Suspension sedimentation of coarse silt and sand in northern Skagerrak: textural and mineralogical trends

(1)

Suspension sedimentation of coarse silt and sand in northern Skagerrak: textural and mineralogical trends

AIVOLEPLAND&RODNEYL.STEVENS

Lepland,A.&Stevens, R.L.1996:Suspensionsedimentatio nofcoarsesilt andsand in northern Skagerrak: textural and mineralogi caltrends. Nor.geol.unders.Bull.430,35-46.

Theminorcoarse-silt andsand fractionof muddybottomsediments at74sitesin northern Skagerrak isinterp reted tohavebeen mostlytransport edinsuspension, aided by frequent reworkingand thebuoyancy of large aggregates.

Theoverall sediment fining trend towardsthe Norwegian Trench is consist ent wit h changes of bathym etry.

Sedi ments inshallowshelfsett ingsare mainlytransport edby acombinati on oftractionand near-bottom suspension processes.Near-bottom suspension processespredom inateon thetrenchslopes,whereasin the bottom of the trenchthereis amorebalanced combinatio nofnear-bottom suspensionandhemi pelagicsuspension depositi on.

Heavy-m ineraldistributi on shave weakgeographic trend s, due tothenatural variability,and relativelynon -selecti ve andrando m characterof the processestransport ing coarse-siltandsandinsuspension.Only biotitehas a welldeve- loped distri butio n pattern andarelati vely good correlation wit h grain size, probablyrelatedto single-partic le trans- porttogethe r wit h fine silt.Wit hconsiderat io n for the inte rpretat ive limitation s oftheindividualheavy-min eraldis- tributi on s,it is valuable to evaluateseveral mineraltrend s simulta neously,as wellasthe grain-size trend s,in order to inter pretpatte rns. Transport fromsout hern areasand fromthenorthisindicated.

AivoLepland" &RodneyL.Stevens',(7)Departme ntof Geology,Earth SciencesCentre,GoteborqUniversity,5-41381, Goteborq, Sweden,(2)Institute of Geology,Eston ianAcademy ofSciences,EE-OOO lTallinn,Estonia.

Introduction

The grain-size distribution of the bottom sediments in the Skagerrak is clearly controlled by the circulation pattern and basin bathymetry, resulting in different geographically associated transport and deposition mechanisms (Olausson 1975, van Weering 1981, 1982, Jerqensen et al. 1981,Fait1982,Stevens et al.1996). The transport to the Skagerrak and Kattegat of suspended sediment derived from shallow areas of the southern and centralNorth Seais well documented (van Weering 1981, Eisma&Kalf1987a,b,Eisma&Irion 1988),and ca.50-70%

accumulatesin the areasof low current velocity,especially in the Norwegian Trench and in the surrounding areas with water depths greater than 200 m (Eisma & Kalf 1987a, Eisma1990,Salge&Wong 1988,North SeaTask Force 1993). This predominantly suspension-d erived depositioninnorthernSkagerrakisrespon siblefora very fine and uniform sediment texture,consistingnormallyof morethan 90%clay and finesilt.Thesed iment grain-size distribution ischaracterisedby a majo rclaymode and a slig ht mode in the coarse-silt fraction (Stevens et al.

1996),whereas the content of very coarse silt andsand is negligible(1-2%). The bimodalityof these sedimentshas been related to a combination of suspension and tracti - on processesfor the deposition of the fine and coarse fractions,respectively(van Weering 1981,Pederstad et al.

1993). The coexistence or alternation of these two mechanisms isnot well understood (van Weering et al.

1993).

A principle objective of this paper is to addre ssthe transport and depositional processesof the very-coarse-

silt and sand population (coarser than 6<jl), which is consi- stently presentin the very fine-grained sediments of the Norwegian Trench. The int erpret at ion of fine-sediment transport is also necessary since we suggest that suspension processes strongly influence both fine and coarse fractions. The emphasis upon coarse silt and sand is moti- vated because in these fractions the mineralogy can be routinely and quantitatively documented using petro- graphic microscopy .The combined treatment of sediment mineralogy and texture is used to evaluate the sediment transport pathways and processes.This approach is especially im port ant in complex environments where single parameters are less diagnostic.

Materials and methods

The sam ples investigated within this study were collec- ted from 74 locations in northern Skagerrak (Fig. 1) during two cruises, in 1992 and 1993, with the Bergen University research vessel 'Hakon Mosby', arranged by Bergen University and the Geological Survey of Norway.

Surfacesediments were subsampled within the upper 7 cm from short co res of approximately 50 cm,obtained usinga modifiedNiemi stbcorer (Niernisto 1974).

Parti cle-size analyses of the samples were carried out using wet sieving at'/ 2-<jlintervals for the fractions coarser than 5.5<jland pipette analysisfor the finer fractions (see Krumbein &Pettijohn 1938, McManus 1988,for stan- dard procedures). The pre-treatment of the samples included the oxidation of organicmatter with hydrogen peroxide, subsequent washing and final dispersion in

(2)

36 Aivo Lepland & RodneyL.Stevens

saON

...,...~...,ikm

o

20 40

0.05 M sodium hexametaphosphatewhile rotation tumb- ling for4 hours.

Thecharacterisation of sedimentmineralogy is based upon the qualitative analysisof x-ray diffractio n using powderedsam ples andthe quantitative microscope ana- lysis of the heavy-mineral concentrate (density > 2.9 q/crn'),separated using sodium polytungstate (Callahan 1987).Graincounts were performedwit hin a narrow size range of the very-coarse-siltfraction (4.5-4<1» in order to limit hydrodynamic differences.The fine-sandfraction (4-

3<1» istypicallyused for heavy-mineralanalyses(Mort on&

Hallsworth 1994),but a finer fractionwas chosen in this studybecause of its relatively higher content compared with fine sand.Ribbon-traversecounts of up to 800 grains (t ranslucent heavy mineral s, opaque minerals,carbon a- tes,micas) were made in order to include at least 300 translucentheavy mine rals (Krum bein&Pettijohn 1938).

