Polar Biology

, Volume 31, Issue 10, pp 1191–1203

Natural succession of macroalgal-dominated epibenthic assemblages at different water depths and after transplantation from deep to shallow water on Spitsbergen

Authors

    • Section Seaweed BiologyAlfred Wegener Institute for Polar and Marine Research
    • Centre for Tropical Marine Ecology
  • Markus Molis
    • Section Seaweed BiologyBiologische Anstalt Helgoland, Alfred Wegener Institute for Polar and Marine Research
  • Christian Wiencke
    • Section Seaweed BiologyAlfred Wegener Institute for Polar and Marine Research
  • Nelson Valdivia
    • Section Seaweed BiologyBiologische Anstalt Helgoland, Alfred Wegener Institute for Polar and Marine Research
  • Annelise S. Chapman
    • Biology DepartmentDalhousie University
Original Paper

DOI: 10.1007/s00300-008-0458-4

Cite this article as:
Fricke, A., Molis, M., Wiencke, C. et al. Polar Biol (2008) 31: 1191. doi:10.1007/s00300-008-0458-4

Abstract

In the current study, we investigated the primary succession of seaweeds over different time periods at different water depths. Furthermore, we followed the succession of field-grown benthic communities of different successional age, developing on ceramic tiles, prior to and after transplantation from 8 to 0.5 m water depth. The transplantation simulated changes associated with the break up of sea-ice cover, e.g. light regime or wave exposure. For this purpose, we transplanted 12 and 21-month old communities, grown at 8 m water depth, together with a set of sterile tiles, onto rafts, floating in 0.5 m water depth. Our results describe for the first time the succession of macroalgal communities in the Arctic and give important insights into the effect of disturbance of differently aged communities. The primary succession at 0.5 m water depth was mainly driven by Bacillariophyta and filamentous green algae like Urospora sp. and Ulothrix implexa. Twelve-month old communities at 8 m water depth are dominated by members of the Ectocarpales (Phaeophyceae), like Pylaiella littoralis, P. varia, and Ectocarpus siliculosus and the green alga U. implexa, whereas the 21-month old community showed a higher cover of the green algal class Ulvophyceae and sessile invertebrates. After transplantation to near surface conditions, species composition of the communities changed, but this effect was differently strong between communities of different age.

Keywords

ArcticDiversityMacroalgalRecruitmentSublittoralSuccessionHard bottomCommunityTransplantation

Introduction

Arctic seaweeds are strongly affected by seasonal changes of sea-ice cover, water turbidity, and extended periods of darkness, which can last up to 4 months (Weykam et al. 1997; Bischof et al. 2002; Aguilera et al. 2002; Wiencke et al. 2007). To cope with these severe environmental conditions, polar seaweeds have developed physiological adaptations. They predominately grow and reproduce in late winter and spring. Their saturation points for photosynthesis and compensation points for growth are low, enabling them to occur over a broad depth range (Kirst and Wiencke 1995; Weykam et al. 1996; Wiencke et al. 2007).

Polar regions, especially the Arctic, will be strongly affected by global warming (Hassol 2005; Piepenburg 2005; Svendsen et al. 2002), which may also alter the underwater radiation regime either through reduction in sea-ice cover (in terms of both temporal and spatial extent) or through inflow of turbid melt water. This may additionally reduce the salinity of surface waters. Moreover, Arctic regions are affected by enhanced levels of damaging ultraviolet-B radiation (UVB, 280–320 nm) due to stratospheric ozone depletion (Kerr and McElroy 1993; Hassol 2005). Any abrupt change in the underwater light regime is bound to affect photosynthetically active primary producers the most. Especially the early developmental stages, like spores and propagules, are very sensitive to light stress, specifically ultraviolet (UV) radiation. Yet their shade-adaptation allows them to grow under the adult canopy (Coelho et al. 2000; Roleda 2006).

To date, most studies concerning Arctic seaweeds have addressed individuals or populations, investigating taxonomic, physiological, and autecological aspects. Only very little is known about ecological responses to different light regimes at higher levels of organisation, i.e. communities (Lotze et al. 2002; Molis and Wahl 2004; Dobretsov et al. 2005) and, even more fundamentally, about benthic community succession at higher latitudes (Zacher et al. 2007; Wahl et al. 2004). More information on species succession is required, to better understand the magnitude and direction of environmental impacts, like disturbance, grazing, or climate change, on the composition, diversity, and ecosystem function of benthic communities in the Arctic. Gathering such basic information on community ecology is especially relevant for communities at polar regions, as recruitment, reproduction, and growth appears much slower at Arctic than at temperate habitats (Dunton et al. 1982; Newell et al. 1998; Barnes and Conlan 2007), resulting in several fold longer recovery periods after perturbations (Beuchel et al. 2006). Benthic organisms from Spitsbergen have been shown to respond to fluctuations in regional climate patterns (e.g. McMahon et al. 2006). At Kongsfjorden, seasonal changes in salinity and water transparency, due to the melting of pack ice and glaciers during spring and summer, accompanied by strong variation in the light regime, will have profound effects on the species succession of macrobenthic assemblages.

To date, almost no knowledge on primary and secondary succession of seaweed communities in the Arctic is available. For this purpose we constructed freely floating rafts and fixed sterile ceramic tiles on them to study primary succession. To study the effect of environmental change as it may be associated with the break-up of sea-ice, field-grown communities of different successional age that developed on ceramic tiles were transplanted from 8 to 0.5 m. Our study gives important new information on succession of macroalgal-dominated benthic assemblages in the Arctic.

