Depth affects the population dynamics on a soft coral?dominated reef on the Great White Wall, Fiji

Depth affects the population dynamics on a soft coral-dominated reef on the Great White Wall, Fiji

Nile P. Stephenson, Katie M. Delahooke, Charlotte G. Kenchington, Jone Waitaiti, Alice A. Ball, Victor E. Bonito, Andrea Manica, & Emily G. Mitchell

Depth_affects_the_population_dynamics_on_a_soft_co_compressed Download

Abstract

Soft corals (order Alcyonacea) are an important component of tropical coral reefs, and often form locally abundance dense carpets¹. Some soft coral species are prone to bleaching and heat stress like scleractinian, reef-forming corals, but others appear resistant to bleaching stressors². Despite the importance of soft corals in tropical reefs, there is limited knowledge regarding their demographic ecology³. Spatial point process analysis (SPPA) enables inference of ecological processes underlying spatial distributions of individuals in a community by investigating the fit of models of known processes to observed spatial patterns⁴. SPPA can also quantify how these underlying ecological processes change with environmental gradients (such as depth)⁵. We used SPPA to investigate the population ecology on the Great White Wall (GWW), Fiji a vertical reef dominated by Nephtheidae soft corals alongside other filter feeders and algae⁶. We found that a combination of reproductive and habitat association processes best explained the spatial patterns on the GWW⁷. We found that the best-fit underlying ecological processes changed with depth, and that Nephtheidae soft corals became more dominant on deeper parts of the wall, likely due to a combination of reproductive lability, fast growth rates, and increasing suitable habitat associated with filter-feeding⁸⁸⁸⁸. Even where the Nephtheidae were most dominant, other filter feeders were able to persist⁹. This mixed community composition could be because the Nephtheidae feed in periods of high current, and temporal niche partitioning in feeding could be permitting biodiversity on the GWW, allowing for hydroids and sponges to coexist with the Nephtheidae¹⁰.

Introduction

Tropical coral reefs are major contributors to marine biodiversity and provide economically important ecosystem services, such as reducing wave action and erosion, tsunami protection, ecotourism, and subsistence fishing¹¹. Tropical coral reefs are experiencing dramatic ecological collapses due to anthropogenic climate change (via increased storm frequency, ocean warming, and acidification), destructive fishing techniques, and overharvesting, leading to a decline in hard coral populations, and consequentially, those ecosystem services provided by hard corals¹².

A major component of the benthic community within tropical coral reefs are the soft corals (order Alcyonacea)—soft bodied octocorals with an internal skeleton composed of sclerites¹³. Similarly to hard corals, soft corals are primary consumers, promoting biodiversity via mutualistic benefits to a range of species, produce economically important biocompounds, and support fish biodiversity¹⁴¹⁴¹⁴¹⁴. In contrast to hard corals, some soft coral taxa, such as those in the family Nephtheidae, retract when not feeding and so do not provide the same coastline protection via disruption of wave action¹⁵. However, certain soft coral taxa are resistant to many of the stressors prevalent in the decline of hard coral reefs, such as pollution and sedimentation, temperature-related bleaching, ocean acidification, and storm damage¹⁶. This resistance has led to persistence or even dominance of soft corals in some areas where hard corals are declining¹⁷. Conversely, some soft coral populations bleach and experience population declines congruent with those of hard corals in response to climate change¹⁸.

Some soft coral species have been designated as invasive or range shifters due to rapid substrate colonisation arising from fast growth rates, high competitive capacity through allelopathy, removal via colony migration permitting overgrowth of hard corals, and lability of reproductive mode¹⁹. Therefore, soft coral-dominated communities have the potential to be an end point of phase-shifts for tropical coral reefs as hard corals decline due to anthropogenic stressors which soft corals can tolerate²⁰. However, soft coral reefs would likely provide a different set of ecosystem services compared to hard corals²¹. It is therefore important to understand underlying ecological processes contributing to soft coral demographic ecology in areas where they are dominant in order to inform conservation and management strategies for tropical coral reefs²².

