The phytoplankton in Lake Baikal included autotrophic picoplankton (APP), nano- and microphytoplankton. The phytoplankton biovolume, the Chla and other phytoplankton pigments were distributed very heterogeneously in Lake Baikal, thereby indicating variations of both phytoplankton abundance and composition. Differences were found between the open basins (South, Centre, and North) as well as between the near-shore (Maloe More) and river-inflow (Selenga Delta, Barguzin Bay) sites.

Regional distribution of APP, nano- and microphytoplankton: The whole lake median of the total phytoplankton biovolume was 0.61 mm³ L^-1, of which 78 % was nano- and microphytoplankton and 22 % APP. The median APP biovolume over the whole lake was 0.097 mm³ L^-1 and significant differences for the total APP biovolume were found in the following order: Selenga Delta > South, Centre > North (Fig. 16). Thus, the APP comprised only 11 % of the total phytoplankton biovolume in the North, but 61 % in the Selenga Delta (Fig. 16). Eukaryotic APP dominated in the South, whereas cyanobacterial APP dominated in the Centre, North and Selenga Delta (Tab. 2).

The median nano- and microphytoplankton biovolume was 0.4 mm³ L^-1 (Fig. 16). No significant differences were found between individual regions, either based on biovolumes (Fig. 16) or cell numbers. Bacillariophyceae dominated the nano- and microphytoplankton assemblage all over the lake in both years with 70 – 90 % by volume (Tab. 2), while all other groups contributed less than 10 %.

In 2002, nano- and microphytoplankton species were investigated in greater detail than in 2001. Therefore, the results presented here will be limited to 2002. On average 53 % of the nano- and microphytoplankton biovolume resulted from endemic species (see Appendix C - Tab. 1 for endemic species). Between 9 and 14 nano- and microphyto-plankton species were identified at the different sites and depths. The mean reciprocal Simpson diversity index for the whole lake was 3.96 (Shannon-Wiener 1.92) ranging at individual sites from 2.1 to 3.6 (Tab. 2), but no significant differences or trends between the regions, stations and depths could be found for the diversity indices.

Fig. 16. Regional variability of temperature, Chla, carotenoids, and biovolumes; bars represent means with 95 % C.I. and boxes medians with interquartile ranges (25-75 %). A map of Lake Baikal showing the sampling sites is given for orientation. The APP biovolume vs. total phytoplankton biovolume ratios as well as the sum of carotenoids vs. Chla are given as percentages. The number of samples (n) for temperature and pigments was 43 - 52 in each of the three open basins and at Selenga Delta (Sdelta), 22 at Barguzin Bay (Bbay) and 6 at Academician Ridge (Aridge) and Maloe More (MM). The number of samples (n) for biovolumes was 9 - 12 in each of the three open basins and Selenga Delta.

All species were tentatively grouped into functional groups (Tab. 2) according to the scheme of Reynolds et al. (2002). The dominant functional group (by biovolume) was that of vernal bacillariophycean blooms typical of oligotrophic lakes (A), due to the dominance of Cyclotella species (Tab. 2). At a few sites the dominances changed. In the South (5 m) the dominant functional group was that usually found at the start of summer stratification in oligotrophic conditions (E) (Tab. 2). This group indicated slightly higher nutrient availability than the functional group (A). Also the groups (S1) and (Z) were found in the Selenga Delta, which indicated summer stratification (Tab. 2).

Tab. 2. Regional variation of diversity, functional groups and of the contribution to total biovolume. Only those species were listed here, which contributed more than 1 % at any of the sites.

Besides Cyclotella species, flagellata, most of them belonging to Chrysophyceae, were numerous across the whole of the lake (Tab. 2). The contribution to the total biovolume of these flagellates was nonetheless higher in the South and Selenga Delta than in the North (Tab. 2). A high number of Koliella longiseta (Chlorophyta) was found besides Cyclotella in the South and Centre, as well as a high number of Nitzschia acicularis in the Centre (Bacillariophyceae; Tab. 2). In the North Cryptomonas spp. (Cryptophyta) were found among the flagellates in high concentrations as well as Rhodomonas pusilla (Cryptophyta), which dominated the cryptophycean species at all other sites (Tab. 2). Furthermore, in the North the contribution of Cyclotella species to the total biovolume was higher than at all other sites unless Academician Ridge (Tab. 2). At Academician Ridge, Cyclotella cells were very numerous and the biovolume per Cyclotella cell was much higher than at the other sites. Their diameter reached 150 µm. The number of Aulosira sp. (cyanobacteria) cells was the highest at both sample sites within the large Selenga Delta, but they were much less abundant (> 5 times) at 5 m than at 10 m and 30 m (Tab. 2). Potential grazers of these filamentous cyanobacteria (the ciliophora Strombidium sp.) were numerous at 5 m but not in the deeper layers. The respective Aulosira species differed morphologically from A. implexa and A. laxa, which up until now were the only Aulosira species described for Lake Baikal (Bondarenko 1995) and has thus not yet been described for Lake Baikal. The description will be done elsewhere.

Regional differences of temperature and Chla: The mean July temperature across the whole lake (including all sampled depths) during the three expeditions was 8° C (Fig. 16, Fig. 17). A significant decline in temperature was found along the transect South > Centre > North (Fig. 16), with the difference between Centre and North only given at 90 % C.I. The Selenga Delta and Barguzin Bay had the highest temperatures (Fig. 16, Fig. 17). Nonetheless, the surface water (upper 5 m) in the Maloe More was also warmer compared to the open basins (Fig. 17).

The Chla concentration was significantly correlated to temperature. The whole lake average Chla concentration (2001-2003, including all sample depths) was 1.71 ħ 0.19 nmol L^-1 (95 % C.I., n=222; Fig. 16). The Chla concentrations in the three open basins (South, Centre and North) were significantly lower compared to those at the river inflows (Selenga and Barguzin) and within the open basins the North had a significantly lower Chla concentration than the South (Fig. 16). The warm surface waters of the shallow Maloe More were Chla-rich compared to the open basins (Fig. 17). The interpolation of the Chla results showed furthermore the transition from Chla-rich near-shore regions to Chla-poor open basin regions (Fig. 17).

Fig. 17. Map of Chla (left) and temperature (right) by nearest neighbouring interpolation from direct water samples taken from surface waters (0.5-5 m).

Fluorescence horizontal transects (c. 3 m water depth) performed in July 2002 and July 2003 during ship travel confirmed the interpolation of the Chla and temperature distribution (Fig. 18). From those transects the changes within the basins and towards the river inflows and near-shore regions could be continuously tracked and these transects confirmed the extremely heterogeneous Chla distribution and temperature gradients along the lake (Fig. 18). From the Ushkanin Islands (in front of the west coast of the Svyatoi Nos peninsula) along the west coast of Svyatoi Nos towards the Barguzin inflow, for example, the temperature increased from 10° to 15° C and the Chla concentrations doubled at the same time from 1.6 to 3.2 µmol L^-1 (Fig. 18-1). From the North of Svyatoi Nos along the east coast of Olkhon Island the profiles indicated a clear shift from higher temperatures and Chla concentrations near Svyatoi Nos to low temperatures and Chla concentrations in the open water of the Central basin; and temperature and Chla concentration increased again nearing the southern end of Olkhon Island (Fig. 18-2). A profile from the harbour of Listvianka (South basin) to the North of the Selenga Delta showed furthermore the increase of temperature and Chla concentration nearing the Selenga Delta (Fig. 18-5).