Materials and methods

Study area

The experiment was performed in Kongsfjorden (78°N, 11°E), an Arctic fjord, located on the north-western coast of Spitsbergen (Svalbard). A summary of the physical environment is given by Svendsen et al. (2002) and Hanelt et al. (2001). Approximately 70 macroalgal species have been recorded from the Svalbard region to date (Weslawski et al. 1993; Vinogradova 1995), with only very few endemic Arctic species. Because of extreme environmental stresses, i.e. sea-ice formation and iceberg scouring, seaweeds occur almost entirely in the subtidal. On Spitsbergen, the upper subtidal zone (to 2.5 m depth) is characterised by annual or pseudoperennial species which can survive the winter as microscopic stages or rhizoidal cushions (Hop et al. 2002; Wiencke et al. 2007). Concerning the sessile invertebrates, more than 450 taxa of invertebrates have been recorded on hard substrata in Kongsfjorden (Hop et al. 2002), whereas about 100 species of motile and sessile invertebrates are recorded from the study area as associated with marine macroalgae (Lippert et al. 2001).

Experimental site

Our study was conducted close to the Ny Ålesund harbour (78°55′N, 11°56′E). Although the Ny Ålesund International Research and Monitoring Facility is intensively used for research, we know from many years of observation at remote sites in the fjord that the study area still represents the general situation along the fjords coastlines. The colonization site for the ceramic tiles used as artificial substrate was located at the old pier, adjacent to the Old Power Station. The transplant experiments were performed in approximately 500 m distance, just east of the new pier. At both sites, the substratum consists of soft sediments, interspersed with gravel, rocks, and remnants of old concrete structure. The subtidal vegetation at the colonization site is dominated by kelps (Alaria esculenta, Laminaria digitata, Saccharina latissima, and Saccorhiza dermatodea) and various smaller undergrowth algae (Palmaria palmata and Dictyosiphon foeniculaceus).

Measurements of abiotic factors

Incident photosynthetically active radiation (PAR) was continuously measured throughout the experiment by a LI-COR™ data logger (LI-1000, Li-Cor, Lincoln, USA) equipped with LICOR 190 SA quantum sensor (cosine corrected) that was installed on a permanent structure in the village of Ny-Ålesund at approximately 4 m height. Instruments were checked and data were downloaded weekly.

Additionally, PAR, ultraviolet A (UV-A), and ultraviolet B (UV-B) radiation were measured at the study site at 0.5 and 5 m water depth on seven dates during the experiment between 11:00 a.m. and 2:00 p.m., using a quantum meter LI-COR data logger (LI-250A, Li-Cor, Lincoln, USA) equipped with an underwater PAR sensor (LI-192) and a spectroradiometer Ramses (ACC-UV, TriOs, Germany) with a hyperspectral UV-A/UV-B (280–571 nm) sensor. Diffuse vertical attenuation coefficients of downward irradiance (Kd) were determined using the following formula (after Kirk 1994):
$$ {\text{Kd}} = { \ln }\left( {{{{\text{Ed}}_{(z{\text{2}})} } \mathord{\left/ {\vphantom {{{\text{Ed}}_{(z{\text{2}})} } {{\text{Ed}}_{(z{\text{1}})} }}} \right. \kern-\nulldelimiterspace} {{\text{Ed}}_{(z{\text{1}})} }}} \right) \times \left( {z_{\text{1}} - z_{\text{2}} } \right)^{ - {\text{1}}} $$
where Ed(z1) and Ed(z2) are the respective irradiances at z1 = 0.5 m and z2 = 5 m water depth.

Parallel to the underwater light measurements, water turbidity was determined, using a Secchi disk (Hydrobios, EU Norm EN27027). Surface water temperature and salinity were determined with a digital thermometer and a handheld refractometer (Winopal, Germany), respectively.

Experimental design and set-up

Using a randomized block design, the experiment examined community composition, biomass, and diversity in relation to community succession and sudden changes of environmental conditions. A total of five wooden rafts (blocks), 420 × 47 cm, floating at a depth of 0.5 m, were positioned in a row, with an individual distance of approximately 10 m. Rafts were able to move freely around their longitudinal axis, allowing for flexibility in tidal, wind, and water motion, as well as avoiding drifting icebergs. A polyvinylchloride (PVC) panel, 43 × 48 cm, was horizontally fixed to each raft.

Five ceramic tiles were attached reversibly to each PCV panel, using Velcro-tape. For the study of the primary succession (primary succession community, PSC) sterile ceramic tiles (9.6 × 9.6 cm2) were fixed horizontally onto rafts. For the study of secondary succession, we used benthic communities grown on ceramic tiles at a depth of 8 m for 12 (mid succession community, MSC, since May 2005) and 21 months (late succession community, LSC, since August 2004). Ceramic tiles were fixed horizontally onto wooden planks using rubber clamps. Each plank, carrying four LSC and seven MSC tiles, respectively, was attached to a steel wire rack (2 × 2 m), fixed to the ground. On 12 and 13 May 2006, communities of different age (MSC and LSC), were recovered and transplanted after 3 days under laboratory conditions onto rafts floating at a depth of 0.5 m, thus causing a sudden change in environmental conditions, especially in the light regime.

