Recovery of deep-sea meiofauna community in Kaikōura Canyon following an earthquake-triggered turbidity flow

Abstract

Turbidity flows can transport massive amounts of sediment across large distances with dramatic, long-lasting impacts on deep-sea benthic communities. The 2016 M_w 7.8 Kaikōura Earthquake triggered a canyon-flushing event in Kaikōura Canyon, New Zealand, which included significant submarine mass wasting, debris, and turbidity flows. This event provided an excellent opportunity to investigate the effects of large-scale natural disturbance on benthic ecosystems. Benthic meiofauna community structure before and after the event was analysed from a time series of sediment cores collected 10 years and 6 years before, and 10 weeks, 10 months, and 4 years after the disturbance. Immediately after the 2016 event abundances of all meiofauna dramatically decreased. Four years later the meiofauna community had recovered and was no longer distinguishable from the pre-event community. However, the nematode component of the community was similar, but not fully comparable to the pre-event community by 4 years after the disturbance. Community recovery was systematically correlated to changes in the physical characteristics of the habitat caused by the disturbance, using physical and biochemical variables derived from sediment cores, namely: sediment texture, organic matter, and pigment content. While these environmental variables explained relatively little of the overall variability in meiofauna community structure, particle size, food availability and quality were significant components. The minimum threshold time for the meiofauna community to fully recover was estimated to be between 3.9 and 4.7 years, although the predicted recovery time for the nematode community was longer, between 4.6 and 5 years. We consider the management implications of this study in comparison to the few studies of large-scale disturbances in the deep sea, in terms of their relevance to the efficacy of the marine reserve that encompasses Kaikōura Canyon, along with potential implications for our understanding of the impacts of anthropogenic seafloor disturbances, such as seabed mining.

Keywords: Deep Sea; Disturbance; Meiofauna; Recovery; Resilience; Submarine canyon; Turbidity flow.

Figure 1. Map of sampling locations.

Location of sampling sites in Kaikōura Canyon overlayed on canyon flushing-induced bathymetric changes. (A) Magnitude of erosion and deposition (seafloor change) within Kaikōura Canyon caused by the canyon flushing triggered by the Kaikōura Earthquake, measured by the differencing the pre- and post-earthquake bathymetry data sets (Mountjoy et al., 2018). (B) Location of the time-series of multicorer sampling sites (yellow circles = sampled in 2010, late 2017, and 2020; purple circles = sampled in 2006 in addition to other time points; green circles = sampled in early 2017 in addition to other time points) within the head of Kaikōura Canyon. Inset shows the location of Kaikōura Canyon (star) relative to New Zealand. Some of the red (erosional) banding evident along the bottom reach of Kaikōura Canyon is an artefact of higher levels of uncertainty in bathymetric differencing for overlapping multibeam coverages (for more detail see Mountjoy et al., 2018). Image source credit: Bigham et al. (2023b) CC-BY 4.0.

Figure 2. Non-metric multidimensional scaling (nMDS) plots of meiofauna and nematode community structure.

Non-metric multidimensional scaling (nMDS) plots of community structure: (A) meiofauna, (B) meiofauna centroids, (C) nematodes, and (D) nematode centroids before the turbidity flow and at 10 weeks, 10 months, and 4 years after the disturbance in Kaikōura Canyon. For meiofauna centroids (B) data from 10 and 6 years before has been combined into a single “Before” centroid. Similarities were calculated from zero adjusted, square root transformed fauna abundances for both community levels. All stress values are below 0.2, indicating that the plots are acceptable representations of the similarity patterns.

Figure 3. Distance-based redundancy analysis (dbRDA) plots for meiofauna and nematodes.

Distance-based redundancy analysis (dbRDA) plot visualising in two-dimensions the relationships between variation in community structure for (A) meiofauna and (B) nematodes (6 years before, and 10 weeks, 10 months, and 4 years after the turbidity flow event in Kaikōura Canyon) and environmental variables examined by the DISTLM analysis. Only variables with a Spearman rank correlation greater than 0.2 are displayed. Vector lengths are proportional to their contribution to the overall variation.

Figure 4. Scatter plots of key environmental factors.

Scatter plots of the most important environmental factors identified by the DISTLM analysis for structuring meiofauna and nematode communities before and after a turbidity flow in Kaikōura Canyon. (A) The percent total organic matter (% TOM), (B) nitrogen (%N, C) the ratio of molar carbon (C) to nitrogen (N), D) Chl a (mg g ${}_{\mathrm{sediment}}^{-1}$), (E) ratio of Chl a to phaeopigments, (F) the skewness of grain size, and G) the percent of grains less than 16 µm. Each dot represents a single core. The dashed line indicates when the turbidity flow in Kaikōura Canyon occurred.

Figure 5. Plots predicting time to recovery for meiofauna and nematodes.

Plots showing three hypothetical models of population growth (linear, exponential, and logistic) used to predict the time to community recovery (indicated by the grey area on the plot; the minimum threshold of 79% or 46% similarity is the within-group similarity of the pre-turbidity community structure) for: (A) the meiofauna and (B) nematode communities in Kaikōura Canyon.

Figure 6. Plot of juvenile nematode percentages through time.

Plot showing the average percentage of juvenile nematodes from sites K2 and K3 at each time point. The dashed line indicates when the turbidity flow in Kaikōura Canyon occurred.

Figure 7. Illustrated schematic showing the changes in the meiofauna community through time.

Schematic illustration showing of the relative abundances of the key taxa identified by the meiofauna SIMPER analysis that characterised the changes in the meiofauna community before and after the turbidity flow in Kaikōura Canyon. Solid arrows connect time points. One individual represents an average abundance of 1–10 ind./10 cm², two individuals represent an average abundance of 10–100 ind./10 cm ², three individuals represent an average abundance of 100-1000 ind./10 cm², four individuals represent an average abundance of 1,000–2,000 ind./10 cm², and five individuals represents 2000+ ind./10 cm². Fauna illustration credit: Elise Littell.