Fig. 18. Horizontal Chla (green) and temperature (blue) profiles by underway measurements (c. 3 m water depth) in July 2002 (red) and July 2003 (yellow).

Regional distribution of carotenoids: The total carotenoid concentration was significantly correlated to Chla (r²= 0.94, p< 0.001, n= 222) and was 85 % of the Chla concentration considering molar ratio (60 % considering weight ratios; Fig. 16). As was shown for the Chla variations, the total carotenoid concentrations were significantly lower in the open basins than at the river inflow sites and the North also showed significantly lower carotenoid concentrations than the South (Fig. 16). The ratios of carotenoids that collected light (such as fucoxanthin) to carotenoids that protected cells against high light (such as diadinoxanthin) was not correlated to the total carotenoid vs. Chla ratio (data not shown). Only at Maloe More the very low carotenoid vs. Chla ratio clearly resulted from lowering protecting carotenoids, indicating thus low light acclimation of the phytoplankton at that site.

Fig. 19. Regional variability of the marker pigment concentrations for (A) Bacillariophyceae+ Chrysophyceae (fucoxanthin and Chlc), (B) for Chlorophyta (lutein and Chlb), (C) for Cryptophyta (alloxanthin) and for Eustigmatophyceae (violaxanthin), and (D) for cyanobacterial APP (zeaxanthin and ß-carotene). Combined data set 2001-2003 (July) was used for the calculations. Error bars represent a 95 % C.I. Abbreviations: SDelta – Selenga Delta, BBay – Barguzin Bay, ARidge – Academician Ridge, MM – Maloe More. Lines do not indicate trends between two neighboured points, but serve to visualise regional differences.

Taking a lake average (July 2001, 2002 and 2003), the dominant carotenoids were zeaxanthin, fucoxanthin and lutein. Fucoxanthin (Fig. 19A), Chlc (Fig. 19A), diadinoxanthin and diatoxanthin concentrations showed no significant differences between the open basins, but significantly higher values were found at Barguzin Bay and at Academician Ridge. Chlb and lutein showed significantly higher concentrations in the South than in the Centre and North, as well as at Academician Ridge and Maloe More (Fig. 19B). Alloxanthin (Fig. 19C) and peridinin did not show significant variations between the open basins and the remaining regions. Violaxanthin concentrations showed significant decreases in the order South > Centre > North (Fig. 19C). Zeaxanthin and ß-carotene were significantly highest in the Selenga Delta, and the South showed significantly higher concentrations than the North (Fig. 19D).

A discriminant analysis (Wilk´s Lamba=0.08, p<0.001; Fig. 20) indicated, that considering the pigment composition, Academician Ridge and Barguzin Bay (root 1, p<0.001) as well as Maloe More (root 2, p<0.001) were most distinct from all other regions. The major pigment within root 1 was fucoxanthin (standardised canonical coefficient = 0.89) and the major pigment, influencing root 2, was violaxanthin (standardised canonical coefficient = -1.17), indicating major influence of Bacillario-phyceae+Chrysophyceae at Academician Ridge and Barguzin Bay and of Eustigmato-phyceae at Maloe More.

Fig. 20. Discriminance analysis separating the seven regions according to the pigment distri-bution. See Fig. 19 for abbreviations.

Estimation of phytoplankton composition using marker pigments: Accessory pigments can be used to estimate the composition of the phytoplankton assemblage (cf. Fietz and Nicklisch 2004, Appendix A). Based on the whole data set of 2001 to 2003 (n= 222) significant molar Chla/marker pigment ratios were calculated using multiple linear regression. This resulted in the following equation:

Chla = 1.26·Fuco + 1.62·Allo + 3.0·Chlb + 0.61·Zea` + 6.49·Viola` (Tab. 3A), whereby Chla was the total Chla from the five phytoplankton groups, “Fuco” (fucoxanthin) was the marker for Bacillariophyceae+Chrysophyceae, “Allo” (alloxanthin) was the marker for Cryptophyta, “Chlb” was the marker for Chlorophyta, “Zea´” was the marker for cyanobacterial APP and “Viola´” was the marker for Eustigmatophyceae. Zea´ was the cyanobacterial zeaxanthin and Viola´ was the eustigmatophycean violaxanthin only (cf. Fietz and Nicklisch 2004, Appendix A for calculations; cf. Fietz et al., submitted, Appendix B for having considered Eustigmatophyceae to be a common member of the Baikalian phytoplankton). The coefficient of determination (r²) was high (0.97; Tab. 3A) and the calculated Chla matched the measured Chla with a mean error of 6 % and a maximum error of 24 % in all regions and years.

Fig. 21. Contribution of main chemotaxonomic phytoplankton groups to total Chla. Abbreviations: ARidge – Academician Ridge, Bacill.Chrys. – Bacillariophyceae+Chrysophyceae, BBay – Barguzin Bay, Cyanob. – cyanobac-terial, Eustigmato. – Eu-stigmatophyceae, SDelta – Selenga Delta.

The contributions to total Chla of the chemotaxonomic groups varied between the regions (Fig. 21). Whereas Bacillariophyceae+Chrysophyceae contributed only 20 % to the total Chla in the South and Selenga Delta, its contribution was higher (45-50 %) in the North and Barguzin Bay and highest (70 %) at Academician Ridge. The contribution of Cryptophyta was small (< 15 %) all over the lake. The contribution of Chlorophyta to the total Chla was highest in the South (40 %) and lowest at Academician Ridge and Maloe More (< 10 %). The contribution of Eustigmatophyceae was about 20 % in the South and Selenga Delta and was highest at Maloe More (45 %). Possibly several Chrysophyceae contributed to the total violaxanthin at Maloe More so that the eustigmatophycean violaxanthin was overestimated at that site. The contribution of the cyanobacterial APP was highest in the Selenga Delta (30 %) and lowest in the North, Academician Ridge and Maloe More (<5 %).

* cyanobacterial zeaxanthin only (see text)
** eustigmatophycean violaxanthin only (see text; see Fietz et al., submitted, Appendix B for Eustigmatophyceae)
*** In a preliminary study (Fietz and Nicklisch 2004, Appendix A) it was mentioned that caloxanthin, a zeaxanthin transformation product, possibly coeluted with alloxanthin in some samples. In the present study, the correlation between α-carotene and alloxanthin, both contained in Cryptophyta, was significant. Therefore we assumed that alloxanthin in almost all samples did not coelute with caloxanthin. Few outliers from the significant alloxanthin/α-carotene relationships were omitted in the multiple linear regression calculations of the present study.

A) Whole lake

molar ratio

P

partial c.

95 % C.I.

P

r²

Chla / Fucoxanthin

1.26

< 0.001

0.81

0.12

< 0.001

0.98

Chla / Chlb

3.00

< 0.001

0.69

0.42

Chla / Alloxanthin^***

1.62

< 0.001

0.37

0.54

Chla / Zeaxanthin^*

0.61

< 0.001

0.82

0.06

Chla / Violaxanthin^**

6.49

< 0.001

0.84

0.57

B) Regions

molar ratio

P

partial c.

95 % C.I.

