BCRA 14-3-1987

8/21/2019 BCRA 14-3-1987

http://slidepdf.com/reader/full/bcra-14-3-1987 1/52

a Science

The Transactions o the British

Cave

Research ssociation

BeR

I

olume 14

Number

December 987

I

M2 Cave Picos

de Europa Spain

B.C.R.A. Symposium abstracts

Ankarana Caves Madagascar

Caves

of

Glomdal Norway

Caves and Mines of Hawkstone Park

Leptospi ros is among Bri t ish

Cavers

8/21/2019 BCRA 14-3-1987


ave

Science

The

Transac t ions of

the

Br i t i sh

Cave

Research

covers

a l l

aspects of spe leologica l

sc i ence ,

inc luding

geology,

geomorphology

,

hydrology

,

chemistry

, physics ,

archaeology and

biology

in

t he i r

appl i ca t ion

to caves .

I t a l so

publishes

a r t i c l e s

on t echnica l

mat te rs such

as explora t ion

, equipment , diving ,

surveying

,

photography and documentat ion

,

as

well

as expedi t ion

repor ts

and

h i s to r i ca l or

biographical

s tudie s

.

P

apers

may be

read a t

meetings

held

in various par ts of

Bri ta in

, but they may

be

submit ted for

publ ica t ion

without being

read .

Manuscripts should be sent to the Edi to r , Dr T . D. Ford, a t the Geology

Department , Univers i ty o f

Leices te r

,

Leices te r LEl

RH .

In tending authors are

welcome

to

contac t

e i t he r

the

Edi to r

or the

Product ion

Edi to r

who

wi l l

be

pleased

to adv i s e in any cases of doubt concerning the prepa ra t ion

of

manuscrip ts .

NOTES FOR

CONTRIBUTORS

These notes are intended to help the authors

to prepare their material in the

most

advantageous

way so

as

to expedite publication and to reduce

both

their

own and

edi torial

labour. I t saves a

lot of time i f the rules below

are

followed .

All material

should

be

presented

in a format

as

close

as

possib le to t ha t of CAVE SCIENCE s ince

1985.

Text should be typed double-spaced

on

one

s id e

o f

the paper only

.

Subheadings with in an

art i c l e should fol low the system used in C VE

SCIENCE; a

system

of primary,

secondary, and i

neccesary,

t e r t i a r y subheadi

ng

s s hould be c lea r ly

ind icated .

Abstrac t : Al l mater ial

should

be accompanied

by

an abs t rac t

s t a t ing

the es sent i a l

re su l t s

of

the

invest igtation

for use by abst ract ing l ibrary

and other serv i ce s . The abs tra c t may a l so

be

publ i shed in CAVES AND CAVING.

References to

previously

published work

should be given in the standard format used in

C VE SCIENCE.

In

the

text

the

statement

referred

to should

be

followed by the

re levant

author s

name and

date

(and

page

number , i f appropr i

ate

in

brackets . Thus :

(Smith, 1969,

p . 42 )

.

All

such

references c ited in the text should be given in

fu l l , in alphabet ical o rder, at

the

end . Thus:

Smith, D.E

. 1969

.

The speleogensis of the Cavern

Ho l e . Bu

eting

Yorkshire Caving As s oc . , Vo l

7

p.

1-63.

Books

should

be

c i t ed

by

the

author,

date , t i t l e ,

publ isher and where published.

Periodical

t i t l e s should

be abbreviated in

standard

style

, or , where

doubt ex i s t s

, should

be

written

out in fu l l .

Acknowledgements: Anyone who

has

given a

grant

o r helped

with the in v es t ig a t io n

, o r

with

the prepara t ion of the

ar t i c le ,

sh ould be

acknowledged

br ie f ly

. Contr ibutors in

Univers i t ies and other ins t i tu t ions are

reminded

that grants towards the co s t

of

publicat ion may be

av a i lab le

and

they should make the appropr iate

enquiries

as

early

as poss ible .

Expedit i

on

budgets

should inc lude an

element to help

publication

, and the edi tor should

be informed at

the

time

of submission .

I l lus tra t ion: Line diagrams and drawings must

be in BL CK ink on e ither clean white

paper

or

Card , or on t racing paper or such mater ia l s as

kodatrace.

Anaemic

grey ink and penci l wi l l not

reproduce

I l lus t r a t ions

should

be

designed

to

make max imum use of

page

space . Maps must have

bar scales only. I f photo - reduction i s

co n

t emplated a l l

l ines and l e t t e r s must be

larg

e

and

thick

enough

to

a

ll

o w

for

the ir reduction.

Let ters mu s t be done by s t enc i l

,

l e t raset

or

s imil r

methods

, not handwritten . Diagrams should

be numbered

in sequence as

f igures

and

referred

to in the te x t where necessary by inserting

Fig. 1)

e tc in

brackets . A fu l l l i s t of f igure

capt ions

should

be

submitted

on

a

separate

sheet .

PhotographS

are

welcome. They must be good

clear

black

and

white

prints , with sharp

focus

and

not too much contrast ; prints about 15 x 10 cm (6

x

4 inches) are best ; i in doubt a se lect ion may

be submi t ted . They should be numbered

in

sequence ,

but

normally not referred

to

n the

t

ex t .

A fu l l

l i s t

of plate

captions , with

photographer credits where

relevant

should be

submitted

on

a separate sheet .

Tables : These should not

be

included

in the

t ex t but should

be

typed, o r c lea r ly handwri t ten ,

on separate sheets .

They

sh oul d

be

numbered 'in

sequence ,

and

a l i s t of capt ions i necessary,

should

be

submitted on a separate sheet .

Approxima t e

locat ions

for tables pl

ates

and

f igures sh ould

be marked

in penci l in

the

manuscript

margin, unless already

clear from

the

tex t .

Co

pyr

ight: I f

any

te

xt

,

diagrams or photos

have

been published elsewhere , it

i s

up to the

author to c lear any copyright or

acknowledgment

matters.

Speleological expedi t ion s have a

moral

ob l i ga t ion to produce r

ep o r t s

contractual in the

cases o f rec ipients o f awards from the Ghar Pa r au

Foundation . These shou l d be concise and

cove '

the

re su l t s o f the expedit ion

as

soon as poss ible

a f ter the

return from overseas

,

so that

l a t t er

e xpedi t i o ns a r e informed for their planning .

Personal anecdotes

should

be kept t o a minimum ,

but us e

f u l advice such as location of food

suppl ies

,

medical

services e t c . may be included .

Authors may order reprints of their contr ibut ion

for

the i

r own private use . The orde r must be

noti f ied to the

Editor

a t the time

of

submission.

I f you

have any

problems

regarding

your

material , please consul t e i the r of the Editors in

advance

o f

submission.

8/21/2019 BCRA 14-3-1987


ave Science

TRANSACTIONS

OF THE

BRITISH

CAVE

RESEARCH ASSOCIATION

Volume 14 Number 3

December

1987

Contents

Geology and speleogenesis of the M Cave System,

Western

Massi f , Picos de

Europa,

Northern Spain

Kevin J Senior

B.C.R.A.

Cave

Science

Symposium, November 1987

Abstracts

The

Crocodi le Caves of

Ankarana :

Expedi t ion

to

Northern Madagascar , 1986

Jane

Wilson

Caves of Lower Glomdal, Rana, Norway

Simon

B o t t r e l l

Caves

and

Mines

of

Hawkstone

Park ,

Salop

Terry Middleton

Leptosp i ros i s

among

Br i t i sh Cavers

C A Se l f

e t

a l

Cover: The M Cave,

Spain .

By Kevin

Senior

Editor : Dr. T.D. Fo

rd

Geology

Dept. , Leices ter

Unive r s i ty Le ices t e r

LEI

7RH

93

105

107

121

125

131

Production

Edi tor : Dr. A.C. Waltham, Civ.

Eng.

Dept . Trent Polytechnic Nott ingham NGI 4BU

Cave

Science

i s pub l i shed by the Br i t i s h Cave Research

Assoc ia t ion

and i s

i s sued

to a l l

pa id

up members

of the Assoc ia t ion .

1987 subsc r ip t i on r a t e s a re : Individua l £10.00,

I n s t i t u t i o n

or Club £12.50

Subsc r ip t ions should

be

sen t t o

the

Membership Secre ta ry:

D.

Stoddard, 23

Claremont Avenue,

Bishopston, Br i s t o l

BS7

8JD

Individua l

cop ies

and

back

numbers

of Cave Science are obta inable from:

B.C.R.A. Sa les

20 Woodland Avenue,

Westonzoyland Bridgwater ,

Somerset TA7 OLQ

Copyr ight the

Bri t i sh Cave

Research Assoc ia t ion 1987.

No pa r t of t h i s publ ica t ion

may

be

reproduced in any other pub l i ca t i on used in adve r t i s i ng s to r ed in an e lec t ronic r e t r i e v a l

system, or o the rwise used for commercial purposes , withou t the pr i o r wr i t t en

consent

o f the

authors and of the Assoc ia t ion .

ISSN 0263-760X

8/21/2019 BCRA 14-3-1987


8/21/2019 BCRA 14-3-1987


CAVE

SCIENCE

Vol. 14, No

. 3 ,

December 1987

Transact ions

of

the Bri t i sh

Cave

Research Associat ion

Geology and Speleogenesis of the M2 Cave

System

Western

Massif

Picos de Europa Northern Spain

Kevin

J .

SENIOR

Abstract :

The

geological his tory of the picos de Europa has

been

par t i cu la r ly

favourable from

a

speleologica l point

of

view with 1200 metres

of

carbonates

thickened

by

over thrus t ing in 2000 metres

of

r e l a t ive re l i e f .

The

M cave

system, in the

Western

Massif, has been

explored to a

depth

of

-

986

metres.

The

upper par t of the M cave or ig ina ted as a s ingle small

phrea t ic passage which descended s teeply to the major phrea t ic

condui ts

below.

Two

models

of the

ear ly development

of

the cave

are

discussed. The

f i r s t

model ha s a

bathyphreat ic system modified

by

vadose entrenchment

while

the second proposes t ha t passages in the upper cave

are para-phrea t ic

and formed wi th in an extensive vadose zone. Incl ined ramps in the upper

cave are vadose fea tures formed where the base of each pi tch has re t rea ted

headwards

and cut

downwards, Glacial

meltwater

i s

proposed

a s the erosive

agent of

the

pi tch

-

ramp systems. Sediments

preserved

wi th in

the

lower

passages s t rongly resemble Red Permian mater ia l s

and

demonstra tes the

existence

of

a post -Carboniferous

cover

dur ing the

ear ly development

of

the

cave

.

INTRODUCTION

The

picos de Europa mountains of northern

Spain are divided into three

massi fs by

spectacular gorges which dissec t the range

from

south to north

(Fig

.

1 .

In the Eastern Massif

severa l major cave systems have been discovered

and

explored by

Lancaster Universi ty Speleological

Society

LUSS) and the

Seccin de Espeleologa

Ingenieros

Indus t r i a les (SEII)

(Sef ton,

1984),

while

in

the Central

Massif

groups of French

cavers have discovered major systems

such

as Torca

de Urr ie l lo and Sima del Trave (Beno i t

, 1985) .

The Western Massif

i s divided

in to northern and

southern

sec t ions

by

the

r idge of

the Picos

de

Cornion

.

Oxford

Universi ty

Caving Club

OUCC)

have been

working

to the nor th of the

r idge

for

many years where

they discovered

Pozu

del Xitu

(Singleton, 1981). However,

the region

to

the

south

received r e l a t ive ly l i t t l e a t ten t ion

from

speleo l ogis t s unt i l 1983

when the

area

was

explored by a combined team from York Universi ty

Cave and Pothole Club YUCPC) and the SEII from

Madrid.

The same group has

returned to

the area

each

summer

since 1983 and has focused

a t ten t ion

on the cave numbered M2 af te r it s discovery in

1984 in the Vega Huerta

area

to the south of Pena

Santa.

Figure 1 The th ree massi fs

of

the Picos

de Europa

.

•

I DfIID

ovadonga

93

GEOLOGY

The

three massi fs are composed almost

en t i r e ly

of

limestones which were

la id

down

during

the Carbon

i fe rous

Period of between

290 and 345

mil l ion years

ago.

In the

Western

Massif

approximately 1200 metres of carbonates were

deposi ted with very few impermeable hor izons so

that

most

formations are su i tab le for cave

development

(Fig. 2).

The

geological h i s tory of

the Picos de Europa since the beginning of the

Carboniferous

resu l ted in

a region with large

speleologica l potent ia l and the de velopment

of

cave systems such as M2

i s in t imate ly

re la ted to

the

geological s t ruc ture and l i thology.

The

Carboniferous Period and Variscan Orogeny

During the

Carbon

i f erous Per iod the Picos de

Europa occupied

a

palaeogeographic pos i t ion

which

was

favourable

for

carbonate deposi t ion.

While

neighbouring

regions experienced increas ingly

unstable condit ions the picos de Europa remained a

r e l a t ive ly s tab le province

in

which

carbonate

deposi t ion predominated

for

55 mil l ion years .

The

picos de Europa

was

located on

the

nor theas ter n

margin

of

the Cantabr i an Zone, a palaeogeographic

uni t

dis t inguished by Comte

(1959).

The province

was

bounded to the

west

by

the

Ponga

Nappe

8/21/2019 BCRA 14-3-1987


Lithology

Format ion

Figure 2

The

genera l s t r a t i g r aphy of t

he

Wes te r n Ma s s i f . ,.. -? - ;-black shales

Modified a f t e r Farias (1982). t i t

' l imestone

Age

turbidites

LEBENA Fm.?

/

?

-

r

? - -=b ='-a-c-k-:: -g-r-

ey-

,-th-i-n-'y-

C KAS IMOVIAN bedded,bioclastic

I

e S ~ 9 n e s shales

VEGA HUERTA

CARBANAL Fm.

A

000m

R

B

. .

A ........, .c>KIAN T 1.-- r

ed--.....

w1th

0 Istostromes

m r

.........

v

.

beds mass

iv

e, pale ,

UPPER

w

o

~ PODOLSK IAN ~ ~ ~ ~ ~ ~ ; a s t i c

V

KASH IRSKIAN c:C> :::: _ _

: - : - - : - - . - - : - . . . - _

MEMBER

O;-LO:;:::WE=R:-- - - t 8

MEMBER

£ .

N

I

F

E

R

o

U

S

1

-- - - grey, bioclastic 1s t.

V f R E I S K ~ A N ?

with chert shales

BASHKIRIAN

massive, grey

li mestones

VALDETEJA Fm.

-III

O ,C

1000m

o ? ?

1 111 11

SERPUKHOVIAN -

black,thlnlybedded

BARCALiANTE

= 6

1st

. often laminated

Fm.

l J ~

- - chert

.O.EVONIAN

ORDOVICI

AN

Om

CAMBRIAN

VISEAN

TOURNAISIAN

t AMt;NN IAN..

AREN I

GIA

N?

TREMADOC I

AN

POTSDAM IAN?

ACAD IAN

Province and to the south by the Pisuerga - Carrion

Province .

At

the beginning of the Carboniferous

Period

uniform

marine condit ions e x i s t ed throughout the

Cantabrian

Zone resu l t ing

in the

widespread

deposit ion of a sequenc e of

red ,

nodular

l i mestone s

with

chert and th in shales , the

Genicera

Formation (Fig. 2) .

At

the beginning of

the Namurian the black, Cal iza de Montana was

deposited but in the

Ponga

Nappe and

Pisuerga

- Carrion Pro v inces

sedimenta t ion

became

increa

s ing ly

dominated

by t e r r igenous

material .

Carbonate deposit ion continued in the

Picos

de

Eu

ropa

Province however, with

intraformational

brecc ias

and slump de pos i t s the only indications

of

tectonic

disturbances .

94

r nOdullr 1st.

k-wPcite

pa e limestones

calc. sandstones

quar

tz

i

te

BARRIOS Fm.

sandstones ha les

OVILLE

Fm.

The ca us e

of

the

i n s t a b i l i t y

was the

no r th - western movement

of the

African co

ntin

e n ta l

p late toward

s

the

Eur-Americ a n p

l a t

e . The

V

ariscan

or ogenic be l t

developed along

the

l ine

of

col l i s ion . The

de ta i l s of the Variscan Orogeny

are

complicated and

seve

r a l mode l s ha ve been

propo

sed

(see Windley, 1984

for

a review) .

The

main per iod

of de

f

ormation in

t he

Cantabrian

Zone

occurred during the

Astur

i a n Phas e

of the Variscan orogeny, a t the end o f the

Moscov ian (about 295 Ma). Sediments we re thru s t

in t o

the

core of a de veloping a rc uate mount

ain

chain i.

e .

into th

e Can

ta

b r ian

Zone) from

t

he

northwe s t w s t and so u t

hwes

t . The p icos de

Europa Province

was

the l a s t to be af f ected

by

deformation because

of

it s pos i t ion a t t

he

View t o

the north fr o

m

base

c

m

p

a t Vegaban

o .

The escarpment on

the r ig h t i s

th

e s o

uthern

edge

of t he

Fronta l Nappe

Unit .

The

main peak

i s Pena Santa and

the

en

t rance of M2 i s l oca te d c l o se

to

the

i

mp

o s ing so u t h fa ce. The

bottom

o f M2

i s

a

bou

t

300

metres

below t

he le v

e l o f

base

c

amp.

8/21/2019 BCRA 14-3-1987


nor theas tern

margin

of the Cantabrian Zone,

away

from

the encroaching

th rust sheets . The

main

phase

of

deformation within the Picos

de

Europa

Province began in the Kasimo

v i an and cont inued

unt i l the Autunian Martinez-Garcia,

1981).

Palaeozoic sediments,

predominantly

Carboniferous

carbonates , were th rus t

from the

north

- nor theas t

to

the

south-southwest

over the

Pisuerga-Carr ion

Province

Maas,

1976;

Marquinez,

1978;

Far ias ,

1982). The

massively bedded l imestones moved

as

a

ser ies

of competent sheets with very l i t t l e

in ternal

folding. Each

th rust

sheet

was pushed

over the one

emplaced

illUTlediately

to

the south

which

grea t ly increased the

t o t a l

thickness

of

the

carbonate

sequence.

Maas,

1976),

es t ina ted

that

the l imestones moved

a t l eas t

20km

because

the

Pisuerga-Carr ion fac ies

is

so

d i f fe ren t

from

that

of the Picos de Europa

.

Post-Var iscan Events

During the Permian, Mesozoic and ear ly

Ter t i a ry ,

a

sedimentary sequence was deposi ted

over the

picos de

Europa

l imestones.

This

is

inferred

by the presence of Permian deposi ts

5

25 m

2000

5

Los Moled izos

I

Pei\a Santa Nappe Unit

Frontal Nappe Unit

C

r.:' ' 71 Vega

Huerta -

A

Carbanal Series

R

B

o

N

I

F

E

R

o

U

S

Picoa

de

Europa

m

.

D

Upper Member

Lower Member

Callza de Montana

Valdeteja Fm.

Barcaliente Fm.

Genicera Fm.

Upper Devonian

Pisuerga Carrion

Province

LA :A Devono-Carboniferous

Black shales sandstones

a al,nm .raltA

Figure ) The Geology of Vega of Huerta .

Largely

a f t e r

Farias

1982).

M2 I8 , I 34, A15 and A10) are the major

caves

discovered

to date

within down-faulted regions, as

a t

Sot res

Marquinez, 1978; Smart, 1984), and

by

the present

drainage

pat te rn

which

must

have

been superimposed

on

the Carboniferous l imestones

from a

post -

V

ar i scan cover.

Post -Var iscan deformation

was

extensional

in

charac te r with fault-bounded basins

cont ro l l ing

the pat te rn of sedimentat ion. In l a te Eocene and

Miocene t imes,

however,

nor thern

Spain experienced

a fur ther per iod

of north-south compression, t h i s

t ime

re la ted

to

the Pyrenean Orogeny. The

compression ended about 38

Ma

ago

and

probably

in i t i a ted the upl i f t of the picos de Europa

massi fs

Le Pichon Sibuet , 1971; Boi l lo t

Depeuble 1982;

Vegas

Banda, 1982). The Pyrenean

deformation

af fec ted

the Carboniferous

l imestones

l a rge ly by re -ac t iva t ing exi s t ing

Variscan

and

post -Variscan f rac tures . These are

par t i cu la r ly

important cont ro ls on cave development

because

they

are

l ess

well

sea led

by veining

and

provide

the most open routes for

water in to

the

massif .

N P P E

Perla

Santa N

I

depth

o

500

l000m

Figure 4 Diagramatic sect ion

through

the

Fronta l Nappe Unit .

o

2km

,

95

8/21/2019 BCRA 14-3-1987


The

Geology of

Vega Huerta

The Western Massif has been mapped

by Far ias

(1982) al though only the

eastern

par t

of

the

range

i s

covered (Figs. 3 and 4) . Far ias

ident i f ied

four

major Variscan t h ru s t

sheets (or

nappes).

Each i s over thrus t

on

to the

one immediately

to

the south

so t ha t

the dip of the th rus t s genera l ly

increases

from south

to north. Vega

Huerta

i s

located on the

f i r s t

(most southerly) nappe to be

emplaced which i s

ca l led the

Frontal

Nappe

Unit.

The

highest mountain

in the

massif , Pena Santa, i s

located

a t

the

f ront

of

the

second nappe,

the

Pena

Santa Nappe

Unit .

Within

the Frontal Nappe Unit

the dips

of

the beds

i s

genera l ly 0° to 20°

al though

the

dip reaches 30° a t

the

southern

escarpment

and s teeper dips

occur

near major

fau l t s .

This cont ras t s with the area to the

nor th

and around

Treviso

where the dip

is

ra re ly less

than

45°.

All the l imestones of

the

Frontal

Nappe Unit

are

potent ia l ly cavernous

al though the

Vega

Huerta-Carbanal Ser ies which has a

s igni f icant

shale content , contains only collapsed shaf t s a t

the

surface with

no negotiable entrances

found to

date .

Erosion

of

t h i s formation has produced

windows

in to

the under lying Picos de Europa

Limestone

Formation

within which the

caves

I8 and

I34 are

developed (Fig. 3).

Several

major post -Var iscan

fau l t s

cut

the

Fronta l

Nappe

Unit.

These

t rend east-west

and

southeast-northwest and three come together j u s t

to the

south of

the

entrance of

M2 There are

also

many

l e s s per s i s t en t

f rac tures , not

shown

in

Fig.

3, but

which are

clear ly

important cont ro ls

on cave development.

CAVE

DEVELOPMENT

WITH SPECIAL

REFERENCE

TO

M2

The

Frontal

Nappe Unit d i f fe r s

from

the

nappes emplaced

l a te r

in

the region north of Pena

Santa)

in

tha t the l imestones r es t on the rocks

of

the

Pisuerga-Carr ion

Province.

These

are l a rge ly

black

shales ,

sandstones

and conglomerates and

therefore

provide

an

impermeable

basement

to the

l imestones. The Rio Cares

has

cut complete ly

through the Frontal

Nappe

and now f lows some 300

metres below the base of the

l imestones.

Therefore

vadose

condit ions

must have

exis ted

throughout

the Frontal

Nappe

for a

considerable

per iod

(Fig. 4). At the presen t t ime, water

draining to the base

of

the Frontal Nappe can be

96

expected to flow down the dip of the basal

th rust ,

approximately towards 010°. In par t of the area

explored

by

OUCC between Pefia Santa and Ar io,

there appears to

be

be a perched water

- tab le

which

causes the caves

to

terminate in sumps

a t

an

a l t i t ude

of

about

1300 metres (Roberts,

1986).

The

existence

of

the

M2

cave system, which

descends almost

to the

base

of

the

l imestones,

demonstrate the absence

of

a water

- tab le

a t 1300

metres in the

Frontal Nappe.

As

Smart (1984) commented, cave development

would

have

been

in i t i a ted

with

the

exposure of

the

Carboniferous

l imestones from

beneath the

post -Var iscan

sedimentary

cover .

At

present there

i s l i t t l e informat ion to date

t h i s

event but

erosion of the post -Var iscan

cover

would have both

accompanied and

followed

the

up l i f t

of the Picos

de

Europa. Major cave systems probably developed

while the

post -Var iscan

cover

supported large

drainage

catchments

and concentrated water in to a

few

s inks .

Such

systems

wil l

be termed

cover

systems in t h i s paper. As the cover was eroded

however,

an

inc reas ing

number

of

smaller

catchments

would have

developed as

new

f rac tures

were exploi ted. Many of the exis t ing sinks and

cave passages

would

have been

abandoned.

The glac ia t ion

of the

Picos de Europa

dur ing

the Quaternary

had a

number of

important

af fec t s .

Glacial per iods are charac ter i sed by both erosion

and

deposi t ion

so

many

establ ished

depressions

were undoubtedly choked with

debris .

However,

cave systems were probably formed a t t h i s new time

by

g lac i a l meltwater

sinking

in to

the l imestones

as

proposed

by Smart (1986) fo r many

caves in

the

Eastern Massif. Essent ia l ly the glac ie r s provided

large

catchments

and concentrated water in to

cer ta in s inks . The Picos

de

Europa must have been

a region of

high re l i e f

before

the Quaternary

glac ia t ions ,

so the

majori ty of the post -Var iscan

cover

may have

been

eroded (and major

caves

developed)

before

the onset

of

glacial

condi t ions .

Since the r e t r ea t

of

the glac ie r s the

surface

of the Frontal

Nappe has probably changed

very

l i t t l e

except

that

most of

the larger shaf t s

have

gradual ly choked with scree . At present there are

hundreds of

shaf t s developed in

the surface of the

Frontal Nappe but very few lead to negotiable cave

passages.

The

l a rge

number

of

f rac tures

means

that

each one has only a

small catchment

area,

of ten receiving water only from the

snow-plug

within i t . These f lows are

insuff ic ien t

to keep

T

op camp below

the sou th

face of

Pena Santa The main face

i s

composed

of l im s ton s from

the

Barcaliente

Formation which are

th ru s t over the

disrupted Vega

Huerta

Carbanal Series

8/21/2019 BCRA 14-3-1987


cave passages

c lear

of

the debr i s

loosened

by

f reeze- thaw ac t ion in the winter months. The

ent r ances

to

the

cove r sys t ems are almost

cer ta in ly eroded

or

buried, al though deeper

passages may have been

preserved.

The

maj

or cave

systems

discovered to date

are

usua l ly those

with

entrance

passages which

ha ve be en crea ted

or

enlarged by

me

l twater and wh ich have then escaped

i n f i l l bec

ause

of a f o r tunate loca t ion or because

t hey have high- leve l

routes

over the blockages

crea ted by debr i s from the

surface Laverty

and

Senior ,

1981).

THE M2 CAVE

SYSTEM

The

entrance of M2 i s located about 400

metres

west o f

Vega

Huerta beneath

the

imposing

south face of Pena Santa . In th i s loca t ion ice

accumulati on was probably le ss extensive

and

le ss

prolonged

th an on the

nor th- fac ing

slopes

of

the

areas explored by aucc and LUSS. Never theless,

remnants of terminal mor

a ines

ex i s t in the major

depressions

south of Pena Santa

and

glacia l

rounding is

evi

d ent

in the

l imestone

r idge

conta in ing the entrance

of M2 The

entrance

is

on

a minor

fau l t

dipping

65°

to 020° which forms a

l i n e a r

depression

along

t he long axis of

the

r idge

. A poss ib le explana t ion f o r t h i s loca t ion

i s that crevassing

occurred where the

ice rode

over the r idge

so

that

supra-g lac ia l

and

en-g lac ia l meltwater was direc ted to the base of

the

glacier , where t then sank i n to

the

M2 cave.

Looking

up

the

main

phreat ic

conduit

tow ards Road

to

Nowhere The passage s

about

15

metres

n

diameter

and has

a gradient

of

about 15 .

97

The phreat ic tube jus t downstream from Ken Hil l Gallery.

Here t is mostly developed above a bedding plane

within the

Valdeteja Formation ownward solut ion was

l imited

y

sediments

Wi th in

the cave ind iv idua l passages

can

be

seen

to explo i t var ious geological cont ro ls but by

fa r

the

most important are f au l t s . The term

f a u l t i s used in the fo l lowing sec t ions only

where movement

can

be

c le a r ly

detec ted .

The

term

f r ac tu re

i s

used

in

a general sense to describe

a l l planar f is sures

and

includes f au l t s ,

jo in ts

and

maj

or

bedding

planes. In

general ,

the

orien ta t ion

of the

cave passages

Fig.

5)

suggests

that

the

overa l l t rend

i s cont ro l led by

two se t s

of

f rac tures

which

s t r i k e approximately 020°

and

140°. Fractures with th i s o r ie n ta t io n are mostly

re la ted

to the per iod

of extensional deformation

which

followed

the

Variscan

orogeny.

These

f rac tures are

the most

important

cont ro ls

below

Non- Stop Drop.

Above t h i s

point the

cave i s

developed down

the dip

of f rac tures s t r ik ing

approximately 280°.

The

Upper Sect ion of M2

The

entrance

f au l t runs

in to

a more prominent

reverse

fau l t

dipping

65°

to

040° which i s f i r s t

encountered

j u s t

beyond

No

Eighth Fig .

5

and

Fig.

6).

This second fau l t

def ines the roof of the

pi tches between Watford Gap

and

Ivan s

Other

ar i face .

