Can niche plasticity mediate species persistence under ocean acidification?

Cipriani, Vittoria1 ; Goldenberg, Silvan2; Connell, Sean1; Ravasi, Timothy3; Nagelkerken, Ivan1

Published Jul 25, 2024 on Dryad. https://doi.org/10.5061/dryad.x0k6djhtq

Data files

Jul 25, 2024 version files 130.36 KB

Abstract

Global change stressors can modify ecological niches of species, and hence alter ecological interactions within communities and food webs. Yet, some species might take advantage of a fast-changing environment, and allow species with high niche plasticity to thrive under climate change. We used natural CO₂ vents to test the effects of ocean acidification on niche modifications of a temperate rocky reef fish assemblage. We quantified three ecological niche traits (overlap, shift, and breadth) across three key niche dimensions (trophic, habitat, and behavioural). Only one species increased its niche width along multiple niche dimensions (trophic and behavioural), shifted its niche in the remaining (habitat), and was the only species to experience a highly increased density (i.e. doubling) at vents. The other three species that showed slightly increased or declining densities at vents only displayed a niche width increase in one (habitat niche) out of seven niche metrics considered. This niche modification was likely in response to habitat simplification (transition to a system dominated by turf algae) under ocean acidification. We further show that at the vents, the less abundant fishes have a negligible competitive impact on the most abundant and common species. Hence, this species appears to expand its niche space overlapping with other species, consequently leading to lower abundances of the latter under elevated CO_2. We conclude that niche plasticity across multiple dimensions could be a potential adaptation in fishes to benefit from a changing environment in a high-CO₂ world.

README: Can niche plasticity mediate species persistence under ocean acidification?

https://doi.org/10.5061/dryad.x0k6djhtq

This dataset originates from a study investigating the impact of ocean acidification on a temperate rocky reef fish assemblage using natural CO2 vents as analogues. The dataset covers various niche dimensions, including trophic, habitat, and behavioural niches. The study focused on how fish niches are modified in response to ocean acidification, assessing changes in breadth, shift, and overlap with other species between the acidified site and the control site.

Description of the data and file structure

Raw_single_niche_data

The “Raw_single_niche_data” dataset consists of seven spreadsheets, each sharing two essential columns: 'group' and 'community'. These columns are crucial for subsequent analysis using the SIBER framework.

group = species

Common = common triplefin, Forsterygion lapillum
Yaldwyn = Yaldwyn’s triplefin, Notoclinops yaldwyni
Blue_eyed = blue-eyed triplefin, Notoclinops segmentatus
Blenny = crested blenny, Parablennius laticlavius

community = treatment

C = control
V = CO2 vents

Description of the seven spreadsheets:

Isotopes - the dataset includes ratios of 13C/12C and 15N/14N expressed in the conventional δ notation as parts per thousand deviation from international standards. Stable isotopes were derived from a total of 251 fishes collected across three years of sampling.

iso1= δ13C

iso2= δ15N
2. Stomach volumetric - The dataset includes estimated volumetric measures of stomach contents, where the volume contribution of each prey category relative to the total stomach content (100%) was visually estimated. Data were collected between 2018 and 2019. The stomach content was analysed with this method for common triplefin, Yaldwyn's triplefin, blue eyed triplefin and crested blenny. There are 19 prey categories.
3. Stomach count - All prey items were counted in 10 prey categories: copepods, ostracods, polychaetes, amphipods, gastropods, bivalves, tanaids, mites, isopods , and others. Digested items that were not identifiable were excluded from the analysis. The stomach content was analysed with this method for common triplefin, Yaldwyn's triplefin and blue eyed triplefin.
4. Stomach biomass - The dataset includes calculated biomass derived from the mass of prey subsamples within each category, multiplied by their count.
5. Habitat - The microhabitat occupied and habitat orientation (horizontal, angled and vertical) was recorded using free roaming visual surveys on SCUBA (February 2018).

Microhabitat types:

t. = turf algae <10 cm in height

ca. = erect calcareous algae

cca. = crustose coralline algae

b. = bare rocky substratum

sp. = encrusting fleshy green algae

cobble. = cobbles (~0.5–2 cm in diameter)

Type of surface orientation:

hor = horizontal

angle = angled

vert = vertical
6. Behaviour - Behavioural variables quantified from underwater footage and expressed as rates per minute. The behaviours are: swimming, jumping, feeding, attacking and fleeing from an attack.
7. Aquarium: Data from an aquarium experiment involving Forsterygion lapillum and Notoclinops yaldwyni, showing the proportion of time spent in available habitat types to assess habitat preference in controlled conditions. Time in each habitat type and spent in activity was derived from video recordings of 10 minutes and expressed as a proportion of total observation time.

Common = common triplefin, Forsterygion lapillum

Yaldwyn = Yaldwyn’s triplefin, Notoclinops yaldwyni

Common.c = common triplefin in presence of Yaldwyn’s triplefin

Yaldwyn.c = Yaldwyn’s triplefin in presence of common triplefin

turf.horizontal = time spent on horizontal turf substratum

bare.horizontal = time spent on horizontal bare substratum

turf.vertical = time spent on vertical turf substratum

bottom = time spent on the bottom of the tank

swimming = time spent swimming

aquarium.wall = time spent on the walls of the tank

switches = numbers of changes between habitats

Unified_overlap_dataset

The “Unified_overlap_dataset” consists of ten spreadsheets, each sharing “id”, “year”, “location” and “species “column (with few exceptions detailed). These first columns need to be factors for analysis using the Unified overlap framework. We used the R scripts provided in the original study (Geange et al, 2011), as detailed in the manuscript. Data for control and vents are in separate data sheets, with C = control and V = vent.

