aun,nb as!l!gn lsa,u !S a1 anb s!pue) lanuue apA3 ... xnlj $ 8 ~ pageu!gsa '~se.quo3 Ag *uo!g3npsad woge!p ienuue 8upnp pasn !S akgg jo %gz pue Allenuue ...
Can. J. Fish. Aquat. Sci. Downloaded from www.nrcresearchpress.com by UNIVERSITY OF MICHIGAN on 12/19/13 For personal use only.
vsn '909zE Td "PPOId J" "") SsTL jema3u! aq3 Jo qanm '(5361 aqsl3q3g f ~ ~'Ie6 331 ~ a y n d s!seq ) 6 . pue sapaqsad $0 v.mrngr;edaa :ssaJpp$i jnsssq,., @TBUUEBHBuo p a l ~ m skns!qmpo~d ~ ~ mspnp u!qsns 03 pzu~nb=%~ TS ~ ~ 6'a~nqraxmb$. oAso~~aoq~q JO ssew aq) jo ~ J O W10 % ~ 'rn8y.pg~~ 6 aye? u~ oAlpnurm~ pala yamasaa Iqraarnuosrlaq s a vadg ~ a~q ~jo 19 -ON nopnqpjnoa - ~ a n a 3 as!~pa3wps~dr ~ jle a js %66 3~q3papmpsa ( ~ ~ 6 jsq 1) p u no~spuagl[ ~ y 3 ~ a s a as s v p~a J g ay3 jo ~ g 'ON v uorlnqyagaao3, -ph6ss6aaasu~ *ucBpaanpo~dlnapuadap nlgluaarn 03 aapo u! Agy~nnn~ p a ~ ~ l a n a 8aqa Jsnw ~ (!%aJO zsrj~sqna8onq) s r n s ~ u ~ 8 ~ 0- 3 ~ a q" ~( 1 ~ law=%oJS 6~ pm aysgays~)A 3 1 l l q e g ~(as) ~ ~ ~e q r s snoa3gls JO SIS=%~ 01u1 p a ~ e ~ o & 08%3 ~aq) ~ JO gscoau 3~93pajm3s p a ~ i o s s pAq pq!rn!g Lyg~uos~as &on sl ra0~13npmd816aopz!p 3gau -wgda pqs 1ua3xa 9x.p 03 q.g~n0.18mojelp paJgjnmys s ~ 3upeol q -nomap amy esyIs jo %aaqshl~aywaqaoa8orqu s s s ! p n ~ ~ 'u12%~qa~~g a v q ng ' ( ~ 8 4 1'le 'UOl]EJ=%gr§UO3 (d) smoqdsoqd pas~a~ana ]a mmlg,) sal3ads p8p naaMJaquo~ga~adms3 js am03 jneaeh noa~~sodwo3 A1~unmsnnsa aaogqraejdo1Aqd no -3no a y a3uanaul ~ m3 L J ~ ~ E ~ ~gnapgnu E A E ul sasuamjjl ssnan~nas p y pug pnz d qjoq JO sag3As j-e~lm=%q3o=%%oyq %JQJ
-ssgwoge!p ap uo!y~nps~d el a.~!essa~gu !S np ?g!j!q!uods!p el ~ueqw&s!l ua uo~3ueldogAqd ap sa~adsasap a.quuos!es uo!ssamns "1 ap jel(ns$J al .ins Jangju! ga asnanbe assew el suep !S np Ja!uuos!es guswass!d~neddeun q aauaw gua~nadd np ga !S np anb!u!y~o$3o!qa8el~A~a.1 ap xneg sag la luawauuo!s!~o~dde,j9 x n v ~sal aAlua sa3uaaqjjsp sai '!sup+j .s!oj aun,nb as!l!gn lsa,u !S a1 anb s!pue) lanuue apA3 np slnm ne s!oj sasnaqwou ap ?s!qy-~ ~ n a d a] '(,-e.$'0) aaual snld !S ssp us!ge-lgu$3?sel y jaodde~aed nea,p auuolo~el suep guauklap!de.i na!l e r] np auJagu! uo!geJau -?Baa el an$ auusp lueg3 'allanuue a ~ ! e y a du o ! l ~ n p o ~ eldap sano3 we ?s!l!gra d np % L>aa nea,( suep aguasadd np jigeuugsa xnlj al '~iguo3Jed *saawoge!pap allanwe IP~OJd ap alpnuue ass& el ap % LIP anb q u m q au us!pnps~del ap smss ne ?s!1!gn !S np % g e la nea,l suep aguasa~da ~ l ! s ap aisanuue alejog assew eg ap % ~z a!uJnoj ~ n a d !S np J ! J & L B B xn ! ~~S a1 j ~ .~usw!p?sap sa~~esu! saJgole3ap us!geqnDu!,l y a q ~ sluaw 8 !pas sal suep (sad) alqnlos jspeiy a~sqdsoqdnp la agnoss!p aq!s el ap xnlj sal ?!g!~uenb3uop e u g . u e S ! q x ~3el a] suep sJuaw -!pas sap u o ~ ~ e ~ u ~ 2 ? ~ p p aJau!waag?p 3 ~ ~ o dapuuye ! ~ an$!w!q3o?8o!q ~ apA3 wp saluesadu~o3saline xne ~a~edwo sals .mod sa~~sn3el sluaw!pas sal suep Id) a~oqdssqdnp la a~&s!l?s q apsjgew!psa xnlj sa~?u!kk~.ia$?pe uo
