,1 I —S ,I S ,I S S , I S 1 / S i: '( b ' I ' ' 'I S ' ' ' - ' % I I -I dI ('1 _D - 9, - - 0

-p -(5,) /

I S /1 '7 I 'I, S F I ci' 4 S t S S S S S I S U S II- S. 0 6 • w

0 a---... cq E E U 0

' (I, z .. LU0 - 'U I '0 I' z / .-, I < 'U \ I - •_o 0 U.' I, 'U • 'U - a ,.z, I _O - \1 d -- / -S

N ON 00 0 00 0 0 0 0 ' N '0 N

(w—*) IOflPIAIPUI so ad s,sojod ou uoo

'-3 U) . 4 C..4 0 . t: Lii c: c . '-4 CD I-' \O '0 - -4 tr --3

tjtIjtii I-I i-' Ia C.. f3 I-4 CD • I I I • I I I I I I I t:i0 0 0 cZcno IC) CD '-a Cl.

S. I--a ( 0 0 0 0 0 0 0 0 0 0 0 I_I. • CO 0 0 I-a pa U) CD C!) C) I-' CD I I I I I I I I • • I I- I 00 0 1-4= o kzj crio I '-a I-3 li—i

4) CD H CD laJ Ui 0 0 0 0 0 0 0 0 0 0 Ui . cD is-a 0 I-a 0

o 4-43

U) I-' Ui (.0 r'.) I- I-I • S S • S • I I • . S i-i •..frJ 0 0 0 0 0 0 0 I-40t ID '1

CD '-4 I C!) CD Ic::; ) - 0 Ui UI UI I\) 'i Ca - 0 0 0 0 0 0 0 0 0 Ui L'i. 00 0 0(,j0 CD

I_i I_a I-i a - - '--a c - I--a c U' CD S S • 5 I . . • I • I-40 0 I--a 0 Ui 0 0 0 0 0 CD CD Lii '-a Cl) I-I I_a I-' '-3 O\ -)0 0 00 00 0 0 UI0 0\UI C\0 0 0 L'3Ui 'ii CD • • • • . • . • . U) 0 0 0 0 0 0 -4 0 0 I—i Pa I-a I-a CD I-' 0 I-' I- r',) I--' t) I'.) UJ I\) (.0 - C/) '-3 CD 1-I CD

a. Cl) (

o I-' I-' I-' '-0 - 0 0 0 U-I 0 0 (.o - '-3 0 0 0 0 0 0 0 0 0 0 UJ Ui Co S S • • S S • S • • 0 0 0 0 0 0 0 0 U' 0

o 0

(o--a)pasa;ui SMIL1S % It) 0 It) 0 N. It) C4 0 o -- 'I) - - - - S ------____,

-, S - - N. - - - 13.

.. I ::_ - - -) - I , -- / z I 0

- I - '0 -

— S. b / / / / / I I!) 1 In / z 'La 0 / 'U 0 z I- 'U In I u / o U / '-5','

C4— CO OCO 0 '1 i C (-)IonpIAI puI ;sotj iad sausoiod ou -100- peaks in incidence but the numbers of parasites per host individual were very variable. Similar results were obtained for digeneaxis, with the exception of January 1977 (when only 2 shrews were ermiried, one of which was infested), and. July 1977 (one sample eimined., but no parasites were found).

S • minutus was caught at different times of year, and. the results of infestation by parasites are shown in Table 3-7. Rowever, the - numbers of shrews caught and. autopsied. were too m1l to give siifioant results of a seasonal nature. A fuller account of the parasites of S. minutus is given by Grainger & Fairley (1978) 'who ¼ found infestation levels of most d.igeneans and cestod.es were lower in overwintering shrews.

All N. fod.iens were caught at the sane time of year, consequent1y no seasonal data are available. b) Incidence of infestation in different age classes of shrews The incidence of infestation of the j-uvenile, sub-adult and. mature adult shrews eainined. is given in Table 3-. 8. TIus includes data from the Alice Holt Forest and. The Wi1d.emes/Monks Wood. study areas separately for S. araneus and. combined for S. ainutu,and. from Monks Wood and the Alresford cress beds combined for N. fod.iens.

The parasites in S. araneus from Alice Bolt Forest had. higher levels of infeatation In adults than in juveniles and, with the exception of cestodes, infestation was lower in sub-adults than in other age classes. It might be expected that juveniles recently out of the nest would. not be heavily infested as the source of most of the parasites encountered in this study is invertebrate prey items which act as intermediate hosts. In winter, intermediate hosts may C-4 0 C-. c c

cD I-' I-. -4 -4 -4 —1 S.

bt'i cD I I I I I I I I I I I ço a I-. c+ (Des

CD 0000a:oo::aao Wa CflI pa CD C) CD 0 -' '-I) ''III 0 0 0 0 0

o o 0 0 0 8 8 0 8 8 0 0 8; I I I I I I I I I I I I I I0

o o 0 0 0 0 0 0 0 0 0 0 0

o I I I I - 0 'Ji 0 n 0 0 _71 b0

'-3

r LJ I-' 1\) I- I-' I I- -

I I- lJ I-J i-' i-' OOOOO0000000Ogo o 0 0 0 0 0 0 PxI:3 o 0 0 0 0 0 0 0 (a t Cfls- 0 I r-04 i.- ($c-

L.J LI I-I Lxi 1L1i I:).o Co (12 ($2

-4

C) I-i -2 p. 0 . -3 CD • • 0 a. o '.o CQ CD CD (12 o 0 '-I Co t.. i-i r 0 I\) _ S . . Co 0 0 %J1 I 0 C CD (11 'i (12 (a CD (a

'0 \0 O 0 %Ji 0 J) 0\ S . . '0 -2 (13 o- -.3 -3 r) 0 -.2 CD p. - - Ca I-i (13 CD -C$2 OP. O(D . - 0 P1 C/) CD (a (a ($2 - 1%) 0 - -4 0 . '0 '0 (12 (a (0 C) -.3 0 Co I-' • 0 < o '0 ¼.fl (12 CII H3 5 (a - 0 -.3 0 Co CD • 5 '%•fl I-n Cl] (a I-_n I-_n p 0 0 0 S CD 0 0 riD p. I-J 0 0 N.) N.) o 0 - • S • 0 0 0 Cl] (I] P 0 0 0 0 ID (0 0 b(a I--n -3 (/2 0 Ji 0 0 S •0 0 (a - riot be as abundant, and so occur less frequently in the diet of shrews.

This could result in a decrease in infestation of overwintering sub-adult shrews. A similar situ.ation was found with S. araneus from

The Wilderness/Monks Wood where sub-adults were less heavily infested. with dJ.geneans than adults and. juveniles. Adults were the most heavily parasitised group at Alice Holt Forest, probably as a result of an increase in the diversity of the diet and an increase in activity of the shrews in the breeding season. In summer, when the population of shrews was large, infestation zniit be expected to be high owing to the ease of spread of infection.

The results for S • minutus and. N. fod.iens are not conclusive owing to the small number of shrews erimined, althoui sub-adulta showed. a decrease in the proportion infested, compared with juveniles and. adults. c) Infestation of different sexes of shrews The incidence of infestation of mal3 and. female S. araneus,

S. minutus arid N. fodien is shown in Table-3--9.

S. araneus Wilo.ernes 3/ Alice Holt Forest Monks Wood. s• minutus N. fodiens M F M F M F M I Total ey rnined 64 52 17 24 11 12 8 2 AT0DES

Porrocaecum ap. 33.8 13.3 11.8 4.2 0 0 0 0

Staznmerinema soricis 63.4 58.3 11.8 4.2 27.3 50.0 37.5 50.0 DIGEANS Brachylaimus ap. 53.5 40.0 35.3 41.7 0 0 12.5 0 CESTODES 76.1 73.3 70.6 54.2 54. 6 66.7 75.0 50.0 Table 3-9: Incidence of infestation (9) of male & female shrews - 102 -

Results of seasonal levels of infestation of male arid female S. araneus from Alice Holt Forest are shown in Fig. 3-24. There was little difference in levels of infestation between the sexes, althou,gh a higher proportion of males tended to be parasitised than femalea in S. ararieus, and vice—versa in S. minutus.

iv) Discussion The results of this study show that there is a correlation between the levels of infestation by parasites of S. araneus and the population density of shrews, but contrary to Buckner & Yudin's findings, the decreasing levels of infestation are associated with a decrease in population density of shrews. Consequently, infestation by some parasites at least tended to be lower In overwintering sub-adults. Porrocaecum ap., the example quoted by Bucicier, was found to be present at all times of year in Alice Holt Forest shrews, but tended to decrease In intensity and incidence in antumn wnen the shrew population was declining in numbers. In The Wilderness and Monks Wood, Porrocaecum sp. was not commonly ibund, and. yet a marked seasonal fluctuation in shrew numbers was evident.

The results of this study also substantiate the work of Lewis (1968) in Wales and Kisielewska (1963) in Poland who found that levels of infestation of helminth parasites in shrews decreased in winter. Kisielewska found that infestation by the commoner cestodea, Cjanotaenia crassiscolex and. Hymenolepis scuti gera was higher in juvenile and adult S. araneus but significantly lower in overwiritering sub-adults. This she attributed to a change in diet of shrews caused by the hibernation or inactivity of the intermediate invertebrate hosts of these parasites. This was reflected in this study by the numbers of cestod.es per host individual but not in the percentage of the population infested.

w ii .' -L 1< 1 , 1-'

- N

t lu. 4 co I •0 z

d ....' - U, 4 I

'1 'I

L4 I ' 2 U- I! C40 0 E 3 I 0- V - / 0 z . I 0- •lI .1 0 oe 1' 2 i 4- 0 0. I < i 2 I ' z , z , 6 8 L ___.L.. -.) 00 0

SGjISOJOd Cu uooye - 103 -

Lewis found no difference between the degree of infestation of shrews from different localities or different habitats, except for the cestodes C. crassiscolex, which was more numerous at his Skomer site, and H. singularis which occurred in larger numbers at Aberys t'wyth.

He also found. no siguificant differences in levels of infestation between the sexes of shrews, but a seasonal variation was rioted, especially of the d.igenean Brachylaimus oesophagei in accordance with the present study, and also of the two cestodes named above. All three parasites have one or two intermediate invertebrate hosts, and. the density and distribution of these hosts miit well have an effect on the infestation of the shrews which fed on them. B. oeso phagei infestation was low in winter, between October and. February, as in this study, and. so also was the population of the gastropod. Zonitoides excavatus, the intermediate host. Both intermediate host and parasite increased in intensity and incidence in spring and summer, according to Lewis. Likewise, C. crassiscolex showed a marked decrease in infestation in winter when its host, the gastropod Oxychilus helveticus was few in number. The reason for a lower level of infestation by parasites in The Wilderness/Monks Wood. may be a low density of intermediate hosts, particularly gastropod. molluscs which,in fact, were not often recorded in the diet of shrews in this study area.

It seems, therefore, that the seasonal infestation of

S. araneus and S. minutus is correlated with the density of shrews and the density and. degree of infestation of gastropods or other intermediate hosts containing the larval stages of the parasites.

When intermediate hosts are few in number shrews are less likely to locate and. eat them, and so infectinn is avoided. It could be argued that the diminishing levels of infestation by parasites which accompanies the decreasing population sizes of shrews in autumn

and winter is evidence that the parasites kill the more heavily infested hosts, thereby causing the decline in population density.

However, infestation levels were found to commence a decline three to four months prior to the reduction in population numbers of shrews, 3.ndicating that parasitism cannot be a major, or even a

contributory, factor in controlling popuJation sizes of shrews in the areas investigated during the present study. - 105 -

SECTION D: SIThIMARY

Diet of Shrews Analysis of faecal pellets from s:hrews in scrub—grassland. and. deciduous woodland. gave the following results:— a) S. araneus 1. A wide variety of invertebrates were eaten by shrews. 2. The greatest diversity of prey was taken. in winter. - 3. The major prey types in all seasons were adult coleopteraxis,. insect larvae, araneids and. opilionid.s and isopods. Intermediate in importance were lunibricids and. adult dipterans. Other groups, including hemipterans, chilopods and snails occurred. less

frequently in the diet. 4. The size of prey ranged from collembolans and mites less than 5mm in body length to lumbricid.s and. insect larvae greater than

20mm long. 40. of prey types were less than or equal to 5mm, but very few were smaller than 3mm. b) S. minutus 1. A wide variety of invertebrates were eaten but lumbricida were absent from the dieb of this species. 2. Ma3or prey types comprised adult coleopterans, araneids and. opilionid.s and insect larvae. Hemipterans and adult dipterans

were taken in hier proportions than by S. araneus but isopods and chilopod.s occurred. infrequently. Seasonal data were too few to give conclusive results.

3. 57.4% of prey items were less than or equal to 5mm, but few

were less than 3mm. o) N. fodiens 1. The small number of samples examined revealed that equal

proportions of staphylinids, adult dipterans, opilionids, isopoda chilopods and. lumbricids were eaten. - 1o6 -

Food Availability

Estimation of prey availability by vegetation/soil cores and pitfall trapping in scrib-grassland revealed, the following:-

1. The most numerous prey types greater than 3mm in body length in vegetation/soil cores were adult coleopterans, lumbricids and. ararieids and opilionida.

2. There was a predomirnce of small invertebrates of 3-5mm long, mostly staphylinids.

3. Pitfall trapping revealed. seasonal trends in numbers of prey, with peaks of activity on the ground surface by most prey in slimmer and. autumn. 4. Vegetation/soil cores showed small seasonal variations in prey types but differences in numbers between the years sampled tended to override these variations.

5. Estimates of biomass and numbers of prey per m 2 showed a difference between the years sampled, and. a peak in winter month

The major contributors to biomass were lumbricids and. to numbers were adult coleopterana.

The Relationship between diet and. food. availability

1. Little correlation was found between seasonal availability of prey types and their occurrence in the diet o1 5. araneus. There was no correlation between diet and. activity c,f prey on the ground surface.

2. Only one prey type (adult coleopterans) showed. an increase in frequency of occurrence in the diet with an increase in abundance.

3. The percentage composition of adult coleopterans, adult d.ipterans and insect larvae in the diet of minutus did reflect I the activity of these prey on the ground surface.

4. It is suggested that shrews feed primarily on the most commonly occurring prey types, such as adult coleopterans, which are abundant throughout the year, and supplement the diet with other invertebrates - 107 when available. Some degree of selection of prey, particularly insect larvae, may also occur.

5. It is concluded that there is no shortage of food in autumn and winter which could account for a population decline of shrews.

Climatic Conditions

Records of meteorological data and. captures of shrews from the

scrub—grassland and. deciduous woodland study areas showed the - following results:-

1. A relationship between numbers of captures and temperature and rainfafl. was apparent, with a tendency for high numbers when

temperature was high and. rainfall was low (as in summer), and low numbers when temperatures were low and rainfall was high (as in winter).

2. Changes in temperature and. numbers of captures showed similar seasonal trends but were not synchronous.

3. It is suggested that weather conditions influence the activitT of shrews on the ground surface rather than mortality.

4. A correlation was found. between seasonal numbers of captures arid. daylength, particularly in autumn and. winter, but a much closer correlation was apparent between seasonal dayleEgttl and temperature.

It is suggested that day-length does not greatly affect the behaviour of shrews.

Parasites

A study of the internal parasites of shrews from t'e deciduous woodland and scrub—grassland habitats provided. the following results :-

1.. The intensity and. incidence of infestation of S. araneus by digeneans, cestodea and nematodes showed a seasonal cycle, with a tendency for peak occurrences in summer, when shrew numbers were high, and reduced occurrences in winter when shrew numbers were low. - 108 -

This is attributed to the ease of spread of infection in summer, enabled by increasing density and activity of shrews arid certain invertebrate prey which are intemned.iate hosts.

2. Adult S. araneus were found to have higher levels of infestation than juveniles and sub-adults.

3. Levels of infestation were not found to differ siiificant1y between the sexes of shrews.

4. The numbers of S. minutus and. N. focliens autopsied were too small to provide seasonal results of parasite levels.

5. It is suggested that parasitism is not a factor in controlling population density of shrews, and does not contribute to the autumn decline in numbers. - 109 -

CHAPTER 4

BIOENERGETICS STUDIES ON SERiS - 110 -

SECTION A: PDOD CONSUMPTION & ASSIMILATION

PART I: INTRODUCTION

There is a wealth of conflicting information about the metabolic rate of shrews, based largely on measurements of oxygen consumption, which was generally assumed to be high in these small, highly-active animals on the principle that the smaller a mammal is the higher its metabolic rate must be. Pearson (1947) found a correlation between body weight and metabolic rate, but the high degree of indi.viduality tended to mask this tendency. However, he did report that the long-tailed shrew, Sorex cinereus cinerev^ had the highest metabolic rate of all the small mmm1 recorded. Norrison, Ryser and. Dawe (1959) subsequently reported. that the basal metabolic rate of the masked shrew, Sorex cinereus, was no higher than other small mammals, and Hawkins, Jewell & Tomlinson (1960) found. no difference in metabolic rates of shrews and. laboratory mice of the same weight. Like Pearson, Gebczinski (1969) found. that shrews had. the highest level of oxygen consumption of all the mammals investigated, and. this was confirmed by Vogel (1976) for the Soricinae and. by Gebcznska and Gebczynski (1965) who found that oxygen consumpt n by Neomys anomalus and. N.. fod.iens was 22) greater than oxygen consumption by rodents.

This apparently high metabolic rate was not due simply to smaller body weight, for a comparison of oxygen ccnsumption by Sorex araneus (G.ebczynski 1965) and. by the har7est mouse of a similarly small size (Gorecki, 1971) still found the former to be consid.erably higher. French et al (quoted by Grodzinski & Wur4er in Golley, Petruzewicz & Ryszkowcki, 1975) found a decrease in - 111 -

Average Daily Metabolic Rate of small mammals with an increase in

body weight and although they observed a considerable overlap between oxygen consumption of larger insectivores (such as the shrew Blarina brevicauda) and the smaller rodents of approximately

20g body weight, shrews mostly had. a higher metabolic rate than rodents. Recimond & Layne (1958) also found that the metabolic rates of certain tissues (including liver, kidney and lung) of the

shrew Cry1)totis parva were not high on the basis of size of the animal alone, and suggested that the high rates of oxygen consumption were due to extrinsic factors such as nervous

stimulation. This is also suggested by the work of Kend.eigh (1945), Pearson (1948), Morrison et al (1953, 1959) and Layne & Redmond. (1959) who found that body temperatures of shrews were no higher than in other mammals.

Fewer reports are available on food consumption and.

assimilation in shrews as an indication of metabolic rate, and. their relationship with body weight and. temperature. There is no doubt

that food consumption by shrews is high, as has been confixmed. by

Hamilton (1930) and Morrison, Pierce & Ryser (1957) for American shrews and by Tu.pikova (1949) and Mezhzherin (1964) for Euroasiatio shrews, and. numerous others cited by Pernetta (1976a). Rudge (1965) found British shrews consumed between 65% and 12CPA of their body

weight daily, Wi. iii smaller shrews of 5-8g consunwig a higher proportion of their body weight (9901°) than animals of 9-12g (72%). Hawkins & Jewell (1962) found. no fundamental difference between the basic metabolic rates of shrews and small rodents and. concluded that

the voraciQus appetite of shrews was due to a difference in activity,

and the intake of food with a high water content and. a low calorific value, rather than a higher metabolic rate. They found that the net

calorie intake per gram of body weight was approximately the same in -132- shrews and. laboratory mice of similar size.

In view of the conflicting infonnation on the metabolic rate of shrews, a study of food consumption and assimilation efficiency in different British species was made and. their relationship with body weight and. temperature investigated under laboratory conditions using specially designed metabolic cages.

PART II: METHODS i) Maintenance of Shrews in captivity

All five species of British shrews were successfully maintained in captivity for considerable periods of time. To prevent fighting they were kept singly, in large plastic bins measuring 370mm by 550mm. They were provided with a substratum of sawdust and hay bedding which was changed only once every three to five weeks to avoid undue disturbance. Food and water was available to them at all times • Fresh food was provided three times weekly in the font of blowfly (Calir phora sp.) larvae and pupae, occasionally supplemented by minced beef, mealworms and other invertebrates when available. Blowfly larvae and other live invertebrates were placed in steep sided glass jars from which they were unable to escape, but into which the shrews could climb to feed.. In this way, an abundant food source was always available without need for- replenishment each day.

The shrews were kept at an ambient temperature of 20°C with a lighting regime of approximately 12 hours light and 12 hours dark.

Light cons.sted of natural daylight augmented. by fluorescent tubes.

Shrews in the wild were found. to survive a maximum of 13 months, but in captivity they were able to live considerably longer as - 113

Table 4-1 shows, and. some even survived two winters under these conditions.

Species Max. No. Mouths No. Winters in cantivity Survived Neomys fod.i ens 17 2 Sorex araneus 18 2

S. ininutus 16 1 Crocidura ru.aaula 18^ 1+ C. suaveolens 12+ 1+ Table 4-1 ii) Metabolic Cages Accurate measurements of food consumption and assimilation rely on the recovery of all uneaten food. remains and. faeca3. naterial and. so a metabolic cage was designed for the study of

energy budgets of small lnPniTnals and. which could also be used. for food preference tests and. activity studies. A photograph of the apparatus is given in Fig. 4-1. It consisted. of three perspex chambers: a nest-box in black perspex, an activity pen and. a. feeding area connected by tunnels. The whole apparatus measured 965mm long by 300mm wide. The floor of the metabolic cage was of stainless steel mesh with holes 5mm square throui which faecal pellets could. fall into a tray for co1lecton but which did. not permit escape of the shrews. The activity pen and. the feeding area both had removable wire mesh tops, and the neat-box a removable perspex top, to allow access. The whole apparatus could. be dismantled. for cleaning between experiments.

Bedding was provided in the form of cotton wool in the nest-box. Food and. water were placed in steep sided glass jars to avoid

spillage and escape. The food provided was blowfly pupae which had. Fig. k-i : The Metabolic Case.

nest box B : activity pen C : feeding box D : connecting tunnel . mesh base F : collecting tray 1

4 • ±

•.. . 'j . ,,-• .

H' 4

U-

' -114- been kept wider cold cond.itior.s to delay the emergence of adults.

Shrews were habituated to the apparatus for 24 hours at the beginning of each experiment, and experiments were carried, out for three to four consecutive days during which measurements of food. consumption and faeca.]. pellet production were made at 24 hourly intervals. Each day a fresh supply of pre-weighed food was provided and. remaina of the previous day's supply were removed for weighing. Shrews often stored food in their nests, so the nest-box was searched for remains of prey items and. any food remaixie which had f4len into the tray below were also removed for weighing.

Mcst of the faecaJ. pellets fell into the tray beneath the apparatus but any food adhering to the wire mesh floor, perspex sides or amongst the nesting material were also collected. The food. dishes were searched for faeca]. pellets but only rarely did. shrews d.efaecate in their food. or water. Faeces and. urine are produced. simultaneously by shrews and the collection of the two fractious separately proved. impracticable as only very small quantities of urine are produced. Consequently, faeces and urine were collected together and. both classed as faeces in estimating assimilation.

Daily food consumption was measured, together with faeca]. production, the wet weight of the latter being taken initially before it was dried at 60°C to constant weight for use in estimates of assimilatioa. Control jars of food which were not given to the shrews were weighed each day to take account of the slight weight losses incurred as a result of water evaporation from the blowfly pupae, and. allowances were made for this in subsequent calculations.

Shrews were also weighed before and. after each experiment, the mean of these weighings being used. in subsequent calculations. - 115 -

Experiments with the metabolic cages were carried out on both males and females of all five species of British shrews:

Sorex araneus, S. minutus, Neomys fod.iens, Crocidura russula and.

C. suaveolens. Food. consumption and. assimilation were investigated

at different temperatures. Experiments were conducted indoors at

temperatures of 17-27°C, and. outside but under cover at temperatures of 0-9°C in winter. Experiments were duplicated for certain individuals at different temperatures. The mean values for each - riirnal during a 3-4 day experimental run were calculated. Daily food. consumption varied little during the experiments but if it failed to stabilise during the three day experimental period following the 24 hours habituation period. in the metabolic cages, the experiment was prolonged until a stabilised food. intake for a three-day period. was obtained..

