Ontogenetic habitat shifts in Nassau grouper: growth, predation risk, and the minimize μ/g hypothesis

Introduction

• Study investigates why juvenile Nassau grouper Epinephelus striatus shift habitats during ontogeny (early life stages) on Bahamian coral reefs.
• Uses an optimality framework: mobile animals balance growth (for energy gain) and predation risk (mortality) across habitats.
• Three competing hypotheses tested via field experiments and cost–benefit analysis:
  • (1) Maximize growth rate g.
  • (2) Minimize mortality risk μ (predation).
  • (3) Minimize the ratio μ/g (trade-off between mortality risk and growth).
• Focus on off-reef nursery habitats: algal habitat (interstices of macroalgal clumps) vs postalgal habitats (areas outside/interstices, adjacent to macroalgae and complex microhabitats).
• Ontogenetic shift observed around 50 mm TL: juveniles move from algal nurseries to postalgal habitats, then to patch reefs and deeper offshore reefs.
• Key prediction tested: habitat use should minimize μ/g if trade-offs exist; results supported this in a size-dependent way.
• Relevance: clarifies how size-dependent predation risk and growth opportunities shape distribution of mobile marine species; implications for population dynamics and conservation in coral reef systems.

Species, Habitats, and Life History

• Nassau grouper (Epinephelus striatus) is a large serranid, important commercially in the Caribbean.
• Juvenile recruitment occurs off-reef in macroalgal beds (Laurencia sp.) where they stay for ~2 months post-settlement as small juveniles (TL ~35–40 mm).
• Ontogenetic habitat shifts observed:
  • Algal habitat: interstices of macroalgal clumps; refuge and possible food source for very small juveniles.
  • Postalgal habitat: outside/interstices but within macroalgal beds; includes coral, rubble, sponges, edges of Laurencia, etc.; used by mid-sized juveniles (~50–75 mm TL).
  • Later shifts: late juvenile shift to patch reefs and then offshore reefs.
• Field sites: Great Exuma Island, Bahamas; CMRC, Lee Stocking Island. Tidal creeks near CMRC with ~7–200 m^2 macroalgal beds; temperatures 20–35°C during experiments.
• Key macroalgae: Laurencia sp. providing structure and potential prey refuges; other habitats include Porites porites, seagrass Thalassia testudinum, sponges, crevices, and solution holes.
• Sizes studied (approximate TL):
  • Small: 35–40 mm TL
  • Medium: 50–55 mm TL
  • Large: 70–75 mm TL
• Off-reef nurseries and ontogenetic shifts are common in reef systems; this study contributes quantitative tests of optimization models for marine nursery habitat choice.

Theoretical Framework and Hypotheses

• Optimal habitat choice is driven by changing mortality risk and growth potential with size:
  • Foraging/growth vs predation risk often trades off across habitats.
  • Simple formulations predict animals should move to habitats that maximize net benefits (growth) or minimize risk, or balance both as μ/g.
• Key equations:
  • Growth rate in habitat h for size s: g_{s,h}
  • Mortality risk (predation) in habitat h for size s: bc_{s,h}
  • Ratio of mortality risk to growth rate: R{s,h} = rac{bc{s,h}}{g_{s,h}}
• Hypotheses tested:
  • Maximize g: shift to habitat with largest g_{s,h} across habitats.
  • Minimize μ: shift to habitat with smallest bc_{s,h}.
  • Minimize μ/g: shift to habitat with smallest R{s,h} = bc{s,h}/g_{s,h}.
• The study also considers that diurnal vs nocturnal habitat use could bias growth estimates if patterns differ between day and night; thus nocturnal-diurnal checks were included.
• The underlying model mirrors classic foraging–risk trade-offs (e.g., Werner & Gilliam 1984; Gilliam & Fraser 1987) and has been extended to ontogenetic habitat shifts in marine systems (e.g., Sogard; Utne et al.; Gilliam, Fraser, etc.).

Methods Overview

• Study design combines field cages to quantify habitat-specific growth (g) and tethering experiments to quantify habitat-specific predation risk (μ).
• Three size classes were tested to bracket the ontogenetic shift from algal to postalgal habitats.
• Study sites: two off-reef tidal creek sites (B3 and B4) near CMRC on Lee Stocking Island, Bahamas.
• Field durations: 1995–1998 with multiple field seasons; growth experiments conducted winter, spring, and summer 1996 for small, medium, large sizes respectively.
• Nocturnal/diurnal habitat use tested in laboratory assays to validate day–night patterns when fish are confined to either algal or postalgal habitats.

