Automated and quantitative characterization of multi-scale benthic habitat and associated biological communities of an unknown southeast Pacific seamount

This study presents the first quantitative, multi-scale characterization of the previously undocumented Solito Seamount in the southeast Pacific, using an integrated framework of high-resolution geophysical data and ROV observations to map distinct megahabitats and reveal how geomorphology, substrate, and depth-dependent oceanographic conditions structure unique benthic megafaunal assemblages.

The Gist

Imagine the deep ocean floor not as a flat, empty desert, but as a bustling, three-dimensional city built on underwater mountains. These mountains are called seamounts. They are like underwater skyscrapers that rise thousands of feet from the ocean floor, creating unique neighborhoods for sea creatures to live in.

However, most of these underwater cities are unmapped and unknown. This paper is like a team of deep-sea explorers who finally got a drone to fly over one specific, mysterious mountain in the Southeast Pacific (dubbed "Solito Seamount") to create the first detailed "real estate map" of the place and see who lives there.

Here is the story of their discovery, broken down into simple concepts:

1. The Mission: Mapping an Unknown City

The scientists didn't just want to take a few photos; they wanted to understand the entire layout of the mountain.

• The Tools: They used a high-tech ship that sent sound waves down to the bottom (like a super-powered sonar) to create a 3D topographic map. This is like using a laser scanner to map the shape of a building without touching it.
• The Drones: They sent down two underwater robots (ROVs) named SuBastian to swim along the slopes and take 4K video. Think of these as the "street-level" inspectors walking around the city blocks to see what the ground is made of and who is living there.

2. The "Neighborhoods" (Megahabitats)

The computer analyzed the shape of the mountain and automatically sorted it into five distinct "neighborhoods" or Megahabitats. Imagine a mountain with a sharp peak, steep sides, deep valleys, and a flat base:

1. The Ridge Crest: The sharp, rocky top of the mountain. It's like a jagged mountain peak.
2. The Ridge Slope: The steep sides just below the peak.
3. The Basal Slope: The lower, gentler slopes where the mountain meets the plain.
4. The Valley: The deep, low-lying trenches between ridges.
5. The Flat: The vast, flat ocean floor surrounding the mountain.

The computer did this mathematically, looking at how steep the ground is and how rough it feels, rather than a human guessing.

3. The "Ground" (Substrates)

Once they knew the neighborhoods, they looked at the "flooring" in each one. Just like a house can have hardwood, carpet, or tile, the seafloor has different materials:

• Bedrock: Bare, hard volcanic rock (like a concrete sidewalk).
• Sediment: Soft sand and mud (like a carpet).
• Coral Rubble: Piles of broken dead coral skeletons (like a pile of bricks or rubble).

They found that the "flooring" changed depending on the neighborhood. The steep peaks were mostly hard rock, while the deep valleys were covered in soft sand.

4. The "Residents" (The Animals)

This is where it gets exciting. The scientists found that where you live depends on what your floor is made of.

• The "Hard Rock" Apartments: On the steep, rocky slopes, they found a fancy community of "suspension feeders." These are animals like corals, sea fans, and crinoids (feather stars) that stick to the rock and wait for food to float by in the current. It's like a high-rise building where everyone has a balcony facing the wind.
• The "Rubble" Zone: In one specific area on the upper slope, they found piles of broken coral. This wasn't just trash; it was a nursery! A specific type of sponge (Stelletta sp.) loved living here. It's like finding a specific type of bird that only nests in old, broken tree branches.
• The "Sand" Zones: The soft, sandy areas had fewer animals, mostly those that could burrow or move around easily.

5. The "Weather" Factor (Depth and Oxygen)

The researchers also noticed that the "city" changed as you went deeper.

• The Shallow Dive: Had simpler communities.
• The Deep Dive: Had a much more diverse and complex mix of animals.

Why? Because of the "weather" of the deep ocean. The deeper dive went below a layer of water that has very little oxygen (the Oxygen Minimum Zone). Below that layer, the water is rich in oxygen and nutrients, acting like a fertile garden that supports a much wider variety of life.

Why Does This Matter?

This paper is a big deal for three reasons:

1. First Look: This is the first time anyone has made a detailed, scientific map of this specific seamount. Before this, it was just a blank spot on the map.
2. The Blueprint: The method they used (combining sonar maps with robot videos and computer math) is a new, objective way to study the deep sea. It's like moving from drawing a map by hand to using GPS and satellite data.
3. Conservation: We can't protect what we don't know exists. By showing that this mountain is a busy, diverse ecosystem with unique "neighborhoods," the scientists are giving conservationists the data they need to protect these areas from threats like deep-sea mining or fishing.

In a nutshell: The scientists used a digital map and a robot camera to discover that an underwater mountain in the Pacific is a complex city with different neighborhoods, different types of ground, and unique animal communities that depend entirely on their specific address and the oxygen levels of their neighborhood.

Technical Summary

1. Problem Statement

Despite the ecological importance of seamounts as biodiversity hotspots, the vast majority remain unmapped and poorly characterized, particularly in the southeast Pacific Ocean. This region is known for high marine endemism and ecological isolation but suffers from a lack of systematic exploration. Existing habitat mapping methods often rely on subjective, rule-based classifications (e.g., user-defined thresholds in Benthic Terrain Modelers) that lack reproducibility and standardization. Furthermore, there is a critical need to understand how multi-scale environmental heterogeneity (from mega-scale geomorphology to macro-scale substrate) structures benthic communities in data-poor deep-sea regions. This study addresses the gap by characterizing a previously undocumented seamount ("Solito Seamount") located between the Nazca–Desventuradas Marine Park and the Juan Fernández Archipelago.

