.. ICES C.M. 1986/C:8 Ref. Biological Oceanography Cttee., Sess. Q ANNUAL CYCLES AND PHENOMENA ON OTHER TIME SCALES IN TEMPERATURE, SALINITY, NUTRIENTS AND PHYTOPLANKTON AT HELGOLANJ REEDE 1962 - 1984 I' .. by • G. Ra1Qch ; J. Berg, Institut für Meereskunde der Universität Hamburg, Troplowitzstr. 7, 2000 Ha~burg 54 and • E. Hagmei er, Biologische Anstalt Helgoland, Meeresstation Helgoland, 2192 Helgoland ABSTRACT The long-term time series of' temperature, salinity, plant nutrients and phytoplankton biomass at Helgoland Reede in the German Bight are investigated with respect to prominent time scales by spectral analysis of variance. This analysis was based on 23 years of data from 1962 to 1984. • , I The data are presented in the form of annual cycles by projecting the data into one year, to demonstrate the dominant features of the annual dynamics. It turns out that the cycles of diatoms and silicate are inversely related to each other, more than are diatoms and phosphate. The annual cycles of the nutrients nitrate and nitrite are inversely related to the flagellates# biomass. The Elbe river outflow, which is directly related to the nitrogen compounds, seems to prepare the situation for relatively hig1 flagellates# biomass development during summer after diatoms, decrease due to the depleted silicate. Thus silicate seems to limit the diatoms# development in early sum~er at Helgoland, but not phosphate or nitrate, and later on flagellates take over the leading role. Changing to the frequency domain, it turns out that the outstanding frequency in many of the spectra is the annual frequency. There most of the variance is found for the parameters sea surface temperature, phosphate, and nitrate. Salinity, nitrite, ammonia, silicate; the phytoplankton entities, and Elbe river discharge exhibit the bulk of their total variance in the frequency band 1 <f<10 cycles/year. For phosphate, nitrate, and nitrite the spectrum shows constant (or even decreasing) variance for frequencies smaller than 1/year, while the variance decreases, similar to temperatur~, for frequencies larger' than 1/year. Most of the parameters depend on frequency with a decay rate of about -5/3 for frequencies > 10 cycles/year. Nitrate and . silicate exhibit a stronger decay with a rate of about -2.12, Elbe river discharge with a rate of -3.43. In the frequency band 1 ,1 "f < 10 cycles/year most decay rates. are weaker than -5/3 , (except for phosphate and nitrate), between -0.5 (for diatoms) .and-1.5. , ; 1. Introduct ion · The Helgoland Reede time series are amongst the longest and most · complex chemical and biological data sets on the European ,continental shelf. The time series were started in 1962 by the Biologische Anstalt Helgoland and include aperiod of rising : human impact on North Sea coastal waters. · The Helgoland data set provides a unique possibility to ,investi· gate, on which time scales the essential processes within the 'coastal marine plankton system take place. Due to.the lack of · sufficiently long time series, only very few investigations exist : about time scales in chemical and biological data sets andif, they usually refer only to very few state variables (e.g. TONT & : PLATT, 1978). · As is known, in an isotropie turbulence field the spectrum : decreases exponentially with wave number by a decay rate of -5/3. The first one applying. this theoretical result to the interpretation of chlorophyll variance spectra was PLATT (1972), · for time scales of hours. Since then many authors obtained wave : number. spectra for, chlorophyll measurements on the basis of : spatial sections. Statements about the spectral distribution of : spatial variance have sometimes been transferred to time by assuming the validity of the "frozen field " hypothesis (OKUBO, 1980; PLATT &DENMANN, 1975). Results for time seal es larger than : 1 day are hardly given (e.g; TONT &PLATT, 1978). I 1 The Helgoland.data set allows for direct investigation of time • scales in the range of, say, 5 days to 23 years. Long-term trends : occurring on time scales from 2 to 23 years have been already : presented and discussed by BERG &RADACH (1985), RADACH & BERG (1986), and GILLBRICHT (1986). Here we discuss mainly events and variances on the various shorter time scales in more detail. , It is undecided whether. an exponential decrease of variance with : a decay.ra~eof -5/3 should be expected for the frequency ranges and location investigated here~ In any case, it is striking that · manyof the parameters exhibit a decay of variance with a rate of , about -5/3 for frequencies larger than 10 cycles/year. 2. Material and methods , Surface samples between Helgoland main island and the "Düne" (54 ; 11.3-N, ,7 54.0-E)) have been collected since 1962, first on : three days, later on five days per week. After, temperature measuiement, samples were analysed for salinity and the nutrients ,phosphate, nitrate, nitrite, ammonium, and silicate; Phytoplank0 0 , ton stock was estimated by the