The number of individuals of each species used. in the metabofle cage experiments is shown in Table 4-2, together with the total number of 3-4 day runs. The animals of indetermined sex were juveniles which escaped. or died. before their sex could be confirmed.

a Total no. runs Number used. at temps. of runs Nale Female 18-20°C 0-9°C Total Neomys fodiens 7 7 3 17 14 31 Sorex araneus 17 1 2 15 20 35 S.minutus 6 4 1 11 2 13 Crocidura ruasula 1 3 2 6 3 9 C. suaveolens 2 2 0 4 3 7 Table 4-2: Numbers of shrews of different species used in metabolic cage exoeriments - LJ.0 -

PART III: RESUI/S i) Food consumption by shrews in standard cages and in metabolic cages. A comparison was made between food consumption by shrews under normal conditions of captivity, when they were kept in large bins, and in metabolIc cages. Tests were carried out under similar conditions of temperature and lighting. Table 4-3 shows the figures obtained for N. fod.xens and. S. ararieus from which it was concluded that there was no significant difference between results (P>o.i) and. that data obtained from shrews in metabolic cages were representative of the food requirements of rnmrnals under near-natural conditions.

Mean Consumption (g.wet 'rt/day) % Body weight of food. taken per day Specles Metabolic cages Bins Metabolic cages Bins

Neonrs fodiens 7.56 (0.76) 6.26(0.11) 51.41(9.36) 48.82(1.18

Sorex araneus 6.4e (1.43) 5.98(0.97) 79. 57(18.11) 77.71(12.6 Table 4-3: Conparison of food. consumption of shrews in metabolic cages and in bins (with standard deviation).

ii) Sexual differences in food. consi2mption

The mean daily food consumption (wet weight) expressed as a percentage of the body weight of males and females of four different species at a variety of ambient temperatures is shown in Table 4-4. Only one male S. arax was available for the experiments and so the sexes of this species cannot be compared. There was considerable individual variation it' daiiy food consumption within each sex, especially in S. ininutus but no statistically significant difference between males and females in any of the species was found (P)o.1), despite the apparent differences in Table 4-4. Neither was there any significant sexual difference in assimilation. Consequently, results for the two sexes were combined in future experiments. -

N. fodiens S. ininutus C. suaveolens C. russula M F M F M F M F

Mean food consumption 55.29 46.83 130.91 113.31 52.24 58.00 37.98 55.3 as % of body weight(S.D) (7.42) (5.70 ) (14.85) (17.z.4)(21.76)(l4.10) (7.29) (4.3 Mean body w-t. g. 13.87 14.97 4.00 3.75 8.53 7.46 8.19 7.7 (s.D.) (1.72) (1.89) (0.45) (0.65) (0.82) (0.19) (0.27) (0.4 Table 4-4

in) Differences in food. consumption & assimilation according to species & body weiit Results obtained for each of the five species of shrews in the metabolic cages are shown in Table 4-5. Food consumption is expressed. in terms of wet weight from which dry weight equiva1ents were calculated on the basis of blowfly pupae having a mean water content of 70.2%. Assimilation and assimilation efficiency (the percentage of the food. digested and. assimflated) were calculated. using the dry weights of food. consumed. and. faeoal products voided. daily, according to the following formulae:- - Assimilation = Consumption - Faeces & Urine Assimilation x 100 Assimflation Efficiency (%) = Conmption

The results of daily food consumption and. assimilation efficiency for the different species of shrew for a mean temperature range of

12-17°C are si,mmised in Table 4-6, where the species are listed in order of decreasing body weight. Food intake was highest in the larger shrews, Neomys fod.iens and. Sorex araneus as would. be expected. on the basis of body weight alone. However, the loweeb intake in terms of g wet weight of blowfly pupae as exhibited by Crc.idura russula whose daily consumption at 3.92g was statistically significantly lower than that of the much smaller S. minutus ( Pc0.05). Both crocidurid species bad. food intakes in the same order of magnitude as S. minutus

It') tn I-i. 0 0 (I) I0 c0 OQa J. Co I-, -4

N N ¼j I-' , '0 o co N I I IsCD lt N Ioi Ii - 0 0 -4 04+

C/) ¼3 .*) 0% N N I-I I- 0 .. '53 e CD CD I_I I_a -3 co '*J, ¼) • • Co Co .:- .p- • • • • S • • CoN • lCD \ø) 0 '0 '0 - I-a Co co - N '0 - 0% _53 0 N I-' % —S —S o 0 o • • o 0 N N t.-i 14+0' %J) - _) 0 %0

n i) \ • • • • • 0% CD - _S.1 • • 5 • 5 Co '0 '0 '.n co N N N Co '0 o 0' CD ,_J $_J —S. —S 0 I_-I 0 I_a c1114, • • • • I-I I-a I--a dO • • o 4+ '5') 0' - N -p. ..— %__# o -.- • o .5.- 5.—._ 5.__

I_i ,_J I-' I_J 10 CD • S • S N N N • 0' 0 - I-' • I I.-' 0 -53 -3 _tSi -S -S 1':C —S o 0 o 0 'U Iom • S o 0 IQ'i-1-' • S (D N .p. o 0 %__ 5- '5,' -p. 4+4+ H 5— .5— *5

o 0 0 0 o 0 • . S • o 0 o o • S • S • N N N S C) N N -p. th l-J CD I(DIJ. '0 U I-' I.j -S —5. __S v—S lCD - o 0 0 0 o 0 o 0 • S S S • S o 0 cQc+P. o 0 0 • S .i) o 0 o 0 5- 5.__ 0' I-a '0 '0 5-_ 5.__ %__ .- .5—v .

'-I, N N I_a I • I I I

I.Jt

- '0 co '0 - • Co S • • CD 0 N - N- o. —.5 -P- Co Ii —5 —S S Ln I I-' N co \O l"'d 0 • 0% 0' N • • C.,. N • • 0' 0 CD '0 (DOMa CD - 's-" -' a) -5, S.— 8 %__ 5-_ C.. 5- I_J pJ I_a 1%) p.-' • . • . lb m 0' CD 0' f CoP . ,_a - N —S —S o 0 o 0 • S J N I) ,p o 'o H U .5— 5--

0' c c CD -..] ,' '0 ICDP - 'Sn N .p. ji-s,com • • • • • S ¼i -53 'Sn o j3 0' i '..) 10 '-5. —S —S N • S 0 %.jt CD-' • S • o.- S - 0 0% tj .1:• 0 [04+ 11 5__ 5.__ '0 I-J -53 t) U 5— •__ 5_. 5— i__S 0

5.—. - .LJ.O - despite being twice the weight of the smaller species. This is further emphasised in Fig. 4-2 which shows food consumption by shrews of different body weights, from which it is apparent that food consumption and body weight are not directly related. Of particular interest are the results of consumption within a species where there is much variation between individuals, regardless of body weight.

The relationship between food oonsuxnption,as a percentage of body weight, and body weight ishown in Fig. 4-3 where it can be seen that, if species are compared, food consumption generally showed an inverse relationship to body weight, with the extremes being exhibited by

N. fod.iens and S. minutus (mean 51.23% and 125.49% respectively). However, this relationship is only evident for the three British ma1n1c.nd species (. minutus, S. a.raneus and N. fodiens): The two crocidurid species had food consumptions which were significantly lower than those of S. araneus of equivalent size, and approTi-Tnate more closely the results for N. fodiens. That food consumption is not directly related to body weight is further emphasised by a comparison of individuals with different body weights within a species, where there is no indication of decreasing consumption, as a percentage of the body weight, with increasing body weight.

Assimilation efficiency of shrews ranged from 72% to 8 and again, if mean values are compared for the three British mainland secies ,there is an indication of decreasing efficiency with increasing size: tie assimilation efficiency of N. fodiena was significantly lower than that of S. araneus and S. minutus (P

4-kew +lv ryy ry j, +h h,rhr TrI

. - C v 0 '-I - .- a E '. -J Z(I,USU() o•o 14 I 0 :• N 0 a

a a 0 0 '0 00 I a 0 a

0 0 01J 0 o a 0 -c 0) 5 0

S • • 4 - .4 > . -U •! • 0 .414 . S •.. ••4. ••S. S ._ '0 • V I 0 0000 0 0000 0 0 0

I I N N - , ., (6) uolldwnsuoD poo' C V z C)u.I -0 0 C ZuUO u•o ii

0 1CM a I a a '0

ci ci a a DO ci a a ci • a : • •• S • • S 44l 4 •• •4 •% 44 • • • S • S. I! S U • '0

0 0 00 00 0 Oo 0 .0 0 0

-i CM 3 0 0 0 0 0 CM (04) uoqdwns:o) p001 - 119 -

Shrews of different ages were used in the experiments but no correlation between food consumption or assimilation efficiency and age was apparent. It seems that the variations in food consumption

and assimilation efficiency within a species are due to the inherent differences in metabolic rate between individuals.

iv) Food consumption & assimilation at different temperatures

Table 4.$ shows the results of food consumption and assimilation under summer temperatures (in the range 18-27°C) and winter temperatures -

(0-9°C). Assimilation efficiency did not differ significantly between the two temperature ranges. Contrary to expectation,

however, there was not a single example of food cnnsuinption by any

species increasing significantly with a fall in temperatures in most cases it remained the same but,for S. araneus a reduced consumption was recorded at lower temperatures which proved to be statistically

significant (Pc0.0$). Further Investigation of food consumption by

shrews of all species outside in their normal conditions of captivity

in bins during the winter confirmed this observation, even when ambient temperatures decreased to -1.0°C.

Considerable differences In water content of the faece* were

found during experiments cnnducted in warm and cold conditions • This can be attributed to the drying effect in warm conditions, and to the

predominafitly humid and often very wet environment during experiments

conducted outside in winter. A mean water content was calculated from data collected from metabolic cage experiments in warm dry cnnd itions.

This was in the range :32-56% (see Table Li_5) and is probably representative of the water content of the faeces.

PAflT IV: DISCUSSION

A trend in increasing food consumption in terms of percentage intake of the body weight, with decreasing body weight has been shown - .L%l -

for the eriea N.fodiens, .aran and S. minutus,but that food consumption is not solely governed, by body weight is shown by the two crocid.urid species whose food. consumption was far below the expected on the basis of size alone. Siiiilar results were obtained by Vogel (1976) who found oxygen consumption of Crocidura russula was similar to that of African crocidurids and was much lower than for the

European species S • araneus, S • mi.nutus and N. fodiens. Hunkeler &

Hunkeler (1970) also found that food consumption of four species of African shrews, including crocidurids, was considerably lower than

European and American shrews. Vogel (1976) suggested from these results that the Crocidurinae are characterised by a low metabolic rate and the Soricinae by a high metabolic rate, especially the tribe Soricini. with the tribe Neomyini being intermediate. If food consumption as a percentage of body weight can be taken as an indication of metabolic rate, then results of the present sttviy substantiate those of Vogel, with the crocidurids ati the larger N. fodiens having a much lower % intake than the Soriolds,

Vogel (1976) attributed the difference between metabolic rate of the crocidurids and the soricids to evolution in geographical isolation under differential climatic conditions. This could explain why Dritish crocidurids are found no further north than the Channel

Islands and Isles of Scilly where the climate is more equable than on the Britien mainland • However, this is contradicted by the work of Dxyden, Gebczynski and Douglas (1974) with the tropical musk shrew Suncus jur1nus. This species was also found to have a lower metabolic rate than temperate shrews but when acclimatised to cold temperatures it was able to increase its metabolism to overcome this stress.

Differences in assimilation efficiency between the species

investigated in the present study were apparent, ranging between

71.7% and 86.2%. Tho highest efficiencies were found in the two - .12.1. -

Sorex species, which were significantly greater than the crociduricla

and the larger N. fodiens. Pernetta (1973, l976a) obtained an assimilation efficiency of 90% for C. suaveo1es but attributed this high value to the exclusive diet of blowfly pupae which shrews eat by opening the exoskeleton and licking out the contents without

ingesting much indigestible chitin. In the wild,where a large variety of prey is taken, chitin intake may be considerably greater. Hawkins & Jewell (1962) obtained assimilation efficiencies for S • araneus and N. foci lens in the or1er of 87-92% which is also higher - than the results of the present study but, with the exception of C. russula, all these results are higher than the 711. for

laboratory mice of similar sizes (abstracted front Hawkins & Jeweli

1962).

Shrews in the present study showed no increase in food

consumption with a decrease in temperature and this agrees with the results of Rudge (1965) for N. focilens, S. minutus and S. araneus kept outside between 6 and 20°C. Rudge did not report a significant reduction in food consumption by S. araneus in cold conditions as reported here, but his temperature range was higher than that of the present study. These results contradict the findings of Wolk (1969) who reports that the lower the temperature the greater the food consumption. She kept S. araneus in outdoor cages and studied their consumption of dried minced meat on a seasonal basis. She found that consumption by young shrews was statistically significantly greater in winter months than in summer and autumn. The lowest consumption as a percen.age of the body weight was shown by old adults in summer and this also increased in winter. She also reported that the body weight of the shrews decreased in the winter, so that the increase in percentage consumption could be attributed to the decrease in weight of the animals rather than a direct response to the change in temperature, but in the light of results in the present study this seems unlikely. In the present study a large rn1mber of -

individuals were used who8e weight had remained constant under laboratory conditions at 20°C so that a functional response to temperature was recorded, rather than a morphological one.

Failure to increase food consumption in winter could indicate a functional or a behavioural response (in terms of reduced. predatory aCtivitie8), to temperature changes, as well as a morphological response in terms of seasonal weight changes, which have already - been noted • Associated with functional response are the findings of Gebczynski (1965) who reported that oxygen consumption , at a range of temperatures ,was lower in S. araneus in winter than in summer. This reduction in metabolic rate could be caused by the decrease, not in total body weight or size, but by decrease in weight of internal organs such as heart, liver, kidneys and. spleen found by

Pucek (1965) aM an increase in the insulating properties of the fur to allow this (Cebczynski & Olszewski, 1963). Paradoxically, Gebczynski (1971) found the oxygen consumption of S. minutus increased below 15°C but seasonal measurements showed a slight decrease in winter compared with summer when taken at 20°C.

The apparent reduction in metabolic rate of S. araneus in winter might also account for the decreased activity of shrews at this time of year as reflected by numbers of captures in the present study. LOxton et a]. (1975) found that although the diurnal cycle of activity of captive S. araneus was similar in winter and summer, the overall activity was lower in winter. Nezhzherin (1969) suggests that the decrease in body weight of shrews in winter may be a trend towards their energetic optimum and 'saves the population' in conditions of food shortage: if a decrease in body size does not increase metabolic rate then selection for small size would occur when food. resources are low. If tropical shrews can increase metabolic rate to cope with conditions in temperate latitndes, it is possiblo that temperate - ]23 -

shrews under cold conditions could also modify their metabolic

rates to reduce their energy requirements. This could explain the failure to increase food consumption while still maintaining the same body temperature which is neither significantly reduced in

cold conditions (Gebczynsld, 1977) or in times of reduced activity such as sleep (Redmond & Layne, 1958). A decrease in body weight

and size could reduce the total daily food intake and hence the number of prey items required and the energy required to catch them. If decreasing body weight is also accompanied by a tendency for increasing assimilation efficiency this would have a distinct selective advantage in winter. A comparison between the species

showed the small S. minutus to have a significantly higher assimilation

efficiency than the much larger N. fodiens, but there was no significant difference between assimilation efficiency of the

intermediate S. araneus and either 5. minutus or N. fodiens, Furthermore, a tendency for increased assimilation efficiency with decreasing weight was not apparent when individuals of the same species were conpared. The suggestion that there is a selective advantage in reducing body weight in winter because it reduces absolute food requirements also presupposes that there is a direct

relationship between body weight and food consumption, but, as has already been described., this is not so. A reduction in body weight

within a species does not necessarily reduce food consumption. It seems unlikely, therefore, that the decrease in body weight of shrews in winter, amounting at most to 3g per individual in S • araneus, is sufficient to either produce an increase in assimilation efficiency or to reduce absolute food requirements.

Reduction in body size increases tI surface area to volume

ratio which in turn increases heat loss, and Gebczynski & Olezeuski

(1963) aM Gebczynski (1969) found the highest heat losses of t.

S. araneus occurred in autumn ani winter, despite the greater insulation provided by the winter coat. This could be a distinct disadvantage in reducing body weight in winter.

In conclusion, it is suggested that the winter weight loss of shrews does not confer a selective advantage on them. However, if shrews are able to reduce food consumption at low temperatures (as is suggested by the results for S. araneus in the present stxiy) by other means, such as reducing activity ani conserving heat by remaining for longer prriods in the nest, this would have an important influence on survival at low temperatures.

Mean Food Mean Food Consumption Mean Mean Body Consumption as % of body Assimilation Species Weight g (g wet wt.) wt(wet wt) Efficiency %

Neomys fodiens 14.144 (2.01) 7.32(1.21) 51.23(8.714.) 77.51+(6.77) Sorex araneus 8.37 (1.11.7) 7.19(1.86) 8689(2l.l7) 84.92(4.81) Crocidura russula 8.18 (o.6z) 3.92(0.57) 48.26(8.70) 71.68(3.52) C. suaveolens 8.07(0.82) 11. 35( 1.29) 54.71(17.68) 82.05(2.114) S. minutus 3.95(0.50) '1..96(O.90) l25.45çL6.62) 86.23(5.47)

Table 4-6: Mean daily food consumption & aseimilation0efficiency of shrews (with staniard deviations) atl2-17 C - ]25 -

SECTICH B; CALOHIFIC & WATER CCNTENTS CF SHREWS & THEIR DIF PART I INTRUCTION

Shrews take a wide variety of prey items as the diet analysis in Chapter 3 has already shown. They tend to feed primarily, but not exclusively, on the most abundant invertebrate types available, such as adult coleopterans, and supplement the diet with other, less numerous types such as chilopods. The laboratory studies of

Crowcroft (l95a), Rudge (1965) and Pernetta (1973) have shown that captive shrews show preferences for certain prey items on the basis of palatability: for example, millipedes are apparently distasteful and are not eaten and, as faeca]. analysis has revealed, occur extremely rarely in the diet of wild shrews, and likewise with certain isopods. However, it is not known whether any additional selection of prey occurs on the basis of food value or whether the occurrence of certain prey items in the diet is purely a function of thir abundance and palatability. More information is available on the energy values of shrews themselves and their assimilation efficiencies (Hawkins & Jewel]., 1962; Myrcha, 1969; Pernetta, 1973, l976a) but apart from the work of Hawkins & Jewel]. (1962) little attempt has been made to relate this to diet.

In this study, calorif Ic values and water contents of known prey items taken from The Wilderness study area .were determined,

together with the calorific values of shrew carcases and faeca].

material, in an attempt to correlate diet with food. value and energy utilisation by shrews.

PART II: ME'THS

The calorific values of the prey items most commonly found In

the diet of shrews in The Wilderness were determined together with

the standard laboratory food of Calliphora (blowfly) pupae and larvae using a Phillipson microbomb calorimeter (Philhipson, 196k). A Gallenkainp ballistic macrobomb calorimeter was used for the determination of calorif Ic values of faecal pellets obtained in

the metabolic cages and of carcases of S. araneu q trapped from Alice Holt Forest, where larger samples were available. The carcasos were first fat-extracted (see ].atvr) and the calorific value of fat,

determined in a microbomb calorimeter, and fat-free carcases were

measured separately.

The material for bombing was first weighed and then dried to - constant weight at 6o°c and Its water content calculated. It was then homogenised using a glass pestle and mortar and,when necessary, frozen in air tight glass vials to await bombing. Samples of material were taken and re-dried immediately before being bombed. Depending

upon the amount of material available, the ca].orif Ic value of a

minimum of five samples was determined in the microbomb aid three In the inacrobomb. Where results were inconsistent, more samples were used. Results were calculated as kca]/g di weight of ash-free material.

PART III, RESULTS

A list of potential prey items occurring in The Wilderness is shown in Table Li_7 with their respective water contents and ca].orific values. Water and calorific cantents varied between groups within the major taxa investigated, but a summary of the mean values for the main groups is presented in Table 4-8, in order of increasing water content aid decreasing ca].orif Ic content. Results of calorif Ic values of certain items are comparable with those quoted by Cummins &

Wuycheck (1967), the slight differences probably pertaining to

(I) sample size, which in many cases was larger in the present study aM (2) the condition of the invertebrates when collected, for example

egg..bearing females have a higher calorif Ic value than males (Cuminina

& Wuycheck, 196?) and their numbers in the sample would tnerefore be

expected to affect the results. PREY ITEN Mean % Mean Kcal/g Water Content dry weight (S.D.) ADULT COLEOPERANS Carabids 62.2 6.11 (0.23) Tachyporus spp. 6.12 (0.45) St aphyl inid.s 65.8 5.70 (0.22) Elat erid.8 60.6 5.34 (0.z.) Coccinella sp. 54.8 6.49 (0.07) Cantharid.s 68.6 5.88 (0.24) HEMIPERANS Cercopid.s 66.8 5.62 (0.15) Mirid.s 5.63 (0.21) ADULT DIFFERANS 5.07 (0.23) ) ADULT LARVA&- Lepid.opterans 76.7 5.54 (0.38) Tipulicis 84.0 5.17 (0.13) Calliphora ep. 69.7 5.97 (0.19) Calliphora sp. ( pupae) 70.2 5.84 (o.i) Tenebrio molitor *63.0 +6.31 ARANEIDS 76.1 5.37 (0.17) MYRIAPODS Lithobiomorphe 75.1 5.76 (0.22) Geophilomorphs 73.6 5.99 (0.13) ISOPODS Armadillidium vulgai'e 63.6 3.87 (0.22) Oniscus asellus 67.5 4.06 (0.31) Porcellio scaber 70.9 4.18 (0.24) MOLLUSCS Anon hoTtensis 82.7 5.4]. (0.09) Other slugs 83.9 5.12 (0.16) LUMBRICIDS 84.4 5.49 (o.i9)

Table 4-7: Calonific values of invertebrate prey items common in the diet of Shreds

* from Hawkins & Jewell (1962) + from Slobotikin & Richman (1961)

t Mean Calorific Value Mean Water Content (%) (kcal/g dry weight) Adult coleopterans 62.4 Adult coleopterans 5.94 Hemipterans 66.8 Myriapods 5.88 Isopods 67.3 Insect larvae 5.77 Myriapod.s 74.4 Hemipterans 5.63 Insect larvae 75.2 Luinbri cid.s 5.49 Araneids 76.1 Araneid.s 5.37 Molluscs 83.3 Molluscs 5.27

Luxnbricids 84.4 Adult dipterans 5.07 Adult dipterans - Isopods 4.04

Table 4-8: Water contents & calorific values of maor prey types 127

The lowest calorific values were recorded for diplopods

(3.1t kcal/g) and for isopods (Li.Ol+ kcal/g). The former occurred very rarely in the diet of shrews but are included for comparison. Calorific value is not necessarily related to water content, for although adult co].eopterans had. the highest calorific value and the lowest water content, isopods had a low calorific value despite a low water cnntent. Calorific value is more directly related to the proportions of low-energy chitin and high-energy fat possessed by the different invertebtate prey items • Both isopods and millipedes have a high chitin content whereas many insect larvae, such as Tenebrio molitor, have high fat cnntents.

The calorific values of shrew carcases, faecal pellets and the standard laboratory and experimental food of blowfly pupae are Ehown in Table 4-9. The calorific value of blowfly pupae was determined minus the exoskeleton which is not eaten by shrews. The calorif Ic value of the faecal pellets iS shown to be similar in all three species of shrews investigated and was considerably less than that of the prey.

The energy contents of shrew carcases together with the daily food intake and daLly egestion of faecal products in terms of kilocalories,are shown in Table 4-10. The calorific value of fat-free carcases was determined for S. aran only. This value was used to estimate the energy content of carcases of this and other species, including S. minutus and N. fodtens for which information on water and fat content was available from autopsy data, and of C. russula and C. suaveolens whose fat and water contents were assumed to be similar to those of S. araneus.

Consequently, calorific values per gram of carcase of different species of shrews are not directly comparable. The energy content of the daily food intake and faecal production are based on mean

- _

dry weights of material from Table 4-j in the previous section dealing with metabolic cage results.

The energy content of the daily food intake and faecai.

production ,ard the assimilation efficiency, show similar trends to

those already discussed in the previous section for the different

species, as might be expected. Assimilation eff3ciencies in terms

of calories are found to be 3-6% higher than those based on weights of food intake and faecal production.

Calorific value No. of (kcal/g dry weight individuais Total No. of ash-free material) ec.mpled Samples S. araneus

Fa.free carcase 513t1 (0.70) 2.5 7.5

Fat 9.11.5 (0.23) 10 39 Faecal Pellets

N. fodiens 4,70 (0.11.8) 5 15

S. araneus 4.79 (0.4.5) 9 27

S. ininutus 11.79 (0.611) 2 6

Food

Calliphora without cases 5.89 (0.07) - .5

Table 4-9: Calorific values of shrew arcases, fat, faeces & laboratory diet (S.D.)