Methods: Growth Rate (g) Estimation

• Cage experiments to prevent movement between habitats and exclude predators:
  • Cage design: circular, radius 0.6 m, height 0.7 m; bottom open; top/side mesh 6.35 mm; anchored to substrate; 1.2 m^2 area per cage.
  • Cages deployed in macroalgal beds with 40–60% macroalgal cover; placed ~10 m apart; six replicates per habitat treatment per site.
  • Treatments:
  • Algal: access confined to interstices of macroalgal clumps (Laurencia). Achieved by placing Laurencia inside a mesh bag inside the cage.
  • Postalgal: access limited to postalgal microhabitats; interstices of macroalgae blocked by mesh bag.
  • Control: access to both algal and postalgal habitats (no mesh restriction beyond cage enclosure).
  • Fish sizes and count: 1 fish per cage; six replicates per treatment per site; two study sites (B3, B4).
  • Duration: 6–7 weeks per experiment; weekly checks; growth measured as tl (mm) gained over time; daily growth rate g = (TLfinal - TLinitial) / days caged (mm/day).
  • End-of-experiment macroalgal assessments: percent cover and displacement volume of macroalgae in cages; macroalgal volume measured for algal and postalgal treatments (Bag-based for algal/postalgal; suction sampling for controls).
• Experimental design and analysis:
  • Randomized complete block ANOVA with Site as fixed block and Habitat as main factor (Algal, Postalgal, Control).
  • Post-hoc contrasts: (1) control not different from the habitat where the size class naturally occurs; (2) control and natural habitat have higher growth than the alternative habitat for that size class.
  • Homogeneity of variances tested with Fmax; normality assumed via standard ANOVA checks.
• Size classes and expected natural habitats:
  • Small fish (35–40 mm TL): typically algae-dwelling; natural habitat is algal.
  • Medium fish (50–55 mm TL): typically postalgal habitat.
  • Large fish (70–75 mm TL): typically postalgal habitat.

Methods: Mortality Risk Estimation (μ)

• Mortality risk quantified via tethering experiments (predation proxy).
• Tethering setup:
  • Tether through lower jaw with 30 cm of 0.009-inch monofilament; tether length gives ~2,800 cm^2 movement area.
  • 20 juvenile Nassau grouper per size class per site per trial; 10 tethered in algal habitat and 10 tethered in postalgal habitat; 10 m apart within macroalgal beds with 40–50% macroalgal cover.
  • Algal tethered fish: inside macroalgal clumps (Laurencia sp.). Postalgal tethered fish: adjacent to macroalgae but prevented from using interstices.
  • Check every 2 hours during daylight; terminated by sunset (~1815–1915); end-status recorded as present or missing (assumed mortality if missing).
  • Each 12-hour tethering experiment conducted twice within a five-day window for each size class; TLs at tether release: small ~39.0 ± 3.0 mm; medium ~53.5 ± 2.3 mm; large ~73.0 ± 2.9 mm (means ± SE).
• Mortality risk measurement:
  • μ_h,s is the proportion dead per day in habitat h for size class s, inferred from missing fish at end of tethering trial.
  • Comparative tests between habitats use loglinear G-tests to evaluate differences in missing fish numbers between algal and postalgal treatments for each size class.
• Tethering artifacts checks (simple vs higher-order artifacts):
  • Lab observations to detect adverse tethering effects on health/behavior (n ≈ 10 tethered vs 10 untethered, ~1.5 weeks).
  • Field artifacts check: compare tethered vs untethered behavior near macroalgal clumps; assess escape tendency; test tether breakage with predator-exclusion cages.
  • Additional test: test tethering effects on movement rates; standardize movements per hour; compare between tethered and untethered using t-tests.
• Purpose: determine habitat-specific μ for small, medium, and large fish in algal vs postalgal habitats.

Methods: Nocturnal vs Diurnal Habitat Use (Laboratory Check)

• Rationale: field observations of diurnal habitat usage may differ from nocturnal usage, potentially biasing growth estimates if fish switch habitats at night.
• Experimental setup:
  • Laboratory aquaria with macroalgal clumps (Laurencia sp.) of ~400 mL displacement, cleaned of prey items.
  • Clumps placed in a 154-L aquarium to simulate algal patches; a separate, separate area allowed access to postalgal microhabitats; prey-free algal mimics to ensure prey availability was controlled.
  • One fish per aquarium; 10 fish per size class; observations during day (0800–1800) and night (2000–0600) under red light to minimize disturbance.
  • Prey removed from algae to avoid food distribution confounding habitat choice; aim to isolate habitat preference from prey availability.
• Observation method:
  • Hourly checks of fish position relative to algae; assume fish outside 5 cm of algae if not observed under algae.
  • Night observations used red-light to minimize behavioral disturbance; daytime observations limited to ~20 s per aquarium.
• Outcome: assess whether nocturnal vs diurnal patterns align with daily habitat use observed in the field for each size class.