technique,of,UTERMöHL (1958).Con- : versions to carbon follow STRATHMANN (1967). Phytoplankton is subdivided into diatoms and the rest, maily consisting of flagellates. Details of the data of single years can be takenfrom the annual reports of the BIOLOGISCHE ANSTALT HELGOLAND (1962-1984); 2 • 3. Annual cycles Todemonstrate the importance of the annual period compared to smaller periods t all 23 years of data are plotted into one year (Figs. 1a-12a). In Figs. 1b to 12b the mean annual cycles are presented t together with a confidence band of a width of two standard deviations. Thus the mean value for. every day lies within this band with an error probabilityofmuch less than 0.1 %. In the following the term "var iability" always means absolute variabil ity. • Water temperatures and salinities off Helgoland are strongly influenced by wind t moving and mixing of water masses of. different origin t and, bY. river runoff, precipitation t .and heat flux (BECKER; 1981)~ In winter time; occasionally drift ice is transported from coastal areas as far as Helgoland. Surface temperatures (Fig.1) follow adefinite, annual cycle. Summer maxima are reaching 16 to 20 ·C. Variability in summer is lessthan in winter timet where minima are mostly higher or lower thän the calculated mean values~ Surface salinity (Fig.2) has also a typical mean course during a year t .but large deviations from mean values occur. In general, salinities in March and May are higher and in April lower than published for the period 1965-1975 (WEIGEL, 1978). Much of the variations may originate ih different meteorological forcing t causing the water with important freshwater ädmixtures to touch Helgoland or to flow northward along the Wadden Sea. To judge the situation at Helgoland it is important to consider the discharge of the Elbe river (Fig. 3), as an indicator describing the continental, run off. The annual cycle iS,dominated by isolated discharge everits~ The meari cycle represents only a paar impression ofthe dynami~ of discharge. It has its maximum in April, .its minimum in August. Variability has its maximum in January, its minimum in October. In summer strang discharge events occur. very rarely. It is reasonable tocompare the discharge of the Elbe river with surface salinity near Helgoland t and indeed, for time scales larger than two years a clear inverse relation can be established (BERG &RADACH, 1985). The amount of nutrients in solution is the result of supplies by rivers and remineralization;arid reduction by primary production and currents leaving the area~ The different nutrients exhibit different annual cycles. . The concentration of phosphate (Fig~ 4) is nearly co~stant from October, till the end of March. Its decline in April and May is reflecting the uptake by multiplying diatoms. Comparing phosphate consumption .with phytoplankton stock (Fi~.10) we see,hard]y.. anY relation (see also GILLBRICHT, 1986). TO,explain this we must be aware of the fact that production and not standing stock is the correct quantity to compare with phosphate decrease; Thus we can assume e.gj a high grazing pressure on the algae ! This state of low concentrations (0.25 m mol P/m 3 ) remains. unchanged until,July. "Then phosphate increases again., Variability is strongest during late summer and early Jall, while it is smallest during .the phase of depletion in May/June~ In times of rising phosphate contents in the water (beginning in July), 3 : excretion and decomposition of organic matter and reminerali: zation are providing more phosphate than the algae are able to , consume. : The annual cycle of,nitrate (Fig. 5) has its peak values in January/February and in May, with a relative minimum in March.' The high level of.about 25 units decreases from May,to July down · to less than 5 units. The concentration stays at this level until October. Then the concentration increases again until January to : the winter level. Variability decreases steadily from January to September (excluding April), and it increases again in the same : steady way to the winterly range ofvariabil ity~ The high values ,of nitrate in winter and spring seem tO,have a direct relation to : the Elbe river, discharge. : Similar restilts were obtained earlier , by LUCHT & GILLBRICHT (1978). · Nitrate distribution indicates more 1nfltience of river input than phosphate content. This can explain why the effect,of the phyto: plankton spring bloom,is registered, later than in the phosphate curve. Recovery of nitrate in solution ,is rather slow; partly : because excreted nitrogen-compounds .will be reused before : oxydation, partly because remineralization is passing more steps , and needs a longer time than that of phosphorus~ The annual cycle of nitrite (Fig. 6) has its highest values in December/January. From January,on nitrite decreases until March, . has a local maximum in April/May, and then decreases again 'strongly until June. It remains.on this level until September and · increases. lateron steadily to the