PART IV: DISCUSSION The items occurring most frequently In the diet of S • araneus in The Wilderness were adult coleopterans, which are shown to have

the highest calorific value and the lowest water content of all the prey items examined. The other most common pray items Included

insect larvae, araneids and isopods. Insect larvae had a high

energy value with a water cnntent of 75.2% while isopods had the lowest energy value but also a low water content and araneids occupied an intermediate position. Lunibricids were taken in 4 -3 C) C) it') it') I 10 1 0 Ia Co 1-4 F fl l- (, : r IJ 1p) I 0 I- 03 Ii'l Ip l ft-h I i i 103 Ito 103 ItO 03

I-J *03 co a, ! S S S S 0 ,_J '0 ) .? CD .n 3 4 '4

a I I • . a N %.j 10303D CD e-

03 o 4 4 (0 03 • S S 03 03 S

N I-' N 0* • S S S

a 1* .Jt I.j '..j '..j S • 0 5 S ).-• 0,

'-4 I-' F-' %.3 0' 0's - O\ 0 Io S • S S S ' '0 o '0 - CD p

I-' I-' P- N

1* o 0 N I-' 0

I-' I-' 0 I-' N 03 • S S S S lb—S C) s 0 F.-, -.3 I.J% 3 S_ C CD Ij F-' '.)

0 CD - •sØ 0, 0, 0's •-•3 N CX) N 10W • • • • • ItIHØ o- N .-.a j C's '0 N ".0 s IWO I P l(D I-' ; F—.O c. - J29 - considerable numbers at certain times of year and ,despite having a very high water content ,they also possessed a high calorif Ic value in terms of kcal per g of dry weight. The other items, such as hemipterans, adult dipterans, myriapods and molluscs )oCCUrred less frequently in the diet.

Associated with the energy value of the food items is the chitin content which represents the major indigestible component and which therefore reduces the available energy value of the food. The quantity of chitin ingested partly depends on the size of the preys with large Items such as carabid. beetles and araneIds,the fleshy parts are devoured first ar'd the limbs are discarded, but with smaller items such as Tachyporus spp. the whole animal is eaten (pfrsonal observation).

Shrews in The Wilderness ate large numbers of adult coleopterans, which mostly comprised staphylinid.s which have small elytra and large fleshy abdomens, and. Insect larvae which also have large fleshy abdomens. These represent high energy foods and not only were they the most numerous of the macro-invertebrates in the soil and amongst the vegetation, but also, with the exception of lumbricids, had the highest bioinass at most times of year. It is not surprising,therefore,that these items comprise the dominant components of the diet of shrews.

It seems likely that their predominance in the diet is a function of their abundance rather than their selection by shrews, owing to the large numbers of isopods and. other prey Items with considerably lower calorif Ic values ar4 higher water contents which are also taken.

No account was taken in the determination of calorif Ic value of

S • araneus carcases of possible sexual or seasonal differences for It was the purpose of this study to present an overall mean value. Myrcha (1969) measured the calorif Ic value of S. araneus, S. minutus and

N. fodiens and although she found no sexual differences she does report higher calorific values of carcases in winter and earl,y spring -'-.J" - than in summer and early autumn in S. araneus,ancl to a much lesser extent in S. minutus. The differences ranged from If.80 to 5.19 kcal/g dry weight,which she says were statistically significant, and her results were considerably lower than the 5.55 kcal/g obtained in this study. S. mi.nutus had similar values to S. araneus3 and N. fodiens was slightly lowfr. Gorecki (1965) also found low values for S. araneus,of the order of 14.k5-4.68 kcal/g,and reports no seasonal differences, possibly because his data did not cover a complete year.

The daily calorific intakes by shrews in the present study are compared with the results of some other workers in Table If-il, and are found to be considerably lower. However, there is little difference in assimilation efficiency, in terms of calories,between the results of this study and the findings of Hawkins & Jewell (1962)

& Pernetta (1973, 1976a).

The differences In results of calorific intake between the studies can be attributed to the nature of the diet aM the experimental conditions in each case. In the present study the! diet consisted exclusively of blowfly pupae of which the exoskeleton. is rejected by shrews and only the softer parts inside are eaten. This provides easily assimilated energy with little waste in the form of indigestible chitin, which might be expected to result in a lower consumption of this food type than of a more chitinous food such as Tenebrio larvae, or of unscoured earthworms. If this were so, a lower assimilation efficiency would be expected to accompany the higher consumption of more chitinous material, but this does not appear to be the case. C) C) Z U) U) C) U) U) It!) Is . S • C) U) S S • S S S S S S (d Iu II CD 0 0 CD )-4 p IH

CD Cs. •< $.a. CD CD cf CD CD CD CD 0 CD CD c+ CD I_J 0 o p, i-a CD CD CD

1.1. CD C) CD

iCD Ii-' P 3 * 2 2 I 3 2 Ii I() 0 gilt to -. ii rI €3 a 'iIo 11 €3 10

p.3 I_i I-a I-' II i-J I-' i-' i- 1%) 0% - %O N - 0' UI fs.) .:- '.A) j.-a 0 • S S S S • S S N co co so ,J so . p. 0 '0 s..) 0% i-a I I I I '0 N I-a N Q3 • S S •N co 0 CDc

p o 0 0 N I-' i-a I-' I-' 3 I.-' • . a • • S • • S S o' so co '.a UI I-' p. I-I CD 0) UI I-' U '0 I-' I-' 1) Q • I-a. $ I I co $ I I 1)

o ) I-J €3 N CO '-.) o CD '-a.

i;:: I I - i '? '0 so o N

i!J

€3 0 * . CD 2 2 S S Cs. CD P '0

sg' a .3 '0 -'3 ' .,I Not only did Hawkins & Jewefl (1962) and Pernetta (1976a) obtain higher values for food consumption, both in terms of dry weight of food consumed and daily calorific intake,but they also quote calorific values of the food which are higher than those of the present study. For example, Hawkins & Jewell (1962) give a calorif Ic value of 6.2 kcal/g for blowfly larvae which contrasts with 5.8ii and 5.97 kcal/g for pupae and larvae respectively in the present study, and tIs could partly explain the differences in results between studies. The calorif Ic values f or the faecal pellets of N. fodiens obtained by Hawkins & Jewell (1962) were lower, at 3.7 kcal/g dry weight , compared with those of this study (4.7 kcal/g), although Pernetta's results for C. suaveolens were high at 4.3-4.8 kcal/g.

The energy intake by shrews in this study, in terms of calories, was found to be similar to those obtained by Hawkins & jewell (1962) for the harvest mouse and laboratory mice of equivalent weight (see Table 4-li) although these rodents were fed on a diet comprising only 12.5-14% water compared with the 70.2$ for naggot pupae. This confirms that the energy requirements of shrews are not excessive compared with rodents of similar size, and that the high rates of metabolism obtained by othrr workers may be due to the activity and, as Hawkins & Jewell (1962) observed, the water content of the diet, causing theni to consume a greater wet weight of food with a high water content bhan with a lower water cnntent. Pernetta (1976a) reports an apparent anomaly in terms al dry weight equi.valent of food consumed by C. suaveolens which he found to be greater in the presence of an extrrnal water source than witho.t it, coupled with a higher calorific value of faeces and a lower assimilatinn efficiency in the presence of external water. Pernetta attributed this to the tendency fvr rectal feeding (coprophagy) which he observed in the absence of water. No allowance for coprophagy was - 132 - made in the present stiy, but it is possible that it could affect the results of assimilation efficiency, producing higher efficiencies in shrews which indulged in this habit. - 133

SFETION C PRODUCTTON & SEASONAL WEIGHT CHANGES

P&RT I: INTRODUCTION The reproduction of S. araneus and S. minutus has been well'

documented by Brambell (1935) and Brambell & Hall (1937) using

autopsy material, and Dehnel's (1952) account of the breeding of

S. araneus includes details of mating shrews in captivity, their reproductive behaviaur and the growth of the young. Crowcroft (195t.a) also reported the growth rates and morphological changes - of captive reared S. araneu, and Ansell (l9O) describes some

aspects of the post-natal development of an African crocidurid, Crocidura bicolor. However, owing to the difficulty of breeding shrews in captivity ,there is little detailed information about the development, survival and behaviour of young shrews.

Between weaning and the sub-adult stage attained, in the autumn, young S. araneus gain weight but rarely exceed 7.5g,which in captive-

reared animals was achieved soon after independence. Over the

following winter cons1derable changes in size and weight occur. chapter 2 has already revealed how wild shreis undergo a reduction in body weight in winter prior to the spring jump in growth to

maturity, but there is little information about simIlar weight changes in captive shrews. The possible causes for the decline In numbers of shrews in aiitumn and the reduction in weight of remaining ones in winter, have been explored in Chapter 3, but no positive conclusions were drawn.

The maintenance of shrews in captivity for periods of several

months for experimental purposes provided a suitable opportunity

to study the growth of young shrews and to monitor seasonal weight

changes under different conditions. - 131)'

PART II: ?'T[HODS

During the present study, litter and embryo sizes were recorded fro'n autopsied shrews obtained from the study areas, and the opportunity to monitor the post-natal growth and. development of the young was afforded. by four pregnant shrews which produced

litters in the laboratory. In this way some information on production, growth and behaviour of shrews was obtained.

During the course of experiments on captive shrews, individuals

were maintained both under constant conditions in the laboratory and

in outdoor enclosures where they were subjected to natural weather

conditions, inc)uding rain, frost, snow and extremes of temperature. All shrews were provided with ample food. They were weighed. periodically and the weight changes of juveniles/sub-adults between

summer/autumn and spring are reported and the results from the two regimes of captivity compared.

PART III: RESULTS & DISCUSSION 1) Production a) Litter size & Gize of pre.natal embryos

Nine piegnant female S. araneus were amongst the autopsied shrews from Alice Bolt Forest and The Wilderness & Monks Wood, plus

one N. fodiens from Airesford, Hants. No S. minutus bearing young were caught. All th pregnant shrews were captured between May and September. A summary of the number and size of embryos from each

female is shawin Table l -l2, where body length refers to the distance between the crown and the rump, and body width equals the maximum distance between the belly on the ventral surface and the vertebral column on the dorsal surface - -

Mean Length x Body weight Month Of Nuniber of Width of Spec lea of female(g) Embryos (u) Capt... Emb'os .araneua May 8.k6 7 5x 3 10.25 7 10 x 5

12.37 8 kx2

10.75 9 kx2

10.75 9 6 x3

June 11.50 9 3 x3

July 8.31$. 6 9x6

August 9.11.6 7 9x5

September 12. 1$.

Mean 10.11.7 7

_f'od lens August '5 kx3

Table 1$12

The mean number of embryos prr pregnant S • araneus were seven

and the largest embryo found at autopsy waS 10mm x 5mm in a Utter of

seven. Brambell (1935) found litter sizes of 1-10 Individuals, with

a mean of six. pernetta (1977a) found litter size decreased as the breeding season progressed, with a mean of 7.5 in May aM 1.0 in October, but insufficient data were available in the present study to reveal any significant differences in mean litter size throughout the summer.

Three female S • araneus ?hich were not mated in captivity, produced litters of four, six and nine young respectively in the laboratory. One C. suaveolens which did. mate in captivity produced a

litter of three young. - 136 b) Post-natal developrcnt & growth of young shrews

The young of both S. araneus and C• suaveolens were born blind and naked after a gestation period of at least 21 days. Each young is estimated to have weighed 0.5g with a body length, excluding the tail, of approximately 15mm at birth • Growth was very rapid, particularly in the first 10-15 days of life, as shown in Figs. 4-k and 14._5 and the photographs in Figs k-6A & B. After nine ci aye the fur of S. araneus young became evident as a soft grey down, and by 11 days the teeth showed their chmracteristic yellow tips. By iLl. days of age the young had grown to 5-7g,after which body weight levelled off; they were covered in short, fine greyish fur and the eyes were beginning to open but they were still cnnfined to the nest. By 16 days the eyes were fully open and by 18 days the young of one litter were nearly as large as the female and ventured cit of the nest fcr short periods independently of the female, although she was observed to drag them back to the nest on occasions. At this stage, the young were still suckling and no invertebrate prey was tak. At approximately 21 days old they were first observed to eat b].owf) y larvae but still made attempts to sickle; threat calls were emitted between the female and the young. During the weaning period a decrease in body weight was apparent. By 22-25 days they were fully weaned and began to increase in weight once more. They began occupying separate nesting sites where bedding was provided but constructed only very rudimentary nests. Aggressive behaviour was evident between the young and between the female and the young, with squeaks and scuffles occurring both in the nests and in the enclosure, when it seemed there was competition for nesting sites and for food. Within 25 days after birth the young were completely intiependant.

The shrews whose growth was monitored for longest included one litter of S • araneus born anti reared in the laboratory at 2O°C anti one N

0 CW)

N '0 N

N N N 0 N 1' 'I) 0 -0

0 a

a, S

0 Co C') CM (6) H6'M I(po9 0 I')

N N

0 N Ce

r— U) >% a w

a, 0) N I-

0 cd.

w O)

(ww) L146ua1 Xpo Fag. -6A : Young S.draneus.

I Day Old. Body length I8rnm.Weight O.42g.

5 Days Old. Body length 3Omrn.Ueight 1.2g. -I.. Fig. 1 -6B : Young S.nraneus.

8 Days Old

15 Days Old

61 71 10 CE74TIMETRES

r r I' II - 137 - litter of S • araneus born ani reared in outdoor enclosures where corxIitions were considerably colder amti wetter, with temperatures decreasing to 8°C. The young of the other two litters died after

7-9 days owing to the death of the females. The growth of the S. araneus brought up indoors was more rapid than those outside, and these obtained juveni1e/subadu1t weights of approximately 7.Og by 1k days after birth, while those outside had reached only about

5..5g by the same stage ,aM even after a month had not increased to 6.0g. Whether this was Iue to the colder corkitions or to the difference in litter size is not known; those born outside were in a littrr of 6 while those inside were a litter of only four. It is possible that the female with the larger litter as unable to fèdd them all adequately.

c) Food consuinption by nursing females Food consumption by nursing females during the nine days following the birth of the young is summarised in Table 1l_l3. - S • araneus S • arane'is C • suaveolens (8-25°cJ (zoc) - (2C) - Litter size 6 9 3 Mean daily food consumption 15.0(1.01) 11 .5(1 .96) 5.32(0.07) (g wet weight) Approximate % of body 12 110% 49% weight consumed

Table Li._13s Food consumption by nursing females (S.D.)

Food consumption was considerably higher at this time than for other

erirews of equivalent size, discussed in a previous section. C. suaveolens typified by a lower intake than S. araneus anyway, is

here shown to have a food consumption less than half that of the

other species. - 138 -

d) Be1ocation of nesting sites During the nursing period, the S. araneus in the outside enclosure moved her young several times to different nesting sites.

The young were born in a box provided, in which the female made a

nest of dried leaves. Three days later all six young were moved to a new nest beneath a large piece of bark O.5in away from the first site. the original nest was soiled and damp but whether it was this or

disturbance which caused the move is not known. After a further 14 days, by which time the young were large and active with their eyes

open, but not weaned, yet another nest was established in a shallow U-shaped burrow in the soil. This was approximatei 100mm deep with two entrance holes and a central circular nesting area supplied with

dried leaves.

e) Retrieval of the young,& fostering of C. suaeolens young, by S. araneus

Before they were 18 days old the female S. aranej would retrieve her young from the open area of the cage and carry or guide them back into the nest. On several occasions the young were removed from the nest for weighing and replaced in the open area of the cage. The female quickly located their whereabouts and pro..eeded to return them to the nest one by one until all were replaced. When the young were very small she carried or dragged them by the tail or any other part which she could get hold of. As they grew older and too large to carry she dragged them by the scruff of the neck. This behaviour continued even when the young began venturing out of the nest independently, but not once they were weaned • There was never any evidence of caravanning behaviour as has been reported for C. russula and C. suaveoleris (Corbet & Southern, 1977), C. bicolor by Ansell

(1964), Suncus etruscus (Fons, 1974) aM for S. araneus (Harper, 1977).

Retrieval behaviour was also shown by S. araneus for the young of 1 39 -

C. suaveolens. The female C. suaveolens died leaving three young of

6 days old and of the same age as one of the S. araneus litters. The crocidurid young were placed yelping in the cage, but not in the nest,

of the S. araneus in an attempt to foster them, whereupon the female S • araneus investigated them and carried each one back to her nest. The young survived for a further two days, so the fostering must have been successful, but then the foster parent died and the young could not be maintained.

r) Survival young No conclusions can be drawn concerning the relationship between

the percentage survival of young wild shrews and litter size on the limited information presented here • The death of two females before

the young were weaned may have been caused by excessive disturbance

or to an unbalanced diet. Prior to the death of one female, two of the smallest of her litter of nine young had already died • Of the six S. araneus brought up under semi-natural conditions outside, only three survived to independence. It seems unlikely that a complete litter of nine would survive because of the inability of the female

to feed them all, whereas a litter of 4-5 would sfand a better chance

of survival. It is suggested that there are three most vulnerable

stages in the post-natal development of the young 1) the first week of life in a large litter (6-9) when growth is rapid and variable between

individuals such that the smallest and weakest succumb in competition

for food with stronger siblings )ii) weaning, when a decrease in weight

was observed, possibly caused by the inability of' the female to continue

feeding the young, and deaths may result in the transition to an

invertebrate diet; 3.ti) independence leading to competition for nesting

sites, territories ar4 food, when death mnay result from agonistic

behaviour, and predation of inexperienced young shrews. - 14 0 -

It ma be concluded from this work that female S. araneus

produce litters of up to nine young with a mean of seven of which only three or four survive to independence. Pot.natal growth of the young is very rapid and independence is achieved at 23-25 days after birth, by which time the young have increased in size to resemble sub..adults. Survival of the young depends upon their

ability to compete with siblings in the nest for food, upon successful weaning and upon the ability to compete for nesting sites and food, and - avoid predation. The number of young surviving in the wild is an

indication of this success or lack of it, Dehnel (1952) records females producing three or four litters during their life-time in captivity. Even if two litters are produced inthe wild with a mean of seven young per litter and a survival of 5(,one female could produce seven young. With an approximately equal sex-ratio (described in Chapter z) about half the yaung produced would be females with breeding potential, giving a population increase of 2..5 each year. The studies

of population sizes and survival in The Wilderness in Chapter 2 have already indicated that such an increase does not occur and that mortality is particularly high at the juvenile stage of a shrew's life.

i) Seasonal weight changes

The weight changes of individual S. araneus sub-adults are shown in

Fig. 11-7A & B. These include four shrews which were kept in outdoor enclosures until December but thereafter were moved inside for odd periods (denoted by empty triangles). All four showed a decrease in weight from September/October to December outside which parallels the decrease exhibited by wild shrews • A spring increase towards maturity was found both in shrews kept outside and inside: this commenced in

January which is earlier than in wild individuals, except for S.a.8M which did. not gain any appreciable weight until March. This contrasts

with the results of Pucek (1964) who found that shrews caught in the summer and. maintained in captivity inside at 10-20°C during the winter, A LEGEND • Shrews outclooa Shrews indoors

21 I6 U o 18 15 0 a; I- D 15 4 -C 4- 4- 0 0) 12 13 C a; a- E9 12, a; 1-6 11.0

0 —3 1 —6 I

4- -c 0)

> -u 0

U,,, Uf

Fig. 4-7 A The seasonal changes in body weight of captive individual S. araneus (S. a.) outdoors and indoors, and Eeasonai changes in outdoor temperature and cFayIcih. LEGEND • Shrews outdoors B a Shrews indoors

J7Ff D7L1

- 16 Nf1M . 154M

; 13•

C N.L3M

IC I -,

F IFfJ I,,

Fig. 4-7 B The seasonal changes in body weight of captivP individual s araneu (S. a.) and N fodiens (N f ) kept outdoors and indoors. 1111. - did not undergo a winter depression in body weight, neither did they exhibit the spring jump in weight to maturity. However, his shrews caught in winter did show an increase in body weight to maturity which also commenced much earlier than in wild nnes. His results for summer-caught shrews may be comparable to those obtained from animals maintained inside at 20°C for the duration of the winter.

Long.term measurements are only available for two . araneus ,but a weight change is apparent in these two sub-adults which was of a completely different form from shrews maintained outside. Rather than decreasing to December, weight increased but fell slightly in January/

February. S.a. 2M then exhibited the spring jump in weight to maturity but S.a. kM continued to decrease throughout the spring.

Included in Fig. Li-7A are the maximum and minimum temperatures recorded in the enclosures at the times of the weighings 3 and seasonal. changes in daylength. The weight decrease of S. araeus accompanied a decline in temperature but although minimv.m temperatures continued to fall to as little as -6.0°C no further decrease in weight occurred. Even when minimum temperatures were no more than 1°C, body weights were increasing towards maturity. 1978 was characterised. by a particularly cold, early spring but this did not affect the timing of maturity in captive shrews outdoors wh3 ch pursued a pattern of growth similar to those ir the wild. A much closer correlation is apparent between seasonal body weight changes and daylength.

Of the N.fcd in captivity none were maintained outside for any length of time and the weights of those kept ins.de showed considerable variation, as Fig. k-7B shows. In Nf. 2, Nf. 3 and Nf.11 a common trend is evident, resembling that of S. araneus kept inside, when weight increased fromthe summer onwards but suddenly decreased in

January. A fall in body weight was also evident in two N. fodiens

(Nf. 3M and Nf. kM) which were placed outside in February and. - 1112 subsequently decreased in weight into the spring. N.f.1M behaved differently,however, and showed a. decline in weight in November and January even though it was maintained inside.

The body weights of S. minutus, C. russula an]. C. suaveolens were also investigated. S. minutus showed no significant weight change either inside ,or outside where one individual was kept under cover but was subjected to natural temperature and light regimes.

Both C. russula and C. suaveolens kept inside showed no 9marked change in body weight between the autumn an]. winter but exhibited a spring increase. Two placed outside showed no appreciable change in weight an]. a spring increase was not apparent. It is not known whether wild crocidurids undergo the weight decrease inwinter in temperate climates which is so characteristic of the soricids, for no extensive field st4ies have been conducted on this species in

Britain.

The body weights of captive shrews showed considerable variation, as is evident from Fig3. L1-7A & B • This situation was also observed by Pucek (1964) who found that the body weights of S. araneus in captivity fluctuated almozt from day to day. During a period of

L1214 days body weight may increase by 2O-2 of the initial value but these rapid changes he observed more frequently inwinter than in summer. Body weight also increased at the time of moulting but no such correlations were made in the present study. The weight changes of shrews in captivity may be compared more satisfactorily by a consideration of the trends exhibited by the total number of shrews rather than selected individuals. The collective weight distributions of captive S. araneus and N. fodiens are shown in Fig. Z1_8. S. araneus maintained outside showed a trend. of decreasing weight in the winter an]. inrreasing weight toviards maturity in spring, despite the considerable individual variation. Shrews riiantained. inside shoied no such change

LEGEND G indoors co1e outdoors animals

N fodiens 180-189 170-179 160-1&9 150-159 140-149 130-13-9 120-129 0. __ 110-119 100-10.9 0I VI VI 90-9c 0 0 S araneus 140-14 9 - 13.0-139 0) • 12'0-129 110-11 9 100-109 90- 99 80-89 70-79 6 0-69 50-59

40-49 - I S 0 N D J F M A 1977 1978

Fig. 4-8 The seaonaI weight distribution of groups of captive S araneus and N çodlens kept outdoors and indoors. - .L1J - and weit distributions are more clumped, both for S. araneus and N. fodiens. The spring increase shown by S. araneus was not revealed by N. fodiens,eithef because captivity upset dcveloxient or because the onset of maturity occurs 1.ater than April when the final weighings were made.

The irregular fluctuations in weight of captive shrews maintained in solitary confinement inside, with constant temperature and lighting conditions and an unvaried diet, indicates an upset of the natural seasonal rhythm. This is further illustrated by the tendency of shrews to increase fat deposition to a highly exaggerated extent even after short periods in captivity (persond]. observat5 on). This makes comparison of weight changes of laboratory kept shrews with those living outside in near-natural conditions difficult. Those kept outside did. exhibit the weight changes found in wild shrews. A winter decrease in body weight was apparent, if to a lesser extent than in wild shrews, even though food was at all times abundant. This seems to

indicate that food. is not the limiting factor in the growth of shrews in the wild, and. the winter decrease in weight cannot be attributed to a lack

of food • This furthrr reinforces the coiiclusions in Chapter 3 where no

depletion in food resources in winter was revealed.