The 8 presented contour maps of mineral percentages refer to the content with inthe heavyconcentrates.These percentages are calculated differently to avoi d depen- dencyon the variati onsof predominant minerals. Biotit e content sare givenas percentagesof allcounte d grains;

carbonate contents are percentagesof allgrains except micas;magnetite and hematite contentsare percent ages of opaque minerals plus translucent heavy minerals;

hornblend e, epidote gro up, garnet and pyroxene con- tentsarepercentages ofthe translucent heavy minerals.

In 20 selected samples, the proportionsof terrigenous (mainly quartz and feldspar)and bioge nic(mainly forami- nifera and shellfragm ent s)comp onent sin the sediment were determinedby count sof200 sand grains.

Results

Sediment characterand grain size

The uppermost50 cm of the sediment coresfrom northern Skagerra kare quite homogeneous,character ised by an alm osttotal lack ofapparent sedi ment arystru ctu-

GU-BULL430.1996

59"N

Fig.1.Sample stations and generalised bar- hymerry in northern Skagerrak. Arrows showrhedirection ofthe main oceanograp- hic surfa ce currents, rhick arrows corres- pon drostrong flo w and thinarrowsroweak flow(after Svansson 1975).

resexcept for a down-core increasein consolidati onand adecrease in water cont ent. The top 10-15cm of most cores consist of very loosematerial. Thisunconsolidated materialatthe sedimentsurface is absent in samplesori- ginating from the southwestern area where relatively coarse sedi mentsare common.

In spite of the predominance of clay and fine-siltparticlesin northernSkagerrak,a limited amount of sand«4

<1» and verycoarse silt (4-5<1» is always present in these

sediments.The sedimentgenerallycontain slessthan1% sand,and inmost areasof the NorwegianTrench deeper than400 m the sand fractiondecreasesto less than 0.4%

(Fig. 2a).Sand content decreasesfrom 8 to 0.4%north- ward,downthe sout hernslope,and from 3to0.4%inthe northeastern endof the trough. Thesand sizeis generally limi t ed tothe very-fine-sand fraction,main ly wit hi n the interval 3.5-4 <1>, which constitutes 50-70% of the total sand fraction.In most of the area only occasional grains offinesandor coarser grains occur.

Grain count s ofthe sand fraction show the predomi- nanceof terrigenous (quartz and feldspar) components in most of the study area.Terrigenouscomponents tend to bemost abundant in the relatively coarsesedimentsin the south, the nort hw est,and along the Norwegian mainland,wherethey consti tute 70-85%.In the deepest area ofthe NorwegianTrenchthe contentof biogen iccompo- nents increases, accounting in some samplesfor up to 75%of thesandgrains.

Thecont oursfor very coarse silt(4-5<1» followthe same pattern as for sand,but the contentis approximatelyone order of magnitudehighe r(Fig 2b). Inthedeepest partof the Norwegian Trench the bottom sediment consists of 1-1.5%verycoarse silt,andtow ards the Norwegian main- landthe valuesgraduallyincrease to 2-3%.A pronounced down-slope decrease of the very-coarse-silt content occurs on the northeastern and southeastern slopes of the No rweg ian Trench, where the values change from

>10to<2%. Therelat ive abundance of the 4-4.5^<1>fraction appearsto deviate slightly from the systematic trend ofdecreasing coarse-fraction abundance(seen only wit h

(3)

NGU-BULL430,1996 AivoLepland&RodneyL.Stevens 37

11°00' E 10"30'

10"00' 9"30'

9"00'

~7.-

. ~ ~~\('

~ ^· ^W? ^. _./~ ^.

^~~

^. ^. ^. ^.

^.VERY COARSE SILT^~

^. ^. ^. ^. ^. ^' ^.

4-5<jI (%of total) b

8"30'

57°40'I----,---,----c----,----~'--__,-'

58"00' N 59"00'

58"20' 58"40'

11°00' E

10"30' 10"00'

9"30' 9"00' 8"30'

a

SAND <4<j1 (%of total)

57°40'l -_ - , -_ _--,-_ _----, , - -_ _

~

58"00' N 59"00'

58"20' 58"40'

11°00' E

CLAY >8<jI

(%of total)

10"30' 10"00' 9"30'

9"00' 8"30'

d

57"40'1 - - _-,--_ _ --,- 'T"":"_ _..., -' -_ _ _ , - '

58"00' N 59"00'

58"20' 58"40'

11°00' E

10"30' 10"00'

9"30' 9"00'

c::=S-:-\0-

-3 ~

- ~ ~/

_, /[~~ ,j

~~20

VERY COARSE SILT AND SAND

<5<jI(%of total)

8"30'

c

57°40'' - -_ -,.-_ _ ---, ,--_ _,--_ _-'''i--'-_- ,- '

58"00' N 59"00'

58"20' 58°40 '

Fig.2.The distrib utionof(a) sand«4rp);(b)verycoarsesilt(4-5rp);(c)sand and very coarsesilt«5rp);and(d) clay(>9rp),

half <\J intervals, Fig. 3).This results in a slo p e break or

slight peak at 4-4.5<1>on the size distribution curves.