We tested the effects of successional age (fixed factor, three levels, PSC, MSC and LSC), and duration of exposure to “near-surface conditions” (fixed factor, 3 levels, 0 [no-transplant], 4, and 8 weeks) on species composition, diversity (species richness and evenness), and biomass of macrobenthic assemblages. Two PSC and MSC, and only one LSC-tile were fixed onto each raft (Fig. 1). LSC were sampled only once, after 8 weeks, whereas PSC and MSC were sampled twice (after 4 and 8 weeks). In order to have within-block replication and hence be able to quantify the within-block variability of our dependent variables, we arranged PSC and MSC across four blocks each, with a single replication in at least one of the—randomly assigned—blocks (Underwood 1997, Fig. 1). No within-block replication was used for the LSC, because not enough tiles were available.
https://static-content.springer.com/image/art%3A10.1007%2Fs00300-008-0458-4/MediaObjects/300_2008_458_Fig1_HTML.gif
Fig. 1

Position of the tiles on the rafts (I–V); primary succession (P), mid succession (M) and late succession (L) community. Tiles sampled after 4 weeks coloured white, after 8 weeks grey. The striped tile got lost

Sampling of communities

Three tiles of the LSC were recovered and analysed for their species composition in May of 2005, i.e. with an age of 1 year. Species composition and percentage cover of the whole community tile were analysed. On 12 and 13 May 2006, all MSC and LSC community-tiles were recovered of 8 m water depth and stored in the Kings Bay Marine Laboratory for 3 days under low light and ambient water conditions in a flow-through system before deployment on the experimental rafts (Fig. 2). During transport (≤15 min), tiles were stored in opaque closed containers to avoid damage through excessive solar radiation or desiccation. PSC and MSC were destructively sampled after 4 (for each community: n = 5) and 8 weeks (for each community: n = 5) of exposure time to surface conditions, on 12 June and 10 July, 2006, respectively. LSC were only sampled once, i.e. on 10 July 2006 after 8 weeks of exposure, and with one tile being lost (= 4; Fig. 1).
https://static-content.springer.com/image/art%3A10.1007%2Fs00300-008-0458-4/MediaObjects/300_2008_458_Fig2_HTML.jpg
Fig. 2

Experimental set-up and settlement tiles: a Colonisation site, the old pier, b Raft with communities growing on tiles, c Late succession community tile (21-months old; ×12 magnification); microscope images of species grown in various communities: dAcrosiphonia sp. (×400 magnification), eUlothrix implexa (×400 magnification), fUrospora sp. (×400 magnification), gDevaleraea ramentacea (×100 magnification), hPylaiella littoralis (×400 magnification), iDermatocelis laminariae (×200 magnification), jDesmarestia viridis (×200 magnification), kBalanus crenatus (×12 magnification), lObelia dichotoma (×12 magnification)

To prepare samples for analysis, tiles were carefully cleared of sediments in the laboratory, by rinsing them gently with ambient seawater, using a wash bottle. To avoid edge effects, only the central part of the tile (31 cm2, 34% of total tile area) was assessed.

Analysis included estimation of percent cover of (1) Bacillariophyta and of (2) individual macrobenthic taxa (individuals >0.5 mm), which were measured using a stereo microscope (12×; Zeiss, Stemi SV6) with an ocular counting frame (100 subdivisions), covering 1 cm² of the tile. For each community (tile), five randomly placed frame counts (5 cm2 total) were analysed. Percentage cover was estimated at 5% intervals. Due to the three-dimensional community structure, total percentage of all species could exceed 100%. Additionally, for better determination, five sub-samples were taken per counting frame and analysed using a microscope (100, 200, 400 and 630× magnification; Zeiss Axiolab). Species, that occurred only once in the counting frame, i.e. covering ≤1%, or were found only in the sub-samples, were assigned 1% cover. All encountered individuals were identified to the lowest possible taxonomic level. To determine the biomass of a community (dry mass), all organisms were removed from the sampled central part of the tile using razor blades, dried at 60°C to weight constancy, and weighed to the nearest 0.001 g (Mettler AE50).

Using percentage cover data, different biodiversity indices were determined. Species richness (S), which describes the total number of taxa and evenness (\( J = - \sum\nolimits_{i = 1}^s {{ \log }\left[ {p_i } \right]{ }p_i { \log }\left[ S \right]^{ - 1} } ^{ } , \)) where pi is the cover of species i divided by the total cover of S species, and which describes how evenly individuals are distributed across species. Data were analysed using PRIMER™ version5 (Clarke and Gorley 2006). For further species identification additionally sampled tiles were transported to AWI Bremerhaven, Germany and cultivated in a constant-temperature room at 5°C and long-day light conditions (18 h per day).

Statistical analysis

Univariate dependant variables (total cover, biomass, and diversity measures including species richness and evenness) were analysed using analysis of variance (ANOVA). One-factorial ANOVAs were performed to test for significant differences between sampling dates (factor: week) and among communities of different ages but exposed to near-surface conditions for the same time (factor: age). A three-factorial ANOVA was performed to test the effects of age (fixed), week (fixed), and raft (block, random) on the dependent variables. Furthermore species were grouped in Bacillariophyta, Chlorophyta, Phaeophyceae, and Rhodophyta, as well as sessile invertebrates, and were compared by one-way ANOVA, in order to determine their contribution to the different communities. Prior to the analyses, percent cover data were arcsine-transformed. Homogeneity of variances was tested using Cochran’s test, and variances homogenised by log- or square root transformation if necessary. Data remaining heteroscedastic despite transformation were analysed with the nonparametric Kruskal–Wallis test. Tukey’s test of Honest Significant Difference (HSD) or Kruskal Wallis multiple comparisons were used for post-hoc tests. Data were processed using STATISTICA™ software (StatSoft), version 7.1.