Methods: Data Analysis Details

• Growth data (g) analyses:
  • For each size class, test if g differs among habitats (Algal, Postalgal, Control) using randomized complete block ANOVA with Site as blocking factor.
  • Check homogeneity of variances with Fmax test; if violated, interpretation cautious.
  • Orthogonal contrasts for a priori hypotheses:
  • H1: gcontrol ≈ gnaturalhabitat, and gcontrol, gnaturalhabitat > gotherhabitat.
  • Size-specific expectations: Small: algal < postalgal, control and algal alike; Medium/Large: postalgal < algal? (specific ordering reported in study).
• Macroalgal cover and volume checks:
  • End-of-experiment macroalgal percent cover and displacement volume measured per cage.
  • Tests for differences among habitat treatments within each site and size class using Ryan’s Q test following ANOVA.
• Nocturnal artifacts checks:
  • Compare day vs night habitat use frequencies using G tests (loglinear) by size class.
• Mortality risk data (μ) checks:
  • Compare numbers missing between algal vs postalgal tethering treatments for each size class using loglinear G tests.
• μ/g ratio analysis:
  • Compute habitat-specific μ/g for each size class:
  • For small fish: μ/g lower in algal habitat; for medium/large fish: μ/g lower in postalgal habitat.
  • Use a one-tailed randomization test (Manly 1997) to test the hypothesis that the habitat minimizes μ/g relative to the alternative:
  • Generate 5,000 random reallocations of μ and g values across habitats for each size class to form a null distribution of ΔR = (μalg/galg) − (μpostal/gpostal).
  • Compare experimentally observed ΔR to the null distribution; reject null if 95% of simulated ΔR values are on the opposite side of zero relative to the observed sign.
• Randomization test details (Fig. 5):
  • For each size class (small, medium, large), compare ΔR_obs to the 5,000 simulated ΔR distributions; significant at α = 0.05 if observed value lies outside the central 95% of simulated values, consistent with directional expectation (negative for small; positive for medium/large).

Results: Nocturnal vs Diurnal Habitat Use

• Nocturnal vs diurnal patterns in the lab mirrored field day use:
  • Small fish spent more time in macroalgae at night (≈99% observations) than during the day (≈87%): G-test shows P = 0.001 (n ≈ 22).
  • Medium fish: day 60%, night 67% (within algal clump when observed); large fish: day 44%, night 36%.
• Overall, nocturnal use did not reverse the diurnal pattern sufficiently to invalidate the experimental design; diurnal/night patterns were similar enough to justify confining fishes to their typical habitats in both day and night for growth experiments.

Results: Size- and Habitat-Specific Growth (g)

• Growth rates overall:
  • For all three size classes, growth rates were higher in postalgal and control habitats than in the algal habitat (galgal < gpostalgal and gcontrol ≈ gpostalgal > g_algal; significant differences detected via ANOVA and contrasts).
• Site effects:
  • Small fish showed site-by-habitat interaction: growth at site B3 was higher than at B4 (B3: ≈0.23 mm/d; B4: ≈0.18 mm/d).
• Caging artifact checks:
  • Macroalgal cover and volume varied modestly among treatments and sites, but these were not associated with spurious differences in growth among habitat treatments for small/medium fish.
  • End-of-experiment macroalgal cover differences:
  • Small fish: algal cover ≈50.2% (Alg). Postalgal ≈50.0% (Postal). Control ≈60.5%.
  • These differences were not sufficient to explain the observed growth differences; displacement volume differences across sites were within natural ranges.
  • Macroalgal volume differences across sites: B3 had higher macroalgal volume than B4 for medium fish; nonetheless, growth differences aligned with habitat treatment rather than volume alone.
• Overall conclusion for growth: small, medium, and large fish all grew fastest in postalgal or control habitats; algal habitat restricted growth for all sizes, though the magnitude and site dependence varied by size.