P

r²

South

Chla / Fucoxanthin

1.17

< 0.001

0.77

0.33

< 0.001

0.99

Chla / Chlb

3.28

< 0.001

0.95

0.37

Chla / Alloxanthin^***

1.55

< 0.05

0.33

1.50

Chla / Zeaxanthin^*

0.79

< 0.001

0.96

0.07

Chla / Violaxanthin^**

1.64

<0.005

0.45

1.10

Centre

Chla / Fucoxanthin

1.57

< 0.001

0.94

0.18

< 0.001

0.99

Chla / Chlb

3.18

< 0.001

0.87

0.57

Chla / Alloxanthin^***

1.17

< 0.001

0.55

0.57

Chla / Zeaxanthin^*

0.57

< 0.001

0.77

0.15

Chla / Violaxanthin^**

3.35

< 0.001

0.52

1.76

North

Chla / Fucoxanthin

1.12

< 0.001

0.64

0.38

< 0.001

0.94

Chla / Chlb

7.36

< 0.001

0.73

1.96

Selenga Delta

Chla / Fucoxanthin

1.39

< 0.001

0.69

0.45

< 0.001

0.98

Chla / Chlb

3.87

< 0.001

0.70

1.24

Chla / Alloxanthin^***

1.48

0.08

0.26

1.69

Chla / Zeaxanthin^*

0.66

< 0.001

0.86

0.12

Chla / Violaxanthin^**

3.45

< 0.05

0.32

3.16

Barguzin Bay

Chla / Fucoxanthin

0.79

< 0.001

0.68

0.41

< 0.001

0.97

Chla / Chlb

5.77

< 0.001

0.85

1.71

Chla / Violaxanthin^**

 

9.75

< 0.001

0.79

3.61

Tab. 3. Molar ratios of Chla vs. marker pigments and their respective statistics calculated by multiple linear regression (A) for the complete data set (July 2001, July 2002 and July 2003) and (B) regionally differentiated.

The factors calculated for the whole lake do not account for the variances of differential phytoplankton compositions in the individual regions of the lake (as indicated by the regional variations of the species and marker pigments, Fig. 19). Therefore, individual Chla/marker pigment ratios were calculated for each region (Tab. 3B). The splitting of the data set, however, limited the calculation of significant factors in the North and also in Barguzin Bay (Tab. 3B). In most cases the regionally differentiated factors did not differ significantly from those calculated for the whole lake data set (Tab. 3A,B) and, therefore, the use of the whole lake factors can be recommended.

Interannual variability: Significant differences in the Chla concentrations were found in each basin when the three years were compared, particularly in the North (Fig. 22A). In 2001 the Centre had significantly the lowest Chla concentrations, while in 2002 concentrations were lowest in the North (Fig. 22A). Similarly, significant interannual differences were found for each of the carotenoids, for Chlb (Fig. 22B), for Chlc (Fig. 22C), for the sum of carotenoids and for the carotenoids vs. Chla ratios (data not shown), particularly in the North and Centre.

Fig. 22. Interannual variability of Chlorophylls: (A) Chla; (B) Chlb, and (C) Chlc. Error bars represent a 95 % C.I. Abbreviation: S Delta – Selenga Delta.

High interannual variability was also found for the cyanobacterial and eukaryotic APP biovolume, particularly in the South and Centre ( Fig. 23A). Nevertheless, the North had lowest APP biovolumes in both years and the Selenga Delta the highest (Fig. 23A). Additionally, high interannual variability was found for the nano- and microphytoplankton biovolume, particularly in the North, where a much lower biovolume was determined in July 2002 compared to July 2001 (Fig. 23B). This change in biovolume was mainly due to shifts in the species composition that means shifts in the biovolume/cell ratio (mainly of Bacillariophyceae, data not shown). As has been shown for the biovolumes (Fig. 23A,B) and for chlorophylls and carotenoids (Fig. 22), the estimated contributions of the distinct chemotaxonomic groups differed also between the three investigated years (Fig. 23C).

Fig. 23. Interannual variability of (A) eukaryotic and cyano-bacterial APP biovolume; (B) nano- and microphytoplankton biovolume; (C) contribution of chemotaxonomic groups to total Chla (see text for calculation); Abbreviations: Bacillarioph. – Bacillariophyceae, BacillChrys – Bacillariophyceae+Chrysophycea, Cyanob. APP – cyanobacterial APP, Eukary APP – eukaryotic APP, Eustigmato. – Eustigmato-phyceae, S Delta – Selenga Delta.
Note: see Appendix B for having considered Eustigmatophyceae to be a common member of the Baikalian phytoplankton.

Temperature and chlorophyll depth profiles from January 2001 to December 2003 near Bolshye Koti (see Fig. 16 for location) are plotted in Fig. 24. A thin convective layer occurred under the ice in February, where temperatures below 0.5° C were found down to 25 m (Fig. 24). The temperature increased to 3° C between 50 m and 100 m, and remained between 3 and 4° C from 250 m down to the lake bottom (Fig. 24). Therefore, this period showed an inverse stratification, with mixing restricted to a shallow layer under the ice. Every year wind-induced mixing spanning over the upper 250 m occurred from May to June, after ice-break up, and another in November (Fig. 24). During spring overturn in 2001 Chla concentrations up to 0.5 nmol L^-1 were found down to 100 m and up to 0.25 nmol L^-1 down to 250 m (Fig. 24). During the spring overturn in 2002 Chla concentrations over 2 nmol L^-1 were found down to 100 m and up to 0.5 nmol L^-1 down to 250 m, while during summer stratification the Chla concentration was low (<0.2 nmol L^-1) below 50 m (Fig. 24).

Fig. 24. Temperature and Chla depth profiles (0-250 m) from January 2001 to December 2003 showing mixing and stratification over the season and regions. Values at the time of ice break-up (April) have been interpolated. From Mid-April to end of August 2003 data below 50 m were not available (hatched rectangle).

Fig. 25. Vertical variability of temperature, pigments (see Fig. 16 for bars and boxes). The APP biovolume vs. total phytoplankton biovolume ratios as well as the sum of carotenoids vs. Chla are given as percentages. The number of samples (n) for temperature and pigments was 21 at 0.5 m, 110 at 5 m, 43 at 10-20 m, 36 at 20-30 m and 6 at 45-85 m. The number of samples (n) for biovolumes was 10 - 18 at each depth.

The average July temperatures of the lake (2001-2003) decreased significantly with the depth in the order: 0.5 m, 5m > 10-20 m > 20 – 30 m, 45 – 85 m (Fig. 25). Stratification and mixing might also result from conductivity in oceanic and also some freshwater systems, but in Lake Baikal conductivity varied only between 100 to 110 µS cm^-1 (corrected to 20° C) between the individual regions as well as with the depth (Fig. 26). Mixing conditions varied among the different basins and regions. High-resolution temperature and conductivity depth profiles showed that in the North mixed conditions prevailed, while a weak stratification developed in the South and Selenga Delta. Stable stratification with a broad epilimnion developed in Barguzin Bay (Fig. 26). There, the conductivity showed a second maximum at 20-25 m, probably indicating the influence of subsurface water currents induced by the river (Fig. 26). The Chla concentration in July showed a maximum at 16 m in the South, whereas it was rather homogeneously distributed in the North and Centre. It decreased with depth in the Barguzin Bay and most prominently in the Maloe More (Fig. 26).

Fig. 26. Depth profiles (0-30 m) of (A) temperature, (B) Chla, and (C) conductivity.

