Conchology of endangered freshwater pearl mussel: conservation palaeobiology applied to museum shells originating from northern Finland

2007·6.1 MB·English

by  Samuli Helama

Boll. Malacol., 43 (9-12): 161-170(2007)
Conchology of endangered freshwater pearl mussel:
conservation palaeobsology applied to museum shells
originating from northern Finland
Samuli Helama* (El), Jan Kresten Nielsen* & llmari Valovirta0
* Department Abstract
ofGeology, P.O. Box64, Shellsof Finnish freshwater pearl mussel (Margaritifera margaritifera)were analyzed by means oftaphonomy
00014University and sclerochronology. The samples originated from two small rivers In northern Finland, River Kotioja and
ofHelsinki, Finland, River Saukko-oja. Taphonomical analysis showed thatthe mussels died recentlyand thattheshellswere likely
[email protected],
0 Corresponding to beopened violentlythrough pearl hunting. Internal shell growth Incrementswere studied on annual basis.
A(ut)hor Annual shell growth increments showed life-long trends, attributable to factors internal to growth. The
trends were removed individually from each series and the resulting time series were statistically compared
# StatoilASA, Exploration within and between the two rivers. The temporal shell growth variability was distinctly similar within the
&Production Norway, P.O. datasets of both rivers, evidenced by inter-series correlations. This correlativity indicated shared growth re-
Box273, NO-7501 sponse to external factors (i.e. climate and hydrology) and further justified the construction of mean river-
Stjordal, Norway specificshell growth chronologies, sclerochronologies. The construction ofthe chronologieswas demonstrat-
ed by means of sclerochronological crossdating, indexing and chronology stripping. The two chronologies
° Finnish Museum
ofNatural History, werecompared bytheirgrowth variabilityvisuallyand statistically. Thecomparison showedthat(1)theampli-
Invertebrates Division, tudes of the growth variations were not much different between the rivers, and, that (2) the two scle-
P.O. Box 17, 00014 rochronologies could be linked in time domain. These results Indicated that(1) the habitatsofthe two popu-
UniversityofHelsinki, lationswere insignificantlydifferentowing totheenvironmental variability magnitude, and, that(2)the exter-
Finland nal control on thegrowth variabilityinvolvedthesame(i.e. regional)climaticfactors regardlessofthe river.
Riassunto
II bivalve dulcicolo Margaritifera margaritifera è una specie a rischio di estinzione in tutto il suo areale di
distribuzione. Capire le relazioni ecologiche di questa specie con l'ambiente è essenziale per organizzarne
la protezione. Informazioni sulla crescita del bivalve possono essere ricavate dall'esame delle strie di accre-
scimento, esaminate in sezione trasversale. Inoltre, l'analisi tafonomica può dare indicazioni su cause e
tempi della morte degli individui. In questo lavoro, si esaminano due campioni di conchiglie provenienti dai
fiumi Kotioja e Saukko-oja, nella Finlandia settentrionale, conservati nel Museo Finlandese di Storia
Naturale. Allo scopo di mettere In luce la storiapreepostmortem degli esemplari, sono state condotte os-
servazioni tafonomiche e sclerocronologiche. L'identificazione degli incrementi interni di crescita è stata ef-
fettuata su base annua, separatamente per il materiale provenienti dai due fiumi. I dati tafonomici indica-
no che gli esemplari sono morti in modo violento e in tempi recenti. Frequenti fratture sulla parte dorso-
posteriore sono interpretate come causate dall'apertura forzata delle valve, per la raccolta illegale di perle.
La morte improvvisa e simultanea dei bivalvi è dimostrata anche dall'ultimo incremento di crescita, ecce-
zionalmente stretto e corrispondente allo stesso anno fra i vari esemplari. Per ¡I materiale di entrambe le
località, sono stati messi in evidenza gli incrementi annui medi di crescita. La crescita delle conchiglie sem-
bra essere avvenuta con lo stesso tipo di variabilità (annuale ed a più lungo termine) in entrambi fiumi,
I
suggerendo che i due ambienti hanno subito variazioni ambientali di ampiezza simile. Sulla base della let-
teratura, si può ritenere che il fattore primario che sta dietro la variabilità nella crescita sia la temperatura
estiva. I dati sclerocronologici mostrano una correlazione statisticamente significativa fra le due popolazio-
ni: ciò implica che esse possono essere correlate cronologicamente. Come prospettive future, si potrebbe-
ro raccogliere dati sclerocronologici secondo una più fitta maglia regionale, al fine di migliorare la cono-
scenza dell'ecologia della specie e la sua dipendenza dai fattori ambientali in fiumi diversi.
Key words
Taphonomy, sclerochronology, growth increments, Margaritifera margaritifera, Lapland, dendrochronology
Introduction the habitat requirements and other ecological features
of M. margaritifera is of great importance in the protec-
Freshwater pearl mussel (Margaritifera margaritifera) is tion plans. In northern Europe, and especially in
an endangered species, which has suffered from pearl Finland (Valovirta etal., 2003), theearlier threatwas pri-
hunters and aquatic contaminants (Bauer, 1986, 1988). marily pearl fishing while the modern threats are asso-
Conservation plans and ecological study programs we- ciated with the drainage of peatlands and silvicultural
re conducted in several countries in order to enhance treatments that change sedimentation rates and water
the potentiality of the species to survive 21st century quality of the rivers. Negative factors include eutrophi-
threats (Young, 1991; Araujo & Ramos, 2001; Skinner, et cation, river dredging and the construction of submer-
al. 2003; Valovirta et al, 2003). Increased knowledge of ged dams for the purposes of fishery. M. margaritifera
was protected in Finland since 1955, but many of the post-mortem and mortal-histories of shell specimens.
populations are currently not reproducing due to poor Internal shell growth structures were examined to dis-
waterquality (Valovirtaetal., 2003). play the annual growth increments of the bivalves.