The weight changes may be associated more with temperature

fluctuations aid changes in daylength,partiularly if shrews outside are compared with those kept exclusive).y inside .t 20°C which,although

they did show fluctuations in weight,did not exhibit the normal cycle of weight change. That temperature is not the prirae limiting factor

in growth,however,is indicated by two observations:i) some shrews

kept exclusively at 20°C did. exhibit a weight decrease at some time

during the winter/spring and ii) the weight of shrews outside began

to increase in January aid continued into the spring despite the

prolongati.nn of the cold, wet weather. That temperature does not - - wholly affect weight change is also demonstrated by animals taken

inside for periods after living outside for the groater part of their captivity these continued to show a normal weight increase to maturity in spring, such as was not undergone by all the S. araneu'3 or N. fodiens maintained exclusively inside.

It may be concluded that the seasonal weight changes of shrews cannot be attributed to the availability of food resources. Temperature ami other climatic factors may influence body weight but are not of prime importance, for althnugh a decrease in body weight occurs simultaneously with a fall in temperature, the spring increase towards maturity commences before an appreciable rise in ambient temperature occurs • The close correlation between body weight and daylength suggests that daylength may be an important influence. Daylength and temperature changes may, therefore, provide a cue for changes in body weight but further work is required to investigate the precise conditinns uhSer which these changes occur before the mechanism can be explained. - 145 -

SECTION D: SEASONAL WATER & FAT CONTENTS

PART T INTRODUCTION It has been suggested that the decrease in body weight of

overwintering shrews, described in Chapter 2 and in the previous

section,is produced by changes in dimension of the cranium and certain intrrnai organs (Pucek, 1963, 1965) and also by a decrease

in body water content (Myrcha, 1969). However, little is known of

the seasonal fat content of shrews and its role in affecting changes

in body weight. Certain anomalies are apparent when losses in body

weight are correlated with water and fat contents, for exainp].e,

Myrcha (1969) says that a decrease in body water content accounts

for most of the total weight loss of overwinterir.g shrews, aixl at the same time reports an increase in fat content of shrews inwinter. In contrast, Buchalczyk and Korybska (19611.) found fat content of shrews

was greater in summer than in winter. Moreover, it was observed. in the present study that captive shrews kept in corstant , warm conditions

with abundant food increased in weight, regazUess of season to an extent not observed in the wild which seemed to indicate that

excessive fat deposition was occurring.

It was decided, therefore, to undertake a study or water and. fat

contents of shrews, particularly S. araneus ,in an attempt to discover whether seasonal changes do occur and, if so, hcrt these changes are correlated with changes in total body weight.

rr II: IEPffODS Two different methods were employed for extracting the fat from autopsied carcases in ordsr to determine its quan.ttty. Initially, a simple ether- shaking technique was used but this was superceded. by a modification of the standard ,Soxhlet method when the apparatus became available. For both methods the carcases, with guts removed, were dried to constant weight at 60°c (48-58 hours). Water content3 were calculated and the dried carcases were prepared for fat extract ion.

1) Ether-shaking technique This method of estimation was based on the recovery of the fat from extracted carcases. Each carcase was cut up into small pieces, ground in a pestle and mortar and soaked in 100% ethanol for 30 minutes. It was then placed into a lOOnil volumetric flask and approximately 30m1 of diethyl ether was added. The top of the flask was sealed over with aluminium foil to prevent evaporation and spillage. Flasks were placed in a mechanical shaker for 6 hours after which the ether and fat were decanted through filtcr paper into pre-dried and. weighed lOOmi beakers and left to evaporate in a fume cupboard. The beakers, containing the fat, were dried at 6o°c for 2 hours and weighed. The extraction procedure was repeated with fresh solvent until a constant weight of fat was obtained. This usually required three extractions of six hours each. ii) Soxhiet method In this method the difference between the initial weight of the carcase and its weight following extraction was taken as a measure of its fat content • The apparatus is described by Sawicka-Kapusta in

Grcdzinski,KLkowski & Duncan (1975), Each dridd carcase was cut up into small pieces and wrapped in white laboratory tissue which was stapled at each end to make a small package. Fch package was soaked in 100% ethanol or 30 minutes to denature the proteins and prevent them being washed out with the fat during extraction. Between three and four packages,deperxling upon their size,were placed in each Soxhiet apparatus and extracted with ether solvent over a waterbath at k5-50°C. The length of extraction depended upon the fat content of the carcases. Extraction was continued until there was no trace of yellow fat in the ether in the bowl at the base of the apparatus and

- 1117 -

the packaged carcases had achieved constant weight. This usually

involved three 8-hour periods between which the solvent was replaced. Packaged carcases were left soaking in solvent overnight between

extraction periods. Following extraction the packaged carcases were

left in a fume cupboaxi for the ether to evaporate ,and dried for two hours at 60°c. Carcases were then removed from their wrapping aM.

weighed ,and the percentage fat content of the dried body weight

calculated.

Water and fat contents were determined mostly in S. araneus co].&ected from Alice Holt Forest, but other S. arareus together with some S. minutus and N. fodiens were used, when available, from other study areas • In additinn, some detrrminations were made of the water and fat contents of a small number of shrews which had been maintained.

for some time in captivity. Numbers of individuals of the different species used in water and fat determinations are &huwn in Table 4-14.

Fat Content

Soxhiet Ether- Species Location Water Content Method_ shaking Method Wild

Alice S.Holt araneus Forest 160

S. araneus The Wilderness & Monks Wood 34 311.

S. minutus Alice Holt Forest & The Wilderness 19 19 -

N. fodiens A1resfox, }Iants. 5 S- Captive

S. araneus 8 7 -

S. minutus 4 L. -

N. fodiens 5 6 -

C. russula :3 3 - C. suaveolens 2 2 -

Table kltf, Numbers of shrews used in water & fat extractions

- -

PART III RESULTS 1) Comparison of the two methods of fat extraction A comparison of the Soxhiet and ether..shaking methods of fat

extraction revealed that the results from the latter were uniformly slightly higher, and these differences were statistically significant (p o.oi). Table Li_15 shows the results obtained from carcases which were divided into two equal portions, one portion being treated by the ether-shaking technique and the other by the Soxhiet method.

The higher fat figures obtained by this method may be explained by the fine nature of the material. Carcases were ground up for

extraction and it is possible that some of the finer material escaped

through the filtering apparatus into the fat-collecting receptacle

below. As most of the results were obtained by the Soxhlet method,

the 'esttlts from the ether-shaking technique were not included in the estimates of fat contents of different species, age classes or captive

shrews • However, all arcases for the first six months of the seasonal data were treated by the ether-shaking method and so are included for a comparison of mean monthly means with a correction factor f -0.73% based on the experimental comparison of the two methods.

Table 14-15 Fat Content

Ether..snak1n Technique Soxhiet Difference

5.50 4.40 1 • 10

7.42 6.42 1 • 00 5.51 5.30 0 • 21 6.63 4.96 1.67 5.42 5.12 0 • 30 6.95 1.50 4.6]. 4.34 0.27 1.140 6. ]. 4.79 0.27 5.24 4.97 0.60 5.83 5.23 •5.98 0.14 6 • 12 Mean differcnce - 0. 33 O71 - L9 - ii) M1d Shrews The mean water and fat contents of males and females of all the species examined are presented in Table 4-16. There was no statistically significant difference between the sexes of each species in either water or fat content. In the case of S. araneus which was obtained from two different sites, no difference in water content was found but the mean fat contents,as a percentage of dry body weight, of all individuals from The Wilderness & Monks Wood were statistically significantly higher than those from Alice I101t Forest. - There was no significant difference between the mean water contents of the species examined but differences in mean fat content were apparent. The highest mean fat content,as a percentage of dry body weight,wae found in S • minutus and the lowest in N. fodiens ,andboth were significantly different from S • araneus from The Wilderness and Monks

Wood. The relationship between body weight and fat content of Soxhiet- extracted shrews can be seen in Fig. 4-9. A comparison of the fat content of the three species shows a decreasing fat cnntent with increasing body weight, but this trend is only evident where the differences in body weight are well marked: within a species, such as araneus and S. minutus ,there was no decrease in fat content with an increase in body weight.

The mean water aM fat contents of the different age classes of shrews (both sexes combined) and the timeoi year at which they occur are shown in Table 4.-17,together with their body weights. No seasonal data wCrt obtained from N._fodi.ens so the results are limited to

S • araneus from Alice Holt Forest and The Wilderness & Monks Wood and S. minutus (all sites combined). The water contents of sub-adult

S • araneus from both Alice Holt Forest and Wilderness/Monks Wood in winter were statistically significantly lower than those of juveniles in summer, but this is not correlated with body weight as there was no significant difference between water contents of juveniles and the a 0 '- 'i j

CD CD CD

Co CD CD CD

#_%

I_J I-' p a CD CD '.3 CD 0 Co Co ci. C,

o 0

p. 0% CD CD S.

0 0 pi

Ui 0 Os 0% \0 • Os - • S • S 0) 0% ¼.O 0) ci. C 0 -4 — ,—. F—' I.i I a Ui

4. -4 LpJ 0% — '— -- -__ '-

CD o Os 0% 0% Cs - - Ui • . S I I-' 0 -3 03 CD C) '.o o ci F—., s F— 10 CD Ui 3 0% tO ci. • • • 'D CD - t o t 0 i.-' 0 0 '— -___ .— —

I. a 0% 0% Os 0% '-a - 0% C, • S S S Ui Is) tJ 0) cj 0) i.- I-' CD F— F—., ID Ui s) - f\) CD • . . - I') .P I-i CO 'i — -4 r-4 _ —

- Os Ui • S S S - 0 0'. 0) -3 ..o C) .--' — —s -a o i- a a a • • S czo 0) '0 Is) 0) '0 .__ ___ '—s I lcD Os Ui - • • • • CD C) g '0 LJ as p F-' i-a o i.- 0 0 CD • . S S 0 0) 0 0'. —s-

- 0'. Ui • . S S Is) 0) 4-a -3 o co 0) i-a .—' ___% a— C, o i-' 0 0 • • S S '0 Ui 0) 0) 0 CjJ Ui .,— —.. __ '- LEGEND • S araneus (Alice Ilolt Forest). O lOOr 0 Sminutus A N fodiens.

0 0 0

C 0 4- 0 C S 0 U 0 I. 4- • . 0 • $5 U- A 10 S • 0 • • . 0'

.eA. .

4 A

3-0 S 4 .

2.0'— I. I I I I I I I I 06 08 10 12 14 16 18 2<) 22 2.4 26 28 3<) 32 34 3-6 38 40 Body Weight (g.drywt)

Fig 4-9 The relationship between fat content (as a percentage of the dry body weight) and body weight of wild shrews. U) I)' D' pq .$. CD I CD I.. I-I I-' CD CD Cl. CD CD CD

p CD C'. CD ij U) I-I ,-.. CD 1 P

Ii 03

CD Ii

CD Q% C c I • • C) 8 _ 0 1.-i CD C) — 0% . * C) C%)r) I-'Lri P'CD CD I CD • . • • • oc CDCD 1' _ _.i NZx Ji ... . •c4

6 IC!) o. o-r o. j$ CD 0 • S 5 • . • CO 'O QO% I•-".JJ .p CO %- S.-

CD •0 -.]. .o • F. I- k..)

CD IC) N -S 0 ;3 .4 0 .) N i P CD CD I •o • . • • . Cs) - 0 . r co l CD CD (5) Nj 0 0 iS . S.. '- • & o CD -S I-jt N CD CI) • • U U • S o o i-a C'.) 0 w . jj 0 ' .. S..' a

I CD N N I- • S •

. I . I-I' °' I' i-'u o •0 • S • • • • '0t) '- co co i IC,. CD Ic'.Ii .5-

I-'-2 Oci • • • • CD U)t*) 4-cD la.1 %O -... 5-. 5- t . - 1)0 - larger mature adults. No such difference was apparent for S. minutus but the numberof sample8 examined was insufficient to reveal any trend.

The apparent differences in mean fat content between the age classes in Table 4-17 were not,in fact,statistically significant. There was also no difference between the fat content of four pregnant females bearing between four and seven embryos and other non-pregnant females at Alice Holt Forest. However, the fat content of the two pregnant females from The Wilderness and Monks Wood, both of which had nine well- developed embryos, was higher,at 6.21%,than non -pregnant females from the same sites. The seasonal changes in mean monthly water and. fat contents of S • araneus from Alice Holt Forest are prebented in Table 13_18 and. shown graphically in Fig. 4-10 and Fig. 4-11. Mean body weight is included in each graph for comparison. From Fig. 4-10 it is apparent that there is a winter decline and a summer peak in water content of shrews and these coincide with the respective decrease and increase in mean body weight. This is particularly evident for the years 1975, 1976 and the first half of 1977. The latter part of 1977 and the first few months of 1978 do not show this trend, possibly owing to the paucity of data.

Despite no significant difference in fat content between age classes of shiews • there is evidence of a seasonal change in fat content as a percentage of dry body weight, as Fig. 4.-].1 shows, and this also coincides with changes in body weight. The lowest fat contents occurred in the summer and the highest in autumn and early winter (between SEptember and becember) although the level of % fat content differed from year to year. The differences in fat content between shrews in

August and December 1976, and August and September 1977,show a statistically significant increase in the autumn and winter (PO.05).

A statistically significant decrease in fat content was found in spring and summer when results for December 1976 and May l976 and September

"r; 4. 0 V\• 0

I I

it: 4,

a, • C) C\• 0a 0 4. 0 %- a 0% N t- 0% 0% t,- N • • • • r4 I-I 'U4 4 '0 4. a) 'o 0 • • r4 a a r4 4 0 o

a 44 0 11) 4., 4,a, 0 .4.,

a; IF 4, N '\ 0 '0• •-• • a N u N 1-4 4. cc • • 0% • • cc rl 'U4. N U r1 N a) .-I 4 0'• 4.4.a oa ua ,- u v-4 0 'Qs 4. a, >4 I-I r4 P. Ps • (0 f-4r4 • to c • ; ! o 'i 0 0 Ci 0 .c, •.— s- LEGEND ij- Water content 0-0 . (data from The Wilderness Monks Woodi 690 +

68-C 110

670

10-0

66-0

.1- C a, 650 9-C 0 U 4-

w 640

63 0 0

62 70

60.0

590 1 1 I I '—'5-0 J A SOND J FMA 1975 1978

Fig. 4-10 The seasonal changes in mean percentage water content and body weight (with standard errors) of S araneus from Alice Holt Forest. LEGEND .-. Fat content (SozhIet). .'

1-0

10-0

4- 90 C. a) 4- C 4- 0 - U 0) 4- a) 0 8 U- 0 > 0

7-0

5-0

4-0

Fig. 4-11 The seasonal changes in mean fat content (as a percentage of the dry body weight) and body weight, with standard errors, of S araneus from Alice Ilolt Forest. - 151 -

1977 and April 1978 are compared (PtO.05). The differences between

December 1976 and May 1977 are not sIgnificant, possibly owing to the small number of samples.

These seasonal changes in body Mater and fat content coincide with seasona]. changes in body weight but it has already been demonstrated that,within a species, they are not directly related (see Table 1f-17 and

Fig. 4-9). It seems, then, that seasonal changes in water and fat contents of shrews occur independently of body weight. iii) Captive Shrews

The water and fat contents of captive shrews which had been maintained. in constant cenditions in the laboratory (at 20°C on a diet of blowfly larvae) are compared with wild shrews in Table 4-19. Water contents of captive shrews appeared to be lower than those of wild. members of the same species but only In the case of S. minutus was this difference statistically significant. The small number of captive shrews examined and the wide Individual variation makes comparison difficult. The differences found could be attributed to their condition at death and the length of time before they were collected and frozen prior to autopsy. When collected, all carcases were stored frozen in sealed polythene bags to prevent water loss. The fat contents showed large differences between captive and wild shrews, with values for captive irAi'criduals being between seven and ten times greater than for wild. ofles in all species examined • No results were obtained for wild C. russula and C. suaveolens. The accumulation of fat was independent of the length of captivity for instance, after a period of one month in captivity fat content had reached 38.&. in

S. minutus and between 14.37,ard 65.8% in S.araneus. After four months In captivity fat content of S. minutus had reached 4O. and of . rancn was 9.3-40.7%. It seems that fat content increases to a maximum value under optimum corditions within the first one or two months of captivity, but that there is a great deal of Individual CD 0 -'

CD C) c+ CD 0 -' ,- £) CA CD tO CD CD

'-3

$4 C) I-' CDC) a cx' - C..., cx'

I—I \0 J1 0 - • . lc+a

- o\ '._J, _,1 U' '0 tQ I-' cx' c • S • • • CD C) '0 \O (-..J - 0 CD o - 0 C...' ' __ -' -S C) '-3 t..) l- £\) L.J -3 0 • S S • • i cx, cx' o a -I 0\ CD 0 .— '- '- .— S.- cf CD

CD 0 H3 -is. 1') - 1.) C...) r.) 0 '0 M 0 • S S S S r, - i....' 0 - CD r'. r - 0 I\) C) -S #- __ -s 0 LJ I-' '-.0 -.3 • I-' S • 's.D c+ • I_i (_..., • CD -' - sO' —a N) 5- I-i S..- s-.' Ui 5-

'0 N. a • S S II-J 0 0 cx' t cx' cx' O\ 10 p-S 02 L-1 d S S.- 0\ 0'. 0'. - - c-f I.J • • • CD Ui Ni 0 Ui cx I-' CD ___S. — C)

0 - N) I_J c-f '-I 's_n --3 -3 CD c-f S.- S._ S.-.•

CD CD c-f

'sO Ui 03 CD I-, 0 '-J* %__ s.__ 5- c-f - - variation.

Discussion

The most striking results in this study are the apparent seasonal cycles of water and. fat contents in 8• araneus ,which have also been demonstrated by Myrcha (1969) in Poland. No sexual differences were found in either study. Myrcha, too, found a high water content of shrews in summer which decreased steadily in autumn and. winter and rose again in spring, a.rd similarly with S. minutus and N. fodiensMyrcha (1969) states that the decrease in water content could account for most of the reduction in body weight in winter and suggested that the decreased water content led to a reduction in cellular metabolism. However, the change in water content between sunuer juveniles nd overwintering sub-adults in this study had a mean of only 2.95-3.89%,which is not sufficient to account for the decrease in weight of sub-adults in winter. For example, ii' juveniles with a mean weight of 6.Li6g lost 3.89% of their body weight by tissue dehydration they would only decrease to

6.21g, and those with a mean weight cf 7.l3g would only decrease to 6.85g which is considerably greater than the inan body weights of

5.50-6.31g of sub-adults recorded in winter. Often, young shrews weighing 7.5g in autumn were found to decrease to aá little as 5.5g in December/.Lrnuay. This would require a total weight loss of 26.7%. The greatebt decline in percentage water conLent recorded was only

6.2 between Septeiiiber and. December 1975 which could account for little more than a quarter of the weight loss shown by S. araneus in the same months. The greatest decline recorded by Myrcha (1969) was only . This casts considerable doubt upon the significance of the decrease in water content of shrews inwinter.

A seasonal cycle in faL content was also demonstrated by Myrcha

1969) for S. araneus, 3. 'ninutus and N. fod lens. In suruJer, values - -

for S. araneus were constant, but from September onwaztls fat content increased to a maximum in January and then decreased to reach a

minimum in March before rising to the summer level in May. Fat

content in the present study reached a maximum between September and December. ?'yrcha obtained similar results for S. minutus but with a minimum in October and, maximum in December. !4yrcha found that fat

content .was greatest in shrews from her more northerly study area. This was also noted in this study where shrews from The Wilderness &

Monks Wood had higher values than those from Alice Holt Forest, - approximately 100 miles due south. Pucek (1965) studied the brown adipose tissue of S. araneus and also found greatest q ,uantities in

winter and lowestin summer. However, Buchalczyk & Korybska (196k)

dissected out the accumulated brown adipose tissue from the scapula

region of shrews and ,despite considerable sexual and annual variationa) found high fat reserves in summer and low reserves in winter for young adults and. high quantities in spring and summer for old adults. AccoHing to Myrcha (1969) brown adipose tissue is the only form of fat reserve found in shrews in natural conditions. A seasonal cycle

in fat content was also found in the rodents Appdemus flavicollis and

Clethrionoinys glareolus by Pucek (1973) but this took a different course in successive years, sometimes being high in winter and sometimes low,

Two questions arise from the study of seas ens]. fat content of shrews. Firstly, is there a significant seasonal fluctuation in fat content at all? Relevant to this point is the way in which the shrews were collected f,r analysis. In the present study shrews were collected from Longworth livetraps in which they had died. It is possible, Since no comparison of the effects of different trapping techniques on fat contents was made, that the values obtained may not be indicative of the true values because of the starva Lion and stress of shrews prior to death and subsequent analysis. If tis were so, however, it seems - 1514. -

unlikely that residual fat contents should be greater in winter than

in summer, for under cold conditions fat could be expected to be

utilised. in metabolism. Moreover, Myrcha (1969) found no significant difference between fat contents of shrews caught by live-trapping and snap-back trapping.

Secondly, if a seasonal fluctuation in fat content can be assumed

to be genuine how significant Is it' It has already been mentioned

that the seasonal cycle in water content of shrews Is of a very small inagnitude,and this may also be said of the changes in fat content which

in this study were less than 2% between the maxiniuin and minimum values recorded, and only as found by Myrcha (1969). Although the changes

have been found to be statistically significant, their importance in the overwintering of shrews would appear to be negligible when

compared with the values for fat cnntent in captive shrews.

That captive shrews are able to accumulate large quantities of brown adipose tissue was also found by Pucek (19614.) who correlated It with body weight and length of captivity, with maximum values being

recorded after 3_Lf months of captivity. No such correlations were made

in the present study,for shrews were found to accumulate large fat reserves regardless of body weight and after periods in captivity of only one or two months. Accumulation of lipid has also been observed

in other small mammals, including Apodemussy1vatcus, Clethriononiys glareolus and Microtus agrestis kept under laboratory conditions for

short periods (Ferns & Adams, 1974). The xeason for this accumulation

is probably threefold: 1) abundant food ii) constant, warm temperatures

arid iii) decrease in the effort required to obtain food leading to

reduced activity. Despite such artificial conditions, this does show that undrr optimum conditions shrews are capable of storing fat far in excess of the values ever recorded In wild shrews. - -

Myrcha (1969) suggests that shrews accumulate fat in the autumn for use in winter when enorr expense is high because of low temperatures, and in spring when tbere is an increase in gro.ith and activity .d.th the onset of the breeding season. In,the light of the results of fat storage in captive shrews,however, it seems unlikely that the extremely small fat reserves found in wild shrews are of any particular value to them. Moreover, the fact that maximum values of fat content were recorded into the depth of winter indicates that fat reserves are not utilised at this time of year as they are in summer when values were - lower. A reduced activity or a high calorif Ic value diet might explain the higher winter values, but no particular seasonal change in the diet was in fact recorded.

It Is concluded that seasonal fluctuatiors in water and fat contents of shrews recorded in this and other studies are real, despite the possible errors resulting from the technique of collectinn of carcases for analysis, but that these fluctuations are not of sufficient magnitude to be of any importance. The variations between studies may be attributed to the different methods of treatment of the caxcases.

Water content was found to decrease simultaneously with body weight in winter but was not sufficient to account for the decrease in body weight. Fat content increased in autumn and winter, but to such a small extent compared. with captive shrews, as to be insignificant in its value as a fat reserve to overwinteri ng shrews • A higher fat content in winter compared with summer Indicates that fat was not metabolised. under cold. conditions, - 15c -

SECTION E: STUDIES ON THE ACTIVITY OF S. ARANEUS

PART I: INTRODUCTION

The activity patterns of shrews, including S. araneus, have been studied both under laboratory conditions (Tupikova,

1949; Cro.icroft, 195L1.b; Gebczynski, 1965; Loxton et al, 1975) and under field conditions using krapping results as a measure of act.vity (Jansky & Hanak, 1960; Shillito, 1963b). Comparatively little s known, however,of the relationship between activity and temperature or the seasonal differences in levels of activity, although results of trapping seem to suggest dissimilarities between winter and. summer.