The combined data set of all samples collected during the July expeditions 2001 and 2003 revealed a significant decrease of the Chla concentration and the sum of carotenoids below 20 m compared to the 5 m and 10 – 20 m samples (Fig. 25). The percentage sum of carotenoids compared to the Chla was also lowest (74 % molar ratio) in the 20 – 30 m layers (Fig. 25). At several stations, deep Chla maxima were found in depths of 45 m, 50 m and 85 m. Within these deep maxima the Chla concentrations were as high as above 20 m (Fig. 25). The sum of carotenoids was lower in these deep layers than above 20 m and the percentage molar ratio was only 69 % (Fig. 25). At 45 – 85 m the ratio of ´light-collecting´ vs. ´protecting´ carotenoids was highest (data not shown), indicating that the lower total carotenoid vs. Chla ratio in greater depths resulted from reduced protecting carotenoids. Thus, as expected low-light conditions prevailed at the great depths, but high light conditions down to 10 – 20 m.

The total APP also decreased in the 20 – 30 m samples compared to the 5 m samples, whereas no significant changes could be found for the nano- and microphytoplankton (Fig. 25). There were some tendencies in both years that biovolumes differed with depths within the individual regions. The relatively shallow Selenga Delta for example showed a maximum of the APP biovolume at 16 m depth and of the nano- and microphytoplankton at 5 m (Tab. 2). Several species showed also distinct relationships with depth. Aulacoseira baicalensis for example formed a deep maximum in 85 m water depth in the South (Tab. 2).

* Highest values within a set for each canonical variate.
Note: The stratification was determined as 0=mixed and 1=stratified according to the temperature profiles carried out with the CTD-probe and/or submersible FluoroProbe.

 

Regional distribution

Seasonal distrib.

Canonical loadings

Canon. loadings

 

Root 1

Root 2

Root 3

Root 4

Root 1

Root 2

Set 1, environmental variables

Latitude

-0.52

-0.67^*

0.10

0.52

-

-

Water Depth

-0.51

0.44

-0.71^*

0.20

-

-

Stratification

0.99^*

-0.05

-0.15

-0.05

-0.73^*

-0.68

Temperature

0.53

-0.47

-0.20

-0.68^*

-0.74^*

0.67

Set 2, phytoplankton group contribution

Total Chl a

0.87^*

-0.40

-0.01

0.08

-0.51

-0.57^*

Bacillario.+Chrysophyceae

-0.28

-0.44

0.74^*

0.16

0.39

-0.57^*

Chlorophyta

-0.06

0.92^*

0.09

0.14

0.44^*

-0.22

Cyanobacterial APP

0.46^*

-0.02

-0.29

-0.07

-0.91*

0.07

Eustimatophyceae

0.16

-0.12

-0.30

-0.84^*

-0.33

0.38^*

Cryptophyta

-0.62^*

-0.10

-0.50

-0.02

0.16*

-0.04

 

Root 1

Root 2

Root 3

Root 4

Root 1

Root 2

Eigenvalue

0.51

0.27

0.10

0.03

0.75

0.20

Canonical Correlation (r)

0.72

0.52

0.31

0.18

0.86

0.45

Significance (p)

0.00

0.00

0.00

0.06

0.00

0.30

Percentage variance, set 1

0.45

0.22

0.14

0.19

0.54

0.46

Percentage variance, set 2

0.24

0.20

0.16

0.13

0.26

0.14

Redundancy, set 1

0.23

0.06

0.01

0.01

0.40

0.09

Redundancy, set 2

0.13

0.05

0.02

0.00

0.19

0.03

Tab. 4. Canonical correlation analysis of set one comprising the environmental variables, and set two comprising total calculated Chla and the respective percentage contribution of each phytoplankton group to the total calculated Chla. See text for calculation of the contributions.

Driving factors for phytoplankton distribution: Canonical correlation analysis (CCA; Tab. 4) indicated that total Chla as well as the percentage contributions of cyanobacterial APP and Cryptophyta to the total Chla were correlated with the first canonical variate, which explained 24.5 % of the variance. The percentage contribution of Cryptophyta to the total Chla was negatively correlated to the first canonical variate. Considering the environmental data the first canonical variate showed the strongest correlation to the stratification (Tab. 4). The second canonical variate was related to the contribution of Chlorophyta and to latitude. The third canonical variate was related to the contribution of Bacillariophyceae+ Chrysophyceae and to the water depth. The fourth variate was not significant. Although CCA does not provide direct covariation between the two sets of variables, we may state for example, that cyanobacterial APP showed highest contributions where the water column was stratified, whereby the temperature was secondary. Chlorophyta dominated in the low latitude regions, probably related to the insolation, as temperature was not of great importance. Bacillariophyceae+ Chrysophyceae dominated, where the water depth was high, that means rather in the open basins than in the near-shore or river inflow regions.

Fluorescence depth profiles: As has been suggested from the CCA, mixing and stratification were very important factors for the differential phytoplankton group distribution. Thus, in situ fluorescence depth profiles, showing the mixing/stratification regime of the water body by recording the temperature and the total Chla were analysed in July 2002 and July 2003 (Fig. 27, Fig. 28) and August 2002 (Fig. 29). Basically these profiles strengthened two trends: The first one was a trend from near-shore to offshore, well visible for example from the Barguzin transect where the regime changed from a stratified one to a mixed one with the increasing distance from the shore towards the open Central basin (Fig. 27). Another trend in July was the one from the South to the North. While in the South the water body was almost always stratified, the water body of the North was mixed (Fig. 28). Nonetheless, this trend was much less expressed in August. Then, almost all stations were stratified, even those of the North (Fig. 29).

The August depth profiles exhibited great heterogeneity for the Chla maxima even between neighbouring sites, which in Lake Baikal are, however, still a few kilometres away from each other (Fig. 29). In the South, the epilimnion spanned up to 50 m and the Chla maxima mostly occurred up to that depth (Fig. 29). Only in the central part of the South basin, where water depth reached up to 1400 m, the epilimnion was limited to 20 m (Fig. 29). Then, the Chla maximum was found between 10 and 20 m within the thermocline and the maximum Chla concentration was much higher than at all the near-shore stations. In the North several Chla maxima were also found in depths varying from 10 to 50 m (Fig. 29). The Chla maxima occurred often within the thermocline, as has been shown for the deep-water stations in the South (Fig. 29). Some of these trends were already indicated by the direct water samples (Fig. 26), but using continuous measurements, the resolution was strongly increased. Using direct samples, the risk persist to overlook deep Chla maxima.

However, it is worth to be mentioned here, that these data are uncorrected for humic substances and not calibrated including all dominant phytoplankton groups. Nonetheless, preliminary correlation analyses of A. Nicklisch (HU Berlin, Germany, pers. comm.) indicated that total Chla concentrations measured with the FluoroProbe fitted well to the total Chla determined by HPLC (r² = 0.96).

Fig. 27. Chla and temperature depth profiles assessed with the submersible fluorimeter in July 2002. Transect from near-shore to off-shore (from Barguzin Bay to eastern shore of Olkhon Island).

Fig. 28. South to North transsects of Chla and temperature depth profiles in July 2002 (red circles, profiles on left side) and July 2003 (yellow circles, profiles on right side).