Invaluable information about bivalve ecology can be Increment width records from individual specimens
obtained from conchological studies. M. margaritifera is were constructed and combined into river-specific shell
a long-lived mussel, reaching a maximum life span of growth increment chronologies. The two rivers were
more than hundred years, and forms shell increments compared by the sclerochronological statistics. Our
with an annual periodicity (Bauer, 1992; Mutvei et al., sclerochronological aims were (1) to present two new
1994, 1996; Dunca, 1999; Ziuganov et al, 2000; Dunca & shell increment chronologies for northern Finland, (2)
Mutvei, 2001). Internal growth structures are visible in to demonstrate the identification of the annual growth
the cross-sections of the shells and serve as quantitative increments, (3) to examine the inter-annual to decade-
records that span over the life time of the individual. scale behaviour of the shell growth and (4) to compare
Samuli Shell growth increments can be used to infer the ecolo- the river-specific records to seek the potential signature
gical requirements and environmental controls of diffe- in growth for ecological and environmental characteri-
Helama, rent species. For example, Jones (1981) studied the in- stics of the habitats.
ternal growth increments ofSpisula solidissima along the
Jan New Jersey coast and found that the growth variability Material and methods
indicated sea surface temperatures and success of juve-
Kresten
nile recruitment. Tanabe (1988) used internal annual Shells of Margaritifera margaritifera (Linnaeus 1758) we-
and microgrowth increments ofPhacosomajaponicum for re collected from the proximity of two rivers in nor-
Nielsen age and growth rate determinations in Japanese coastal thern Finland. River Saukko-oja is situated in the mu-
& area and presented a linkage between the increment nicipality of Salla in south-eastern Lapland and River
llmari growth and tidal rhythms. Seire et al. (1993) studied in- Kotioja runs in the municipality of Taivalkoski in
crement growth rates fromcross-sections ofMacoma bal- Northern Ostrobotnia (Fig. 1). Fourteen shells were
Valovirta thica and their relationships to environmental factors found at River Saukko-oja (July 27th 1984) and seven
and pollution along the southern shore of the Gulf of shells at River Kotioja (July 13th 1979) (Tab. 1). All the
Finland. Dunca et al. (2005) studied the internal growth specimens were found as empty shells at the time of
increments of M. margaritifera in Sweden and demon- collection, lying on the surface sediments on the river
strated the relationships between the growth, the sum- sides. The exact year of death of the animals is thus
mer temperatures and the aquatic pollution. Bivalves unknown.
that grew in a given site or region were influenced by
the same climatic and aquatic factors. Increments are
thus expected to exhibit common growth signal and a 200£ 300
number of individual growth series can be combined
into one mean chronology that better reflects the popu-
lation growth variability (Marchitto et ah, 2000; Helama
etah,2006a).
Here we study the collection ofshells ofM. margaritifera
that were previously collected along two rivers in nor-
thern Finland. The samples now belong to the collec-
tions of the Finnish Museum of Natural History, Uni-
versity ofHelsinki. The usage ofthe museum or subfos-
sil/fossil shells is ofparticular importance in the case of
endangered species, such as M. margaritifera, to preser-
ve the populations that still exist. Palaeoecological tech-
niqueswere applied to the analysis ofthe historic skele-
tal remains of species that are threatened with extinc- Fig. 1. Map showing the shell collection sites as squares, River Kotioja
tion. Such a biological approach has been coined to as (K)and RiverSaukko-oja(S).
conservationpalaeobiologybyFlessa (2002). Fig. 1. Ubicazione dei siti di raccolta (quadrati) nei Fiumi Kotioja (K) e
Taphonomical analysis was performed to reveal the Saukko-oja(S).
SiteXStatistics Samples Years SD ms ARI
RiverKotioja 7 24 0.201 0.253 -0.311
RiverSaukko-oja 14 36 0.275 0.238 0.353
Tab. 1. Descriptive statistics for the sclerochronologies from River Kotioja and River Saukko-oja. Calculated statistics were the total samples size
(Samples), temporal length ofthechronology(Years)with at leastfoursamples, standard deviation (SD), mean sensitivity(ms)and firstorderautocor-
relation(ARI).
Tab. 1. Dati statistici descrittivi per le sclerocronologie dei Fiumi Kotioja e Saukko-oja: numerototale dei campioni (Samples), durata della cronologia
162 (Years)conalmenoquattrocampioni, deviazionestandard(SD), sensibilità della media (ms)eautocorrelazionedi primoordine(ARI).
Macroscopic and microscopic analyses ofthe shells emphasize the growthvariations atthepopulationlevel
and it is advantageous to compare each individual
Specimens were studied to obtain an overview of their growth series to the series of mean growth (master
preservation. The stage of pre-mortem bioerosion and chronology) instead of single individuals (see Helama
dissolution as well as post-mortem bioerosion, abra- et al., 2006a). Previously, river-specific chronologies ba-
sion, disarticulation and fragmentation were examined sed on the annual growth increments of alive-collected
through macroscopic inspection (see Nielsen, 2004). The Margaritifera margaritifera displayed positive correla-
purpose of this taphonomical analysis was to estimate tions within and between sites in Sweden (e.g. Mutvei
thepost-mortemand mortal-history thespecimens. etal., 1994; Helama etal., 2006a).