(!s)
.uos]3npoad wo$e!pdo) Adessa~auAl!l!qi?l!eAF? !S 8u!$!w!l Aq uo!ssa33ns sa!3ds uolyue1do)Aqd pusseas jo awo3Ino aqg axanlju! pue ssew JaJeiMaq] u! uo!galdsp !S leusseas 01 peal ue3 d pue !s jo sale8 8u!l~A3a~ le3!waqma8o!q aqg u! pue sale~Apddns aqg u&s! sa~uadajj!p'sn y l ' a ~ u sAluo pasn s! s~ seamqm apA3 lenuue ue 8u!anp saw!g Aueu pasn aq ue3 d '(L-~A-e)*o) !S j s usge~aua8a~ Jawols q3nw 443 q l p ~pa~ed - W Q ~ 'uunl03 J ~ ~ Vaqg M u! Alp!de~s ~ n m s d go us!ge~aua2s~ ~eu~alu! aqg asnemg .uo!g~npo~d h e w ! ~ dpnuue aog pan!p!ln d aqg jo % L >PUP A l p ~ u u eJagem aqg u! $ le$slgo sseu aql js % 1.p Aluo Alddns ues sluawpas aye1 w o ~ j xnlj $ 8 pageu!gsa ~ '~se.quo3Ag *uo!g3npsad woge!p ienuue 8upnp pasn !S akgg jo %gz pue Allenuue JaleM s y u! ~ e3!1?sjo s e w legog aql jo % L z Alddns ue3 sguawpas UQJJ xnjj !S palew !gsg -saJoD ~uaws pas ~ e q ul!o uo!yqn3u! aql Aq paJnseaLu aJaM s)uaw!pxi UOJ) saxnlj ($as) sn~oqdsoqdala!lxaJ alqnlss pue (!s) w ~ ! palzlossa s -ueS! -q3!,q ay e7 u uo!~e~auaSa~ guaw!pas j s a3uel~odw!ayl a]e2?psa~u! 03 spA3 le3!waq3oa80!q aq) U! sluausdw o ~ Jaqgs 4gp~p a ~ e d u uayl o ~ pue pau!wdalap aaaM (d) sn~oqdssqdpue eml!s jo saxnu luaw!pas ap!~~-aye1 pageu!ls3 a~ go * ~ E - J :SP~-82s ~ 'genby ~ o y~s ! j me3 'ue8sq3!,q aqel u?8 u ! l ~ A ~(euwu? a~uelssdw!:s)uaw!pss w o ~xn(g j sn~oydsoqdpue exl!! '$861 'ayqakg~s'7 ' 2 pue 'Aal8!n(-~ 'V 'W"1 'a ' A 4 1 ~ ~ 3
:s~uaw!pa~ uorj xn j snroydsoyd pule le31
Can. J. Fish. Aquat. Sci. Downloaded from www.nrcresearchpress.com by UNIVERSITY OF MICHIGAN on 12/19/13 For personal use only.
using a Mark HV Soutar box corer aboxd the RIV Laurentian. Only cores with no visual evidence of disturbance were selected %OH~ k d y . Estimates of Si and SRP flux across the sediment-water interface were made by incubating intact cores and measuring nutrient concentrations in overlying waters over time. The rate of change of solute concentration in overlying water through time was taken as net release from sediments. Cores for measurement of Si flux were incubated over 24-36 h during the linear release phase (Andrews a d Hmgrave 1984). Cores used to study SRP flux were incubated over much longer time periods (30-48 d) in order to obtain measurable SRB fluxes (Quigley and Robbins 1986). All cores were incubated in darkness with mild aeration of overlying water to prevent development of concentration gradients. Cores were incubated at 5°C for Si flux md at 6°C for SRP flux. These incubation temperatures were at or near in situ temperatures.