Id: sample number

Year: year the data were collected

Location: North (n) or South (s), site location

Species: fish species

Common = common triplefin, Forsterygion lapillum
Yaldwyn = Yaldwyn’s triplefin, Notoclinops yaldwyni
Blue_eyed = blue-eyed triplefin, Notoclinops segmentatus
Blenny = crested blenny, Parablennius laticlavius

We used the same data as per previous section.

Isotopes C and Isotopes V:

iso1= δ13C
iso2= δ15N

Diet V and Diet C:

For stomach content: we used only volumetric stomach content data as inclusive of all species of interest. It is not raw data, but we used the reduced dimension obtained from nonmetric multidimensional scaling (nMDS), thus the 2 columns resulting from this analysis are vol1 and vol2. Raw data are in the datasheet Stomach volumetric in the “Raw_single_niche_data” dataset.

Habitat association C and Habitat association V / Habitat - C and Habitat - V

For Habitat association, the columns are id, species, habitat and position. The habitat association for each species is categorical based on habitat occupied and position (e.g., turf - vertical). Information for Crested blenny were extracted from the behavioural video recordings (with each video being a replicate). The dataset is then linked to Habitat cover in both control (C) and vent (V) sites to determine the choice of the habitat based on habitat availability. Therefore, the habitat cover only presents the percentage cover of each habitat type at control and vent.

Habitat:

turf = turf algae <10 cm in height

ca = erect calcareous algae

cca = crustose coralline algae

barren = bare rocky substratum

sp = encrusting fleshy green algae

cobble = cobbles (~0.5–2 cm in diameter)

sand = sand

Position:

hor = horizontal

angle = angled

vert = vertical

Behaviour C and Behaviour V: Behavioural variables quantified from underwater footage and expressed as rates per minute. The behaviours are: swimming, jumping, feeding, attacking and fleeing from an attack.

Reference:

Geange, S. W., Pledger, S., Burns, K. C., & Shima, J. S. (2011). A unified analysis of niche overlap incorporating data of different types. Methods in Ecology and Evolution, 2(2), 175-184. https://doi.org/10.1111/j.2041-210X.2010.00070.x

Methods

We used a small hand net and a mixture of ethanol and clove oil to collect the four species of interest (Forsterygion lapillum, Notoclinops yaldwyni, Notoclinops segmentatus and Parablennius laticlavius) at both control and vent sites over four years.

For stable isotope analysis, white muscle tissue was extracted from each fish and oven-dried at 60 °C. The dried tissue was subsequently ground using a ball mill. Powdered muscle tissue from each fish was individually weighed into tin capsules and analysed for stable δ¹⁵N and δ¹³C isotopes. Samples were combusted in an elemental analyser (EuroVector, EuroEA) coupled to a mass spectrometer (Nu Instruments Horizon) at the University of Adelaide. We then analysed the isotopic niche in SIBER.

For stomach content analysis the entire gut was extracted from each fish. Using a stereomicroscope, for count and biomass, all prey items in the stomach were counted first. For each prey category, well-preserved individuals were photographed and their mass was calculated based on length and width. The average mass per individual for each category was then multiplied by the count to determine total prey biomass. For the volumetric method, the volume contribution of each prey category relative to the total stomach content was visually estimated (algae were accounted for). Digested items that were not identifiable were excluded from the analysis. Each stomach content dataset was reduced to two dimensions with non-metric multidimensional scaling (nMDS) to be then analysed in SIBER.

To assess habitat choice, visual surveys were conducted on SCUBA, to record the microhabitat type and orientation occupied by Forsterygion lapillum, Notoclinops yaldwyni and Notoclinops segmentatus. The resulting dataset comprised a total of 17 distinct combinations of habitat types and surface orientations. The dataset was simplified to two dimensions using correspondence analysis (CA) for subsequent SIBER analysis.

Fish behaviour was assessed using GoPro cameras both in situ and during controlled aquarium experiments. In the field, recordings lasted 30 minutes across 4 days, with analysis conducted using VLC. Initial acclimation and periodic intervals (10 minutes every 5 minutes) were excluded from analysis. In controlled aquarium settings, individuals of Forsterygion lapillum and Notoclinops yaldwyni were observed both in isolation and paired. Their habitat preference, surface orientation, and activity levels were recorded for 10 minutes to assess behaviour independent of external influences. Both datasets were dimensionally reduced for analysis in SIBER: non-metric multidimensional scaling (nMDS) was applied to the in situ behavioral data, while principal component analysis (PCA) was used for the aquarium experiments.

Unified analysis of niche overlap

We quantified the local realised niche space for each fish species at control and vent along the four niche classes, adapting the data as follows:

isotopes (continuous data): raw data.
stomach content (continuous data): reduced dimension from the volumetric measure of the previous step.
habitat association (elective score): habitat and orientation preference linked to Manly’s Alpha association matrix.
behaviour (continuous data): raw data.