In order to examine the internal shell growth structures, Growth synchrony was examined herein using visual
shells were cut, cross-sections of the cut surface polis- comparison and quantified by the Pearson product-mo-
hed and the annual growth lines observed under com- ment correlation coefficient (r) (Holmes, 1983). Pearson
puter-integrated microscope system. Following the me- correlation measures a statistical equation describing
thods described by Dunca & Mutvei (2001), one valve the linear relationship between two timeseries. Conchology
of each specimen was cut from the umbo to the ventral
margin perpendicular to the winter lines and along the of
axis of minimum growth. Complete growth records for Time series analyses - growth trend modelling
most species are found along the axis of maximum Prior to actual crossdating, individual growth series endargered
growth. However, Dunca & Mutvei (2001) counted were indexed by detrending the initial measurement se-
exactly the same number of increments in the axes of ries. Individual series of annual M. margaritifera shell
minimum and maximum growth sections with better growth increments exhibit a trend as a function ofonto- freshwater
visibility in the minimum growth section. The sections genetic age of the organisms, demonstrated previously
were ground (800 and 1200 grit metallographic grin- though the investigations of internal (Mutvei et ah,
ding paper), polished (3 pm diamond paste) and then 1994; Dunca, 1999; Dunca & Mutvei, 2001; Helama etal., pearl
etched in Mutvei's solution at 37-40°C for ca. 25 min, 2006a) and shell surface (San Miguel et al., 2004) incre- mussel:
carefully rinsed in de-ionized water and allowed to air- ments. Such a trend, commonly referred to as a 'growth
dry (Mutvei et al, 1996; Schòne et al., 2005). While the trend', is largely due to ontogeny of the organism and,
acetic acid dissolved the carbonate, glutaraldehyde fi- accordingly, target of removal prior to analysis of com- conservation
xated the organic matrix, and alcianblue stained and fi- mon growth signal (e.g. Fritts, 1976; Helama et al., 2004,
xated the mucopolysaccharids and glucosamids in the 2006a). In the present study, growth trends were model-
shell. This treatment resulted in an excellent three-di- led using regression lines and modified negative expo-
mensional preservation of the growth structures with nential curves (Fritts et al., 1969; Fritts, 1976) and subse- palaeobiology
distinct, etch-resistant, blue-colored winter lines. Final- quently removed by dividing the observed annual
ly, annual growth increments were viewed under a re- growth value by the expected growth value of the mo-
flective light binocular microscope and digitally photo- delled curve. Model withbetter fitwas chosen as a final applied
graphed. Widths of all the increments were measured model ofontogeny (Fig. 2). Resulting ratio-based growth to
from the outer shell layer, with 1 pm precision, perpen- indicesweredimensionless.
dicularto the growth lines. Growth indices of shell increments were crossdated museum
using all available growth time series for each site. Site-
Time series analyses - detecting the growth specific chronologies were averaged using all crossda- shells
synchrony ted seriesby arithmetic mean.
originating
Prerequisite of any sclerochronological time series ana-
Sclerochronological statistics
lysis is the procedure called crossdating (Douglass,
from
1941; Fritts, 1976; Marchitto et al., 2000; Helama et al., Many aspects in the study of shell growth increments,
2006a). Due to common environmental and climatic for- i.e. sclerochronology, are highly similar in theoretical
cing factors, the wide and narrow annual growth incre- basis to the study of tree-rings, dendrochronology northern
ments are expected to occur in synchrony. This syn- (Fritts, 1976). As a matter of fact, the aforementioned
chrony may occur over large spatial scales and can be principles of crossdating and growth trend modelling Finland
used to assign correct dates to shell increments. There- were originally developed by dendrochronologists
fore, the year-to-year variability in the growth rates (Huntington, 1914; Douglass, 1941) and lateradopted to
may be synchronized over one specific site or over wi- sclerochronology during the past decades (Marchitto et
der areas if the growth of the organisms reacted simi- al., 2000; Helama et al, 2006a). According to dendroch-
larly to the shared climatic or aquatic factors. As indivi- ronological theory, the site characteristics can be detec-
duals of the same species are expected to respond to ex- ted from the behaviour of tree-ring series, that is, by
ternal factors similarly, the theory of crossdating pre- means oftree-ring statistics (Fritts etal., 1965).
dicts that the synchrony between individuals is most li-
kely to occur atspecies-level. Furthermore, averaging of Statistics that can be related to ecological or environ-
a number of crossdated individual time series into one mental gradients are for example the measures ofvaria-
site and species-specific mean chronologyis expected to bility: standard deviation and mean sensitivity (Fritts et 163
Fig. 2. Growth trend modelling exemplified for
two Margaritifera margaritifera (L.) shell increment
series. Regression lines and modified negative ex-
ponential curves were fitted to initial series of in-
crement width observations (upper plots). Model-
led growth trends were removed by dividing the
observed values by modelled values. Resulting di-
mensionless index series no longer exhibit long-
term growth decline(lowerplots).
Samuli
Fig. 2. Modellizzazione dell'andamento di crescita
esemplificato per gli incrementi di crescita in due
Helama, conghiglie di Margaritifera margaritifera (L). Le li-
nee di regressione e le curve esponenziali negative
Jan modificatesonostateadattateai datidelleserie ini-
ziali di incremento in ampiezza (grafici in alto). Gli
andamenti dicrescita modellizzati sonostati rimossi
Kresten
dividendo valori osservati per valori modellizzati.
i i
L'indice adimensionale risultante non mostra più il
Nielsen Cytogeneticage Ontogeneticage rallentamento di crescita a lungotermine (grafici in
basso).
&
llmari al, 1965; Fritts, 1976). Mean sensitivity (ms) was deter-
I
mined byFritts (1976) as
Valovirta
2 XM-X
ms X i ,)
7
1 f=l
where x is the tree-ring index (here, growth index of
shell increments) of year t in the series possessing n
tree-rings (here, increments). Compared to standard de-
viation, mean sensitivity measures the growth variation
between consecutive years (i.e. variability at high fre-
quencies). Standard deviation and mean sensitivity we-
re used here in comparison of the results of River
Saukko-oja and River Kotioja. In addition, the first or-
derautocorrelationswerecalculated.
Results
Taphonomical analysis
All the shells were preserved in pristine condition with
intact periostracum and major portions of ligament.