The

cave c le a r ly developed down

the

dip

of

the

fau l t plane

and

a small

phrea t ic

tube i s

preserved

a t the head

of Watford Gap, i t s t rend

determined

by

a small

phrea t ic

tube which

approximately follows

the

s t r i k e of

another

f rac ture . The

Amapolo Series

i s probably

a

separate system which has been in te rsec ted by

down-cutting

of the

canyon in the

main

cave.

Scal lops

in

the

Amapolo Series ind ica te a flow

direc t ion f r om we

s t

to eas t .

The

pi tches developed along the second

fau l t

plane

are l inked

by ramps

which

r i se

from the base

of each pi tch to the head of the next . Such ramps

are corrunon in M2 and are important fea tures of

many Picos

caves.

Fig . 7

surrunarises the

morphology of the ramps and

shows that

they are

vadose fea tures formed by

pi tch

re t rea t . They are

not

l i tho log ical ly

cont ro l led

and

are not segments

of phrea t ic

tubes in te rsec ted by vadose canyons.

Each

ramp

marks the foot

of

a

pi tch

where t

8/21/2019 BCRA 14-3-1987


\C

co

~

~ n t r a n

P2

~

::

..

r ~ · . · ·

Amapolo

Ser

ies

IP49

o

100

Ivan s Other

Orj

fice

200

Play the White

Man

300

~ ~ W e t a m p

p ~ V Y · R a m p

EXTENDED

EL

N

\Non-Stoporo

p

EVATION

o

50

100

200

metres

YUCPC SEll

1987

8CRA

58

Location

; . o

S8 W ,43°12 N

I

t i

tude

:

1980

metres

Vega

Huerta

Picos de

Europa

N. Spain

600

700

800

900

ooo

Entrance

Sump

.

o

50 100

200

M2

metres

POl Ode Cuetalbo)

650

c:d-

EI Gordo

P12

Highway Star

P34

Pitch

CROSS-SECTIONS 2 XMAIN SCALE

t 8 ~ t

• I. •

b e

e

f 9

cJ

iJ

[:Ii

k I

m

Choke

823

Figure

5 1987

Survey

of

M2

Multiple phreat ic

levels near

Ken

Hi l l

Gallery have been omitted for

c la r i ty

8/21/2019 BCRA 14-3-1987


re t rea ted headwards and cut downwards.

The

gradual decrease

in width

of both the ramp and the

canyon towards the base of the pi tch

may

be

evidence

of a decreases

in

discharge

over t ime

.

Measurements

of ramps

in two

caves,

M2 and Ml03,

have

revealed

mean gradients varying

between

30 °

and 34°

which suggested

that

the headward erosion

ra te i s more than 1.5

times t ha t

of

downcutting.

The larger

s ize of the pi tches compared to the

canyons demonstra tes the grea te r eros ive power of

fal l ing water compared

to

channel f low. The

canyons

are

of ten

too

narrow

to

t raverse

so

the

only negotiable route

through

these pi tch-ramp

systems

i s

usual ly

down

the pi tches then back

up

the ramps. The nature of erosion a t pi tches was

studied by Brucker

e t a l (1972)

who

measured

a

decrease in carbon-dioxide pressure

in water

as

it

fe l l down shaf t s . They

at t r ibu ted

th i s change to

de-gassing of

dissolved carbon-dioxide from the

th in

f i lm of

water flowing

down

the wal ls .

In

M2

a t

present , very

l i t t l e water fa l l s through the

pitch-ramp system .

In

normal summer

flow

condit ions only a th in

f i lm of

water flows down

the wal ls

and the wetted area

i s

r es t r ic ted to

a

small

sec t ion of

the shaf t below the i n l e t canyon.

In the highest flows

observed

to date the i n l e t

stream

disperses

into

a f ine spray as

it fa l l s

which causes an increase in the wetted ar ea on the

shaf t . I f the erosion ra te

i s

equal over the

whole

wetted

area ,

these observat ions

in fer that

the length of a

ramp depends

on

it s

age

whereas

the width of a ramp i s re la ted more

closely

to the

mean

discharge.

A

consequence

of

predominantly

low f lows

i s

t ha t

the wetted area on

a

pi tch

approaches t ha t of

a canyon

so that

pitch-ramp

systems tend to

evolve in to

s teep ,

narrow

canyons

(Fig

.

7b,

stage 3) .

Further work

i s

required to determine

the

signif icance

of

the ramp angle in par t i cu la r with

respect to

the angle

.of the dipping sur face ,

usual ly a

f au l t

plane, beneath which the ramps

develop. Potholes

commonly develop

a t the

base

of

the pi tches

because

the mechanical erosion of

swir l ing

sediment in the

plunge

pool d r i l l s

downwards a t

a

greater

ra te than

the ex i t

canyon

can

cut down (Ford,

1965). The

absence of

potholes

a t

the

base

of the pi tches in

pi tch

-ramp

systems

in fe rs

t ha t

the

ex i t

canyon

could always

cut down a t the

same

ra te

as the base

of

the

pi tch . The

simplest

explanat ion for t h i s

i s t ha t

there was

insuff ic ien t sediment

to

create

potholes , a conclusion which

suppor ts

the ear l ie r

suggest ion

that g lacial meltwater was probably

involved in the formation

of

the upper cave.

At

Motorway Services a small s tream enters

and f lows down the tor tuous Play the White Man

canyon. This passage

i s

similar to No Eighth

in

that

it

cu ts

through massively

bedded

l imestones

\

Pitch-ramp systems,

A-B,

C-O

F-G

developed

below faults

f

f

f .

- - - . r - . . . . - - - l . ~ , W a t f o r d

Gap

i original phreatiC or para-phreatic

p s s g e

f,

N.rrowcanyons,B-C

IE F

developed

on

I

nks

betwe.'I

pitch-remp

8yStMnL

,

Figure 6 Diag ramati c

se t i on o f

M between

the entrance and the

Dry Ramp.

99

a

concave

steps

/

b

---a .. - - - - - - -

,

,

\

Plana of P2 ramp

,

w -

,

,

Figure 7 a) The

general

morphology

of pitch-ramp

sy

stems.

b)

Sc hem

at i c development of ramps r l

and

r2 due

to

retreat

of pi tc hes

Pl and

P2

.

to l ink

one

ser ies of

fau l t -gu ided

pi tches with

another .

On such

l inks a

narrow and meandering

canyon

has

developed

because it

i s

so le ly

the

r esu l t

of canyon

incis ion.

No

pi tch has

re t rea ted

along

it to form

a

ramp (Fig. 6).

The

passage

enlarges

beyond

Play

the

White

Man as

the stream cascades

down

the dip

of a

f au l t

wi th

ca lc i t e

and

dolomite

veining ( the Wet Ramp)

to

sink

a t the

bottom in

an immature r i f t .

The

way on

i s

up the Dry Ramp, which

i s

between

2 and

3

metres wide and r i ses approximately

25

metres

to

the head of Non-Stop Drop.

The

Dry Ramp has also

formed

as

the

r esu l t

of

pi tch

r e t r ea t beneath yet

ano t her

fau l t

(Fig.

6)

but the ramp-forming pi tch

has been abandoned due to

a

capture

down the Wet

Ramp.

Below

Gareth s pi tch

the

cave changes

di rec t ion towards the northwest.

The

· new t rend i s

determined by a phrea t ic tube which r i ses and

f a l l s along another

f racture

inc l ined a t 70° to

040°.

A

deep vado

se canyon i s developed

below the

tube and

there

has been extensive modif icat ion

by

col lapse . Near

Chamber

of Orrors ,

for example,

two para l l e l

open

f ractures

and one

f i l l e d

with

ochre,

dolomite

and shale have contr ibuted to a

massive col lapse .

Below

Chamber

of Orrors the tube

cont inues

in the roof

of the passage and cons tant ly changes

di rec t ion and

cross

-s

ect ion

as it explo i t s

successi ve

f rac tures . There

are severa l p i tches

along

t h i s sec t ion of the

cave

but

only in the

case of Blind Pot , an obvious capture, does the

tube cont inue

over

the head of the pi tch .

At

the

other pi tches

the tube

must

have descended steeply

down the dip of the

control l ing

f rac ture .

Subsequent modificat ion of t h i s dip segment

under

vadose

condit ions has

removed

signs of

i t s

or ig ina l phrea t ic morphology, however

the tube

i s

always encountered again

a t

the base of the

pi tch

where

it

i s

developed

along

a

su i tab le f rac ture .

The Lower

Sect ion of

M2

At

the

Undescende

d Pitch the dimensions of

the

ca

ve

increase and

a

large meandering canyon

leads

to

the

junct ion with

the

lower sec t ion

of

M2

8/21/2019 BCRA 14-3-1987


a t Ken Hil l Gallery.

Remains

of phrea t ic tubes

more than 3 metres in diameter

revea l

the or ig in

o f

t h i s

sec t ion of

the cave

although there has

been

subs tant ia l modificat ion by col lapse . Two

major

phrea t ic

condui ts

meet a t Ken Hil l Gallery.

One

is the

Tea-Time

Ser ies

which has

suf fe red

two

captures , one

wi th in

Ken

Hi l l

Gallery

where the

Tea-Time Ser ies

enters

and another

a t the l imi t

of

'upstream' explora t ion . These two routes

join

and

then choke a t 823

metres.

The second major

phrea t ic

i n l e t

l i e s

in

the

roof of

the

canyon

between

The

Undescended Pi tch and Ken Hil l

Gallery. Near the bottom of Ken Hil l Gallery,

below the camp, a s tream jo ins the main gal le ry ,

f lowing

in

a meandering canyon. A small

phrea t ic

tube

in the roof

diverges

from

the

canyon

in an

upstream di rec t ion and terminates a t a pai r of

perched

sumps. In a downstream di rec t ion from Ken

Hil l Gallery, inc i s ion

by

the i n l e t s tream

gradual ly increases and eventual ly the l a rge

phrea t ic

passage in the

roof becomes

inaccessible .

The

s tream fa l l s down some wet pi tches (another

r e l a t ive ly recent

capture)

and eventual ly

disappears in to

Nicky's

Rif t .

Fortunately the

phreat ic passage can be

regained

above the Road to

Nowhere

af te r which it again

descends, clear ly

developed

down the

dip

near

and

a t

the base

of

the

Valdet

e ja

Forma t ion (Fig. 8) . The main phrea t ic

condui t

explo i t s both

fau l t s

and bedding

planes

but

it s

general t rend

seems to

be

cont ro l led by

the

bedding

of the massive Va lde te ja Formation.

In Ken Hil l Gallery and adjoining passages

there are considerable thicknesses of sediments

which

are

of

f ive

main types:

1. Unsorted,

l imestone boulder beds with

rounded

boulders up to 0.7 metres in length. These make

up the

largest

volume

of sediment.

2. Well

sorted,

well

rounded,

imbr icated

l imestone

pebble beds

.

3. Cross- laminated and cross bedded

sands

and

s i l t s

of

many colours with lenses

of unsor ted

sandy

gravels , the

mater ia l

loosely

termed g r i t

in

Fig. 9.

4. Well rounded, brown to red, gravels and sandy

gravels . These are an equivalent l i thology to the

lenses

of unsor ted,

sandy

gravel

( g r i t ) within

the cross bedded

sequence,

however they have a

much wider

dis t r ibu t ion .

5. Dark

brown

clay.

This i s found on top of the

cross-bedded sequence

and

i s

the

mater ia l

f i l l i ng

the f i s sure

in

Fig. 9.

Figure 9 Sketch of the sediments below the

climb to the Tea-Time Ser ies .

S

?

.

\

\

\

\

\

N

Figure 8

The

r e la t io n sh ip of

the ma

i n

phre a t i c condui t

t o

the

geology near Last Big Chamber showing how the chamber

develo ped

by

so lu tion a t the bottom o f a phre a t i c

lift.

The sandy

gravels

are remarkable sediments.

They are extremely porous and water moving

through

the deposi ts

has

produced spec tacu la r

leaching

pat te rns . Some beds are weakly cemented by

ca lc i t e

but most are unconsol idated and form

extensive ta lus

slopes

in

Ken Hil l Gallery.

The

cros s-b edded sequence of

s i l t s , sands

and

sandy-gravels appears to be the oldest

sediment

preserved and

represen ts

a l te rna t ing still and

energet ic periods.

Most

of

the

f ine

material ,

par t icu lar ly the

blue-grey s i l t , i s found a t

the

bottom

of the

sequence. The

sandy gravels are

found cut t ing across the beds beneath and become

more extensive towards the top

of

the sequence.

One lens

has clear ly been deposi ted in

a

channel

with a s teep bank showing

t ha t

the

phreat ic

condui t

was

a t l e s t

par t ly

drained

when

the l a s t

of the

coloured sediments were being deposi ted.

The

unsor ted

nature

of

the sandy

gravels and

the

chaotic inc lus ion of s i l t

i n t r ac l as t s

suggested

that

these mater ia l s were deposi ted

suddenly

a t

the

confluence of a

fas t - f lowing i n l e t

( the

Tea-Time Series?) with a r e l a t ive ly still

r iver

o r

lake

within

Ken

Hil l Gallery.

The Carboniferous succession

in

the Picos de

Cornion

(Fig. 2) contains no formation which could

have provided these

materials .

They are most

s imi lar to

the

so-cal led

'Red

Permian'

formations

l ike the Sotres Formation of the

Eastern Massif

(Marquines,

1978,

Martinez -Garc

i a , 1981, Smart,

1984) or the Labra Formation (Maas, 1976). In the

western Massif , a small

outcrop of

'Red Permian'

i s preserved

to

the

north

of

a mountain ca l led

Valdepino, about km west

of

Vegabano,

but

a t

an

a l t i t ude

of only 1400 metres.

The M

cave

sediments

provide the f i r s t d i rec t evidence of a

fissure

f i l led with

brown

alit

pale yellow laminated sands with

orange

silt

lam inae

ark brown yellow grit

coarse, red laminated sands red silt

pale brown y e l ~ o w ~ s ~ ~ n d ~

~ ~ ~ ~ ~ ~ ~ ~

~ ~ ~ i j ; I I I B l l i l l ~ ~ ~

ellow sil t

I

gM

brown sand

grey grit

cross- laminated red , U ' ~ _ _ ,--

coar grit

wit

leaching patterns

red sand

yellow sand

_

r t

fine

yellow sand brown

grit

2

o

metr

yellow sand

silt = = = = = ~ ~ ~ ~ i i ~ I I I ~ ~ I ~ i i ; ~ i ~ i ~ ~ ~ ~ ~ ~

oarse brown sand _ _ cross laminated. red

aand

laminated yellow brown silt - - -

.

__ fine red sand

pale

blue

-g rey silt

with some

~ ~ ~

. : .

brown laminae - :::::--:...: :

brown grit

load structures

100

8/21/2019 BCRA 14-3-1987


Leaching

patterns in u n s o r t ~

sa n

dy gravel

s , Ken

Hill Gallery.

Cross-bedded

sediments

below

the

climb

to the

Tea-Time

Series

in Ken Hill Gallery. The

photd

covers par t of the

area

shown in

Fig.

9. Numbers indicate se d iment sampling

s i t e s and

the c l a y f i l l e d f i s su r can be

seen near

no 18

The

tape

i s

1.5 metr

e s long.

101

post -Var iscan

sedimentary cover

a l t i t udes

in

the Western

Massif.

Since the lower

sec t ion

of

M descends from nor th to south the

source

of

the

cave sediments

may

have

been

to the

nor th

of

the

Pena Santa r idge.

The dark brown clay which covers

the

coloured

sediments

in

Ken Hil l Gallery i s

also

found

between the

sediments

and the wal l

of

the

cave,

and

i n f i l l i ng f i s sures .

This

shows tha t

the

coloured

sediment dr ied ,

shrank

and

cracked

before

being

re-submerged

a t

the

t ime

the

dark-brown

clay

was deposi ted.

The

rounded l imestone

pebble

beds and

the

boulder beds

appear

to pos t -da te the coloured

arenaceous sedimen ts. This re la t ionship i s most

c lear ly

seen in the

Last Big

Chamber

where the

pebble

beds can

be

seen

on top of paral lel-bedded

yellow s i l t s and f ine sands.

Discussion

Passages with a phrea t ic morphology

occur

a t

-100

metres

(Watford Gap), -140 metres

(Amapolo

Ser ies ) ,

between -425

metres and

-575

metres (base

of

Gareth s pi tch

to Undescended Pi tch) and

within

the lower sec t ion of

the

cave (-700 to

-986m

) .

The re la t ionship

between these

passages i s not

c lear but, as mentioned previously,

the

Amapolo

Series

i s

probably

a separa te

development. The

small

s ize

of the

high- level

tubes suggests

t ha t

the

upper par t

of M did

not

or ig ina te

as

a cover

system I f the

highest

phrea t ic

passages are

in te rpre ted as having

formed

beneath

a

regional

water - table then

base

l eve l in the Rio

Cares

and

Rio

Dobra

must

have been

some

1100 metres higher

than

t is

today.

Higher

base l eve l s have been

proposed to explain

the

presence

of phrea t ic

passages a t high a l t

udes

wi th in

Pozu del

Xi tu

(Laverty

and Senior , 1981) and

Sima

56 (Smart,

1984, and 1985).

The

fac t tha t

the

phrea t ic tube

in

the

upper sec t ion of M

exploi t ed

f ractures

of

var ious

or ienta t ions

en- route to the

lower par t of

the

cave shows tha t

the

loca l

hydraul ic

gradient

a t

the

t ime was towards

the

lower cave

and

not

towards e i ther

of the

gorges. Therefore

the

lower

par t of M must be a t

l eas t

as

old

as

the upper

cave.

I f

upper

and

lower

cave

or i g i na l l y formed

as

par t

of the

same

phrea t ic

system,

M

i s an

example of a bathyphrea t ic

cave (Ford

and Ewers,

1978). These authors

descr ibe

such

caves

forming

where

groundwater

hydraul ic

gradients are s teep

and the

water t ab le remains high because

f i r s t l y

8/21/2019 BCRA 14-3-1987


the

f i s sure

network

is immature

and secondly

the

dis tance

between

point of

inf lux and the

resurgence i s grea t .

This in terpretat ion

impl ies

that

the

main phreat ic tube in the lower cave must

r i se

again,

perhaps

by

as much as 500

metres,

to

it s

former

resurgence

l eve l

. According to

the

model of

Ford

and Ewers

(1978)

bathyphreat ic

caves

evolve

towards deep-phrea t ic then water - tab le

caves as the f i s sure network matures and

shorter

routes

are exploi ted

between the

point of inf lux

and

the

resurgence.

Such

passages are not seen

in

the upper par t

of

M2 and the

single

phrea t ic tube

which cont ro ls the t rend

of

the upper cave shows

no s igni f icant

inc rease in

s ize

with depth. These

observat ions suggest

that phreat ic

condit ions

exis ted for about the

same

length

of

t ime

throughout the whole

of

the upper cave. The upper

cave

was

poss ib ly

abandoned a t an ear ly stage then

re -ac t iva ted some t ime l a te r under

vadose

condi t ions .

However, the

e a r l i e s t

vadose canyon in the

upper cave

i s

everywhere developed below the small

phreat ic tube. I t has not

explo i ted o ther rou tes

so the period of abandonment was

not

long enough

to al low a

s igni f icant

increase in the

permeabi l i

ty

of

the f i s sure network

by

the t ime

vadose condit ions become establ ished. Within the

bathyphreat ic model, a very rapid drop

in base

level

seems

to

be

required

to

explain these

observat ions.

The phrea t ic morphology

of

the

e a r l i e s t

passage in the upper sec t ion

of

M2 may, however,

be

an

extreme example of a

para-phrea t i c passage

(Tratman, 1957).

A para-phrea t ic

passage

i s one

formed where l a rge volumes of water dra in through

a network

of

low permeabil i ty f i s sures within the

vadose zone. Local

phrea t ic

condit ions would

have

exis ted as

long

as

the

volume of

water

exceeded

the capac i ty

of

the passage.

I f

an immature

f i s sure network i s assumed, such condit ions may

have

exis ted

long

enough for phreat ic

tubes to

have formed on the more hor izonta l sec t ions of the

immature cave. Vadose condit ions would have

become es tabl i shed

f i r s t

on the more ver t i ca l

segments where

capac i ty to t ransmit

water was

grea te r . This

model i s proposed

because

it would

explain the

morphology

of the

phrea t ic

passage

in

the upper cave without invoking a sudden

drop

in

base l eve l . The regional water t ab le could

have

been

f a i r ly

s tab le , probably close

to

the level

of

Ken Hil l Gallery, a t the t ime the upper cave

was

in i t i a ted .

Both models require an immature f i s sure

network but the condit ions in the

aqui fer

are

Feldspathic conglomerates of the

Red

Permian north of

Valdepino.

Rocks

l ike

these

are

the l ike ly source for the

coloured

sediments

in M2

qui te d i f fe ren t ,

phrea t ic

in the

case of the

bathyphreat ic model

and vadose in the

para-phrea t ic model. At present there

is

insuff ic ien t

evidence to prove

or

disprove e i ther

of

these in terpretat ions .

Further explora t ion may

show

whether the phrea t ic r i se r

in fe r red

by the

bathyphrea t ic model ac tua l ly

ex i s t s .

The lower sec t ion

of

M2 was c lear ly a major

phrea t ic

condui t

and acted as the focus for

severa l subter ranean st reams.

The

large

passages

and

sediments

in

the

lower cave

may

be

remnants

of

a pre-g lac ia l cover -sys tem , formed wi th in the

phrea t ic

zone, and

explo i ted by

caves

in i t i a ted

under

g lac i a l

condit ions.

Torca de Urr ie l lo in

the Central Massif appears to be

similar

to M

although the author has no personal exper ience

of

the cave. From

the publ ished

survey,

the

upper

sec t ion of

Torca

de

Urr ie l lo

has

developed

down

the general dip of the beds and major s t ruc tures .

This sec t ion

of

cave l inks in to a much l a rger , and

probably much

older

development or iented along

the

s t r ike .

I t may not be coincidence

that the lower

sec t ions of M2 and Torca de Urr ie l lo are both

a t

s imi lar

a l t i t udes .

102

The

t rend of the

lower passages in M2 gives

no

clue to

the

loca t ion of the ancient

r i v e r s

resurgence

but

presumably

it

was

in

the Rio Cares

or

Rio

Dobra. The s tream which enters the

lower

cave

a t

the

bottom

of

Ken

Hil l

Gallery s inks

in to

Nicky s Rif t and f luoresce in dye has proved

t ha t

the

same water

emerges

again near the terminal

sump . From there

the

water has

been

t raced to

the

canal

de

Capozo some

4km

to the

eas t

and

approximately 350 metres lower than the terminal

sump (Lloyd,

pers . comm.).

The major

east-west

fau l t located j u s t

to

the

south

of M2 probably

conducts

the

water to t h i s

resurgence and may

prove to

be

a

key

fea ture

cont ro l l ing

the t rend of

the unexplored continuat ion

of

the main

phreat ic

passage.

Fur ther explorat ion i s

required to

determine the des t ina t ion

of

the

ancient

r iver

which formed the main

phreat ic

condui t .

The geological

sec t ion

through the

Frontal

Nappe (Fig. 4) shows t ha t perhaps 300 metres

of

limestone

may

ex i s t beneath the

sump

a t

-986

metres.

This in terpretat ion

i s

based on the fac t

that

the top of the

Barca l ien te

Formation

i s

seen

within the cave a t an a l t i t ude of about 1100

metres (-900 metres depth) while along the s t r ike

to

the east , in the Canal

de Capozo,

the

Barcal iente Formation reaches 400 metres

in

thickness

because

of

over thrus t ing (Fig. 3).

The

dip of the Barcal iente Formation in the lower

passages

i s

about 5°

to

the

south and it i s

probable

that

the

same

r es i s tan t

l eve l

i s

responsible for the terminat ion

of

the cave a t the

-823

metre

choke, in Nicky s Rif t and a t the f ina l

sump.

CONCLUSIONS

The

discussion

above

and

the

conclusions

summarised below are

based

on observat ions made

dur ing the 1986 and 1987 expedit ions to M2. I t i s

hoped

that

sediment analyses

wi l l

be completed

dur ing 1988 which wil l ident i fy the source rocks

with more

cer tain ty

and poss ib ly provide a da ta

for

the

depos i t s .

The prel iminary

conclusions

are:

1.

The

geology of the Frontal Nappe Unit provides

dif feren t condit ions for cave development compared

to

the

o ther

nappes because the

l imestones have an

impermeable basement.

2. The

ramps

found in

M2

and many

other picos

caves are vadose fea tures formed

by

pi tch- r e t r ea t .

Ramps form where sediment supply i s low which

prevents

potholing a t the base

of

the

pi tches.

Glacial meltwater i s proposed as the l i ke ly agent

responsible for

the

erosion of the pi tch-ramp

sec t ions .

3. The phrea t ic tube in the upper par t

of

M2

exploi ted various,

low permeabil i ty f i s sures to

t ransfer water

down the

loca l hydraul ic

grad ien t

into

the

lower

sec t ion of the cave. I f the upper

cave or ig ina tea beneath the regional water t ab le

the

whole

of

M2

i s an example of a bathyphreat ic

cave. The

phreat ic

tube in the upper cave may,

8/21/2019 BCRA 14-3-1987


The Last

Big Chamber.

The th in

beds

vi s ib le i n the far wal l

are

within the top p ar t of the

Barca

l

i ente Formation.

however , be a para - phrea t ic pa s sage fo r

me

d within

an

extens ive

vadose zone .

4 . T

he

lower sec t ion of M2 acted as a maj

or

phrea t ic

condu i t

an

d formed

before post

- Variscan

sed ime n

t s

w

er

e c omp l e te

ly

e r oded from

th

e

Ca rbon i f erous m

es tones

.

5.

The sedimen t s in Ke n Hil l Ga l l e ry

are

probably

der ived f r om Red P

erm

i an

fo

r m

a t i

ons now erode d

f rom most o f the Western Ma s

s i f .

ACKNOWLEDGEMENTS

I would

l i ke

to

thank the members of York

Universi ty

Cave and Pothole

Club and

the

Seccin de

Espeleologia

Ingenieros

Indus t r i a l e s who ca r r i ed

out the explora t ion

and

survey of the M2 cave system. Without t he i r s k i l l

t h i s

paper

could

not

have

been wr i t ten . Specia l thanks are due to a l l

those who helped with photography or car r ied sediment and

rock samples from the cave, to

Andy

Kemp for t h i s help in

surveying

ramps, to Katherine Force and Ivan Walton for

ass is tance

with

the sampling and photography

of

the

sediments

and

to

Enrique

Vulture Fdez.

Giber t

who shares

my

cur ios i ty and obtained useful maps and

air-photographs .

The

expedi t ion

could

not

have

taken

place

without the generous

suppor t from the

many

sponsors l i s t e d

below

and I am

par t i cu la r ly gra t e fu l to the Royal Geographical Socie ty for

both

f in an c ia l

suppor t and helpfu l advice . Fi nally I

would

l ik e to thank

Dr . Peter

s ~ r t

for hi s c r i t i c a l

comments

and

enthusiastic

encouragement.

The expedit ion sponsors were: Alpinex , Caravan, Ever

Ready, Gil l c hr i s t Foundation , Lucy Foods, Lyon Equipment,

Maxiprint, Mornflake, Phoenix

,

Robert Saunders, Royal

Geographical

Socie ty ,

Springlow Foods

,

Spor ts

Council

/

Ghar

Para

u Foundation, Tate Ly le Twinings , Universi ty of

York

A th le t i c Union, The

Vice-Chancellor

Universi ty of York,

Wilderness Ways.

REF

ERENCES

Benoi t

P . 1985. Ed .

Les Picas

de

Europa. Spelunca

Supplement. N o . 9 , pp . 35 - 46.

Bo i l lo t ,

G., Dupeuble,

P.A.

and Malod, J . 1979.

subduction and tectonics on the

continental

margin

off nor thern Spain

.

Marine

Geology. Vol. 32 ,

pp.53 - 70.

Brucker,

R.W.,

Hess

, J.W .

and White,

W

B. 1972.

Role of

Vertical Shafts in the

Movement

of Ground

Water

in

Carbonate Aquifers . Groundwater. Vol. 10, pp.6-14.

Compte,

P. 1959. Recherches sur l e s s t e r r

ains

anciens de

la

Cordil lere Cantabrique.

Mem

Ins t . Geol.

Min.

Espana . No. 60,

pp

. 1 -

440

.

Far ias ,

P. 1982.

La es t ru c tu r a d e l s ec to r Central

de

los

Picos de

Europa

. Trabajos

de

Geologia. Vol. 12,

pp. 63-72.

Ford,

D.C.

1965. Stream Potholes

as

Indicators of Eros ion

Phases in

Limestone Caves.

Nat ional

Speleologica l

Socie ty

Bu l le t in .

Vol. 27, pp.27-32.

103

Ford, D.C.

1978.

The

Devel

opment of Limestone Cave

Systems

in the Dimensions

of

Length and Depth.

Can.J .Ear th

Sci. Vol. 15, pp.1783-1798. .

Ju l ive r t , M 1978. Hercynian Orogeny and

Carboniferous

palaeogeography in northWestern Spain:

a

model of

deformat ion-sedimentat ion r e la t io n sh ip s . Zei t s chr i f t

der Deutschen Geologischen Gesel lschaf t . Vo l

. 129,

pp.565-593.

Laverty,

M

and Sen io r , K. J . 1981. The geology of the

Aria ar

ea and cave

development

in Pozu d e l Xitu.