A negative correlation expectedlyoccurs between the coarsest« 5<1» and finest (>9<1>= clay) fractions (Table 1, Figs,2c and 2d),but notably not in the deepest part of the Norwegian Trench,where the highest abundance of clay and the lowest abundance of very coarse silt and sand do not geographically coincide with each other.In comparisonwithother grain-sizeclasses,the claycont ent shows the most pronounced relationship wit h water depth (r= 0.653,Table 1).Coarse silt is morenegatively correlatedwit h bothwater depth andclay thanis fine silt.

In other words,the changes in the clay content in northern Skagerrak are largely compensated by opposite variat ion sin the coarse-siltcontent,whilethe proportion of the fine siltremains relativelystable.

Mineralogy of theheavy-min eralconcentrate

The components of the heavy-mineralconcentrate (refer- red to below as heavy minerals) are divided into four major groups: micas, carbonates, opaque and translucent. Micas are predominant in most extracts, and among them biotite is more abundant than muscovite.Biotiteis the only mineral which correlates relatively well with sediment grain size(r

=

^0.433,^{Table 2).}The highest biotite contents (30-35%) occur in the deepest part of the NorwegianTrench and to the northeast(Fig. 4a).Biotite contents gradually decrease towards the Norwegian mainland, to 10-15%. On the Danish side of the Norwegian Trench,lower biotite percentages occur in the northeast« 20%)and southeast« 10%),whereas in the central part of the study area, 30% isocon lines extend almost perpendicular to the slope.

The carbonate grains are mainly dolomite, but since the optic properties of calcite and dolomite are poorly

(4)

38 AivoLepfand&RodneyL.Stevens NGU-BULL 430,1996

Table1.Correlatio ncoefficients forthe grain-sizefractionsand water depth.

Table 2.Correlationcoefficients for themean grainsizeofthe sedimentsand the separateheavy-mineral components.

Fig.3.Grain-sizefrequency distributionsshowingtheoverall finingtrend from the slopes ofthe NorwegianTrench(a,b,c)towardsits bottom(d).

Sediments of the Norwegianslope(b)and trench battom(d)showthe rela- tively greater abundan ceofthecoarse-siltfractiondueto the biog enic deposition in theseareas.The resulting curvebreak is indicated by the arrows.

Sing le grainsof leucoxeneand pyrite occur in a few samples,The hematitecontent is consistentlybetween 4 and 8%in the central part of studied area.It increases to the east,at a few sites to >12% (Fig.4c)_A generallydecrea- sing trend of hematitecontents can be follo wedtowards the north and southwest,The magnetite distribution is much more variab le,with severalzones oflow« 7%)and high(>11%)content orientated roughlyperpendicularto the bathymetriccontours(Fig, 4d),

Translucentheavy mineralsare characterisedbya predominance of hornblende,which accounts for45-60%of all minerals in this group.The hornblende distribution pattern is fairly irregular (Fig, 4e),changing between 45 and 60%over relativelyshort distances,Hornblende contentis generally higher(>55%)in the nort hern and southeastern parts of the study area, whereas low values occur in patches along theNorwegia n Trenchand to the south.The second most abundant catego ry is the epido - te group,namely epidote,zoisite and clino zoisite, which are considered together because of their transit iona l optic properties,In most of the area the epido te group variesunsystematically bet we en 22 and30%,reaching a maximum value (>32%)inthesouthern part and mini- mum values«20%)in thewestern part(Fig,4f),One map ispresented for all pyroxenes because of the low contents of individual minerals of both mon oclinic and orthorhom bicvarieti es.Among pyroxenes,hypersthene and augite areoft en most abundant,alt houghenstatite and diopside also occur in themajority of samples.The pyroxene contentvariesrath erirregul arlyfrom less than 5%tomorethan 11%,com monly ea.8%(Fig,4g).Aslight decrease towards the Norwegian mainl and isobserved, Garnet ismainlyrepresented by acolourless variety,and the total content isusually betwee n4 and 10% (Fig, 4h).

Zircon and sphe ne(titanit e) are present in mostof the sam ples.Thecontentsrangefrom lessthan 1%upto 7%, normally 1-2%.Tou rmalin e,stauro lite,kyanite,andalusi- te,sillimanit e,rut ile,apatite and barite weresporadically found in a lim it ed num ber of samples, usually in trace amounts,

The correlation coefficient s, which are mainly wit hin the range 0,2to -0.2,indi catea lackof sig nific ant relationships between the different minerals or between mineral varieti es and grainsize(Table2).The slight correlatio n between bioti te and grain sizeand hornbl ende's weak negative correlations wit h epidote, pyroxene and garnet are theonlyexcept ions,The correlations between epidote,pyroxeneand garnet are verylow,showing similar values, roughly one order of magnitude lower than their correlatio n with hornblende . This indicates that among translucent heavy minerals there are no strong associationsin the population as awhole,The content changes of horn blende are randomly compe nsated by epidote,pyroxeneorgarnet,but noneofthese minerals has priority.

8

Pyr Ep Hem Hlb 20.0 1

lC ~..,

10.0 ~ r' ~

5.0 -1

{f:ri

^Z

"*^2.0

j

^r ^{L I '}_-r» ⁽^I ^{DANI SH}SLOPE

1.0 I~-y/

05

~ iF

0.1

~ ¥f __ _~ --

9 3 6 8

20.0 1

d

>~

1

10.0 -1

5.0 J

t A

^.'