The aim of this study was not to investigate in detail thc activity patterns of shrews, which have already been described by other workers, but to compare activity under different temperature regimes in an attempt to investigate any seasonal changes which might account for the low numbers of shrews trapped in winter.

PART IT • MITHOD

The activity of S. araneus was investigated using time-lapse photography. Each shrew was maintained in the standard, lid-less bin (described in Section A, 1(1)) to which it .as accustomed, containing a black perspex nest-box with bedding. Sawdust was placed in the bin,togethr with a small amount of hay sufficient to provido some cover without hiding the shrew from view. Water and food (blowfly pupae) were placed in jars at the opposite end of the bin from the nest-box.

Filming was carried out u'3ing a Vinen Mark 3, 16mm Scientiiic Tizie-lape Camera set to give a time-lapse interval of 15 seconds between exposures and coupled to a short duration electronic flash. - .1.)( -

The Camera was set up above the bIn so that the whole enclsuro was in the field of view, Shrews showed no reaction to the flashes, continuing to feed and explore for considerable periods in the open without any apparent disturbance.

Three shrews were each filmed for two consecutive 2hour periods and at two different temperatures. Activity was investigated in the laboratory,where shrews were maintained at 20°C,and outdoors in a shed )where they were protected from wind and rain but where temperatures ranged from 1-9°C. Filming in the laboratory at 20°C was carried out In May when ambient temperatures outdoors were also quite high, and outside at 1-9°C in February under wintrr conditinns. Natural lighting was used, which in the laboratvry was augmented by fluorescent tubes,and outdoors was provided via windows in the shed. All the shrews used were maturing adults.

Following processing, each film was projected on to a screen using a Specto l6mri Motion Analysis Projector, and all the frames in which a shrew appeared were recorded. The total number of appearances by the shrew in each st,ccessjve 30 minutes was summed and this was taken as a measure of the activity outside the nest. Shrews spent their time outside the nest in feeding and explor) ng and rarely remained inactive when in the open. When a shrew was outside the nest but failed to move in successive frames it was not considered in this instance as being active.

Part lilt Results

The results for all the shrews filmed were combined for each temperature regime, and the mean number of frames in which a shrew appeared In each successive period of 30 minutes during the whole of * 158 -

a 21f-hour period was calculated. Fig. 4-12 shows the activity of

S. araneus 3both in terms of total number of appearances and total

number of minutes in each 30minute period spent outside the nest box in cold, winter conditions and. in warm, summer conditions.

Shrews were active throughout the 24-hour period, except at the lower temperatures when there was no activity betweeno7.30 and 09.30 huurs. Shrews under both temperature regimes were significantly more active during hours of darkness than in daylight (PcO.Ol) and

this was particularly evident at 20°C in May when the dark period was comparatively short.

The major differences between the activity under the two temperature regimes was not so much the cycle of activity as the level of activity. Fig. 4-it suggests that shrews were more active during the day at 20°C than they were at l-9°C,and similarly during the night, but when activity during daylight and darkness between each temperature regime are compared the differences were not,in fact, statistically significant. Hovever, when activity over the whole 2Ll.*hor period is compared shrews were significantly less active at low temperatures than at high temperatures (PaO.Oi): shrews at 1-9°C spent an average of 18.8% of their time in the 24 hour period engaged in some activity outside the nest (feeding, exploring and grooiing), whereas at 20°C activity rose to 27.&. The peaks of activity were also consistently shorter at lower

temperatures: at 20°C the minimum and maximum duration of activity

in any 30-minute period was one minute and 18 minutes respectively,

while at 1-9°C it was zero and 14 minutes respectively. X4IAIpy jo S4flUIw

I- hr 0

E I-

suooeddy O •°N - 159 -

PART IV DISCUSSION

The main criticism of the metIod employed in this study is

of tne interval between exposures the film thich was sufficient

to enable a shrew to vacate the nest-box and retreat into it again between frames an& so not appear on the film. However, shrews

rarely appeared in a single frame, tending to have activity periods extending for several minutes at a time when they appeared in successive frames, showing no aversion to the u1ahes. Consequently, the results are assumed to be representative of the daily activity

of the shrews under investigation.

The results of the activity of S. araneu at 20°C show similar results to those of Crowcroft (l95th),with a major peak between

8.3Opm and 4.30 am during darkness, a minor peak in the morning between 6am and lOam,and reduced activity in the afternoon. At

1-9°C major activity again took place at night and diminished towards daylight, with another peak between 9.30 ard lO.3Oam followed by reduced activity before a progressive increase prior to dusk.

Despite the dark period being longer during the filming in February than in May, the activity of shrews at 1-9°C in February was still considerably lower than in May.

The reduced levels of activity at low te1peLatures shown in this study confirm the findings of Jansky & Hank (1960) who reported daily activity to vary with season, and Gebczyn' 1ti (1965) who found that the pattern of activity was the same in winter ac in autumn and

spring but that there were seasonal differences assctated hith the

intervals between peaks and. their extent, wit1' greatest activity in

summer. Loxton et al (1975) also found that 1 although the diurnal

cycle of activity of captive S. araneus was similar in winter and

summer, the overall activity was lower in winter. It ha already

been shown (in Chapter 3) how temperatures beloi r ground in winter are - 160 - considerably higher than on the ground surface, so if activity on the ground surface is reduced, heat loss can also be reduced. This may in turn reduce ener requirements 3 allowing a fall in food consumption in cold conditions (which has been demonstrated for S. araneus in a previous section). Seasonal changes in activity could be of great importance to the oveiwintering success of shrews, and could also provide an explanation for the reduced numbers of captures of shrews in winter. - 161 -

SECTION F• SUMMA1Y FOOD CONSUMPTION & AS3T.flILATION

E,cperiments carried out on the food, consumption and assimilation of captive shrews in metabolic cages gave the following results.

1. No sexual differences in food consumption were found.

2. Absolute daily food consumption decreased with size inthe series

N. fodiens, S. araneus, S. ini.nutus but was lowest in C. russ•

Daily consumption as a percentage of the body weight showed an increase with decreasing weight in the series N. fodiens (mean 5$), S. araneu

(mean 87%), 5. minutus (mean 126%). The crocidurids had the lowest consumption (48-5%). Within a species, daily food consumption was not related to body weight.

3. Assimilation efficiency decreased with increasing body weight in the series S. minutu (mean 86%), S. araneuR (mean 85), N. fodiens (mean 7) but was lowest in the crocid.uride (72-32%). Within each species,assimilation efficiency and body weight were not related.

J. Assimilation efficiency did not change with ambient temperature and

neither did. daily food consumption, except in 5,. araneu when

consumption was reduced at low temperatures.

5. It was concluded that the metabolic rate of shrews,in terms of food consuinption ,is not directly related to body weight but is characteristic of each species anti of each individual within a species. It is suggested., therefore, that a decrease in body weight of shrews in winter does not necessarily reduce absolute food requirements.

Wter cc1 itent &caiorfic va1ue The results obtained of the water contents and calorific values of known invertebrate prey iteris and of the clorific values of shrews and. their daily eror rei1remerLts are listed. below. 1. Adult colccpterars, the major prey of wild ehrews,had the highest calorific vaJue (mean 5.9kca1/g d.cy weight) and the loiest water content - 162 -

(mean 62%) of all prey items. Insect larvae, also important in the diet, had a calorific value of 5.77 kcal/g dry weight. Isopods had. a relatively low calorif Ic value (mean l. .0Z} kcal/g) arid water content

(mean 67%). It was concluded that the diet of shrews is a function of' prey abundance rather than selection for food value.

2. The calorific value of S. araneus carcases had a mean of 5.55 kcal/g dry weight. Dally food consumption amounted. to 0.814-.2.31 kca]/g dry weight, with the lowest values for N. fodiens a.nd the highest for

S. minutus. Faecal production was in the range 0.71-2.30 kcal/g dry weight/day, and was lowest in S. ininutus and highest in Nfodiens.

Energy intake by shrews was found to be similar to that of the harvest mouse arid laboratory mouse of equivalent weight described in other studies. It was concluded that the energy requirements of shrews are not excessive compared with other small mammals.

Production, growth & seasonal weight chges Studies on the production and. growth of shrews and the seasonal weight changes of captive shrews revealed the followIng results:-.

1. The mean litter size of wild. s. araneus was seven (maximum nine).

The size of laboratory born litters was 4-9 in S. araneus and 3 in C. suaveolers.

2. The gestation period fr S. araneus was at least 21 days. Post- natal growth or th young was rapid: within 14 days they had reached

5-7g and by 22-23 days they were fully weaned.

3. Growth of young S. araneus outdoors at 8-25°C ias slower than indoors at 20°C.

4. Food consuwption of nursing S. araneus was approximately 1.11 times that of non-nursing females.

5. Retrieval of young displaced front the nest arid re-siting of nests was carried out by female S. araneus.

6. It is suggested that only 3-k young of any litter survive to weaning and that weaning is a very vulnerable stage of a shrew 'a life. -. 163 -

7. Young S. araneus outdoors showed a decrease in body weight in winter despite abundant food,and a spring increase to maturity.

S. araneus kept ic1oors at 20°C showed a spring increase in weight but no winter weight loss. S. ininutus showed no appreciable weight changes either outdoors or indoors, but data were few. N. fod underwent a decrease in weight outdoors but no change indoors. The crocid.urids showed a spring increase in weight indoors but no change outdoors.

8. All shrews kept indoors at 20°C generally exhibited irregular - fluctuations in weight,with a tendency for fat deposition.

9. It was concluded that the winter weight loss of shrews is not caused by food shortage: the exact cause remains urknown but changes

in temperature and day].ength may provide cues for seasonal weight changes.

Water & Fat contents Studies of the water and fat contents of wild and captive S. arafleus, S. minutus and N. fodiens and captive . russuJa and . suaveolens produced the following resu1is:-

1. No sexual differences in water and fat contents of wild shrews were found.

2. Fat contents of wild shrews differed between localities: shrews

from the northern-inost site having the highest fat contents.

3. When species were compared, no diffeencesin water content of wild shrews ware fnund but fat content, as a percentage of d17 body weight,

was highest in S. minutus (mean 6.8) and. lowest in N. fodiens (k.20$). Vithin a species, fat content and body weint were rot related.

k. No difference in fat content between age classes of wild shrews was found. Water contents of sub-adult S. araneus in winter were lower than summer juveniles but there was no difference between juveniles and

adults. - 161f -

5. Mean water contents of S. araneus showed a winter decline and a summer peak, coinciding with body weight changes. Mean monthly fat content was lo,t in suminet and higher in autumn and winter. These seasonal differences were not of sufficient magnitude to account for the loss in body weight in winter or to be of any survival value to shrews.

6. Fat contents of captive shrews kept at 20°C were in the range 23-4 of dry body weight,ami in S. araneus, S. minutus and N. fodiens were 7-10 times greater than in wild shrews.

Act iv it1

A comparison of the activity of captive S. araneus at 20°C (in summer conditions) and at 1-9°C (in winter conditions) by time-lapse photography showed the following results.

1. Shrews were generally active throughout the 2Ll-hour period.

2. Activity was greatest during hours of darkness in both temperature regimes.

3. Activity over the whole 2I-hour period was lower at 1-9°C than at 20°C. The duration of activity in any 30-minute period was also shorter at 1-9°C.

Li.. It is suggested that the activity of shrews is reduced in winter and this could explain the reduced number of captures. - i6 -

CHJtPThR 5

STUDIES ON ¶I1E )RAGING & BVRROWING BEEAVIOUI OP S. ARANEUS - 166 -

SECTION A: ITRODUCTION

The diet of shrews has been shown to be highly varied and comprises not only large numbers of invertebrates active on the ground surface, such as staphylinid beetles, araneids and adult d.ipterans, but also a substantial number which are either strictly soil-dwelling, such as the majority of lumbricids, or lie quiescent beneath the ground surface for part of their life-cycle, such as lepidopteran and tipulid. larvae and pupae. Little is known of the foraging strategy of shrews, in partioular the location of prey beneath the ground surface. That foraging is extremely rapid and efficient is illustrated by the followin field record: one S. araneus liberated from a trap was observed foraging amongst short grass and moss and. within 15 minutes had located and eaten one earthwoin, atiplid larva, a large lepidopteran larva and a slug. Pernetta ( l97, 197Th) found thal prey detection was aided by auditory and. tactile stimuli: shrews would. orientate towards tubes containing active carabid beetles but not to quiescent pupae.

uilliam (1966) examined the histological structure of the eyes of moles and concluded that at best they function only as light detectors, and Pernetta (1977b) suggests that the same is true for shrews, in which case they are of limited use in foraging. Auditory, tactile and visual stimuli are of little use when prey is hidden or lies quiescent below the ground surface. Holling ( 1958 , 1959) suggested that scent is important in prey location by American shrews which were able to detect the odour emanating from sawfly cocoons buried in shallow sand. However, little is known of the ability of shrews to locate hidden prey beneath the ground surface in the wild, or of the relationship between surface arid sub-surface foraging.

Diet studies described In a previous chapter have shown no major seasonal differencs in feeding habits of shrews, a1thou - 167 - large numbers of quiescent insect larvae are eaten in winter.

Faecal pellets were collected only from shrews active on the ground surface, and this, coupled with the rapid throughput rate of food. in the gut, did not permit a definite conclusion as to where in the soil/vegetation profile the majority of prey items were being caught. However, studies on food availability do suggest that although prey is abundant throughout the year, there is a tendency for reduced surface activity by invertebrates in winter when they migrate down into the soil profile. Trapping results also reveal a tendency for shrews to be less active on the ground surface in winter, and tis could be correlated with a change in foraging strategy, with a tendency to increase subterranean foraging in cold. winter conditions.

Likewise, little is known of the burrowing behaviour of shrews in the wild, although they are thought to inhabit subterranean tunnel systems. Shrews liberated from traps in an open area where they can be observed are quick to locate a hole in the ground into which they retreat. Thirther investigation reveals such holes to be natural cracks in the ground or parts of more extensive small mammal tunnel systems. Both Adams (1912) and Crowcroft(].955) found S. araneus in captivity to be active burroweze and Hamilton (1931) reports the ability of Blarina brevicaiLda to construct subterranean tunnels as well as shallow surface runs in leaf litter, but there is no information on the extent of burrow systems created by

S. araneus under natural conditions.

The aim of this study was to investigate the foraging ability of shrews, provided with non-active prey buried beneath the ground. surface, and. to collect incidental information on their burrowing tendencies under semi-natural conditions in outdoor enclosures. - 168 -

Two major investigations were carried out in association with the foraging of shrews, namely, the effects of depth and density of prey on predation levels. - l9 -

SECTION B • NETHODS

PART I: THE E)PERINENTAL PEt'TS

Three enclosures were designed and placed outside in the

grounds of Westfield College. Each enclosure was divided into

three compartments or pens, l.22m by O.8in in size. The design

of the pens is shown in Fig. 5-1. The walls 'ere constructed from sheets of rigid P.V.C. 8mm thick, measuring O.7m high.

The base comprised stainless steel mesh with an aperture size of 5mm by 5mm, through which shrews could not escape. To protect the pens and. their occupants from cats and other intruders, hinged tops

were made of galvanised wire mesh with an aperture size of 20mm by 20mm. The enclosures were sunk into the ground to a depth of 500mm

and the same depth of garden soil was placed inside each pen,on top of the wire mesh. On the surface of the soil, a covering of leaf

litter was provided.

PART II: MAINTENANCE OF SHREWS IN TEE PENS

Shrews were placed individually in the pens. At one end of the pen a nest box with hay was provided, together with an abundant supply of food and water. This 1ivg_ea was left undisturbed. for the duration of the experiments. No other protection against the weather was provided. When not participating in experiments shrews were fed on blowfly larvae or pupae placed. in jars under cover for protection against rain. It was desirable that the shrews should be encouraged to search for food buried in the soil, but not die of starvation in the process, so during foraging experiments additional food was provided in the form of minced beef: this was eaten by shrews but they showed a preference for the experimental food of blowfly pupae. Fig. 5-1: Outdoor Enclosures

Emptr enclosures showing mesh base (and a 15mm ruler at the bottom for the scale).

J..

Enclosures containing soil, and leaf litter, and the two corers. .- ___ ---. -

I - - .; - - -

r _____ - -. r+Il:p. IEJWIIh LI" ll

.t 9•1,

. j i.-. - -.c.

Ii. - 170 -

Each shrew was kept in the same pen for the duration of the

experiments and was allowed at least one week to habituate to Its surroundings before experiments were commenced.

PART III: TEE APPARATUS & EXPERIMENTAL DESIGN

Pt- grid was constructed with a sheet of 'Netlon' which fitted inside each pen, covering the whole ground surface

area. It was divided into 50mm by 50mm squares, each equare with its own co-ordinates. Small marker sticks were placed through the mesh of the grid into the soil beneath at required co-ordinates, after which the grid, was removed, leaving the markers to indicate the points at which food was to be buried or recovered from. The co-ordinates governing the distribution of food points were determined from a set of random numbers, a different set of numbers being used for each experiment.

Once the food points had been determined and marked out, prey items in the form of blowfly pupae which had been kept cold to prevent emergence of adults, were introduced into the soil at certain depths at each point. A metal corer of 25mm diameter was used to remove a plug of soil to the required depth and blowfly pupae dropped into the hole. The hole was filled with soil and. th marker stick removed. When all the food points had been set up the whole ground

surface of the pen was gently compressed, rakd. over and the leaf

litter redistributed to prevent identification of the disturbed

areas by the shrews.

The pens were left undisturbed for two days to encourage

tho shrews to forage, after which the food. points were again - .l.(.I. -

identified using the grid and marker sticks. A plug of soil was

removed from each food point using a corer with a diameter of 75mm

to allow for any movement of the pupae by earthworms, shrews or

other means. The soil plug was sorted in an enamel dish and the blowfly pupae searched for. The search was facilitated by their

reddish-brown colour which was quite distinct amongst the grey-

brown soil. The numbers of pupae recovered in this way were

recorded. All results were expressed in terms of percentage predation of pupae which referred to the percentage of the total number of buried, pupae which had been taken by shrews. i) Control exoeriments were carried out before shrews were placed in the pens to test the efficiency of the experimental design and.

to insure that the failure to locate pupae at a food point was

indeed indicative of predation by shrews, and was not due to

predation by othec animals such as carnivorous beetles, or to an

inadequate recove:ry process. To discover if any pupae hatched into

adult flies during the course of the two-day experiments, some were

observed in jars in the laboratory but were never found to emerge

during the experimental period.

ii) Predation a different soil depths: pupae were placed singly

or in groups at a certain number of food points at different soil depths to investigate the ability of shrews to locate hidden, non-active prey. Groups of five pupae were used, despite the rarity of stch an ocoirrence in the wild, for a number of reasons. It was feared that uneaten pupae buried singly would be difficult to locate at depths where lar8e soil cores had to be taken during the recovery process at the en of each experiment, and. therefore lead to error.

Moreover, not all pupae are equally palatable to shrews and burial of a single moribund pupae would. not encourage predation: this was overcome by using groups of pupae. Following burial of the prey, it was nec'ssary to compact the soil to simulate natural conditions - 172 - which, a1thoui performed carefully, would crush the prey items beneath, so groups of pupae were buried to increase their chances of survival. Subsequent experiments showed that recovery and. survival of single pupae was possible and further trials were conducted to investigate the recovery of single prey items by shrews. Only ten food points were provided to insure that their proriunty to each other was not the cause for their detection.

Pupae were buried at depths of between 20 and 160mm. No trials

were conducted. with pupae placed on the surface of the soil because it was feared that displacement of the prey items might be caused by

wind, rain and movement of the leaf litter in addition to predation

by shrews, and produce inaccurate results.

iii) Predation at different prey densities: pupae were placed singly at a constant depth in the soil but at different numbers of food

poiz, ts. A depth of 50mm was chosen because all ehrews were able to forage to this depth, and it insured that pupae not recovered. at the end. of the experimental period had. been eaten and not simply moved or mislaid. as a result of the normal exploratory movements

made by the shrews.

To reduce the opportunity for shrews to learn from the experiments where to locate buried food, no trials were run at the same depth or density on successive occasions, and trials were

interspersed with several days 'rest' when no pupae were buried.

iv) Varnished pupae:

Trials were also conducted to investigate the effects of odour of prey on the detection by shrews. Pupae were given two

coats of varnish which were allowed to dry thoronghJ.y in an

attempt to prevent the odour escaping from them. An equal number

of treated and untreated pupae were buried singly at 50mm depth

at 10 food points. - 173 - v) Burrowing activity: In addition to the foraging experiments, infoi'mation was collected on the burrowing activity of shrews i-n the pens.

Al]. experiments were carried out using S. ararieus which were placed in the pens shortly after their capture from the wild. - 174

SECTION C: RESULTS PART I: CONTROLS

Control experiments with pupae buried in groups of five and singly at 10 points between depths of 20 and. 120mm showed 1O(% recovery after 2 days in all cases with the exception of one trial where recovery of groups of pupae at 50mm depth was 96%. Control pupae which were selected with the experimenta' pupae but were kept in jars in the laboratory were never found to emerge into adult flies within the two-day trial period, although they did emerge after approximately five days.

PART II: PREDATION OF BURIED PUPAE BY SEREMS

The foraging activity of a shrew in its pen was often evident as diggings in the soil. Such diggings were not necessarily sited directly above a hidd.9n food. point but were found scattered, apparently at random, over the soil surface. They usually comprised email 'scuffle' diggings of 30 - 100mm diameter and up to 100mm deep. 1) Predation at different soil depths.

Shrews were able to locate pupae, in groups of five, between the soil surface and. a depth of 160mm. The mean percentage predation at different soil depths, togather with the standard. error is shown in Fig. 5-2k. A total of 43 trials were conducted with a minimum of 4 trials, using differert shrews, at each depth.

Percentage recovery declined with increasing depth: predation at 20mm depth reached almost l0C, fell Lo less than 50% at 80-120mm and was no more than ø'/ at 160mm. Discovery of a food point where five pupae were buried resulted in all the prey at that point being recovered and eaten in practically all cases. A No.triols i2 4 9 6 groups of pupae (- igle pupa. (b-

C 0

0 Depth of Prey (mm) -o

I- a- ______B 100 9Mo- 0 8M 0\ 2 M- '_ -

6 - - 40 - - -..'

0 20 40 60 80 100 120 140 Depth of Prey (mm)

Fig. 5-2 Predation by captive S araneus of pupae burled at different depths. A mean predation of pi.pae in groups of five and singly (with standard errors) and B predation of pupae in groups of five by individual shrews. I

- ij') -

The recovery of single pupae buried at different depths

showed the same trend, as can also be seen in Fig. 5-2A. A

minimum of 3 trials at each depth were conducted. At 20mm depth, percentage predation approached 10C but by 120mm it had declined

to under 3CP/o. No pupae were recovered at 160mm in any trial. The overall recovery of single pupae tended to be lower than that of

groups of pupae.

The results are complicated by the great individual variation which remained high regarclJ.ess of the number of trials, and which makes comparisons on a statistical basis impracticable. Not only was there great inter-individual variation ;n overall foraging ability, but also intra-trial variation where tha performance of individuals in the same trial repeated on several occasions &Lffered. The examples of greatest variation shown by individuals in the same

trial conducted on different ocoasons is shown in Table 5-1.

Shrew Number .2ecovery of pupae Depth tnr pn diIferent occasions

20 64, 88, 98

80 60, 10

80 414 38, 90

120 iN 16, 56 Table 5-1

The pe:rcentage predation of pupae at different soil depths by individual shrews is shown in Fig. 5-2B. De'pite the variation between individuals a trend is still appaient o'' high leve]a of predation of pupae hidden near the ground surfe, which declined as depth increased. Only three shrews of the six tested found

pupae at l60mm, and of theae only one, SaI, rhich is depicted in

the graph, recovered as many as 30% of the pupae which entailed

location of only three of the -ten food points. - •1' ii) Predation at different prey densities.

The mean percentage predation of single pupae buried at

50mm in the soil in different densities is shown in Fig. 5-.3A.

A total of 48 trials were conducted with each density being tested

at least three times. A density of one pupa per pen was tested using 15 different shrews. A trend of increasing predation with

increasing density of prey was evident, despite the g:reat amount

of individual variation. 1 hen density reached 20 pupae per pen

recovery approached 10 /o and individual variation was low.

Between densities of one and 10 inclusive the mean recovery by

shrews was only 50-6% and individual variation was high. Examples

of variation between results from the same individual but on

different occasions are shown in Table 5-2.