Proceedi

ng

of the

Oxford Universi ty Cave Club. No.

10, pp . 41-50 .

Le

Pichon,

X. and

Sibuet

, J . C. 1971. Western extensio n

of

boundary between

E

uropean and Iber ian p lates dur i

ng

the Pyrenean orogeny. Ea

r

th and

Pl anetary

Science

Letters . Vo l . 12, pp.82-88 .

Maas , K.

1976.

The

geology of

Liebana,

Cantabrian

Mountains ,

Spain

; deposi t ion

and

deformation in a

f lysch

area.

Leidse Geologische

Mededelinge

n.

Vol.

49,

pp.379-465.

Marquihez,

J .

1978. Estudio geologico del

sector SE de

los

Picos de

Europa (Cordil lera

Can

t ab r ica ,

W de

Espana). Trabajos

de

Geologia.

Vol. 10, pp.295 - 315 .

Martinez Garcia, E. 1981 . El Paleozoico de la Zona

Cantabrica Oriental (Noroeste de Espana. Trabajos de

Geologia. Vol. 22, pp.55 - 56.

Roberts

, S. 1986.

Combined

Survey of

the

Top Camp Caves.

Proceedings of the

Oxford Univers i ty

Cave

Club. No.

12,

pp . 55 - 56.

Sef ton M

1984.

Cave

explorat ions

around Treviso, Picas

de

Europa, Northern

Spain.

Cave Sc ience . vol .

11,

pp.199 - 237.

Singleton, J . Ed.

1981.

Pozu del Xitu. Proceedings of

the

Oxford Universi ty

Cave Club. No.

10.

Smart , P.L.

1984.

The

geology,

geomorphology

and

speleogenesis

of

the eas t e

r n

m ass i f s Picos de

Europa,

Spain

.

Cave Science.

Vol. 11, pp

238

- 245 .

Smart,

P.L. 1985.

Cave

development

with

special re ference

to

Sima

56 in

Sima

56 . Univers idad P ol i t ecn ica de

Madrid, Madrid.

87

pp.

Smart, P .

L. 1986. Origin and

development of

g lac io -k ar s t

closed depress ions in

the Picos

de

Europa,

Spain.

Zei t s chr i f t fur Geomorphologie. Vol.

30,

pp 423-443.

Tratman, E. K. 1957. A nameless

stream: suggested

new

term

.

Cave

Res. Grp.

Newslet ter . No.6 , pp

. 68-69 .

Vegas,

R.

and Banda,

E. 1982.

Tecton ic

framework and

Pyrenean e v

olut ion

of

the Iber ian Peninsula. Earth

Evolution Sciences .

Vol . 4, pp . 320-343.

Windley, B.F. 1984. The

Evolving

Cont inents . Wiley,

Chichester , 2nd

ed . , 399 pp .

Received November 1987

Kevin

J Senior

39 Rockleigh

Road

Basset

Sou

thampton

SOl 7AQ

8/21/2019 BCRA 14-3-1987


8/21/2019 BCRA 14-3-1987


C VE

SCIENCE

Vol 14, No.3 ,

December 1987

Transact ions

of the Bri t i sh

Cave


B.C.R.A.

Cave

Science Symposium November

1987

Abstracts of papers presented the to

meeting

Manchester

Polytechnic,

November 1987

SUB WATER TABLE QUARRYING

AND AQUIFER DEROGATION

IN

THE MENDIP

HILLS

S.L.

Hobbs;

Universi ty

Geography Department,

Bris to l

In the

f i r s t County Development

Plan

for Somerset

the

Mendips

were

def ined

as

an area of Special

Landscape Value and quarrying

was

l imi ted to the

l ess aes the t i c eas te rn par t

of

the h i l l s . To

fur ther reduce t he i r visual impact

t

was

recommended that

quarr ies

proceed downwards

construct ive quarrying) ,

ra ther

than sideways

through

h i l l s

(destruct ive

quarrying) ,

with

pumping being used to keep workings dry where they

in te rcepted the

water t ab le . Where

pumping i s

required,

derogat ion

of

the

aqui fer may

occur

with the

resu l tan t

drying up of springs and supply

boreholes

in

the local i ty . The

problem

of

predict ing how much water wil l be

in te rcepted by

a

quarry is

complicated because kars t i f i ed

l imestone aqui fers of ten conta in so lu t iona l ly

developed

conduits which can t ransmit

r e l a t ive ly

l a rge volumes

of

water

qui te rapidly.

Water storage

and movement

in

the saturated

zone i s

being

examined by carrying out slug

and

dye

di lu t ion

t e s t s on boreholes

in

the area.

Further knowledge i s gained from pump t e s t s on

quarry

sumps and analys is

of

long

term

water

level

records. These t e s t s , along with examination

of

quarry faces

should

help to

predic t the

l ike l ihood

of

l a rge

condui ts being developed

and

thus help

to

e luc ida te

the long term ef fec t of quarry

dewatering.

GEOLOGY

AND

SPELEOLOGY

OF LOWER GLOMDAL NORWAY

Simon Bot t re l l ; Department of

Ear th Science,

Leeds

Universi ty

The lower par t

of Glomdal, Rana, N. Norway,

contains a l a rge number

of

phreat ic

cave segments.

Many

of

these

are

now dry foss i l

caves, but

others have

been

invaded by modern drainage

and,

in

some cases , ex tens ive ly modified.

The caves

are

formed in

highly deformed

marbles

and

the

geology of the area is

br i e f ly

described.

The

extent to

which

geological

fac tors

have

af fec ted

speleogensis has been inves t iga ted by a

combination

of

surface and underground geological

mapping. This demonstrates t ha t both l i tho logical

and s t ruc tu ra l

fea tures of the

marbles

have

exerted s trong cont ro ls on the genesis and

morphology of the caves, and the nature and

importance

of

these cont ro ls

i s

discussed.

The

caves of Lower Glomdal r e l a t e to an ancient

drainage

pat te rn and give a fragmentary picture of

the hydrology

of

the kars t area

a t

that time;

in terpretat ion

of the palaeohydrology

of

the area

can,

therefore,

on ly be

t e n t a t i

ve.

Scal lop

morphology

in the caves indicates

a predominant

sou ther ly palaeoflow. I t i s suggested

that

the

caves

car r ied

drainage from

an ancient proglacial

lake,

of which

Glomdalsvatnet

i s

a

modern

ves t ige.

The

present underground lake

ou t l e t

i s therefore

an

act ive

remnant

of

the ancient

system,

the

r es t

of

which has been

t runcated by

r i ver capture

fo l lo wing the erosion of the modern

r iver canyon.

105

SPELEOTHEM DATES AND RATES OF LANDSCAPE EVOLUTION

IN

THE ENGLISH MIDLANDS

Peter

J .

Rowe Timothy C.

Atkinson,

Austin; School of Environmental

Universi ty of East Anglia

Timothy

J .

Sciences,

Where

t

crosses

the Carboniferous

Limestone

outcrop in the south-west Peak

Dis t r i c t ,

the

River

Manifold has c u t a deep gorge.

G.T.

Warwick

ident i f i ed remnants of former val ley

f loors

and

suggested that

the

gorge

had

been

deepened in s ix

successive s tages . Short

caves in

the sides of

the

gorge

are t runcated

fragments

of

former

phrea t ic tubes in which

the

Manifold formerly

flowed

beneath

the

gorge f loor , as

t

does

in

dry

seasons today. We

have

used U-series dat ing

and

palaeomagnetic propert ies

to measure the age of

speleothems

in these

caves

and

in fer

the maximum

possible

value for the mean ra te

of

downcutting of

the gorge f loor . The oldes t and most important

s i t e ,

Elderbush Cave

contains

speleothems thought

to have been deposi ted during the Olduvai event

1.67-1.87Ma), implying t ha t

the present r e l ie f

has ev o

lved

over

a t imescale of

a t l eas t

2 mi ll on

years .

The

implicat ions of

t h i s

f inding for the

evolut ion

of

upland

Bri ta in w ll be

br ief ly

considered.

CARBONATE

DISSOLUTION

BY

GROUNDWATER MIXING

IN THE BLUE HOLES OF THE BAHAMAS

P.L.

Smart, F.

Whitaker,

R.J.

Palmer; Department

of Geography, Universi ty of

Bris to l

J.M. Dawans; Koninklijke / Shel l Explorat ie en

Produkt ie Laboratorium, Rijswijk,

Nether lands.

Extensive cavern systems Blue Holes) are

developed along

maj or

f ractures paral le l ing

the

bank margin on

South

Andros Is lands

in

the

Bahamas. Enhanced groundwater flow occurs in

these

cave

systems, and

also from them

in to the

adjacent carbonate aqui fer ,

giving

considerable

potent ia l for carbonate disso lu t ion

by

groundwater

mixing.

In - s i

tu

ob

se rva t ions

of

the

cave

wall

morphology, and petrographical

examination

of

wall-rock samples

show that

extensive and highly

pervasive

carbonate dissolut ion

i s

curren t l

y

occurr in g a t the base of the

f reshwater

l ens ,

where mixing

with

the

under lying

deep

sa l ine

groundwater occurs. Geochemical

studies indicate

that

the brackish waters from t h i s zone of mixing

are undersaturated with respect to ca lc i te ,

permit t ing l imestone dissolut ion a t an apparent ly

high rate

.

Signif icant

product ion of

carbon

dioxide and

hydrogen

sulphide by

bac te r i a l

ac t i v i ty also occurs in t h i s zone. Transmiss

ion

of

carbon

dioxide

in to

the under lying

deep sa l ine

groundwaters

by dispersion and

di f fus ion ,

may

also

allow disso lu t ion

of limestone by deep sa l ine

groundwaters. Carbonate disso lu t ion by

groundwater

mixing

i s

probably

a

s igni f icant

porosi ty-creat ing

process

in

emergent

carbonate

s t ruc tures .

8/21/2019 BCRA 14-3-1987


RADON LEVELS IN

PEAK

AND POOLE'S CAVERNS,

DERBYSHIRE

S .E. Fle tcher ,

J . Gunn; Geography Department,

Manchester

Po ly technic

D. Prime; Geography Department, Manchester

Universi ty

Uranium-bearing minerals

contain radium which

decays to

the

radioac t ive

gas

randon-222.

Radon

in

tu rn

decays through four

shor t - l ived

radon

daughters ,

to lead-210. High

concentrat ions of

radon and i t s

daughters may be present in poor ly

v

en t i l a t ed

mines and caves.

Radon

may also be

brought in to

such

areas in

solut ion

in water . I t

i s considered

t ha t the chief heal th hazard

i s from

radon

daughters

and there

i s

evidence

that

they

can give r i se to lung cancer .

Recent l eg i s l a t ion

requires areas

having

potent ia l radon daughter

alpha energy concentrat ions of 0.03 working

level

(WL) and 0.1

W

should be

designated

respect ive ly

as supervised and

cont ro l led

areas (a working

level

of

0.1 corresponds

to an

ac t iv i ty

concentrat ion of

3700

Bq / m' ).

Radon daughter alpha

energy

concentrat ions

in

the

sect ions of Peak and

Poole ' s

Cavern 's

which

are v i s i t ed

by

t ou r i s t s have been monitored

i r r egu la r ly

since

June 1987

using

both the

es tabl i shed

Kusnetz

method

and

a

'Radon

Sn i f f e r

wl:ich

has

the

advantage

of not requi r ing a mains

power

supply

but

the disadvantage

of

needing a

longer

count ing

time.

In

Poole s

Cavern

levels

have

f luctuated

both

in

t ime

and spa t i a l ly through

the cave, ranging from 0.17 - 0.98

WL.

Variat ions

in Peak Cavern have been much more dramatic with a

minimum of 0.005 and a maximum

of

9.1 WL The

or ig ins

of

the

radon and the causes of

the

var iab i l i ty are present ly under

inves t iga t ion .

I t should be

s t ressed t ha t even

the highest

of

these l eve l s poses no th rea t to t ou r i s t s

or

to

occasional

v i s i t o r s .

However, the r i sk

increases

wi th the number of

hours exposure and could

be

s igni f icant

for cave guides or

speleogists

making

frequent v i s i t s to high radon areas . Perhaps a

more

s igni f icant

problem i s

that unless radon

l eve l s

can

be reduced

or an

exemption

obtained

both caves

may

ul t imate ly

be

designated as

control led

areas .

At

bes t access

would

be

r es t r i c t ed ; a t worst

t

could be los t al together

which

would be

disastrous in

view

of t he i r

sc ien t i f i c

and

spor t ing value. Moreover, t seems

l i ke ly tha t

other

caves

ins ide and outside the

Peak

Dis t r i c t

may also have high

radon

l eve l s .

Hence, there

i s

an

urgent

need for a comprehensive

monitor ing

programme.

UK SPELEOTHEM GROWTH AND WORLD SEA-LEVEL

D. Gordon;

Department

Geography,

Bris to l

Universi ty

Calc i te

speleothem

growth in the UK i s thought to

occur predominantly dur ing

'warm'

in te r s tad ia l and

i n t e rg l ac i a l

periods.

Analysis

of

the

data from

uranium

ser ies determinat ions has ident i f ied the

presence and

ages of three Upper Pleis tocene

i n t e rg l ac i a l s and s ix in te r s tad ia l s .

In order

to

t e s t the assumption of

cl imat ic cont ro ls on

UK

speleothem

depos i t ion , t h i s record has

been

compared with the high sea stand record from

uranium

ser ies

dated

coral reef terraces in

t rop ical

and subt ropica l regions

of the world.

The

presence and ages of

eight

Pleis tocene

high

sea

level

stands

have been

ident i f ied .

Spec t ra l

ana lys i s

has demonstrated that between 70000 and

140000

BP

there i s a high cor re la t ion

between

the

UK

spleothem

and world cora l reef dated records.

( i . e. dur ing the

l a s t i n t e rg l ac i a l complex when

speleothems were growing in Bri t i sh caves

there

were also high sea

level

stands

in

the

t rop ical

and sub-tropical

regions of

the world).

PALEOMAGNETIC

STUDIES

OF

SEDIMENTS

FROM

PEAK CAVERN,

DERBYSHIRE

L.

Thistlewood, M.J.

Noel;

Department

of Geology,

Uni

ver s i ty of Sheff ie ld

T. D. Ford; Department of Geology, uni vers i ty of

Leices ter

To

date

a t o t a l of

four sediment

sec t ions from two

closely

re la ted phrea t ic tubes have been sampled.

Storage

t e s t s

ind ica te

t ha t on

the

whole

samples

car ry a s tab le na tura l remanent

magnet isat ion

(NRM). Alternat ing

f i e ld

demagnet isat ion a t 10 or

15 mT removes a smal l viscous

component

of

magnet isat ion leaving

NRM

vectors with normal

polar i ty

inc l ina t ions . These appear to

record

a

pseudo-per iodic secular

var ia t ion

of

the

geomagnetic f i e ld

agains t

depth in

the

sect ion.

Two

sediment

columns in the

same

var ia t ion and

that

of magnetic

suscep t ib i l i t y

with depth. I t

i s

now

hoped

to extend t h i s cor re la t ion to

the

palaeomagnetic

record obtained from a recent ly

sampled

sediment exposure in

a

nearby passage

developed a t

the

same

cave

l eve l . S t re tch ing

of

secular var ia t ion

records

wil l al low r e l a t ive

contemporary

sedimentat ion

ra tes to

be

calculated

for

dif feren t par t s of the cave

system.

AN

INVESTIGATION

OF

PULSE-WAVE

AND

DYE-TRACER

TECHNIQUES

USING

LABORATORY MODELS

P.

Hardwick,

J .

Gunn;

Department

of Environmental

Geographical

Studies ,

Manchester

Polytechnic

D.P.

Butcher; Department

of

Geographical

Sciences,

Huddersf ield Polytechnic

This

repor ts an

inves t iga t ion of

pulse -

wave and

t r acer techniques which are

used in

kars t

hydrogeomorphological

studies

to determine the

nature

of

inaccess ib le kars t drainage systems.

The

study

used

l abora tory

models which were based

upon

the

knowledge and

exper ience of

the

authors

and

on a

f i e ld

study of

the

P6-P8 condui t system

which forms par t of the Cast le ton kars t aqui fer

in

the White Peak

of

Derbyshire. The f i e ld study

i s

to

be document as a subsequent

paper .

Two cont ras t ing l abora tory

models were used.

Fi r s t ly ,

a

variable

storage model which

was used

to

inves t iga te

the

ef fec t s

of vadose storage on

dy e - t r acers /

pulse-wave

re la t ionships .

Secondly,

a

phrea t ic

zone

model,

which was used to

inves t iga te t racer behaviour

within

flooded

conduits .

Resul ts indicate that the

generat ion of

the

lag

ef fec t s

by

vadose

zone storage

inva l ida tes

the

assumption,

impl ic i t in the f i e ld technique,

t ha t

dyes

or

other

t racer

substances

used

to tag

pUlse

-

wave

experiments

wil l

remain

in

int imate

associa t ion

with

the

pulse-wave as

t i s

transmi

t ted through the vadose zone. Since

the

assumption also

forms the basis of phreas

volumetric

ca lcula t ion ,

these

r esu l t s show t ha t

pulse-wave techniques can produce highly

variable

resu l t s

dependant

upon

the nature

of

the vadose

zone

s torage . However fur ther

work

needs

to be

done

both in

the laboratory and

f i e ld

to quant i fy

the re la t ionship

between t racer

lag and s torage .

The phenomenon

of

upstream dye

migrat ion

was

observed in

the l abora tory phreas.

This affects

both the t iming and

dis t r ibu t ion

of

dye recorded

a t the resurgence. Leading to the

detect ion

of

bimodal t r acer chemographs

from a

single conduit ,

t h i s inval idates

the

assumption that

mult iple

t racer

peaks mul t i p l e hydrological

routeways.

More research both

in the

l abora tory

and f i e ld

i s

suggested, in

par t i cu la r

not ing

that the

technique

should

be redefined

in

condui ts

of known

morphology

before

being applied to

unknown

conduits .

106

8/21/2019 BCRA 14-3-1987


CAVE SCIENCE

Vol. 14,

No . 3, December

1987

Transact ions

of the

Br i t i sh Cave


The Crocodile

Caves of Ankarana:

Expedition

to

Northern

Madagascar

198

Edi ted

by

Jane WILSON

Abstract : Ankarana

is

a small but sc ien t i f i ca l ly

unique

l imestone

massi f ,

notable

for

i t s

huge

r iver caves where Nile Crocodi les f ind refuge from

drought and

hunters .

The

98km

of cave passages not only provide

some

spectacular caving but also comprise diverse habi ta t s for a

range

of

cave

fauna as

well

as important palaeontological and archaeological s i te s .

Although

the

amount

of new

passage found during

the

1986 expedit ion was

disappoint ing,

the

sc ien t i f i c

f indings are

highly

s igni f icant .

INTRODUCTION

The

Ankarana Massif

comprises

a small but

spectacular

area of

t s ingy (pinnacle

kars t ) which

i s

mentioned in

many

books

on

Madagascar. The

Guide

Bleu

for Madagascar

(Hachette

1968)

contains

a

deta i l ed descr ipt ion

of

the

sacred caves

and

mentions the

crocodiles

which l ive

in some

of

the

caves.

Blanc (1981)

wrote of crocodiles in the

subter ranean r ivers of Ankarana, and Paul ian

(1981) described the

caves a t

Ankarana

where

bats

were hunted

.

Blanc

(1984)

suggested

that

Ankarana

s

caves

are

v i r tua l ly

the l a s t st ronghold

for

Madagascar s

crocodi les . Ankarana has

been

a

Special Reserve for 30 years but l i t t l e sc ien t i f i c

work has been done there .

An

inventory

of Madagascar s caves

(Decary

and Keiner

1971) highlighted Ankarana

as an

in teres t ing

area

even

before the major f inds

of

t h i s decade.

Twenty years of met iculous searching

by

Jean Radofilao and

col leagues

(Ravelonanosy

and

Duflos 1965;

Duflos

1966,

1968

;

Radofilao

1977)

and two French expedit ions (Peyre e t

a l

1982,

1984) have resul ted in the discovery and survey

of

93km

of

cave

passage.

Ankarana

boasts

Madagascar s longest cave systems. Indeed

Ankarana s caves

must be amongst

the

most

extensive in Africa .

The

l a rge subter ranean

r i vers

of

Ankarana make the

caves

qui te

dif feren t

to most

t rop ical

caves and provide

a grea t

divers i ty

of

cave habi ta t s su i t ab le

for a

range

of

cave-adapted animals. The

r iver caves

form a

natural

i r r iga t ion system which al lows r i ch canopy

fores t s to f lour i sh

ins ide the

Massif whi ls t

outside

the

karst area the vegetat ion of the

savannah

becomes

parched dur

i ng the

s ix

month dry

season

.

The subterranean

r ivers ,

then

,

suppor t

the

r i ch fauna exis t ing above

ground

a t Ankarana

(see

Wilson e t

a l

1987)

as

well as ac t ing

as

a

reservoi r which ensures a

r e l i ab le water supply

to

local vi l lages

downstream. A

se lec t ion

of

papers

Transport

to our nkarana base

camp

was by

bul lock ca r t .

The

photo shows

a l i n e of forest

growing

above

a subterranean

river and

in

the background

the northern

nkarana

l imestones . Phi l Chapman)

107

have

been

publ ished on the

geology (Rossi ,

1973,

1974),

hydrology

(Rossi ,

1975,

1976a) and the

vegetat ion zoning

(Rossi

1976b)

a t

Ankarana .

The

1981 Southampton Universi ty Expedit ion

gave one of i t s leaders

a

t an ta l i z ing glimpse of

the

biological

and

speleologica l wealth hidden

within the

spec tacu la r

pinnacle

kars t

and

convinced

us

that

the area

was

worthy of more

thorough a t ten t ion .

The

resu l t ing

re turn

t r i p in

1986 showed Ankarana

to

be even r i cher than our

predic t ions (Chapman e t

a l

1987b).

HISTORY OF

EXPLORATION

Jean Radofilao

The

Ankarana massif contains an extensive and

complex system of cave passages, but most

of t he

cave entrances

are

di f f i cu l t to enter ,

and

even

to

f ind, due to col lapse .

There

are a few with huge

entrances,

access ib le

a t

the

base

of

the c l

i

f f

.

These

obvious

caves

(Andraf

iabe,

Ambatoma

nj amana

and

the

Mananjeba subter ranean

r iver)

have

bee

n

known by the

local people

for a long time; they

have explored them with palm l ea f torches.

Ins

i de

are

remains

of

f i r es and

f

i r ed

earthenware pot

s .

During

the

colonia l e ra , severa l Frenchmen

vis

i ted

these

caves, guided by the

loca l s ;

they

wrote

picturesque accounts for the lo c a l

paper

s , but

made

no

sc ien t i f i c observat ions. Ear ly work was

reviewed

by de Saint

Ours

(1959 ) .

The

f i r s t rea l

cave

explora t ion

was done in

1963

when

Jacques

de Saint Ours accompanied

by

G

Coquet surveyed 2. km in the Gro

t

te

d

An

d

ra f i abe

and

est imated

that

they had

explored

more than

5km. Almost every year

between

1

964

and 1

972

,

building upon informat ion

provided

by J

de

Saint

Ours ,

the wri te r ha

s

explored the caves

of

Ankarana, leading a

small team backed

by

th

e

Universi ty

of

Madagascar. I n i t i

a l l y

we

cont inued

the explorat ion of la Grot te d Andraf iabe and then

searched for

others . We explored

Antsatrabonko

,

8/21/2019 BCRA 14-3-1987


North adagascar

o 1

2 Km

11

108

The

dry pinnacle

kars t

tsingy i s

over 200m

above

the

surrounding

savannahs and canyon bottoms.

This photo,

taken

above the

Second River Cave

shows rich

gal lery

for s t in the

foreground

growing above a subterranean

r iver)

and the

Canyon Fores t ier

s t retching away to the

l f t

horizon Some

tsingy s covered

in scrubby, thorny, xerophytic

vegetation

Jane

Wilson)

8/21/2019 BCRA 14-3-1987


NK R N M SSIF

FIG 2

NORTH MAI l\GASCAR

o

2 4Km

~ ; ; ; ; ; a = s ; ; ; ; = = , , , 1

Mila in te ty ,

Ambiky,

and the big r iver

caves. When

these

lack of

money, the

was over

40km.

the s t a r t of Ambatoharanana

system and severa l

smal ler

explorat ions were stopped

by

to ta l passage length surveyed

After

1972 the wri te r cont inued the

explora t ion ,

usual ly

alone.

The l a rges t caves

surveyed dur ing t h i s period were

Andetobe,

Antsiroandoha and

Ampandriampanihy (north) . By

1980

the to ta l of explored passages had

reached

75km .

From 1981 severa l foreign expedit ions

came

to

Ankarana and made t he i r cont r ibut ions to

explorat ion (whils t

the

wri te r

continued alone

dur ing the r es t of the dry

season) .

These

were:

August

1981

a

team from Universi ty

of Par is VI

and a team from the Club

Martel of

Nice

and

Club Alpin

Francaise (Peyre e t a l

1981).

September 1981 a team from Southampton Univers

i t y

(Adamson e t a l 1984)

August 1982, 1984

and

1985 the team

f r

om Nice

returned (Peyre e t a1 1982, 1984,

1985)

August

- October

1986

a new

mul t i -d i sc ip l inary

team (cavers and sc i en t i s t s ) from

England th is

report)

As

a

r esu l t

of

a l l t h i s work the

t o t a l

passage

length explored a t Ankarana by the end of

1986

reached

98km.

THE GEOLOGY

AND

GEOMORPHOLOGY

OF ANKARANA

J .

Radof i lao, P.

Chapman,

D.

Checkley,

S. Hurd and

R.

Walters

The Ankarana massif

nor th of Ambilobe

and

is

s i tuated

about

75km

about

south

30km

of

109

. - ' . .

' - {

Legend

ave

River

Oiff

myon

Road

r

Track

- . / ~ = ~

Path

'-

-

Antsiranana (Diego Suarez) . I t

comprises

a

block

of

Middle

Jurassic

(Bajocian and

Bathonian)

limestone about

30km

long by 8km wide and

i s

bordered

on

the

north-west by a c l i f f face up to

200m high,

the Ankarana

Wall (Fig 2).

The

surface

of

the plateau

i s

very di f f i cu l t to c ross ,

as

it

comprises

unstable

blocks

of

needle- l ike

limestone pinnacles up to 20m

high,

separated by

deep

crevasses. This pinnacle

kars t ,

known

loca l ly as

t s ingy, suppor ts a unique xerophyt ic

vegetat ion .

The

massif is t raversed by

deep,

s t r a igh t , ve r t i ca l ly - walled, f au l t -

control led

canyons, the majori ty of which run

NW SE

almost

at r igh t

angles to

the

Ankarana Wall. They

are

nearly always l i t t e red with

huge

l imestone

b l

ocks

and

the

wider

canyons are

th ick ly

fores ted ,

especial ly

where

basa l t i c

lavas

have flowed in

to

form a

subs t ra te .

The

Caves

Passage dimensions vary

from

very small

crawling-sized

tubes

to the more

common

huge

corr idors whi

ch

reach 50m

width in

Grot te

d

Andraf iabe. Most of the biggest passages

run

NE-SW

para l l e l to

the

Ankarana cl i f f s .

Most

of the explored

passages

are act ive or

semi-act ive and s i tua ted close to the

l eve l

of the

foot

of the c l i f f .

Sumps

are qui te

common

and

some sec t ions

have navigable underground r iver s ,

the most notable

being

the

4km

boat t r ip necessary

to

t raverse

the length of the

River

Styx (G5).

Higher, l ess act ive passages

are

smal ler and

much

l ess

numerous. Access to some need

a r t i f i c i a l climbing

techniques. These higher

l eve ls are very well decorated

and

formations,

8/21/2019 BCRA 14-3-1987


debris ,

volcanic

ash

or

clay of ten block the

passage. La

Grot te

de Mila in te ty G6)

i s

probably

the bes t decora ted

cave.

Collapsed

Caves

A

huge

collapsed cave

was

discovered during

explorat ion

in

1986. This was made

up

of severa l

al igned

col lapses, with

l inking

res idua l

arches .

This fea ture

i s

very

d i f fe ren t

from the

f au l t

canyons since

the

col lapses are

al igned

over

very

shor t d is tances

and

are of ten s inuous ,

and

t he i r

depth

i s

clearlY much l ess than

that

of the

fau l t s . This

collapsed

cave

must have been huge

and probably

developed along a jo in t

higher than

the cur rent ly ac t ive streamway. Examination of

i r photos shows that there

i s

no other comparable

collapsed cave.

Very big collapsed chambers

or

i n t e r i o r

p l a i n s , bordered on a l l s ides by 100m sheer

c l i f f s , ex i s t in

the hear t of the plateau;

access

to

these

is genera l ly poss ib le only

through

the

caves. The bigges t

is

the Manily,

in

the southern

par t

of the

pla teau . I t

i s

oval : 800m x 600m

and

100m deep . In

the

cent re

i s

a

cur ious

ca lc i te

mound.

Severa l

large

col lapses lead

off

t and

we

suspect

t ha t

t i s a l l

that

remains

of a

vas t

ch a

mber with a volume around 50 mil l ion cubic

metres. The bottom of t h i s and other col lapses

always

contains an

accumulation

of lapiaz

blocks .

Some

are also fores ted

or occas iona l ly

marshy.

Some

are t raversed by

a r iver

or stream.