';f.2.0 - TRENC H BOTTOM

Dol Mag

7 8

St 5 3 20.0 l b

1

10.0 -1

n=74 waterdepth clay(>90)

clay(>90) 0.653

fine-silt (6-90) -0.467 -0.549

'coarse-silt'(4-60) -0.501 -0.827

v.coarse silt(4- 50) -0 .455 -0.779

sand(<40) -0.29 1 -0.596

n=74 mean St 0.433 Dol -0017 0.220 Mag 0.182 -0.104 -0.051 Hem -0.263 0.433 0.053 -0.140 Hlb -0.056 0.148 -0.213 -0.159 0.275

5 6

Q

distinguished in grain mounts theywere not subd ivided.

The content of carbonates generally increases from the northeast « 10%) to southwest (>20%),with one inter- mittentzoneof lowervalues«10%)in between(Fig.4b).

The trends crossthe grain-size and bathymetry contours.

The opaqu e mineralsare mainly represented by the magnetite-ilmentiteand hematite-Iimonitegroups(refer- red to below as magnetite and hematite, respecti vely).

Q

~

" C

! , r

"*^2.0_1.0 J_-:

₁₁ _-'

-.

r ^fi

^NORWEGIAN_{SLOP E}

• J r;?;/

0.5 -1

1 ft ,;

J

:~. If(

1Y;\

0.1-J •_~...:..-J, • _

:: : 11 ~ 4 ~

"*⁰²¹

o~

^[

I(~

UPPER SLOPE^DA^NISH

..,

0.1.J;-: ---,---..,

Ep 0.067 0.025 0.248 0.129 Pyr 0.139 -0.172 0.003 0.132

0.125 -0.508 -0.175 -0.475 0.057 Gar 0.138 -0.190 0.101 -0.053 -0.136 -0.551 0.001 -0.041

(5)

NGU-BULL430,1996 AivoLepland& Rodney L.Stevens 39

Discussion

Fine-grainedsedimentation

The sediment samples are not time or event specific,since they include deposition over numerous years and give a time-integrated, net reflection of sedimentation. Therefore, the additional influence of bioturbation is expected to further homogenise and integrate the sam- pled sediment,as is assumed for the surface subsamples.

Although we do not have informationregarding possible size-specific enrichment by biogenic processes, we do not expect this to have an overriding effect, and the grain-sizedata interpreted below do not seem to require such an explanation .

The unsorted character of the laboratory-d isagg rega- red, inorganic, fine-silt and clay fractions of the bottom sediments(Fig .3) is consistent with sedimentation largely from flocculated suspensions. Van Weering et al.(1993) have shown that large amounts of suspended sediments are transported within the bottom nepheloid layer in the

Skagerrak.The sediment concentration (transmission)of this layer was reported to decrease towards areas of increasing water depth, consistent with continued sedimentation along the paths of prevailing oceanographic currents.In the deepest portionof the Norwegian Trench, no distinctive differences in transmissionwere observed separating the intermediate and bottom nepheloid lay- ers, suggesting the predominance of hemipelagic suspension deposition(cf.Stow 1985) in this area.

It has been demonstrated that the flocculation process (including biogenic and inorganic aggregates of several types,cf.Syvitski 1991)is not size-selective and that similar contributions from all size fractions result in relatively flat frequency distributions (Kranck 1975, Kranck &

Milligan 1991).The only portions of the inorganic grain- size distribution that are consistentlysorted in suspension transport are the coarsest fractions where the distribution curve falls off sharply (Fig .3). This may indicate that minor single-particle transport occurs simultaneously with the predominant floc transport, which allows the hydraulic sorting of coarse fractions. Alternatively,larger, heavier floes enriched with coarser particles might also

10"30' I

10"00' 9"30'

9"00'

,

6"30'

b

56"00' N 59"00'

~ 'O

J " · . · .c

. .

