HomeMy WebLinkAboutSUS5004TEMPERATURE PREFERENCE OF JUVENILE ARCTIC CISCO
(COREGONUS AUTUMNALIS)FROM THE ALASKAN BEAUFORT SEA,
IN RELATION TO SALINITY AND TEM PERATURE ACCLI~~TION
by
Robert G.Fechhelm
LGL Ecological Research Associates,Inc.
1410 Cavitt Street
Bryan,Texas 77081
William H.Neill
Department of Wildlife and Fisheries Sciences
Texas A&M University
College Station,Texas 77843
and
Benny J .Gallaway
LGL Ecological Research Associates .Inc.
1410 Cavitt Street
Bryan,Texas 77801
..1 h lA flG I'~
_H;"riC 11"-'
A N ,)
707 ;..SHiH
AJ"C}1 C<AGE.AI(9'50 1
March 1982
2
ACKNOWLEDGEM ENTS
We wish to express our gratitude to Dave Norton of the Outer
Cont inental Shelf Environmental Asses sment Program's Arctic Projects
Office (or his support during all phases of the research.Thanks are
also extended to the members of the Waterflood Monitoring Program survey
team --Bill Griffiths,Dave Schmidt,Brad Adams,Terry Carpenter,Rob
Dillinger a nd Dennis Hensel --who provided the fish for the experiment;
to Scott And erson for his statis tical a d vi ce ;to Chuck Davis for his
help in c ons t r uc t i ng the test apparatus;to Bonnie Bower for drafting
the figures;and to the s t a f f s of LGL Ecological Research Associates
and LGL Alaska for their help a nd encourag ement.
This study was funded partially by the Bureau of Land Management
through i nt e ra g e ncy agr eement with the National Oceanic and Atmospheric
Administration,as part of the Outer Continental Shelf Environmental
Assessment Program.
3
ABSTRACT
Ho ri z on t a l - t he rmal - gr ad i ent apparatus of previously undescr ibed
design was used to d etermine the temperature preference of juven ile arctic
cisco,(oreganus autumnalis,as a function of acclimation temperature
and acclimation-test salin ity.Me an preferred t emperature ranged from
11.5 C for fish acclimated to 5 CI5 ppt to 15.4 C for t he 15 C/15 ppt
acclimat ion group .Estima ted final temperature preferenda were 13.5
C at 30 ppt and 15.6 C at 15 ppt.Preferred temperatures at 5 and 30
ppt were lowe r (PeO.05)than that at 15 ppt .Qual ita tive observat ions
of fish i n t he accl im at ion ta nks suggest ed that phys iolog ically optimal
temperatures of j uvenil e arctic cisco also exceed 10 C over the salinity
range of 5-30 ppt.
These re s ul t s a re cons istent wi t h the summe r d ist ribut ion o f arctic
c isco in the Ala s kan Beaufort S ea :fish c oncent r a t e near shore in a narrow
band of relatively warm wat er of moderate sali ~ity .
Key Wo rds :T emperature preference ,arctic cisco,Be a "fort Sea,salinity,
summer d istr ibut ion,b ehavior.
4
INTRODUCTION
The arctic cisco,Coregonus autumnalis,is one of the most abundant
anadromous fishes in Alaskan arctic waters.From spawning areas in the
Mackenzie River system (Hatfield et al.1972;O'Neill et al.1981)and
possibly other major river systems of the North Slope region,juveniles
migrate into the nearshore waters of the Beaufort Sea where they live
during the ice-free period from mid-June to early Sept ember (Craig and
"aldorson 1980).Freshwater runoff along with intense solar heating
make these shallow coastal habitats relatively warmer and less saline
than deeper offsho re waters (Truett 1980).Nearshore areas also serve
as primary summer feeding grounds for arctic cisco due to shoreward and
longshore transport of invertebrate-rich ocean waters (Griffiths and
Dillinger 1980).
The abundance of arctic cisco and other anadromous species in the
warm-water areas during summer has prompted speculation as to the effects
of temperature on the migration and distribution of these fishes.In
response,we conducted a laboratory study of the temperature preference
of juvenile arctic cisco as a function of both acclimation temperature
and acclimation-test salinity.Although a multitude of temperature-
preference data has been published (Coutant 1977),none is available
for this species nor any other high-arctic coregoninc.Temperature
preferences were measures in a horizontal-thermal-gradient apparatus
of prev;ously undescribed design.