Species composition (based on percent cover of taxa) was analysed by analysis of similarity (ANOSIM) using PRIMER™ software, version 5 (Clarke and Gorley 2006). Prior to this analysis mean values of the within-block replication of each community were determined, to provide a balanced data set-up. ANOSIMs were based on Bray–Curtis similarity indices, calculated from percent cover data, previously fourth-root transformed to down-weigh the importance of dominant taxa (i.e. the Bacillariophyta) within samples. SIMPER was used to determine the relative contribution of single taxa responsible for differences in specific species composition among treatments. Non-metric multidimensional scaling (nMDS) was used to produce rank-order similarities among different communities. MDS plots gave two-dimensional representations of ordinations, with stress values of less than 0.1 corresponding to good visual representations with no real prospect of a misinterpretation (Clarke and Warwick 2001).

Results

Environmental conditions

At midnight, surface photon fluence rates were 136 ± 63 μmol m−2 s−1 throughout the study period. The daily maximum value varied strongly depending on the weather conditions. On cloudy days, maximum values were 527 ± 158 μmol m−2 s−1, on sunny days 1,071 ± 161 μmol m−2 s−1. Water transparency strongly decreased in early June but then increased again as indicated by the changing secchi depths (between 9.5 and 2.3 m; Table 1) and the changes in the vertical attenuation coefficient (between 0.45 and 1.0 for PAR; Fig. 3). At the end of the study period water transparency was again very low as indicated especially by the high Kd values. These changes reflect the inflow of turbid meltwater lowering the salinity in the uppermost water layer (0.05 m) from 37 to 19 on 9 June. Sea surface temperature increased from 0.6°C in May to 6.2°C in July (Table 1).
https://static-content.springer.com/image/art%3A10.1007%2Fs00300-008-0458-4/MediaObjects/300_2008_458_Fig3_HTML.gif
Fig. 3

Vertical attenuation coefficient of downward irradiance (Kd) of the photosynthetically active radiation (PAR), ultraviolet-A radiation (UVA) and ultraviolet-B radiation (UVB) during the experimental period, calculated from measurements at 0.5 and 5 m water depth. Measurements were made near the experimental site on seven different dates during the experiment

Table 1

Water temperature [°C] and salinity [g/1,000 g] at different water depth, and secchi depth (Hydrobios, EU Norm EN27027), determined on seven different dates during the experiment between 11:00 a.m. and 2:00 p.m

 

May 23

June 9

June 17

June 21

June 22

July 4

July 19

Water temperature (0.05 m)

0.6

4.7

4.7

5

5.7

6.2

5.8

Salinity (0.05 m)

37

19

29

30

31

35

30

Salinity (0.5 m)

ND

ND

32

31

31

36

31

Secchi depth (m)

9.5

2.3

5.8

5.5

6.5

6.8

2.5

ND not determined

General patterns of species composition of epibenthic assemblages

Overall, we found 16 different taxa, including 9 species and 7 taxa, containing more than one indistinguishable species. Some of these species could be identified after further cultivation in Bremerhaven (Table 2; Fig. 2). Benthic Bacillariophyta were very abundant. A diverse community of single and band forming Bacillariophyta (e.g. Cylindrotheca sp., Gyrosigma sp., Licmophora sp, Fragilaria sp., etc.) was recognised but not identified to species level nor was the abundance of individual species estimated. Especially the LSC constituted mainly of single Navicula-like Bacillariophyta living in branched tubes. The green alga Ulothrix implexa occurred in all communities, whereas the hydroid Obelia dichotoma was only found in MSC, and the serpulid polychaete Circeis spirillum only in LSC. In contrast to the older successional stages (mid and late succession communities) grown at 8 m water depth, the green alga Urospora sp. was only found during primary succession (PSC, Table 2) grown at 0.5 m water depth.
Table 2

Mean (±SD) percentage cover of taxa encountered on settlement tiles of communities of different successional age (PSC, MSC and LSC) prior to and 4 and 8 weeks after transplantation from 8 to 0.5 m water depth

Taxon (with authority)

Species included/successional age

Primary succession (PSC)

Mid succession (MSC)

Late succession (LSC)

4 weeks

8 weeks

0 weeks

4 weeks

8 weeks

9 month

0 weeks

8 weeks

1 Bacillariophyta

Cylindrotheca Rabenhorst 1859, Gyrosigma Hassall 1845, Licmophora C. Agardh, Fragilaria Lyngbye 1819

39.2 ± 27.3

41.0 ± 16.4

64.3 ± 13.8

78.8 ± 9.5

50.2 ± 12.6

++

72.7 ± 4.6

56.3 ± 12.6

Chlorophyta

 

0.6 ± 0.2

36.7 ± 9.6

26.6 ± 7.7

1.6 ± 0.3

4.7 ± 1.1

+

21.3 ± 5.0

12.3 ± 2.9

 2 Acrosiphoniaceae

Acrosiphonia J. Agardh 1846, Spongomorpha Kützing 1843

1.3 ± 2.1

2.2 ± 2.4

0.2 ± 0.4

0.8 ± 1.8

+

6.5 ± 6.7

6.7 ± 5.8

 3 Ulothrix implexa (Kützing) Kützing 1849

 

0.6 ± 0.4

19.8 ± 5.5

18.4 ± 7.7

0.4 ± 0.3

2.8 ± 3.1

?

0.5 ± 0.6

0.4 ± 0.3

 4 Ulothrix flacca (Dillwyn) Thuret in Le Jolis 1863

 

5.9 ± 3.2

0.2 ± 0.4

?