. . .

~~~~

^.^j ^.

^tf ^d ^s ^.:/6' ^. ^-,

.

' ,~

CARB ONATES (%of allheavymineralsexceptmicas)

57"40· ~

^_ ^_ ^-r-r-'_ ^_ ^...,...-_ ^_ ^.

i f

56°40'

56"20'

10"00' 10"30' 11°00' E

BIOTITE

(%of allheavy minerals)

9"30' 9"00' 6"30'

a

N 59"00

57"40·' - -_ --::r_ _--,_ _ -t-r-_ _ - , -_ _ -4L_--,-~

56"00' 56°40'

56"20'

N 59"00'

56"40'

56"20'

56"00'

c

HEMATITE

(%of translucent and opaque heavy minerals)

N 59"00'

56°40'

56"20'

56"00'

d _._/

.

J ' , ..--

'-... '

.~s .

.

^' ."~

. .

(-

'~'l\~'

.: '::): 8~ · ' .

.~,~ ) .

MAGNETITE

(%of translucent and opaque heavy minerals)

57"40'L.--::r-:-:---,----,---.----4-L--,-~

(6)

40 AivoLepland&RodneyL.Stevens GU-BULL430, 1996

N

5gooo'

e

58"40'

58"20'

58"00'

"'\r2

~

⁵⁵^-

. ' ( SO

.'~.'

. ~~.

i ;~'\Jjro ~

'%~,

HORNBLENDE (%of translucentheavy miner als)

N 59"00'

f

58"40'

56'20'

58"00'

.r~. tJ

' p~'l-"~

"/ .

^, .

\~?X~;:

J f O.Uzj ·

'. ~ ~ ' . . _(1~'

~J

_.~ ' .

"

EPIDOTE

(%oftransluce nt heavy minerals)

10'00' 10' 30' 11"00' E 9"3D'

9"00' 6'30'

57·40''--- - r -- - - , - ---,- ---,- --4-.L----r--'

11·00' E 10'00' 10'30'

go30'

,

goOD' 8"30'

57·40'' - - - , - ---,- ----,---r- - -4-L - --,.--.J

N 59"00

58"40'

58"00'

PYROXE NE

(% of translucent heavy minerals)

N 59"00

58"40'

58"20'

58"00'

h .~. ) ^.

• • ' . 6~

. ~ . \q? .

~ ^; ^~ ^.. ^~

.

C~

.

^4'8

4 · ·· ·

^{" 4 ;} ^' ^"

(, . . . . r •

• • • B ' \ '

· .~8·

~ GARNET

(% of tran slucentheavy minerals)

1eco 10' 30' 11"00' E 9"30'

6'30'

57"40'' - - ---,-- - ---,-- - --.-- - -..,..- - -4....L-- --,---.J

10'00' 10'30' 11"'00' E 9"30'

9"00' 6'30'

57"40'-,-'----,c----,- - - -- --,.- - - 4....L..- --,.--.J

be hydraulically sorted. Single-particletransport is interpreted to be reflected,at least,by the systematic increase of mica towards the deeper portions of the Norwe gian Trench.Flocsort ing as a control of the silt-sized micais unlikely since theflocculation processwouldnotincorpo- rate micapreferentially into smallerfloes,whichare inter- pretedto be increasingly represented in calm and deep areas, as discussed below.

Single-particletransport and hydraulic sorting proba- blyalsooccurtoa limited exten t withi n fine-silt sizes,but this is maskedbythe predo mi nance of unsorted flocculated particles. Studies by Kranck & Milligan (1985) have show n that singl e-particl etransport can beeffective for sizes as small as 5-10 urn.Eisma (1993) hassugg ested that the coexistence of floes and single particles insuspension,as indicated by variable Silt/clay ratios,is related to the sediment concentration and is favoured only in verylow concentrations« 0.3 rnq/l ),whereflocculation is limited because of low collision frequency.The suspended-sediment concentration in the area of the Norwegian Trench is approximately at thiscrit ical concentration(Eisma&Kalf 1987b),theor eticallyallowing the

transportoffine-siIt-sizedsingleparticles.

The relationship between thefine-silt (6-9<jl) and clay contents in the sedim ents of the Norwegian Trench changes with the relative grain size(Fig. Sa).An overall posit ivecorrelation (clay increase is coupledwith fine-silt increase) exist s for relatively coarse sediments « 45%

clay) and a negative correlation for fine sediments.The consistent increase of clay and fine silt in coarse sedi- mentsindicates that both clay and fine silt occur in floes and are not influenced by differential sorti ng inareas of relativelygreaterturbulence. The oppositetrend for rela- tivelyfinersedimentssuggests that a selective sorti ng of fine silt doesoccurin some areas,presumably indicat ing the simultaneous deposition of fine silt as both single part iclesand floes. The comparison of relationships between clay and three sub-fractionsof fine silt (1-<jl intervals)show sthathydraulic sorting is most eff ect ivein the coarsestportionof finesilt (6-7<jl; Fig. Sb),whereas the evidence for sorti ngin the finer sub-fractions(7-8, 8-9<jl;

Fig.Se,d)is lesswelldefined,but aweakly defined'break point' is interpreted to successively shift toward higher clayconte nt s.

(7)

NGU-BULL430, 1996 Aivo Lepland&RodneyL.Stevens 41

b

40

60 40 50

30 40

"0' 3 ^"0' 30

C C

~ ~

0\ r-

I I

\0 \0

....

₂₀ 20

'Uj 0 0

0

Q)

•

~c

10 10

clay >9<jl ("10) clay>9<jl ("10)

c d

40- 40

. .. .-

o

M·... -'

. ^- ^{, . .}

~ ... ~ . ... ..- .~";' .

-- -- .. . ....

^...

-- .. . . . .

^...

-,

o

. _:- _{_-1"-'} ..

^~-~-;~.~

^, _- _0: __ "'.... ^. .

_• _-:r _._... • •• "

~.

. ^,.

"

~ 30-

~ 00 r--I

20-

10-

I

30 40^I clay >9 <jl ("10)

I

50

I

60

"0' 30

C

~

0\

I 00

20

10

_--- ---

I

30

I

40

clay >9<jl("10)

I

50

I

60

Fig.5.Relat ionshipsofclaytotota lfine silt,6-9 11.(a),and claytothesub-fractionsoffine silt,i.e.6-7 11(b),7-81fJ.(c),and8-91fJ.(d).Thenegati veslopeisindica- tive ofhydraulicsortingof the fine-siltparticlesin relat ivelyfine sediments(>45%clay;a).Themosteffectivehydraulicsortingoccursin thecoarsestportionof fine-silt(b),whereassortingislessdeveloped inthefinersub-fractions(c; d).Thetrendsare visually estimated,solidlinesimplyingstrongevidence anddashed linesimplyingweakevidence for hydraulic sorting.

Most grain-sizedistribut io ns can bedivided int otwo slightlyoverlapping subpopulati onsin whicheitherfloes or sing le particlespredominate.The transit ion between these two is marked bythe truncation pointon thedistri- bution curve where the coarse end fallsoff.Most of the sediments in northern Skagerrak have this truncation point at about 7 <1>, but in the deepest portion of the Norwegian Trench it is at 8<1>.and inan area of higher current velocityto the south at 6<I>(Fig. 3).Normally the floc subpopu lation makes up 75-85%of the total sediment, and thetruncat ion-point shift from 6<I>to 8<I>iscoupled with an overall fining ofthe single-particle subpopulat i-