2. Methodology

The study employed an integrated, automated, and hierarchical quantitative framework combining geophysical data with in situ biological observations.

• Study Area & Data Acquisition:

  • Location: Solito Seamount (Chilean EEZ), investigated during the Schmidt Ocean Institute's 2024 expedition (FKt240108).
  • Geophysical Data: High-resolution multibeam bathymetry (100 m grid) and backscatter intensity (50 m grid) were collected using a Kongsberg EM 124 echosounder.
  • Biological Data: Two Remotely Operated Vehicle (ROV) dives (SO643: ~1,054–1,497 m; SO645: ~456–860 m) provided HD/4K video transects.
  • Sampling: Systematic frame grabs were taken every 10 seconds. Standardized quadrats (1–5 m²) were established for benthic fauna assessment, resulting in 134 quadrats (66 for SO643, 68 for SO645).

• Analytical Framework:

  • Megahabitat Mapping (Mega-scale):
    • Calculated terrain derivatives: Slope, Bathymetric Position Index (BPI), Profile/Planform Curvature, Terrain Ruggedness Index (TRI), and Vector Ruggedness Measure (VRM).
    • Dimensionality Reduction: Principal Component Analysis (PCA) was applied to nine abiotic variables to reduce collinearity and capture variance.
    • Clustering: Unsupervised k-means clustering was applied to the principal components to objectively delineate distinct megahabitat classes without user-defined thresholds.
  • Substrate Characterization (Macro-scale):
    • ROV video frames were visually annotated to classify substrates into five types: Bedrock, Bedrock with sediment veneer, Sediment–rock transition, Sediment, and Coral rubble.
  • Statistical Analysis:
    • PERMANOVA: Used to test the significance of dive, megahabitat (nested in dive), and substrate (nested in megahabitat/dive) on benthic assemblage structure.
    • SIMPER (Similarity Percentage): Identified specific taxa driving compositional dissimilarities between habitat and substrate types.

3. Key Contributions

• First Quantitative Map: Provides the first quantitative habitat map of the Solito Seamount, a previously undocumented feature.
• Objective Methodology: Demonstrates a reproducible, automated workflow using unsupervised machine learning (PCA + k-means) to define megahabitats, overcoming the subjectivity of traditional rule-based classification.
• Multi-Scale Integration: Explicitly links mega-scale geomorphology (derived from acoustic data) with macro-scale substrate types and biological assemblages, offering a mechanistic understanding of deep-sea ecosystem structuring.
• Baseline Data: Delivers essential baseline information for a region facing increasing anthropogenic pressure and limited biological knowledge.

4. Key Results

A. Megahabitat Classification

The unsupervised clustering identified five distinct megahabitat types based on geomorphological and acoustic properties:

1. Ridge Crest: Shallow, high relief, high ruggedness, positive BPI.
2. Ridge Slope: Steep, structurally complex, slightly deeper than crests.
3. Basal Slope: Intermediate depth, moderate slopes, geomorphic transition zone.
4. Valley: Depressed topography, negative BPI, low ruggedness, sediment accumulation.
5. Flat Seafloor: Low relief, abyssal plain environment.

B. Substrate Distribution

Five substrate types were mapped. Hard substrates (Bedrock, Bedrock with veneer) dominated ridge crests and slopes, while soft sediments dominated valleys and basal slopes.

• Dive SO643 (Deeper): Heterogeneous mix; ~60% bedrock with veneer, ~20% exposed bedrock.
• Dive SO645 (Shallower): Dominated by exposed bedrock (>80%); coral rubble was a minor but distinct component (~12%).

C. Benthic Community Structure

• Drivers of Variation: PERMANOVA confirmed that substrate type (nested within megahabitat) had the strongest effect on community structure, followed by megahabitat type and the dive itself.
• Dive SO643 (Deeper): Showed high diversity and complex structuring.
  • Ridge Crest: Dominated by reef-building scleractinians (e.g., Solenosmilia, Enallopsammia) and branching octocorals.
  • Ridge Slope: Mixed assemblages of antipatharians, gorgonians, and actiniarians.
  • Drivers: A broad suite of taxa contributed to dissimilarities, organized primarily by megahabitat.
• Dive SO645 (Shallower): Simpler patterns with fewer taxa driving differences.
  • Coral Rubble: Formed a distinct habitat characterized by the sponge Stelletta sp.
  • Drivers: Substrate effects were pronounced; Stelletta sp. was the primary driver of dissimilarity in coral rubble areas.
• Depth/Oxygen Influence: The distinct biological differences between dives likely reflect depth-dependent structuring driven by water mass changes (Oxygen Minimum Zone vs. well-oxygenated Antarctic Intermediate Water).

5. Significance

• Conservation Planning: The study identifies Solito Seamount as a priority candidate for marine spatial planning due to its high habitat heterogeneity and role as a potential biogeographic "stepping stone" connecting the Nazca-Desventuradas and Juan Fernández systems.
• Methodological Advancement: The proposed framework offers a scalable, objective approach for deep-sea habitat mapping in data-poor regions, moving away from expert-biased classifications toward data-driven, reproducible models.
• Ecological Insight: The research highlights the critical role of substrate heterogeneity (specifically coral rubble and hardgrounds) in supporting diverse megafaunal assemblages, including ecosystem engineers like cold-water corals and sponges.
• Future Directions: The study underscores the need for more extensive ground-truthing to transition from unsupervised to supervised classification models (e.g., Random Forest, CNNs) for even more robust habitat mapping.