% Recovery on Density Shrew Number different occasions

1 AM 0,100

10 iN 20, 100, 100 10 N 0,10,20

10 2M 40, 90, 100 Table 5-2

The percentage predation of pupae at different densities by

individual shrews is shown in Fig. 5-3:B. Duplication of trials with a density of one pupa per pen revealed a high degree of

variation in the results, with shrews recovering the single prey

item on one occasion, but not on another. A mean of 5CY'/

predation is plotted for each individual in Fig. 5-3B, which

represents the mean result of the two trials performed with each

shrew. A mean of 66.7% was obtained from all the trials performed

with pupae buried at this density, and this is plotted in Fig. 5-3A. A

15 3 12 6 6 6 No trios, q#% %

C 0 4.. ) 0 0 Prey Density

I.. - B

\0 0\

Prey Density

Fig. 5-3 Predation by captive S. araneus of pupae buried singly at 50 mm depth in different densities. A mean predation, with standard errors,and B predation by individual shrews.

.. 177 -

iii) The effects of weather, season & proxm1ty of food. points on predation levels.

No correlation between percentage predation of pupae and

weather conditions was found. Most trials were conducted during

the summer of 1977 when warm, dry conditions predominated. There was no evidence of a change in foraging ability, indicated, by the

percentage predation of pupae, as autumn approached arid conditions

became colder and wetter. No trials were conducted in conditions

of frost or snow. AU the individuals used were of approximately

the same age, that is they were juveniles or sub-adults. Neither

was there any effect on predation caused by the occasional close

proximity of the food points, particularly in the experiments performed at d.ifferent depths. Food points in close proximity

were not located. more frequently than those more evenly distributed,

iv) Varnished pupaP. Laboratory experiments showed that there was no difference in

palatability of varnished and unvarnished pupae, for shrews would.

eat both vrhen placed together in a jar in their enclosures. Three trials using different shrews were conducted with varnished and. unvarnished prey bur,ed singly at 50mm depth. The results are shown

in Tab)e 5-3, from which it is evident that there was no sigiifican difference in predation between treated and. untreated pupae. In

one trial, all pupae, both treated and untreated, were taken within

the two day trial period. In trials where predation did not reach

lOOYo, treated. and untreated. were taken in approximately equal proportions. Shrew Number AN A BM

Varnished. pupae 100% 60% 80%

Unvarnished. pupae 100% 80%

Table 5-3: Percentage pr-dation of varnished & unvarnished pupae by shrews. - 17&i - v) Burrowing activity

Shrews exhibited very little burrowing activity apart from small 'scuffle' diggings and. holes made during and after the foraging experiments. After three weeks spent in the pens three shrews showed no more burrowing activity than surface runs made amongst the leaf litter and. between one and. three small burrows 30mm deep at the side of the pen and beneath the nest-box. When the foraging experiments were terminated, six shrews were maintained in the pens for between two and four months after which the enclosures, hitherto undisturbed, were tnorouily searched for signs of burrowing. Even after four months burrowing was minimal and comprised small holes near the surface, surface runs amongst the leaf litter and between one and three small U-shaped burrows in each pen. The U-shaped burrows, usually constructed near the edges and corners of the pen or beneath the nest-box and. water bowl, were

30-120mm deep and 90-200mm long. They were smooth with use but rarely contained nests. Shrews, if disturbed from the regular nesting site,often retreated into these burrows wheic they were completely hidden. Shrews always nested in the box provded, and if hay bedding was not included, a nest would be cons &ructed ftom dried leaves augnented. by small stones and lun'ps of soil carried into the base of the box. Occasionally, a shrew' would change the nesting site to one of the U-shaped burrows, with an entrance hole et each end and a nest of leaves in the centre, but such burruws never exceeded 120mm in depth.

In addition to the U-shaped burrows, three shrews each constructed a single surface burrow situated away from the nest-site and consisting of little more than a shallow impression in the soil, less than 30mm deep and 120mm wide and. covered over with leaf litter. - 179 -

In these holes a large hoard of blowfly pupae was found,

accumulated after the foraging experiments had. ended.

Burrowing activity was also observed in captive shrews in glass—sided aquaria where pupae were buried in the soil. The

snout was used to probe, small stones and. lumps of earth were removed using the teeth, and the fore—feet aided digging. In this way shallow holes were made and pupae recovered to a depth of 80mm, but no tunnels or burrows were constructed into which the shrew could retreat, even when no alternative cover was provided for a nesting site. - 180 -

SECTION D: DISCUSSION

The results of this study indicate a remarkable ability of

shrews to locate hidden, inactive prey. Whether such intense digging activity as that revealed by these experiments is found

in the wild is not known. Capti.ve animals in a confined space

may well use up surplus energy by repeated exploratory digging over the same area.

The recovery of pupae at considerable depths by shrews may ha1e been permitted by the degree of compaction of the soil around

the food points which, out of necessity, was not as great as would be expected under natural conditions, Digging to any depth in the wild is severely restricted by the compaction of the soil, particularly when dry, and the amount of vegetation forming a dense mat on the surface and extending with root systems into the soil below. Moreover, most invertebrate prey types, including coleopterans, isopods, araneids and even insect larvae, inhabit the

surface 80-l2Oiriof soil and little besides earthworms are found as

deep as 160mm, and. so it is unlikely that shrevrs would need. to dig

to any great depth for food even if they could detect hidden prey in the soil below the ground surface. Therefore, the results of

this study are not necessarily indicative of the digging inclinations of shrews in the wild, but do reveal something of their ability to locate hidden prey without the usc of tactile, vLsual or auditory stimuli.

The precise method by which hidden prey is detected is still unknown, but there are indications of olfactory cues, coupled with random searching. Holling (1958, 1959) reports that small mammals, including the shrews Sorex cinereue_pinereus and Blarina brevic-tuda dpotd.e, could detect cocoons of the Eu.ropedn pine - 181 -

Sawfly by the odour emanating from them, and the strength of the odour perceived could be changed in experiments by varying the depth of sand covering them. As a result, he found a decrease in predation with increasing depth of the buried cocoons. This was also found in the present study, both where blowfly pupae were buried in groups and singly. The greater concentration of odour emanating from groups of pupae could explain the better perfozmance of shrews in recovering pupae buried in groups rather than singly deep in the soil. This is further substantiated by the work of

Jennings (1976), who found that the detection of tree seeds and peas by Apodemus sylvaticus was positively c'orrelated with the weight of the seeds, and that those coated with vegetable oi'. including linoleic acid, not only increased their palatability, but also their detection. Rolling (1955, 1958 , 1959), also found that shrews could dctect the difference between parasitised arid imparasitised sawfly pupae, eating only the latter, and suggested this was due to scent. However, laboratory experiznents with shrews kept in glass sided terraria containing so1 in the present study revealed that the shrews underwent intense digging activity but were unable to locate all pupae buried singly or in pairs until several hours had. elapsed, even when the prey items were buried at a depth of only 20mm. It seems doubtful, then, that detection of hidden prey by shrews occurs totally by olfactory means, and. this is substantiated by the results of Taylor (1977) in studies of Southern grasshopper mice (On ychomys torridus). He found that these mice captured more mealworms when they were buried in a dispersed pattern than when they were aggregated, and. he attributed this to the greater ease of finding the first prey and. a more efficient style of search for dispersed prey. In the present study, however, predation of clumps of pupae was higher than that of single pupae, particularly between depths of 50mm and. 160mm. - 182 -

Without further evidence of the olfactory bility of shrews,

the results of the present study, together with those obtained by

Holling, in the investigations of predation of food. items buried at

different depths, could be attributed to a random searching

technique employed by shrews. Shrews would be unlikely to locate pupae hidden deep in the soil, simply because a greater area has to be searched, and groups of pupae might be easier to find by virtue of occupying a marginally greater area than single pupae. That olfactory stimuli are not the only means by which prey is located is further Indicated by the lack of correlation between location of the pupae and. the proximity of the food points, and the equal predation of pupae treated with varnish and those left untreated,

Even if the varnish failed to prevent the odour emanating from the pupae completely it might at least have been expected to reduce it or replace it by another unfamiliar scent.

The increasing levels of predation of pupae with the increase in density also seems to indicate a random search, for the greater proximity of pupae and the larger the area occupied by them would. increase their chances of location. If detection occurs purely by scent it might be expected that at a constant depth, to which shrews had proved capable of foraging, all hidden pupae would be located.

This might occur regardless of density, provided the number of prey did not exceed. the daily energy requirements of the shrew. At higher densities where supply exceeds immediate demand, predation may continue to increase as surplus food i.e hoarded: captive shrews were often observed to establish food caches.

Increasing predation with increasing density of prey was also observed by Holling (1959) who found that numbers of sawfly cocoonu opened by shrews inthe wild, increased. until a maximum was reached - 183 - which corresponded with the maximum number consumed daily by captive shrews. This was also reflected by tne stomach contents of Peromyscus maruculatus bairdi: when cocoon densities were high their occurrence in the diet was correspondingly high and. vice- versa. In the present study predation reached approximately l0C)°A at the higher densities of buried blowfly pupae, but the numbers provided (a maximum of 50 over two days) were not in excess of the daily food requirements which are at least 50 pupae per day, arid. consequently no levelling off of predation was observed when supply exceeded demand. ) such as was found by Holling (1959).

The interpretation of the results in the present study was complicated. by the high degree of individual variation, an observation which was also noted by Crowcroft (1955) when investigating burrowing behaviour of shrews. This may be attributed to several factors: (i) the difference in ability of shrews to locate buried pupae, (2) the difference in inclination to search for food when an alternative food was provided, which may have been more palatable to some shrews than to others and. (3) the random nature of the searching which proved successful only in some individuals on certain occasions. However, most shrews exhibited intense searching activity in the outdoor pens which was evident from the my-riad of small diggings and constant distu.rbance of the leaf litter. Such intense searching behaviour was also remarked upon by Southern (l954 and. Crowcroft (1959) in foraging experiments on house-suce. Orowcroft found that offered a choice of feeding stations, a house mouse never took its entire day's supply from any one of them, bat made use of them all. Pernetta (197Th) found. that shrews in laboratory experiments tended. to re-investigate the same area once a prey item had. been caught there, and. this is confirmed. by the present study, where groups of pupae were buried. Where a - ].81I. - food point was located all the pupae were removed by the shrew.

No hoarded pupae were found when the pens and. nest-boxes were searched at the end of each experiment, and. so it was assumed that all the pupae located were eaten, and as supply did not exceed daily demand, as mentioned above, this seems a reasonable a8sumption.

Despite such intense activity shown by shraws in foraging

experiments, they showed little inclination or ability to create burrow systems. Hamilton (1940) remarked. that the weak feet of

Sorex fumeu made digging difficult and that their efforts to do so are weak and. ineffective, and. this could equally be said. of

S,, araneus judging by observations made in the present study. Even after four months spent undisturbed in outdoor pens shrews had merely constructed. small U-shaped. burrows within 90-120mm of the ground surface. This contrasts with the burrowing activity of Apodemus sylvaticus: an individual maintained in a pen for a month constructed a tunnel 0.24m deep and 0.6Cm long.

Crowcroft (1955) observed burrowing behaviour of S. araneus in captivity arid found. digging activity to be intense and. gained. the impression that autumn juveniles engaged. in more burrowing activity than summer adults, which was not observed in shrews kept in outdoor pens in the present study. He says nothing, however, of the length or depth of burrows or whether they were used. for nesting, but they cotdd not have exceeded. the 4 inches of soil, he provided. for them, which compares with the depth of those observed in this study, where up to 300xnnd.epth was available. No observations were made on the burrowing activity of S. minutus in this study, but Crowcroft ( 1955) remarks that this species showed no inclination to burrow but ,ould use tunnels of S. araneus. Noomy-s fodiens inhabits 185 - tunnel systems in the wild, but observations of one individual in an outdoor pen showed burrowing activity to be no greater than that of S. arancus, although Croweroft (1955) found it displayed more digging activity than S. ararieus, but again makes no mention of the extent of the burrows.

All the British shrews are known to use subterranean burrows, and. observations of wild, shrews liberated. from traps confirm their readiness to retreat vertically into small holes and. crevices. This, coupled with their poor ability to create extensive burrows, indicates a use of tunnel systems constructed. by more adept burrowers such as Apod.emus sylvaticus. The extent to which the substratum is permeated by burrow systems of small mammals is unknown but if old, possibly abandonned burrows are included with those in regular use, their number aid. availability is probably grossly underestimated.

To conclude, foraging experiments reveal a rcinarkable ability of shrews to locate quiescent prey hidden at considerable depths in the soil, without the use of visual, tactile or even apparently auditory stimuli. It is suggested that location of non-active prey is aided by olfactory- cues, particularly over short distances, but of equal importance is the extremely thorough but random searching undertaken, as indicated by (i) the shrews' inability to locate all food points at low densities even at compacatively shallow depths, arid (2) their failure to always locate food. points in close prorimity to each other. More work is required before the precise means of prey location can be defined.. Their adeptness at digging associated. with foraging does not compare with their poor ability to construct burrow systems and it is suggested that tunnels of other small mammals are adopted by shrews, and. possibly modified.

These may then be used for nes tang and. foraging in the manner of - 186 - moles, particularly in winter when adverse weather conditions are prevalent. - 187 -

SECTION E: SUMMARY

An investigation of the foraging arid, burrowing behaviour of

S, araneus in outdoor enclosures provided with soil and. leaf litter revealed the following results:-

1. Predation of blowfly pupae buried singly and in groups declined with depth from 1O(/ at 20mm to 30-50% at 120mm and O- at 160mm. Predation of pupae in groups was higher than single pupae. 2. Predation of single pupae at 50mm depth tended to increase with increasing prey density.

3. Disguising the odour of pupae with varnish cUd not reduce predation. It is suggested that foraging is carried out by thorough, but random, searching, assisted by olfactory cues.

4. Great variation lit foraging ability of shrews was found, both between individuals and. within individuals on different

occasions.

5. Burrowing activity apart from foraging was minimal, and. consisted only of $mall U-shaped burrows 30-120mm deep in which nests were occasionally made. Food caches were made beneath leaf litter. - 188 -

CHAP'rPR 6: FINAL DISCUSSION - 189 -

The results of the investigations reported in this thesis,

and the issues which have arisen front then, havA already been

discussed in some detail in the relevant chapters and so will not

be dwelt on further. This final chapter serves to emphasise and

discuss some of the more salient findings of this study, their

relationship with each other and their possible implications, with

particular reference to the overwiritering of shrews.

The results of this study confirm that a marked seasonal cycle

in captures of shrews urs, with high numbers in summer, a rapid

decline in autumn, low numbers in winter, and an increase in spring.

But, a study of the seasonal population dynamics of shrews revealed

that this apparent cycle is not of such great magnitude as had

previously been supposed, and that the causes of the apparent

fluctuations are rather different from previous assumptions.

In the summer laige numbers of juveniles are born into the

population and cause a rapid increase In density arid nuwbers of

captures. In the autumn, however, a sudden and very rapid decline

In the numbers of captures Is apparent. Closer study of the

population dynamics using capture-mark-recapture techniques

revealed that several factors contributed to the so-called

'autumnal epidemic'. Following breeding, old adults disappeared

from the population this occurred througi'out the suntmer, bet.ieen

July and October, but it is probable that most survived to breed

at least twice and were still alive in late Atigust and epteznber.

Senescence may have contributed to the mortality of t!ese adults, but It is unlikely that their death was hastened by climatic factors,

since their disappearance from the population occurred prior to the

advent of harsh winter conditions. It may be competition for food

and nesting sites, and agonistic behaviour which accelerates the - 190 -

death of these shrews during the suruner and early autumn.

Death of old adults is not the only cau'e of the 'autumnal epidemic', for the steep decline in numbers of captures cannot be accounted for by mortality of adults alone. Mark-recapture work revealed that large numbers of juveniles also disappeared from the

population and construction of a survivorship curve md icated that

about 50% of juveniles died within the first tvJo months of life.

What causes this high mortality rate is unknown: it seems unlikely

that environmental factors contribute much to the decline in numbers of juveniles since no shortage in food supply was found and, again,

the population decline commenced prior to the advent of winter conditions.

Likewise, parasitism was not found to be a significant factor in the mortality of shrews. Shrews are subject to high levels of

infestation by a variety of enioparasites, which reached a peak in adults in summer, but generally decreased in autumn and winter. It is possible that high levels of infestation contribute to the death of old

adults in the ummer and autumn following breeding but not of juveniles

or sub-adults, since these age classes generally had the lowest levels

of Infestation.

Several features of the behaviour of juveni) e shrews indicate that sociologici factors and predation have important Influences on mortality. Young wild shrews were vexy active on the ground surface and it has already been suggested that curing summer and autumn they are occupied in establishing themselves in hone ranges and nesting sites. Consequently, juveniles probably spend a considerable time on the ground surface during both day and night, which makes them particularly vulnerable to predators, particularly in view of their inexperience and the dying off of vegetation in the autumn which 1 - - reduces cover. There was no attempt in this study to assess predation levels in the study areas but the results of other studies do indicate that shrews fall prey to a number of carnivorous mammals and birds. Being less agile than woodmice, and possibly less timorous than bank voles, shrews are probably easy catches for such predators. Southern (1954b) showed, that although the major prey of tawny owls was rodents, shrews were taken regularly throughout the year and there was a tendency for them to be taken more frequently in summer and autumn than at other times of year. Shrews form a comparatively small fraction of the diet of weasels and stoats, but are nevertheless still eaten (Moors l975 Tapper, 1976). Other predators of shrews include foxes (Forbes & Lance, 1976; Richards, 1977),barn owls (Corbet & Southern,

1977) and possibly kectrels.

An important featrre of shrew behaviour is their solitary and aggressive nature, and this could also have a significant effect on mortality. When population den3ity is high, as in summer and early autumn, shrews may be under considerable stress. Laboratory studies of the growth and development of shrews suggest that the juvenile stage is a very vulnerable stage of a shrew '5 life, particularly the weaning period: juveniles tended to decrease in weight during weaning and this may be iccentuated in the wild by the problems of foraging for food. Competition between shrews is probably great at this time and could give rise to stress: not only is there competition for resources such as food and nesting sites, but there may also be the stress arising from numerous contacts with other shrews in the process, both directly through pnysical contact and indirectly through vocal communication. Laboratory investigations in the present study showed that agonistic behaviour commences at a very early age, but whether such behaviour ever results in death remains unknown: it is unlikely - 192 -

that shrews kill each other in the wild, although they do fight fiercely even when unconfined in the wild (personal observation).

Bearing in mind the narrow line between life and death for a shrew,

however, any harrasement which prevents feeding for any length of

time could be critical.

One of the most significant finds of this study is the

relatively high survival rate of shrews over winter. It had been

assumed that the low numbers of captures In winter indicated a high

mortality at this time of year. It has already been emphasised

that mortality of young shrews is high but occurs in late summer

and early autumn prior to the onset of winter weather conditions.

Capture-Mark-Recapture studies revealed that once shrews entered

wintcr survival was high thereafter, and the number of shrews

overwintering was considerably greater than was indicated by the

number of captures. This introduces the question of how shrews

overwinter in adverse conditions, and what causes them to be

untrappable af this time of year. Their survival seems all the more

remarkable bearing in mind that they seeni to make no provision for winter in terms of Inbernation or fat storage. Despite the views of

Myrcha (1969) the present study shows conclusively that fat depositic'n in wild shrews does not occur to any significant extent.

This is even more interesting in the knowledge that captive shrews readily lay down fat reserves which suggests that conditions for wild shrews are never condusive to fat deposition, and that there is some feature of their ecology or physiology which does not permit it.

That food sipply is the limiting factor is ruled out by the failure of captive shrews outdoors to undergo fat deposition even with abundant food. Experiments suggest that temperature is an important factor, but even this seems unlikely for laboratorykept shrews at 20°C laid down fat whereas wild shrews in summer did not. It may be that fat - )93 -

depoltion by these solitary, captive shrews in the labordtory Is

a feature of the constancy of the environment coupled with low

levels of activity al lack of 2trcss resulting front tne absence

of contacts with other shrews and predators. More work is required

to study the factors affecting fat deposition by shrews.

Of great importance to the survival of shrews over winter is food supply. It has been shown that food resources do not undergo a shortage in winter which indicates that this may not be such a

rigorous time for shrews as had previously been supposed. Foraging studies have shown that they have a remarkable ability to locate prey, but it is still not known how foraging is affected by harsh weather conditions, jarticularly frost. However, such conditions are also accompanied by a downward migration of invertebrates away from frost and it was apparent in the present study that surface activity of prey was reduced in winter when large numbers of' invertebrates, particularly beetles, were found below the ground surface. It is possible, therefore, that shrews find abundant prey underground during winter and need: not forage much on thi? ground surface. Studies on the activity of wild and captive shrews showed a tendency for reduced activity in winter which could be associated with the vertical distribution of prey. Not only does this explain the decline in number5 of captures in winter, but it could also be of great sigflitiCaflCG in the overwintering success of shrews. Meteorological data show that conditions underground, even by a few centimetres, are very different from those on the ground surface and can remain free from frost. By reducing activity oh the ground surface and adopting a predominantly subterranean existence, heat 1353 could be kept to a mininum.

A reauction in activity may permit a reduction in food requirements

jnce energy s not wasted in foraging and searcl'ing for mates durng - 191+ - winter months, and laboratory studies did &ow that food consumption of S. araneus at least, were significantly reduced in cold winter conditions. Associated with activity and feeding is food caching which many captive shrews indulged in. Not all shrews exhibited this habit but during the activity studies at low temperatures many shrews showed intense activity in food storage, and observations revealed epeated. visits to food dishes until all the food had been removed to the nest. Subsequently, brief visIts were paid to the food dish but prolonged activity outside the nest was not observed, and neither was feeding. Food caching commenced in shrews immediately following their capture and so it seems likely that it does occur in the wild, as in the manner of moles, but more work is required to establish the truth of this. A strategy of food caching would reduce the need for continuous foraging, and even if only a day's supply of food was laid up at a time, it would reduce heat ]oss and energy wastage.

Perhaps the greatest mystery pertaining to the overwintering of shrews is the loss in body weight and how this ts manifested. The present study has indicated that changes in water content and other documented. morphological and anatomical changes are not of sufficient magnitude to account for this weight loss, despite the views of Myrcha

(1969). It is possible that the major part of the weight loss is caused by a reduct.on in overall protein content which results in a decrease in musculature in a similar way to the seasonal weight changes in starlings found by Ward (pers. comn.), This presents another aspect of shrew biology still to be researched.

A further interesting feature of the winter weight loss of shrews is its cause. The loss in weiht of captive shrews outdoors in winter, despite a plentiful food supply,indicates that food availability is not instrumental in the process of weight loss. Related to this is the - 195 -

failure of shrews to increase food consumption in cold conditions,

and S. araneus exhibited a decrease in consumption, as mentioned

above. This suggests that temperature is an important factor but how it affects food consumption and weight changes of shrews remains unknown. It is possible that low temperatures inhibit feeding by producing a form of hypothermia which slows down the activity of the animal, thus making it physically incapable of either catching

or eating food frequently enough to maintain its body weight. This

suggests a situation similar to the 'digestive bottleneck' described by Kenward & Sibly (1977) for woodpigeons on brassica and clover diets: they suggested that pigeons foraging on shrubby brassicas can fill the crop very quickly but feeding is limited by the greater effort required to digest this food, whereas clover is easily digested and assimilated but requires longer periods of foraging to fill the crop. There is no evidence of a reduction in metabolic rate or body temperature which could. cause a decrease in assimilation rate and a

'digestive bottleneck' in shrews, but a reduction in activity produced by low temperatures could result in a 'feeding bottleneck'. However, the sudden weight increase of shrews in spring, even when temperatures are still quite low,suggests that this may nct be so.

Winter represents a transition period, between the birth of a shvei, and the first few months spent in establishIng itself, and the time when it has to breed. Provided it does survive to breed, the method of overwintering is not important and it can afford. to reduce

it activity and rest up until the spring. Having such a short life, it is imperative that, as soon as conditions become favourable (that

Is, in spring and early summer) each shrew is ready to breed so that as many young as possible can be produced in the brief time spent as an adult. Shrews mature very quickly in spring but in order to be ready to breed they must respond to a cue or cues, often some time - 196 - before an appreciable rise in ambient temperature occurs. Thus, the rapid weight Increase of shrews in spring, when reprcductive condition is being attained, probably occurs regardless of ambient temperature, since unlike the situation in autumn and early winter, weight change was not closely correlated with changes in ambient temperature. weight changes in autumn may, therefore, be correlated with temperature and/or daylength changes which affect the hormones governing 9rowth & behaviour (including feeding behaviour) but in spring may result from an increase in the pnotoperiod or some other cue affecting hormone levels. Changes in hormone levels probably stimulate not only an increase in weight in spring, but also an increase in activity on the ground surface, which Is of great importance in establishing home ranges (and possibly territories) and locating mates. Much more work is required to Investigate the physiological aspects of overwintering by shrews.