on.The com montruncat ion point at 7-8^<I> for predomi- nantlyfloc sedi mentation is more fine grained than the

truncation point observations made in environments

wit h greate r suspended -sediment concentrations (Amazon subaqueos delta:Nittrouer et al.1985,Kuehl et al.1988.SanFrancisco Bay:Kranck & Milligan 1992,and Quaternary glaciomarine settings: Stevens 1991). The truncation-point position is most likely related to both the suspended sediment concent rat io n and the energ y level of each particular environment,cont rolling thesize of floes and single part iclesthat can be kept insuspensi- onand transported.

(8)

42 AivoLepland& RodneyL. Stevens

Largerfloes tend to have largerind ivid ual components (Kranck 1975) and, in spite of decreasing density with increasing floc size (Dyer 1989), the settling velocity increasestow ardscoarsefloc sizes.Theassociati on between flocsize,component size and settling velocity may lead to flocsorti ng and a related change in the sizedistri- butionif timeand consistent transport condit ion spermit. Theflocsorting is also dependent on floc stabilitysince largerfloes aresensitiveto breakageand would not surv i- ve thetransport in relatively turbulent environments.The sensitivit y for breakage applies main ly for very large aggregates,such as macroflocs or 'marine snow',as discussed below,but in the very-fine-g rained sedimentsof northern Skagerrakthe changes in floc size occurmost probably withi n the microfloc type and these floesare very stable (Eisma1986,Eisma &Kalf 1987b). Assuming relatively consistent floc stability,the sorting of increasingly finer fractionsandthe truncation-pointshift observed fordifferent fracti ons(Fig. 5)is relatedtothe selective depositionof coarse floescoupled with a fining ofthe sing le-parti cle subpopulation .

Coarse-silt and sandsedim enta tion

The fine grainsize and unsorted character of the top 50 cm of sediments in the Norwegian Trench and adjacent areas support earlier conclusionsregarding the predomi- nanceof suspensionsedimentationand the limited influence of bottomtransportonsediments in thisdeep part of the Skage rrak(van Weerin g 1981,van Weering et al.

1993).How ever,these sediment s alwayscontainatleast 1-2%of very coarsesilt and fine sand,the size fraction that in the rest of Skagerrakand in most other environments is interpreted to be primarily transported and deposited by tractionprocesses.

There are four mainoptions for interpretingthevery- coarse-silt and sand sedimentol ogy in the Norweg ian Trench.Thefirsthypothesis isthe alternationofsuspensi- on and traction depositional mechani sms. Against the background of suspension deposit io n, infrequent and short episodes of signifi cant bottom-current activities might explain the limited amounts of coarse silt and sand.This alternative isconsistent wit h oceanog raph ic docu mentation ofthetem poraland spatialvariabilityof the bott omcurren ts (Sva nsso n 197 5,Rod he 198 7).It ha s also been shown that relatively str on g bottom-current velocities (15-20 cm/s) occasion ally occur even in the deeper port ionsof the Norweg ian Trench (Rodhe1987), andthesecurrents couldperhaps beresponsibl efor traction transport of coarse silt and sand. Neverthe- less,the loose characte rofthe surface sediments inthe Norweg ianTrench doesnot support a frequentbot t om- currentactivity that wouldbe able to transport coarse silt and finesand.If coarsesilt and fine sand have beenpri- marily transported bytraction processes,selective coarse-particle depositi on and erosional effects would be expected to be region ally significant and occasionall y recor ded in the sedi ment colum n despite the signal

GU-BULL 430,1996

modificationdue to bioturbation.

The second alte rnat ive involves anothe rtype of bottom current that couldmobilise verycoarse silt and fine sand for down-slope transport,that is,turbidity-current flow. This process has not been considered in earlier Skagerrak studies, presumably because of the lack of obvious sedimentarystructures. Investigati ons of mudd y, low-density tur biditeson cont inentalmarginsand basin slope s have demons tra tedthat turbid itesdo not always need to be graded;nor do they distinctl y deviate from the background sedimentation of simila rly fine-grained sediments(St anley1985,McCave&Jones 1988,Jones et al.1992).Blanpied&Stanley(1981)have show n thatuni- form turbiditicmud hasa sig nificant ly low er conte nt of forams and other biogenic debris than in suspension- deposited hemipe lagic mud. These uniform turbidites, whichnormally containless than 0.5%sand,haveapre- dominance of terrigenou s component s (Blanpied &

Stanley1981, Joneset al. 1992),comparabletosediments from the Norw egian Trench.Although gravityflow smay occurin thissetting,forinstanceinconnection with high sedimentation rates on the Danish slope of the Norweg ianTrench,no directevidence has yetbeendocu- mented.

The third hypothesis isthat all grain sizes,inclu din g verycoarsesilt and sand,aretransported insuspension.