Inner shell surface was unaffected by dissolution and
other possible deteriorating processes. In contrast, the
umbonal region was corroded to a severe extent, which
probably took place during the life of the specimens.
This was commonly seen in still living freshwaterbival-
ves. Fractures were common next to the posterior mu-
scular scar, cutting the dorso-posterior margin (Fig. 3).
Also, theventral margin contained sporadic fractures. A
knife cutting the muscular meat and separating the val-
vesfrom eachothersmay havecaused thesefractures.
Fig. 3. PhotographsoftheMargaritiferamargaritiferashell showingthe Overall, the excellent preservation of all shells sugge-
fracture(arrow)cuttingthedorso-posteriormargin (a), acrossthe poste- sted that they were collected alive recently, and have
riormuscularscar(b).
only shortly been transported in natural ways before
Fig. 3. Una conchiglia di Margaritifera margaritifera con frattura (frec- they were unearthed for the museum collections. Pre-
cia) sul margine postero-dorsale (a), e attraverso l'impronta muscolare
164 posteriore(b). vious studies have shown that the preservationalcondi-
tions of shells are not necessarily related to the strati-
graphical age of the shells (e.g., Carroll et al., 2003).
However, the shells ofMargaritifera margaritifera appea-
red to be recently collected alive before theybecame ta-
phonomicallyalteredbyacidic rain water and bioturba-
tion, e.g.,byplantroots.
Annual growth increments
Widths of the annual growth increments were measu-
red perpendicularly to winter lines (Fig. 4 a). The main
difficulty of measuring the increments was to distin-
guish the actual winter lines from disturbance or poten-
Conchology
tial reproduction lines that were sporadically observed
between the winter lines. Misinterpretation of different
growth breaks and corresponding growth lines would of
N
lead to shell increment time series withbiased temporal
control and over- orunderestimated ontogenetic ages. endargered
These misinterpretations were avoided using the follo- Fig. 4. Cross-section of the prismatic (P) and nacreous (N) layers from
wing observational criteria: (1) winter lines canbe follo- tshuerrvoeunntdraeld pbayrtsuopfpotrhteinsgheelplowxiyth(E).frAangnmueanltairnycrpeemreinotsstrwaecruem m(Pera),suhreerde
wed through the outer prismatic layer and penetrates perpendicular(black bars)tothewinter lines(a). Winter linecutsthe in- freshwater
into inner nacreous layers without disruption, (2) win- terface between nacreousand prismatic layerstypicallywith a small rip-
ter lines often cut the interface between the nacreous ple(arrow)(b).Thespecimenoriginatesfrom RiverSaukko-oja.
pearl
and prismatic layers as a small ripple (Fig. 4 b), (3) in- Fig. 4. Sezione trasversale inglobata in resina epossidica (E) dello strato
tra-annual growth lines (ifvisible) are considerably nar- gplriia,smcaotnicopor(Pz)ioenimdaidrpeerpieorsltarcaecoo((PrN)). Gnleilliancrreegmieonnteivaennnturaallieddielclraescciotnachsio-- mussel:
rower alongside winter lines than the ones alongside no stati misurati perpendicolarmente (barrette nere) alle linee invernali
disturbance lines, (4) disturbance lines are often asso- (a). Le linee invernali intersecano l'interfaccia tra lostrato madreperlaceo
ciated with fractures penetrating from shell surface into e quello prismatico, producendo (freccia) una tipica ondulazione (b). conservation
L'esemplareprovienedal FiumeSaukko-oja.
prismatic (but not nacreous) layer (Fig. 4 a). These ob-
servational criteria were supported by the simultaneous
crossdatingofthe shell growth increments (seebelow). Kotioja (Fig. 6). According to crossdating, there were
Mean number of annual increments in the specimens three shell growth increment series with correlation co- palaeoblology
from River Kotioja and River Saukko-oja were 19 and efficient with master series below 0.4 in the River
29 respectively. Kotioja population. We used this as an arbitrary level
for acceptance of series into the chronology: the three applied
Growth synchrony between the shell growth series possessing low correlations with the master chro- to
Indices nology were not included in the final, stripped, scle-
rochronology from River Kotioja (Fig. 6). museum
The growth synchrony between individuals was remar-
kable when the index series were aligned by their last Parameters of growth behaviour shells
identifiable growth increment. This implied that the
shells collected from River Saukko-oja died within the Shell increment statistics showed no significant diffe- originating
same calendar year (Fig. 5). The same observation was rence between the site characteristics (Tab. 1). Standard
made forshells from River Kotioja. deviation was somewhat higher for River Saukko-oja
from
The widths of the last identifiable growth increments but the mean sensitivity was higher for River Kotioja.
were relatively narrow in each sample, and since there Both the standard deviation and mean sensitivity are
could be seen no sign of winter lines in the uttermost measures of variability, but the mean sensitivity is ex- northern
part of the ventral margin in the cross-section of the pected to measure more effectively the year-to-year va-
shells, it was likely that all the studied mussels had riability whereas standard deviation accounts for varia- Finland
stopped growing duringthe actual growing season. bility at all time scales. This implied that the inter-an-
Individual sample series of shell growth index series nual growth variability was somewhat more intense in
were compared separately with the master chronology River Kotioja, however, the overall growth variability
(arithmetic mean of all other series) (Fig. 5). Statistical was greater in River Saukko-oja. This may be, at least
comparison supported the results from the visual com- partly, due to the shorter length of individual sample
parison about the inter-series growth synchrony (Figs. 5 series in River Kotioja dataset: long-period growth va-
and 6). Judging from the relatively high correlations riations may not havebeen preserved in the chronology
between the series in River Saukko-oja, it is likely that since the variations at timescales longer than segment
the mussels probably lived in a shared habitat thus not lengths were removed in the growth trend modelling
too far from each other. Growth synchrony appeared to (see Cook et al., 1995; Helama et al., 2004). First-order
be present although somewhat less evident in River autocorrelations were rather low and insignificant for 165
River Saukko-oja. In this chronological position, the
Approximatecalendaryear
194U 1950 1960 1970 1980 highs and lows of shell growth correlated significantly
(Fig. 7). Correlation was evident with growth variations
A at inter-annual to decadal time scales. However, the
ii
3o‘ preliminary nature of the obtained correlation should
0Ql) be emphasized. Inclusion of more samples would be
0
Qo. preferred in the future work to elongate the chronolo-
0TO gies temporally and to strengthen the growth signal in
$
—1\ both river-specific chronologies to ensure estimated
C0l growth variations inboth rivers.