2.3 ± 3.0

1.1 ± 2.0

 5 Ulvophyceae

Ulva Linnaeus 1753, Monostroma Thuret 1854

-

0.2 ± 0.3

0.8 ± 1.3

0.7 ± 1.5

+

12 ± 16.8

4.2 ± 3.7

 6 Urospora sp. Areschoug 1866

 

0.1 ± 0.1

15.6 ± 6.3

0.4 ± 0.9

   

Phaeophyceae

 

22.9 ± 9.2

1.1 ± 0.5

7.9 ± 1.9

++

4.2 ± 0.5

12.0 ± 3.9

 7 Laminariales

Laminaria saccharina (Linnaeus) J.V. Lamouroux, Alaria esculenta (Linnaeus) Greville 1830, Saccorhiza dermatodea (Bachelot de la Pylaie) J.E. Areschoug 1875

0.03 ± 0.1

+

0.3 ± 0.8

 

 8 Chordariaceae

Chordaria flagelliformis (O·F. Müller) C. Agardh 1817, Dictyosiphon Greville 1830, Stictyosiphon sp.Kützing 1843, Litosiphon sp. Harvey 1849, Chorda filum (Linnaeus) Stackhouse 1797, Halosiphon tomentosus (Lyngbye) Jaasund 1957

1.1 ± 1.3

4.4 ± 7.1

+

1.3 ± 1.7

9.9 ± 5.2

 9 Ectocarpales

Pylaiella littoralis (Linnaeus) Kjellman 1872, Pylaiella varia Kjellman 1883, Ectocarpus siliculosus (Dillwyn) Lyngbye 1819

22.7 ± 18.9

3.0 ± 6.7

++

1.5 ± 2.4

 

 10 Desmarestia sp.

Desmarestia aculeata (Linnaeus) J.V. Lamouroux 1813, Desmarestia virids (O·F.Müller) J.V. Lamouroux

0.2 ± 0.4

0.3 ± 0.8

 

 11 Dermatocelis laminariae Rosenvinge 1898

 

0.4 ± 0.4

1.3 ± 1.5

 12 Phaeophyta II

 

0.5 ± 1.1

0.3 ± 0.8

0.8 ± 1.0

Rhodophyta

 13 Devaleraea ramentacea (Linnaeus) Guiry

 

1.5 ± 1.7

0.1 ± 0.1

+

4.6 ± 6.7

0.2 ± 0.2

 Sessile invertebrates

 

6.1 ± 1.8

1.1 ± 0.5

7.9 ± 1.9

1.2 ± 0.5

2.4 ± 0.4

 14 Obelia dichotoma Linnaeus 1758

 

3.6 ± 3.9

1.1 ± 2.1

12 ± 13.1

0.9 ± 1.7

 15 Balanus crenatus Bruguière 1789

 

2.5 ± 2.9

4.8 ± 9.1

-

1.1 ± 1.6

1.0 ± 1.2

 16 Circeis spirillum Linnaeus 1758

 

0.2 ± 0.4

0.5 ± 1.0

Taxa are numbered and subsequent resolution to species level (through growth in culture) are listed in the next column. n = 4–6. Taxa were grouped as Bacillariophyta, Chlorophyta, Phaeophyceae, Rhodophyta, and sessile invertebrates. Data of 9 month old late succession community (9 month LSC), taken on 21/22 May 2005) are based on presence–absence

++ Dominant, + present, - absent, ? not identified

Primary succession at 0.5 m water depth

After 4 weeks of succession, Bacillariophyta and two filamentous green algae (U. implexa and Urospora sp.), recruited onto the ceramic tiles and increased in abundance over succession time. After 8 weeks of exposure, an additional green alga of the family Acrosiphoniaceae was found. Therefore the total taxa number increased from 3 to 4 between weeks 4 and 8 (Table 2). Overall, there was a significant change in species composition within time (ANOSIM: R = 1.0 P = 0.029; Table 3). Species richness (one-way ANOVA: F1,8 = 12.00, = 0.01; Fig. 4), evenness (one-way ANOVA: F1,8 = 160.05, P = 0.00; Fig. 4), and percent cover (one-way ANOVA: F1,8 = 5.5, P = 0.045; Fig. 4) increased significantly in the 8-weeks-old epibenthic assemblages. Surprisingly, no significant change in biomass (one-way ANOVA: F1,8 = 2.57, P = 0.15; Fig. 4) was observed.
https://static-content.springer.com/image/art%3A10.1007%2Fs00300-008-0458-4/MediaObjects/300_2008_458_Fig4_HTML.gif
Fig. 4

Mean species richness (±95% confidence interval), evenness, percentage cover, and biomass (dry weight) of the primary succession communities (12-months old) after 4 and 8 weeks exposure at 0.5 m water depth (n = 5)

Table 3

Results of ANOSIM comparing different communities by testing the effect of the factor exposure to “near-surface conditions” (week, fixed, 3 levels, 0 [no-transplant], 4, and 8 weeks) on species composition, and results of SIMPER (assessment of the contribution >8% of single taxa to among-group dissimilarities) for communities of different successional age (PSC, MSC, and LSC)

 

Week 0–4

Week 0–8

Week 4–8

R

P

R

P

R

P 

Primary succession community (PSC)

NA

NA

1.0

0.029

 C Urospora sp.