The demographic ecology of sessile organisms can be determined through analyses of the spatial pattern of individuals in a population²³. The spatial patterns of organisms within communities reflects underlying ecological processes that have occurred throughout an organisms’ life²⁴. Four sets of processes are known to determine the patterns of sessile organisms: (1) interactions with the environment; (2) reproductive processes; (3) interactions within and between species (e.g., competition, facilitation); and (4) density-dependent mortality processes²⁵. Spatial point process analysis (SPPA) is the analysis of patterns of points (here, each representing a single organism), revealing the most likely ecological processes underpinning the observed spatial pattern²⁶. For instance, reproductive aggregations can be modelled by Thomas cluster (TC) processes, while aggregations due to associations with local habitat heterogeneities by heterogeneous Poisson (HP) processes²⁷. SPPA has been used to reveal mortality processes, habitat interactions, and competition in hard coral communities, as well as density-dependent mortality processes and facilitation in deep-sea soft corals, and reproductive aggregations in gorgonians²⁸. However, SPPA has not, to our knowledge, been applied to tropical soft corals to date²⁹.

Methods

Field site

In this study, we focussed on the GWW, a vertical section of a barrier reef in the southwest of the Somosomo Strait which runs between Vanua Levu and Taveuni Islands, Fiji (Fig. 1a), and is dominated by white Nephtheidae soft corals that give the site its name³⁰. The GWW is the ideal site to try to understand soft coral population dynamics because it is a well-known, accessible reef with high densities of soft corals within safe diving limits³¹. We accessed the site by scuba diving and used video to map between the top of the wall at 17 m to a depth of 33.5 m³². Alongside, the white Nephtheidae soft corals, other abundant morphospecies were a species of colonial hydroid (visually similar to Myriopathes), lemon sponges (Leucetta chagosensis), colourful

Dendronephthya soft corals, algae (possibly Halimeda spp.), and adult Tubastraea sp. (an azooxanthellate hexacoral with a hard skeleton) ³³³³³³³³. All these morphospecies except the algae are filter feeders³⁴.

Fig. 1: a) Location of the Great White Wall (GWW; orange star). b) zoomed out map of Vanua Levu and Viti Levu islands, Fiji, and c) broader geographical context of Fiji. Arrows in a indicate current direction depending on tides. Blue area indicates barrier reef. d) Tubastraea on the GWW, e) Lemon sponge, f) Dendronephthya, g) white Nephtheidae on the GWW, h) algae, i) hydroid colonies. Scale bars for e, g, h, and i ~1 cm; scale bars for d and f ~10 cm³⁵.

Data collection

We used structure-from-motion photogrammetry derived from video footage of the GWW in Agisoft Metashape (v2.1.1) to reconstruct the GWW community³⁶.

Video footage (20 videos, total length=4251 s) and photographs (64 images) were collected by scuba divers moving in a boustrophedonic, lawnmower pattern from the deepest section to the top of the wall³⁷. Footage could only be collected in periods of low current³⁸.
Video footage was segmented at a capture rate of 1 frame/second, yielding a total of 4,348 frames³⁹.
Blurred, obstructed, and out of focus frames were removed, and the remaining 3,601 frames were imported into Agisoft Metashape⁴⁰.
Photos were aligned to form a point cloud, which was then used to produce a 3D model⁴¹.
The model was scaled using spaced laser dots and objects of known size⁴².
A 2D orthomosaic (photomap; Fig. 2a) was produced and exported⁴³.
Individual organisms were outlined and identified on the 2D map to a morphospecies level⁴⁴.
The marked-up map was extracted using a custom script in R, resulting in a final digital dataset of the spatial position and morphospecies ID for each specimen⁴⁵.

In order to investigate if morphospecies were responding to broad-scale background topography, we constructed a digital elevation model (DEM) from the 3D mesh⁴⁶. We decimated and smoothed the mesh to remove any topography caused by the corals themselves, preserving only the finer 3D topological structure (e.g., shallow caves and crevices)⁴⁷⁴⁷⁴⁷⁴⁷.