5
MATERIALS AND ~IETIlODS
Arctic cisco (83-136 mm)were taken by fyke net from the Beaufort
Sea near Prudhoe Bay ,Alaska,during July and August 1981.The fish
were flown tl ·,day of their capture to a laboratory in Fai rb anks.Water
temperature and salinity during transport and during an in ~tial 48-hour
holding period at the laboratory were maintained at leVE ls similar to
those recorded in the field at time and place of capture (i.e.,5-8 C
and 10-25 ppt).
Groups of 20-25 fish were held in 450-liter filtered aquaria under
constant light.(At the latitude --approximately 700 N --and season
of fish collection,day length was 24 hours.)Dechlorinated tap water
RsupplementedwithInstantOcean salts was used in all phases of the
experiment.Conditions of acclimation were organized in a 3x3 design
salinities of 5,15 and 30 ppt (~2 ppt)versus temperatures of 5,
10 and 15 C (~0.5 C);however,owing to numerous logistical problems
the 5 ppt/15 C acclimation group was never tested.Groups were brought
to their specific acclimation conditions by incrementally adjusting
temperature and salinity at the rates of 2 C/day and 5 ppt/day,respectively.
Fish were maintained at their final acclimation levels for a minimum
of ten days prior to testing.During the acclimation period they were
fed to satiation 2-3 times daily on a mixture of commercial freeze -dried
euphausids and brine shrimp.Once testing of a particular group began,
fish were fed to satiation 20-30 minutes before the beginning of each
I.emperature-preference trial.
Fig.I
Fig.2
6
Test Apparatus
Horizontal thermal gradients were formed in an elongate chamber
made from a 5-m length of transparent polyvinyl chloride (PVC)Excelon R
pipe with an internal diameter of 102 mm.Lying within this primary
structure,and extending its entire length along the bottom,were three
smaller tubes --a 12.7-mm-diamter PVC pipe flanked b y two 15.9-mm-diameter
titanium pipes (Fig.I).A 50-mm-wide slit in the top of the primary
tube permitted the investigator free access to any portion of the chamber.
Gradients were established by pump ing coolant (ethylene glycol)
through one titanium pipe while simultaneously pumping hot water through
the other in the opposite direction (Fig.I).This countercurrent arrang-
ment for heat exchange was augumented by heterogeneous insula tion of
the titanium pipes:the upstream third of each was bare;the second
third was spirally wrapped with 6.4-mm-thick clear vinyl so that the
proportion of bare conductive surface progressively decreased;and the
downstream third was completely insulated with vinyl wrap.This system
produced linear to slightly sigmoidal gradients as great as 18 C (Fig.2),
with specific gradients obtainable through adjustments in the temperature
and flow rate of the liquid within each heat exchanger.Compressed air,
bubbled from the central PVC line via 0.5-mm holes at 20-cm intervals,
prevented cross-sectional thermal variation in addition to providing
aeration.A screen of plastic mesh prevented fish from contacting the
heat exchangers and aeration pipe.
Each of four such gradient tanks (mounted one above another )was
marked off at intervals of 50 em,to g ive 10 stations.A copper-constantan
thermocouple submerged at each station was int"rfaced with a Baily
Instruments Inc.digital thermometer (Model BAT-12;display accuracy
7
0.1 C)to providp data on watpr temperaturp.Florescent light rpflected
off a whltp backround pr~vidpd low-level,uniform illumination from the
side of thp tank opposite the observer.
Experimental Procedure
A single fish was netted at random from the desired acclimation
group and placed in a pre-formed gradient at the temperature corresponding
with that of acclimation.The salinity in the gradient was homogeneous
and equal to the accl imation salinity (~1 ppt).In order to accomodate
initial disorientation that might have led to the fish rushing headlong
into areas of stressful temperature,small blocking nets were placed
in the gradient at points equivelent to the acclimation temperature ~3.0 C.
The blocking nets were removed after 30 minutes and the fish was given
an additional 90 minutes to habituate to the test apparatus.The fish's
position in the gradient tank was then recorded in tenths of a division
between each marked station (e.g.,1.6,3.7,etc.)every 2 .minutes
for 60 consecutive minutes.Gradient temperatures were recorded at the
beginning of each trial and after the 10th,20th and 30th observations.