ResuIts and Discussion Sediment Fluxes
FIG. 1. Stations sampled in Lake Michigan. Stations U1, GB5, and C37 me in Green Bay and stations GTB2, GTB3,and DW are in Grand Traverse Bay.
recycling of BSi is known to result from sediment regeneration (D'Elia et aB. 1983; Calvert 1983). However, in Lake Michigan the proportion of internal recycling that occurs in sediments versus the water column has not been determined. The objective of the present study was to evaluate the importance of Si and soluble reactive phosphorus (SWP) fluxes from sediments for internal cycling of Si and P in Lake Michigan. Areal Si and SWP fluxes were calculated to determine (1) the proportion sf Si and P that is regenerated mmally from sediments a d (2) the importance of sediment flux as a source of Si and P utilized annually for primary production. Differences in the biogeochemical cycling of Si and B anad the influence of internal recycling of both nutrients on primary production are discussed.
Materials and Methods Intact sediment cores were collected from sites in Green Bay a d in the open waters of Lake Michigan (Fig. 1). Stations were selected to provide a range of sediment type (Chill 1981) and to sample different areas within the lake to estimate spatial VXiability of sediment fluxes. Sediment cores were collected using a Wildco gravity corer from the RIV Myss's m d the R/V Laurentiwn in 1983 and 1984, respectively. Cores from stations SJKGH, CH2, and BH3 in 1983 and 1984 were collected using a Kahl Scientific box corer aboard the RlV Shenenhsn. In 1985, all cores were collected Can. 9. Fish. Aquad. Sci.,
$101.
45, 1888
Si fluxes across the sediment-water interface in cores from Lake Michigan ranged h m 2.3 to 10.3 mg S i O , = ~ m - ~ y r - ' (Table 1). The r a g e in Si flux was similar to Si fluxes reported from previous studies on the Great Lakes (see Quigley and Wobbins 1984). Because the source of Si flux from sediments is fmm dissolution sf diatom silica, Si flux should be related to sediment concentrations of BSi. Si flux generally increased as a function of BSi concentration in the 8- to 1-cm sediment interval (Fig. 2) and leveled off at higher BSi concentrations as Si flux from sediments becomes limited by diffusion (Conley 1987). A significant fit of a saturation cuwe to the data from depositional basins was obtained (r2 = 0.96, B < 0.001, n = 23). Si fluxes from the newshore station SJKGH were not included in the calculation of the saturation curve because Si fluxes from station SJKGH largely result from the dissolution of a thin layer of BSi on a compacted sediment surface (Conley 1987). Sediment fluxes of S W ranged 2 orders of magnitude among all stations (Table 1) and up to a factor of 3 between years at station K34. Sediments incubated from oligotrophic lakes at low temperatures under oxic conditions often have undetectable rates of SWP release, and occasionally sediments show a net sorption of SWP resulting in negative fluxes (Kamp-Nielson 1974; Holdren and Amstrong 1980). Measured SRP fluxes reported here are typical for an oligotrophic, aerobic, cold environment (Niimberg et al. 1986; Quigley and Rsbbins 1986). Si:SRP molar ratios calculated from sediment Si and SWP fluxes were considerably higher than the idealized Redfield ratio for seston of 16:1 (Wedfield et al. 1963) and ranged from 460: 1 to 6600:1 for ratios that could be calculated (Table 1). Because Si fluxes reported here were in the same range as reported elsewhere, the Si:P ratios were higher due to the relatively low SRP fluxes. Internal Recycling The importance of BSi regeneration from sediments can be estimated by computing a lake-wide estimate of Si flux. The lake-wide Si flux from sediments was estimated using the saturation relationship (Fig. 2) to predict sediment Si flux from mean suficial BSi concentrations from different basins (Conley et al. 1986). The predicted Si flux from a particular basin was
TABLE1. Mean Si and SRP sediment fluxes, stand& deviations, and the Si:P m l a f ratio of fluxes for cores. The number of cores used in the measurement of Si md SRP flux is given in parentheses.
Can. J. Fish. Aquat. Sci. Downloaded from www.nrcresearchpress.com by UNIVERSITY OF MICHIGAN on 12/19/13 For personal use only.