It has already been stated that no seasonal differences in body temperature have been recorded which could cause a reduction in metabolic rate and activity sufficient to produce a state of torpor in shrews, but the methods of obtaining these results should be considered before a final conclusion is made. Measurement of temperatures of such a small mammal presents great practical diffLcu1tIs and .j.t seems unlikely that it can be done without causing great disturbance to the animal. A rise in heart rate through stress could easily result in an Increase iii temperature and overall nietd.bolio rate, giving misleading reu1ts. Can such measurements, taken from an active, excttable beast, possibly reflect the situation in the wild? It is suggested that more information about the seasonal body temperatures and metabolic rates of shrews is required in order to elaborate on their overwintering strategy. - 197 -

Considerable doubt has already been cast on the survival value to shrews of a reduction in weight in winter, since food consumption and body weight c.re not directly related. It is possible that within an individual, with a constant metabolic rate and level of activity, overall food requirements are reduced by a decrease in body weight but this cannot be assumed • It is suggested that the weight loss of shrews is not a strategy for overwintering but that it is the result of changes in temperature and related environmental conditions which affect their physiology and behaviour at a time when activity can be reduced to a minimum.

It had been the intention of this study to formulate an energy budget for a population of shrews, but this has not been possible since some of the parameters of the budget remain unknown. Estimations of oxygen consumption, for example, were not made in the present study for it was hoped to abstract an estimate from the wealth of results currently in the literature on this subject. This did not prove practicable for several reasons: measuring the oxygen consumption at rest of such a highly and contnuowly active animal presents similar problems to those already mentioned for temperature measurements, and because of this there is considerable variation between studies. Moreover, authors rarely elaborate upon the methods and, more especially, the conditions employec3 during such estimations, which makes comparisons betw3en studies difficult and abstraction of data unreliable. This accontcafes the difficulty of using information available in the literature as a base for commencing new studies, and the problem facing short-term ecological research projects today. For example, some information on the diet and. food consumption of shrews is already available in the literature but uiicertainty remains as to its validity when applying it to a new study, since the characteristics of environments, ponulations arid even indivdual differ so much. CoIsc..,ItL"ktc,. & cii.dr*i,L o tw..z- YCp&ttLO IS - 198 - necessary before new answers to old. problems can be produced.

To conclude, shrews are hardy beasts and food supply, parasitism and climatic factors do not contribute significantly to their mortality in autumn and winter when a decline in population density is found. Overwintering survival of shrews is high and it is suggested that this success is due to a strategy of reduced activity and food consumption which may be assisted by a lowering of body temperature and metabolic rate inducing a state of semi-torpor for short periods. It is suggested that the decrease in body weight of shrews in winter is correlated with temperature changes which affect foraging activity and food consumption but that it has no particular survival value in itself. More work is required to elucidate the overwintering strategy of shrews. In particular, we need to know much more about the subterranean activities of shrews in the wild which necessitates the use of tracking devices. Linn &

Shilhito (1960) produced a ring to fit around the ankle of a shrew which could be rendered radioactive so that movements of the animal could be traced with a Geiger-Muller counter, but the Bize of the ring and the ainomt of radioactivity contained in it were so small, in order to avoid harm to tne shrews, that tracking was not easy and could only operate over small distances. Radio-tracking may provide a useful and harmless method of study in the near, transmitters small enough tor a shrew to carey are available already and attempts are being made to reduce the battery to a suitable size for use with very small mammals. Further inl'orznation is also required on the ecology and behaviour of S. minutus, the British crocidurids and. N. fodiens, particularly the latter, about which very little is known. - 199 -

BEFERENC - 200 -.

Adams, L.E. (1910). A hypothesis as to the cauco of the uturnna1 epidemic of the common and lesser shrew with some

notes on their habits. Manchester Memoirs 10: 1-11

Adams, L.E. (1912). The duration of life of the common and lesser shrew with some notes on their habits. Mem. Proc. MaTlchr. Lit. Phil, Soc. 56: lilO

Mains, L.E. (1913). The life-history of British shrews. Wild. Life, Lond. 1: 82-87

AndrzejewS4,R., Fejgin, H. & Liro, A. (1971). Trappability of trap-

prone and trap-shy bank voles. Acta theriol. i6 0l-4l2

Andrzej.e"s.,,R., Petrusewicz, K. & Waskiewicz-Gliwicz, 3. (1967). The trappabi)ity of Clethrionomys glareolus (Schreber, 1780) and other ecological parameters obtained by the CMR

capture method. Ekologia Poiska (A) 15: 709-725

Ansell, WFSH. (1961f). Captivity behaviour c*nttpost-natal development of the shrew Crocidura bicolor. Proc. zool. Soc. Lond. L2: 123-127

Aulak, W. (1967). Estimation of small mammal density in three forest

biotopes. Ekologia Poleka (A) 15, 39: 755-778

Aulak,W. (1970). Small mamma]. communities of the Bialowieza Nadona1

Park. Acta theriol. 15, 29: k65-i

Brambefl, F.W.R. (1935). Reproduction in the common shrew (Sorex araneus L.)

Phil. Trans. Roy. Soc. Lond. B. 225: 1-62

Brainbell, F.WR. & Hall, K. (1937) . Reproduction of the lesser shrew

(Sorex minutus L.) Proc. zool. Soc. Lond. 1936-37: 957-969 - 20]. -

Brown, L.E. (195L1). Small mammal populations at Si]woocl Park Field Centre, Berkshire, England. J. Mariml. 35: 161-176

Brown, L.E. (1956). Field experirients on the activity of the small mammals Apodeius, Clethraorornys & Microtus. Proc.

zool, Soc. Lond. 126: 549-564

Borowski, S. & Dehnel, A. (1952). Materaly do Biologii Sor3cidae.

Ann. Univ. N. Curie-Skiod. C. 7, 6 305-1448 (I'2 Polish

with German sununary)

Buchalczyk, A. (1961). Variation in the weight of the internal organs

of Sorex raneus Linnaeus, 1758. I Salivary glands. Acta therj.ol, 5, 16: 229-252

Bucha1cyk. A. & Korybska, Z. (1964). Variation in the weight of the

brown adipose tissue of Sorex araneu Linnaeus, 1758. Acta theriol. 9, i4 193-215

Buckner, C.H. (1964), etabo1ism, food capacity & feedii,g behaviour in four species of shrews. Can. J. Zool 42; 259-279

Buckner, C.H. (1969). Some aspects of the popu1â.ti.rn ecology of the Common threw, Sorexramwu, near Oxfoid, Englind. J. Mammal. 50: 326-332

Burt, W.H. (1943). TerritorIality & home range conept as applied to mammals • J • Ma nir1. 214. : 3!4 6-35

Chitty, D. & Kempson, D.A. (1949). Drebai.1ng Email 'animals & a new design of live trap Eco1c 30: 536_5t!2

Churchfield, J.S. (1979. A note on the diet of the Eulopean water-

shrew, r'ysf -'_hico, Shaw, in Britain. Notes

from the Mamm-tl Soc] .y in J. Zool., Lond. in press. - 202 -

Corbet, G.B. & Southern, H.N. mis. (1977). Harxlbook of British Mammals • Blackwells, Oxford.

Crowcroft, W.P. (19511a). An ecological study of British shrews. D. Phil. unpubi. thesis, Univ. of Oxford

Croucroft, w.P. (19.54b). The daily cycle of activity in British

shrews, Proc. zool, Soc. Lomi. 123: fl5 -729

Croworoft, W.P. (1955). Notes on the behaviour of shrews. Behaviour 8: 63-80

Croworoft, P. (1959). Spatial distribution of feeding activity in the

wild house-mouse musculusL.) Annals of Appl.

Biol. 117,1: 150-155

Cummins, LW. & Wuycheck, J.C. (1967). Calorific equivalents for investigations in ecological energetics. Pymatuning Lab, of Ecol. Univ. Pittsburgh, Mirneo.

Dehnei, A. (1949). Studies on the genus 2C L. Ann, Univ. M. Curie- Skied. 4,2: 18-102. In PoUch, English summary.

Dehnel, A. (1952). The bio1o' of breeding of tie common shrew, rex araneus,in laboratory co1itions. Ann. Univ. M. Curie-Skiod. C. 6: 359-376.

Dzyden, G.L., Gebczynski, N. & Douglas, E.L. (1974). oxygen consumption by nurLng & adult musk shrews. Acta. theriol. 19, 30: 453-461

Evans, P.C. (1942). Studies of a small mammal popUation in Bagley Wood, Berkshire. J. Anim. Ecol. lii 182-197

Ferns, P.N. & Adams, M.G. (1974). The effects of laboratory confinement on lipid deposition in wood mice, bank voles & field voles. J. Zool., Lond. l74 524-528 - 203 -

Flowerdew, J.R. (1976). Ecological methods. Marnxrial review 6, 4: 47-83

Fons, R. (1974). Le repertoire comportexnental de la Pachyure Etrusque, Suncus etruscus (Savi, 1822). Ia Terre et la Vie. 28 1:31-157

Forbes, T.A.0. & Lance, A.N. (1976). The contents of fox seats from western Irish blanket bog. J. Zool., Lond. 179: 224-226

Gebczynska, Z. & Gebczynski, N. (1965). Oxygen consumption in two species of water-shrews. Acta theriol. 10, l3:209-21

Gebczynski, N. (1965). Seasonal & age changes of the metabolism and activity of Sorex araneus Linriaeus, 1758. Acta therio].. 10, 22: 303-331

Gebozynaki, M. (1969). Daily heat losses in the common shrew (Soricidas)

in d.fferent seasons. (In: 'Energy metabolism of farm animals' ed. Blaxter, K.L., Kielanoweki, J. & Thorbek,

G. Europ. Assoc. for Anim. Prod. Pubi, no. 12). Oriel Frees Ltd. pp. 397-400

Gebczynski, M. (1971). The rate of metabolism of the lesser shrew. Acta therlo). 16, 20: 329-339

Gebczynzki, M. (1977). Body temperature in five species of shrews. Act3 therol. 22, 35: 521-530

Gebczynski, N. & Olszewski, J. (1963). Katathermometric measurements

of Insulating properties of the fur in small mammals.

Acta theriol. 7, 19: 369-371

Golley, F.B., Petrusewicz, K. & Ryszkowsk&, L. (1975). Small mammals: their productivity & population dynann.cs. I.B.P. Handbook no. 5, Cambridge Univ. Press - 204 -

Gorecki, A. (1965). Energy values of body in small mammals. Acta theriol. 10, 23: 333-352

Gorecki, A. (1971). Metabolism & energy budget in the harvest mouse, Acta theriol 16, 15: 213-220

Grainger, J.P. & Fairley, J.S. (1978). Studies on the biology of the Pygmy Shrew, Sorex minutus in the West of Ireland. J. Zool., Lond.. 186: 109-141

Grodzinski, W., Klekowski, R.Z. & Duncan, A. (1975). Methods for Ecological bioenergetics. I.B,P. Handbook no. 24 Blackwell Scientific Publications, Oxford.

Gurnell, J. (1975) . Notes on the activity of wild woodmi.ce, Apodemus sylvaticus, in artificial enclosures.

J. Zool., Lond. 175: 219-229

Gurnell, J. (1976). Studies on the effects of bait and sampling intensity on trapping & estimati.ng wood mice. podemus sylvaticu. J. Zool., Lord. 178: 91-105

Gurnell, J. (1978). Observations on trap response in confined populations of wood mice, demus_sy1vaticus

J. Zool., Lond. 185: 279-287

Hamilton, w.J. (1930). The food. of the Soricidae. J. Mammal 11: 26-39

Hamilton, w.j. (1931). Habits of the short-tailed shrew

Blarina brevicauda (Say). Ohiu J. Sd. 31: 97-106

Hamilton, w .j. (1940). Biology of the smoky shrew (Sorex fumeus Miller). Zoologica N.Y. 25: 473-492

Harper, R.J. (1977). 'Caravanning' in Sorex species. J. Zool., Lond.

183: 54]. - 205 -

Harrison Matthews, L. (1972). British Mamra1s. Ncw Naturalist Series. Collins, London.

Hawes, M. (1977). Home range, territoriality & ecological separation in syinpatric shrews, Sorex vagrans & Sorex obscurus

J. Mammal. 58 (3): 35t1.-.367

Hawkins, AE. & Jewell, P.A. (1962). Food consumption & energy requirements of captive British shrews & the mole. Proc. zool. Soc. Lond. 138:137-155

Hawkins, A.E., Jewell, P.A. & Tomlinson, G. (1960). The metabolism

of some British shrews. Proc. zool. Soc. Lond, 135: 99-103

Hayn, D.W. (1950). Apparent home range of çyotus in relation to distance between traps. J. Mammal 31: 26-39

Hodge, CIA.H., Seale, R.S. & Burton, R.G.0. Geology & Soils (In: Monks Wood, A nature reserve record ,ed. Steele, B.C.

& Welch, B.c.)

Rolling, C.S. ( 1955) . The selection by certain small mammals of dead, parasitised and healthy prepupae of the European sawfly,

Neodiprion sertifer (Geoff.). Can J. Zool. 33: 1104.419

Holling, Cs. (1958 ). Sensory stimuli involved in the location & selection of sawfly cocoons by small mammals. Can.

J. Zool 36 633-653

Rolling. C.S. (1959) . The components of predation as revealed by a

study ci the small marunal predation of European sawfly.

Canadian Entomologist XCI 293-332

Hunkeler, C. & Hunkeler, P. (1970). Besoins nergtiques d.e quelques

crocidures (Insectivores) de Cte d-'Ivoire. La Terre

et l.a Vie 24: 9-456 - 20 ..

Jansky, L. & Hanak, V. (1960). Aktivitat der spitzmause unter naturlichen Bedingungen. Saugetierk. Mitteil 6, 8: 55-63

Jennings, T.J. (1976). Seed detection by the wood mouse Apodemus sylvaticus. Olkos 27: 174-177

Jewell, P.A. (1966). The concept of home range in mammals Symp. zool. Soc. Lor4. 18: 85-109

Kangur, R. (195L.). Shrews as tree seed eaters in the Douglas fir region. Ore. St. Bd. For. Salem. Res. Note 17

Kendeigh, S.C. (1945). Body temperature of small mammals. J. Mammal 26: 86-87

Kenward, R.E. & Sibly, R.M. (1977). A woodpigeon Columba palumbus feeding preference explained by a digestive bottleneck. J. appi. Ecol. 14: 815-826

Kikkawa, J. (1964). Movement, activity & distribution of the small rodents Qlethriononomys glareolus & Apodemu sylvaticus

in woodland. 3. Anim. Ecol. 33: 259-299

Kisielewska, K. (1963). Food. composition & reprodction of Sorex raneus in the light of parasitological research.

Acta theriol. 7, 9: 127-153

Kubik, J. (1951). Analysis of the Putawy population of Sorex araneus

araneus L. & S. minutus minLtu L. Ann. Univ. M. Curie-

Sklod. C, 5, 11: 335-352

Layne, J.N. & Redmond, J.R. (1959). Body temperatures of the least shrsi, Cryptotis parva florid ana (Merriam, 1895). Saugetierkde Mitteil 7: 169-12 - -

Lavrov, N.F. (1943). Contribution to the biology of the common shrew (Sorex araneus L.). Zool. Zh. 22: 3613611. ifl Russian, English summary.

Lewis, J.W. (1968). Studies on the helminth parasites of voles & shrews from Wales. J. Zool., Lond. 154 313-331

Linn, I. & Shillito, J. (1960). Rings for marking very small

mammals. Proc. zool. So0, Land. 134: 489-495

Lofts, B. & Murton, R.K. (1968). Thotoperiodic & physiological adaptations regulating avian breeding cycles & their ecological significance. J. Zool., Loryl. 155: 327-394

Loxton, R.G., Raffaelli, D. & Begon, M. ( 1975). Coprophagy & the diurnal cycle of the common shrew, Sorex araneus. J. Zool., Land. 177: 14149_453

Macfadyen, A. (1955). A comparison of methods for extracting soil arthropods. (In: Soil Zoology ed. Kevan, D.K.Mc.E.) Lond. pp. 315-332

Mezhzherin, V.A. (1958). On the feeding habits of Sorex araneus &

Sorex minutus. Zool. Zh. 37: 9/48-953. In Russian, English summary. flezhzherin, V.A. (1960). Population density of the common shrew (Sorex araneus L.) and its changes over 17 years.

Zool Zh. 39, 7: 1080-1087. Tn Russian with English summaiy

Mezhzherin, V.A. (1964). Dehne].'s phenomenon and its possible

explanation. Acta theriol. 8, 6: 95-114 In

Russian with English summary - £UO -

Mezhzherin, V.A. (1969). Energetics of popu1ation & evoluti.on of the shrews (Sorex, Insectivora, Maxnmalia). (In: Energy flow through small maLrnnctl populations. Proceedin.gs of IBP Meeting on Secondary Productivity in small mamnal populations. ed. Petrusewicz, K. & Ryszkowski, L.)

Michielsen, N. Croin. (1966). Intraspecific & interspecific competition in the shrews rec_araneus L. & Sorex minutus L. Arch. NerlaMaises de Zool. 17: 731714.

Middleton, A.D. (1931). A contribution to the biology of the common shrew, Sorex araneus. Linnaeus. Proc. zool. Soc. Lond. 1931: 133_11+3

Moors, P.J. (1975). The food of wea.e1s (Mustela nivalis) on farmland in north-east Scot].aM. J. Zool., Lond. 177: 455-461

Morrison, P.R., Pierce, M. & Ryser, F.A. (1957), Food consumption & body weight In the masked and short..tailed shrews. Amer. Midi. Nat. 57, 2: 493-500

Morrison, P.R., Ry'er, F.A. & Dawe, A.R. (1959). Studies on the physiology of the masked shrew, pre1v cinereus. Physiol. Zool. 32: 256-271

Morrison, PR. & Tietz, W.J. ( 1953) . Observations on focd consumption & preference in four Alaskan L1anunal • Arctic 6: 52-57

Murton, R.K. & Kear, j . (1978). Photopeciodis'n in aterfow1: phasing of breeding cycles & zoogeography. J. Zoo1. Lond. 186: 243-283

Myrcha, A. (1969). Seasonal changes in caloric value, body water & fat In some shrews. Acta. theriol. 14, 16: 211-227 - 209 -

Mystkowska, E. & Sidorowicz, J. (1961). Influence of the weather on

captures of Microma'nmalia II. Insectivora. Acta

theriol. 5, 18: 263-273

Newstead, R. ( 1947). Food of the pygmy rr lesser shrew (Sorex miriutus Linn.) Northw. Nat. 22: 103-101+

Niethammer, J. (1956). Das Gewicht der Waldspitzmaus, Sorex araneus Linn, 1758, iniJahres1auf. saugetierk Mitteil.1+,

14: i6o-i65

Pearson, 0.P. (1947). The rate of metabolism of some small mammals.

Ecology 28: 127-134

Pearson, 0.P. (191+8). Metabolism of small mammals, with some remarks

on the lower limit of mammalian size. Ccience, N.Y.

l08 44.

Pe1ik.n, J. (1955). PoznamkSr k bionomli popuiac.i nkterch naich

drobnycn ssavcu. Rozprevy Cesk Akid. d 65,

1: 1-63. In Polish. Cited by Kisie1e.ska (1963).

Pernetta, j.C. (1973). Field & laboratory ex .perimants to determine the feeding ecology and behaviout of Sorex araneus & S. minutus D. Phil. unpubi. thesis, Univ. of Oxford.

Pernetta, j.c. (1976a). Bioenergetics of British shrews in grassland Acta theriol. 21, 33: 1+81-497

Pernetta, J.C. (1976b). Diets of the shrews Sorexaraneu3 L. & Sorex ininu

L. in 1ytham grassland. J. Anim. Ecol. 45, 3: 899-912

Pernetta, j.c. (1977a). Population ecology of Brttish shrews in grassland. Acta theriol. 22, 20: 279-296

Pernetta, J.c. (19771)). Anatomical & behavioural specializations of shrews in relation to their diet. Can. J, Zool. 5: 11442_11,.53. - ._a, -

Petrusewicz, K. & Andrzeewski, I. (1962). Natural history of a free-. living population of house-mice (Mus musculus L.) with particular reference to groupings within the population Ekol.Polska A 10: 85-122

Petrusewicz, K. & Macfadyen, A. (1970). Productivity of terrestrial animals: principles & methods. IBP Handbook no 13.

Blackwell Scientific Publications, Oxford.

Phillipson, J. (1964). A miniature bomb calorimeter for small biological

samples. Oikos 15: 130-139

Plabt,W.J. (1976). The social organisation & territoriality of short-

tailed shrew (Blarina brevicauda) populations in old-field habitats. Anim. Behav. 24: 305-318

Pucek, M. ( 1973) . Variability of fat & water content in two rodent species. Acta theriol. 18, 2: 57-80

Pucek, Z. (1959). Some biological aspectz of the sex-ratio in the common shrew (Sorex araneus araneus L.) Acta theriol.

3, If, 43-73

Pucek, z. (1963). Seasonal changes in the braincase of some representatives of the genus Sorex from the Palearctic. J. Mammal L4, 523.-536

Pucek, Z. (1964). Morpho1ogica1 changes in snrews kept in captivity Acta theriol. 8, 9: 137-166 pucek, z. (1965). Seasonal & age chang ?S in the weight of internal orgus of shrews. Acta theriol 10, 26: 369-438

Pucek, Z. (1970). Zeaso1 & age change in shrews as an aiaptive prore's. Symp. zool. Soc. LoM. 26: 189-207

Quilliam, 'v.A. (1966). The mole's sensory apparatus. 3. ZoOl.)LOTld. 149: 76-88 - -

Randolph, S.E. (1975a). Seasonal dynamics of a host-parasite systems Ixodes triangii]ic"p (Acarina: Ixodidae) and its small

ziaxnrnal hozts. i Anim. EccI 14)4, 145J4J49

Randolph, S.E. (197.5b). Patterns of distribution of the tick Ixodes

trianguliceps Birula on its hosts. J. Anim. Ecol I: 451-474

Redmond, JR. & Layne, J.N. (1958). A consideration of the metabolic rates of some shrew tissues. Science N.Y. 128: 1508-1509

Richards, D.F. (1977). Observatinns on the diet of the red fox (Vuipes in South Devon. J. Zool. LOnd. 183: 495-504

Rudge, M.R. (1965). Feeding & metabolism of the common shrew. Ph.D. unpubi. thesis, Univ. of Exeter

Rudge, M.R. (1968). Food of the common shrew, Sorex_araneus,in Britain S. An.m. Ecol, 37: 565-581

Schmidt-Nielsen K. (1972). How animals work. Cambridge University Press

Seber, G.A.F. (1973). The estimation of animal abundance & related para'ioters. Griffin, London

Shilhito, S.F. (1960). The genera]. ecology of the commnn shrew sorex

araneus L. Ph.D. unpub].. thesis, Univ. of Exeter

Shillito, J.F. (l963). Observations on the range & movements of a

woodland population of the common shrew, Sore'_araneus L.

Proc. Zool. Soc. Loud. l40 533-546

Shillito, J.F. (1963b). Field observations on the growth, reproduction &

activity of a woodland population of the common shrew,

Sorex araneus L. Proc. zool. Soc. Lond. 140k 99-114 - 212 -

Skarn, U. ( 1973). Spring moult ari the onset of the breeding season of the common shrew (Sorex araneus L.) in Central Finland. Acta theriol. 18, 23. 1+3458

Slobodkin, L.B. & Richman, S. (1961). Calories per gin. in species of

animals • Nature, Lond. 191: 299

Southern, H.N. ed. (].954a). Control of rate and mice. Vol 2: Rats.

Oxford, Clarendon Press.

Southern, H.N. (195t4b). Tawny owls & their prey. This 96, 384-408

Southwood, T.R.E. (1966). Ecological methods. Chapman & Hall

Stickel, L.F. (19 .5ti). A comparison of certain methods of measuring ranges of small mammals. 3. Ma.mmunal 35, 1: 1-15

Tanton, MT. (1965). Problems of live-trapping & population estimation

for the ocdmouse Apodemnus sylvaticus (L.) J. Anim. Ecol

34: 1-22

Tapper, S.C. (1976). The diet of beasels, Mustela nivalis & stoats, Mustclt erminea during early summer, in relation to

prtdation on gamebirds. J. Zool., Lond. 179: 219-224

Taylor, B.J. (1977). The value of clumping to prey. Oecologia 30: 285291+

Tupikova, N.Y. (1949). The diet natuie of the daily cycle of activity of shrews front the central region of U.S.S.R. Zool. Zh. 28: 561-572

Vogel, P. (1976). Energy consumption of European & African shrews.