O
0
ÜJ
(0/)
CL Discussion
II
Samuli
Information from taphonomy
Helama, and sclerochronology
Jan Museum collection of subfossil freshwater pearl mussel
(Margnritifera margaritifera) shells was studied. The set
Kresten of samples, originating from two small rivers. River
Kotioja and River Saukko-oja in the northern Finland,
Nielsen was examined by means of taphonomy and sclerochro-
Approximate calendaryear
& nology. Previously, the conservational biology of the
species have largely beenbased on morphometrical stu-
limati
dies, in particular the relationships between the shell
length and ontogenetic age (Bauer, 1991, 1992; Hastie et
Valovirta
ni., 2000; San Miguel et al., 2004). In the course of this
study, we have demonstrated how carefully examined
information from taphonomy and sclerochronology can
be used for the purposes of conservation biology and
ecology of the species. In the forthcoming sections, the
t 1 1 1 1 1 1 1 1 1 1 1 1 r
River Kotioja
Samplenumber
Fig. 5. Visual comparison of growth in individual sample series (thin
black line) and master chronology (thick grey line) after growth trend
modelling (Fig. 2). Temporal variation in sample size in RiverSaukko-oja
chronologyshownasgreyarea. Statistical comparison ofgrowth insam-
pleseriesand masterchronologyquantified bycorrelationcoefficients(r)
(histograms).
nFiegr.e)5.e cCroonnforloongtioa vmiesduialae(dleilnleae csrpeesscsietagrinigisee)riedoipnodivliadumaloide(llliinzezeazsoitotinlei 1MIMMI1 SIntirtiiaplpdeadtadsateatset
dell'andamento di crescita (Fig. 2). La variazionetemporale della dimen-
sione degli individui rispetto alla cronologia del Fiume Saukko-oja è rap- 1 21 31 41 51 61 71 81 91 110 111 112 113 114
presentata dall'area grigia. L'istogramma mostra il confronto statistico Samólenumber
fra la crescita individuale e la cronologia media attraverso il coefficiente
dicorrelazione(r). Fig. 6. Statistical comparison of growth in sample series and master
chronology(arithmetic meanofall otherseries)quantified bycorrelation
coefficients(r)(histograms) in River Kotioja. Sampleseriesof initial data-
both chronologies, in fact, the River Kotioja chronology set(grey histograms) having correlation lowerthan 0.4(dotted horizon-
showed a negative autocorrelation. tal line)with master chronologywere removed (Samples 7, 11 and 13).
Stripped dataset (black histograms) yielded master chronologywith en-
hanced common growth signal, indicated by higher intercorrelations
Inter-river comparison between the remaining samples series and master chronology (mean
Sample vs. Master correlations for initial and stripped datasets were
Sampled shells were provided to museum as dead spe- 0.445 and 0.526, respectively).
cimens and thus the actual year of death was not Fig. 6. Confrontostatisticofra la crescita individualeed ivalori di cresci-
known. As demonstrated earlier, however, individuals ta media, tramite il coefficiente di correlazione (r) nel Fiume Kotioja. Le
serie individuali(istogrammi grigi)aventi un correlazione più bassa di 0,4
were believed to have died at same time in each river. (linea punteggiata) rispettoallacrescita mediasonostati esclusi(campio-
Comparison between the two sclerochronologies could ni 7, 11 e 13). I dati dopo l'esclusione di questi campioni (istogrammi
in this work reveal their relative temporal position. It neri) hanno fornito una cronologia media con un migliore segnale co-
seems conceivable that the population in River Kotioja mune di crescita, indicato da una più alta correlazione tra le rimanenti
serie individuali e la cronologia media (correlazione rispettivamente pari
166 would have died 10 years before the population in a0,445e0,526, primaedopol'esclusionedeicampioni).
presentresults willbe related toprevious studies on the
conchologyofM. margaritifera.
Taphonomical interpretation implied that the shells had
likely been exposed to post-mortem conditions very re-
cently and that the valves were originally opened vio-
lently. Sclerochronological analyses showed that the
mussels had died during the same calendar year.
Moreover, the widths of the last increments in the stu-
died shells were exceptionally narrow without indica-
tion of winter-line thus indicating incomplete growth
during the last year of their life. These two lines of evi-
dence, taphonomical and sclerochronological, comple-
Conchology
mented each other in pointing to a simultaneous and an
abrupt death of the mussels, likely in connexion to ille-
of
galpearl hunting.
endargered
Growth variability and temporal synchrony
Prior to the temporal comparison of growth variability,
the growth trends were modelled. The life-long trends, freshwater
the growth trends, ofthe mussels show growth declines
from the juvenile life-stage onwards, in some of the ca-
pearl
ses the linear regression was actually better fitted to the
series than the modified negative exponential curve of mussel:
Fritts et ai. (1969) (Fig. 2). These results are consistent
with the results from the internal shell growth incre-
ments of Swedish M. margaritifera (Dunca 1999) and the conservation
increments of Spanish M. margaritifera observed from
the shell surface (San Migueletai., 2004). Fig. 7. Sderochronologies for River Saukko-oja and River Kotioja; thin
Comparisonbetween the indexed shell incrementseries black line representsthetotal growth variability, thickgrey lineemphasi-
revealed considerable growth synchronywithin the stu- zgneisfitchaentvlayriwabhielintytahterdeecwaadaslati1m0e-sycaelaersl(aag).bCehtrwoeneonlogtiheeslcasotrriedleantteidfiseid- palaeobiology
died populations (Figs. 5 and 6). Observed synchrony growthyearsin RiverSaukko-ojaand RiverKotioja populations(b).