Since thecurrentvelocities in northernSkagerrakarenot stron g enough tokeep coarse siltandsandin suspension asindividual particles,these grainsmight be incorporated within large but low-density aggregates, termed macroflocs when predo m inant ly inorganic particles are invo lved (Eisma 1986,Eisma&Kalf 1987a,b) or 'marine snow'wit h mainl yorganicparticles(cf.Alld redge&Silver 1988). Depending largely on the organic cont ent, the effective density of the floes wit h incorporated coarse particles can be significantlyreduced,enhancingthepos- sib le suspension residencetimeand allo w ing transport over long distances into calm-water environments.

St udies offloc densitiesand sizeshave shown a general tendency for decreasingdensitywit hincreasing floc size (Dyer 1989, Kranck & Milligan 1992, Lick et al. 1992).

Recent in situ measurementshavedemonstra ted that in add ition to organic matterand clay-sized minerals,floc- culat ion may involve grain sizes up to coarse sand (Kranck&Milligan 1992,Kranck 1993).

However, the macroflocs or 'marine snow' that are capable ofcarryingcoarse particl esare sensitiveto shear and are stable only in relatively calm waters(velocities lessthan 15 cm/s;Eisma1986).VanWeering et al.(1993) have shown the extensive development of bottom nepheloid layersintheSkagerrak,andit is believed thatthis near-bo ttom layer is the main medium of suspended sedimenttransport.Because ofshear stressesin the bottom neph eloid layerthetransport of very large aggregates('marinesnow')could belimited,but the more stable macrofloc s may be largely preserved and transpo rted int o theNorw egianTrench. If'marinesnow'is involvedin coarse-part icle transport,itisprobably formed andtrans-

(9)

NGU-BULL430,1996 AivoLepland &RodneyL.Stevens 43

ported by hemipelagic processes, presumably incorporating sediment that has settled down from turbulent transport higher up in the water column.Favourable con- ditions for incorporating coarse particles during 'marine snow' formation may be rare,but this mechanism is still viable consideringthe low coarse-grain percentages. The involvement of hemipelagic sediment transport processes in the deepest area is indicated by the composition of the sand fraction.Consistent with the diminishing impor- tance of sediment supplied by near-bottom suspension transport,the sand fraction of this sediment has a predominance of forams and other biogenic debris of pelagic origin (50-75%,cf.Blanpied & Stanley 1981).

The forth alternative hypothesis also involves flotation of coarse particles, in this case within the poorly consoli- dated and relatively mobile,surface muds.In areas supplied with coarse silt and sand,the formation of a mobile surface mud during periods of quiescencecould incorpo- rate these coarsercomponents due to mixing by bioturbation or occasional resuspension.Subsequentresuspen- sion and movement of these mobilesediment s, including the relatively light mud aggregateswith coarse grains, could help explain the occurrence of coarse-grained components in deep and calm-water settings.Although speculative, this mechanism is supported by the sediment character and the known variabilityof near-bottom turbulence.

Thus, although the fine-grained, poorly sorted sediments in the Norwegian Trench are generally typical for marine suspension deposition, specific considerations seem necessary to explain the coarsest fractions.

Although obscured by bioturbation or by the predominantly muddy texture, low -fr equency turbidity currents and occasional bottom-current traction transport cannot be excluded within this setting.On the other hand, coarse-particle buoyancy is likely through macrofloc or'marine snow'formation, allowing transportin the mobile surface layer, the near-bottom nepheloid layer and the upper watercolumn.

se-silt size.As discussed above, the source of buoyancy, possibly provided by low-density macroflocs or 'marine snow', is necessary to keep heavy minerals of coarse-silt size in suspension for long distance transport. If the heavy-mineral transport is largely controlled by floes, the hydraulic sorting of heavy minerals cannot be well deve- loped, as is indicated in this study by their very low correlation with grain size (Table 2). Still, the heavy-mineral distributions should be source related and informative of sediment transport pathways.

The interpretation of the heavy-mineralcontour maps (Fig.4a-h)with respect to sediment transport pathways is complicated because of several features: 1) variable mineral contents over short distances,2) relatively non- selective transport, probably in flocculated suspensions, and 3) very low heavy-mineral proportions of the total sed iment (normally less than 0.1 %). The variability of the mineral contents in our data (below) does not generally allow identification of distinct, adjacent populations at the 95% confidence level (Folk 1974, Bridgland, 1986).

This is logicalsince the regional sediment sources are not strongly contrasting, and provenance inter-pretations have only been possible for large,generalised provinces (Bengtsson & Stevens this volume).

On the other hand,despite the low heavy-mineral concentrations and their variability, certain geographic trends were noted above. Most important,since the concentrations of severalheavy-mineral species preferentially increase towards the north, south and southeast away from the Norwegian Trench,the supply of these minerals is interpreted to be by sediment transport in the opposite directions (Fig.6).lnspite of the relatively non-conclusive evidence behind the transport interpretations of individual minerals, the occurrence of similar directional indi- cations from several minerals strengthens the probability of these conclusions.In addition, the predominant pathways areconsistent with the transport directionssepara- tely interpreted from grain-size trends (Stevens et al.