level of January. Variability : is in, winter about 5 times as large as ,in, summer. Nitrite is :recorded in amounts one order of magnitude lower than nitrate. Nitrate and ,nitrite have 1ndeed different annual cycles: The ,phase of depletion , is for :nitrite about one month earlier than ,for nitrate. Thehigh ~alues of nitrite in winter and early spring coincidewith high values of the Elbe discharge. This may : indicate slow processes of oxydatiori to nitrate in cold water~ Ammonia (Fig. 7) is another transitional nitrogen compound; ,excreted by animals; taken' up by plants and oxydiied by bacteria. There is no definite annual cycle, deviations from a meari value :around ,7.5 m mol/m~ are rather large. A predominance of :degradation inthis time,of thc year should result in slightly 'higher values during fall~'The data seem to back this. ,Total" inorganic nutrient-N maps essentia.lly, the annual cycleof :nitrate, after adding the fairly constant background of ammonia 'of about 7~5 m mol N/m 3 (Fig;8)~ :For ,silicate (Fig. ,9) the annual cycie starts. 1nwinter and in ,earlY,spring (until April) with the highest concentrations. Silicate content drops, distinctly "dudng, April arid May; :corresponding to the spring diatom bloom. However, as with the ,uptake, of phosphate, a quantitative connectionto the standing 'stock,of phytoplankton cannot be.established. The silicate con:sumption, ,as that of phosphate. in spring~ is ten times more than :expeCted fro~ the amount o( plankton counted; ,Although sma 11 :flagellatesneeding phosphate can be destroyed by fixation or :concealed by detritus, it seems impossible to overlook a large 4 • number of silicate containing diatoms. Consequently consumption and replenishment of phytoplankton should be much quicker than expected. The low concentratioris are observed until September. Then, the concentration increases again to winter values., Obviously there is averystrong variability in winter (about 4 times that 'of summer). The high values in winter coincide with high amounts of discharge of the Elbe river. • ~ Phytoplankton stocks (Fig. 10), presented by their carbon content, appear tri dev~lop ih .lhe ~classical" form for our latitude with, two maxima in April/May and August, but this picture is really formed by the superposition of many short-termed bloom events. ,While diatoms,are dominating during the spring.maximum, in summer time large stocks of flagellates are of equal importance.The standing stock in winter is about one to twoorders of magnitudes smaller than the standing stock in spring and summer; Variability is - on a logarithmic scale,-fairly const~nt, of about orie order of magnitude (i.e. the factor of 10). Diatoms (Fig., 11) appear in several maxima between April. and September, with contributions of different species, . ~hose abundance is partly predictable by their demands for light and temperature; Variability is least in winter, (one ,order of magnitude) arid greatest ih summer (about two orders of magnitude). Flagellates (Fig. 12) occur with small, naked forms in iarge numbers but with low carbon contents all the year round. Additionally, numerous dinoflagellates can be observed in summertime •. From fall to spring a biomass. of20 - 30 mg C/m 3 ,is not exceeded. Carbon values during blooms of Ceratia reach 1000 mg C/m 3 ! To summarize, the annual cycles of diatoms and silicate seem to be inversely related. The same is true for diatoms arid phosphate. Nutrients-N of nitrate and nitrite are inversely related to flagellates- biomass, Elbe discharge andnitrate/ nitrite run parallelly. Therefore, the nutrient input of N-components by the river discharge reaching Helgoland prepares the basis for the Diatoms are not build-up of flagellates- biomass in summer. äbl~ to fullyuse the nitrogen load of the water. It is in June~ when blooms of flagellates first.occur, that nitrate and nitrite fall ,off tO.the lowest levels. ,At this time silicate; which is vital for diatoms but not for flagellates, has decreasedalready to very low values. Thus it seems that silicate represerits the limiting factorfor diatoms, but not phosphate.ornitrate. As the diatoms make only limited use of ,theplentiful phosphate and Nnutrients, flagellates take over th~ leading role and form the dominant amount of biomass; 4. Sp'ectral :'analysis Although there are several gaps of varying lengths in the series, the material is principally suited to investigate time scales from several days to 23 years. After interpolation we performed spectral ~nalysis on the full time series. 