implied that the growth of the mussels was influenced Fig. 7. Sclerocronologie per il Fiume Saukko-oja ed il Fiume Kotioja. La
by same climatic and hydrological variables and that linea sottile nera rappresenta la variabilità di crescita totale; la linea gri- applied
the individuals responded to external variability simi- gia spessa mette in evidenza la variabilità a scala decennale (a). Le cro- to
larly. The shell increment growth ofM. margaritifera and annonliogpieersil'cuolrtriemolainnocirnemmeanntioerdai csriegsncifiitcaatainvnauqalueaniddeontci'fèicuantortirtaarldaopdoìpo1-0
its correlation to climate has previously been studied lazioni del FiumeSaukko-ojaequella del Fiume Kotioja (b), museum
among the Swedish populations (Mutvei et ai., 1994;
Dunca, 1999). According to Dunca et at. (2005), the an- growth increments ofM. margaritifera were significantly shells
nual shell growth was positively correlated with sum- increased after an anthropogenic widened of pH and
mer temperatures in Swedish rivers with decreasing cli- food supply in Swedish rivers. Since the exact calendar originating
matic correlation in polluted rivers. More precisely, the years of the increments were unknown, the climate-
highestcorrelation (0.59) between annualM. margaritife- growth and environment-growth correlations could not
from
ra shell growth increments of the compound dataset be drawn in the present study. Even without rigorous
from northern to southern Sweden was found for June climate-growth-correlations at local-scale, the observed
through August temperatures (Schòne et at. 2004) and growth synchrony alone bears implications for the northern
the daily shell growth was found to co-vary with the chronology construction and inter-population compari-
growth season temperatures from April to October sons. The results from these approaches are discussed Finland
(Dunca et at. 2005). Moreover, the spatial variation in below.
the number of sub-annual shell growth increments of
M. margaritifera corresponded to number of days with Crossdating the increments
water temperatures higher than + 5 °C from north to
south Sweden (Dunca & Mutvei, 2001). It could be thus The presence of the growth synchrony is prerequisite
hypothesized that the observed shell growth variability for any sclerochronological study in which the principle
(Figs. 5 and 7) is positively controlled by summer tem- of crossdating is applied as a tool for data quality con-
peratures also in the present study region. Other con- trol and relative dating of each shell growth increment
trolling factors include food availability and the pH (Marchitto et at., 2000; Helama et at., 2006a). In the pre-
(Mutvei et ai., 1996; Mutvei & Westermark, 2001). For sent study, the process of crossdating was demonstra-
example Mutvei etat. (1996) found that the annual shell ted and validated visually as well as statistically. To 167
quantify the agreement between the series, we used ter-population correlations have previously been pre-
Pearson correlation coefficient as measure of common sented for M. margaritifera in Swedish rivers, implying
growth signal. This statistic may not be the only viable thatthedifferentpopulations exhibitparallel growthre-
statistic to provide similar estimates but its usage has sponses to regional climate (e.g. Mutvei et al., 1994;
been verified previously (Marchitto et al., 2000; Helama Helamaetal., 2006).
et al., 2006a). Moreover, the common signal, as quanti- If one aims to assign absolute calendar years to growth
fied by correlations between the sample series and ma- increments, it could be suggested that the presented
ster chronology (Figs. 5 and 6), was at a level compara- sclerochronologies could be crossdated with live collec-
ble with those described for trees in the dendrochrono- ted mussel shell growth increments or tree-rings. Whe-
logical literature (see Cook & Briffa 1990). However, as reas the collection of live specimens could be harmful
already emphasized by early developers of approach, for surviving populations, local tree-ring chronologies
the correlations or an equivalent statistic may serve as a could actually provide suitable biochronological coun-
Samuli crossdating guide, but careful visual inspection of the terpart for shell growth increments. In this context,
samples should guide the final decision as to whether a Helama et al. (2006b) were able to correlate annual shell
Helama, sample is correctly crossdated (Douglass, 1941). Given growth increments ofocean quahog (Arctica islándica L.)
that the environmental signal will likely be stronger for and the tree-rings of Scots pine Pinus sylvestris L.)
(
Jan some sites than for others, there isno ground for setting along the coastal zone of northern Norway. This sug-
a statistical limit for what constitutes correct cross-da- gests that the correlation between tree-rings and fresh-
Kresten
ting (Wigleyetal. 1987). water shell growth increments may also be possible.
Such inter-species comparison remains as future target
Nielsen Biochronological correlation for sclerochronologies with improved sample size and
& temporal length.