Temperature between adjacent thermocouples and between observations taken
at a single thermocouple was assumed to vary linearly.Temperatures
observed or calculated (if between the~mocouples)for each positional
observation were tabulated and the median taken as the preferred tempera-
ture for that par ticular trial.Frequency distributions based on tempera-
ture (1 C increments)and position (50-cm increments)were also calculated.
Fish from a particular acclimation group were tested in a minimum of
two different gradient tanks which were oriented in opposite directions,
and gradients within specific tanks were varied among trials in order
to detect any bias,other than temperature,that may have affected spatial
distribution.Analysis of var iance,Student's t test and Duncan's
multiple-range test (Ostle and Mensing 197~)were used to evaluate
d ifferences amoung acclimation groups.
8
Fig.3
Table 1
9
RESULTS
Temperature Preference Trials
Individual cisco tended to generate monomodal frequency distributions
with respect to temperature.Strongly platykurtic distributions (i.e.,
those with a moment coefficient of kurtosis greater than 1.0 and position-
ally covering more than 807.of the gradient)were removed from the data
base because the median temperature in such cases more likely reflected
the distribution of gradient temperatures than the fish's temperature
preference.Most of the seven (of 110)trials that were rejected on
this basis involved fish that appeared highly stressed and unable to
adapt to the test apparaLus.The results of three other fish were also
discarded because their distributions were sharply truncated at either
end of the gradient.
Plots of sample variance against standard fish length indicated
that size had a negligible effect on temperature-frequency distribution.
The possibility of temperature re-acclimation during the course of any
experimental trial was discounted because examination of consecutive
observations prov ided no indication of consistent drift in the t emperature
at which fish were observed.Variation in preferred temperature among
gradients and test tanks proved to be non-significant (P>O.10);therefore,
data within acclimation groups were pooled for furthur analysis.
Mean temperature preferendum ranged from a high of 15.4 C for fish
acclimated to 15 C/15 ppt to a low of 11 .5 C for the 5 C/5 ppt acclimation
group (Table I,Fig.3).Arctic cisco acclimated to 15 C preferred
temperatures that were h igher (PCO.05)than those preferred by either
the 5 C or 10 C acclimation groups;however,there was no significant
·..
10
(P>O.OS)difference in thermal preference between fish acclimated to S
and 10 C. The mean temperature preferendum was signifcantly (P<O.OS)
higher than the temperature of acclimation in all groups with the
exception of those acclimated to 15 C. The final temperature preferendum,
de fined as the point at which the temperature-preference trendline
in ~ersects the 4So diagonal (Fry 19 47),was graphically estimated to be
lS .6 ~at IS ppt and 13.5 C ac 30 ppt (Fi g.3).
Preferred temperatures at salinity extremes of Sand 30 ppt were
lower (P<O.OS)than that at IS ppt (Fig.4).
Behavioral Observations
Behavior of the fish in the acclimation tanks suggested that physio-
logically optimal temperatures exceed 10 C.Fish acclimated to 15 C
appeared alert,perceptive and proved to be extremely elusive during
capture attempts.When slowly pursued about the holding tank the majority
of individuals tended to form cohesive,well-organized schools.Such
characteristics suggest a rather sophisticated degree of both muscular
and neurological capability consistant with a poikilotherm operating
under optimal thermal conditions.Frenzied feeding activity and a
comparatively high rate of food consumption (fish fed to apparent
satiation act ively accepted food within two hours were undoubtedly linked
to elevated metabolic rates induced by the high temperature.Of
approximately 4S fish acclimated to IS C for a period of 10-20 days,the
only fatalities involved individuals that jumped out of the holding tanks.
Conversely,arctic cisco acclimated to S C appeared lethargic and
lacked the locomotory agility noted in their IS C counterparts.Tenuous
schools,consisting of loose aggre~ates of 6-8 individuals,persisted no
longer than several seconds.Feeding behavior was casual,and thes~fish,
11
once satiated,refused food for at least the next 6-8 hours.A mor t al i ty
rate of approximately 0.5 fish/day was noted in all of the 5 C acclimation
groups.
These qualitative differences were very apparent when co mparisons
were made between the 5 and 15 C and between the 5 and 10 C acclimation
groups.Differences between the 10 and 15 C groups were subtle except
with regard to schooling behavior,which was substantially more
conspicuous in the 15 C groups.There were no mortalities in the 10 C
acclimation groups.
