Figure Legends
Figure 1 Shift in kelp forest biomass and depth distribution over 25 years of Arctic warming at our study site Hansneset in Kongsfjorden, Svalbard. The habitat for kelp communities along Arctic fjord systems declines as the increase in glacial melt intensifies the phenomenon of coastal darkening which prevents the depth extension of key kelp species despite elongation of the open water period. This illustration is based on data from of Hop et al. (2012), Bartsch et al. (2016) and the present study.
Figure 2 Study site of the depth transect at the Hansneset kelp forest and location of the AWIPEV-COSYNA underwater observatory in Kongsfjorden, western Spitzbergen. The maps were plotted using the R package ggOceanMaps (Vihtakari, 2022) in R version 4.2.2 (‘Innocent and Trusting’, R Core Team, 2022).
Figure 3 Fresh weight (FW) m-² of biomass dominant macroalgal species and groups along the depth gradient at Hansneset, Kongsfjorden (Svalbard) over time as indicated by different colors (mean ± SD; 1996/1998: n = 2 (0 m), n = 3 (5 – 15 m), n = 4 (2.5 m); 2012/2013: n = 3 (0 and 15 m), n = 6 (2.5 – 10 m); 2021: all depths n = 3).
Figure 4 Leaf area index (LAI) of macroalgal species and groups along the depth gradient at Hansneset, Kongsfjorden (Svalbard) in summer 2012 and 2021 (mean ± SD, n = 3) as indicated by different colors. Original sample size differed between 2012 (0.25 m²) and 2021 (1 m²). In 2012 no kelps were documented at 10m and 15m due to the small frame size (0.25 m² frame).
Figure 5 Changes of abundance and lower depth distribution (relative to chart datum) of biomass dominant brown algae species over time at Hansneset, Kongsfjorden (Svalbard). Absence of data between 2 - 7m in 1996/98 and 2014 is marked in grey. Changes in lower distribution limits are highlighted (red line) and semi-quantitative abundance classes are indicated by color intensity.
Figure 6 Relative age-class distribution of all kelps and single kelp species along the depth gradient (2.5 – 10m) at Hansneset, Kongsfjorden (Svalbard) in summer 2021. Numbers above stacked columns indicate the mean number of kelp individuals per m² (n = 3). The age of kelp individuals was subsumed into four age classes indicated by color code.
Figure 7 Age and density per m² of adult kelps (≥ 1 year) at 2.5m (a, b) and 5m (c, d) at Hansneset, Kongsfjorden (Svalbard) between summer 2013 and 2021. Colors indicate kelp species (mean ± SD; 2013 and 2021: n = 3).
Figure 8 Individual holdfast (a), stipe (b), blade dry weight (DW) (c) and blade:stipe DW ratio (d) of adult kelps (≥ 2 years) collected from 2.5m and 5m depth in Hansneset, Kongsfjorden (Svalbard) in summer 2021. Boxplots represent the median (50thpercentile), the interquartile range (25th to 75th percentile), and whiskers the lower (5th) and upper (95th) percentile. Colors of boxplots indicate the three kelp species and single data points from 2.5m (blue circles) and 5m (red triangles) are presented. Significant differences between species are marked with different capital letters (non-parametric Kruskal-Wallis test with pairwise Wilcoxon rank sum tests with Bonferroni correction, A. esculenta:n = 75 (2.5 m); n = 43 (5 m); ‘Digitate Kelps’: n = 65 (2.5 m); n = 6 (5 m); S. latissima : n = 29 (2.5 m); n = 4 (5 m)).
Figure 9 Individual blade %carbon (a), %nitrogen (b) and C:N ratio (c) in a random subsample of adult kelps (≥ 2 years) collected from 2.5m and 5m depth at Hansneset, Kongsfjorden (Svalbard) in summer 2021 (A. esculenta: n = 11 (2.5 m), n = 15 (5 m); ‘Digitate Kelps’: n = 15 (2.5 m), n = 10 (5 m); S. latissima : n = 15 (2.5 m), n = 7 (5 m)). Visualization same as Figure 7.
Figure 10 Blade carbon (a) and nitrogen (b) of adult kelps (≥ 2 years) per m2 at 2.5m and 5m depth at Hansneset, Kongsfjorden (Svalbard) in summer 2021. Colors indicate kelp species (mean ± SD; n = 3).
Figure 11. Long-term change in PAR and turbidity at the AWIPEV-COSYNA underwater observatory off Ny Ålesund, Kongsfjorden. Weekly means originate from the daily vertical water column profiles recorded between 11m to 1m depth (+/- tide). The deviation of the observed weekly PAR and turbidity values to the expected PAR and turbidity values is shown. The expected value (zero line) was calculated as the mean of the calendar week across all years. A simple linear regression model (dashed line) was used to analyze the trend in PAR and turbidity change over the years. Only the photosynthetic active period (summer light condition) from week 8 (March) to week 44 (October) was considered.
Appendix 8. Absolute weekly PAR and turbidity values recorded at the AWIPEV-COSYNA (Coastal Observing Systems of the Northern and Arctic Seas) underwater observatory off Ny Ålesund, Kongsfjorden. Data for the winter period from week 45 (November) to week 7 (February) not shown.