Fig. 29. Depth profiles of Chla (green lines) and temperature (blue lines) in August 2002. See Fig. 28 for legend, whereby in August maximum measured depth was 100 m and maximum temperatures were 18°, 16° and 14° C in the South (including Selenga Delta), Centre and North, respectively.

Intense monitoring was conducted from May 2002 to June 2003. Variations from January 2001 to December 2003 are plo than 2001 and 2003 with a mean summer (July-September) temperature at 5 m water depth of 12° C, compared to 8.5° C in 2001 and 8° C in 2003. Peak temperatures reached 17.4° C in 2002 compared to 15.4° C in 2001 and 11.6° C in 2003. The ice cover lasted shortest in 2002 (1 month) relative to 2001 (2.5 months) and 2003 (2 months). In every year the integrated Chla below the ice was as high as in summer. Generally four peaks could be discerned (around March, June, July and September), but their intensities varied interannually. The spring peak 2001 was probably missed due to the inaccessibility of the sampling station. The spring peak 2002 was by far the highest within the three investigated years.

Seasonal dynamics of APP, nano- and microphytoplankton: During the intense monitoring 2002-2003, only spot samples could be counted with the epifluorescence microscope for APP. The amount of APP was high in July (about 0.7 mm³ L^-1) and decreased towards autumn (less than 0.2 mm³ L^-1; Fig. 30B). The APP started to increase again by the end of March (Fig. 30B). Nevertheless, the presence of APP in February and March (about 0.1 mm³ L^-1), when the lake was covered with ice, indicated an important APP formation under the ice. Light microscopic estimations of the total APP during the whole season supported the conclusions drawn from epifluorescence spot checks (Fig. 30B).

The median annual nano- and microphytoplankton biovolume was 0.07 mm³ L^-1. In spring the nano- and microphytoplankton biovolume and the cell number were high, while the diversity was low (Fig. 30B). This spring bloom (2002) was dominated by Stephanodiscus meyerii and Aulacoseira baicalensis (Bacillariophyceae) that belong to the indicators of vernal blooms usually found in mixed, mesotrophic conditions (B; Tab. 5). At the end of May to the beginning of June the nano- and microphytoplankton biovolume and cell number decreased, while the diversity increased to a maximum in Mid-June (Fig. 30B). The diversity was, therefore, highest at the time of the minimum biovolume and cell abundance.

The nano- and microphytoplankton biovolume was then low during summer and autumn, but the cell number, in contrast, was in summer as high as during spring (Fig. 30B). Towards the beginning of July (2002), the contribution of Bacillariophyceae decreased and Chlorophyta (Koliella sp. and Monoraphidium sp.) became important (Fig. 30B; Tab. 5). The dominant functional groups changed to mixed, oligotrophic (E) or mesotrophic conditions (X₂; Tab. 5). In July the assemblage comprised mainly Koliella longiseta (Chlorophyta) and Rhodomonas pusilla (Cryptophyta) (Tab. 5). According to the functional groups the large number of chrysophycean flagellates indicated the start of summer stratification (E) (Tab. 5). Bacillariophyceae were rather rare at that time of the year (Fig. 30B; Tab. 5). Small algae such as R. pusilla and Chrysophyceae (chrysophycean flagellates as well as Chrysidalis sp. and Dinobryon spp.) predominated also during the whole August and September (Tab. 5). Asterionella formosa (Bacillariophyceae) reached a mass development (10-fold increase within one week) in autumn that was unusual for the Bolshye Koti site (Tab. 5).

Under the ice of the next year (2003) the biovolume and cell number were low, but the diversity was as high as in summer (Fig. 30B). The assemblage was dominated by species preferring cold, mixed and enriched conditions (C), such as A. formosa (Bacillariophyceae) (Tab. 5). Shortly before ice-break-up Synedra acus (Bacillariophyceae) and Gymnodinium baicalensis (Pyrrophyta) became dominant (Tab. 5). The former prevail in vernal mixed conditions (D), but, in contrast to larger Bacillariophyceae, such as Cyclotella, they might not be sensitive to stratification. Gymnodinium species are mostly sensitive to mixing and dominated in stratified epilimnia (L_M; Tab. 5). Both groups indicate conditions of nutrient availability. The bacillariophycean maximum expected after ice-break up 2003 was probably missed, because sampling had to be stopped during ice-break up until a time when conditions permitted ship access to the sampling location.

Seasonal dynamics of phytoplankton pigments: The biovolume maximum did not correspond to the chlorophyll and carotenoid one (Fig. 30B). Shifts towards Chla-rich phytoplankton groups prevailed in summer, while large and Chla-poor cells dominated in winter and spring. Thus, the Chla vs. nano- and microphytoplankton biovolume ratios were < 6 nmol mm^-3 in spring (May-June) but 22 nmol mm^-3 on average in summer (July-September) and 8.5 nmol mm^-3 on average under the ice (February-April).

After the ice break-up at the end of April 2002, Chla showed a first maximum in Mid-May (2 nmol L^-1), decreasing strongly from the end of May to the beginning of June (0.9 nmol L^-1; Fig. 30B). The sum of carotenoids was about 90 % of Chla (based on molar ratios and 60 % based on weight ratios) during the time of spring mixing. Secchi depth was low (10 m) at the Chla maximum but increased while chlorophylls and carotenoids decreased (Fig. 30B).

Tab. 5. As Tab. 2, but for seasonal variation.

Fig. 30. Seasonal monitoring during May 2002 to June 2003 (at the long-term monitoring station of SRIB 2.8 offshore from Bolshye Koti, South Basin) of temperature, Secchi depth, biovolumes and selected pigments.(A) temperature at 5 m water depth and Chla integrated over 40 m water depth (the suggested euphotic zone) from January 2001 to December 2003. The highlighted area designate the period of the intensive monitoring, detailed in section (B): Intensive monitoring from May 2002 to June 2003. On the left side: Secchi depth, temperature, total APP biovolume (epifluorescence microscopic spot counts + light microscopic estimation), total nano- and microphytoplankton (NMP biovolume and cell abundance), reciprocal Shannon-Wiener diversity index, and composition of the phytoplankton community based on their share to total Chla (see text for calculations). On the right side: total Chla as well as marker pigments and biovolumes of the respective phytoplankton groups.

The clear-water phase gave way to another Chla maximum in Mid-July (2 nmol L^-1; Fig. 30B) and the sum of carotenoids increased up to 160 % of Chla. The Chla concentration was high during summer, when temperatures rose to a maximum of 16° C and Secchi depth was less than 10 m (Fig. 30B). The carotenoids were about 100 % of Chla during the summer stratification. From the end of summer to autumn the temperature and Chla concentration and percentage carotenoids decreased while the Secchi depth increased (Fig. 30B). From February to March 2003 the Chla concentration increased under the ice and Secchi depth was decreased (Fig. 30B). The sum of carotenoids was low (70 % of Chla) under the ice.

The distinct pigment changes likely reflected the variations in the phytoplankton groups. Similar to the biovolume of the Bacillariophyceae, all related pigments, such as fucoxanthin, Chlc and diadinoxanthin showed maxima in May after ice-break up (Fig. 30B). In summer, Bacillariophyceae biovolume decreased while the related pigments decreased to a lesser extent, marking shifts towards smaller, pigment-rich Bacillariophyceae or towards Chrysophyceae (which also contain those pigments). Microscopic counts confirmed the shift towards Chrysophyceae (flagellates, Tab. 5).