Within temperature-acclimation groups,differences in performance
among salinity groups were not detected.
12.,.
DISCUSSION
Thermal preferenda of arctic cisco varied between 11.5 and 15.4
C,depending on the temperature of acclimation and the salinity at which
the fish were acclimated and tested.Given th~t p referred temperatures
are typically highest for juveniles of a species (Coutant 1977)and under
conditions that simulate summer photoperiod/seasonality (Sullivan and
Fisher 1953;Zahn 1963),our results may represent maximum values o f
temperature preference for arctic cisco.
Current literature suggests,for f ish in general,a strong correl-
ation between preferred temperature and the optimum for physiological
performance (Brett 1971)in such terms as maximum potential for growt~
(Jobling 1981),maximum sustainable swi.nming speed (Fry and Hart 1949;
Brett 1967),maximum distance moved as a result of electrical stimulation
(Fisher and Elson 1950),and,most importantly,maximum scope for
activity (Brett 1964;Beamish 1970a).Thermal optima reflect the com-
bination of an underlying positive effect of temperature on biochemical
reaL~ion rates and inherent physiological limitations (e.g.,increasing
net cost of oxygen delivery)which come into playas temperature approaches
the upper lethal limits of the organism.From this perspective the
enhanced survival,schooling,feeding and locomotory performance noted
in arctic cisco acclimated to 10 and 1~C qualitatively corroborate the
thermal preferenda (11.5-15.4 C)determined from the gradient experiments.
The tendency for arctic cisco to select highest temperatures at
intermediate salinities is similar to that reported for the threespine
stickleback,Gasterosteus aculeatus (Garside et al.1977).In the latter
case,intermediate salinities were those isosmotic for the species,
suggesting that thermoregulatory behavior may compensate for osmotic
-
13
stress.The potential severity of such stress is indi cated by the obser-
vations of Rao (1968):in terms of oxygen consumption,the cost of osmo-
regulati on for rainbow trout (Salmo gairdneri)reached 20-274 of total
metabolic demand as environmental salinity diverged from isosmocity.
At salinity extremes,the selection of a lower temperature would reduce
standard metabolism and partially offset the elevated oxygen demand created
by o smotic loading.Higher oxygen concentrations at lower temperatures
might also prove beneficial in supporting increased metabolism.Our
experimental temperature gradients were accompanied by relatively linear
dissolved oxygen gradients ranging from approximately 9 ~g02/liter at
20 C to 15 mg02/liter at 5 C.Farmer and Beamish (1969)likewise found
that oxygen consumption rates for Tilapia nilotica were lowest under
isosmotic conditions.Yet,this species,when acclimdted from 15 to
30 C,showed a preferred-temperature trend essentially opposite that
of arctic cisco and threespine stickleback:selected temperatures were
lowest at intermediate salin ities (Beamish 1970b).Such contrasting
results lead one to consider the importance of habitat and niche diversity
when assessing the species-specific effects o~temperature and salinity
interactions.
Regardless of the variat ion induced by salinity and acclimation
t emperature,the selected temperatures.of juvenile arctic cisco either
approach or exceed the upper limits of the thermal spectrum typically
available to them during the summer season.Moderately saline (18-25
ppt),ne arshore waters of the Alaskan Beaufort Sea reach an annual max imum
temperature of 10-12 C during the month of August (Craig and Haldorson
1980),although temperatures as high as 15 C have been reported (D.Schm idt,
pers.corom.).Given these environmental conditions and the results of
I
14
our gradient experiments,t emperature would seem to playa role in the
summer distribution of this species.Valtonen (1970)presumed a similar
role of thermal preference in the tendency of juvenile Coregonus nasus
to occupy warm nearsho re waters along the coast of the Bay of Bothea,
Finland.
During the summer,arctic coastal waters not only prOVide the highest
tempe ratures local ly a va i l a bl e ,but also generate intense trophic support
for t he f ish species t ha t oc cupy them.This simultaneous occurence of
elevated temperature and abundant forage no doubt confers an ecological
advantage i n terms of grow th potential.Jobling (1981),using a n
accumulation of publ ished data for 49 s pecies of fish,concluded that
there is good correlation between prefe rred temperature and the temper-
ature that promotes maximum growth.Assum ing this relationship holds
for a r c t i c cisco,juven il es a r e ecologically and physiologically positioned
to make optimal use of their limited feeding season.
