Spatial and temporal distribution of Sargassum muticum (Phaeophyta, Fucales) in Dråby Vig, Limfjorden

Introduction

Sargassum muticum is a brown algae originating from Asia. For the past 25 years it has been known for its invasive colonization of European waters (Rueness, 1989). In Denmark, the first specimen of S. muticum was found in Nissum Bredning, Limfjorden (St. 1 figure 1) in 1984 (Christensen, 1984). Today S. muticum is the most dominant macroalgae in Limfjorden (Stæhr et al. 1998) but only few studies have focused on its small scale distribution pattern (Knudsen, 1995). Mapping distribution of organisms is the core of the subject of biogeography (Seddon, 1971; Tivy, 1982) but mainly recognized as a descriptive and explorative procedure. However inference of processes creating the observed pattern is possible if the scale of mapping fits the scale of the operating process (O´neill et al., 1986; Wiens, 1989). In recent years the awareness of spatial dependence in regional variables (e.g. distribution of organisms) - and its implication for using interpolation routines, defining proper sampling units, and obscuring p-values from statistical tests - has outlined a new focus on basic mapping techniques and spatial analysis in subtidal communities (Kendrick, 1992; Bell et al. 1995; Vidondo et. al. 1997). The work presented here looks at the patterns of distribution at several spatial and temporal scales starting at the scale of a locality (100´s of m) and zooming down to microscale (cm) levels.

Figure 1: The study site Dråby vig, Limfjorden (St.10 enlarged). The highlighted red areas correspond to the different mapping areas (not drawn to scale) outlined in the following pages. Mapping of circle points are described at page 2, red line is on page 3, square on page 4-7 and black lines on page 8. 

Distribution of S. muticum along a depth gradient (1989-1997)

Figure 2: The abundance of S. muticum in relation to year and depth.

Above graph is based on data collected in midsummer in connection to the National Monitoring Program (NMP) station 10 in Limfjorden (Krause-Jensen et al. 1995; Limfjordsovervågningen, 1996). An analysis of these data in relation to S. muticum at the level of the Limfjord system is presented by Stæhr et al. (1998). The variable "percent cover of hard substrate" measured by the NMP was multiplied by the category "percent hard substrate" to make their data comparable with own results. The category "percent hard substrate" refers to a subjective judgment in the sampling process of whether the substrate is suitable for macroalgal settlement, survival and growth.

Figure 2 shows three important aspects. First the dynamic nature of the invasion and establishment from 1989-1992 is clear. Second the algae becomes dominant from 2-4m depth. Third the percent cover of the area is very high for a single species in such a subtidal system almost forming a monoculture from 2-4m. However these results - emphasizing vertical pattern in large depth intervals - does not outline the distribution pattern very precisely (e.g. the mapped horizontal scale is subjectively based - and not explicitly specified). In trying to improve the resolution of the observed depth-dependence a more detailed transect mapping was carried out. The location of this transect is indicated by the red line along the abscissa on the figure and is explored on page 3.

Distribution of S. muticum along a transect (September 1997)

A transect was laid perpendicular to the coastline and mapped for percent cover of S. muticum, depth and 5 different substrate classes. The figure first explains the reason for the low abundance observed in the depth interval from 1-2m noticed on page 2 - since mud and sand obviously is unsuitable for settlement and growth. Second, the high abundance in the first few meters from the coastline, where small and medium size stones are present, is the product of the annual recruitment (propagules settled around August) and is composed of centimeter long individuals. However only few recruits survive the autumn and winter season, due to exposure and sand abrasion (this explains the low values on figure 2 in this interval). Third the mapping demonstrates a more continuos (soft curvature), patchy (‘ups’ and ‘downs’ pattern) and detailed distribution compared to figure 2. Fourth, the low percent cover compared to figure 2 (2-4m) is due to seasonal lifehistory (Wernberg-Møller et al., 1998). Fifth, this mapping technique again emphasize vertical aspect and neglects the horizontal variation. Due to the limits of point 4 and 5 a hectare was mapped intensively during 1997 and the red line superimposed on the abscissa demarcate the position of the mapped hectare described on page 4-7.

Distribution of S. muticum in a hectare from 2.7-3.4m (May-November 1997)

_

_

_

SCUBA divers visually estimated percent cover of S. muticum around 68 fix-points in a 100x100m square. The sampling grain was ca. 2m² and each point was estimated by two divers reducing the subjective element of the cover estimate. Maps were produced by surface interpolation using a weighted distance model in a Geographical Information System (IDRISI).The red values are the means for the hectare. Three patterns of variation are shown. First there is a seasonal component illustrated by the difference in colors between maps. This seasonality reflects the life strategy with rapid growth in spring and summer and a drastic change in autumn when the algae shed its laterals (cf. Wernberg-Møller et al, 1998). Second there is a patchy element (cf. p. 4 & 5). Third there is an abundance gradient from down position of maps (near coastline) to top. This aspect is highlighted on page 7. The low mean values compared to figure 2 reflects scale-dependency. All three variation types emphasize the limitations of interpolating values estimated from techniques described on page 2 and 3.