5 . ; Because we were interested in an as high resolution of the 10wfrequency ~art ofthe variance spectra ~s possib1e; we tried to , avoid any 10ss of. data. Therefore the Fast-F6urier-Transform ~ith the necessity of N being apower of 2 cou1d,not be used. For the same reason; the usua1 technique of partitioning the series ,for • getting a smoothed spectrum was not app1ied~ We didspectra1 'analysis with a real Fourier transform after prime fact6r sp1iti ting. :The dots in the figures of the variance spectra represent means 'of five (separated) variance va1ues. The unit 6f variance is the ; square of the unH of the parameter in question. The confidence interval ,(JENKINS & WATTS, , 1968) of the chi-square distributed mean variances for now 10 degrees of freedom,is plotted for an 'error probability, of 5% in the lower 1eft edge of the,figures. :Because ofthe logarithmic sca1e, this interval is valid for all points of the spectrum. The variance represented 6y certain frequency ranges' were :obtained by summing up the variance va1ues in the original '!spectra (not shown)within.these ranges (Parieva1s theorem). The :regression 1ines in the figures were ca1cu1ated as least square fits in the double 10garithmic sca1es a10ng 1n(variance) , , .. , = ' ... ' a * 1n(frequency) + "I In the spectra of sa1inity; ,frequency of 14.7 days (i.e~ ,showing up. This frequency :region, where the M2-tide is 6. nitrate, and nitrite thea1iasing 24.7 cyc1es/year) is significant1y is the resu1t of daily samp1ing in a of greatest importance~ ,5. Interpretation of spectra From t~e specira we ~an,le~rn, on w~ich sca1e~ the ~iffer~nt processes - physica1, chemieal; or bio10gica1 - mayact and dominate the deve10pment of chemica1 and bio1ogica1 parameters in the sea. Here spectra were ca1cu1ated for the time,series of, temperature, sa1initY,.nutrients and phytoplankton (Figs. 13-24). The spectra exhibit different.structures. The.outstanding frequency in many of the spectra is the annua1,frequency, which is demonstrated by the ca1cu1ation ,of the pcirtions,of variance is percentagesof total variance of the spectrum within different frequency bands (Tab.1)~ , , I ' I I ' The variance of temperature (F1g., 13) at He1go1and is near1y exc1usive1y found in the anriua1 period, ~hich cou1d mean that the annua1 heating cyc1e dominates all other processes 1ike advection of water masses with different temperaturesand fresh water inf10w. In c6ntrast, variance of sa1inity (Fig. 14) is determined to 21% by the annu~l ~yc1e; arid 34% of,the total .variance is found in the range on 1 to 10 cyc1es / Year. Long~term processes are even more important (25%) than the annual frequency. ,This is very similar to what the variances of Elbe river dischargeteach (Fig. 15). 36% of variance is due to periods,either1onger ,or periods shorter than the annual ~eriod, suggesting that salinity may be determined by river runoff. In the range of shorter time scales than 0.1 year salinity shows 18%, but river discharge on1y 5% of variance" demonstrating that discharge events are not hap~ pening on such short scales. Thus salinity must be determined 6 by different processes on these scales. The aliasing frequency (24.7 cycles/year) is significant in the salinity spectrum (see also BECKER &KOHNKE, 1978), but not in the temperature spectrum. This would occur, if e.g. temperature would be predominantly determined by the heating processes, acting on larger spatial scales than the processes regulating salinity, which are mainly connected to the river inflow. • Phosphate (Fig. 16) and nitrate (Fig. 17) contain 52% and 41% of their variance in the annual period, whereas nitrite (Fig. 18), ammonia (Fig. 19), and silicate (Fi~. 20) have more than 35% of their variance in the range of 1 to 10 cycles/ year. Phosphate is hardly determined by processes on long time scales (8%), in contrast to nitrate (20%). Ammonia (Fig. 19), as an exception, has nearly no contribution of the annual frequency to total variance, and 61% of its variance is determined by processes on time scales less than 1 year. Long-term contributions to variance are small for nitrite (16%) and silicate (13%) • The bulk of variance in the plankton series (Figs. 22 - 24) is contained in the frequency band between 1 and 10 cycles/year and, for diatoms only, in the higher frequencies (37%). Flagellates (Fig. 24) behave very different from diatoms (Fig. 23): the former contain only 2% of total variance in the annual frequency, 66% in the range from 1 to 10 cycles/ year, and only 9% on shorter scales, whereas diatoms exhibit nearly the same part of variance in the two higher frequency ranges (36% and 37%) and only 11 and 16 % in the frequency ranges larger than or equal to one year. Thus we must assume that the variance spectra of diatoms and of flagellates must be based on different dynamics. parameter variance in % of total variance in the different frequency bands f<1 • sea surface temperature salinity phosphate nitrate nitrite ammonia total inorganic nutrient-N sil icate Elbe discharge phytoplankton diatoms flagellates 1 ~ f <10 annual frequency f=1 10 ~ 2 25 8 20 16 35 95 21 52 41 36 4 2 34 26 28 41 37 1 18 14 11 7 24 22 13 36 11 35 35 23 21 16 13 14 5 31 37 23 2 31 39 36 39 36 66 9 9 f Tab. 1: Portions of total variance in different frequency bands 7 :The Helgoland Reede time series exhibit for most of the parameters a decay in the short period part of the spectrum resembling that of temperature (Tab.2). It is only Elbe river discharge, silicate and nitrite which show a much stronger decay in that part of the spectrum (the rates are -3.43, -2.13, and -2.11, resp.). For Elbe discharge this means a strongly decreasing ,importance of short term events, for scales less than a month. :For the other two parameters the situation is not clear. 