57"40' '---,-- - --,-- - -,--- - -,--- - -'T-...L-- ---,--- '

Fig.6.TransportdirectionstowardtheNorwegianTrenchcorresponding to specificheavy-mineralcomponents.

.'q}.~~"

»:

.,~ . ~hornbl ende

. c=;>garnet

c=f>hematite . . . magnetite

~.pyroxene ,.... ..epidote

,.... dolomite

11·00' E 1rroo' 1cr30'

9"30' 9"00' 8"30' 58"00'

58"20' 58°40' N 59"00'

Heavy-mineral compositions have been widely used in provenance studies (van Andel & Poole 1960, Morton 1985, Singh et al.1993, Bengtsson & Stevens this volume). Also,the progressive modification of the heavy-mineral suite along the sediment transport pathway,where heavier minerals are preferentially deposited and lighter ones are carried further, is a well documented process (Komar & Wang 1984,Trask & Hand 1985, Li & Komar 1992). Most studies of provenance or selective heavy- mineral sorting have dealt with sandy environments where sediments were transported predominantly by traction processes. The structureless, fine-grained sedi-

mentsof northern Skagerrakdo notin d ic a t e bottom trac-

tion transport.On the other hand,the documented current velocities(Rodhe 1987) do not support the suspension transport of heavy mineralsas single particlesof coar- Mineralogical interpretations

(10)

44 Aivo Lepland&RodneyL.Stevens NGU-BULL 430.1996

NW

===:>

hemipelagic

near-bottom suspension

traction ^S ^E

<=

.-:

-: '.~

^'<,

~A> "~-: -: . '.

•- : - :'.

~ '~

^{. :}

~~;/~ ~P~!::: .: . .

' I '.' %2

20~ t- 1: .'. 1

10 ~ /. 05

5 Q1

%~ ^{. .}^{. .:}^.^.^.^.^{. .} ~~tJ ³ ⁴ ^{5 :} ^{7 8 9}

05 ..... 5 traction.

013 45 6 78 9

20~ %~

near- bottom suspe nsion

4> 10 0.5

near -bottomsusp e nsion %5~ ^{0 1}

34 5 6789

05 4>

o1 ne a r-bottomsuspension 3 4 56 78 9

4·

near-bottom susp ensio n.

hemipelagic

Fig.7.Transportand deposi tion almecha- nismsas indicated bythe sediment par ticle- sizedistribution, by the concentration sof water-column suspended matter(Eisma &

Kalf 1987b),and by transmission measure- ments(vanWeering et al.1993).The shallo w shelfhas traction-deposited sand and coar- se silt comb in ed with near-bottom suspensi- ondeposition(bo ttomnepheloid)ofunsor- ted clay and fine-siltflocs.Near-bottomsus- pension mechan ismspredom inat e on the trench slopes.Inthedeepestportionofthe NorwegianTrench thecont ribution tothe minerogenic sediment from hemipelag ic depositionbecomesmorepronounc ed,as is indicated bytherelativeabundanceofcoar- se-silt of hem ipelagic origin (biogenic debris).Onthegrain-sizedistribut ion dia- gramsthethin linesarefromtheadjacent positionsforreference with respecttodown- slopechanges.

1996) and with trend s in mineral-magneti c parameters (unpublished data).

Particular caution is appropriate wit h regard to the possibly dependent relationships between concentrati- onsof different translucent heavy-mineral species since they aresummed to 100%,but thereare nostr ong correlations,positive or negative,observed betwe en minerals (Table 2).Neverth eless,significant changes inthe major mineralvarietieswould beexpectedto change percentagesof ot her components,for exam ple,the pro nounced dow n-slope decreaseof hornblende onthesouthwestern side of thetrench coincideswith an opposite trend of epidote (Fig.4).The similarlackof correlatio n between the mineralspecies(orgrou ps)and grainsizealso sugg ests that hydraulic sorting hasnot controlled the coarse-silt mineraldistribu ti onsin the fine-grained depo sits.This is consistent withthe int erpr etedim port ance of suspension transport and the influ ence of flocculation. In contrast, stron gmineralogic sortingeffects arecomm on lyassocia- ted wit h traction transport of sandy sedim ent s(Trask&

Hand 1985,Li&Komar 1992), asis interprete d inthe adj a- cent,shallower areas to thesouth(Bengt sson&Stevens this volume).

The dist ributio n of biotite in the coarse-silt fracti on (hyd raulically equivalentto very-fine-s ilt quartz,Doyleet al. 1983) does,however,indicatethat suspension trans- portcan provide consistentmineralogictrends related to source character and sorting during transport if these minerals are not exclusivelyincluded infloes,Increasing

micaconte nt s suggest that sing leparticlesedimentation becomes significantin the central Norweg ianTrench due to the diminished supply by nepheloid suspensions.

Because of its flatcrysta l form,micais hydraulicallymore mobile thanot herminerals of the same size, and itiscon- sideredto bea sensitive hydraulic indicator ofsediment depo sit ion and transport cond it ion s(Doyle et al. 1983, Komaret al.1984).The relativelyhig hercurrent velocities limit mica deposit ion on the slope s of the Norwegi an Trench and inshallow areas,resulting in its winnowing and deposition in calmerand deep erareas.

A minimal disturbance of the mineral dispersal pat- terns is a prerequisite if source interpretations are to be made upon relatively weak patternsof low-percentage com ponent s,asthose discussedabove.We expect that suspension processes themse lvesare variablewith time and location, so that much of the irregularity of the heavy-mineralcontents isinherent tothis environment.

The generalsediment sources indicated bythe transport direct ionsare logicalwith respectto the principle oceanographic currents(cf.Figs.1&6).However, in our inter- pretationof transport pathwaysthe dow n-slope directions are more pronouncedand areimportant as areflecti - on of the netsediment accumulation .

Conclusio ns

Sedimen tgrain-size trendsin nort hern Skagerrak arecon-

Suspension sedimentation of coarse silt and sand in northern Skagerrak: textural and mineralogical trends