Tectonics

The Ankarana

Massif

probably behaves as a

r ig id

block. Ear th movements

in

the

Jurass ic

resul ted in

the

downward

displacement of

the

west

of the massi f .

This

continued to the Cretaceous

w ith a

genera l

t ipping of

the

massif to the west .

Movement

of the basement

beds

i s

thought

to have

occurred a t t h i s stage.

The f i r s t fau l t ing brought about

the sheer

200m Ankarana Wall border ing

the W

of

the mass i f .

Simultaneously , the massif

developed

a shallow

syncline

with

4

dips to

both eas t

and

west

with

an axis

approximately lkm

eas t

of

the c l i f f wall

.

This

per iod

of ear th

movement

also produced

a

ser ies

of

fau l t

scarps

approximately para l l e l

to

the

main

c l i f f s

.

Volcanic

ac t iv i ty

in

the

Pliocene formed the Montagne d

Ambre

volcano and

probably the canyons

. The

pr inc ipa l cave systems

were

t runcated by t h i s fau l t ing .

The

massif a t

t h i s stage was th rus t upwards

and the

NW SE

syncline

accentuated. These

movements

also led to

jo in t ing across the massif

in

the W

-

SE di rec t ion .

Cave

development

has f requent ly

followed

t h i s

the

dominant jo in t di rec t ion . All major canyons are

of t ec tonic

or ig in being al igned along major

jo in ts

or

f aul

t s

. None appear

to be due to the

Simon

Fowler entering

the

River

Styx

he

Russian

Mountains

inside

the

llkm

Grotte d Andrafiabe

(Paul

Stewart)

collapse of cave passages .

In

the nor th of Ankarana,

Quaternary

volcanic

ac t iv i ty

formed

a number

of small

ash cones

which

can

be

seen today. The

lava penetrated

some

of

the canyons

and

blocked

a good number of the

entrances in the

north.

In

places

the

lava has

backed up agains t the

wall .

Lava

flows

are

l ess

evident in the south.

Hydrology South of the Antenankarana River

The

Mananjeba

River (Fl-3)

cont inues

to flow

during

the

dry season,

providing

a r e l i ab le water

supply

for the

vi l l ages downstream

of Ankarana.

I t also floods frequently so the massif here

has

undergone s t rong

so lu t iona l

ac t iv i ty and

only

res idual l imestone

blocks

remain.

The f i r s t of

the three blocks i s the

largest .

All underground

flow in t h i s southern

par t ,

depends

purely

on the

Mananjeba

River.

I t

divides into

three main

branches,

c lose

to

the pla teau; the southern one

sk i r t s the southern

f lank

of the

th i rd block

and

does

not

sink

a t

a l l ;

the

middle

one

flows under

the

second

block for 2.7km

in

the very l a rge

passages of

the

Grot te d A mbatoharana

(Gl)

(which

i s

l8.2km

long) ; the nor thern

branch flows for

600m through

the

f i r s t block, in

par t

of

the

Grotte d Ampanriampanihy (Gla)

and

by jo in ing with

the water flowing from

the

north, form

the

Antenankarana River .

110

Hydrology North of the Antenankarana

River

The

main

kars t region

i s north of the

r iver .

Numerous

small r ivers

sink in

the massif

in

the

nor th and

east ,

but only

two flow in

the

dry

season; a)

Andranots is i loha

(meaning

r iver

without

a

s ta r t )

comes

from the Montagne d Ambre and

sinks

in to impenetra

ble boulders

in

the

extreme

north

of

the massif

.

Only

some

of the water disappears ,

the r es t re jo ins the Andranomandery which f l

ows

to

the

west

of

Ankarana.

b) The

Besaboba

runs

of f

the h i l l s

in

the eas t and divides

in to

numerous

branches, most of which disappear in to

ear th- f i l l ed hollows. The only two penet rab le

8/21/2019 BCRA 14-3-1987


AREAS

AROUND CAMP

o

s inks

(G17,

G18)

both go for

about

400m

to

a

sump,

and

are only

act ive

in

very

heavy

ra ins .

L i t t l e i s known

about

the hydrology

in

the

nor th

and

t

i s

di f f i cu l t to i n t e rp re t . However,

t appears

that

the

waters from

the

nor th form the

two large r ivers

and

the i r assoc ia ted caves in the

Canyon Fores t i e r

G14 and G1S) and in canyon CS.

We assume

that

these

appear

again as the r iver in

Campsi

te

Cave

G13) and

eventua l ly fur ther south

in

the

caves of Andetobe (G9),

Antsatrabonko

(G8),

and the

Andrafiabe

gorge

(C2).

These

then join

to

form s ingle , strong flowing

r iver ,

from the But tes

Chaumont, to the resurgence a t Grotte du Rois

(GS).

since

much

of

the

r iver

is calm,

deep

and

lO-20m wide,

t i s navigable .

Flow

ra tes

(measured a t the surface)

a t

both the But tes

Chaumont

and the

resurgence were approximately

lScums

/ s . I t

appears,

then,

t ha t

the

major

north-south

drainage

follows

the

base of

the

sYl lcline

about

1km

eas t

of

the

Ankarana

Wall

and

that

most

of

the

t r ibu ta r ies lead to t h i s master

system.

CAVE AND

ISOLATED

FOREST

EXPLORATION

R Walte r s

Our

explora t ion

ef fo r t s

concentra ted upon

searching unworked

areas of

the massif and

looking

for l inks between

mapped

sec t ions of

cave. The

areas neglec ted by

previous French expedit ions

were

largely in the

nor th and

eas t

of Ankarana

where

much

of

the

water sinks and very few caves

are known.

We

also

hoped t ha t

caves would provide

access

to many of the sunken

fores t s which

the

bio logi s t s were keen to inves t iga te .

Areas of s tudy and explora t ion were se lec ted

using

aer i a l photographs,

taken

in 1949

and

avai lable

from

FTM the

Government

map

shop

a t

Ampasamieto,

Antananarivo. We

paid

par t icu lar

a t ten t ion to obvious

r iver

s inks , s teep-wal led

depress ions

where erosion debris

was

unl ike ly to

have blocked cave entrances and deep gorges, since

gorges

elsewhere in

the massif cut major

cave

systems . The l a rge i so la ted fores t s of the

cent ra l region (E4-E6)

were

considered worthy of

the

bio logis ts a t ten t ion

and a

great

deal

of t ime

and ef for t

was

spent

t ry ing

to gain

access

to

th em.

Canyon Fores t i e r

Here

we wished

to complete the explorat ion

of

the vas t depress ion

beyond

the

French

discover ies

of River Caves

One and

Two (GlSa, G14a).

En route

both ca

v

es

were explored and

although no fur ther

discover ies were made, both yielded much to

i n t e res t

the

b i o l

o

gis t s . The

canyon

area

close

to

the River Cave

One

appeared to flood

to a

depth of

30m confirming

an observat ion

(Jean Radofilao,

pers. c om., 1986)

that

during the wet

season

2 m

storms

some

of

the

Ankarana c n y o n ~

are

completely

flooded.

A rock

shel ter

discovered between the Fir s t

and Second River Caves, la Grot te Trans Sept ,high

in the

north-east

wall of

Canyon Forest ier

yie lded

r i ch depos i t s

of

rodent bones.

River Cave

Two

was

only

about 3km

from

our

base

camp, but

took around

f ive hours to reach.

Thick

vegetat ion

and

lack of water made

explorat ion slow

and unpleasant .

Beyond River

Cave

Two, a

complete

search of the depress ion

revealed many

small caves; these

were

a l l

a t

relat ively high levels

in

the

wal ls of the

depress ion.

Areas

around

the Camp

The f i r s t

canyon (CS), j u s t north of

the

camp, conta ined two r ivers which had been

thoroughly explored

by the

French ( f igure 3).

Many

small caves

where found

in

the

c l i f f s a l l

along

the canyon but a l l these were small and were

20-30m

above the

water levels in

the

neares t

r iver

caves.

A

rock she l t e r in

the nor th -

eas t wall of

Canyon CS which we cal led Kitchen Cave or

Lakaton ny Lakozia ( f igure 4)

provided

f inds of

pot te ry , hearths and bones to i n t e res t

archaeologis t s . At

the eas tern

end of

the

canyon

an oddly

s i tua ted

dry r iver bed in

a

mixture

of

mud and boulders .

I t

i s some

Sm

deep and 10m

across and

runs

across the canyon for

i t s half

kilometre width.

The

r iver appears to have

s inks

a t both

ends

in boulder p i l es .

A

small

entrance

in the base

of the

dry

r iver

bed

near

the

easterr .

end

leads via a 3

0m climb

to the

water level where

a sump and a

boulder choke ha l t

progress

(see

discuss ion) •

LAKATON NY

LAKOZIA

a

firepl ce

o pottery

BeR A

o

5

10m

111

8/21/2019 BCRA 14-3-1987


A deep depress ion nor th

of

the camp (E9 )

looked promising.

However

, the canyon leading t o

t

proved to be very hard going with very th ick

vegeta t ion . I t s

walls

conta ined many small ca ve s

but these

a l l connected with

the surface only

IO -

20m

above and proved to

be

of no speleologica l

in te r es t . The

depress ion

i t s e l f

i s

very deep but

i t s

walls are

mantled by

extensive

debris which

crushed our

hopes of gaining

access

to any

of

the

maj or caves. The cent ra l area showed signs of

f looding,

but there

was

no

indicat ion that

the

water

comes

from

an

under lying cave.

One

small

shaf t , a t l eas t 20m

deep,

was

found

in

the W W

corner , but there

was no

draught . I t looked dry

and

unpromising

and

was

not descended.

South

Table Massi f Area

This area (f

igure

S) acquired

i t s

name from

the

prominent

l a rge rectangular block of l imestone

about

3km

long

by

O.Skm wide.

Aeria l

phot

ographs

suggested

t ha t

the

en t i r e Antanatsimanaja River

(F7) sinks j u s t south -

eas t

of the Tab l e .

However,

our

explora t ions revealed

that

the

r iver

does

not sink but merely

breaks

up

into

smal ler

r ivers

in

a

wooded

area; these reform

to

flow

south

and

jo in the Mananjeba River

(Fl)

The

s teep -

s ided

canyons

around

the Table

are not deep

enough

to reach

the

cave -

bear ing

limestones and

no major

discover ies were made.

One new

cave

(G20)

was

s lrveyed

in

the

eastern

canyon. I t

was very

well decora ted

and

had a

strong

draught , but

ended in a ca lc i t e choke af ter

only

290m ( f igure

6). The dry course of

a

r iver

was followed north in

the

western

canyon to the

end of

the

Tab le but t

normally

sank in an

impenetrable mud choke.

Grotte d Ampondriampanihy - Nord

We

were

te

mpted to explore t h i s neglected

area

by two

fea tures

:

the

sinks

of the

Besaboba

River (G17,

G18) and the Grot te

d Ampondriampa

nihy

-

Nord

(GI6).

The f i r s t sink

proved to be a spectacular loca t ion but the cave

was to ta l ly blocked

by

debr i s af t e r only 20m.

During our

searches

for

the Grot te

d Am pondriampanihy

-

Nord

(GI6) another

r iver was

discovered

beneath

boulders

in

a depress ion to i t s

south. I t

i s

presumed

that th i s

flows

to

the

Grot te

d Am pondriampanihy

-

Nord as no other ou t l e t

was seen. The

Grot te

d Ampondriampanihy-Nord

i t s e l f

i s a most impressive

cave.

I t s entrance,

which i s 60m wide

and

SOm

high

i s

hidden

in

a

small depre s sion a t

the

extreme southern end of

the mass i f .

The

entrance

narrows

to a

passage 30m

high

and

20m wide which

leads

af te r 700m to a 700m

long canal . To avoid poss ib le encounters with

crocodi les , negot ia t ion

of t h i s

par t of the cave

requi res a boat since there i s no dry land for i t s

en t i r e length.

A

fur ther 2km of similar

large

cave

leads to a boulder obst ruc t ion .

An ascending

passage on

the

r igh t

leads

up to a depress ion

(E7a) which had not been

fu l ly

inves t iga ted . The

Fig 5

===

=

km

Fog

-

td

SVFI

RW

Cave NQG2

EAST ANKARANA

Gnd , W 1.9 5' IJ.J

s

12

57

15

Length 29

2m

0

It

1

,

10

20

.,

.,

o

)

3()n

cave requi res a fu l l sur vey but t ime did

not a l low

us

to complete th i s . We entered many passages

from

a

large passage j u s t on

the other s ide

of

the

canal , but

surveys

drawn by the

French were ra ther

inaccurate

and gave

no clues as to

how much

of

the

cave

passages

we

were

explor ing

was new. The

depress ion

was very deep with

the

r iver

crossing

the

base

of t

beneath

boulders , but

no

continuat ion

of

the

cave was

found. A

route

to

the nor th led to a

fur ther

depress ion

(E7b)

where

the

r iver was

found again.

This

was par t i cu la r ly

in teres t ing as

t

showed that

the

r iver

flowed

north. However, the r iver emerged from an

impenetrable boulder

p i l e

only to s ink again in a

cave which was sumped af t e r a few metres .

The Northern

Area

During two

t r ips to the area

north of

the

camp,

we

only

succeeded

in f inding caves

which

had

already

been explored

by the

French. The

despera te lack o f

surface water

made th i s area

very di f f i cu l t

to

inves t iga te a t

the height

of

the

dry season.

Isolated Forest

Areas

The

depress ions

to

the

south of the end

of

the

Route Grim

( the t rack

which led to our base

camp) were extens ive ly explored ( f igure 7). From

the depress ion a

deep

canyon

running

north-west

looked very promising. However, extensive debris

along

the canyon walls blocked any

exis t ing

entrances. A small high- level cave , the Grot te

des Araignees, was

discovered and

sur

veyed

( f igure

8). I t

comprised 220m of

low

passages connecting

a

ser ies

of

chambers.

A narrow canyon

was

discovered to

the

south-west

of the f i r s t depress ion.

This

l ed , v

ia

severa l collapsed

areas and

shor t caves, towards

the i so la ted fores t depress ion (E4), one of

the

big cent ra l

depress ions

and a t a rge t for

the

bio logi s t s

.

The caves

and col lapses

became

known

as

Les

Grot tes Effondrees

(G12).

Time

did not

permi t us to f ind

the

way to E4.

However

soon

af t e r most

of the cavers

had l e f t

Ankarana,

Jean

Radofi lao to ld us that

he

had reached

i t .

112

8/21/2019 BCRA 14-3-1987


\

,

Grottedes

Araignees

ANKARANA

Grid :W

49 OS

36

5

12· 55

42

L.eogth : 187m

Caves

and Collapses

of

The

Route Grim

1 2 m

RW. Jl

LOCATION

o

O.25kM

Other Caves Visi ted

During

the course of the expedit ion, t r ips

were made into Grot te d Antsatrabonko (G8), Grot te

d Ants i roandoha (GIl)

and the

Grot te des

Rois

(GSa)

where

the

local kings are

in te r red

Discussion

The

dry r iver in the f i r s t canyon,

CS,

near

the

camp,

i s

in teres t ing

because t

provides

fur ther

evidence tha t

the canyons

f lood

to

high

levels

on occasions . The

dry r iver bed

l i e s about

30m above normal

water level

in

cave

G14.

The

drainage

in

the area of the Grotte

d Ampondriampanihy (GIG) seemed to di f f e r from

our

i n i t i a l predict ions

A

poss ib le drainage pat te rn

i s

shown

in f igure 9. The

flows

are a l l minute in

the

dry season

and

so

dye

tes t ing

was

not

prac t icable

Clearly the

s i tuat ion in

the wet

season

i s

very dif feren t

and drainage

pat terns

may

be completely a l te red

The r iver

which s inks in

the wet

season

a t G17

and

G18 appears to be very

big indeed

and

yet

the area

around the sink shows

only

f a i r ly

local

f looding and the

r iver does

not

back up. I t seems

that

the

r iver does not go to

GIG but fur ther into the massif probably

to

the

area

around

depress ion

E8. This

suggest ion was

supported by

more

rece:lt aer ia l photographs than

those taken in 1949 which are avai lable from FTM

in Antananarivo. These photographs indicate

a

large

r iver

dra in ing

in to

depress ion E8 which i s

s i tua ted d i rec t ly

north of depress ion

E7.

I f

t h i s

i s

so, then water is draining north through t h i s

part of

the massi f suggest ing a

more complicated

pat te rn than

has

been suggested

to da te

113

HYDROLOGY

IN THE

G. 6 AND E7 8 AREA

(Proposed)

Conclusions

<

The

cave explorat ion

ef fo r t s covered

many

areas

with in Ankarana which

had not been

previously inves t iga ted

The French

suggest ion

tha t

new caves would

only

be found

through

a

great

deal

of

hard

work was cer ta in ly proved to be t rue

However,

a

few cave discoveries were made

by

Expedi t ion members

and a

number

of

tan tal iz ing

leads have

been

l e f t

Routes in to

the cent ra l

depress ions

have

now

been found

and

the dra inage

predic t ions

in

the

East

might

yei ld cave passages

in depress ion

E8.

Any future t r i p must

invest igate

these.

8/21/2019 BCRA 14-3-1987


BATS

Mick McHale

The

Chiroptera

of

Madagascar

have

a f f i n i t i e s

with both the

African

and Asian

faunas,

Rhinolophus

i s absent from Madagascar but the

giant Pteropus f ru i t bats are present and

considered a

de l icacy

there. Few bio logi s t s have

worked on the bats of

Madagascar, an

except ion

being Dorst (1948).

The

f i r s t

study of the

ca

ve - frquenting bats

of

Ankarana

was

by

members

of

the

1981

Southampt

on

Universi ty expedit ion

(Adamson e t a l 1984;

Wilson 1985). We

wi shed to

ex

tend the 1981 study

by

co l l ec t in

g those species

r oo s t ing in caves and compare them to those which

co u

ld be t rapped

f ly ing

in the adjacent

f o

res ts .

Ankarana

has

a varied

and in teres t ing

bat

fa un a and the caves

pro

v i de an important refuge

for many species .

In

some areas bats enjoy a fad y

(taboo) s t a tu s and are not molested,

but

in

severa l

regions

of Madagascar

including

Ankarana

the

largest

species are considered much t a s t i e r

than

lemur

meat Whils t f rugi vo rous bats may ra id

local f ru i t t r ees ,

the insect ivorous

species

contr ibute

to

loca l agr icu l ture

by

reducing insec t

pes t populat ions. Future

studies

might do well to

quant i fy the

th reat

to bats posed by hunting and

the extent to which they are suscept ib le to

habi ta t des t ruc t

io n .

The cumbersome

Alcan

Harp

Traps proved

to

be

very

ef f ic ien t

in

t rapping fores t - f ly ing

bats .

In

cont ras t ,

despi te

repeated

e f for t s , we caught only

one

bat using

the compact and readi ly

por tab le

mist

nets .

Representat i ve s of each

bat species

were

k i l l ed

by neck

dis loca t ion and

preserved in

formalin (10 )

or

local

alcohol

(rum ) .

Preserva t ives were changed af ter three weeks.

Bats were observed in a l l caves v i s i t ed by

ex

pedi t ion members; the

most

common

was the

t iny

insec t ivorous

Miniopterus

minor. These ro o s ted in

caves

s ingly or in groups of

about

nine

individuals .

Th ey could be e as i ly

removed

by hand

from

the walls

or

low

roofs

of caves because

they

seemed to be

in

a

s t a t e

of torpor .

The

bats were

col lec ted towards the end

of

the s ix month l

ong

dry season

when

there

was

a dear th

of

insec t food;

t i s

poss ib le

that

the

bats

apparent

lowering o f

metabolic

r a t e was

in

response

to

shor tage of

food

.

Trouessart 's Trident Bat (Triaenops furculu s )

female; this

is a species endemic t o

Madagascar

(Ben

Gaskell)

Another cave

-

roost in

g

species

which occurs

a t

Ankarana, but

l ess commonly than

M. minor i s

Miniopterus i n f l a tu s

afr icanus .

This proved to be

the f i r s t record of t h i s an imal from Madagascar

(J .E.

Hil l ,

pers .

comm

1987).

The large

Hipp

o

s ideros commersoni

roost in

hundreds in the deeper (dark zone) sec t ions of

the

Grot te d '

Andrafiabe and

skeletons found in other

caves imply

they roost in l esser numbers

elsewhere

in the massif . Paul ian (1981)

wrote

of the

ki

l l in

g of Rousettus

a t

Ankarana and we cer ta in ly

found

evide

nce

that

the

local

people

l igh t f i r es

in the

caves

to

dis turb the bats ,

then

hi t

them

wi th

s t icks

as they t ry to escape. These bats ,

wh ich have a 60cm wingspan,

turned out

to be

Eidolon he1vum

dupraneum which

has not previously

been

rec ord

ed

as a cave-roost ing species

(Martin

Nicoll pers .

comm 1986).

I n the l a rge entrance

chamber of

la

Grot te

d 'Antsi roandoha we

discovered

a roost

of severa l thousand of these f ru i t bats;

they

also

seem

to

roos t in

crevices

in

the canyons

within the mass i f . Faeces beneath these colonies

implied

t hat

these bats subs i s t almost en t i r e ly

upon a

single type

of f ru i t ;

t h i s appears

to be

the ebony Diospyros

sp

. A r i ch , var ied and unique

invertebrate

cave fauna i s

based

upon f oo d brought

in to

the

caves

by

ba t s . Any

eco log ica l

changes

in

Ankarana

s

fores t s would not only endanger the

bats but also the

in teres t ing

and

special ised

ecosystem

which

r e l ie s upon them

.

Bats col lec ted from

t raps

in the fores t were

general ly

dif feren t

from

species

found roos t ing in

caves. Some

species were

found roost ing

deep

ins ide cave s in

the

Dark

Zone

(DZ),

while

others

roosted with in

s ight

of dayl ight

in

the

Threshold

(ThZ)

or

Entrance

Zones (EZ).

Chiroptera

recorded

from Ankarana

Species

Eidolon helvum dupraneum

Hipposideros

mme rsoni

Mini opterus minor manavi

Location

Grette

d A.ntsiroandoha

Gret

te d

Antsiroandoha

Grotte d

Andraf

iabe

Campsite

cave

Fir s t River ave

Second

River

Cave

Grotte 1 Andraf iabe

Grott.e d

Antsatrabonko

other caves and rock shelters

also

fl y ing

in forest

Miniopt

erus

inflatus

afr icanus Campsite cave

Tadarida j uqular is

Myotis goudoti

Emballonura atrata

Triaenops furculu s

Triaenops

persiclls rufus

i rst

River

Cave

Grotte

d Andrafiabe

forest

forest

forest

forest

CROCODILES

Paul Stewart

Zone

EZ

02

02

Th2

ThZ

ThZ

ThZ;

OZ

ThZ;

OZ

ThZ

ThZ

ThZ

The

Mal agasy

populat ion of

the

Nile

Crocodi le,

Crocodylus ni lo t icus ,

is genera l ly

considered to

be

under considerable

th reat .

The

subter ranean r ivers of the Ankarana

Massif

have

been c i ted

by

Blanc

(1984) as one of the l a s t

refuges for the species on the i s land and Ankarana

harbours

what appears

to be

the only populat ion

worldwide were indi v iduals are

regular ly

found

within a cave system . I t

i s

also one of the few

s i te s

in the worl

d

where 6m

long crocodi les st ll

survive.

114

The

crocodi les reach t he i r grea tes t dens i t i es

a t Ankarana in

the southern

caves of the River

Styx

which f lows

down

through

the massif from

the

nor th . I t appears

t ha t

the crocodiles r e t r ea t

into these caves to escape

the

drought in the dry

season

(May

to

Octobe

r ) , when areas of the

southern

marshes and r ivers

dry up.

The

crocodiles are

hunted

unsystematical ly in

t h i s

region for spor t

by

foreigners and for leather and

perhaps meat

by

some loca l s . Despite repeated

searches,

no evidence of

crocodi les

was

found

in

the caves

of

the northern

par t of the massif .

8/21/2019 BCRA 14-3-1987


The Nile Cr

ocodi les

f ound

in

the

Ankarana caves

were of

normal

pigmentat ion Ja ne Wils on

Their absence

might

be explained by

the fac t that

these caves

are i so la ted from the southern systems

by

long

sumps

and

white water

sec t ions

and

also

lack

su i tab le basking

s i t e s

in the open. The

nor thern

subterranean r ivers

are

also considerably

colder than the southern waters . For example,

water

in

the northern Second

River

was

a t

21°C,

whereas water

flowing

through the southerly

Ambanemero Faul t

cave system was

from 25 to 27°C,

I t is

l ikely that the

crocodi les

feed while

in the

southern

caves as the water temperatures

correspond

to

the low

end of

the ac t i

ve

temperature

range

for the species (26-27°C). A

v

ar i e ty

of prey is avai lable including crabs,

shrimps, crayf i sh , and f i sh

.

The larger

crocodi les may

not

only prey on smal ler

indiv iduals but also upon the large ee l s

descr ibed

in

the sec t ion

below.

OTHER CAVE FAUNA

Jane

Wilson

More

than 80 of Madagascar 's

animals

and

plants

are

pecul ia r to the

i s land

and the

cave

adapted

animals

(Troglobi

tes

which have

evolved

from

t h i s

unique

and

r i ch fauna have

the

scope for

being even more

in teres t ing

and qui te

dif feren t

to

that

of any other

kars t region. Ankarana is

unique for o ther

reasons

. I t i s

probably

contains

the most extensive

t rop ical

cave systems

in

Central

Africa

an

d the

s ize of

i t s

subter ranean

r ivers provide unusual t rop ical

cave

environments.

I t

i s also

an essent ia l

dry

season

refuge for

cer tain

epigean species .

Flooding

dur ing the wet

season seems to be on

catast rophic

proport ions for

some

canyons are submerged

under 30m

of

water .

And

t

is

in teres t ing

to

consider

the

effects

th i s

might

have

upon

specia t ion , which seems to

be

going

on a t

a

great

ra te

in

some

caves:

fur ther

s tudies are sure to reveal much more.

Although

the l i t e r a ture on the cave dwelling

animals

of Madagascar i s

qui te

extensive, t i s

pr inc ipa l ly l imi ted to

taxonomic

descr ip t ions

(Paul ian

1961;

Decary and

Keiner , 1970;

Remil let ,

1973) .

No cavernicoles had been recorded from

Ankarana un t i l 1981

(Wilson 1985) .

The en t i r e

food chain within

any cave

ecosystem i s abso lu te ly dependent upon

energy

from

the outside

environment.

At Ankarana food is

provided

by ba t guano,

mater ia l washed

in by

f loods

and

to a much l esser extent swif t

and

lemur

excreta .

This rel iance

on the outside

is

par t icu lar ly

re levant

a t

Ankarana, where

deforestat ion

i s threatening

the food supply

of

both f rugivorous and insect ivorous

bats .

Deforestat ion

i s

also allowing increasing so i l

erosion and redeposi ted sedimants could

poss ib ly

block

cave passages.

This was seen

(by

RW)

in the

Grot te

d'Analamisondro

t r a in the

nor th of

the

massif where

a

high

level

passage

some

5m

by

8m

was

completely blocked with

mud.

l15

Within Ankarana's

98km

of

surveyed

cave

passages, there

i s a wide

range of habi ta t

types:

large subter ranean

r ivers (where the

crocodi les

aes t i vate , high humidity

dry chambers, guano beds

and

chambers desiccated

by warm

through draughts

with

only

50 humidity. Cave temperatures

are

high: a t

the terminal co l lapse of

la

Grot te

d'Antsiroandoha t

was

24°C, for

example. The

existence

of

many varied

habi ta ts ,

even within

one

cave, explains

the

divers i ty

of

fauna. I t

i s

perhaps surpris ing that

th i s

r i ch

area had

been

neglected by speleologis t s and

bio logis ts .

Anima l s

were col lected

from

the i g h ~

r i ches t

caves v i s i t ed

. Many of the species l i s ted from

the

Andrafiabe

and

Antsatrobonko caves (Wilson

1985) proved to

exi

s t

in

other Ankara caves

but

very

many other

invertebrates await

ident i f icat ion

.

The

1981 col lec t ions revealed a

surpris ing

number of shrimp species: nine inhabi t ing two

caves

(Gurney 1984) and

the

1986

col lec t ions

included a t lea s t

two more

new records . I t seems

l ike ly

t he t

Caridina and Par i s ia shrimps form

large species complexes a t

Ankarana,

with

t roglobi tes ,

t roglophi les and

epigean species

represent ing var ious stages

in troglomorphic

evolut ion

. Shrimps known from Ankarana so

far

are:

Par i s i a dentata

P.macrophthalmia

P.microphthalmia

Caridina

xiphias

C.parvocula

C norvestica

C.ni lo t i ca

Car dina crur i sp inata

c.typus

C. unca

C.isaloensis

Other Caridina and Par i s ia species await

descr ip t ion .

The

crayfish Macrobrachium

moorei

also

exis ts in

the Ankarana streams but

sh

ow s no

cave

adaptat ion .

Other

in teres t ing

f inds

included

a new

species

of

t rog lobi t ic mill ipede and

probably the f i r s t Schizomid

to be

col lected

in

Madagascar (Legendre 1972). An

Amblypygid,

probably Charinus madagascar i ens i s was

common

in

the ent rance and threshold

zones

of a

number

of

Ankarana caves; t h i s species

is

probably endemic

to

the

massif . Most caves had

a res ident

eel ,

of ten

over

1.5m

long.