Chlorophyta and its related pigments also showed some discrepancies (Fig. 30B). While the Chlorophyta biovolume was low in spring 2002 and 2003, lutein and Chlb were high (Fig. 30B). In summer, both pigments increased faster than the biovolume, suggesting the influence of chlorophycean APP with high pigment vs. biovolume ratios (Fig. 30B). The summer maxima and the winter decrease of zeaxanthin and ß-carotene correspond to the course of the cyanobacterial APP (Fig. 30B). Except cyanobacterial ones, pigment concentrations were also high under the ice and pigments of Cryptophyta (alloxanthin) and Pyrrophyta (peridinin) even showed maxima.

Estimation of phytoplankton composition using marker pigments: As has been shown for the regional variability, the contributions of distinct chemotaxonomic groups to the total Chla can be calculated using the marker pigments. Based on the factors listed in Tab. 3A for the whole lake average, contributions have been calculated for the period of the intense monitoring (Fig. 30B): Bacillariophyceae+Chrysophyceae had highest (up to 90 %) proportions of total Chla in spring, autumn and winter, Chlorophyta and Eustigmatophyceae in early summer, and cyanobacterial APP in summer; Cryptophyta contributed low proportions to total Chla throughout the year.

The calculated Chla matched the measured one with a median error of 2 % in spring (May-June), of 13 % in summer (July-September), but of 38 % in fall (October-December) and of 31 % under the ice (February-March). It matched the measured Chla in June 2003 again with an error of only 5 %. As the applicability of the factors differ with the phytoplankton composition, the regional factors, accounting for these heterogeneities (Tab. 3B), were also tested. The best fit for the summer community was reached using the Selenga Delta factors (Tab. 3B) with a median error of only 10 % between calculated and measured Chla. Best fits for autumn (Selenga Delta factor) and under ice (Centre factor) communities, in contrast, implied errors greater than 25 % and thus, these communities were different from all others.

Driving factors for seasonal succession: Canonical correlation analysis (CCA) showed that stratification and temperature dominated the first canonical variate (Tab. 4). The percentage contribution of Chlorophyta and cyanobacterial APP to total Chla were highly correlated to this first variate, as well as, although negatively, Cryptophyta. The second variate was not significant. Taken together the CCA for regional and seasonal variations indicated that temperature and stratification were of major importance for the phytoplankton development in Lake Baikal and potentially influence predominantly small cells such as cyanobacterial picoplankton.

Through the upper water column (water samples), the deeper water column (sediment traps) and also within the water to sediment interface (top sediment slice) major lipophilic photosynthetic pigments known from freshwater samples were detected by the HPLC-aided analysis (Tab. 6). Characteristic fluorescence and absorption chromatograms are shown in Fig. 31 for the water samples (Fig. 31A), the 40 m trap (Fig. 31B), the 1400 m trap (Fig. 31C) and the top sediment slice (Fig. 31D).

N°

Pigment

40 m

trap

1400 m

trap

(µmol m^-2 month^-1)

1

Chlorophyll a

12.1

2.89

2

Chlorophyll b

0.53

0.11

3

Chlorophyll c

2.63

0.41

4

Chlorophyllide a

2.13

0.68

5

Pheophorbide a

40.8

6.74

6

Pheophytin a

6.5

7

Pheophytin b

0.075

8

Pyropheophytin a

0.32

9

Fucoxanthin

11.6

1.55

10

Violaxanthin

0.05

-

11

Diadinoxanthin

0.57

0.26

12

Diatoxanthin

1.44

2.00

13

Alloxanthin

0.84

0.04

14

Lutein

0.40

0.06

15

Zeaxanthin

1.38

0.1

16

ß-Carotene

0.39

0.007

Tab. 6. Major lipophilic photosynthetic pigment fluxes into the 40 m trap and into the trap at the lake bottom. Values for the lake bottom (1400 m) were extrapolated from the curve fittings shown in Fig. 32 and Fig. 33 for all pigments with exponential decrease. For pheophytin a, pyropheophytin a and pheophytin b, which did not show a significant decrease with depth, 40 m and lake bottom values were calculated as averages of all traps. Chla included allo-, epimers and other derivates; pheophorbide a included all pheophorbide a derivates and pheophytin a all pheophytin a derivates.

Pheopigments, degradation products of chlorophylls, were found only in sedimented material (Fig. 31, Tab. 6). Four different pheophorpide a -like pigments were found (Fig. 31). The occurrence of those pheophorpide and pyropheophorbide derivates was not correlated with the depth. Thus, all of them are grouped as “pheophorpide a” in the following text. All pheophytin a -like pigments were also grouped to “pheophytin a” in the text below. Pyropheophytin a could be clearly differentiated from pheophytin a and is discussed separately (Fig. 31, Tab. 6). The chromatograms of pigment extracts of the sediment traps and sediment slices also contained several unidentifiable components (unnumbered peaks in Fig. 31). They were strongly degraded pigments, which were not included in our quantitative comparisons.

Fig. 31. Typical fluorescence (410 nm / 670 nm excitation / detection wavelength, left side) and absorption (440 nm) chromatograms (right side) of (A) the water column, (B) the 40 m trap, (C) the 1400 m trap and (D) the top sediment slice below the mooring. Numbering of the peaks confers to pigments listed in Tab 6.

Transfer fluxes: During 16 months of deployment 239 g m^-2 dry matter (DM) settled in the 40 m trap, with an average flux of 14.9 g m^-2 month^-1 (Tab. 7, Fig. 32). The content of total organic carbon (TOC) was 21.9 % at that depth and that of total nitrogen (TN) 1.6 % (Tab. 7, Fig. 32). The resulting atomic ratio of TOC vs. TN (C/N) of 15 indicated that the sedimented material resulted from the autochthonous production by suspended phytoplankton and that terrigenous input is likely to be negligible at that site. The amount of pigments gathered during the 16 months deployment in the 40 m trap was 193.1 µmol m^-2 for Chla and 797 µmol m^-2 for chlorophyllide a+pheopigment a (Tab. 6). It is worth noting that the replicate samples of the 40 m trap deviated strongly (coefficient of variation: 60.5%), whereas the coefficients of variation for the replicate samples in the traps below varied from 2.5% to 15.5%. Pheophorbide a was the most prominent degradation product in the 40 m trap (Tab. 6, Fig. 32). With respect to TOC (total organic carbon) only 3.9 µmol g^-1 Chla but 9.47 µmol g^-1 pheophorbide a were found (Tab. 8). The lowest ratio was found for pyropheophytin a/ TOC (Tab. 8).