Ata theriol. 21, 13: 195-206

Wolk, E. ( 1969). Body weight & daily food intctke in captive shrews.

Acta theiiol. 14, /4: 35-47 - 213.-

Yalden, D.W. (197(1. ). Population density in the common shrew, Sorex araneu. Notes from the Mammal Soc. 28, reprinted

from J. Zool., Lond. l3, 2: 262-26(1.

Yudin, B.S. (1962). Ecology of shrews (genus Sorex) of Western

Siberia, U.S.S.R. (In: Problems on the Ecology,

Zoogeography & Systernatics of an!mals Vol 8) pp: 33l311..

Trans. no. 17, 973, Joint Pubi. Res. Serv. USA. Copy Elton Library, Oxford).

Zablotskaya, L.V. (1957). Pilfering of conifer & linden seeds by shrews. Prioksko-Terrasnyy State Reservation 1. In Russian, cited by YixIin (1962). -

APPENDIC

— 2.5 —

0 Q-4ç) cDOIC 0I'-ç) -3 mmoo 1 H,.i.,-CD cnomb c+ o-.-?bO +CD+Io c+J1Up 0 Ojd -H'rJ o -'CD CD<(DIO CD) H CD 0 J ca I-' 0 I_M d cl ' 0 d 0 I-'. i-i. 0 'j I-' I 3 td . () 1 0 $i H CD i-i rJ Q P. c d c+ H CD C) I m 0' p 000 r-CD0O OCDPdCD CD 0 0 H CD ) COO) OO'AV IS c+P co 00) 'cO ) ', + CD I-' I.J S I-' d W I(D (Il (0 .-... HH CD COdin + II 0'CD H(D • • H I 0'0' 0' J.&% CD 'to p 0' CD CD C) 0' H CD CD(D'AV+ 11V''

'- __ •__%__ 0O0OoOONOOON0O0O00 000H0OO HOOHN)O0'0 000 ON000N0000'..JO0 '00OH¼)O 0000NHO0OOI-NO HO0t')O¼.0 QL)0HHL,0H . OiCJ 000000H 000H0000000 00000HO 000N000 N OOHOOOOHOOI-0OOOOI-' 0000000 00000NO Iá 00 HH I-' 00 N 00 0 ) 00 L) 000 00 H H 000 ON000LA)O ) 0000NNOHOONN0000000H 00H0N0 OOOHOHt3 0 0 0 N 00 N 0) 0 0 H L) 0 0 0 O0 0 0 '0000000 0000HOO 0 N H 00No'OH 0)00 HO NO 000NHOI-' ¼) 00000-H'30H04O00')0H0 O\0000NH ONHN000 0'H 00 N H 0 L)H 0H H 0 L3 HN H 00 -'J ¼.) H0 N ON'3H)¼). -rN 00 H OH 0000 H t')Lj 00 H OO 00 0OH)4H )N t) HHH'0 0NN4000UH N 0 0 N H ON0)'00 4 0 00 H H N H C' H H H H N 00 H H N N00OH. 0)0N'0H '0

00-HHC0NO00H0N0H0NHN 0OH0H0 OHOH)N 000N0U00NN00H00 N000000 ONHNH0 0) 00 0 C N H . 0 N 0 - 00 . 0 H N 40 0 L O_.) 0 0)

Appendix 3-1: Aralysis of faeca]. pellets of araneus from The Wilderness (No. samples containing prey item listed). - 216 -

APPENDIX 3-2

A note on the diet of the European water-shrew, eomys fodlenB bicolor Shaw

by J.S. Churchfield Department of Zoology, Westfield College, Hampstead, London, N.bl.3. - 217

There have been a number of observations of the feeding

habits of the water-shrew, Nomy todiens, in c'aptivtty (eg.

Cranbrook, 1959; Hawkins & Jewell, 1962) but information concerning

its diet in the wild is sparse. Brewster (1966) observed a water-

shrew catching freshwater snails and Pernetta (1976) reports the occurrence of an adult smooth newt, Triturus vulgaris, in the gut of

a specimen of N. fodiens. In Poland, Buchalczyk & Pucek (1963) describe the storage of water snails, dytiscid beetles, frogs

(Rana tentperaria and R. arvalis) and fish (Esox lucius and. Lota1ot) by N. fodiens in the wild, and Wolk (1976) lists a variety of freshwater insects, molluscs, amphibians and fish which are eaten by this species in winter months. However, the extent to which the food. resources of the aquatic habitat are exploited by N. fodiens in

Britain is not known.

Captures of T .fodiens using Longworth traps were made between

August and November 1977 in two different habitats; one, commercial water-cress beds in Hampshire and the other an area of mixed. deciduous woodland some distance from water in Cambridgeshire.

The facca]. pellets produced by these shrews in the traps were collectcd. and analysed for food remains. Only invertebrate remains were found and these were identified with the aid of a reference collection of potential prey items taken from both areas. A total of 10 samples

(each sample comprising between three and five pellets from a single shrew) were analysed, 7 from the water-cress beds and 3 from the mixed deciduous woodland. Although this is a small number of samples, a considerable variety of invertebrate remains was found. (see Table 1).

The degree of fragmentation of the invertebrates made quantitative estimates impracticable, although in one sample a total of 9 Simulium larvae were identified.. - 218 -

The greatest variety of invertebrates in thc' diet of .hres from both localities wa of terrestrial origin and similar to that found in the diet of the common shrew, Sorex raneus, by this author, with the notable exception of a round-backed rillipede which was found in one sample from N.fodiens but has yet to be recorded in the diet of S. araneus. The most common occurrences were luinbricids, isopods and dipterans (both larvae and adults). Aquatic as well as terrestrial invertebrates appeared in the diet of N. fodi.ens from the water-cress beds. Here, the most abundant aq.zatic invertebrates found were the freshwater crustaceans Gammans pulex and. Aellus aguaticus; fragrents of these species were identified in the faecal pellets of 3 water-shrews. Other aquatic invertebrates recorded in the diet included two occurrences of Simulium larvae and one of a

net-spinning trichopteran larva.

The intermittent and localised occurrence of N. fodiens and its migratory habits (Shillito, 1963) make it a difficult species to study in the field. There have been many reports (generally unpublished)

of water-shrews being found considerable distances from water, and the

results of this study confirm that these shrews ar not nsce.sarily

restricted to aquatic habitats in order to obtain food • tater-shrews

in terrestrial habitats are subject to competition from othe.c species

of shrews, but theli ability to catch small fish, aiiphibian and.

invertebrates underwater (personal observation) per'nits then to

exploit prey itexnb associated with swift1y-flowir streams (including

water-cress beds) without competition from arty other insectivorous

mammal. - 219 •-

NUMBER OF SAMPLES

Water-cress Beds Mixed Woodland PREY ITEMS fl)ENTIFIED 123 Lf .567 1 23

Terrestrial Invertebrates

Dipteran adults x x x x

Dipteran larvae x x

Tipulid. larvae x

Hemipterans x x

Staphylinids x x

Isopods x x x x

Geophilomorphs x x

Lithobiomorpha x

Diplopods x

Opilionids x

Lumbricid.s x x x x xi x

Aquatic Invertebrates

Simulium sp. larvae x x

Trichopteran larvae x

GaTnrnarus pulex x x x

Asellus acjiaticus x

Table 1: Analysis of faecal samples from Neo'odiens•

A cross denotes the presence of the listed pre,r items in the sample. - -

Refireuceq

Brewster, R.W. (1966). Hunting behaviour of the Water Shrew.

Essex Naturalist 31: 377-378

Buchalczyk, T. and Pucek, Z. ( 1963) . Food. storage of the European Water Shrew Neorys fodiens (Pennant, 1771). Acta

theriol. 7: :376; 377.

Cranbrook, the Earl of (1959). The feeding habits of the Water Shrew, pmys fodiensbicolor Shaw, in captivity and the effect of its attack upon its prey. Proc. zool, Soc.

LoM. 133 245-249.

Hawkins, A.E. and Jewell, P.A. (1962). Food. consumption and enerr

requirements of captive British shiews and the mole. Proc. oo1, Soc. Lond.. 138, 1:137-155.

Pernetta, J.C. (1976). A note on the predation of the smooth newt, Triturus vulgaris, by the european water-ehrew, Npornys -fodiens bicolor. J. Zool. Lond.. 179,

2:215-216.

Shilhito, J.F. (1963). FIeld oboor',ations on the water shrew, Neoriys fodiens. Proc. zool. Soc. Lond. 140: 320-322.

Wolk, K. (1976). The winter food. of the Euripean water-shrew.

Acta theriol. 21, 6 117-129. - 22]. -

8 ')H t '0 Ut NJ) COC\Ln CoH OLn S S S a S • S S S x N 1) 0 - .4 Cx) N) CE H - • - '0 tj - ) L.) 0 0% '0 Q 44 4+ $4 44 *4 If 44 44 4 4 44 44 4 .4 if *4 If ii. *4 44 44. •f 4- t1 ) ¼. 4-' '0 0' N) 0 H 0 I-' 0' Lr - H F-' H 4 N) Lxi -3 I • • 0 . a • • • • • a • • • 0 • • • a • N) CD • 0% - '0 '0 N) . N • N) Lrt '0 N) CO > .. 0' H H CD H a) 0 (7' CO ) 0' -4 4 '0 '0

•00 . • --'O• • HOa • • 000a • a HN).)'0. • . a •4-jN) • • HH• a N CO - 0 0 - 0' - N Cx) N) 0 - '.) -.i .4 N) - .4 i-i 44 *4 *4 *4 44 *4 *4. 4-4 *4- *4 14 *4 *4 $4 44 $4 $4 *4 p -f i-f d 0 0 OOHN)000 OOC) 00 N)) N)000 • • • , • • . • a • • • . . a • . . • • 121 0'CD N 0'-. -4 0 I-' 0% N) ('-3 N) ¼) - '0 4 C' 0'. 0 0' N 0) N) 0 CD 0 H N Cx) 4 0 '0 N Lri L,*. 0 '-I) O OHN00000 O)O 000 HO H000 I • S S • S a S • S S S • • • S S Jc, - - 0'-. CD (7'. N) N) 0 N) H CD 0 H .n n 1' *4 4 4. 44 44. 4+ *4 4+ *4 *4 *4 14 (4 44 44 $4 000000 0 N) 0 0000000 I I. • • • • • • I • I I • S • S • • a • • • • Iz,I-3 !(D - - 0" '0 N) N N '0 1-" '-.) N) N s.-.) ') (\.) - I cl 0 0 0 -4 '.j 0 0 '0 - '..) ¼n 0 4 N) N I& Ico 0-.. N) N H N) CD - 0'-. N) 0 • L.) N) 3 0 - N) • • S S S S S • S S • S S S S S S • S S S O - - 0% N - 0) I-' N) 0 '0 N) - Ct' I-' 0) - 0 4 C/i 44 *4 *4 14 '4 4+ $4 '4 *4 *4 *4 4* 4-i. *4 4+ 4+ 44 44 1+ *4 44. .0 Lxi H0O HOHOHO H N)O0 0000 OF-'Hf-' • S S • • • a • . . • S • S S S S S • • . 121 H ¼n '0 - '-..) 0 '0 CO - H N) CX) 0) 0'. 0) N) 0 0 0' - H CD 0 .4 '0 N '0 0) 4.j CX) H H C ¼.n '0 N N H

4-' H I-' I-' I-' I-' H 4--' 0 O U 0 H N 0' L) N ) F-' 0 '0 O'0 La) 0) 'd • . e • • • • • • • • . . . • . a a • • a I-I - O N 0'. . - 0 0) H 0 CD - N) %.j 0 .4 0 Lri -) C) Ct' 44 *4 *4 *4. *4 44 1-. 4 + + *4. 44 $4 + + *4- *4 *4. 14 $4- + F-I Lxi HLa) NLa)N)La) N)HOH NO 0 HHHH HNN)H 0 • • S S S • S S S S • S • • S S S S S S S 0' -'0 I-' .ji (7'. H 0 CO Laa CO\O Lr* 0% - La) . 4 CD N '0 H J2 O N) Lj -.4 -ç •- 0'. 9 I-' H 0 La) .-4 - (0 CD CD '0 -ç'

- 0 -.i N) C' - N) 0' 0'. 0 - - La) .4 0' I-' '0 L.) 4 O ) H • S • S S S S S S • S • • • • S • • • S • (.1) O 0) - (7' 0 0\ - 0 - 0) N) 0 4.1 La) '--3 r N) La) N) '4 0% Ci 14 *4 44 i-i- I-i- i-f 1* (-4- '4 4-4. 44 1 $4 4-4- 1+ 14- (4 *4 *4 1+ *4 •N S00 • HHHOS • • HOOS S • HN• S HS N)N3S S S N)N.)S • • HHHOS S S N) Ln La - Q) La C' -4 '0 La '0 H H H N) - La) La) '0 Cl) CO H 0 4 0 La) 0 0) 0% Lit H H 0 Ln H H

NO OHO HNHOO 000 HO H4--'OF-'F-'O o . a a • • • a a a a a a a . a a • • S 11-4 N) N) 0) C) CO 0 0 CD'0 - 0'. - H .4 '.0 0'. - La) 1-' CX) I 44 (4 4 -4. 44 44 4+ 4-4. *4 4 '-4 *4 4-4- $4 4 4 4- *- 00000000000 000000000 S S S • • S •CD .N La)S • L.S S COS 0) SLa) S N -S I La)• • H N)S -S -ç- O 0 ..4 Lit -) Ln - 0 N) Lit 0 H La) C' - H N) 10) C) 0 La)• S I-'S H• I-'a ()• H• 0S 4-'S 0S 0 S 4-'S I-'S 0• 0S 0 S0 0S 0S 0S 0 50 CC)5 (11 N . C'. - -4 N N) N) 4 - C' 0 N) 0' Ln Cl'. '..*) H I-' L.) La) C. 4* 4 4+ *4 *4 *4 t* 44 *4 15 *4 -4 4'. 4+ *4 (-r 44 4 4 4 s-I O 0 0 0 0 0 0 0 0 0 0 0 C) 0 0 C) 0 0 0 0 0 • S S • S S S • S S S a S a S S S S S • a La) -4 (7' 0'-' N) -4 N) '0 N) N) '.jt La) I-' N) I-' H N) I-' I-' H I-' .4 4j 0 0) - La) 0 - N) Lit I-' N) La) N) 4 0" H 0 0 Lit

La) N) H 4--' H H H La) - La' H H H H 4 0 '0 0' Lit C'. N) 4--' '0 % '0 H - H N) 4'.0 • . a a • . • a S S • S S • S • S S S S S N) N) C' N) 0 CD -P N) Lit - N) C' 4 0' .4 - C' La) CD 0' i* I 44 44 44 4* *4 44 44 4-5. 4- 44 $-j 49 *4 14 44 # 4+ 4-'.- *4 14--I N N) N) NJ N) 0 H NJ 'C 4 - La N) N 0 N N N) • • • S S • • S S S S S S S S S a S • S a N) '0 0) 0) 4 0'. CD '4) Li La) Lit I-' 'C) H La) '..jt 4-' Lit NJ '.0 O 4 4 0 0 La) H H - '0 Lit 0) Lii '.0 '0 H -r 0 0% 0 C). 1 Cl)

H HHHHHHHHH o,samp1e Lit Lit Lit tj'. Li' ¼t U'. Lii L) Li'. Li'. Lit C) 0 0 0 C) 0 0 0 0 -

— —

H r O 0 C 0 H 'C '0 0 0 N. CD a 0 CO Hr-4 N V\(I H N I-I 0' C\ O 0 '0 H 0 0 0 4- '0 '0 ('\ 0 ( 4. r4 0 H H H N.

H H N. H 4- 0 CP (4 CD 0 0 H H H C'\ W\ N H i-I N N H C '0 e-1 Co N C O r\ '0 4- 0 0 0 0 4- '0 r' O' 'C 0 C) C\ 0 N N H 4- H Co 4- N H U) CD N. 4- 'C Q' -0 0 0 0 C 0' 0 H 4- CD '0 4-HO' 0 H CD '0 N- '0 0' 0 N. 0 ( C 0 C 0 0 0 ,-4 0 4-HN. N C') H C- 00C4.N-HN04-0OOO4-HCDO 4- HHH

N. '0 c '0 U) C'\ 0 C 0 0 C"'0 0' 0 4- 4- '0 0 N. N NC'1 N 0" H 4- '0 0 '0 4- 0 '0 0 CO 0 0 N. C'- 0) H '0 0 H-

4- N- H 0' ( 0 \ 0 N 0 0 C\ 0 'C 4- 0) 0 HHrI '.0 N HN

4-CoHCr\000C\00N.00H0 0 HHHN

'0 0 c-I \0 0 C) (4 0 '0 0 0 N. 0 0 0' 0 0 H 4-

N- N. C C'- C'\ Cr C'\ 0 0 (' 0 0 0 H 0 C'- 0 0' H i-I H 0 4- H GD H '.0 0 '0 0 0 0 -* N. U) C\ N c-I H %1\ H H H H '0

CO ( 0 CD 0' H N. 0 0 C'- 0 H r-1 N- 0' H H 4- '0

E4 4- 4- N U) 0' C 0'. H C '0 0 '0 0 CV\ H '0 1.) N O U\ N. N H N CD H C 0 '0 H N H H '4 4- 0' H '0 0) 0 - c-I 0 0 i-I c-I 0 r\ '0 0) c-I 4. N H N H H N. U) 4. H c-I c-I H 0'. 0'. 0 CO - c-I 0 0 0 0 ( 4. C) '.0 C'- C'\ H 0' s-I C'. c-I H 4. H c-I N c-I N H N- c-I c--I C'- c-I 0 0 U) c-I '0 ( 0 C\ 4- 0 c-I '.0 c-I N N c-I H c-I '0 '.0

ON.0HHCO0GD N.ON.OHHO4- 4- (0 (0 N H ( H c--I Ci) H z '0 0'. N r\ 0 .i 0 4- N 0 c--I 0 %t\ N. 0 r-44- 4-H N N c-I H N '1 ) c-I H E a w * 43 '-' E"—' ' H C)

tr. > A-1C-4r\ I '-' '-' I) '-.' c '. (' u) C) H 4-' H '-''-.- 43 U) P4 c--f > (0 4S I 0: U) )'4 C) 1.. -c-I 4) c'j '-p.-.- (/) c--f 'c P4 C) I.i (I, .r-I 0 C) -4 4.) C) (0 -c-I H 0 +3 ,rJ P 4' "-4 d P4 c-4 0 P40 .rf 0 0 1-s : c-I S-i rJ -c-f (I) P4 & W .r4 : . -i : c-I . : s-f I Pi .r P .rf - C- '-i •'- 0) 4) o Ps (1) 0 0 F-4 C) p- - 0

- 223 -

O0HO4.000H\OHVN N H H (Y(J C 0 0 0 0 C'\ 0 0 H 0 C 0 '.0 H N0ON C\ 0 '.0 C'3 \ 0 4. U) 4. C H '0 HN'OHHr-I 4.000C$\0'0OOHrl4.NN-H

H 0 0 C) '3 0 '.0 0 U) r-4 C'\ '.0 4. CO H '.0r-I H '0 0 0 0'. 0 0 t-r 0 U) 4. '.0 r'- c a' Nc\ H\N-H N0 0 c' 0 '0 0 0 C '0 N'0 '.0 r' C'. N

( 0 0 0 0 '.0 0 N (\ 0 N- '0 N S N N 4.C'.1 HrI4. 0,

H 0 H 0 H 0 0'. H - N- 0'. '.0 ç 0 H '.0 H r4oc C) H '0 0 0 0 0'. 0 0 r\ H (\ 0'. '.0 H CO .ri H 0 4.4. C, 0 0 0 0 0 0 0 4. '0 4. N- N U) H N U) ('.1

0 0 0 0 '.) 0 0'. (\ 0 C\ '0 4. 0 0 NN i-I N-

H 0 0 4- '0 0 0 \ 4.

H 0 0 4- N- '0 '0 0 N '0 '0 '0 CO H H U) C',' 00000 04-0\04-\0N-c

r-I 0 0 '0 4 4. 0'. 0 ' ' H 8 Hc' c' 0 r'-. H 4- '0 0 '.0OCr\ H 0'. 4. 0\00'. U) H H H cl-I ri % i-I 0 N O H H '0 H H 0 -I 0 N- H H H

4- N- C) '0 0 C) H 4 N- 4- H ('UH 0 H C) 0'. 0 4- 0 H '0 0 0 H 0 0'. 4- H z -. ,_. -S - -.' a)I E c c!l 0 F4 cdt H v-I H '- 0 0 bCI C'-' '/i I'. ' i-f Hi 5-, "-" .1''- -I P'. '.4- I -1 I >4 - '-' I (1)1 d)i --5 -i-I a) (I) IiP - Eir-1icd C) ..-ii w II o 0 'dt(rI I E U) .rlOI o = Hi I .,-iI H = 0 r-J crj HI .1-I 0 q-4IiHI 5--' 0 .cOd)I o H Cs) v-i: %4Cs, •rl tO ,0 1Iviol C): o 43 do uj .i a) •-I ..-i -( c-. 01 ø (1) (1) e:

0.0.r.4O00000OOQ.Zt(\0C)O

HOr1'O O0000r4HOC' 000

t' H 4 H H 0 0 0 . H H 0 0 0 4 0 0 H

N- (\ "0 CZ) '0 H H 0 0 0 C\ 0 0 0 C) C\ 0 0 (\z ç.j H H

( C' N- (' H 0 N. 0 0 H 0 ('Z 0 C'J C) C\ ('Th 0 0 0 i-I r-1 H

HOC H \0 H i-I

HHHH000C'0c0CHO C'00 H '0

(r\ c'00

N- N-- N-0N-00N-00HC'3U\00H'O (-4

HH0c]0OHH 0N-Ot'-r-100 H Hrl

H H

"\ 0 0 '0 0 '0 0 0 r\ 0 ( 0 H c? H 0 0 'OHH c-I

"C) .- %r0000000v'000O0H C0O c-I H H 0 "oo z

P4 ft a, + ' E E E -# ' 0 0 E Cd r-1r1 Cd (-\ lcd -S 4 L i Cd -' '-' a) Cd U) a) H E E

CI) P4: H Cl) tJ\ I 0 Cd i\C G) cd Cd ..- : C-i '..-' ' (0 C) s-i k cdH Ps a) (I, •ti C: •ri 0 0) -i-a 0) .rj (0 ..-4 0 s-I C) 0) Ps+'.i •tj r-1 0 PS Ci 0 •r-1 - P4 H S-c 'ti ..-i 0 ri.0 P4 0) 0) .-4 0) en - H Ps .: HE Ps .: P4 .ni 0 0 0) .4 'D o Ps 0) Cl) 0 L)X E- 0 ç5ç. 00

- -

00 OH 0 OH 0OCH OC 0 H

0 0000r-lr-1 OOH000'O O C'-'

0H000000cst000c'.

0Oc0Oc'J0(\O0O0cD H J H CJ

O0HC'OC'OHC'0OOH

0HH0000c0O0O0OO '0

O0'0OCt'-C'JOC'000OOt\

0 OC'0r-I OH 0 000rI 00 4 H

00 00 HH OH COO OOrl .4.3

0 0 0 0 0 H 0 0 0 H 0 0 0 t- r'- i-1

00'000HH0C\HOO C' cock-' U] C) +3

C) 000 000H0HH0cOr'- •rI

0 0 0 H 0 0 0 rI 0 0 0 0 0 0 C." H C) ,0 OH OH OONOrl O0OOcC1 i-I

-S -S

C) 0 0 C) c3I ri ri r-i ri bCl A s A Vir C) S___ •_ I p4 I 11)1 G oI U 4.) s..' Ih ..4 ri P a] .-4) t)I W o rJI U)I I U) : U) ru cdl O ri o H( Jr4 •r4 o o C) 4.) /) a ..- I r-1 U] U) .rI o o 'TiL C)I tiji v-I CO -i C) tI Ui ol 0 v-I IOO +3 p4 CO H .ri .r4 .ii ( 01 o o (1) U) i- c)