Unfor tunately

these

extremely strong animals fo i l ed a l l

at tempts

a t

capture

so remain unident i f ied.

They are

l ikely

to be a new species .

River Cave Two in

the

Canyon Fores t i e r was of

par t icu lar biological

in te r es t . Regular v i s i t s by

Crowned

Lemurs,

Lemur coronatus . to

dr ink

within

the

cave

have

polished rocks

smooth and

since t h i s

cave water

hole i s

used

by

rare

species such as

the

Fosa,

Cryptoprocta

ferox,

t i s an

excel lent

place

to

study both t rogloxenes

and

epigean

species .

The

cave contains

slowly-moving water

which i s

about

2m

deep a t

the entrance and i s

a t

21°C. Populat ions of eyed and bl ind cave shrimps

and three new

species of f i sh also l ive there .

8/21/2019 BCRA 14-3-1987


Two

Glossogobius

species (Peter Mil ler , pers comm

1987)

have

eyes,

are pigmented and presumably

ar r ived ( jus t

as

t he i r next meal wil l ar r ive)

by

way

of

f lood

waters . In te res t ingly

, some

of these

f i sh

were found

in

a

deep pool

which

was

glazed

over

by a mm layer of ca lc i te .

The th i rd

f i sh ,

probably

a

Gobius sp . ,

i s a

tog

lob i

te

and

was

found

even in

the Entrance and Threshold zones of

t h i s

cave . This is

almost cer t a in ly

endemic to

Ankarana which impl ies t i s r e s t r ic ted to to

r ivers flowing within

a

range perhaps measuring

10

x 30km .

Like

the 39 known species

of

t rog lobi t ic

f i sh t should be considered an endangered

species .

USE

OF

CAVES

BY LEMURS

Jane Wilson

and Paul Stewart

Accessible

drinking water for both people and

animals

i s

a t

a premium a t Ankarana dur ing

the dry

season.

The

most

r e l i ab le

water holes were

a t

cave entrances and lemurs and

other

wild l i fe were

regular ly observed drinking a t

·

the entrances

of

the Second

River Cave

(G14a),

la

Grot te

d Andraf iabe (G7) and

a t

the Syyx River

Cave

(G5).

I t

i s

l ikely that

lemurs

have

neede

d

to use

cave water

holes a t

Ankarana

for

centur ies and

t h i s might explain why the su b - foss i l s i t e ins ide

the Grot te

d Andra

f iabe i s

so

r i ch . All skeletons

were found

j u s t

beyond

the

l imi t

of

l i gh t

penet ra t ion

and most

were

complete skeletons which

must have ar r ived in the

cave while the

carcass

was

r e l a t ive ly

in t ac t . Many were lying on tops of

boulders in po

s

i t i ons incompatible with having

been

washed

in

by

f loods

.

We were unable

to

f ind

any avens nor any evidence

suppor t ing the idea

that

th e bones cou l d have ar r ived by

fal l ing

through a hole in the roof . No

ta lus

cones were

evident

in

the cave

system.

The

most

l ikely explanat ion for the ar r iva l

of the lemurs a t t h i s s i t e i s

that

they had come

into the cave

,

probab

l y

to drink

a t

a

water hole

sub-fossi l ized lemur bones found ins ide

Grotte d Andrafiabe.

Two

skulls

of

Hapalemur

simus are

on the le f t and the sKull

of a

Mesopropithecus

is bottom r ight

(Jane Wilson)

in the

cave entrance, and were

scared

in to

the

cave

by a

predator ( the Fosa, Cryptoprocta

ferox

wil l

take

lemurs and

seemed to

be

hunt ing

Crowned

Lemurs

a t

the Second

River

Cave) . When

fr ightened, lemurs of ten f lee upwards and once

they had f led beyond the l imi t of l igh t

penet ra t ion they would have

been unable

to f ind a

way out .

There

i s a

r e l i ab le

water

source

in the

Dark Zone j u s t beyond the sub- foss i l depos i t s but

t

seems

unl ike ly

that

lemurs could have

navigated

to

t h i s

in

absolute

darkness.

Carcasses

lying in

the

cave s

Dark Zone would

be

unavai lable

to

scavengers and also pro tec ted from the decomposing

effects

of

l ea f l i t t e r

.

This explains

why

more

foss i l s

have

not

survived in

Ankarana

s

i so la ted

fores t s .

None of the lemurs repesented as sub- foss i l s

a t t h i s s i t e st ll

s urvive

a t Ankarana as far as

we know and

Ankarana

s

extant lemur species were

poorly represented in

the

remains .

This

i s

consis tent

with the dr inking hole hypothe s is since

the

entrance

water hole

no

longer contains water

dur ing

the

dry season.

PALAEONTOLOGY

Martine

Vuillaume-Randramanantena

and

R

Ralaiar ison

-

Raharizel ina

The

Grot te d Andraf iabe

In

1981 members

of the

Southampton

Universi ty

Expedi t ion discovered a skul l and limb bones

of

four

Greater Bamboo

Lemurs,

Hapalemur simus

(Wilson

1985). At th is time

museum

skeletal

specimens

of

th is

species

t o t a l l ed

only

about

15

worldwide (Vuillaume-Randriamanantena

, Godf

rey and

Suther land

1985) .

The

s e

f inds implied

that

Ankarana might prove worthy

of a

special

palaeontological survey. When we

v i s i t ed

the

same

cave

in

1986, the sub

-

foss i l s i t e a t la

Grot te

d Andraf iabe,

proved to be

unusual ly

r i ch .

We were able to

col lect bones

of 18

Hapa

l emur

simus which i s a grea t concentrat ion of ske l etons

l fa lemur ra re ly found in su b - foss i l s ta te . We

l e f t other sub- foss i l s in

s i t u

since they were too

securely

cemented

to enable us to remove them,

with the l imi t

ed

equipmen t we

car r ied

, without

damage. Most sub- foss i l s (90 ) were of Hpalemur

simus

.

An

unident i f ied

t i b i a

(a lower

leg

bone)

probably

a pathological lemur

bone was also

col lected

and

skeletons

of Lemur sp. were seen.

Nearby were

skul ls

of

p r o ~ u s

diadema and

Mesopropithicus. Propi thecus

has

not been

recorded from Ankarana previously

a

nd

Mesopropi thecus

which

is now

ext inct , was

known

only from a

very few specimens col lec ted

more than

50

years

ago

in cent ra l and southern Madagasc

a r .

The

f i r s t Mesopropithecus was described

by

Standing

in

1905 and others were found

by

Lamberton in 1937

and

1939. The l a t t e r f inds were

in

the cave s i t e of la

Grot te

d Ankazoabo

on

the

south-west coast of Madagascar (Lamberton 1939,

1946

,

1948). The

1986

f inds

i

ncluded

previous ly

unknown

bones of the Me s opropi thecus which give

valuable new information cont r ibut ing to knowledge

of

the animal s l i f e s ty le (Vuillaume - Randriaman

antena

and

Ralaiar ison - Raharizel ina

1987).

One

of

the

most in teres t ing qu e

s

t ions

unanswered by palaeontologis t s , i s

the

age

of

lemur su b -

fo ss i l s

.

The

mo s t ancient Malag

as

y

s ub -

fo ss i l s

do cumented are from the deepest

depos i t s a t

Ampazambazimba;

these are

about

8000

years old (MacPhee,

Burney

and

Wells 1985).

More

commonly

age e s t imations f a l l

between 2000

and

1000 years

before

present

(Dewar

1984) .

Other

researchers

,

re ly ing

on

ora l evidence and

repor

t s

of ancient

t r ave l l e r s

l ike Flacourt

(1661), fee l

that

the l a s t Madagas car dwarf hippopotamus and

the

l a s t f l i

ght less elephant bi rds

on l y

disappeared a

few

centur ie s

ago. Sadly

there wer e

few

sediments

a t the

Andrafiabe

sub- foss i l s i t e s o

excavat ions would

be unl ike ly to revea

l f

urther

informat ion

on the age

of

the sk e

l etons a l t hough

accurate

da t ing

would

be po s

s ib le i f

some

bones

were

sacr i f i ced

.

The

remains

seem

to

span

a

l a rge

t ime

scale

. Some of the s ub - f oss i l s m

ight

be a

f ew

hundred

years o l

d, others including bi rd an

d

f ru i

t bat bones

are probably a t

most a

few

years

old.

116

8/21/2019 BCRA 14-3-1987


Micromammal Remains

in

Other Caves

We

searched numerous rock she l te r s for

ske le ta l remains and severa l proved to

have

r ich

deposits

of

recent or ig in .

The

r i ches t was from

the

Grot te

Trans-sept , in

the nor th wal l

of Canyon

Fores t i e r . I t

seems

tha t a t

Ankarana

cosmopolitan

rodents

have

not

yet

supplanted the

local

rodents

which are endemic to Madagascar. In the Ankarana

rock s he l t e r s

surveyed,

skeletons of loca l rodents

outnumbered those of

int roduced

species . This

i s

in

grea t

cont ras t to

the bones

found in

the

caves

of

Anjohibe

or

in

caves

in

the

cent re

of

Madagascar. On the cont rary , int roduced

r a t s

and

mice

represent

the

majori ty of species in

these

more southerly caves.

Par t

of

the zoological

work of

the expedi t ion

was

to at tempt a comprehensive inventory of a l l

animals inhabi t ing Ankarana , both in

caves

and

on

the

sur face . We expected to compile a

spec ies

l i s t of

the loca l

small mammals

using Longworth

Traps.

However, even

i f

the

f re ight shipment

conta in ing

the t raps had ar r ived , such a

survey

would have fa i l ed since mo s t small mammals

were

aes t iva t ing

and inac t ive

.

Skele ta l

remains found

in

the rock s he l t e r s therefore

provided

unexpected

informat ion on the ecologica l

i so lat ion

of

forests

within

the Ankarana Massif.

RCH EOLOGY

Jane

Wilson

Madagascar was probably only f i r s t se t t l ed by

man about 2500

years ago.

The f i r s t

s e t t l e r s were

thought to have ar r ived

by

boat

from

Indonesia

and

the

voyage

across the Indian Ocean in a

large

out - r igger canoe was proved to be poss ib le by the

Sarimanok Expedit ion in 1985. Early

archaeological discover ies are therefore

most

unl ike ly , al though some people think tha t there

was a populat ion

in

Madagascar before the

Indonesian se t t lement . When we discovered

pot tery,

f i r e

places

and other evidence of human

set t lement a t Ankarana

we

f i r s t

assumed

t h i s to

be

ev

id

ence

of people

hiding

a t Ankarana

between

1835

and

1838

when loca l

people

were hounded by the

so ld ie r s of Radama I . This may

be the

or ig in o f

some

ar t i f ac t s

but

t

seems probable tha t

the

archaeological

remains

are

f r

om

a

range

of

dates.

The

Ankarana shards show

s imi l a r i t i es with

pot te ry

used on the i s lands

off Madagascar s

nor th-west

coas t s dur ing the XVI XVIII centur ies

(Dr

Hilar ion

Rakotovololona,

Musee de l Univers i t e ,

Antananarivo:

per s

comm, 1987). Dr Bob Dewar

(Universi ty

of Connecticut) fe l t t ha t

the pot te ry

photographed

in

the huge entrance

of

the Grotte

d Antsi roandoha had been del ibera te ly l e f t in

caves

as offer ing s to a sp i r i t

or

as par t

of

a

funerar

y ceremony.

Ankarana seems

to be a r i ch

archaeological s i t e ,

worthy

of spec ia l i s t

a t ten t ion .

CONCLUSIONS

Phil Chapman

The kars t

and

caves

of Ankarana, al though

packed

into

a

small area of l imestone,

rank

with

other grea t kars t i c regions of

the

t ropics such as

the

Mulu Caves

in

Sarawak

or the Guangxi towers

in

China.

The int imate re la t ionship

between

c

ave

waters

of Ankarana

and the

wi

ld l i f e - r ich

fores t s

which

they

suppor t , lends the

caves

a biologica l

s igni f icance

over

and

above tha t

afforded

by t he i r

r i ch cavernicolous faunas.

The caves

also hold

important palaeontological

and

archaeological

remains.

Yet the area faces an

uncer ta in

fu ture .

Local, l a rge sca le

defores ta t ion maybe

increasing

s i l t i ng within

caves

and i s

posing a d i rec t th rea t

to

many fores t species.

Reducing

fores t habi ta t s

i s

l ikely

to inf luence

the

food supply of cave

commun

i t i e s

which depend upon bats , swi f t s , e tc .

Uncontrol led

hunting threa tens Ankarana s

unique

subter ranean

crocodi les .

In te rvent i on i s

r equi r ed

soon i f t h i s unique

area i s

to be safeguarded for

the

fu ture . The

Malagasy

Government

has

the po l i t i c a l

wil l

to

117

protec t Ankarana, but sadly lacks the necessary

resources.

Thus

an

ambitious

conservat ion projec t

i s underway to

extend

the work presented in

t h i s

repor t . In autumn

1987

a

small

team ,

including

some members of the 1986 Crocodile Caves of

Ankarana

Expedition, wil l make a f i lm about

the

area, i t s caves

and

wild l i fe

and t conservat ion

needs.

A

para l l e l

research

projec t ,

involving

mainly Madagascar -ba sed personnel , wil l aim to

draw up a Management Plan

for Ankarana

which wi l l

be

implemented using money ra i sed by

publ ic

appeal

through

in te rna t iona l

conservat ion

organisa t ions .

We hope tha t the

Crocodile

Caves of Ankarana

Expedit ion has

s tar ted

a process

which

wi l l

eventually

lead

to the permanent protec t ion of

t h i s superb

area

for

the enjoyment

of

cavers,

sc i en t i s t s

and the people

of

Madagascar.

KELIFELY REC O

NN ISS NCE

Dave Checkley

The Ke l

i f

e l y pl atea u , wi t h 8000 sq km o f l mes t o

ne,

i s

the

second

largest

potential caving

area

in

Madagascar. I t

i s reputedly nearly

a l l Jurass ic limestone (Balazs 1980), and

i s

approximately

60km

long

by

50km

wide.

The

southern

scarp

r ises

nearly

1000m above the surrounding plain .

Kelifely i s in a f a i r l y remote area , 250km north - west of

Antananarivo

and 600km south west

of

Ankarana. Road access

may

be

di f f i cu l t since several river crossings

are

required

and

the

ferr ies no

longer

operate.

However,

there

are

regular

f l igh t s

fr om

Antananarivo

to Ambatomainty, a

v i l lag e

25km

south of

the massif

with

a good dry

weather

ai rs t r ip .

The

Mahakamba

River

l i e s between

the massif and Ambatomainty,

but

t h i s probably d r ies up

for

some

of

the

year ,

as does the

larger Mahavavy River to the eas t

of

the massi f .

The only people to look for caves in the region

previous ly

were a French team that v i s i t ed the most nor ther ly

area of Kelifely (Peyre e t a l 98 and 1983). They

found

no

s igni f icant caves, but their movements

were severely

r i s t r ic ted by the presence of bandi ts

in the region.

The

French approached

the

massi f from the

town of

5 i tampiky in

the nor th and

did

not reach the

massif

proper . Bandits

are

no longer a

s ign i f i c ant problem.

e chartered a

four-seater

aircraft through

Madagascar

Airtours in Antananarivo at approximately

£400

for

the

afternoon. We

f le

w direct ly to Keli fely across the

dry

plains .

From

Ambatomainty

we

flew

s t ra ight

up

to

and

along

the southern

scarp.

West of Arnbatomainty the scarp i s steep

and cont inuous .

t

has very

fe,,

breaks in

t

and only

occasional patches of

woodland

in the shallow val leys

incised

into

i t .

To

the nor th eas t

o f Ambatomainty

a t the base

of

the

massif

there i s an

incredib le landscape

of

tor tuous canyons.

These winding,

steep-sided

gorges

were

c l ose l y intertwined

over a

huge

area . The

gorges

were perhaps 30m deep and

were

so closely

packed

th a t the area presented a vi s t a of endless

rocky pinnacles running up to

the

Keli fe ly scarp . There was

no water in

any

of these gorges and no obvious caves at the ir

heads in

the

massif . Woodland

did

however obscure many of

these

val ley

heads. At

many gorge

heads were steep c l i f f s ,

but

with no sign

of caves at

their bases. The

red sandstone

in which the gorges

have

formed ( Isalo formation)

must

be

very

soft

to have been

so

heavily eroded

by

the

l imited run

off

from

the Keli fe ly

scarp .

The scarp in t h i s area has a

stepped appearance with small (?l imestone)

c l i f f s

and

intervening grassy

or sometimes

wooded

s lopes .

North of the scarp is a plateau, an

extensive

area of

grassland sloping very

gent ly

northward. These grasslands

are

only

occasionally

punctuated

by clumps

of trees in

shallow depressions and by the winding, tree l ined

river

val leys . e saw

only

one region of apparently broken

l imestone

pavement

and a

few

shallow dol ines . We followed

the

two major

rivers

draining the plateau for their entire

lengths the

Kiananga

and Tsiamadiovolana Rivers) . Their

blue green waters occasional ly plummeted perhaps S m down

great

round

bowls,

forming spectacular waterfal ls . We could

not

be certa in that there were no caves at

the

bases o f these

wooded

bowls, but the

r ive r

gorges

cut out of

these

bowls a t

their far s ides appeared to

carry

the same quantity

of water

as went in

at

the

top.

The

fur,ther north

we went

along

these

t r ibu ta r i e s

of

the Mahavavy

River,

the wider

and more

wooded

t he i r valleys became. However ,

even

in the higher f requent ly

c l i f fed

gorges of these rivers

we

saw not a s ingle cave

entrance . The

r iv e r s

always flowed

on

the surface and

genera l ly there

were very few kars t features .

We

flew

as

f a r

nor th as the

Kasijy Forest .

Although we did not l and a t Keli fe ly we flew

within

20m

of much of the surface of the massif for one and a h a l f

hours . We were unable to ce r t a in ly

id en t i fy

any cave

entrances. I do

not th e r e f ore

bel ieve that

the

area i s

worth

a

major

speleological expedi t ion .

This

conunent

i s

made

in

the knowledge

th a t a small Malagasy team

(based in

Antananarivo)

wi l l vis i t

the

area

t h i s year (1987). I t wi l l

be i n t e res t ing to

see

what they f ind , but the development of

major cave systems in the

area

does not seem

very

l i ke ly .

8/21/2019 BCRA 14-3-1987


MEDICAL REPORT

Jane Wilson

During

the

dry

season

which is the most practicable time

to work at Ankarana, the dearth of

water

on

the surface makes

it

a

remarkably

pleasant and heal thy

place

to be.

I t hardly

rains so tents are

unnecessary, there

are no insects and the

only noxious animals are

the.scorpions and

centipedes. These

animals

do not real ly become

a

nuisance

unt i l the

ear l ies t

ra ins in October,

which

i s about the time th a t the bi t ing

insects appear. The largest and most unpleasant scorpion i s

Groophus palpa tor and the two of

us who

were stung by these

were

in great pain with rigors ,

sweating,

etc for 24 hours

and

unwell

for

severa l

days

subsequent ly .

Pain

a t

the

s i t e

of

the

s t ing

resolved

after just

over a

week but

one

victim

was

l e f t

wi th an anaesthetic

f inger for a month. Powerful

pa in-k i l l e r s preferably opiate analgesics , should be

availab le for the f i r s t

24

hours a f t e r

such a s t ing. We

used

susta ined release Morphine

t ab le ts ,

but parentral drugs might

h

av

e

been more appropria te

.

Sc o rpions favour hiding in dark

corners dur ing

dayl ight

hours

and even

manage

to

f ind t he i r

w y nto rucks ck pockets when t h s been suspended on

tree. Great

care i s needed

to

avoid

st in

gs

which

are

reputedly

r a r e ly fa t a l

on Madagascar (see also Wilson 1987) .

Another important health r i sk a t

Ankarana

i s from

trauma; the

t s ingy

i s sharp

and unstable and

very abrasive.

Strong boots

and

l ea th er gloves helped to pro tect

us

and

we

were fortunate

that

our most

serious

accident was a

l ace r a t io n

from a palm leaf .

Other

problems included t r ave l l e r s diarrhoea and

sickness;

Lasoni l

proved helpfu l

subsequent ly

f

or soothing

over-act ive

anal

sphincters. Several of us were

troubled

by

mouth

u lcer s

which we

eased by

Te e - je1 . Three of us

caught

worms

which

were diagnosed and

treated after

we returned to

Brita in .

We

s tayed hea l thy

because

o f

the easy

condi t ions

a t

Ankarana,

having the

right irrununisations

and malaria

prophylaxis and good

camp

hygiene (with

a

l a t r ine and

compost

heap away from the

l iv in g a r ea ) .

We acqui red most g as t ro

in tes t ina l problems while we were

sampling

the exce l lent

Malagasy food that i s availab le

in

the towns . These usually

se t t led within 24-48

hours

takihg only clear

f lu id s

but one

o f us required a

course

of

Flagyl (Metronidozole)

for

presumed

gia rd ias i s .

Malagasy doctors

and lo ca l

Aides

Sani ta i res

are well t rained and very competent but

the

serv ice they can offer i s ext remely l imi ted due to

shortages

of drugs and medical suppl ies

. A

more comprehensive account

of the medical

aspects

of the

Expedi t ion,

including

a l i s t

of

drugs

taken, i s in

our

repor t (Chapman e t a l

1987a).

ACKNOWLEDGEMENTS

We are

indebted to the

Malagasy author i t i e s

for

permission

to

conduct

this

survey

and to

Drs Lee

Durrell

and

Alison Richard In te rn a t io n

a l Group

of Sc ient i s t s ) ,

Dr

Jon

Pollock,

Shane

Winser,

Diana Hancock, Dr David

Harding, Tony

Wal

tham,

Mike

Sal isbury,

Mrs

A

Trimble,

Ruth

Mindell ,

Ben

Lyon and Dave Ell io t for advice and

contacts

and

to

Simon and

Cather ine

Howarth, Andy

Cairns, ill and David Wilson,

Anthony

Orr, Ian Looker, Howard Jones, Tony Gunn, Ian Redmond

and

Jane

Darr ington for i nva luable

help.

David ar.d Lucy

Andriamparison, Dr Voara Randrianasolo, Benjamin LeNormand,

David Curl and many others

ass i s ted and encouraged

us while

we

were in

Madagascar.

The p ro jec t

rece ived

funding

from

the

Spor ts

Council ,

Ghar Parau ~ o u n d a t i o n

Royal

Geological

Society

Bri t i sh Sugar

Award),

Bri t ish

Ecological

Socie ty ,

Smith

Kline

and French; Duke

of

Edinburgh

Trust, Scient i f ic

Explorat ion

Socie ty , John

Spedan

Lewis Trust , Bri t i sh Museum

(Godman Fund) , Twenty-Seven

Foundation, Royal

Botanic

Gardens, Kew for seed col l ec t ions ) . Sir Peter Scot t

kindly

acted

as

ou

r

pat ron and

Dr

Peter Edwards (Biology Dept . ,

Univers i ty of

Southampton)

and

the

l a te

Dr G T Jef ferson

acted as academic referees . We received

approval

from the

Primate

Socie ty and

the

academic suppor t of

the

Univers i t ies

of

Southampton,

Hull and

Wales.

The fol lowing

companies were

esp ec ia l ly generous in supplying t he i r products :

AJBB

Plas t i c s , Alcan; Alexandra

Workwear;

Bayer,

BDH

Chemicals;

Berk, Boots , BTR Boots; Carton In d u s t r ies ,

Caving Supplies;

Chance

Propper;

CPL

f i lms;

Dearlove Blue

Drums;

Duncan

Flockhar t Co; Endecotts; Fisons , Gold

Cross;

W Gore;

Hampton Prin ting;

Hanimex; In te rn a t io n a l

Fre ight

Forwarders;

I tona ;

Janssen; Kodak

; Leder le ; Lyon Equipment;

May

and

Baker, Mossley Wool Combing and Spinning;

Napp; Nelsons;

Oxoid;

Paramorr

Electronics ,

Pel l ing Cross;

Pfizer;

Phoenix

Equipment, Polaroid;

Prest ige; Roche;

Smith

Kline and French,

Smith

and

Nephew; Surviva l

Aids;

Varta Batteries ; Winit

Boots . We are g ra te fu l to Mr

J E

Hi l l and Joan El l i s for

prompt

ident if icat ion

s of the bats and shrimps.

The

expedi t ion

members were Jane

Wilson

and Phil

Chapman

co-leaders) ,Dave

Checkley, Simon Fowler,

Shiels Hurd,

Mick

McHale, Jean Radofilao (formerly Duflos) (C.U

.R . ,

B.P .O. ,

Antsiranana), Raobivelonoro Ralaiarison -

Raharizelina

and

Martine Vuillaume -Randr iamanantena (Service de Paleontologie ,

Universi te de Madagascar, B.P.

906

,

Antananarivo 101),

Guy-Suzon

Ramangason Department de Botanique, Universite de

Madagascar) , Jean - Elie Ranctriamasy Parc Tsimbazaza,

Antananarivo)

, Paul Stewart and Roo Walters. Dave Clarke,

Sally Crook, Ann Denning,

Ben

Gaskell and Maggie

Hutchings

also

helped

with the f i e ld

work.

REFERENCES

Adamson, A P, Boase, M J

Howarth,

C J, Wilson, J M and

Wilson

M

E,

1984.

Southampton

Universi ty

Madagascar

Expedit ion

1981

. Universi ty of Southampton 136pp.

Balazs, D. 1980.

Madagaszkar karsztvidekei.

Karszt es

Bar lang (Budapest)

vol

1, 25-32.

Blanc,

C P ,

1981

Batrac iens e t Repti les : formes e t couleurs

inso l i t e s . In

Madagascar, Un sancuaire

de

la nature

.

Ed.

P Oberle.

Lechaval ier , Par i s .

Blanc,

C P,

1984

.

The

r ep t i le s .

In

Madagascar: Key

Environments ser ies eds

A

Jo l ly ,

P

Oberle and

R

Albignac. pp 105-114. Pergamon, Oxford.

Chapm

an,

P ,

Ch eckley,

D,

Fowler,

S V, Hurd,

S, Ramangason

, G

-5,

Randriamasy, J -E, McHale,

M

Stewart,

P, Walters,

R, and Wilson, J

M

1987a. Crocodi le Caves of Ankarana

Expedition

Report.

editor R Walters; private

publicat ion .

Chapman, P ,

Checkley,

0, Fowler, S v ,

Hurd,S ,

Ramangason, G

-5, Randriamasy,

J -E, McHale,

M

Stewart, P, Walters,

R

and

Wilson,

J

M. 1987b.

A

survey

and

management

proposa ls for a tropica l deciduous forest reserve at

Ankarana

in

northern Madagascar. Submitted to

Biologica l

Con

se rva t i on

.

Decary, R, and Keiner, A,

1970.

Les cavites soutarraines de

Madagascar. Annales de

Speleologie

vol . 25,

pp409

- 440 .

DecarY ,R

, and

Keiner,

A, 1971. Inventaire

schematigue

des

c

av

i t e s de Madagascar. Annales

de Speleologie

vol . 26,

pp31-46.

Tr

oo

p

of Crowned Lemurs

enter ing

the Second River Cave

to dr ink

(Jane Wilson)

118

8/21/2019 BCRA 14-3-1987


Dewar, R. 1984. Ext inct ions in Madagascar ;

the

loss o f

the

sub-fossi l

fa un a

. In Quaternary Exti nctions: a

prehistoric revolution eds. P S Martin and R G Klein.

Universi ty

of

Arizona Press ,

Tucson;

pp547-593

.

Dorst, J . 1948.

Essai d une

c le f de determination des

Chauvre-sQuris

Malgaches. Mem.

Inst . Sc i . Madagascar A

1

1)

81.

Duf les , J . 1966. Bilan des ex

p lorat ion s

spe leo log iques pour

l 'annee 1965. Revue de Geographie (Universi te de

Madagascar)

vol

9 pp235-252.

Du f 10 s ,

J .

1968. Bi1an des

ex p lo ra t i o n s pour l ' an n ee

1966.

Revue

de

Geographie

(Universite de

Madagascar)

vol

12,

pp121-1 29

.

F1acourt, E

de 1661.

H is to i r e

de

la

Grande

I s l e

de

Madagascar. Gervais-Clouzier Paris,

471pp.

Gurney,

A R.

1984. Freshwater shrimp genera

Carid ina

and

Parisia of Madagascar with descript ions of

four new

species .

J .

Nat. Hist .

vol

18,

pp 567-590.

Lamberto n, C 1936

(1937)

.

Fo u i l l es

paleonto10giques fa i t e s

en 1936

.

Bull .

Acad .

Malagach

e vol

19,

ppl-19 .

Lamberton,

C

1939. Contribution

a

1a connaissance

de

1a

subfossi le de Madagascar. Lemuriens et Cryptoproctes.

Note

4.

Mem . Acad.

Malagache vol 27, pp9-49

.

Lamberton,

C

1946 (1948)

. Cont r ibut ion a

la

connaissance

de

la faune

subf oss i l e

de

Madagascar . No te 20 Bul l . Acad .

Malagache vol 27 , pp30 - 34.

Legendre,

R

1972. Les Arachnides

de Madagascar,

in

Biogeography

and

Ecology

in Madagascar . eds R B a t t i s t i n i

and G Richard-Vindard. Monographiae

Biolog icae

Vol 21 ,

(Junk, The Hague)

pp427-457.