Broad ecological patterns on the GWW

To assess broad ecological patterns in the GWW community, we used Berman-Lawson-Waller tests to investigate the relationship between depth and density for all organisms collectively and for morphospecies individually⁴⁸. These tests compare heterogeneous Poisson models with depth as a predictor against a null model of constant density⁴⁹. We did not conduct any size-based analyses because Nephtheidae soft corals change size with respect to current, so accurate size measures were not possible⁵⁰⁵⁰⁵⁰⁵⁰.

Univariate spatial analysis of common morphospecies

We investigated the population ecology of the six abundant morphospecies on the GWW by means of univariate spatial point process analyses (SPPA)⁵¹. To do this, we used the pair correlation function (PCF) distance measure to investigate model fits between the observed point pattern of each taxon against known spatial patterns⁵²⁵²⁵²⁵². To investigate the effect of depth on population dynamics, we divided our study area into three depth bins: shallow (17-23.5 m), intermediate (23.5-28 m), and deep (28-33.5 m)⁵³. The division was dictated by variation in the topography of the GWW, as the intermediate depth bin contains four shallow caves⁵⁴.

The simplest spatial pattern is complete spatial randomness (CSR), modelled as a homogeneous Poisson process⁵⁵. Non-CSR patterns can be aggregated (PCF > 1) or segregated (PCF < 1)⁵⁶. Aggregation can be caused by habitat associations (modelled by heterogeneous Poisson (HP) processes) or by dispersal-limited reproductive processes (modelled by Thomas cluster (TC) processes)⁵⁷. Processes combined are best modelled by inhomogeneous Thomas clusters (ITC)⁵⁸. Segregation can be caused by intraspecific resource competition or density-dependent mortality⁵⁹.

To test model fits, 9,984 Monte Carlo simulations were run for CSR, TC, HP, and ITC processes⁶⁰. We then used Diggle’s goodness-of-fit test to quantitatively assess differences between the observed and simulated PCFs, with a high

value indicating a good model fit⁶¹⁶¹⁶¹⁶¹.

Fig. 2: a) Orthomosaic of the Great White Wall (GWW); b) Digital elevation model of the GWW; c) specimen map of the GWW. Colour of point indicates morphospecies (lilac=white Nephtheidae, yellow=lemon sponges, light blue=hydroids, dark green=algae, dark blues=scleractinian corals, reds=other soft corals). Scale bars = 5 m. Note that we changed the size of the circles to aid with visibility and so the circles are not to scale⁶²⁶²⁶²⁶².

Results

The mapped GWW community consisted of 25,496 specimens over 377.74

⁶³. The community was dominated by white Nephtheidae (20,978 colonies; 85.82% relative abundance), with an increase in abundance from shallow to deep depth bins⁶⁴. Hydroids (4.24%), lemon sponges (1.79%), and algae (3.62%) were the other most dominant morphospecies⁶⁵.

Tubastraea (1.18%) and Dendronephthya (1.06%) were abundant in the shallow depth bin but hardly present at the deeper depth bins⁶⁶.

Berman-Lawson-Waller tests revealed that the density of all morphospecies together and of each morphospecies individually significantly depended on depth⁶⁷. No morphospecies exhibited a best fit model to CSR at any depth bin⁶⁸.

The white Nephtheidae and hydroid populations were best fit by a TC model at smaller spatial scales and an ITC model at larger spatial scales⁶⁹. For lemon sponges, patterns were best fit by HP and ITC at larger spatial scales and TC at smaller spatial scales in the intermediate and deep bins⁷⁰. The algae spatial pattern was best fit by an ITC model at the shallow depth bin and a TC and ITC model for intermediate and deep depth bins⁷¹. In the shallow depth bin,

Dendronephthya and Tubastraea were best fit by TC models⁷².