0 In the frequency band from 1 to 10 cycles/year the decay rates 'are generally, except for nitrate and nitrate, smaller than -5/3. They are relatively similar for sea surface temperature, and ,total nutrient-N. They are smallest for phytoplankton, especially for diatoms (-0.5). In this range nearly all parameters'seem to ,be determined by processes which do not obey a -5/3 law. ;A more detailed discussion of interrelations of the parameters 'will follow, when the cross spectra and coherence spectra are 'available. 'parameter ,sea surface temperature salinity phosphate nitrate :nitrite ,alTlTlonia :total inorganic nutrient N ,silicate Elbe discharge 'phytoplankton C ,diatoms flagellates slopes in the frequency band 1<f<10 10~f -1.4 -1.8 -1.7 -1.8 -1.8 -1.1 -1.0 -1.6 :"1.7 -1.1 -1.5 -1.3 -2.1 -1.9 -1.9 -2.1 -3.4 -1.5 -1.6 -1.7 -1.5 -0.9 -0.5 -1.3 Tab. 2: Slopes of variance spectra obtained by linear regression analysis for different spectral bands 6. Acknowledgements o • We are obliged to the technicians of the Biologische Anstalt Helgoland for careful collection of the data, especially to E.-H. Harms, P. Mangelsdorf, and K. Treutner. We like'to thank Prof. Dr. M. Gillbricht (Biologische Anstalt Helgoland) and Dr. D. Kohnke (Deutsches Ozeanographisches Datenzentrum) for making :available the data used in this study. Many thanks are also due ,ta my colleague Dr. M. Bohle-Carbonell for his valuable comments 'during the editorial phase and to my student A. Moll for his technical assistance. 8 • 7. References • • BECKER, ,G., 1981: Beiträge zur Hydrographie und Wärmebilanz der Nordsee.- DHZ 34: 167-262. BECKER, G., & D. KOHNKE, 1978: Long-term variations of temperat ure and salinity in the inner German Bight.- Rapp. P.- v. Reun. Cons. int. Explor. Mer, 172: 335 - 344. BERG, J. & G. RADACH, 1985: Trends in nutrient and phytoplankton concentrations at Helgoland Reede (German Bight) since 1962.ICES C.M. 1985/L:2, Copenhagen. BIOLOGISCHE ANSTALT HELGOLAND, 1962 -1984: Jahresberichte. Hamburg. GILLBRICHT, M., 1983: Eine "red tide~ in der sUdlichen Nordsee und ihre Beziehungen zur Umwelt.- Helgoländer Meeresunters. 36: 393 - 426. . . GILLBRICHT, M., 1986: Phytoplankton and nutrients near Helgoland.~ ICES C.M. 1986/L:20. JENKINS, G.M., & D.G. WATTS, 1968: Spectral analysis and its applications.- Holden-Day, San Francisco. LUCHT, F., & M. GILLBRICHT, 1978: Long-term observations on nutrient contents near Helgoland in relation to nutrient input of the Elbe river.- Rapp. P.-v. Cons. int. Explor. Mer, 172: 358 - 360. OKUBO, A., 1980: Diffusion and ecological problems: mathematical models.- Spring:r-Verlag, Berlin Heidelberg New York, 254 pp. PLATT, T., 1972: Lo:al p~ytoplankton abunjance and turbulence.Deep-Sea Res. 19: 183 - 187. PLATT, T., &K.L. DENMAN, 1975: Spectral analysis in e:ology.Ann. Rev. Ecol. Syst. 6: 189 -210. RADACH, G, & J. BERG, 1985: Trends in den Ko~zentrationen der Nährstoffe und des Phytoplanktons in der Helgoländer Bucht (Helgoland Reede Daten).- Berichte der Biologischen Anstalt Helg~land, Nr. 2 (in p~ess).· . STRATHMA~N, R.R. 1967; Estimating the organic carbon content of phytoplankton from cell v~lume o~ plasma volume.- Limnol. Oceanogr. 12, 411-418. TO~T, S., & T. PLATT, 1978: Fluctuations in the abundance of phytoplankton o~ the California coast.- in: E. NAYLOR & G. HARTNDLL, 1978: Cyclic p~enomena in marine plants and animals. Pergamon Press, Oxford • UTERMOHL, H. 1958; Zur Vervollkommnung der quantitativen Phytoplankton-MethJdik. Mitt. int. Ver. Limnol. 9, 1-33. WEIGEL, H.-P. 1978; Temperature and salinity observations from Helgoland Reede in 1976.- Annls. biol. Cons. int. Explor. Mer 33: 35. 9 CENTJCAROf CENTJCAROE· 20 20 10 O-h!~""TFP1~Ill'll'Ir--''!l'IlT~:1!'r--r::rn:-"""'l'"Tln--""T!l7"r.-~'--''I!l'I''Ir-T:l1I''il'--'l'!'r'r--' 1~1I2 SURFRCE TEHPEARTUAE HERN SERSONRL SIGNAL CREWS OF M.B. 'ELLENBOGEN' RND 'RADE' I BAH Fig. 1: Surface temperature at Helgoland Reede~ 23"years of data projected into one year: a) all measurements b) mean annual cycle and standard deviations on the basis of the measurements of each day 0/00 35 a) 0/00 35 30 b) 30 ...... , ' :., ...." , . 25 -tll J stllj-,,",W-'"Imrr--n!mr"--rMllV-T'lI'I1C---nrn-"""'T!!1?"-IU"Ir-"m!'''--r.r1!'r-'''lJ''!''''