MacPhee,

R D

E, Burney,

D A

and Wells ,

N A,

1985. Early

Holocene

chronology

and environment o f Ampasarnbazimba a

Malagasy

subf oss i l

s i t e .

I n te r na t . J Primato l

vol 6

(5) , pp463-490.

Paulian,

R 1961 . La

faune cavernicole .

In: Faune de

Madagascar XIII : La

Zoogeographie

de Madagascar e t des

I les Vo i s ines .

Inst itut de

la Recherche Scienti f ique ,

Tsimbazaza

Antananarivo pp94-99

.

Paul ian ,

R 1981 . Les Mammiferes: vest igues

d 'un

monde

disparu

.

In Madagascar, Un sancuaire

de la

nature. Ed

P

Ober le .

Lechaval ier ,

Paris .

Peyre, J - C

e t a l 1981. Madagascar

- Expedi t ions

speleo logiques: CANOPUS

81

and K el i f e ly 81.

Sp e

1eologie

115 Speleo

Club Martel , Nice

/C lub

Alpin Francais .

Peyre,

J - C Arthaud, G

Bessaguet,

D

Fulcrand, 5,

Martin

M

Radofilao,

J,

Flandin

J -M and

Tessier , F. 1982.

Expedi t i on Speleologique : Madagascar 1982. Club Alpin

Franca i s

/Federat ion

f ran ca i s de spe leo lg ie .

Peyr

e , J

-C e t a l 1983. Kel i f e ly

83 -

Reconnaissance

Speleo10gique. Club Alpin Franca i s

.

Peyre,

J -C, Bessaguet, D Curtare l l i , T, Franco, E

and

Radof i l ao ,

J .

1984. Speleologie :

Special

Madagasca r .

Bul l .

Club

Marte l , Ni

ce/C

lub

Alpin Francais .

Peyre, J -C

e t a l

1985. Ex plo r

a t i o n

Spe leo log ique dans

l Ile

de

Madagascar. Bull .

Club Marte l , Nice / Club

Alpin

Franca

i s .

Radof i l ao , J 1877. B il an des

ex

plo ra t ions spe leologiques

dans l Ankarana

. A

nna

les

Univ. Madagascar. Se r

:

sc i

na t

math. vol 14, pp195-204.

119

Ravelonanosy,

C and Duflos , J . 1965. Bilan des

explorat ions

speleo10giques pour l 'annee 1964. Revue de Geographie

(Universi te

de Madagascar) vol 6, ppl1

7 -

132

.

Remillett , M 1973. Apercu de la faune

souterraine

a

Madagascar. Li v re du Cinquantaire de l Ins t i tu t de

Speleologie Emile

Ra

co

vi tza ,

Buchares t

pp135-160 .

Rossi,

G 1973. Problems morphologiques du karst de

l Ankarana.

Revue de Geographie (Universite de

Madagascar) vol 23.

Rossi , G 1874. Morphologie

et

evolution d un

karst

en mil ieu

tropical:

l Ankarana.

Memoires e t

Documents

CNRS

nov.

se r

.

15:

Phenomenes Kars t iques I I ,

pp279-29S.

Rossi ,

G

1975 Resultats

de

deux annees

de

mesures de

teneurs

en calcium carbonate d un resurgence

karstique

t r op ica le . Spelunca vol 2, pp27 - 28 .

Rossi, G 1976a Karst et d i s so lut ion

des

calcaires

en

mil ieu

t rop ica l . Z. Geomorph.

N. F.

(Ber l in) vol

26

pp124-152.

Ross i , G

1976b.

Problems biogeographiques

de l 'ex t reme

-

nord

de Madagascar. Madagascar, Revue

de

Geographie

(Universi te

de Madagascar) vol 28 pp

133-154.

De Saint

Our

s , J 1959. Les phenomenes karst iques

a

Madagascar . Annales de Spe1eologie vol 14

Standing , G

-F

1905

(1906). Rapports sur l e s ossements

sub- f

oss

i l e s

provenant d ' Ampasambazimba.

Bul l . Acad.

Malagache (an.

se r . ) vol IV,

pp95-100.

Vuillaume-Randriamanantena, M

Godfrey,

L R

and

Sutherland,

M R. 1985. Revision o f Hapa1emur ga11iene.

Fol ia

Primato log ica

vol

45, pp89 - 116.

Vuillaume-Randriamanantena, M and Ralaiarison -

Raharizelina,

R. 1987. Prospection paleontologique dans l Ankarana.

Rapport prel iminaire . Paper presented to Academie

Malagache

on

16 Oct . 1986

and

in press

with

Bull .

Acad.

Malagache

. .

Wilson,

J M

1985

. Ecology o f the Crocodi le

Caves

of

Ankarana, Madagascar

. Cave

Science:

Trans.

Brit.

Cave

Res. Assoc. vol

12,

p(3) ,

pp135-138.

Wilson, J M 1987.

The scorp ion

s tor y .

B r i t i s h

Medical

Journal, Christmas 1987, in press .

Wilson, J

M

Stewart

, P D and

Fowler,

S V.

1988

.

Ankarana:

a

rediscovered reserve

in

northern

Madagascar. Oryx, in

press.

Received

August 1987

Jane

Wilson

c /o Phi l ip Chapman

City

Museum

Br is to l

BSS 1RL

8/21/2019 BCRA 14-3-1987


8/21/2019 BCRA 14-3-1987


C VE

SCIENCE

Vol. 14,

No.3 ,

December

1987

Transact ions

of the

Bri t i sh

Cave Research

Associa t ion

Caves

of Lower Glomdal Rana Norway

Simon BOTTRELL

Abstrac t : The lower

par t

of Glomdal, Rana, conta ins

a

la rge

number of

small

cave

segments

and one

l a rger

cave.

Most

of the

cave passage

i s

of

phrea t ic

or ig in , re la t ing to an

ancient hydrologica l

regime,

and

the caves

are now

abandoned

and

dry.

The

except ions

are caves which

have

been

invaded by

a

modern

streamway and

now form par t

of

the ac t ive drainage system. The most

s igni f i cant ac t ive underground

streamway

flows

for 1100m

northward across

the area and may

be

entered

in f ive separa te caves.

INTRODUCTION

In July and August

1984

a

small expedi t ion

from

the Univers i ty

of East Anglia

UEA)

vis i ted

Glomdal in

Rana,

Northern Norway. We undertook

a

sys temat ic

explorat ion of

the lower

par t of

Glomdal, between the road end a t

Fiskjornmoen

and

the southern

shore

of Glomda1svatnet Figs .

1

and

2). This

area i s

an

extension

of

the

marble kars t

which

forms

the well known caving area higher

up

G10mdal

wi th severa l

large

caves e. g.

Pikhauggrot te , Fossehole t ,

S torbekgrot t e ;

see

Hjorthen 1968) and St . P i e r r e

and

St .

Pier re 1969)) . Although severa l caves

and

entrances

were

known in lower Glomdal,

no

extensive explorat ion had been

undertaken

and

few

of the known caves

had

been

surveyed.

We

aimed

to

produce

a fu l l inventory

of

the caves

of t h i s

area

and

survey as many as

poss ib le

in

the

t ime

ava i lab le .

In the event , the discovery

of

a new

ac t ive cave system in the nor th-eas t of the

area

meant

tha t we

were

unable to cover

the

south

and

south-eas tern areas.

After the UE expedi t ion

l e f t

Glomdal, members of

Wessex

Cave Club

connected two

of the newly discovered caves and

in

1986 Simon Bot t re l l

re turned

to survey

the

connections with the

aid of

members of Speleo-club

Vauban

.

The lower

par t of

Glomda1 i s dominated by the

s teep

s ided

canyon of

the

Glomaga r iver but

on

the

eas t

s ide

there

i s an areas

of small h i l l s up to

270m

high

bounded to

the eas t

by

the

c l i f f s of

the

old

g lac i a l

val ley . Much of

t h i s

area

cons i s t s

of

outcropping marbles which are well

kars t i f i ed ,

with

an

abundance of

open

grykes, shaf t s

and

shakeholes . Most of the

area i s

covered

by

dense

bi rch fores t

and

a l l access

beyond

Fiskjornmoen

i s

on

foot .

In 1984

the

primary

aim was to locate,

explore

and

survey as many caves as poss ib le in

lower

Glomdal

and the ma jor i ty of the avai lab le

t ime

was

spent on these t asks .

In

addi t ion ,

some

dy e -

t racing

experiments

were car r i ed out

and a

prel iminary

geologica l map

of the area was

made.

In 1986 the emphasis of the work was sh i f t ed

to

more

deta i l ed

geologica l

mapping and

s tudies

of

the morphology

of the caves. The

survey

of the

l a rges t

cave

of the area

was completed wi

th the

help of members of Speleo-club Vauban Li l l e ,

Franc

e)

.

GEOLOGY OF

THE RE

Pr ior to the 1984 UE expedi t ion t

had

been

thought

tha t the marbles

outcropping in

lower

G10mdal

were an extension

of the r e l a t ive ly simple

geology higher up

the val ley and we had

ant ic ipa ted tha t the mapping

of

the lower val ley

would be completed during the

expedi t ion.

In

the

event ,

the geology

of the

lower

val ley

was

found

to be

far more

complex.

In 1984 only

a simple

outcrop

map

in the area of the streamway caves was

completed. In 1986 three weeks we r

e

spent on

mapping

the

lower

par t

of the

val ley

and

the

complex s t ruc ture unravel led. Samples were

also

col lec ted from

both

lower and upper Glomdal for

petrographic and

laboratory s tudies .

Only a

br ief

descr ip t ion

the geology

of the area

around the

121

caves i s given

here . I t

i s

hoped

tha t in 1987 the

mapping

wil l be completed

and the

deta i l ed

resu l t s

published Bot t re l l and Lauri tzen, in

prep.) .

Three l i tho logies

were

recognised in lower

Glomdal:

GREY

M RBLES

of almost pure ca lc i t e , with

minor

amounts of quar tz and

mica. Bands are

commonly

1- Sm wide. YELLOW M RBLES

composed of

ca lc i t e

with

var iable amounts of dolomite , mica,

quar tz and

other

minera ls .

Bands

are

between

3

and

30 m

wide. These

marbles

are very

heterogenous and

exhibi t composit ional

banding on

a

var ie ty of sca les from mill imetres

to

about

a

metre , with

some

bands of being composed of almost

pure

ca lc i t e marbles .

MIC SCHISTS

composed of

quartz, mica, feldspar

and

garnet .

The rocks of lower Glomdal exhib i t three

phases of deformation, the

second

dominates the

large-scale s t ruc ture

and hence

the outcrop

pat te rn . The

folding i s

t igh t

to i soc l ina l and

thus

the

groups

of marbles outcrop

as

near ly

para l l e l bands

100-lS0m wide t r ending nor th-south,

and general ly

dip

a t between

30 and

60·

to

the

eas t . This

pat te rn

i s

complicated

by the ef fec t s

of

superimposit ion

of

the three phases

of

deformation, but within the area of the caves

descr ibed

here

the

marbles

can be

considered as

near para l l e l

bands . The folding a lso af fec t s the

Figure

1. Location map

for Glomdal

in Norway.

8/21/2019 BCRA 14-3-1987


thickness

of

the bands, the th ickes t

marbles

being

found a t the

fo ld

hinges and thinning

out

along

the fold l imbs.

The

dominant

f ractures

in

the

area t rend north-south and are near

ver t i ca l .

THE

CAVES

The caves of the area can broadly be sp l i t

in to

two groups, present ly ac t ive

stream caves and

segments

of

foss i l phrea t ic caves.

Active stream caves

The now famous

Glomdalsvatnet out let

cave

(Laur i t zen e t a l . 19 85) l i e s in the nor th of the

areas and was already wel

l

known.

Our

major

discovery

was

a

s tream system

draining from a

sink

a t

UTM50 7757 a t 205m

a . s . l .

to a resurgence a t

UTM507768 a t 129m

a . s . l . This

could be entered as

f ive separa te caves a t the entrances marked in

f igure 2.

The

streamway connect ion

was pos i t ive ly

dye- tes ted

through a l l

the

caves

in

the

Neverslet te-systemet and they are

plo t ted

together

in f igure 3.

The s tream sinks

among boulders

a t

the

head

of

a dry va

l l ey

leading to a

closed

depression a t

the

base

of which a small

entrance

drops in to

the

streamway of KJOKKEN VASKEN

(llOOm).

The cave

follows the stream

in

an inc l ined boulder - f i l l ed

r i f t

which

opens

out

in to

a large

r i f t

chamber

with the s tream

flowing

down a

ser ies

of cascades

to

a

sump.

Opposite

the top

of

the cascades an

upslope crawl

leads

to another

r i f t passage.

Upwards

connects

back

to the roof of the

cascade

chamber, downwards leads to

two sumps

and another

upward crawl into the

mainly

dry

upper se r i es .

The

upper se r i es consis ts of

a

number

of

small

passages leading off

from a

main

walking sized

passage. Immediately before a l a rge wa t e r - f i l l e d

descending tube ( the Moose

Trap)

i s a

crawl

on the

r igh t

leading

to

a

ver t i ca l r i f t and the squeeze

connect ion

to the

Gal ler ies Francais , a

maze of

small

phrea t ic tubes,

and the second

entrance to

the cave. Beyond the

moose

t r ap i s a 20m

high

ver t i ca l r i f t and a number of small passages which

might give fur ther extensions

i f

pushed.

The

streamway i s re jo ined

in DE

MURREN

DE

DOTRES GROTTE

(32m)

(DMD-grotte)

where

a

small

entrance

in the l a rge

depression

to the eas t of

the

path

gives

access

to

a

few metres of streamway

between a

sumped r i s ing and the s tream

disappear ing

in to

a small

f i ssure . The s tream

reappears

a t

the

upstream

sump of

TROLLHOLET

(125m) ,

here

the

streamwa

y i s

entered

via an

8m

entrance

pi tch .

Downstream leads to a sump in a

r i f t passage,

upstream

leads to near surface

boulder chokes and low crawling to the upstream

sump.

GAUDAGROTTA (188m) i s entered by an 11m

entrance

pi tch

in

a

narrow shaf t

leading

to

a

squeeze into

a

large stream passage. Downstream

soon leads to a small

maze

o f act ive

and

abandoned

passages and

the

upstream

sump in a large r i f t

approximately 20m

from the downstream sump

of

Trol lhole t . This

i s the most

in teres t ing

of the

smal ler caves.

o

500

• I I

Metres

N

t

I

I

/

,

,

ktjorna

Fi

1

•

Path

•

Cave entrance

Underground

Flow

Kjokken-vasken

2

De

Murrende

Dotres

Grotta

3

Trollholet

4

Gaudagrotte

5

Dolinegrotte

Figure

2. Map

of study area in

lower Glomdal,

showing

loc t ions o f

entrances

to streamway caves

DOLINEGROTTA

(29m) i s the

l a s t ca

ve

in the

system. A

climb

down

through

boulders a t the

base

of a

l a rge

depression

leads in to

a

shor t sec t ion

of

stream passage sui ted only to the masochist

The stream f ina l ly

resurges among boulders

to the

south of a small lake

a t

UTM507768. Above

the

resur

gence

i s a small dry val ley and a

shor t

dry

cave which may be re la ted to the

system.

Figure

3.

Surveys of caves of the Neverslette

plotted on

UT

National

Grid.

NEVERSLETTE

SYSTEMET, Lower Glomdal, Rana, Norway.

75800

50800 +

76000

+

K j ~ k k e n v a s k e n

50800

+

Water

flow

(Dye tested)

200

+

o

metrea

200

400

600

+

800

6 0 0

G a u d a g r o t t e ~ ?

~ ~

~

~ ~ . ~ ' D o l i n e g r o t t e

---' Trollholet

DMD-grotte

+

800

Grid Interaectlons

at

200m

Intervals on UTM grid.

Plotted from surveys to BCRA Grade

, ,C

y UEA Speleological Expedition 198' ', drawn by S.H.Bottrell

122

8/21/2019 BCRA 14-3-1987


Loc ~ . J r . : o .

0 1 g . 4 >

Survey

· ~ d e

5m

S h o r t

c h o ~ e d

s e q m e r l ~ s

of

r1hr-e A t l c

C.. J

.

v e ~ ~ s s 2

Ol d phrea t Ic ~ a s s a q p Invaderl

t

Jy

I

noder

n

w ~ y .

9

10

1 1

1-

4,-,

Choked r t

l r l g r y ~

Dry

c a v e ~ b o v e · e s ; L ~ r l r

Cho ke d r i

ft

In gr y le .

Anneksgrot

. ~ ?

(see

t e

' :

t.)

.

1:: , 1 ·4

15

16

4rn deep

3m

~ , m deep

T m

:=Jm

3

r

)m

C h C ) ~

e n t r c n \ c ~ ~ s to phr p a t 1

r

t U \ I : - ~ S

I S I . J r o t t r ~ - p V l r ~ : : : , I , l

n o \ ' l n

phrea t i c r ~ l n n a n f l c)()r ed w t h I c e .

~ r g e

remrlant

o f

p h r e ~ t l c

l b ~

n e t ~ J o r ~ with

s e v e r ~ l ( l t h e r

s m ~ l l e r

~ n t r ~ n c e 5 .

I'le'/E'r

1

t t e l n ~ t. C

i.:lVr:.:s :

22m

95m

2

r

) m

20m

I'IS1

4c 1.1.

1

)

010

Ph r eat ic

- e m n ~ n t : . .

F'hre lt .

c: cavE'

r - e m n ~ r ) t .

F hreC\t lc

c<,J.vp

r emn,3n t.

Ph r e a t i

c:

cave

r

emnant..

F'hre.:.;..tir::

c:

. . : ~

\'

. . ~

r emn-3n

t

F·

h

rea :.ic

cavE.'

r'

emnant

F h r ea .t C

V r : - : > rE'mn. :

\nt .

o k k e n - v a s k e

NS1A

4c

G ~ l l e r - F r ~ n c 2 l s ~ n t r ~ n c ~ .

N:3

4,-

tJ

33

c

11:;4·

1-c:

r·j·35

4c:

:;:2

m

1

::5

18: 3.0

29m

OMD -

qro t t e

.

Tro l

lh

o l e t .

n·.3-ud

ag

r

o t t.

a .

Dol

i roegr

o t t a

.

These

caves form a very in te res t ing drainage

with i n

easy walking

d i s t anc e of

Fiskjornmoen.

wi

th

the excep t ion o f the s lop ing rift

streamway

in Kjokken-vasken most of

the caves

are

fo ss i l

phrea t ic passages which have

been

invaded by the

present streamwa

y .

Two

othe r ac t ive

st ream

systems were found

in

the

a rea , one ANNEKSGROTTE (77m, UTM507766,

loc .12

on Fig .4 ) was ca r ry ing

no

st ream in 19

84

and was

consequently not

dye- te s t ed .

In the wetter

w

ea the r

of 1986 it

took

a

smal l s t ream.

The cave

i s

in

two

ha l

v

es

upst ream

and downstream from

a

la rge nor th -sou th rift

on

sur face .

The

upst ream

passage s t a r t s

from a

chamber

a t a s tooping

he igh t

and

becomes progress ive ly smal le r un t i l fur ther

progress in the

r azor - sha rp

marble becomes e i t he r

too

wet

or too painfu l ; t h i s

sec t ion

of the

ca

ve

appears to be

en t i r e l y v

adose i n o r ig in .

In the

downstream cave the passage con t r a s t s completel

y ,

af t e r

descending the surface rift

a

l a rge keyhole

sec t ion

passage

soon

leads

to

a chamber wi t h a

sump pool .

Attempts t o f r ee -dive

the sump

revealed a

la rge completely

f looded passage

beyond.

The o

ther

a c t i ve

streamwa

y f o

rms an

ob

v ious

connec t ion between s inks

a t UTM507767

and

the resurgence 100m nor th a t UTM507768 and

can

be

en te red

via

two smal l caves

between

( loca l i t i e s 7

and

8

and Fig. 4) , where

the

st ream

flows through

old

phrea t i c cave

passage .

Foss i l phrea t i c ca

ve

segments

77

+

metres

76

Rivers

~ ?

Stream sink

Table

1.

Cave lo cal i t i e s

in lower Glomdal .

Figure 4. Detai l

ed

map of

study area showing

cave

l ocal i t i e s noted,

16

17 , 20

1 ~ ~ § t 21

19 ..22

23.

NS2

7

Streams

q V

Resurgence

A

number of sec t ions of old phrea t i c ca

v

es ,

now

dry, were l oca ted and surveyed

during the

exped i t ion . Some of these w

e re

previous ly known

and where

poss ib le

the loca l Norwegian names are

used for these . Most were

s ingle or

branched

tubes , 1m to 3m in

diameter , with l eng ths v arying

from a few metres

to over 15 0

m.

Entrances are

formed

w

here su r face e ros i

o n

has i n t e r sec t ed the

cave passage, and the cave segments e i t he r connect

two

or

more such en t rances or

t e rmina te a t

chokes

or s t a t i c

sumps.

These caves r e l a t e to an ancient

hydro log ica l

regime

and are now abandoned with

varying

degrees of sediment

i n f i l l .

Attempts to

es tab l i sh the

pa leo-cur rent d i r ec t i on i n

these

passages from the morphology of sca l l o

p s

on

the

wal l s was

of ten

inconclus ive , bu t

t he re i s

a

genera l t r end to southe r ly flow d i r ec t i on i n

those

cases . ,here it could be es tabl i shed . All such

caves

found in

the no r the rn

pa r t

of the lower

Glomdal are

sho

wn on

f igure

4 and l i s t e d i n t ab l e

1 .

150

Contours

in

metres

G::)

Closed depression

10

Locality

No.

(see Table 1)

123

8/21/2019 BCRA 14-3-1987


DYE-TRACING EXPERIMENTS

Fluoresce in and

Rhodamine dyes were used to

es tabl i sh

hydrologica l

connect ions in

the

act ive

stream systems.

In

many

cases

a vis ib le pos i t ive

conne

ct ion

wa

s made, but ac t i vated charcoal

detec tors were suspended

in

the

water f low

and

subsequently

used

to confirm

resu l t s

(Atkinson

and

Smart, 1981).

Mo

s t of

the

t e s t s were concerned

wi th the Neverslet te-systemet streamway, but one

w

as performed

bet

w

een

the

surface s

tream

in

the

nor th

-eas t

of the area and

resurgences

on

the

s outh s ide

of

Glomdalsvatnet. All of

the

resu l t s

are

summarised

schemat i ca l ly

in f igure

5.

Neverslet te-systemet

The

connect ion between the

caves

of t he

Neverslet te -systemet was proved conclusively

and

a

t r i bu t a ry

to

the

system was t raced

from the

surface stream

s inking

a t

UTM508761 to a

s ide

passage in DMD-grotte. The

flow t ime

through

the

system

was

somewhere between 2

and

12 hours from

sink to resurgence 11 00m)

and the

v is ib le

pulse

passed quickly,

indica t ing

tha t there are no very

large dead

volumes of wa te r

in the

sumps or

between Dol inegrro t te

and the

resurgence.

NEVERSLETTE-SYSTEMET

UTM508761

S ~ 3

UTM

UTM

507788

DG GG

TH

DMD

KV

507757

.1

.1 .1 .1 .1

1

.2 .2 .2 .2

2

• 3

3

3 3

3

SURFACE STREAM

UTMII087811

. ~ ~ ' ~ ~ ' ~ ' ~ '

U T M : o : ; ~ . , L

. 7 -b

: : i

~ r ' - J J I J J

UTM

, , I

J

O ~ I I ~ V ~

UTM1I07787

P u r ~ : ~ r b l e band.

* Dya Injection point • Po.lllve te.t o Negatlv. t• •

Figure

5 . Schematic

representat ion

of dye tracing

resul ts

in

the

Neverslette

Systemet and

surface

streamway. G

=

Dolinegrotte, GG

= Gaudagrot te ,

TH Trollholet , M

=

M -

grot te ,

KV

= Kjokken-vasken.

Surface streamway

In

the

nor th-eas t of the

area

a

stream

f lows

from a water fa l l

a t

UTM507757 to sink

a t

UTM50 7767

in high wa te r

it

overflows to

the

small lake a t

UTM507768) .

Water in t h i s s tream was t raced to

two

separa te resurgences

a t

UTM507768 and

UTM508679 Fig .4) . One

route

from the

s inks

near

UTM507767 to

the

resurgence a t UTM5077768

i s

obvious but the route followed

to

the

resurgence

a t UTM508769 i s unclear .

The stream f lows through

a

small

gorge

in

a

pure marble

band

from

UTM508765

to UTM508765 to UTM509763

and

t h i s

band

can be

t raced northward to

the

resurgence. I t i s thought

tha t

par t

of the flow of the

s tream

may

follow

an

underground

route

in t h i s

marble band, though

no

s

in

k could be found. Flow t ime from

the

water fa l l

to both resurgences (1200m) was bout 1 hour.

No

connect ion

was found between

t h i s s tream

and the

resurgence in

the

dry

r iver bed

a t

UTM506768;

t h i s

resurgence may be fed by drainage from the marshy

ground to the west .

ACKNOWLEDGEMENTS

The au thor thanks Nigel Crisp, Ian Kitching, Chris Smith

and K

aren

Beatt ie for extensive

help

and cooperation in

the

f ield

and

during

preparation of

th is paper . We would

a l l

l ike

to thank Stein - Erik

Lauritzen

and Tim Atkinson for

the i r

ass is tance

at

various stages of

this

work. We

also thank

a l l

of

the following who generously

provided f inancial or other

assis tance,

without

which

the

work

in

Glomdal

could

not

have

been completed.

1984 UEA Expeditions Committee, Sir

Phil l ip

Reckitt

Educational

Trust, Gilchr is t

Educational

Trust,

Fed

Olsen

Lines, Colm

ans

of

Norwich,

Tate

and

Lyle,

UEA

Student s Uni

on

Se r vices,

UEA

Students

Union

Exploration

Society,

R.E .

Thorns

and Co, Morning Foods, Read and Woodrow Ltd., Farley Health

Products, Trident Maritime.

1986

Bil l

Bishop Memorial Trust, Sir

Phil l ip

Reckitt

Educational Trust, Bri t ish Geomorphological Research Group .

REFERENCES

A

tkinson,

T.C.

and

Smart, P.L. 1981

Art i f ic ia l t racers in

hydrogeology . In A

survey of

Bri t ish

hydrology,

1980

pp.173 - 190 Royal Society, London .

Hjorthen,

P.G. 1968 Grotter og grot teforskning i Rana.

NGU

Smaskrif t No. 9 Oslo .

Lauri tzen S -E . Abbot

J .

, Arnesen R. Cross ley Go

Grepperund , D

. ,

Ive,

A

and Johnson, S. 1985 Morphology

and

hydraul ic s

o f an ac t i ve phreat ic

conduit .

Cave

Science 12;

139-146

.

St . P ierre D. and

St . Pierre

S.

1969

Cave

s

of

Rana

Nordland, Norway . Trans. Cave Res. Gp . G.B., 11; 1-71.

Recei ved

June

1987

Simon Bo tt re l l

Department

of Earth Sciences

University

of Leeds, LS2 9JT

124

8/21/2019 BCRA 14-3-1987


CAVE SCIENCE Vo l . 14,

No.3 ,

December 1987

Transact ions of

the

Bri t i sh Cave

Research

Associat ion

aves and Mi

nes of

Hawkstone

Park

Salop

Terry MIDDLETON

Abstract

:

Hawkstone

Park

contains

a

number

of impressive outcrops of

Tr iass ic

sandstone which

have, over

the years

, been worked

by

man resu l t ing

in

the product ion

of

both

caves

and mines.

These

workings have been

described and where

accessible , surveyed

and a summary of the his tor ica l

and

geological

se t t i ng of the s i t e has been made .

Set within some of the most picturesque

scenery

in

Shropshire

Plain ,

Hawkstone

Park

provides

a fascinat ing and indeed memorable

local i

ty for a

v i s i t .

I t i s of in te r es t to the

speleologi

s t because

of the number of caves which

l i e within i t s boundary . Unlike the natural

caverns of kars t areas, the caves of

the

Park

occur

in

sandstone and most

of them

are

man-made

having

been constructed dur ing the 18th century

purely for the enter tainment of vis i to r s . The

Park

also

contains

to a l imi ted ex ten t

evidence

of

mine working

and

t h i s

together with the

caves

forms the

basis

of th i s

report .

The

Park

, which provides a superb

set t ing

for

a golf course, i s located some l lkm

north - north - eas t of Shrewsbury and km

west - north-west of

the

vi l lage of Hodnet and can

be approached from e i ther

the

A49 or

the A53.

Three h i l l s and an

escarpment

r ise above and

provide a

sharp

cont ras t to the ro l l ing ter ra in of

the

golf

course

and these

are

known respect ively

as Grotto

Hil l ,

Red Castle

Hil l , Elys i

an

Hil l and the Terrace. At the i r

highest point ,

the

wooded slopes r i se to a

maximum elevat ion of 208m a t a spot marked by the

Obeli

s k on t

he

Terrace . All

of

the land wi

th in

the

Park i s private

and anyone wi s hing to

v i s i t

the loca l i ty must

f i r s t

obtain permission

from

the

manager

of the

Hawkstone

Park

Hotel .

GEOLOGY

The rocks of the

Park

are Tria s s ic in age and

comprise, in chronological orser ,

red

Bunter

Upper

Mottled Sandstone,

buff

- co l oured Lower Keuper

Sandstone local ly

known as

Grinshi l l Sandstone

and Keuper Waterstones Pocock and Wray,

1925).