Substrate suitable for S. muticum settlement, survival and growth

The substrate type suitable for macroalgal settlement, survival and growth is dependent on the life history of the algae (e.g. size, growth form) and environmental variables (e.g. sand abrasion, exposure, sedimentation). Therefore a common fixed grain size is not generally a good measure of suitability to different algae. A differentiation of substrate into size classes makes it possible to vary the amount of the potential area suitable for the algae under investigation depending on the above mentioned factors.

Each substrate class was estimated visually as percent cover of area. "Small stones" are smaller than 10cm, "medium stones" less than 20cm and "boulders" larger than 20cm (error bars are ±SD based on mean values of 7 monthly sample procedures - each composted of 68 random samples cf. p.4). S. muticum is a large algae (compared to other Danish species) with a strong holdfast but fragile primary laterals. The canopy creates a high drag under exposed situations and it is therefore not superior on exposed localities. The substrate in Dråby vig is characterized by many small stones. This size class is only partly suitably for S. muticum. In the initial life history phase the weight of the small stones are higher than the buoyancy of the recruits. However as illustrated in the photo the buoyancy of large algae often supersedes the stones and the algae is under stormy conditions moved shorewards. This net shoreward transport mechanism is mainly detrimental to the algae - but has also been suggested as a potential small scale spreading mechanism (Christensen, 1984).
 
 

July: Basal part of a 'jumping' Sargassum muticum settled onto a small stone. When the thallus grows the buoyancy of the vesicles will cause the alga and its substrate  to float away. Later when the primary laterals are shed, the alga will resettle in a new location - not necessarily a favorable one.

Abundance of S. muticum (July 1997) super-imposed on a digital elevation model of boulder distribution

The description of sampling procedures of the two variables are found on page 4 and 5 (color codes are similar to legend on figure 4), and the relevance of boulder distribution compared to S. muticum abundance described on page 5. The map was produced by a PCA analysis in IDRISI on 7 maps of monthly boulder distributions. The first image component showing the "typical" boulder map (Eastman, 1995) explains ca. 90% of the total variation and is the one used above. The residual variation is noise, since boulder distribution is supposed to be constant on the scale of years. High elevation on the figure thus correspond to high boulder concentration.

The figure explain some of the patchy 2D pattern noticed on page 4, e.g. the green top (high S. muticum and boulder values) and dark trough in the lowest corner (opposite i.e. low values). However the overall trend of light colors in the one end and dark colors in the other does not correspond to boulder distribution. This trend is thus explored on page 7.

The map was produced in IDRISI by superimposing a biomass map on a DEM of depth values (c.f. p.4 and 6 for more details). Notice that the z-axis is highly exaggerated to visually elucidate small scale differences. The map was converted from percent cover to biomass by map algebraic procedures based on an empirically estimated relationship for July 1997 (Wernberg-Møller et al., 1997).

The figure describes the biomass gradient reasonably well. Some patches are however better explained according to substrate conditions - cf. Page 6. The potential relationship between irradiance, depth and S. muticum abundance within the hectare should however be interpreted with caution. First the depth difference is small (maximum 0.7m) which makes simple cause/effect conclusions suspicious considering that the algae reach a length of 1.5m in 4 months (Wernberg-Møller et al., 1998). Furthermore transplanted individuals to similar depth regimes did not show significant difference in height after 3 months of growth (unpublished data). It is possible that the settlement success and initial growth is more sensitive to low irradience levels than established individuals. Finally sedimentation levels is correlated with irradience and exposure, since all the variables to some extent are depth dependent. To differentiate the effects of these factors specific experiments and testing should be conducted.

The locations of the three transects (154m, 130m and 100m) are shown on page 4. The transects were mapped continuously and presence and location of S. muticum noted with centimeter accuracy. Figure 8a illustrates the distribution pattern along the transects as density points. These data were pooled into 0.5m intervals and the autocorrelation calculated for each 0.5m lag.

The correlograms show the same trend (although very noisy) indicating an isotrophic distribution. The transects seems to indicate spatial independence at the scale of approximately 80-90m although differing, detrending, fitting and testing models need to be carried out before conclusion can be drawn. Potential small scale processes creating these patterns might be intra- and interspecific competition and propagule dispersal, settlement and recruitment - which is highly localised for Sargassum (Kendrick, 1992).