Although our expe riments implicate tempe rature as an environmental
determinant in the summer distribution of arctic cisco,it is only one
of many factors that can affect population movement .Variables such
a s abundance of forage.shallowness of nearshore waters.substrate
composition or the dynamics of coastal currents may compete directly
with temperature in determining specific patterns of dispersion and
migration.Laboratory st udies have d emonstrated that behavioral thermo-
regulation in fishes can be modified by a variety of biotic and abiotic
factors (Reynolds and Casterlin 1979).Yet,while the exact ecological
role of temperature will depend upon temporal and spatial i nt eg r a t i on
of both species and environmental characteristics,the strong thermal
dependency of phys iological mecha nisms in fishp.s demands i t s serious
·.15
consideration.ln th e case of arctic cisco.pre ference for warm waters
along the Beau fo rt Sea coast wou l d appear to be a sensible adaptive
strategy in that it would enable them to realize their physiological
potential and thereby maximize the probability of successfully coping
with a rigorous environment.
16
LITERATURE CITED
Beamish.F.W.H.1970a.Oxygen consumption of largemouth bass.Micropterus
salmoides.in relation to swimming speed and temperature.Can.J.
Zool.48:1221-1228.
Beamish,F.W.H.1970b.Influence of temperature and salinity acclimation
on temperature preferenda of the euryhaline fish,Tilapia nilotica.
J.Fish.Res.Board Can.27:1087 -1093.
Brett.J.R.1964 .The respiratory metabolism and swimming performance
of young sockeye salmon .J.Fish.Res.Board Can.21:1183-1226.
Brett.J.R.1967.Swimming performance of sockeye salmon.Onchorhynchus
nerka.in relation to fatigue time and temperature.J.Fish.Res.
Board Can.24:1731-1741.
Brett.J.R.1971.Energetic responses of salmon to temperature.A study
of some thermal relations in the physiology and freshwater ecology
of sockeye salmon (Oncorhynchus nerka).Am.Zool.11:99-113.
Coutant.C.C.1977.Co~pilation of temperature preference data.J.
Fish.Res.Board Can.34:739-745.
Craig.P.C.and L.J.Haldorson.1980.Beaufort Sea barrier island-lagoon
ecological processes studies:Final report.Simpson Lagoon.Part
4.Fishes.Fish.Res.Unit 467.In:Environmental Assessment Alaskan
Continental Shelf.Final Report.BLM/NOAA,OCSEAP.Boulder,Colorado.
266 p.
Farmer.G.J.and F.W.H.Beamish.1969.Oxygen consumption of Tilapia
nilotica (L.)in relation to swimming speed and salinity.J.Fish.
Res.Board Can.11:2807-2821.
17
Fisher,K.C.and P.F.El.on.1950.The selected temperature of Atlantic
salmon and speckled trout and the effect of temperature on the response
to an electrical stimulus.Physiol.Zool.23:27-34.
Fry,F.E.J.1947.Effects of environment on animal acti~ity.Univ.
Toronto Stud.Biol.,Ser.55;Publ.Onto Fish.Res.Lab.No.68.62 p.
Fry,F.E.J.and J.S.Hart .1949.Swimming speed of goldfish at different
temperatures.J.Fish.Res.Board Can.7:169-175.
Garside,E.T.,D.G.Heinze and S.E.Barbor.1977.Thermal preference in
relat ion to salinity in the threespine stickleback,Gasterosteus
aculeatus L.,with an interpretation of its significance.Can.
J.Zool.55:590-594.
Griffiths,W.A.and R.Dillinger.1980.Beaufort Sea barrier island-
lagoon ecological processes studies:Final report,Simpson Lagoon.
Part 5 .Invertebrates.Fish.Res.Unit 467.In:Environmental
Assessment Alaskan Continental Shelf,Final Report Principal
Investigator.BLM/NOAA,OCSEAP.Boulder,Colorado 190 p.
Hatfield,C.T.,J.N.Stein,M.R.Falk and C.S.Jessop.1972.Fish r esources
of the Mackenzie River Valley.Vol.1.Dept.of the Environ.,Fish.
Ser.Winnipeg,Manitoba.97 p .
Jobling,M.1981.Temperature tolerance and the final preferendum -
rapid methods for the assessment ~f optimal growth temperatures.
J.Fish.BioI.19:439-455.