Under modern

s t r a t ig raph ica l nomenclature the

Grotto

Hill viewed from the Red

Castle area. Note

the

path

which traverses the base of the

c l i f fs and

the ruined arch which

is

s i tuated by

the

summit

at

the

southern

end of

the

hi l l . The

wooded slopes of

the

Terrace can

e seen on

the

r ight

hand side

of the photograph.

25

Lower Keuper

Sandstone

and the

Bunter

Upper

Mottled Sandstone

form

the uppermost uni ts of

the

Sherwood

Sandstone

Group

whilst

the Keuper

Waters

tones form the

basal

bed of the Mercia

Mudstone

Group

Warrington

e t

a l

, 1980) . Good

c l i f f sect ions of a l l three l i tho logies are

displayed

a t

Hawkstone. The sandstone uni t s ,

which are

extensively

cross - bedded , dip gent ly

a t

an

angle of 6 degrees

in

a north north-wester ly

di rec t ion .

In

consequence ,

the

older , mainly red

sandstone

s

outcrop

a t

the southern

end

of the

Terrace and a t

Elysian

Hil l

wh i

ls t the

over lying

Grinsh i l l

sandstones,

of

which

over

45m a re

exposed,

are seen a t the northern

end

of the

Terrace and a t

Grotto

Hil l . Red Castle Hil l

comprises a small out l ie r of Grinsh i l l s andstone

rest ing

upon the red sandstones

which

continue

beneath the undulat ing land

scape of

the

golf

course. A

good

sect ion of Keuper Waterstones i s

exposed i n the ravine a t

SJ582293

.

Fa u

l t s bound

the

north -

west

and north -

eas t e r ly edges

of Grotto

Hil l and

the

Terrace

see

Bri t i sh Geological

Survey 1: 63360

sheet

138 Wem;

1967

and explanatory

memoir Pocock and Wray

1925).

The Grinshi l l

sandstone, which i s

local ly

a n important building

stone

in

the

area, contains large numbers of

small

res is tant

c rys ta l l in

e nudule s

whi

c h

were

desc r

ibed

by Pocock and Wray as nests of b a r i t e crys ta l s .

However

the nature

of

these

crys t a l s

or

pseudomorphs?) i s d i f f i cu l t to confirm

without

microscopic

or

chemical analysis . These nodules

are

also

pre

s en t , though less abundant, in some

of

the outcrops of the under lying red sandstones .

Both sands tones contain subangu l ar grains and in

several areas

the

s i l i ceous matrix of the bedrock

has

been replaced by

a

calcareou

s cement.

The

presence of the

crystal l ine

nodules, the

var ia t ions in colour of the sandstone,

together

8/21/2019 BCRA 14-3-1987


with the

f luctuat

i ng nature

of

the cementing

mater ia l suggest

that

the

sediments

/ rocks have

been subjected to both penecontemporaneous and

post-diagenet ic changes.

Green

copper

ores of low

grade

may be found

in the

upper

uni t s of the Grinshi l l sandstone.

These have

been

named

as

malachi te

by

previous

authors (Murchison, 1839; Pocock and Wray,

1925;

Dewey and Eastwood 1925; Carlon, 1981)

however

t

should

be

noted

that

dur ing recent f i e ld t e s t s the

ores

f a i l ed

to

show

any

react ion to

d i lu t e

hydrochloric

acid No

deta i l ed chemical

examination

of

the ores

i s known to

have

been

carr ied

out.

Blue copper

minerals , pos s ib ly

azur i t e occur loca l ly in very small amounts. The

ores tend to be found impregnated between s l igh t ly

coarser

uni t s within the

cross-bedding. The

mineral izat ion does not

appear

to be

associated

with, or be loca l i sed by,

fau l t s t ructures .

THE MINES

Although the

copper

ores are of low grade

there

i s

some evidence to

suggest

t ha t they may

have

been mined or

that

t r i a l s were car r ied

out .

Many au tho r i t i es have

suggested

that

Hawkstone

Grotto (SJ.573297)

was a Roman copper mine

(Watkin,

1879;

Kenyon, 1892) and i f

so

t

i s

the

only

mine

a lb ie t

modified subsequent ly,

t ha t

is

accessible in the

Park

today.

I t

has also been

suggested

that

the Roman v i l l a

found

a t Weston

could

have

housed a mine of f i c i a l (Kenyon, 1892) .

However

these

ten tat ive

connect ions

st ll

have to

be proved . Never t heless there i s

ample

evidence

of

the

Roman

occupat ion

in

the surrounding area

.

Roman coins and

bronze implements have

been found

in the r u ins of the Red Castle and more Roman

ar tefacts have been

discovered a t the

h i l l

f

or t of

57

Bury

Wall.s

which l i e s 2km to the south

(Watkin,

1879; Kenyon, 1892). A hammer pick of possible

Roman or ig in has been

discovered

a t

Hawkstone

(Davies, 1935) and the Roman road

connecting

Wroxeter

(Viroconium) with Chester (Deva) was

described in the Victo r ia County History (1908) as

climbing the western extremity of

Hawkstone

r idge ' . There i s also plenty of evidence to

confirm that

the Romans were

ext rac t ing

lead

from areas

nearby

including

Shelve

H il l and the

Stiperstones .

Pigs of

lead

bear ing the

inscr ip t ion

IMP. HADRIANI. AVG (Imperatoris

Hadriani

Augusti) have been found in

the

Snai lbeach area of the

St ipers tones j u s t

south of

Shrewbury

(Davies, 1935).

The workings a t

Llanymynech

Hil l (Adams, 1970),

which

are also

thought to

have

been worked by the Romans

for

e i ther

lead or copper, are l ikened

in

the Victo r ia

County History (1908) to those a t Hawkstone.

However

the

proposed

s imi l a r i ty does seem a l i t t l e

dubious The grot to a t

Hawkstone

was

subsequent ly

enlarged

for

recrea t iona l purposes

and as a

res

u

lt

for the

purpose

of th i s account i t s fu l l

descr ipt ion

has been included in the

sect ion

on

caves.

Both

Murchison (1839)

and subsequent ly

Carlon

(1981)

make reference to a shaf t on the Terrace .

The shaf t is said to be located a t SJ. 576295 and

was

described

as being

1.5m

in

diameter

and

capped

by an iron gr i l l e with sandy spoi l nearby. This

area

can be reached by

following the footpath

up

the h i l l from C7 and C8 to the Terrace. The shaf t

should be located in the area immediately north

eas t of the

junct ion

of th i s footpath

with the

main path along

the top

of the

Terrace

. This

region

i s covered

with rhododendron s a nd the

undergrowth i s so

dense

t ha t a person could walk

d i rec t ly over

the

g r i l l e

on

the shaf t without

Grotto

Hili

T H E

C VES

OF

WKS

TONE

P RK

SJ5729

29

Key: .... l i f f

S t

oo

p s

lo p

Ro

od

< o _ Fo o t path / ... Wood bou nd.r v

57

126

N

.....

/

_ i

o Bu ild ing

• C

ave

29

Figure 1. Map showing the

locat ion

of the caves

found

within the

boundary

of

Hawkstone

Park.

(Clif f

detai ls taken

and

enlarged

from the 1:10560

Ordnance Survey map SJ52NE.

Cave

locat ions

mapped by

author

-

May 1987).

8/21/2019 BCRA 14-3-1987


real i s ing i t

A search for the shaf t during t h i s

survey proved f ru i t l e s s . Disturbed

ground

composed

of

loose sandy mounds and tunnels

cover ing

an area of approximately 180 square

metres was

found

wi th in t h i s

region of

woodland

close to the

boundary

with the adjacent

f i e ld .

However t h i s

dis turbed area

appeared

to be no more

than an extensive

badger s

set The shaf t was

l a s t seen

i n

1952 (Carlon,

1981) and

t

i s

unl ike ly to

be

re loca ted unt i l

the rhododendron

bushes

are

thinned.

More workings occur j u s t beyond

the

boundary

of the Park

a t SJ.587292 (Pocock and Wray, 1925;

Carlon,

1981). The

workings included a small

shaf t

3m

deep connected to an 2 . 4m

level

dr iven

to

the

south

eas t

which in turn l inked to a 6m level

running

south - south-west

to i n t e r sec t

the slopes

of the Terrace. The workings, which were

i n f i l l e d

in

1977,

were sa id to

contain deposi ts of

azur i t e

and malachite .

These 2

sets

of workings

(SJ.576295

and

587292), both

of

which are located on the

Terrace,

seem to

have

been l i t t l e more than t r i a l s . I t

i s

di f f i cu l t

to

date

them,

but

they may

have been

excavated dur ing the 17th or ear ly 18th century

pr ior to the purchase

of

the land

by

the Hil l

family in 1737 (Carlon, 1981). Documentary

evidence of the

dates of

the mining a t nearby

s i t e s

may

be

re levant to

th i s

quest ion.

Sources

s i ed by Warrington (1980)

indicate that

mining

occurred

a t

Weston

and around

Redcast le before

1740,

and that the

par tnership

was formed

to mine

in the townships of Weston-under-Redcastle and

Wixhill in 1697. This ac t iv i ty probably

took

place

a t the

s i te s

noted

by Dewey

and Eastwood

(1925) and

Pocock and

Wray

(1925) between

1

and

2.5km south - west of

Red

Castle

Hil l .

Some

working

a t

wixhi l l

also took

place

dur ing the

mid-19th

century

(Pocock and Wray, 1925).

Carlon

(1981)

also

re fe r red

to leases being

granted in 1897 and

The

arch C3) and the

footpath

at the base

of Cli f f s on Grotto Hill .

1698

for

mining

a t Wixhi l l .

As noted

by

Warrington

(1980),

the documentary

evidence

of

working

of

non-fer rous

ore deposi ts

in

Cheshire

and

nor th Shropshire

in the l a t e

1690 s

contrasts

with the

lack

of such evidence

before

tha t time

and i s

consonant

with the

development

of

i n t e res t

in them af te r the passing of the Mines Royal

Acts.

I t would, therefore , seem reasonable to suggest

tha t

the

mining t r i a l s

found

within

Hawkstone

Park

were dr iven

a t

the

same time

as those of s imi lar

loca l i t i e s

in

the

vicin i ty .

THE CAVES

Much of the landscaping of the

Park

was

car r ied

out dur ing the

18th century by i t s owner

Sir Rowland

Hil l

(Oswald, 1958).

During

t h i s

per iod a number of

cu r io s i t i es

were contructed for

the

benefi t

of

vis i to r s including the Obelisk, the

White

Tower,

a

l ion ' s

den complete with

stone

l ion,

a

hermitage and

severa l caves and

rock

shel ters .

A t o t a l of 11 such caves can

be

found

a t Hawkstone ( f igure 1) and although a l l but

one

of

them

are very

smal l , they

are nevertheless of

i n t e res t as tokens

of the

lengths

to

which the

landowner

was

prepared to

go

to add i n t e res t to

his

Park.

The

caves

can be subdivided

in to

3

categor ies :

1) Former mine

workings

t ha t

were

extended or

enlarged dur ing the

18th

century.

2) Tunnels not associa ted with mining tha t

were constructed

dur ing the

18th

century.

3) Natural alcoves or

recesses in

the

cl i f f s .

These natura l alcoves have been included in

th is

repor t

because

they were named as

caves in

the

old tour i s t guides to the Park and were

mentioned in t he i r i t i ne ra ry . Typical excursions

are

descr ibed

by Leach,1891; Davies, 1894; Haslam,

1985; and in the

Hawkstone

Handbook

for

~ 9 3 8 The

s i t e of each cave i s

shown

on the loca t ion map and

plans of the

larger

systems ( f igures 2 and 3) have

been included with t h i s ar t ic le . A number

of

caves are marked

on the

1: 10560 Ordnance Survey

map (SJ52NE). All are descr ibed below:

C

(SJ.57142940):

s i tua ted

a t

the

western

end

of

the obvious

eas t /wes t

t rending

gul ley on

Red

Castle

Hil l .

I t

comprises

a s ingle

curved

passage

12m in length ( f igure 2) which l inks the gul ley

with the

west-south-west

face of Red Castle Hi l l .

The

tunnel , which is cut

through

red sandstone, i s

0.8m wide and

1.8m

high and

bears a mu l t i tude of

pick

marks.

I t i s also gent ly

incl ined

to the

west

and t rends in an eas t /wes t di rec t ion .

C2

(SJ.57082963): located

in

a small

i so lated

pi l l a r

of

red sandstone

a t

the northern edge

of

the

r idge

of Red Castle Hil l .

I t i s

j u s t 5m long

and

i s S-

shaped in

plan

with

an entrance a t

e i ther

end. I t t rends

in

a nor th west/south

eas t

di rec t ion . The passage

is 1.8m

in

height , 0.8m

in

width

and

is

man - made (see f igure 2) .

C3

(SJ.57252970): a shor t

man-made

arch

5m in

length

with benches

hewn out

of the rock

on

e i ther

s i

de.

I t can

be entered

by

fo l

l owing

the

path

which sk i r t s the south west facing c l i f f of Grot to

Hil l (see

photograph

1) .

The

arch i s 1.2m wide,

2m high and has been driven through a

buff - coloured sandstone .

C4

(SJ.57232985): located a t the northern end

of

Grotto Hil l

and

has 2 entrances.

I t i s bes t

entered

from the

narrow

ravine which cuts

through

the dip slope of the

escarpment. The

cave i s

man- made and l inks

in

an

eas t /wes t

di rec t ion and

i s 18.5m

long. The low

entrance

s i t ed in the

ravine leads to a small

ci rcu lar

chamber with a

cent ra l

pi l l a r and window .

From

the

chamber,

the

passage

becomes

gradual ly lower

and

f ina l ly

emerges

in to

dayl igh t

(see

f igure 2) .

The

l a t t e r

par t of the cave i s wet with much mud on the

f loor .

127

8/21/2019 BCRA 14-3-1987


Hawkstone Caves

Ravine entrance

HAWKSTONE

GROTTO,

Plan

SALOP.

N

C8

SJ

57 9 969

BeRA Grade 50

PL N

-'/

Iab

~

Mag

N

o

10

metres

Key

/

Steep

cliff

Window

Entrance

Pil lar

o

Skylight

•

Blocked skylight

Surface arch

Joint

Surveyed by O John , D. Luke , T.Middleton , February 1987 .

C5

- Hawkstone

Grotto

SJ.57302970):

s i t ua t ed a t the summit of

Grot to

Hil l ,

it can be reached by e i t he r

proceeding

along the rav ine

to i t s

southern

end

or

by fol lowing one of the pa ths up the d ip s lope o f

the escarpment . The

cave

has 4 en t rances

and i s

by fa r the l a rges t

sys tem

in the Park

and

a

de ta i l ed survey

i s

presented he re in

( f igure 3) .

Deposi ts of green

and

a

little

blue

copper

ore

can

be

seen impregnated in the

bedded

buf f -co loured

sandstone

un i t s .

The

passage

which

leads

i n to

the

system from the r avine

i s

2m high and m wide and

t r ends in a nor th wes t / south eas te r ly d i r ec t i on

( f igure

3) .

Oswald 1958) s t a t ed t ha t t h i s

passage was cons t ruc ted

for

Si r Richard

Hi l l to

connect with the

main

pa r t of the G ro t to . The

r e s t of the cave appears to be much olde r

than

t h i s passage and, as a l ready

ment ioned,

a number

of au tho r i t i e s have

suggested

t ha t

it

may have

or ig ina ted

as

a Roman Copper

Mine.

The cave

was

v i s i t ed by

Dr

Johnson during h i s

t r ave l s

Oswald,

1958) .

Soon a f t e r enter ing the cave from the rav ine

the

former

posi t ion of the doorway

can be

seen

cut

i n to the passage

wal l s .

Although

the p le thora of

pick marks in the

passage

tend

to

obscure the

f i ne r geologica l d e t a i l s , a number of near

ve r t i c a l

sediment

f i l l e d f i s su re s can

be

seen on

route to the Labyrinth .

The

pos i t i on

and

t r end

of

these

j o i n t s

have been

shown on

f igure

3.

Small

amounts of water

i n f i l t r a t e

down the penult imate

j o i n t j u s t pr io r

to

the junc t ion with the f i r s t

chamber.

A

hole

on the r i gh t l eads

i n to

the f i r s t

128

20

. - ~

\

C4

Figure 2. Surveys

of

some of

the

smal ler caves

found

with in

the

Park.

(Detai ls recorded by

author

January - May

1987).

TCM Figure

chamber ( ' A

on

f igure 3) which i s approximately

c i r cu l a r and has a diameter of 5.6m. The chamber

con ta ins 2 main p i l l a r s and 2

skyl ight s

one of

which i s blocked. These

support

p i l l a r s vary in

s i ze throughout the cave. However, fo r c l a r i t y ,

they have

been shown on the

survey

as

being

equa l - s i zed. At

the no r th -wes te rn

s ide of the

chamber

l i e s

a

shor t s emi -c i r cula r passage around

a fur ther p i l l a r . The so l id sandstone a t the

en t rance to t h i s

sec t ion

has been carved

to

resemble

a

s tone arch.

At

the

f a r

s ide of

t:1e

chamber a passage leads in to the

Labyrin th

whi l s t

the or i g i na l en t rance passage continues i n to a

second c i r cu l a r chamber ( . B •

on f igure

3) with a

diameter

of

7.

3m.

t

has 4 support

p i l l a r s and

one of these in pa r t i cu l a r has been beau t i fu l l y

carved and f lu ted . Out of the 3

skyl ight s pre sent

only

1 i s

open.

At the no r the rn end of the

chamber a shor t passage l eads

out

i n to dayl ight a t

the

edge

of the

c l i f f

( take ca r e ) . The chamber

a l so

has a

smal l a lcove on its western

s ide . A

passage continues from

the

southern

end of the

chamber i n to the Labyrin th .

The Labyrin th i s a

most

impressive

cavern

with

sha f t s

of

l i gh t ente r ing

from

skyl ight s and

windows

a t

the western end. t

i s

28m in l ength

and 9m in width and con ta ins

seve ra l

rock p i l l a r s .

At

the western

end

a passage l eads ou t onto the

edge

of

the

escarpment

and

the pa th may

be

fol lowed to a carved viewing area .

In

seve ra l

p laces the uneven wal ls of the

Labyrin th

appear to

have

been coated with a l aye r of

green

pa in t

( a l so

noted by Jones,

1974) . At the

southern

end of

the

8/21/2019 BCRA 14-3-1987


The southern entrance

area

o f

Hawkstone ro t to Note the

remains o f a roo f on the

back

w a l l s .

cavern a number of route s l ead through o ld

doorways i n to the Grotto . The fol lowing

descr ip t ion

of

the

s i t e appeared

in

The County

Sea ts of Shropsh i re ' (1891);

the v i s i t o r i s in troduced

to

the magnif icent

Gro t to ,

a vas t

subterranean

cave,

in the midst

of

which i s a spac ious recess f an t a s t i c a l l y

i n l a i d

with

a grea t va r i e ty of s he l l s , f o s s i l s

and othe r cur ious pe t r i f ac t i ons

Or i g i na l l y

the windows of the

Grot to were

f i l l e d

with

painted g la s s , unfor tuna te ly however,

none

of

t h i s remains today. During the 2nd

World

War the Park

was

u t i l i z ed f i r s t by an American

mi l i

t a ry camp, then by a German Prisoner of War

camp

and

f i na l l y by Br i t i s h t roops wa i t ing

for

demobi l iza t ion

.

During

t h i s

per iod

much

of

the

ornamentat ion in the

Grotto

was

destroyed

(J .Jones

pers comm .

) . Nevertheless

some of the decora t ions

still remain and they are

be s t

d isp layed a t the

eas tern end

of

the

Grot to

where numerous s he l l s ,

mainly

l amel l ib ranchs

and

gastropods,

have been

s e t in a cement /plas ter

matr ix

toge the r

with

much

furnace s lag . The remains

of

a t i l e d roof can be

seen in places . The

ru ined arch

which s tands

a t

the top of Grotto Hi l l (ye t another

fo l ly)

l i e s

The

Labyr

i

nth

in Hawkstone

rotto

129

d i r ec t l y above the main passage

which l inks

the

Labyr inth

with the Gro t to . The westernmost

passage

l inking

the Labyrin th with the

Grot to has

a beau t i fu l l y arched roof and it w

as

in t h i s

area

t ha t ba t s ,

a t

one t ime, were qui te common .

Un

for tuna te ly

none o f the

cave

s in the Park

appear

to be used

by the

ba t s today.

Most

of

the

ex i s t i ng cave fauna can be found in

the

blocked

skyl ight s and

the

ca ve spide r

(Meta

menardi)

and

the

Hera ld Moth

(Scoliopteryx

l i ba t r ix ) are

the

most not iceable

spec ie s .

The

Gr o t to may ha

ve

been occupied by a hermit

pr io r to the

18th

cen tu ry and, i f so t h i s could

have

been

the

rea s on why

the

Hi l l family l a t e r had

an

a r t i f i c i a l hermitage complete w ith au t omaton (a

pap ie r

mac

he f igure worked by le v

e r s )

b u i l t with in

the Park.

C6

-

The

Retreat

(SJ. 57372973): a natura l alcove located

a t

the

base of the c l i f f s immediately ad jacen t

to

the

foo tpa th which leads up onto the Terrace

( f igure

1) . t

i s

found

in

a

buff

-

coloured sandstone and

i s only

m wide

and 4m high.

t

i s

i

ncluded in

t h i s r epor t because it was desc r ibed as a cave in

the i t i n e r a r i e s

of

the

Vi c t o r i an

t ou r i s t guides .

A f i s sur e i n f i l l e d with a mottled ca lca reous

f ine

-

8/21/2019 BCRA 14-3-1987


The carved

arch

in

Hawkstone

Grot to

grained

sediment

can seen

This

l ine

of

weakness

development

of

the cave .

C7

a t

back

probably

of

the

aided

cave.

the

(SJ.5 7462957): s i tua ted

in

an i so la ted tower

of

buff - coloured sandstone known both as the Fox s

Knob

or

the

Fox s

Head which was named af te r the

inc ident whereby

a

fox

together with

i t s

pursuing

hounds

leapt

to

t he i r

deaths

from

the

top of

the

outcrop (Leach, 1891; Oswald, 1958) .

The

cave i s

7.8m long and

has

2 entrances which are l inked by

an

S-shaped passage

(see f igure

2) . The

s t r a igh tes t

sec t ion

of cave

follows a

north

eas t / south west

t rending

f i ssure . The passage

i s

2m high

and 0.9 m wide and

descends

gent ly towards

the

nor th

west. There are

no

pick

marks in the

cave but depos i t s

of

copper

and

i ron

ore

minerals

are

evident in the surrounding rock.

C8 - St Francis Cave

(SJ.57492954):

s i tuated immediately

adjacent

to the footpath.

The

cave takes a sinuous route

which t rends

in

a north - north -

eas t / south

south -

west

di rec t ion to

emerge

af te r 18m a t the

far side

of

the r idge ( f igure 2). The passage

which

contains

many pick

marks,

i s

2m high and 1m

wide

and i s inc l ined towards

the

south west.

The

i ron

hinges of

a

former door can

be

seen

se t

in to

the buff-coloured sandstone

by

the lower entrance.

A number of insec t s in

par t i cu la r

cave spiders and

Herald

Moths can be seen cl inging to crevices

between the

roof

blocks .

C9 - Reynards

Banqueting

House

(SJ.57772911):

a

natural

alcove

5m

high,

6m

wide and

2m deep

which has

developed along

a

fau l t

in red sandstone ( f igure 1) . I t i s included in

t h i s

repor t

because

t

was described as a cave

in

the

Victorian tour i s t l i t e r a ture . I t s name

derives from the numerous bones

of

hares, rabbi t s

and poul t ry

which have

been

found

in

the alcove

(Leach, 1891).

CIa

(S J . 57582895): located

in

the red

sandstone

of Elysian

Hil l .

I t is found

in

a low outcrop

which

faces

eas t and

is

approximately 10m above

the footpath which follows the bottom of the

val ley (f igure

1) . The c a v e ,

which i s 1.5m wide

and 2m long,

i s

man- made and

was

used for the

storage

of

tools when

the lower hal f of the val ley

was

a garden.

Cll

(SJ.57812905): located

in

an outcrop of red

sandstone

by the

footpath

immediately

below the

ruined

White Tower

( f igure 1) . I t i s a small

horse-shoe

shaped shel ter

with a bench cut

in to

the

rock

and t i s man-made.

The

cave i s

3m in

diameter and 1.7m high.

This

may

well

be the cave

which

was

mentioned

as being used as

a hiding

place for one

of

the

ancestors of the Hil l family

(Leach,1891).

CONCLUSION

I t

can be seen

from

the account

that

al though

the

caves and

mines a t Hawkstone

are

l imi ted

in

s ize

they are nevertheless of in te r es t and are

worthy of a v i s i t . I t i s unfortunate , however,

that

there

i s

present ly i n su f f i c i en t evidence

avai lable to

enable accurate

da t ing of

the mining

ac t i

v i

ty

. A

deta i l ed survey

of the

whole

of

the

Park i s a t present

being car r ied

out by Shropshire

County

Council

and t i s poss ib le t ha t t he i r

f indings may r esu l t in the s i t e being opened to

the

general public .

ACKNOWLEDGEMENTS

The author wishes to

thank

Dr T D Ford

of

Le ic e s t e r

University· and G Warrington

of

the Bri t i sh Geological Survey

for kindly reviewing

the

manuscript

and making a number

of

valuable comments. Thanks are

also

due to

Jack Jones

the

loca l guide

to

the

Park

Mr K Brazier the General Manager

of

the Hawkstone Park Hotel , who was kind enough to al low access

to

the

s i t e

and to D

John

and D Luke

for

as s i s t ing in

the

surveying

of Hawkstone Gr o t t o .

13

REFERENCES

Ano n. , 19 3

8.

The Hawkstone Handbook -

An

I l lus t ra t ed Guide

to

Hawkstone Park, pp21-24 (publ i shed

by

Wilding and

Son

Ltd,

Shrewsbury)

Adams,

D R

1970.

Survey

of

Llanymynech Ogof Roman Copper

Mine

. Llanymynech, Montgomeryshire.

Shropsh ire

Mining

Club

Account

No

8,

78pp.

Carlon, C

J ,

1981.

The Gallantry

Bank

Copper Mine,

Bickerton, Cheshire.

Bri t i sh Mining,

No

16,

pp14, lS .

Davies 0 1935. Roman

Mines

in

Europe

.

Clarendon

Press

pp159,160.

Davies, R 1894. Hawkstone Park

I l lus t r a ted

Handbook,

ppll ,12 .

Dewey,

H and Eastwood, T, 1925. Copper Ores

of the

Midlands,

Wales, The Lake D i s t r i c t

and the

I s l e of Man. Specia l

Report on

the Mineral Reserves

of Great

Britain Vol

xxx,

p21.

Haslam R

1985.

Rescue of a Masterpiece - Future of

Hawkstone Park,

Shropshire . Country Li fe , May 9th ,

pp1244-1247.

Jones , B, 1974 . Fo l l ies and

Grot tos, pp78-84.

Kenyon, R L, 1892. Bri t i sh

Shropshire .

Transact ions of

the

Shropsh ire Archaeological

Socie ty

Vol 11, p165,167.

Leach,

F (Editor ) , 1891. The

County

Seats o f Shropshi re -

Hawkstone Park

pp8 9 .

Murchison,

R I , 1839. The Si lu r ian

System,

pp38,39.

Oswald, A 1958. Beaut ies and Wonders of Hawkstone. Country

Life , Ju ly

3rd, pp18-21.

Oswald, A 1958. Beaut ies

and

Wonders of Hawkstone. Country

Life , Ju ly

10th, pp73-75.

Page,

W (Edi tor) , 19.08. The Victor ia History of the Counties

of

England. A

History

of the County

of

Shropshi re , Vol

1, pp266,417.

Pocock, R Wand Wray, D A 1925. Geology

of

the

Country

around Wem Memoir of the Geological Survey of Great

Br i ta in , pp40,41,91.

Rawdenhurst T 1784. nt iqui t ies o f Hawkstone. Gentlemans

Magazine, Vol LXVII, p377.

Warrington,

G

1980. Non-fe rrous

Mining in North Shropshi re

and

Cheshire. Shropshi re

Caving

and Mining Club

Journa l

(for 1979), pp9-20

(with

Addenda; ib id , for 1980, p13

(1981)) .

Warrington,

G

Audley-Charles ,

M G E l l i o t t , R E, Evans, W

B,

Ivimey-Cook H C Kent P E

Robinson

P L

Shotton

F

W

and

Taylor, F M 1980. A co r r e la t io n of Tr iass ic

rocks

in the Bri t i sh

Is l e s . Geolog ica l Society of

London,

Special

Report

No

13, 78pp .

Watts , WW 1939. Shropshi re . The Geography of

the

County

of Salop, p226

(publ i shed by Wilding and Son,

Shrewsbury) .

Watkin,

W T, 1879. Roman

Shropshire .

Transact ions of

the

Shropsh ire Archaeo log ica l

Socie ty ,

vol 2,

p322,353.