  

41%

+

 C Ulothrix implexa

  

31%

+

 C Acrosiphoniaceae

  

19%

+

Mid succession community (MSC)

0.99

0.005

0.82

0.005

0.82

0.26

 P Ectocarpales

23%

19%

 

 C Ulothrix implexa

21%

16%-

 

 C Ulothrix flacca

13%

12%-

 

 I Obelia dichotoma

  

10%

+

 

 P Chordariaceae

  

9%

+

 

Late succession community (LSC)

  

0.27

0.052

 

Contributions are averaged across all significant pair-wise treatment comparisons. Data based on fourth-root transformed percent cover data. The direction of the effect is given as + positive, − negative. C Chlorophyta, P Phaeophyceaea, I sessile invertebrates, NA not applicable because settlement tiles with PSCs at week 0 were empty

Succession at 8 m water depth prior to transplantation

The two different old communities differed significantly in their species composition (ANOSIM: R = 0.86, P = 0.002; Table 4). The 12-month old community (MSC) was dominated by species of the order Ectocarpales (Phaeophyceae) and the green alga U. implexa (Table 2; Fig. 5), whereas the 21-month old (LSC) community was mainly composed of species of the class Ulvophyceae (Chlorophyta) and showed a significant higher cover of sessile invertebrates (ANOVA: F1,7 = 8.49, P = 0.02; Table 4).
https://static-content.springer.com/image/art%3A10.1007%2Fs00300-008-0458-4/MediaObjects/300_2008_458_Fig5_HTML.gif
Fig. 5

Mean percent cover of different taxonomic groups (Bacillariophyta, Chlorophyta, Phaeophyceae, Rhodophyta, and sessile invertebrates) found in mid succession (12-months old) communities and late succession (21-months old) communities after exposure to near-surface conditions for zero (no-transplant), 4 and 8 weeks, respectively (n = 6)

Table 4

Results of ANOSIM comparing different communities by testing the effect of the factor ‘age’ (fixed, 3 levels, primary (PSC), mid (MSC), and late successional community (LSC) on species composition, and results of SIMPER (assessment of the contribution >8% of single taxa to among-group dissimilarities) at three times after exposure to near surface conditions (0, 4, and 8 weeks)

 

PSC and MSC

PSC and LSC

MSC and LSC

R

P

R

P

R

P

0 weeks

NA

NA

0.86

0.002

 C Ulothrix implexa

  

15%

 P Ectocarpales

  

14%

 C Ulvophyceae

  

12%

+

 I Obelia dichotoma

  

9%

+

4 weeks

0.74

0.03

  

 C Chordariaceae

27%

   

 Bacillariophyta

17%

   

 C Ulvophycea

12%

   

 C Ulothrix implexa

10%

   

8 weeks

0.67

0.03

0.907

0.029

0.198

0.23

C Urospora sp.

23%

 

18%

 

P Chordariaceae

16%

 

17%

+

 

I Obelia dichotoma

14%

    

C Ulothrix implexa

13%

 

14%

 

I Balanus crenatus

10%

    

C Ulvophyceae

  

10%

+

 

Contributions are averaged across all significant pair-wise treatment comparisons. Data based on fourth-root transformed percent cover data. The direction of the effect is given as + positive, − negative. C Chlorophyta, P Phaeophyceaea, I sessile invertebrates, NA not applicable because settlement tiles with PSCs at week 0 were empty

In contrast to the primary succession (PSC; Fig. 4), no increase in biodiversity over time was observed in MSC and LSC (Fig. 6). The brown alga Dermatocelis laminariae (Chordariaceae) and the polychaete Cireis spirillum were additional members of the LSC (Table 2). Despite a higher mean number of taxa in LSC, there was no significant difference in species richness between MSC and LSC prior to transplantation (ANOVA: F = 0.53, P = 0.48; Table 4). However, evenness was significantly lower in LSC than in MSC (Kruskal Wallis: H1,10 = 4.33, P = 0.04, Table 4; Fig. 6).
https://static-content.springer.com/image/art%3A10.1007%2Fs00300-008-0458-4/MediaObjects/300_2008_458_Fig6_HTML.gif
Fig. 6

Mean evenness (±95% confidence interval; white columns) and species richness (green columns), of communities grown for 12 (MSC) and 21-month old (LSC) at 8 m water depth, prior to and after 4 and 8 weeks of exposure to near surface conditions at 0.5 m water depth. For the LSC, data for the 9-month old successional stage (qualitatively analysed after 9 month succession) were included in the table. Data based on percentage cover (n = 5). Data for the LSC after 9 month are based on presence–absence data (n = 3)

Changes in community structure after transplantation from deep to shallow water

The transplantation from deep to shallow water caused a significant change in the species composition of the different communities (ANOSIM: R = 0.76, P = 0.001; Table 3). After transplantation a general decrease in percentage cover of the total macrobenthic biota was recognised after 4 weeks (Fig. 5). In the MSC most Phaeophyceae, including the predominating Ectocarpales vanished totally from the tiles after transplantation (Table 2). Also for the Chlorophyta a strong decrease in percentage cover was recognised, which caused significant differences between the communities before (0 weeks) and after transplantation (4 and 8 weeks; ANOVA: F2,13 = 41.02, P = 0.00; Table 4).

After 8 weeks of exposure to near surface conditions, the community recovered and the sessile invertebrates, mainly represented by the hydroid O. dichotoma, started to increase strongly in its percentage cover, therefore the communities significantly differed between 0 and 4 weeks, as well as between 4 and 8 weeks of exposure to near surface conditions (ANOVA: F2,13 = 16.5, P = 0.00; Table 4). Moreover, species richness decreased significantly during the first 4 weeks after transplantation and remained until the end of the experiment significantly lower than prior to transplantation (ANOVA: F2,12 = 8.08, P = 0.01; Fig. 6). Evenness also decreased significantly during the first 4 weeks (Kruskal–Wallis: H2,16 = 9.85, P = 0.001) but returned until week 8 to pre-transplantation values (Table 4; Fig. 6).