* Müller et al. 2005

40 m trap

1400 m trap

dry weight (g m^-2 month^-1)

14.9

9.49

TOC (%)

21.87

7.14

TN (%)

1.60

0.76

C/N (mol mol^-1)

14.80

9.91

Tab. 7. Sedimentation and accumula-tion rates of the dry matter, total nitrogen and atomic C/N ratio in the 40 m trap and at the lake bottom. Values for the lake bottom (1400 m) were extrapolated from the curve fittings shown in Fig. 32 (see also Appendix C - Tab. 2. )

40 m trap

1400 m trap

µmol g ^-1

chlorophyll a/ TOC

4.1

4.38

chlorophyllide a/ TOC

1.14

pheopheorbide a/ TOC

9.87

pheophytin a/ TOC

2.25

8.12

pyropheophytin a/ TOC

0.08

0.42

Tab. 8. Ratios of Chla and its degradation products per organic carbon in the 40 m trap and at the lake bottom. Values for the lake bottom were extrapolated from the curve fittings (Fig. 34, Appendix C - Tab. 4.). Data for chlorophyllide a/ TOC and pheophorbide a/ TOC, which did not show significant depth trends, were calculated as averages of all traps

In the 40 m trap fucoxanthin was the dominant carotenoid (Tab. 6, Fig. 33). Other pigments of Bacillariophyceae+Chrysophyceae (Chlc, diadinoxanthin and diatoxanthin) as well as the cyanobacterial zeaxanthin also showed high sedimentation rates, whereas the chlorophyte Chlb and lutein as well as the cryptophyte alloxanthin, sedimented only in low amounts (Tab. 6, Fig. 33).

A principal component analysis (PCA) including the dry matter, TOC and TN as well as all pigments revealed that three components controlled 90.7 % of the variance. The first component (65.5 %) included the depth and controlled dry matter, TOC and TN. The first component controlled also the labile pigments Chla, b and c as well as chlorophyllide a and pheophorbide a and also all carotenoids. The second and third components comprised the more stable pigments pheophytin a, pyropheophytin a and pheophytin b.

The sedimentation to the lake bottom showed a power regression for dry matter, but two-exponential or two first order independent decay regressions were apparent for TOC, TN, Chla, b, c, chlorophyllide a, pheophorbide a and most carotenoids (Fig. 32, Fig. 33, Appendix C - Tab. 2, Appendix C - Tab. 3). In contrast, for the more stable chlorophyll degradation products, pheophytin a, pyropheophytin a and pheophytin b, none of those regression models fitted accurately (Fig. 32). The composite character of the regressions (Appendix C - Tab. 2, Appendix C - Tab. 3) indicated that the degradation passed through two different phases, triggered by different factors.

Fig. 32. Vertical profiles of settling particles in the water column, showing (A) total mass fluxes (dw, g m^-2), (B) organic carbon (C %), (C) total nitrogen (N %) and (D) atomic C/N ratios as well as vertical profiles of Chla, and its degradation products (µmol m^-2). The traps were deployed for about 16 months. The respective regression equation and their coefficient of determinations (r²) are reported in Appendix C - Tab. 2, and Appendix C - Tab. 3. Abbreviations: Chlida – chlorophyllid a, Phbida – pheophorbide a, Pha – pheophytin a, PyroPha – pyropheophytin a.

Fig. 33. Depth profiles of marker pigments from Bacillariophyceae+Chrysophyceae (A), Chlorophyta (B), cyanobacterial picoplankton (C), Eustigmatophyceae and Cryptophyta (D). The traps were deployed for about 16 months. The respective regression equations and their coefficients of determination (r²) are reported inAppendix C - Tab. 3. Abbreviations: Allo – alloxanthin, ß-car – ß-carotene, Diato – diatoxanthin, Fuco – fucoxanthin, Phb – pheophytin b, Viola – violaxanthin, Zea – zeaxanthin.

The first degradation phase occurred within the upper 250 m and was much stronger than the second. Below 250 – 300 m the degradation became visibly lowered. Calculations of simple exponential or decay models for those pigments resulted in much lower coefficients of determinations or insignificant models. Therefore, one simple phase cannot describe the pigment degradation with depth. The functions (Appendix C - Tab. 2 and Appendix C - Tab. 3) should allow reconstructions of initially settled pigments from trap or sediment data in further investigations. The models are nevertheless preliminary, as they do not take into account different degradation extents within the traps.

Different degradation patterns were revealed when chlorophylls and its degradation products were referred to TOC (Fig. 34) instead of area references (Fig. 32). The Chla/ TOC ratio decreased with depth, indicating that organic carbon is more slowly degraded than Chla (Fig. 34, Appendix C - Tab. 4), whereas the pheophytin a/ TOC ratio and the pyropheophytin a/ TOC ratio increased with the depth, indicating the formation of pheophytin and pyropheophytin with depth (Fig. 34, Appendix C - Tab. 4). Best fits for the chlorophyllide a/ TOC ratio and pheophorbide a/ TOC ratio vs. depth were also linear regression models, but they were not significant (Fig. 34).

Fig. 34. Depth profiles of Chla/ TOC ratio (Chla/C), chlorophyllide a/ TOC ratio (Chlida/C), pheophorbide a/ TOC ratio (Phbida/C), pheophytin a/ TOC ratio (Pha/C) and pyropheophytin a/ TOC (PyroPha/C). The respective regression equations and their coefficient of determination (r²) are reported in Appendix C - Tab. >4.

Considering that the 40 m trap was collecting 100 % of the particles that settled out of the euphotic zone, 100 % dry matter would be 14.9 mg m^-2 month^-1. At 100 m 10 mg m^-2 month^-1 dry matter were collected which accounts for 67 %. At the lake bottom an estimate of 64 % of dry matter settled down. The estimated percentages of TOC and TN that reached the lake bottom were, in contrast, only 33 % and 48 %, respectively (Tab. 7, Fig. 32). The organic compounds were obviously more strongly degraded than the non-organic fraction (mainly siliceous valves of diatoms) of the settled material. The loss during the sedimentation was, however, even stronger for most of the pigments. Only 24 % of Chla reached the lake bottom and 21 % of Chlb (Tab. 6, Fig. 33). The distinct degradation products of Chla showed different losses. About 100 % of the pheophytin a reached the lake bottom, but only 16 % of pheophorbide a.

Composition of the settling material: Fucoxanthin, Chlb and zeaxanthin were used to estimate the contribution of Bacillariophyceae+Chrysophyceae, Chlorophyta and cyanobacterial picoplankton to total Chla. According to the factors established in Fietz and Nicklisch (2004, Appendix A-Table 3), the Chla content in the 40 m trap was composed of 87 % Bacillariophyceae+Chrysophyceae, 11 % Chlorophyta and 2 % cyanobacterial picoplankton (Fig. 35). The percentage contribution did not change with the water depth, as the same composition was found in the deepest traps (Fig. 35).

Fig. 35. Contribution to total Chla of the dominant phytoplankton groups in the sediment traps and in the surface sediment. Estimates based on the marker pigment vs. Chla ratios calculated by multiple linear regession (cf. Fietz and Nicklisch 2004, Appendix A for calculations). Abbreviations: BacChrys – Bacillariophyceae+Chrysophyceae, Chloro – Chlorophyceae, Cyano – cyanobacterial picoplankton.

In all traps the calculated Chla concentration – based on the marker pigments – was much higher compared to the measured Chla concentration. On average the calculated value of the Chla concentration was 157 % of the measured value. Pheopigment a, which results from grazing and photooxidation, were not added to Chla because carotenoid degradation products, resulting from the same processes, were also not added to the marker pigments. Adding chlorophyllide a that occurs in senescent cells due to enzymatic lyses, which does not affect carotenoids (Jeffrey et al. 1997), the average calculated Chla+chlorophyllide a concentration was 121 % the measured value. This overestimation could result from an unusual high carotenoid or Chlb/Chla ratio of specific settling taxa or from a higher degradation rate of Chla than carotenoids or Chlb. The second assumption is likely.