Station
Date
Si flux (mgSi02-cm-2eyr-I)
SRP flux (kg P-cm-'.yr-
l)
Molar Si:P
15 July 83 19 July 84 12 Aug. 85 16 July 83 19 July 84 13 Aug. 85 16 July 83 18 July 83 8 Oct- 83 25 Apr. 84 8 Oct. $3 25 Apr. 84 8 8ct. 83 25 Apr. $4 23 Aug. 83 4 May 84 10 Aug. 85 4 May $4 20 July 84 11 Aug. $5 20 July $4 1I Nsv. 84 11 Nsv. 84
o
5o
too
200
250
FIG.2. Measured Si flux vs. BSi concentration in the upper 0- to 1cm interval in c ~ r e s .The squares denote unpublished data (C. E. Schelske, unpubl. data). The line is a saturation cuwe describing the relationship between suficial BSi concentrations m d Si flux (Km= 123, Si flux,, = 16.8; r2 = 0.96, n = 26)
then multiplied by the basin area (Table 2). Depositional basins released l 1.6 x 108kg SiO,, trmsitiond basins released 1.13 x 108 kg SiO,, and nondepositional basins released 2.18 x
108 kg SiO, (Table 2). The resulting annual basin-weighted Bake-wide Si flux is 14.9 x lo8kg SiO,. A basin-weighted Si flux for Lake Michigan by Schelske (1985), estimated from mean Si fluxes from Great Lakes sediments, ranged from 9 x 108to 13 x 108kg SiO,. yr- l , or 1340%lower flux. ScheBske's estimated range may be low because it was based on fluxes determined from psrewatea Si concentration gradients, a method that underestimates Si fluxes in compkssn with intact core flux estimates (Conley 1987). The impact of sediment regeneration on internal system dynamics can be determined by comparing it with other cornpsnents in the biogeschemical cycle of Si. The total mass sf Si in the water c o l u m during the winter maximum is 72 x 108 kg SiO, (Schelske 1985). Therefore, at steady state, on average, 21 % s f Si in the water is internally regenerated from the sediments annually (Fig. 3). By csmpkson, tributary inputs sf Si are estimated t s be 2.0 X 10-g SiO,-yr-I (Schelske 19851, only 13% of the sediment flux. Tributary inputs annually supply only 2.8% s f the total Si in the water during the winter maximum. In contrast with SSn, SRP fluxes from sediments showed no relationship to suficial sediment totd P concentrations, to NaOH-extractable P concentrations in suficid sediments, to depth, or to bulk sediment composition as determined by Cahill (198 1)- Therefore, for the purpose of calculating m areal SRP flux, it was assumed that SRP fluxes occur at a constant rate over the lake bottom. The mean SRP flux determined from cores, 1.82 -b 1.29 ~g Pocm yr - , was multiplied by the area of the lake bottom. The resulting annual lake-wide estimate of SRP flux from sediments is 1.1 k 0.75 x %O%gPeyr- I . The amount sf SRP regenerated from the sediments lake-wide may Can. J. Fish. Aqua?. Ssi.. kd. 45. 6988
TABLE2. Prediction of areal Si flux for depositional basins, transitional areas, and nsndepssitisnal areas in Lake Michigan.
Area Mean swficial BSi ( k n ~ ~ ) ~ concentration (rng~g-')~
Can. J. Fish. Aquat. Sci. Downloaded from www.nrcresearchpress.com by UNIVERSITY OF MICHIGAN on 12/19/13 For personal use only.
Basin Depositional basins Green Bay Grand Haven Milwaukee Northern Southern
1 628 2 030 1 450 10 300 6 9043
Transitional areas
Predicted flux Basin-wide flux ( r n g * ~ r n - ~ . y r - ~( )X~ 108 kg SiO,)*
134 39 55 94
28
8 700
aRomChill (1981). bFrornConley et al. (1986). cDetemined from a saturation curve describing Si flux as a function of BSi concentration in the 8- to I-ern sediment interval (Si flux,, = 16.8, KM = 123; a2 = 0.96, n = 23). dBetemined by multiplying predicted basin flux times depositional basin area.
External Load
Sediment
( y i ' )Flux
(yil)
Annual
Total
UtiEizstion
Mass
FIG. 3. CompiaPison of Si and P dynamics in Lake Michigan.
be higher than reported here because Green Bay has periods of reduced oxygen concentrations (Conley 1983), which may lead to increased SWP sediment fluxes (Bostr6m et d. 1982). However, only a small area in Green Bay undergoes oxygen stress during the summer stratified period (