Results: Mortality Risk (μ) via Tethering

• Predation risk results by size class (two tethering experiments at site B3):
  • Small fish: relative predation risk was significantly lower in the algal habitat than postalgal habitat (G = 3.96, P = 0.05).
  • Medium fish: no significant difference between habitats (G = 0.50, P > 0.05).
  • Large fish: no significant difference between habitats (G = 0.03, P > 0.05).
• Overall interpretation: small juveniles experience higher predation risk in postalgal habitats, while larger juveniles do not show strong habitat-specific differences in short tethering trials.
• Tethering artifacts checks:
  • No evidence that tethering affected fish condition or behavior in a way that biased predation estimates; lab assessments showed no deaths or injuries in tethered fish over ~1.5 weeks; tethered vs untethered movement rates were not significantly different (t-test, P > 0.05).
  • Predator exclusions confirmed that tethering artifacts of simple type were minimal; higher-order artifacts (interactions between tethering and habitat) were assessed by comparing tethered animal behavior across habitats and found not to bias results.

Results: μ/g Ratios and Randomization Test

• Habitat-specific μ/g ratios for each size class:
  • Small fish: μ/g lower in algal habitat than postalgal habitat (suggesting algal habitat reduces mortality relative to growth benefits, favoring μ/g minimization in algae when small).
  • Medium and Large fish: μ/g lower in postalgal habitat than in algal habitat (postalgal provides better growth context with similar or lower mortality risk for larger fish).
• Randomization test for μ/g differences (5,000 permutations per size class):
  • Small: observed Δ(μ/g)alg - Δ(μ/g)postal more negative than 95% of randomizations; significant at α = 0.05, supporting algal habitat as minimizing μ/g for small.
  • Medium/Large: observed Δ(μ/g) larger than most randomizations in the opposite direction, significant at α = 0.05, supporting postalgal habitat as minimizing μ/g for larger fish.
• Figure 4 (habitat-specific μ/g): shows the pattern above for each size, with the algal habitat favored for small fish and postalgal habitat favored for medium and large fish.
• Figure 5 (randomization distributions): demonstrates that the observed μ/g differences are unlikely due to random allocation of μ and g values; the observed values fall in the tails of the simulated distributions, indicating statistical significance.

Interpretation: Ontogenetic Habitat Shifts and Mechanisms

• Overall conclusion: observed ontogenetic shifts in Nassau grouper are best explained by the minimize μ/g hypothesis, not solely by maximizing growth or minimizing mortality alone.
• Mechanisms by size class:
  • Small fish (35–40 mm TL): higher predation risk in postalgal habitat; macroalgal refuge provides safer survival, even if growth is slower there. This leads to an overall lower μ/g in algal habitats for small fish.
  • Medium and Large fish (50–75 mm TL): growth potential is higher in postalgal habitats due to foraging opportunities; predation risk differences are smaller, but growth rate advantages in postalgal habitats yield lower μ/g there.
• Ecological interpretation:
  • Macroalgal refuges provide size-dependent predation protection that is crucial for early survival; as fish grow, the relative value of foraging in postalgal habitats increases, shifting the optimal habitat towards those areas.
  • The results demonstrate a functional link between body size, predation risk, and growth potential that drives ontogenetic habitat shifts in a marine nursery system.
• Broader implications:
  • Supports a general framework where ontogenetic habitat choice can be understood as a dynamic optimization problem balancing survival and growth.
  • Highlights potential sublethal and density-dependent effects of confinement to suboptimal foraging habitats (e.g., slower growth, altered predator–prey dynamics).
  • Macroalgal nurseries, while providing refuge, are relatively patchy; their availability can influence population structure and metapopulation dynamics (source–sink dynamics) in Nassau grouper and similar reef systems.

Discussion: Implications for Population Ecology and Conservation

• The study emphasizes that habitat-based decisions during juvenile stages can shape population trajectories by altering growth, survival, and density dependence.
• Macroalgal nurseries may act as a population bottleneck or a refugial source depending on spatial availability and recruitment patterns; this has implications for management under habitat alteration or climate change.
• The framework and methods offer a template for testing ontogenetic habitat shifts in other marine systems where predator–prey interactions and growth trade-offs vary with size.
• Practical considerations:
  • Conservation strategies should consider protecting macroalgal nursery habitats as potential sources of juvenile Nassau grouper survivorship.
  • Habitat connectivity between off-reef nurseries and on-reef juvenile habitats may be critical for maintaining population resilience.