Despite missing changes in species richness and evenness in the LSC, the red alga Devalerea ramentacea, strongly decreased in abundance, bleached, and nearly vanished completely and the Phaeophyceae, dominated by the Chordariales, significantly increased in abundance until the end of the study (ANOVA: F1,8 = 6.87, P = 0.03).

In contrast to the changes in the macrobiota, the Bacillariophyta increased in abundance after transplantation but decreased significantly between 4 and 8 week of succession in both communities (MSC: ANOVA: F2,13 = 6.84, P = 0.01; LSC: ANOVA: F1,8 = 6.87, P = 0.03; Table 4).

Prior to transplantation, the species composition of MSC was significantly different to that of LSC (ANOSIM: R = 0.86, P = 0.002, Table 4; Fig. 7). Eight weeks after transplantation, differences in biodiversity and species composition became smaller (Table 4; Fig. 7); however, MSC had a significantly more invertebrate cover than the LSC (ANOVA: F1,7 = 8.49, P = 0.02).
https://static-content.springer.com/image/art%3A10.1007%2Fs00300-008-0458-4/MediaObjects/300_2008_458_Fig7_HTML.gif
Fig. 7

Non-metric multi dimensional scaling (nMDS) plot of mid succession (12-months old) communities (open square) and late succession (21-months old) communities (filled triangle) a before (above, n = 6) and b after (below, n = 4) transplantation and exposure to near-surface conditions for 8 weeks. Data based on non-transformed percentage cover data

Discussion

The primary succession at 0.5 m water depth is mainly driven by the Bacillariophyta, which form dense mats and facilitate the settlement of other species, like Urospora sp. (Chlorophyta). Members of the Ectocarpales (Phaeophyceae) occurred as pioneer species, dominating the communities during the first year of succession. Some species were found in deeper waters than previously described, probably due to the relatively high light conditions predominantly in spring and to a high adaptation potential of these species to the highly dynamic Arctic environment. Yet, sudden changes in the environmental conditions strongly affected the diversity and structure of epibenthic communities. Such changes had detrimental effects on the abundance of most macroalgae, were neutral to Bacillariophyta and beneficial to the barnacle Balanus crenatus and the hydroid O. dichotoma, which seem to be dependent on community age and possibly on favourable environmental conditions.

Primary succession

Most of the early settlers were Bacillariophyta, forming dense mats and dominating the communities over the whole exposure period. Such Bacillariophyta mats appear to precondition the substratum for macroalgal propagules, including provision of UV-free space (Vinebrooke and Leavitt 1999), and may facilitate settlement for following species through their production of extracellular polymers (Lam et al. 2005). In our study, the annual species Urospora sp. and U.implexa quickly monopolized empty substratum. Especially Urospora sp. demonstrates as pioneer species a high colonization potential. The species grows and reproduces aseasonally under a wide range of environmental conditions, and is able to use both high and low photon flux rates efficiently, due to its filamentous thallus structure (Littler 1980; Weykam et al. 1996). Moreover, Urospora sp. appeared to be competitively inferior, as it was absent from older communities.

The Ectocarpales group dominated communities after 1 year of succession. Similarly, Pylaiella littoralis, another member of the Ectocarpales, dominated epibenthic communities in a field study from Nova Scotia, Canada, (Lotze et al. 2002), but only during the first 6 weeks of succession. This difference probably reflects the different time scales for recruitment in temperate versus polar regions.

The current study enlarges our knowledge of the depth distribution of macroalgae in the Arctic in several ways: species that were previously described from shallow water (<5 m), including the green alga U. implexa, U. flacca, Acrosiphonia sp., and the brown alga P. littoralis (Svendsen et al. 2002; Wiencke et al. 2004), grew at 8 m water depth. This reflects the relatively high underwater light conditions characterised be relatively low Kd values (Fig. 3) and high secchi depths (Table 1), at least during some parts of the year, enabling growth of species with high light requirements. In addition, these species might be able to acclimate to low light conditions. According to Weykam et al. (1996), filamentous and foliose species from the intertidal and upper subtidal zones, i.e. the majority of the species found in the sampled communities, are characterised by comparatively high maximum photosynthesis and alpha values, enabling them to use both high and low photon flux rates efficiently.

In accordance with existing ecological models for succession (Connell and Slatyer 1977), changes in biodiversity among communities of different successional age were observed during primary succession. Due to the arrival of new settlers, species richness and evenness increased with successional age. In contrast to the mid (MSC) and late succession community (LSC) no significant difference in species richness was found between them while evenness decreased only slightly in LSC compared to MSC. These findings point to a long-lasting succession with low rate of competitive displacement, where only few species show the tendency of becoming relatively more abundant. Especially the older LSC showed a high variance in evenness, which indicates a high variability in the relative abundance of taxa and points to a high variability in species composition. Biodiversity comparisons must be carried out with caution since we used communities that had been settled at different times (MSC and LSC), rather than observing the same community continually. However, comparing genuine succession stages of the LSC (after 9 and 21 months), supports our findings.