O'Neill,J.,C.McLeod,L.Norton,L.Hildebrand and T.Clayton.1981.
Aquatic investigations of the Laird River,British Columbia and
Northwest Territories,relative to proposed hydroelectr ic development
at site A.C&G Laboratories Ltd.,Edmonton,Alberta.122pp.
Ostle,B.and R.W.Mensing.1975.Statistics in Research.Iowa State
University Press.Ammes Iowa.596 p.
18
Rao,G.H .H.1968.Oxygen consumption of rainbow trout (Salmo gairdneri)
in relation to activity and salinity.Can.J.Zool.46:781-786.
Reynolds,W.'!.and H.E.Casterlin.1 979.Behavioral thermoregulation
and the "final preferendum"paradigm.Amer.Zool.19:211-224.
Sullivan,C.H.and K.C.Fisher.1953.Seasonal fluctuations in the
selected temperature of speckled trout,Salvelinus fontinalis (Hitchell).
J.Fish.Res.Board Can.10:187-195.
Truett,J.1980.Beaufort Sea barrier island-lagoon ecological processes
studies:Final report,Simpson Lagoon.Part 2.Physical processes.
Fish.Res.Unit 467.In:Environmental Assessment Alaskan Continental
Shelf,Final Report,Principal Investigator.BLH/NOAA,OCSEAP.
Boulder,Colorado.51 p.
Valtonen,T.1970.The selected temperature of Coregonus ~(Pallas)
sensu Svardeon,in natural waters compared with some other fishes,
p.347-362.In:D.C.Lindesy and C.S.Woods (eds.).Biology of
coregonid fishes.Univ.Hanitoba Press,Winnipeg.
Zahn,H.1963.Jahreszeitliche Veranderungen der Vorzugstemperaturen
von Scholle (Pleuronectes platessa,Linne.)und Bitterling (Rhodeus
sericeus Pallas).Verhandl.Dtsch.Zool.Ces.Huenchen p.562-580.
•
19
Table 1.Mean preferred temperature ~1 standard error (sample size)for
arctic cisco acclimated to various combinations of temperature
and salinity.
ACCLIMATI ON ACCLIMATI ON TEMPERJ.TURE (oC)
SALINITY (ppt)5 10 15
5 11.5 ..0.7 (12)12.8 ..0.6 (15)
.Hl 1°,13 .7 ..0.5 (10)13.8 ..0.4 (14)15 .4 ..0.4 (16)-
3 0 12 . 8 ..0 . 7 (i i)12.7 ..0.6 (12)14.1 .. 0 .9 (10)
20
FIGURES
Fig ure 1.Schematic representation of one of the thermal gradient
systems (above )and a cross -sectional v iew of a grad ient tank (below).
F igure 2.Water temperature versus location (at 50 cm intervals)for
a typical horizontal thermal gradient.
Figure 3 .Preferred temperature (~1 SE)versus acclimation temperature
for acclimation-test salinities of 5 ppt (dotted line),15 ppt (dashed
line)and 30 ppt (solid line).
Figure 4.Preferred temperature (~1 SE)versus accllmation-test
salinity for acclimation temperatures of 5 C (dashed l ine),10 C (solid
line)and 15 C (dotted line).
,".
I Hot I ......SfIReservoirI !
®-,-.,-
••Air.,
t !S I Cold I
+--I Reservoir I
Plastic Mesh
PVC Air Line Titanium Heat Exchangers
20
0
0
I
CD
15
"-~...
CO"-CD 10Q.
E
~
5
,
0 1 2 3 4 5 6 7 8 9 10
Location
20
0 18
°I 16
Q)--...--:::l 14 ~------------~
tV...12 ...........................Q)
c.
E 10t!
8
-c
Q)6......
Q)....4Q)...a.2
o 2 4 6 8 10 12 14 16 18 20
Acclimation Temperature -"C
II)I I I I I I I I I C?...C1.
L -0 C1.-C?I
I
I
I II):~.I -~..
I N ..-
I ,:
I ..---I I~
--0 (I)
N
....en-~~:::1 -II)I~....
•\C\0\0 .--\-......
\....,~
\,-
\,=\..-
....1-.\----II)(,)
(,)<C
I •••I ••••
0 en CO ~N 0 en CO ~N 0
N ....................
00 -aJnJeJadwa.l paJJaJaJd