A conclusion concerning the source of pheopigment a, as has been shown for Chla, is limited, because marker pigments which underwent a similar degradation then pheopigments are missed. The only share which can be calculated is for Chlorophyta assuming that the ratio of pheophytin a to pheophytin b represents the former ratio of Chla to Chlb in the settling material. Then, the factor for the Chla/Chlb relationship (cf. Fietz and Nicklisch 2004, Appendix A-Table 3), can be adopted for the pheophytin a/ pheophytin b relationship. According to this approach 2.6 % of pheophytin a should originate from Chlorophyta.

Fluxes and ratios in the top traps: Two mooring strings, containing 9 traps in the North and 15 in the South, were deployed during two ensuing periods (2001-2002 and 2002-2003). Annual fluxes to the uppermost trap varied strongly between both investigated deployment periods (2001-2002 and 2002-2003) and between the two investigated sites (South and North; Tab. 9A). At both sites (South and North) the Chlas (Chla + chlorophyllide a + pheopigment a) flux was much lower during 2002-2003 than during 2001-2002 (Tab. 9A). The dry matter (DM) and TOC fluxes were lower during the second period as well (Tab. 9A). The Chlas and TOC fluxes as well as the Chlas/TOC ratio were higher in the South than in the North in 2001-2002, but lower in 2002-2003 (Tab. 9A).

The fluxes of Chlbs, Chlc, fucoxanthin (Tab. 9A) and other carotenoids were lower during the second deployment period in both basins similar to the Chlas flux (whereby for the North the 300 m trap has to be considered when comparing both years, because the 50 m trap was lost during the first deployment). The Chlbs/Chlas ratio was much lower (1-3 %), than the Chlc/Chlas and fucoxanthin/Chlas ratios (42-48 % and 18.5-20 %, respectively) in the South and similar results were found for the North (cf. Tab. 9A). These ratios indicated a much greater proportion of the Bacillariophyceae+ Chrysophyceae (indicated by Chlc and fucoxanthin) flux into the top traps than of Chlorophyta (indicated by Chlb) in both years and at both sites.

Fluxes and ratios in the bottom traps: At both sites the Chlas fluxes were significantly lower during the second deployment than during the first (Tab. 9B). Furthermore, the Chlas fluxes, which reached the lake bottom, were significantly higher in the South than in the North. Compared to the top traps, the Chlas fluxes which reached the lake bottom were 80-95 % lower for the South mooring 2001-2002, and both North moorings, but higher in the South mooring 2002-2003 (cf. Tab. 9A,B).

The fluxes of Chlbs, Chlc, fucoxanthin (Tab. 9B) and other carotenoids as well as of dry matter and TOC (Tab. 9B) were (similar to the Chlas fluxes) also significantly higher in the South than in the North. No significant differences were found for the C/N ratios (Tab. 9A,B).

* Müller et al.( 2005); ** Mackay et al. (1998), core ´baik29´
Note: In the South during both deployments the top trap was exposed at 40 m. In the North the top trap was exposed at 300 m water depth in 2001-2002 and at 50 m in 2002-2003. Therefore, in the North (second deployment) the fluxes and ratios at 50 m were given and those found in 300 m were added in parenthesis. As the fluxes and ratios in the deepest traps varied only little, the lowest four traps were combined as "bottom traps" and median was calculated. They were deployed between 1100 and 1400 m in the South and between 650 and 900 m in the North. Within the bottom traps superscript S1, S2, N1, and N2 indicate significant differences (at 90 %, Mann-Whitney-U-test) to South 2001-2002 (S1), South 2002-2003 (S2), North 2001-2002 (N1), and North 2002-2003 (N2).

site

South

North

period

 

2001-2002

2002-2003

2001-2002

2002-2003

A) top traps

at depth

40 m

40 m

300 m

50 m (300 m)

DM

(g m^-2 yr^-1)

174.5

39.9

127.4

30.7 (19.7)

TOC

(g m^-2 yr^-1)

38.2

5.0

10.0

9.0 (2.5)

C/N

atomic ratio

14.8

12.4

11.8

9.9 (12.3)

Chlas

(µmol m^-2 yr^-1)

752.0

43.8

179.6

146.3 (70.8)

Chlas/DM

(µmol g^-1)

4.3

1.1

1.4

4.7 (3.0)

Chlas/TOC

(µmol g^-1)

21.0

8.7

18.0

16.2 (28.3)

Chlbs

(µmol m^-2 yr^-1)

7.2

1.4

2.0

0.77 (1.0)

Chlc

(µmol m^-2 yr^-1)

31.5

2.1

1.9

6.1 (0.03)

Fucoxanthin

(µmol m^-2 yr^-1)

139.2

8.7

12.9

22.3 (0.28)

B) bottom traps

at depth

1.1 - 1.4 km

1.1 - 1.4 km

0.68 - 0.9 km

0.68 - 0.9 km

DM

(g m^-2 yr^-1)

117.8 ^N2

141.0^{N1, N2}

109.3^{S2, N2}

18.7 ^{S1, S2, N1}

TOC

(g m^-2 yr^-1)

9.2^{N1, N2}

12.1 ^{N1, N2}

5.3 ^{S1, S2, N2}

1.7 ^{S1, S2, N1}

C/N

atomic ratio

11.2

13.0

11.0

11.2

Chlas

(µmol m^-2 yr^-1)

203.9 ^{S2, N1, N2}

81.2 ^{S1, N1, N2}

30.2 ^{S1, S2, N2}

6.3 ^{S1, S2, N1}

Chlas/DM

(µmol g^-1)

1.55 ^{S2, N1, N2}

0.64^{S1, N1, N2}

0.26 ^{S1, S2}

0.32 ^{S1, S2}

Chlas/TOC

(µmol g^-1)

20.4^{S2, N1, N2}

6.5^{S1, N2}

5.5 ^S1

3.9 ^{S1, S2}

Chlbs

(µmol m^-2 yr^-1)

2.6^{N1, N2}

2.5^{N1, N2}

0.82^{S1, S2, N2}

0.40 ^{S1, S2, N1}

Chlc

(µmol m^-2 yr^-1)

5.8 ^{N1, N2}

0.93^{N1, N2}

0.26 ^{S1, S2, N2}

0.03 ^{S1, S2, N1}

Fucoxanthin

(µmol m^-2 yr^-1)

22.3 ^{N1, N2}

5.6^{N1, N2}

0.37 ^{S1, S2, N2}

0.17 ^{S1, S2, N1}

C) core tops

section

0-1 cm

0-1 cm

time span

yr

~7*

~7**

DM

(g m^-2 yr^-1)

89*

90**

TOC

(g m^-2 yr^-1)

3.4

3.5

C/N

atomic ratio

8.5

9.1

Chlas

(µmol m^-2 yr^-1)

9.9

2.3

Chlas/DM

(µmol g^-1)

0.11

0.03

Chlas/TOC

(µmol g^-1)

2.92

0.66

Chlbs

(µmol m^-2 yr^-1)

0.54

0.21

Chlc

(µmol m^-2 yr^-1)

0.17

0.06

Fucoxanthin

(µmol m^-2 yr^-1)

 

1.14

 

-

Tab. 9. Fluxes and ratios of dry matter, organic carbon and Chlas in (A) the top traps and (B) the bottom traps of the four investigated moorings, and (C) in the top of cores from below the moorings.