Secondary succession after transplantation to near-surface depth

Communities were clearly altered by transplantation to shallow water. The observed effect was strongest during the first four weeks of exposure to near-surface conditions, resulting in a decrease in species richness and a change in species composition. Furthermore, communities of different age apparently adjusted to the sudden change in light regime in that there was little further shift in species composition after 8 weeks exposure, with only a small increase in sessile invertebrates.

Our transplantation simulated a sudden change in environmental conditions, comparable to sea-ice break-up, causing primarily a change in the underwater photon fluence rate and spectral composition but also in salinity, leading to a change in macroalgal community structure. Such alterations of community composition might be explained by different acclimatisation potentials of the various macrobenthic species (reviewed by Falkowski and LaRoche 1991), as well as species specific recovery processes after disturbances (Sousa 1984; Connell and Keough 1985). To some extent, most algae are capable to acclimate in response to changes in photon fluence rate and spectral composition. But especially taxa with predominant occurrences in relatively deep water (Wiencke et al. 2004), like members of the orders Desmarestiales and Laminariales, and the red alga Devaleraea ramentacea, strongly decreased in abundance or bleached. Species of the brown algal order Ectocarpales, which dominated the communities after 1 year of succession, also disappeared after transplantation.

Our results complement well published data on the susceptibility of macroalgae to changes in the environmental radiation conditions. In their habitat, photosynthesis of the individual species is acclimated or adapted to the specific underwater radiation conditions. Upon exposure to natural sunlight in shallow water, sunlight depleted of UV-B radiation or sunlight depleted of UV-A and UV-B radiation, photosynthesis becomes inhibited to a different degree depending on the radiation regime (Hanelt et al. 1997; Karsten et al. 2001). UV-B radiation has the strongest effect and deep water species are most susceptible. Photosynthesis in species from shallow waters can, however, within limits acclimate to changes in the radiation conditions whereas deep-water species cannot (Bischof et al. 1998, 1999). This does not apply only to macrothalli but also to the unicellular propagation units of macroalgae, as exemplified in kelp zoospores (Roleda et al. 2006a, 2006b, Wiencke et al. 2007, Lüder et al. 2008). Apart from photosynthesis also the DNA is affected by UV-B radiation. On the other hand, there are UV-B protective and repair mechanisms operating which finally determine the sum parametre germination, a factor most relevant for recruitment and succession.

In contrast to macrobenthic taxa, Bacillariophyta seemed to be more tolerant to near-surface conditions, as their abundances did not change in the MSC after 4 weeks of exposure. Similar observations exist from field studies conducted in Antarctica, (Campana et al. 2008; Wulff and Zacher 2008; Zacher et al. 2007). One explanation for a higher tolerance of Bacillariophyta might be the induction of photo-protective substances (e.g. Karsten et al. 2006). Alternatively, shifts in species-composition from species sensitive to UVR and/or high PAR irradiance to more resistant Bacillariophyta species could explain missing transplantation effects. Moreover, indirect effects from protective shading of UVR-tolerant species for UVR-sensitive species were reported from micro-benthic communities (Karsten et al. 1998). In this respect the measured changes in the underwater radiation regime and the decrease in salinity due to inflow of turbid meltwater have to be taken into account. For instance, the hydroid O. dichotoma and the barnacle Balanus crenatus strongly increased in the MSC between 4 and 8 weeks after transplantation, replacing the previously dominating Ectocarpales after their disappearance due to high light stress. Similarly, macroalgal species composition in the LSC was also changed: mainly species of the brown algal family Chordariaceae increased in abundance through new colonisation possibly favoured by suitable environmental conditions.

Total abundance of the sessile invertebrates of the MSC was 8 weeks after transplantation significantly higher than in the LSC. This difference might indicate that the communities of different ages continue to be distinct from one another in their further succession. Therefore, the initial community composition, which depends on the community age, seems to have an important influence on the succession and further development of the community.

Settlement patterns of the barnacle Balanus crenatus and the hydroid O. dichotoma were not explained by space availability only, as none of this species was recorded in the primary succession. Instead, settlement seemed to be regulated by other factors, including the presence of chemical signals (Pawlik 1992), the presence of conspecifics (Wethey 1986), or other indicator species (Raimondi 1988; Khandeparker et al. 2006; Hung et al. 2007). In return, barnacles often facilitate colonization and survival of macroalgae (Albrecht 1998). In our study, older barnacles (Balanus crenatus) were generally covered with various macroalgae, mainly members of the Chordariaceae or Ectocarpales group. This might be explained by the substrate-alteration (as the rough surface of barnacles collected more algal spores and provide shade (Albrecht 1998; Farrell 1991). The absence of members of the Chordariaceae group during primary succession may, similarly, be explained by inappropriate settlement conditions. Overall, our findings correspond to the facilitation model of Connell and Slatyer (1977), which predicts that early settlers prepare the ground for later ones.

Acknowledgments

This work was part of the diploma thesis of the first author and has been carried out at the Ny Ålesund International Research and Monitory Facility. The authors thank the German scientific diving crew under the leadership of Max Schwanitz: Claudia Daniel, Peter Leopold, and Michael Tessmann for assistance in the field, as well as the Koldewey Station team Rainer Vockenroth, Kai Marholdt, and Cedric Couret for support. Thank to Betti Saier, for assistance in the sampling, conducted in 2005. Thank to Ruth Müller for assistance in measuring the environmental data during the experimental time. Thanks for help in identification questions to Mara Schmiing, Jana Wölfel, Ulf Karsten, and Margaret Clayton. We gratefully acknowledge financial support by the AWI.

Copyright information

© Springer-Verlag 2008