Received

June

1987

T Middleton,

3

Pool

Lane

Brocton

Sta f fo rd . ST17 OTR

8/21/2019 BCRA 14-3-1987


CAVE

SCIENCE

Vol. 14

,

No.3 , December 1987

Transact ions of

the Bri t i sh Cave Research

Associat ion

Leptospirosis

among

ritish

Cavers

C A SELF, W I ISKRZYNSKA, S A WAITKINS, J W WHICHER J T WHICHER

Abstract : Leptospi ros is , known

in

i t s most severe form as Weil s Disease,

i s

one

of the

few

se r ious

i l lnesses

to

which

cavers

are

at r i sk .

In

1985

,

a

survey was

made of

150 se lec ted Bri t i sh cavers ,

to

check

for Leptospira

antibodies

in

the

blood. The

sample

featured

some

of

the most act ive

and

exper ienced cavers in

the

country. In

22

of cases there

was evidence

suggest ive of contac t with the

harmless

Leptospira

bi f lexa

and

in 9

with

the paras i t i c Leptospira in te r rogans .

We

found no evidence

of

a regional

bias for in fec t ion with

Leptospira

interrogans

from

any

Bri t i sh

caving

region. Of

the

cavers wi th Leptospira interrogans

ant ibodies , hal f had

caught

the paras i t e

in the

Gunong

Mulu

National Park

(Sarawak).

One

had

been

diagnosed in

the

USA,

while the remainder were mild or symptomless

cases

probably cont rac ted in the

Br i t i sh

I s l es .

INTRODUCTION

The Genus

Leptospira

Man

is

an

occasional host

to

a

group

of

sp i ra l organisms (spirochaetes)

of

the

genus

Leptospira . Leptospires are

free-swimming

and can

be subdivided

into

two major categor ies :

L.

bi f lexa are

saprophyt ic

var i e t i es ving on the

products

of decayed

vegetable

matter

in fresh

water streams and ponds. L.

interrogans

are

paras i t i c

and l ive

in

animal hosts . A wide

var ie ty of wi ld animals ,

par t i cu la r ly

rodents ,

car ry the

parasi tes ,

but even amphibians and

r ep t i l es have been

known

to do

so

(Benenson 1980).

Of

domestic

animals,

ca t t l e ,

pigs and dogs are the

usual hosts . Leptospires are 5

-20 um in length

(Thomas

1979) and in fec t ion

i s

normally

detected

by

the presence of spec i f i c antibodies in the

blood. There are about 180 d i f fe ren t s t r a in s

(serovars)

of L. interrogans which

are

def ined by

prote ins

(ant igens)

on

t he i r

surface

to

which

an t ibodies may

be

raised in

man

or animals.

Serovars with important antigens

in

common

are

grouped together into

23 serogroups. Individual

serovars

may

be spec i f i c for d i f fe ren t animal

hosts .

The predominant

in fec t ing

serogroups found

in

the

Br i t i sh

I s l es are icterohaemorrhagiae

(carr ied

by

the

brown

ra t ) , canica la (carr ied by dogs) and

se j roe , par t i cu la r ly

serovar

hardjo

( ca t t l e ) .

Unti l very

recent ly

hardjo

was c lass i f i ed as

a

serovar of hebdomadis

(Waitkins 1985). The

host

animal i s qui te of ten

asymptomati

10

but wil l car ry

l ep tospi res in high

numbers (

10 organisms

/ g

of

t i ssue)

in

the i r kidneys,

excre t ing

them in the

urine

(Waitkins

1985). In a warm

moist

environment

leptospires

can survive for severa l

months . Condit ions are m

os t

favourable

in

t rop ical

areas, while in temperate count r ies there

i s

a

marked

seasonal incidence of

in fec t ion , most

cases occurr ing

in

l a t e

summer

and autumn.

The

disease Leptospirosis

The

Leptospira organisms usual ly

enter

the

body

through

cuts and

abrasions

of the skin but

may be inhaled

or

enter through the

conjunt ivae of

the eyes.

Both species of

Leptospira may

enter

the

body,

but only L. interrogans

wil l

mult iply

and cause

i l lness . In the pas t

l ep tospi ros i s in

Br i t a in was mainly an

i l lness of

sewer and

mine

workers,

whose

working environments

were in fes ted

by

r a t s . Pest control , protect ive clo th ing and

heal th educat ion have

done

much to

reduce the r i sk

to these people. Those most

a t r i sk

now are farm

workers, par t i cu la r ly dairymen,

who

cont rac t the

i l l ness (caused

by

the serovar

hardjo)

through

d i rec t contac t with ca t t l e

(Waitkins 1985).

The

increasing popular i ty of

water

spor ts has meant

an

increase in

incidence of

casual ly acquired i l l ness

usua l ly

caused

by

the ra t -borne serogroup

icterohaemorrhagiae. In

the t ropics

many of

the

fevers

suf fered by

agr icu l tural

labourers

(e .g.

131

cane cu t t e r s ' disease) are

Leptospira infec t ions

(Wilcocks

and

Manson-Bahr 1972) .

After

entry into

the body

L.

interrogans

mult ipl ies rapid ly in the bloodstream

and

thence

f inds

i t s

way

to

a l l

organs.

The body

responds

by

producing

many

dif feren t

an t ibodies ;

as

the

pat ien t begins to convalesce the

infect ing

serogroup

ant ibody predominates. The l ep tospi res

tend to loca l i se in the l iver and kidneys,

sometimes causing se r ious damage.

Those in

the

kidneys seem

to

be unaffected

by

antibodies

and

are

shed in the urine for

severa l weeks.

The

incubat ion period for the disease i s

normally about

10 days, but can vary

from

3

days

to

3

weeks. The course of the i l lness var ies from

indi vid ual to indiv idual ,

and also depends on

the

infect ing serogroup.

In many

cases the pat ien t

suf fe r s

only a

mild

i l l ness with loss o

appet i t e ,

nausea, headaches, muscular

pains

and

a

low grade

fever . The symptoms subside af te r

a

few days and

are

of ten

at t r ibu ted

to inf luenza or

v i ra l

i l l ness . The infect ion

can be asymptomatic.

There

i s

of ten

a

more

severe

f eb r i l e

(feverish) i l l ness of sudden onset accompanied by

conjunct iv i t i s ,

diarrhoea, mental confusion and

such

severe

pros t ra t ion that

the

pat ien t

may be

unable

to turn

over

in bed without

ass i s tance .

I t

i s frequently biphasic with a f i r s t phase

las t ing

4-7 days. The symptons then subside for a day

or

two

before the fever

returns, of ten with

severe

headaches .

Weil s

Disease

i s the name given to the most

severe

form of

the i l l ness . Jaundice appears due

to l i ve r damage af t e r 4-6 days and there

may

be

haemorrhages in to the skin

or from

the

gut .

The

kidneys

may fa i l .

Most

deaths

occur

a t

t h i s

stage

of the i l l ness but

in

less severe cases recovery

occurs

in

the th i rd

week.

All serogroups respond to t reatment with

an t ib io t i cs ,

par t i cu la r ly

pen ic i l l i n , in the

f i r s t

week

but some

( icterohaemorrhagiae in

par t icu lar)

wi l l

not

respond i

t reatment i s delayed.

Ful l

recovery can

take

severa l

months

and antibodies

remain

detec table for 1-20 years

(Wilcocks and

Manson-Bahr

1972).

The

most ser ious

cases

and the majori ty

of

cases of

Weil s Disease are due to the

icterohaemorrhagiae serogroup;

hardjo

in fec t ions

appear to be l ess severe.

In

the most recent

s t a t i s t i c s

of

the Publ ic Health Laboratory Service

(Waitkins,

1986) there

were

90 cases of

l ep tospi ros i s ident i f ied in

the

Br i t i sh I s l es in

the

year 1984.

37

cases were serogroup

icterohaemorrhagiae, 32

were sejroe serovar

hardjo. Two

pat ien t s

died.

The

increase in incidence in recent

years of

serovar hardjo

in

Bri ta in has

mainly

been

among

cat t lemen who cont rac t the

i l l ness d i r ec t ly .

Few

cases

ar i se

outside the

agr icu l tural

indus t ry .

Forty f ive out

of

for ty

nine water spor ts vict ims

contracted the

severe icterohaemorrhagiae

serogroup

dur ing the per iod

1978-1983 (Waitkins

1985) .

8/21/2019 BCRA 14-3-1987


Leptospirosis

among cavers

Leptospirosis in i t s

most severe form,

wei l ' s

Disease , gained a notorious reputation among

Br i t i sh cavers when, in 1963 , a

Bris to l

pathologis t ,

Oliver

Lloyd, became

ll

af te r a

caving

t r i p down Stoke

Lane Slocker

in

Eastern

Mendip . He was

a t

the

time

not

only

a famous

figure

in

caving c i rc l es

but

also the Hon.

Secretary

of the Mendip (Cave) Rescue

Organisat ion . Upon recovery from his

i l lness

he

published the danger to

cavers

of

r a t

in fes ted

streams (Lloyd,

1964)

and,

with the

assis tance

of

the loca l au thor i ty Rodent

Officer ,

waged a

pr iva te vendetta against the ra ts upstream

f r

om

the entrance

to

Stoke Lane Slocker .

This

cont inued un t i l 1972 when he re t i red from

his

post

with the M.R.O.

Not surpr i s ingly , Stoke Lane Slocker gained

an evi l reputat ion and the

Mendips

as a whole

became

considered a r i sk area . There

seems

l i t t l e ju

s t i f i ca t ion

for th i s , since

in

over 20

years

there

have only been two fur ther cases of

in fec t ion

reported

in

the

caving press

from

cavers

in Br i ta in . One of these (Frankland ,

1978) again

involved

Stoke

Lane Stocker .

The

other (Lloyd,

1978)

was a mild

case which

occurred

in

1974; the

caver

had

vis i ted

several s i t e s

on the Mendips

(Swildons Hole, Woo key Hole, Lamb Leer and

others) ,

any

of

which

could have been the source

of the

in fec t ion

. A

review

of the

in te rna t iona l

caving

l i t e ra ture

for th i s period (1964 -

1985)

fai led

to f ind any repor ts of Leptospira

in fec t ion .

Recent ly

several caving expedit ions to

the

East Indies

have been

af f l i c ted by

undiagnosed

fevers.

The Mulu expedit ions

(Sarawak, Malaysia)

have been par t i cu la r ly t roubled. In 1978 most of

the caving par ty became i l l ; in

1980 four were

i l l ; in 1984 f ive ou t of sixteen had fevers

(Buchan,

1985). The in fec t ion was f inal ly

a t t r ibu ted

to the Leptospira serogroup

pyrogenes

and

sejroe

serovar hardjo (Waitkins, 1986), milder

forms

than

the

ra t -b

orne L.

icterohaemorrhagiae

which i s the main r i sk

to

cavers in Br i t a in .

THE SURVEY AND METHOD OF ASSESSMENT

The

Cavers

In

September

1985,

delegates attending

the

nat ional congress of the Br i t i sh Cave

Research

Associat ion

were inv i ted to

give

blood to be

t es ted

for

Leptospira

ant ibodies

indicat ing

previous

in fec t ion .

A

good

response

from

the

delegates

was

ant ic ipa ted and of the s ix hundred

or so who at tended the

congress,

151

gave

blood.

A caving questionnaire was completed by a l l

but

one of the

donors.

The

study sample i s

thus

150.

Figures 1 to 4 show the

percentage

of cavers (ordinate axis)

in the

samples

that

were found

to f l l into the fo l lowing

categories related to

the i r caving habi ts : -

Fig 1

Number

of years caving

1

5

5

1

5 20

25

ye rs

The 150 sub jects are not thought to be

represen ta t ive of the caving community, as the

group i s strongly biased against novices by vi r tue

of

the

type of person attending

the

congress .

Cavers came

from a l l o v e r

the

country

to a t tend

the

BCRA annual

congress,

so there i s

no

obvious

regional bias , but they tended to be older and

more committed to t he i r spor t . Further bias

arose

from

the fac t that cavers who consider themselves

to

have

been

a t

r isk

are more

l ikely to

want

to

have

t he i r blood

t es ted . The two regions

in

which

cavers

are

known to

have

caught

l ep tospi ros i s

(Mendips and East Indies) may featu re more

prominently

for

th i s

reason. Further bias towards

those

who

have

been

caving

for many years ,

and

those who have caved abroad, seems

l ikely .

Laboratory

t e s t s

The blood samples

were

separated

and f rozen

and sen t

to the Leptospira

Reference Unit a t

Hereford.

Four dif feren t

laboratory t e s t s

we

re

used to detec t antibodies against Leptospira

as

the

t e s t s

have

a

dif fer ing

serogroup spec i f i c i ty

and

sens i t iv i ty .

The

macroscopic agg lu t inat ion

t e s t Mac

A) i s

performed on a

microscope s l ide when

antibodies in

the serum

sample

agg lu t inate a

par t icu late

antigen. The antigen is genus spec i f i c but,

because

the

t e s t i s insensi t ive,

t only

reacts to

high

levels

of

ant ibody,

such

as

are

produced

by

in fec t ion with L. in terrogans, not by simple

exposure

to L. bi f lexa . The t e s t may remain

posi

t i ve for months or years following recovery

from the

i l lness

but

i s r e la t ive ly

insensi t ive,

sometimes being

negative when other

t e s t s are

pos i t ive .

The complement f ixa t ion t e s t CFT) and the

enzyme l inked immunosorbant

assay

ELISA) are more

sens i t ive

techniques

for detect ing antibodies.

The preparation

of antigen

used in

these

t e s t s is

dif feren t

and while both

of

the t e s t s may be

pos i t ive in cases of l ep tospi ros i s

the

ELISA is

more sens i t ive and speci f ic . The CFT may be

pos i t ive as a r esu l t

of

contact

with

unrelated

microorganisms

such

as

Mycoplasma,

Cytomegalovirus

and Hepat i t is and

rarely

due to contact

with

h

bi f lexa . For

these reasons both

t e s t s

are

commonly

employed.

132

The microscopic agg lu t inat ion t e s t MAT) i s

used to

ident i fy

the in fec t ing serogroup

and

detects

antibodies using

l ive

leptospires from 21

dif feren t

serogroups.

EJ

-ve

g

+ve

ri

+ve

ve

30

~ t e r r o q a n s

L.

i n t e r r o g ~

~ l u l

u)

L.

bi f l

exa

35

40

30

Fig 2

< )

1 ~ 2 5

>5

caves year

Frequency

of

caving

8/21/2019 BCRA 14-3-1987


The ques t ionnai res

A

simple questionnaire

was prepared which

asked the fo l lowing quest ions:

a) Number of years caving exper ience .

b) Frequency

of

caving. The answers

were

l imi ted

to

0-10

, 10-20, 20 - 50 and

50+ times

per

year

.

c)

Popula r i ty of the

Bri t i sh

cav

ing

regions:

Yorkshire

/ North, Derbyshire , Mendip / Devon, Wales,

I re land/Scot land. Delegates were asked whether

they

of ten or occas ional ly

caved in

these

regions .

d)

Foreign

count r ies

v i s i t ed

for

caving (excluding

Eire) .

A follow-up questionnaire was sent to

subjec ts whose sera

showed a

posi

ti

ve response,

asking i f they

had

previous ly

been diagnosed

as having

Leptospirosis

or weil ' s Disease,

where and

when they thought

they had caught it ,

and

a ser ies

of

medical

quest ions

concerning t he i r i l lness .

RESULTS

The r esu l t s

of

the

sero logica l t e s t s

are

presented

in Table 1. A pos i t ive r esu l t in

one

or

more

t e s t s

was

found in 46 of the

150

samples, an

incidence of 31%. 33 cases (22% of the sample)

are pos i t ive only

by the complement

f ixa t ion

t e s t

suggest ing

contac t with unre la ted

micro-organisms

or wi th

L.

bi f lexa .

13 cases (9%) were

pos i t ive in one

or

more of

the

other t e s t s ind i cat ing

in fec t ion

with the

pathogenic L. interrogans . A second quest ionnaire

Su b j ec t

Test Comment

Mac A

CFT M T ELI SA

+

No i l l nes s

+ Bad

Flu 1959

(+)

Mulu fever

1978 and

19 8 4

+ + +

No reply

+

Mulu fever

1981

Mulu

fever

1984

+

+

glandular

fever

1982

+ +

Mulu

Fever

1984

+

Mulu

Fever

( malaria )

10

(+)

+

Mulu fever 1984

11

+

Diagnosed in US 1976

12

(+)

Chro

nic

mala i se

1982

13 +

No

i l l nes s

14-46

No i l l nes s

Mac. A., Macroscopic

agg lut inat ion;

MAT Micro

scopic

a gg lu t ina t ion ; CFT.; complement f ix a t io n

t e s t ; ELISA.,

polyvalent enzyme l inked assay.

+ ,

pos i t i ve ; - negat ive;

(+),

weak

p o s i t iv e .

r------------------------------ Tab1e 1 Posit ive

f indings

20

1

o

2

3

>

foreign countries

in

se ro lo g ica l t e s t s

for

Lepto

sp i

r a l

in f ec t io n

Fig

3

Number of foreign

countries caved in

133

was sent to

these

people

and

the resu l t s

are

incorporated

in Table 1 . The

cases which

involved

serious i l l ness

had

mostly

been ident i f i ed

a t

the

t ime as Leptospira infec t ions . All but one

occurred on the Mulu expedit ions, the exception

being an

American

caver who

had been

infected

in

the US

9

years

previously.

The

remaining s ix

cases were mild or asymptomatic

and had not been

previous ly

diagnosed as l ep tospi ros i s .

Only one

of these subjec ts had

caved

in the

t ropics

( in the

Solomon I s lands) . None o f

the

subjec ts

had

evidence of

a

spec i f i c

serogroup

infec t ion

( th i s

i s to be

expected as spec i f i c antibodies

decrease

wi th in

a

few

months of

in fec t ion

and cannot be

detected

af t e r

a year or more) .

n analys is

of

the f i r s t quest ionnaire i s

presented

in Figures

1-4.

Owing to doubt over the

i den t i t y of

ten

subjec ts ,

f ive

of whom were CFT

pos i t ive , these cases appear in

the

to ta l sample

but

not in

the

CFT

posi

ti

ve groups . The

caving

exper ience of those taking par t

in

the

survey

as a

whole and

those showing antibodies

i s compared

in

Fig 1.

Similar

comparisons are made in Figures

2

-4 of caving frequency,

foreign expedi t ions and

regional popu la r i ty in the

Br i t i sh

I s l es .

The

Figures

show

tha t

the

13 cases who

have

been

infecte d with L. interrogans are more experienced,

more ac t ive ,

and cave abroad

more of ten

but

tha t

the region a l

dis t r ibu t ion

i s much the same

as

for

the

main

group

of cavers

.

The

East

Indies

feature

s t rongly as a caving

region

among

the

seropos i t ive

cases

(other par t s

of the

world show no

di f ference) . Two th i rds of the Mulu cavers

t es ted

were

seropos i t ive .

The

same

analys is can

be appl ied

to the

group

who

were

pos i t i

ve

only

by the complement

f ixa t ion

t e s t .

There

i s a

s l igh t increase in caving

ac t iv i ty

but

the

main di f ference i s seen

in

the

regional

popular i ty o f the Bri t i sh caving areas .

DISCUSSION

In t h i s study

pos i t ive

f indings

or.

the

Mac A,

M T

and ELISA te s t s

are

taken to indica te pas t

infec t ion

with L. in te r rogans . The

f inding

of a

po s i t i ve CFT only i s suggest ive of exposure to

genus - spec i f i c

ant igen

of

non-pathogenic

50

40

30

20

1

y

Fig 4 Pop u la r i ty of Bri t i sh

regions,

those often caved in;

D

Y =

Yorkshire,

D = Derbyshire,

M =

Mendip,

W = Wales,

l / S

= I re land

/

Scot land

M

w

l S

8/21/2019 BCRA 14-3-1987


l ep tospi res or to other

micro-organisms

such as

Mycoplasma. These f indings are of considerable

in te r es t ,

despi te the bias

of

our sample, in that

th i s i s the

f i r s t survey

to assess the

r i sk

to a

large group of cavers of

in fec t ion

with pathogenic

l ep tospi res . In addi t ion there

i s

evidence to

suggest wide - spread contac t among cavers

wi th

non-pathogenic organisms such as L. bi f lexa .

The immediate impression gained

from

the

data

in

Figure 1 i s

t ha t

the proport ion of cavers wi th

lep tosp i ral

antibodies

r i ses

s teadi ly with an

increase in caving

exper ience.

I t

i s however

important

to

apprec ia te

that af te r

contact with

leptospi

r es of

e i ther species

ant ibodies do not

remain detec table in

the

blood indef in i te ly .

In

some cases ,

par t i cu la r ly

af te r ser ious in fec t ion ,

they may st ll be

detec table af te r severa l

years

but they are unl ike ly to

be r e l i ab ly

de tec ted

af te r a

decade

or

more. The r esu l t s probably

represent

r e l a t ive ly

recent contact with

l ep tospi res .

In many cases the

or ig ina l contac t

wil l

have

been

years

ago,

but

with

repeated

exposure to leptospires

the antibodies are

boosted

from

time

to

time

and so remain

detectable.

I t i s a l

so

l ike l y

that

exper ienced cavers go

to areas

of higher

r i sk .

For example, a l l the

Mulu cavers

had

10

or

more

years

caving

exper ience

(median

14

1 / 2 years)

as compared

with a median of

8

years

for the t o t a l sample. The higher

incidence among more exper ienced cavers i s

probably genuine and r e f l ec t s

repeated

exposure

and excursion

s

to areas

of higher

r i sk .

The

other se rops i

t

ve cases (excluding

the

American caver) probably re f l ec t

infect ions that

were

cont rac ted

in t h i s

country. In

cont ras t to

the compact group

that comprises

the Mulu

cavers

there i s a wide

range

of exper ience among

th

i s

group (4

- 29 years of

caving

experience) . Both

the

mean

and the median are

about

2 years higher than

for the whole group

of

150

cavers though with only

s ix cases

t

her

e

i s no s t a t i s t i ca l s igni f icance in

these di f fe rences .

There

i s no

evidence that any par t i cu la r

region

in

t h i s country represents a

greater r i sk

for in fec t ion with

Leptospira

interrogan s . Those

who

cave regular ly in Yorkshire (4 in 83) face the

same small r i sk as

those of

Mendip (1 in 33).

There i s no

Mendip Factor

in our resu l t s .

I f the cavers wi th only CFT pos i t ive

r esu l t s

(who may

have had

contac t

with the

genus

spec i f i c

ant igen

of L. bif lexa)

are examined

in

a

similar

way, the

mean, median

and standard

devia t ion for

the data

in

Figure 1 are

almost ident ical

to the

values for the

sample

as a whole.

There

i s a

small

decrease

in

incidence among inexper ienced

cavers

but between

5

and

20 years

the dis t r ibu t ion

i s very

even.

The

frequency

of caving might be

expected

to

be

important and there i s

perhaps

a

very

small increase for th i s

group.

An

expedit ions

prof i le

i s almost

ident ical

to

that

of

the main group

,

even

down

to the countr ies

that

are

vis i ted .

The only s igni f icant

change is

in

the

regional

ana lys i s ,

where Wales

shows

an

increased incidence and Mendip a

markedly

reduced

incidence.

The

regional

incidence of

L. b if

l

exa

exposure

may be explained

by

the

l i f e s ty le

of

the organism

which favours

damp

moorlands and the streams

that

dra in them;

sur face

streams

on

the

Mendips are

shor t

and

few. While

t h i s

f inding i s of some

i n t e res t t

i s

of no consequence to cavers as h

bif lexa

i s

en t i r e ly harmless. I t

i s however

important

to

point

out

that

among

the caving group

the incidence of pos i t ive CFT i s very

high a t

22 ,

while among blood

donors

t i s somewhere in the

region of 0.01%.

There

i s

some evidence

that

antibodies to

L. b if

l

exa are

detec table only

for

a

shor t t ime ,

in that

22 of

our

samples

claimed

not

to cave frequently in any

region; none

had

antibodies .

CONCLUSIONS AND RECOMMENDATIONS

On

a

survey of only

150

cavers our

conclusions have to be

ten tat ive.

Obviously

leptosp

i

ros i s does

represent a

small r i sk to

cavers

but our

sample

comprises the element

of

the

134

caving community most l ikely to be a t

r i sk ;

the

average Bri t i sh caver has had l ess

exper ience and

i s

l ess widely

t r ave l led . A prevalence of

9

in fec t ion

with L. interrogans found

in our

survey

would

not be re f lec ted in the caving community as

a

whole. Even

the f igure of 4 , those cases

believed

to have

been

contracted

in the Bri t i sh

I s l es , i s probably too high.

The grea tes t

r i sk

of in fec t ion

i s

clear ly

in

caves such as

those in

Mulu

.

The East Indies

have

been

known to

be

a

r i sk

area for

l ep tospi ros i s

for

many years . Among

our

sample

as

many cavers had

been

to Indonesia and Papua New Guinea as had

been

to

'

Malaysia, but only Mulu

cavers

were

seroposi

t

ve.

However the in fec t ion

may

not in

fact

have

been contracted within

the

caves of

Mulu

but

around

the camp s i t e , which

was infested

by

rodents . A high incidence

of

in fec t ion

has

been

found in the same area

among

so ld ie r s

under taking

t r a in ing

exerc ises

in the

jungle (S Waitkins) .

The r i sk

to Mulu cavers seems

so great t ha t

the advice offered in the

repor t

of the 1984

expedit ion

(Buchan 1985) is

worth repeat ing

here:

any unexplained

fever

should be

provis ional ly

diagnosed as l ep tospi ros i s and t reatment

with

pen ic i l l i n

s t a r t ed immediately. The Bri t i sh

Army

uses prophylact ic pen ic i l l i n when on

jungle

t r a in ing in Malaysia (even

then 20

of so ld ie r s

are seropos i t ive

afterwards,

suggest ing

in fec t ion

which

was asymptomatic). We cannot recommend

t h i s

pract ice

for caver s as expedit ions are in the

f i e ld for extended periods and long term

an t ib io t i c t reatment a t the doses

required would

be

potent ia l ly harmful.

The

bes t

protect ion

for expedit ions

i s

physical ra ther than medical. Leptospires usual ly

enter the

body

through breaks in the skin. The

wearing of

t rousers ins tead of shor ts

when walking

through the

jungle and the

use

of

gloves

underground

should s igni f icant ly reduce

the n u m ~ r

of

avai lable points

of

entry

for the paras i t e

.

Entry through the eyes

and

no

s e

can be prevented

by care

in

the

choice of water used

for

washing.

Drinking water should

be boiled or chlor ina ted .

Buchan (1985)

recommend

s the use of rubber

socks

to

prevent abrasions

to the fee t

by

g r i t par t i c l es

( the scra tches become

infec ted

and develop in to

the

painful

Mulu foot ) .

This

would

seem to be

an important measure as the fee t are always

in

potent ia

l ly infec ted water.

Compared

with

Mulu, the r i sk

to cavers

in the

Br i t i sh I s l es

i s

very s l igh t , but

the

same simple

precautions ought to

be

taken. In

par t i cu la r

we

would

recommend

the use of gloves (rubber

'washing

-

up'

gloves

are

adequate) for caving.

Many

caver

s

rout inely

use

gloves

in

wet caves but

do not bother

when

explor ing old

mines, where ra t

in fes tat ion

i s

more

l ike ly .

The Stoke

Lane

Slocker

cave should be

considered a heal th hazard.

Because l ep tospi ros i s

shares

common symptoms

with

many other diseases ,

the i l lness

i s

under

diagnosed in t h i s

country. I f a caver

has

a

fever i sh f lu - l ike i l lness , t i s important

to

mention the spor t to the at tending doctor , as t h i s

may help

the diagnosis and

may

indeed

save

the

pa t ien t ' s

l i f e

.

REFERENCES

Benenson, A

S.

1980.

Control

of

Communicable Diseases in Man

(13th

Ed) - Offic ia l

Report

of American

Public Health

Association.

Leptospirosis, pp193-196.

Buchan

J. 1985.

Mulu Medicine. Caves of Mulu 1984 (pub

Brit ish

Cave Research

Association)

pp51-53 .

Frankland J.

1978.

Weil's Disease

in a Mendip

Cave. Bull

Brit ish Cave

Research

Ass o

cia tion

20,

15-16.

Lloyd

0 C. 1964 .

(as

Cherarnodytes) Mendip Notes Wessex Cave

Club

Jou

rnal 8

No

93,

p25.

Lloyd

0 C.

1978.

weil 's Disease

in

Ca

ve

rs . Caves and

Caving,

No 1,

p23.

Thomas C G

A.

1979.

Medical Microbiology (4th

Ed).

Bai l l ie re Tindal l , London.

Leptospira

~ p 9 5 9 6

Waitkins S A. 1985. Update on Leptospirosis . Brit ish

Medical Journal,

vol

290,

pp15 02-1503.

Waitkins S A. 1986 Leptospiros i s in Man

Brit i sh

Is les

1984.

Brit ish Medical Journal, Vol

292, p1324.

Wilcocks

C,

and Manson-Bahr

P E C.

1972.

Manson's

Tropical

Diseases (17th Ed). Bail l iere

Tindal l ,

London.

Leptospirosis

pp596-602.

Received August 1987

Charles A Self

4 Tyne Street

Bristol BS2 9UA

8/21/2019 BCRA 14-3-1987


8/21/2019 BCRA 14-3-1987


8/21/2019 BCRA 14-3-1987


8/21/2019 BCRA 14-3-1987


8/21/2019 BCRA 14-3-1987


8/21/2019 BCRA 14-3-1987