In terms of clustering parameters from best fit TC models, all four abundant morphospecies increased in the density of parent points (?) with depth, but the white Nephtheidae exhibited a disproportionately greater response⁷³. All morphospecies decreased in the mean number of individuals per cluster (µ) with increasing depth⁷⁴. Likewise, all morphospecies (except for lemon sponges) showed a decrease in cluster radius (?) with increasing depth⁷⁵.

Morphospecies	Depth	n	Density (m?²)	Best-fit TC (pd)	Best-fit scale (m)	? (m)	? (m?²)	? (per cluster)
Nephtheidae	S	1,252	9.67	0.534	0.1-1.5	0.52	0.07	145.63
	I	9,304	66.46	0.654	0.3-1.5	0.57	0.53	125.02
	D	11,165	95.17	0.348	0.1-1.5	0.80	0.54	174.72
Lemon sponge	S	288	2.22	0.505	0.3-1.5	0.23	5.43	0.41
	I	96	0.67	0.801	0-0.3	0.07	1.27	0.53
	D	107	0.91	0.985	0-0.3	0.06	2.68	0.41
Hydroid	S	176	1.36	0.609	0.3-1.5	0.29	0.19	7.05
	I	343	2.45	0.234	0-0.2	0.05	0.66	3.73
	D	647	5.52	0.413	0-0.1	0.03	2.03	2.72
Algae	S	145	2.34	0.875	0-1.50	0.14	0.06	18.99
	I	474	1.12	0.339	0-0.3	0.01	0.31	10.86
	D	347	3.39	0.258	0-0.1	0.04	1.04	2.84
Tubastraea	S	129	1.00	0.330	0-1.5	0.19	0.44	2.28
Dendronephthya	S	164	1.27	0.917	0-0.3	0.06	2.41	0.53

Table 1: Summary of univariate PCF analyses for abundant morphospecies on the Great White Wall. S=shallow, I=intermediate, D=deep.⁷⁶

Discussion

Both the white Nephtheidae and hydroid populations were well fit by TC and ITC models at all depths, reflecting the shared importance of reproductive processes and the local environment for these morphospecies⁷⁷. These best-fit models likely reflect the influence of combined reproductive and feeding dynamics⁷⁸. It is likely that the abundance of filter feeders on the GWW capitalises on an abundance of food from nutrient-rich water flowing over the wall⁷⁹. Lemon sponge spatial patterns were also fit well by reproductive models with habitat associations (ITC), but habitat preferences were much more important where food was scarcer in the shallows⁸⁰.

Algae were best fit by reproductive (TC) models at all depth bins, likely reflecting more plentiful light in the shallows, whereas wall topography and water depth created an impediment to light access at greater depths⁸¹⁸¹⁸¹⁸¹.

Tubastraea and Dendronephthya spatial patterns were best modelled by reproductive processes (TC), but these morphospecies were only plentiful in the shallow bin⁸². This could be reflective of competition for space precluding larger morphospecies with slower growth rates at greater depths where overall density is higher⁸³.

The white Nephtheidae were the dominant filter feeders on the GWW⁸⁴. In other locations where soft corals show local dominance, reproductive lability, fast growth rates, and higher competitive ability have been suggested as mechanisms of dominance⁸⁵. On the GWW, we observed fission and budding reproductive modes from photographs⁸⁶. The capacity for Nephtheidae to grow quickly and reproduce using a variety of strategies, regular nutrient influx, and a potentially high disturbance regime are likely the combined contributors to their local dominance on the GWW⁸⁷.

Despite the dominance displayed by the white Nephtheidae, substantial populations of other filter feeders were still able to coexist⁸⁸. One possible explanation is that Nephtheidae soft corals have a temporal feeding niche—they actively filter feed at times of high current, but otherwise retract their polyps⁸⁹. This would permit other filter-feeding taxa to feed at times of low current when the dominant white Nephtheidae are not feeding⁹⁰. Partitioning of time as an ecological resource has been observed in other systems and may be a mechanism for increasing invertebrate biodiversity on some soft coral-dominated reefs⁹¹.

References

Note: The provided document does not include a full reference list. The following works were cited in the text, but their complete bibliographic information is not available in the source PDF.