--' SALINITY PROJECTED INTO 1 Y B.HETSCHER. K.TREUTNER I '11' SALINITY HEAN SEASONAL SIGNRL B,METSCHER. K.TAEUTNEA I SAH SAH Fig. 2: Salinity at Helgoland Reede, 23 years of data projected into one year: a) all measurements b) mean annual cycle and standard deviations on the basjs of the measurements of each day • e • " . .3/~I!C 3000 ", ,~) ""3/~EC " 3000 e b) 1500 o J,M, o ELBE CHARGE RT NEU DARCHAU PAOJECTEO INro 1 Y DTSCH, GEWAESSERKUNOL. JAHRS. J N IM' ELSE CHARGE RT NEU DARCHAU HEAN SEASONAL SIGNRL orSCH. GEWAESSERKUNOL. JAHRS, Fig. 3: Elbe d;scharge at Neu Oarchau, 23 years of data projected into one year: a) all measurements b) mean annual cycle and standard deviations on the basis of the measurements of each day HY G-AT I L MY G-AT I L 2.50 2.50 a) b) 1.25 IM' PHOSPHATE - P PAOJECTED IHTO l Y E.H,HAAHS. P.HANGELSOCRF I SAH PHOSPHATE - P HEAN SEASCNAL SIGNAL E.H.HARHS. P.HANGELSOORF I SAH Fig. 4: Phosphate at Helgoland Reede, 23 years of data projected into one year: a) all measurements b) mean annual cycle and standard deviations on the basis of the measurements of each day MT G-RT 1 L MT G-RT 1 L 60 60 a) .. , b) 30 30 , .... ---- , , , -,-, . ," , ' o-+,J=Rt-j,.,--..,........--.....".......-"""..--r."""'.--....,..,o=--~~- ....."..::...=.;:mr=--=">'T'"-.,.;,~-:;,"""'..--, o IOn 1112 NITRRTE - N NITRATE - N PRDJEC1ED IN10 1 Y E.H.HRAMS. P.MANGElSOCRF I SAH Fig. 5: HEAN SEASONAl SIGNAL E.H.HARHS. P.MANGElSOORF I BAH Nitrate at He1go1and Reede, 23 years of data projected into one year: b) mean annua1 cyc1e and standard deviations on the basis of the measurements of each day a) all measurements "T G-Ar I' L; ~ 14 "T G-Ar I L , " 11 a) b) :" " . ,', 2 2 \ \ ,, , •,, ,, ,. , ..... - ....... -,.,... , , \ \ ', ....... , ... o ..., ......... , '.... NITRlTE - N PAOJECTED INfO 1 Y E.H.HAAHS. P.MANGElSOORF HE AN SEASONAL SlGNAL E.H.HAAMS. P.MANGELSOORF I SAH SAH 6: Nitrite at He1g01and Reede, 23 years of data projected into one year: b) mean annua1 cyc1e and standard deviations on the basis of the measurements of each day a) all measurements • I ." -,'-- , ", ",-- .. ,'''- ......."'u"'--'?P1"--""""'-""'T'[rnv-""T1Inr--.., NITRITE - N 1 ,, ---------------.--_.,----- o -+-J=t-j-.............-""'"..---,"""..---r.H:nr...-...... J'r.U"'N.---r.J'r.Ul""L,.., J Fig. -- ..... __ ." " .... _,.-._ ... -- .... -' -".- " ,, • , ,, , , • KT G-AT I L 30 a) .. .... KT G-AT I L 30 e b) " , 15 ." . . •• 0°:. '. IS , ,,' " .. ............. _-_ .... ' , ..... -...... " .. -_ .. --- ... ' ---.--- ... ........ -....... RHHONIA N PACJECTED INTO I Y E. H.HARHS. P.HANGELSOORF o -h;;'M3r.N;-~nr--r,4JllI'-'11lI!rw---r,ü17-T'1i";;r-~;;;r-"T/llTI!""--rU"ll.-"'T1Ilrr---rtri'fii.-"'Tll1!"l""'--, I AHHONIR - N HE AN SEASONAL SIGNAL E.H.HAAH5. P.MANGEL500RF / SRH SAH Fig. 7: Ammonia at Helgoland Reede, 23 years of data projected into one year: a) all measurements b) mean annual cycle and standard deviations on the basis of the measurements of each day HT G-AT I L HT G-AT I l 60 80 a) • b) I TOTAL INORGRNIC NUTRIENT PROJECTED INTO.l Y N03-N + N02-N + NH~-N N TOTAL INORGANIC NUTRIENT - N HE AN SEASONRL SIGNAL N03-N + N02-N + NH~-N Fig. B: Total inorganic nutrient-N at Helgoland Reede, 23 years of data projected into one year: a} all measurements b) mean annual cycle and standard deviations on the basis of the measurements of each day tlT G-RT I L tlT G-RT I L 110 110 a) b) '. . . . , .. ' . . 20 20 ,"._----, ..... ~' ," ... _ _ .... , . . . ....... ' , ... tI' o SILICATE - SI PROJECTED INTO I Y E.H.HRRHS. P.MANGELSOORF I SAH " .-----_/ -.- SILICATE - SI HEAN SEASONAL SIGNAL E.H.HARMS. P.MANGELSOORF I SAH Fig. 9: Silicate at Helgoland Reede, 19 years of data projected into one year: a) all measurements b) mean annual cycle and standard deviations on the basis of the measurements of each day "f G I L Hf eiL 10· 10' ,~ :"";4a):":iüi~~ti~~j~~~~~~~ii:::~:;;,.,,,;. _c~~~"\·~1,"r·4'i1~~./;:,"'."':':f!:':~." ;..,:.:. =-: ...... ::.~ .....:.. )~"'f"(.I.~··. b) 10' ':,{C;';: ·:.:;'"·:,:':~·.~':"')'t;,~;",,~'rJ~~.~'~·~.e :: ~W~}~{t~(~~~~;'~; 'Xi/\iV/" 'T\' "''''~:'':'~~f;~~~{W~!~0~!~ .. . .. . ...... ...':' . . .. ' .. ' .' ••••• ...... ------_ ... ' " " " , ...... ----- 10' , ' ..... - ...... - .... ' -...... -- .. _---- .... --. .. _---- .... --, ,-' " 0" ", 10·' ..., 10" +J"..N:r--..!"PIr--.""....- ............-"""'...--..."""::--...".".--.."""'-...........-~,....-"='."..-."..'T"---, PHYTOPLANKTON - C PACJECTED INHl I Y H.GllLBRICHT. E.HAGMEIEA. K.TAEUTNEA I BAH PHYTOPLANKTON - C "EAN SEASONAL SIGNAL M~GILLSA[CHT. E.HAGMEIEA. K.TAEUTNER I SRH Fig. 10: Phytoplankton at Helgoland Reede, 23 years of data projected into one year: a) all measurements b) mean annual cycle and standard deviations on the basis of the measurements of each day • e • "' li I L 10' b) 10' ,, ,, ,• 10' , , ,, ,,---.... ... .",,'-- ....... .,- .. ... ' ........ ,, " 10' ,, -- ..... ,, -_.--.-.-.".""-- 10-\ OIATOHS - C PROJECTEO INTO I Y H.GILLBAICHT. E.HAGHEIER. K.TAEUTNER I OlRT0l15 C I1EAN SEASONAL SIGNAL I1.GILLSRICHT. E.HAGMEIER. K.TREUTNER SAH I SAH Fig. 11: Diatoms at Helgoland Reede, 23 years of data projected into one year: a) all measurements b) mean annual cycle and standard deviations on the basis of the measurements of each day "' li I L '. . 10' "T eIL 10' ". ::';:~'i/:~~0.iii~li~~f.~~i;~iit;~:;:.:,;:. .,; .,. .,: · b) JO' ;;.-",> " 10' loe . .~;~~\·t·~~:·~.,;~~~t~·~?