llmari Subsequent to crossdating, the mean growth chronolo-
gies, sclerochronologies, were produced and compared. From conchology to conservation
Valovirta Comparison was carried out by the statistics of mean palaeobiology
sensitivity, standard deviation and the first order auto-
correlation. Previously in Finland, the variability The study ofinternal growth increments is a prerequisi-
(quantified both by mean sensitivity and standard de- te for building reliable sclerochronologies whereas the
viation) of tree-ring chronologies has been shown to in- process itself is a destructive for shells. That is why the
crease from south towards north, that is, towards hars- use of museum or subfossil/fossil shells is particularly
her and more extreme growing conditions (Hustich, important in the case of endangered species, such as
1948; Mikola, 1950; Lindholm et ai, 2000; Helama et al, Margaritifera margaritifera. Collection of specimens alive
2005). could be especially harmful for populations that are no
longer breeding. Conchological studies may help to
The two sites (River Kotioja and River Saukko-oja) are identify the death conditions of the bivalves: fractures
located at similar altitudes and latitudes. The similarity cutting the dorso-posterior margin may indicate violent
of the sclerochronological statistics could therefore be opening ofthe shellsby knife and the last growth incre-
somewhat expected. Hastie et al. (2000) found depen- ment may bear evidence about the seasonal timing of
dence on growth characteristics ofMargaritifera margari- death. Variability and growth synchrony depicted by
tifera and river length in Scotland. Descriptive growth the sclerochronologies can be used as guidelines about
parameter used by Hastie et al. (2000) ("length-at-age") the ecology of the species. This is to apply the principle
was a measure of absolute growth whereas our stati- of conservation palaeobiology (Flessa, 2002) to study
stics (Tab. 1) measured temporal growth variability. the endangered species withoutreducing thenumber of
Accepting the perception of Hastie et al. (2000), that the individuals in theexistingpopulations.
growth characteristics of Margaritifera margaritifera co-
uld be used as river descriptives, it could be concluded Acknowledgements
that the two populations from northern Finland were
growing in similar habitats, when compared by the ma- The shells of Margaritifera margaritifera were collected
gnitudes of environmental variability. If one aims to under the license from The Lapland Regional Environ-
compare the absolute growth rates and to assess their mentCentre and NorthOstrobothnia Regional Environ-
biological relationships to the environment, more detai- ment Centre. We are greatly indebted to Robert Wor-
led information for example about the morphometrical thington who kindly revised the language of the paper.
variables would be needed. Ideally, integration of mor- Our gratitude and thanks goes also to Jürgen Tischack
phometrical and sclerochronological informationwould and Claudio Mazzoli due to their comments on the ear-
be preferred. This remains as target for an improved lier version of the manuscript. Figure 1 was aided by
model. Online Map Creation. The work ofSH was made possi-
Moreover, a reasonable biochronological correlation ble by a postdoctoral scholarship from the Foundation
was observed for the populations that implied that the of Koneen Sààtiò. The Carlsberg Foundation kindly
population in River Kotioja would have died ten years supported JKN through a postdoctoral scholarship
168 before the population in River Saukko-oja. Similar in- (project04-0256/20).
, -
References Helama S., Schòne B.R., Kirchhefer A.J., Nielsen J.K.,
Rodland D.L., Janssen R., 2006b. Compound response of
Araujo R. & Ramos A., 2001. Action plansfor Margaritiferà marine and terrestrial ecosystems to varying climate: pre-
auricularia and Margaritifera margaritifera in Europe. Na- anthropogenic perspective frombivalve shell growth incre-
tureand Environment 117,CouncilofEurope,65pp. ments and tree-rings. Marine Environmental Research, 63:
Bauer G., 1986. The status of the freshwater pearl mussel 185-199.
Margaritifera margaritifera L. in the south of its European Holmes R.L., 1983. Computer-assisted qualitycontrol in tree-
range. BiologicalConservation, 38: 1-9. ringdatingandmeasurement. Tree-RingBulletin,43:69-75.
Bauer G., 1988. Threats to the freshwater pearl mussel in Huntington E., 1914. The Climatic Factoras Illustrated in Arid
CentralEurope. BiologicalConservation,45: 239-253. America. Carnegie Institution Publication 192. Carnegie
BauerG., 1991. Plasticityinlifehistorytraitsofthefreshwater InstitutionofWashington,Washington,341 pp.
pearl mussel - consequences for the danger of extinction Hustich I., 1948. The Scotch pine in northernmost Finland
and forconservation measures. InSeitzA. & LoeschckeV. and its dependence on the climate in the last decades. Acta
(Eds.), Species Conservation:A Population Biological Approach. BotanicaFenìlica,42: 1-75.
Basel,BirkhàuserVerlag, 103-120. Jones D.S., 1981. Annual growth increments in shells of Conchology
BauerG., 1992. Variation in thelifespan and sizeofthefresh- Spisula solidissima record marine temperature variability.
waterpearlmussel.JournalofAnimalEcology, 61:425-436. Science,211: 165-167. of
Carroll M., Kowalewski M., Simòes M.G. & Goodfriend, Lindholm M., Lehtonen H., Kolstròm T., Merilàinen ].,
G.A. 2003. Quantitativeestimates oftime-averaging in tere- Eronen M. & TimonenM., 2000. Climatic signalsextracted
bratulid brachiopod shell accumulations from a modern from ring-width chronologies of Scots pine from the endargered
tropicalshelf. Paleobiology, 29: 381-402. Northern, Middle and Southern parts of the boreal forest
CookE.R. &BriffaK.R., 1990.Acomparisonofsometreering beltin Finland. SilvaFenìlica,34: 317-329.
standardization methods. In Cook E. & Kairiukstis L.A. Marchitto T.M.,Jones G.A., Goodfriend G.A. &Weidman
freshwater
(Eds), Methods ofDendrochronology: Applications in the Envi- C.R., 2000. Precise temporal correlation of Holocene mol-
ronmental Sciences. Dordrecht, KluwerAcademic Publishers, lusk shells using sclerochronology. Quaternary Research, 53
:
153-162. 236-246. pearl
Cook E.R., Briffa K.R., Mero D.M., Graybill D.A. & Funk- Mikola P., 1950. Puiden kasvun vaihteluista ja niiden mer-
houser G., 1995. The 'segment lenght curse' in long tree- kityksestà kasvututkimuksissa. Summary: On the varitions mussel:
ring chronology development for palaeoclimatic studies. in tree growth and their significance to growth studies.
TheHolocene,5: 229-237. CommunicationesInstitutiForestallsFenniae,38: 1-131.