Processes determining distribution patterns operates on different scales creating patches (mosaics), trends and zones. On the largest scale investigated, depth (light) and exposure variables create gradients (p. 2-3) and substrate conditions explains some patches (p. 3.). On a smaller scale substrate and depth probably again explains much of the spatial variation (p. 4-7). On the finest mapped scale also effects of competition and propagule dispersal are likely explanations for observed patterns (p. 8). In this study processes operating at scales beyond the level of the locality can obviously not be inferred. These includes typical differences in water chemistry (nutrients, temperature and salinity) - which in the shallow water system of Dråby Vig is supposed to be relatively constant - and long term dispersal mechanisms (e.g. floating abilities of reproductive laterals). Mapping techniques are generally descriptive in nature and mainly explorative data analysis can be undertaken. However this project has demonstrated the ability to visually explore relationships between S. muticum abundance and abiotic variables, as well as providing the basis for stating precise testable hypothesis. It would thus be relevant to test settlement-, survival- and growth success of S. muticum in relation to combinations of different exposure, light and substrate regimes on the basis of the knowledge obtained from the mapping procedures.

Related posters...

Invasion and productivity of Sargassum muticum (Yendo) Fensholt in Limfjorden, Denmark
Presented at the 32nd EMBS, Lysekil, Sweeden, 16. - 22. August 1997.

Invasion of Sargassum muticum (Phaeophyta, Fucales) in, Limfjorden, Denmark
Presented at the 10th Marine Research Conference, Hirtshals, Denmark, 21.-23. January 1998.

Phenology of Sargassum muticum (Phaeophyta, Fucales) in Limfjorden, Denmark
Presented at the 10th Marine Research Conference, Hirtshals, Denmark, 21.-23. January 1998.

Related pages...

UW-photos from Dråby Vig, Limfjorden

Acknowledgements

The present poster is part of our Master thesis in Environmental Biology at Roskilde University. It was made possible through financial support from The Staff-Student Committee for Biology (Studienævnet for Biologi). We wish to thank Morten Foldager Pedersen (RUC) for superb supervision and constructive criticism and Dorte Krause-Jensen (DMU, Silkeborg) for stimulating discussions and making data from the NMP available.

References

Christensen, T (1984): Sargassotang, en ny algeslægt i Danmark. Urt, Vol. 4 p.99-104.
Eastman, R.J. (1995): Change and time series analysis techniques: A review in Change and Time series analysis Second edition. Eastman, J.R.; McKendry, J.E & Fulk, M.A. United Nations Institute for Training and Research - Explorations in Geographic Information Systems Technology Vol.1.
Kendrick, G. (1992): The role of dispersal and rekruitment processes in structuring mixed species beds of Sargassum (Sargassaceae, Phaeophyta) at Rottnest Island, Western Austratlia. Ph.D. thesis, University of Western Australia.
Knudsen, L. (1995): Årscyklus for Butblæret sargassotang (Sargassum muticum) på en tæt bevokset lokalitet i Limfjorden. Notat fra Bio/consult udarbejdet til Limfjordssamarbejdet v/Ringkøbing Amtskommune.
Krause-Jensen, D.; Christensen, P.D. & Sandbeck, P. (1995): Retningslinier for marin overvågning - Bundvegetation. Teknisk anvisning fra DMU nr.9 Miljø og Energiministeriet, Danmarks Miljøundersøgelser.
Limfjordsovervågning (1996): Vandmiljø i Limfjorden 1995. Ringkøbing amtskommune, Viborg amt og Nordjyllands Amt.
O´neill, R.V.; D.L. DeAngelis, J.B. Waide, and T.F.H. Allen (1986): A Hierachical Concept of Ecosystems. Princeton University Press.
Rueness, J (1989): Sargassum muticum and other introduced Japanese macroalgae: biological pollution of european coasts. Marine Pollution Bulletin Vol. 20 p.173-176.
Seddon, G. (1971): Introduction to biogeography. Ductworth.
Stærh, P.A.; Wernberg-Møller, T. and Thomsen, M.S. (1998): Invasion of Sargassum muticum (Phaeophyta, Fucales) in Limfjorden, Denmark. Poster presented at the 10th Marine Research Conference, Hirtshals (Denmark) the 21st-23rd of January.
Tivy, J. (1982): Biogeography - a study of plants in the ecosphere. Longman Group Ltd. (Second edition).
Vidondo, B.; Duarte, C.M.; Middelboe, A.L.; Stefansen, K.; Lützen, T. & Nielsen, S.L. (1997): Dynamics of a landscape mosaic: size and age distributions, growth and demography of seagrass Cymodocea nodosa patches. Marine Ecology Progress Series Vol. 158 p.131-138.
Wernberg-Møller, T.; Stæhr, P.A. & Thomsen, M.S. (1997): Invasion and productivity of Sargassum muticum (Yendo) Fensholt in Limfjorden, Denmark. Poster presented at the 32nd EMBS in Lysekil (Sweden) the 16th -22nd of August.
Wernberg-Møller, T.; Thomsen, M.S. & Stæhr, P.A. (1998): Phenology of Sargassum muticum (Phaeophyta, Fucales) in Limfjorden Denmark. Poster presented at the 10th Marine Research Conference, Hirtshals (Denmark) the 21st-23rd of January.
Wiens, J.A. (1989): Spatial scaling in ecology. Functional Ecology vol. 3 p.385-397 Substrate growth