·~.',: 'i ~:::;..;;;:,::~ '.; ....~::.~~~~h~~t{.~:.:..-~~;·?\··"::;\·~·:~.C~~~~),:~:~i:~~· :"~~~~.;C":t.~"~' ;r0:~~~\~!~~~~~;~~'{~\Y~7~;:?~';>~).:!~;f:.~~r~~r~~[1~~.~ ',' 10-' :" ....... :. " .... _--_ ....... 10' 10' ,"' '. 10" +JilWN-"TPrr-1iiI!_--I'IlII"....-rIür.r--'T"Tl'Iö:r--rrrn--""Tll;rnr-""T'l!rr-~"T""-TWr.;--IliF.,....-, ,.2 FLRGELLRTES - C PROJECTED INTO 1 Y H.GILLSRICHT. E.HAGHEIER. K.TREUTNER / SAH FLRGELLATES - C HE RN SEASONRL SIGNAL H.GILLBRJCHT. E.HAGMEJER. K.TREUTNER I SAH Fig. 12: Flagellates at Helgoland Reede, 23 years of data projected into one year: a) all measurements b) mean annual cycle and standard deviations on the basis of the measurements of each day VARIANCE I'" 10-1 -. I~ .. '0-' 10" .... .... :.~: .-.":.' :.....:.::. ..-.,.... . .' ~ .~..:: ;.. ~:~:;~.',(.:::.~ ...., ... .~~~ ,.~ .:..-:~~~~':~~ '. • I '~+----r--_--...---r ~ ........-ro.........,...---_-_--.-,..-,~-ro~----'---'-W ~ SURFACE TEMPERATURE PATCHED . :.:;;' ":. __~..............,r---"-T"----r-_~-.--r-.........., ~ W CYClES/Y 5-F. FG-10. P-.05 LG lTt -A'8-LG lXI A- -1.1'15801 8- -0.611315 1I·-0.8:!3 P-0.050 "-'I IIAIIOIl" 11 AT P.II -0.950 LG In -A'8-LG lXI A- -1.663139 8- -1.'I13:!50 11--0. BI1 P·0.050 11-'11 IIAIIOIl" 11 AT P.II -0.301 Fig. 13: LG In .A·hLG lXI A- -1.010319 8. -1.190B6 11·-0.921 P.0.050 11-193 "AIIOO" " "T P.II -0.010 Variance spectrum of sea surface temperature VARIANCE The confidence intervall for each speclral value is illdlcated ill the lower left; lhe regress ion 1ines for the 3 frequency bands f<l. l<f 10. f<10 cycles/year are given togelher wHh the number of values N and lhe significallce threshold of lhe correlatioll coefficlent R. 1.1 I~ .. : I'" I~ .......,r----.---.--_...--........-.-~---...--.....---.---'r-.--.-...,...,..,.----_-_-.---.--.-.....-........, I~-I-----.---.--r-........-r-....... ~ w ~ SAllNITY PATCHED LG ITI-A'B_LG lXI A- -1.8BB:!91 B- -0.961\31 11--0.9'16 P-0.050 11-'1 "AIIOOII" AT -0.950 p." S-f. fG-10. P-.05 LG lTl-A'B-lG lXI A- -1.65'13'19 8- -I. 1\ 5B1'I "--0.110 P'0.050 11-'11 "AIIOO" " AT P.II -0.301 Fig. 14: Lli lTI -A'Bolli lXI A- -0.981321 B- -1.1281'18 "--0.925 P-0.050 ..·193 "AHOO" " AI P.II -0.010 Variance spectrum of salinity ~ w CYCLES/Y VARIANCE I" I" .... ----..-_---: ..... .... .... ..... ..... 1.... +------.,...---..-.......-.,.-.,.................,..---....--.......-...--....-~ .................,---""'T""---.--.;.:....:41~...__.-.-r_---...-__.- ......._._...............'"'T""110' CYCLES/Y 1 , - 1 . . . ELSE CHARGE AT NEU DARCHAU 'x. lG !TI -A-8oLG A- -2.2121D6 8- -0.80I7Q8 11--0.938 '-0.050 M-' AANOO" R AT P.N -0.950 5-F, FG-10, P-.05 lG ITI -A-8oLG 'XI A- -1.9909 8- -'.'768'9 11--0.852 '-0.050 AAMOO" R AT -0.308 '.H Fig. 15: lG m-A-8oLG 'X' A- 0.055806 11- -3.'29901 11--0.911 P-0.050 M-188 IIANDO" R AT P.N -0.010 N-., Variance spectrum of Elbe discharge at Neu Oarchau VAR IANCE 11' The confldellce inlervall fo,' each <,pectral valllc i5 indicated in the 10wer len; lhe regress i Oll li nes for lhe 3 frcqllellcy hands f<l, l<f 10. f<1O cyclt>5!yearare givcnlogelher wilh lhe n",,,ber of villucs N alld lhe silJniricilnce threshold of lhe correlatlon coefficient R. .0' Ie-' I'" I'" '. I.... .... ...... ... ...:...... ~ 10" . ..::.~.;.~.:~: , ".- •• .'., . -01'; ." .~ . . . .•.. .0" I" 101 PHOSPHATE - P PATCHED lG!T, -A'8oLG .XI A- -2. 99929Q s- -0.082232 11--0.213 '-0.050 M-' RANOO" A AT P.M -0.950 5-F, FG-10. P-.05 lG!T' -A'8oLG .XI A- -Z.96830S 8- -1.039637 11--0.618 '-0.050 M-" IIANOO" 11 AT '.H -0.308 Fig. 16: LG lYt -A'8olG 'X' A- -Z.OSQI9Z 8- -1.189383 11-·0.921 '-0.050 H-19Z IIANOO" 11 AT '.H -0.010 Variance spectrum of phosphate ".; .. ' ...' ... .I.:.....,:.. ::-::.~.~~~.~! . ..._.......;:..., : ..~._ ;, ß·... 'I~ '. ;~.l. ..... 'i·f -l-----...--,..---.---..--..--.-....-....----.----..-....---.---.--.--.-r---....---.-_~..._ I_ _ ~--.--,..--.--..-...."r"I,, CYCLESIY 10' VARIANCE ..I . -- I" .... .... •r' I'" .... ....-1f - - - -.....10- .......--.--.-...........,....,.-r---...---.-----..-............-..................---_._-_-....-....-''T_........~r_-- .........-__..-"T__.__.._.~ CYCLES/Y I" lG (TI.A-8.lG lXI A- -D.05555Z 8- -0.6533611 R-·0.926 1'-0.050 N-" RAHDDR A AT ".N -0.950 .01 .01 HITRATE - N PATCHEO .0" S-F, FG-IO. p•• OS lG nl -A-8.lG lXI A- 0.Z00106 8- -1.5688611 R--0.859 1'-0.050 H-III RAHDD" R AT P.N -0.308 Fig. 17: lG!TI -A'8olG lXI R. 0.6511Z95 8· -1.8021116 R·-0.912 1'-0.050 H.19Z RANDO" A AT P.N '0.010 ". Variance spectrum of nitrate VARIANCE The confidence intervall for each specLral value is indicaled in the lower lefl; Lhe regress ion 1ines for Lhe 3 frequency hands f < 1, 1 < f 10, f< lU eye les/year are 9 iven togeLher wi Lh Lhe number of va lues N and lhe si gn i f icance lIll'esho 1d of lhe eorrelalion coefficient R. , ..I 10' I'" .... 10" I'" I'" ....-t----.......- ......- r - " T - - -........-.-r--:---.--__--r-..-.-.--........., ..