Douglass A.E., 1941. Crossdating in dendrochronology. Mutvei H. & Westermark T., 2001. How environmental in- conservation
JournalofForestry,39: 825-831. formation canbe obtained from Naiad shells. Ecological stu-
Dunca E., 1999. Bivalve shells as archives for changes in wa- dies, 145: 367-379.
terenvironment. Vatten, 55: 279-290. MutveiH.,WestermarkT.,DuncaE.,CarellB.,ForbergS.
Dunca E. & Mutvei H., 2001. Comparison of microgrowth & Bignert A., 1994. Methods for the study ofenvironmen- palaeobiology
pattern in Margaritifera margaritifera shells from south and tal changes using the structural and chemical information
northSweden.AmericanMalacologicalBulletin, 16:239-250. in molluscan shells. Bulletin de ITnstitut océanographique,
Dunca E., Mutvei H. & Schòne B.R., 2005. Freshwaterbival- Monaco, n°special,13: 163-186.
vestell ofpastclimates: Buthowclearlydoshells from pol- Mutvei H., Dunca E., Timm H. & Slepukhina T., 1996. applied
luted rivers speak? Palaeogeography, Palaeoclimatology, Pa- Structureand growth rates ofbivalveshells as indicators of to
laeoecology,228 43-57. environmental changes and pollution. Bulletin de ITnstitut
:
Flessa K.W., 2002. Conservation paleobiology. American Pa- océanographique,Monaco, n°special, 14:65-72. museum
leontologist,10:2-5. Nielsen J.K., 2004. Taphonomy in the light of intrinsic shell
Fritts H.C., 1976. Tree Rings and Climate. Academic Press, properties and lifehabits: marinebivalves fromtheEemian
shells
London,567pp. ofnorthernRussia. PaldontologischeZeitshrift,78:53-72.
Fritts H.C., Smith D.G., CardisJ.W. & Budelsky C.A:, 1965. San Miguel E., Monserrat S., Fernández C, Amaro R.,
Tree-ring characteristics along a vegetational gradient in HermidaM., Ondina P. & AltabaC.R., 2004. Growth mo- originating
NorthernArizona. Ecologi/,46:393-401. dels and longevity of freshwater pearl mussels Margariti
(
Fritts FTC.,MosimannJ.E. & BottorffC.P., 1969. Arevised fera margaritifera) in Spain. Canadian Journal ofZoology, 82:
from
computer program for standardizing tree-ring series. Tree- 1370-1379.
RingBulletin,29: 15-20. Schòne B.R., Dunca E., Mutvei H. & Norlund U., 2004. A
Hastie L.C.,Young M.R. & BoonP.J.,2000. Growthcharacte- 217-year record of summer air temperature reconstructed northern
ristics of freshwater pearl mussels, Margaritifera margaritife- from freshwater pearl mussels (M. margaritifera, Sweden).
Herlaa(Lm.)a.FS.r,esLhiwnatdehroBliomloMg.y,,4T3i:m2o43n-e2n56M.. & Eronen M., 2004. ScQhuòanteernBa.rRy.,SciDeuncnecRaeviEe.z,o,F2i3e:b1i8g03J-.18&16.Pfeiffer M., 2005. Finland
Detection of climate signal in dendrochronological data Mutvei'ssolution: anidealagentforresolvingmicrogrowth
analysis: a comparison of tree-ring standardization me- structures ofbiogenic carbonates. Palaeogeography, Palaeocli-
thods. TheoreticalandAppliedClimatology,79: 239-254. matology,Palaeoecology 228: 149-166.
Helama S., Lindholm M., Merilàinen }., Timonen M. & Seire A., Mutvei H. & Dunca E., 1993. Growth rates and
Eronen M., 2005. Multicentennial ring-width chronologies structure of Macoma balthica shells, collected on the north
of Scots pine along north-south gradient across Finland. coast of Estonia, as environmental indicators. Institute of
Tree-RingResearch, 61: 21-32. Ecology,EstonianAcademyofSciences, Publication 1/1993,
Helama S., Schòne B.R., Black, B.A. & Dunca, E., 2006a. pp. 169-195.
Constructing long-term proxy series for aquatic environ- Skinner A., Young M. & Hastie L., 2003. Ecology of the
ments with absolute dating control using a sclerochronolo- Freshwater Pearl Mussel. Conserving Natura 2000 Rivers
gical approach: introduction and advanced applications. EcologySeriesNo.2EnglishNature,Peterborough, 17pp.
MarineandFreshwaterResearch, 57: 591-599. Tanabe K., 1988. Age and growth rate determinations of an 169
intertidal bivalve, Phacosomajaponicum, using internal shell
increments. Lethaia,21 231-241.
:
Valovirta I., Tuulenvire P. & Englund V., 2003. Jokihelmi-
simpukanja sen elinymparistonsuojelun tasoLIFE-Luonto -pro-
jektissa. Luonnontieteellinen keskusmuseo, Finnish Mu-
seumofNaturalHistory,Helsinki,56pp. [InFinnish]
Wigley T.M.L., Jones P.D. & Briffa K.R., 1987. Cross-dating
methods in dendrochronology. Journal of Archaeological
Science, 14 51-64.
:
Young M.R., 1991. Conserving the freshwater pearl mussel
Margaritifera margaritifera L. in the British Isles and conti-
nental Europe. Aquatic Conservation: Marine and Freshwater
Ecosystems, 1: 73-77.
ZiuganovV., San Miguel E., NevesR.J., LongaA.,Fernán-
Samuli dezC, Amaro R., Beletsky V., PopkovitchE., Kaliuzhin
S. & Johnson T., 2000. Life Span Variation of the Fresh-
Helama, water Pearl Shell: A Model Species for Testing Longevity
MechanismsinAnimals.Ambio,29: 102-105.
Jan
Kresten
Nielsen
&
llmari
Valovirta
Lavororicevutoil3 maggio2006
170 Lavoroaccettatoil -tmaggio2007