Key Concepts, Terms, and Takeaways

• Ontogenetic habitat shift: changes in habitat use as organisms grow and their ecological needs change.
• Algal habitat: off-reef macroalgal interstices providing refuge and potential for foraging in juvenile Nassau grouper.
• Postalgal habitat: off-reef microhabitats outside macroalgal interstices but within macroalgal beds; includes coral, rubble, sponges, etc.
• Growth rate (g): mean daily increase in total length (mm/d) under specific habitat and size class.
• Mortality risk (μ): probability of death due to predation per day in a given habitat and size class.
• μ/g ratio (R): the trade-off metric used to predict habitat choice; lower R indicates a preferred habitat under a trade-off between growth and predation risk.
• Randomization test: nonparametric approach (5,000 simulations) to assess whether observed μ/g differences are likely due to chance.
• Site blocking: experimental design feature using two sites (B3, B4) to control for spatial variability.
• Experimental artifacts: checks for confounding effects from caging or tethering (e.g., prey availability, mesh effects, tethering bias).
• Macroalgal refuges: structural habitat features that may reduce predation risk, particularly for small juveniles.
• Source–sink dynamics: metapopulation concept relevant when patch quality and recruitment influence population persistence across habitats.

Summary of Quantitative Details (Selected)

• Sizes studied: small 35–40 mm TL; medium 50–55 mm TL; large 70–75 mm TL.
• Duration: growth experiments last 6–7 weeks; tethering experiments last 12 hours per trial, conducted twice within a five-day period per size class.
• Cage dimensions: radius 0.6 m, height 0.7 m; area ~1.2 m^2.
• Macroalgae used: Laurencia sp.; macroalgal cover and volume quantified at experiment end; B3 had larger macroalgal volume than B4 for certain size classes.
• Growth results (qualitative): gpostalgal ≈ gcontrol > g_algal for all size classes; small fish showed site-specific growth differences (B3 > B4).
• Mortality results (μ): small fish showed lower predation in algal habitat (P ≈ 0.05); medium/large fish showed no significant habitat differences in μ.
• μ/g results: small fish favor algal habitat (lower μ/g); medium/large fish favor postalgal habitat (lower μ/g).
• Randomization test: 5,000 iterations per size class; all size classes showed significant μ/g differences consistent with the observed pattern (α = 0.05).

References to Figures and Tables (as described in the transcript)

• Fig. 1: Study site map (B3 and B4 at Great Exuma Island; reference sites B1, B2).
• Fig. 2: Habitat-specific growth rate distributions (g) by size class and habitat; significant contrasts indicated.
• Fig. 3: Habitat-specific predation rates from tethering experiments by size class.
• Fig. 4: Habitat-specific μ/g ratios by size class.
• Fig. 5: Randomization test distributions for μ/g differences vs observed values.
• Table 1: ANOVA results for growth rates, percent macroalgal cover, and macroalgal volume by size class and habitat.
• Table 2: Summary of size- and habitat-specific growth rates (g), mortality rates (μ), and μ/g; primary mechanism by size and habitat.

Equations (LaTeX)

• Growth rate and mortality ratio per size-class per habitat:
  • Growth rate: g_{s,h} ext{ (mm d}^{-1})
  • Mortality risk: bc_{s,h} ext{ (proportion dead per day)}
  • Ratio: R{s,h} = rac{bc{s,h}}{g_{s,h}}
• Hypotheses (decision rules):
  • Maximize growth: choose habitat h with g_{s,h} = \u2211 ext{max}
  • Minimize mortality: choose habitat h with bc{s,h} = bc{s, ext{min}}
  • Minimize mortality-to-growth: choose habitat h with R{s,h} = rac{bc{s,h}}{g_{s,h}} = ext{min}
• Randomization test statistic (illustrative):
  • Let ext{Δ}R = R{s, ext{Algal}} - R{s, ext{Postalgal}}
  • Null distribution built by permuting μ and g values across habitats 5{,}000 times; p-value is the proportion of simulated ΔR that are as extreme or more extreme than observed.

Takeaway for Exam Preparation

• The Nassau grouper study provides a clear empirical test of ontogenetic habitat shifts through the lens of optimize-arg min μ/g rather than solely maximizing growth or minimizing mortality.
• Key takeaways include size-dependent refugia in macroalgal habitats, growth-rate trade-offs across habitats, and the importance of integrating growth and predation risk to predict habitat choice.
• Methodological strengths: combining field caging with tethering, plus lab nocturnal/diurnal checks, rigorous artifact controls, and a robust randomization test framework.
• Broader relevance: supports general ecology of size-structured habitat selection, predator–prey interactions, and the design of marine reserve or habitat-conservation plans that account for ontogenetic habitat needs.