-I NITRITE - N PATCHED lG nl ·A·BolG lXI A- -Z.0915111 8- 0.308315 H·0.6n 1'-0.050 N.II HRHDD" ART ".N -0.950 I" S-F. FG.IO, P•• OS lG !T1-A-8.lG IX' A- -1.9Z5515 8- -1.1196511 R-·0.893 1"0.050 H.III HANDO" H AT P.N -0.308 Fig. --...._--._......--.---.- . .I 18: e lG ITI -A-8olG IX' R· -1.35'166'1 8· -2.131'159 H··O. 9~ I 1"0.050 N.19Z RRNOD" R RT P.N '0.010 Varianee spectrum of nitrite :-.--_-....-....-"T_............' ,I CYCLES/Y "I .. VARIANCE .1- ... .... .... .... .' •1" .... ....+---...:......,...-----.--.........--.-............---...--...--..-...-.,......,--.-....----.----.--...--.-.....-.,..............,,....----.---...--.--.--.-..............., CYCLES/Y ~. ~ ~ AHHOHIA - N PATCHED LG lTl -A·8.lG lXI A- -0.819029 I- -1.111362 R--0.751 P-0.050 11-'11 RAIIOOll A AT P.ll -0.301 Fig. 19: LG lYt -A'I.LG lXI A- O. 116936 8- -1.8'11262 1I-·0.9n P·0.050 11-192 RAllOo" fI AT P.ll -0.010 Varianee speetrum of ammonia The eOllfidenee inlervall for eaeh speelral vaille is indiealed in the lower lert; lhe regr'cssiOll lilles for '-he 3 frC'l/lIcncy halids f<l, l<f<10, f<1O cyclps/ycar aroe yivcn loyclhcr wilh lhe number of vallIes N alld the sigllifieanee lhreshold of lhe eorrelation eoefficient R• VARIANCE ... ~ 5-F, FG-10. P-.05 LG 1TI -A'8.lG lXI A- -1.0538q7 I- - I. OQ'l023 R--0.988 P-0.050 11-" RAllOOR R AT P.ll -0.950 .1- ~ .- .1"- .... .... .... . . . .: - • • e •• '. o. t_ .. .: ....., ....."". 0. ~ r:~ ~~ • 0.:.. -:_ e• ;. ..;:':~~~.. .. " : 1 . .,...... : •• •1" ....+----...--...---.--.........--.-............---...--...--..-...-.--.--.-....----.---.--...-...--.-..-.-.-r------..--..-....--.--.--.-.-.-, CYCLESIY .1"'. IgII 10' TOTAL JNORGANIC NUTRJENT - N PATCHED LG lTl -A'8.lG IXI A- -0.007198 B- -0.7301'1'1 11--0.931 P-0.050 11_'1 RAllOOll R AT P.ll -0.950 5-F, FG-IO. P-.05 LG lYt -A·8.LG lXI A- 0.28Q'lQ9 I- -1.Q65816 11-·0.850 P-0.050 11-'11 IIAllOoll fI AT P.ll -0.308 Fig. 20: LG lTJ -A'8.lG lXI A- 0.896022 U- -I.U659Q] A--0.922 P-0.05o 11-192 AAllOo" fI AT P.ll -0.010 Variance spectrum of total inorganic nutrient-N 10' 10' VARIANCE .. ••1 ... ----------- .,.. .... . ".- :'.• -. • . : : : - ..... ~.f'••••• • 01':... : ........- __ .:• .... .... ....+-----.----.-............... .0 -.-"?"'""O........ . - - - - - . - - - - . - - . . . _ _ . . . . . - . . . . . . . . ,........""T""T ur' .11' SILICATE - SI PATCHED .0' .0. 5-F, FG-1D, P-.05 LG (Tl -A'hLG 'XI A- -1.32118 8- -1.601559 11--0.996 1'-0.050 11-3 IIAND"" " AT ".11 -0.997 LG 1TI -A-S.LG 'X' A- -0.238823 LG 1TI -A-8.lG 'XI A- 0.816Z83 I- -Z.1099Q 11--0.936 1'-0.050 N-651 IIANo,," 11 AT "." -0.011 I- -1.ZSlq91 11--0.188 1'-0.050 N-n IIANO"" 11 Al ".N -0.339 Fig. 21: Variance spectrum of silicate VARIANCE The confirlence intervall for each spectral value is indicaLed in the lower len; Lhe regt'ess ion I ines fur the 3 frequency hands f<l, l<f 10, f<1O c,ycles/year are givcn togeLher wiLh Lhe number of values N and lhe significance threshold of Lhe correlalioll coefficienL R• ••1 ." .... .... .... .... .... ---.-~......._ _...,..---...---...---~_ . .............., CYCLES/Y 1 .. .., • •-:-."'?"_"""'---"...... •.' ." ... I ....-1------.....-. . . . . . . --.--.. . . -r-------.,.---~ . . . . . . ...._,.---__- ......-.....__._.. . . . . . .~r__--_.._.......-_._ .r .0' l .01 ... PHYTGPlANKTGN - C LG 'TI -A'hLG 'XI A- -1.012003 I- -0.SQ6861 A--0.858 1'-0.050 "-, IlANO"" " AT ... " -0.950 5-F, FG-10, "-.05 LG lTI -A'8.lG 'X, A- -0.10888 I- -0.889133 11--0.655 1'-0.050 "_Ql IIA"OO" 11 AT P.N -0.301 Fig. 22: LG ,TI -A-8.lG 'X, A- -0.03QQIQ I- -1.516189 1I--0.88Z 1'-0.050 N-792 IIANoO" 11 AT P.N -0.070 Variance spectrum of phytoplankton .................. CYCLES/Y .0. .... ... VARIANCE ' . ... .... .. L... . tu • -. ·c.-. -. -. . ~ .r- .... +-----..--~--. .......~....,....,~---..---.----.-..-- ......~...-r----.--_._-T___._...__""T'""r_--__._-__.-_.__._""T'"" ~ ~ ~ OIAl«!"S - C lG '" -A'8-lG ,X, A- -0.176Q32 11- -0.219351 ft--0.35'1 P-0.050 N-. AANlION ft AT P.N -0.950 ~ ................., CYCLES/Y ~ 5-F. FG-IO. P-.OS lG IYI-A'8_lG 'X' A- -0.278205 11- -0.q95766 ft--0.Q66 P-0.050 N-~' ftANOON ft Ar P.N -0.3011 Fig. 23: lG IYI ·A'8-lG 'X' A- 0.810196 11- -1.58179~ ft--0.887 P-0.050 N.792 ftANOON ft AT P.N -0.070 Varianee speetrum of diatoms The eonfidence intervall fur each speelral value is inclieated in the luwer lett; the regress iOll 1ilies for I.he 3 fre'luency hands f< 1, 1 < f to, f (10 cye lec;/year are gi vell logelher wi lh lhe I1lIßlber of values N alld lhe signifieallce lhreshold of lhe eorrelation coefficient R• VARIANCE '.' ... .... .... .... .... .... •r-+------..--~--........-....,....,~---..---.---.-..--..-._.____r---""T'""-~-....___._..._ ~ ~ w FLAGELLAlES - C lG ,Tl -A'8-lG 'X, A- -0.550952 0- -0.192559 ft·-0.812 P.0.050 N." ftANOON ft AT P.N -0.950 S-F. FG-I0. P-.OS lG IYI .A'O-lG 'X, A- -0.079211 0- -1.310211 ft--O.OIQ P-0.050 N-Ql ftANOON ft AT P.N -0.300 Fig. 24: lG ITl .A·O_lG IX' A- 0.529212 87113 A·-0.90Q P.0.050 N.787 AANOON ft AT P.N '0.010 .'.1\ Variance spectrum of flagellates .............,r__--_._-.......-..._""T'""_.__........____, CYCLES/Y ~ ~