%%{init: {"flowchart": {"htmlLabels": true, "nodeSpacing": 58, "rankSpacing": 82, "curve": "basis"}, "themeVariables": {"fontSize": "18px", "fontFamily": "Arial, sans-serif"}} }%%
flowchart TB
SSB["Spawning stock<br/>biomass"] --> EG["Egg<br/>abundance"]
TEMP["Seasonal<br/>temperature"] --> DEV["Development rate<br/>and hatch timing"]
TEMP --> SPAWN["Spawning timing<br/>and location"]
ICE["Sea ice extent<br/>and retreat timing"] --> BLOOM["Spring bloom timing<br/>and location"]
WIND["Wind, storms,<br/>and circulation"] --> TRANS["Egg and larval<br/>transport"]
WIND --> MIX["Summer mixing<br/>and nutrient flux"]
SPAWN --> ELDIST["Egg and larval<br/>distribution"]
EG --> ELDIST
DEV --> LARV["Larval duration<br/>and growth"]
TRANS --> ELDIST
BLOOM --> ZOO["Copepods and<br/>euphausiids"]
MIX --> PP["Summer primary<br/>production"]
PP --> ZOO
ELDIST --> PREYMATCH["Spatial match<br/>with prey"]
ZOO --> PREYMATCH
PREYMATCH --> LARV
LARV --> A0SIZE["Age-0<br/>size"]
ZOO --> A0DIET["Age-0 diet<br/>lipid"]
A0DIET --> A0ENERGY["Age-0 energy density<br/>and total energy"]
A0SIZE --> WINTER["Overwinter survival<br/>to age 1"]
A0ENERGY --> WINTER
SST["Late-summer SST<br/>and metabolic demand"] --> A0ENERGY
SST --> WINTER
ADULT["Adult pollock<br/>biomass"] --> CANN["Cannibalism<br/>pressure"]
COLD["Cold pool<br/>area"] --> PREDIST["Predator<br/>distribution"]
ARROW["Arrowtooth and other<br/>predator biomass"] --> PRED["Predation<br/>pressure"]
ELDIST --> JUVLOC["Age-0 and age-1<br/>distribution"]
JUVLOC --> OVERLAP["Juvenile-predator<br/>spatial overlap"]
PREDIST --> OVERLAP
CANN --> OVERLAP
OVERLAP --> A12SURV["Age-1 and age-2<br/>survival"]
PRED --> A12SURV
WINTER --> A12SURV
A12SURV --> AGE3["Age-3<br/>recruitment"]
classDef spawner fill:#e8f2f7,stroke:#26739b,stroke-width:2.4px,color:#1f2933;
classDef physical fill:#e8f2f7,stroke:#26739b,stroke-width:2.4px,color:#1f2933;
classDef prey fill:#ecf5eb,stroke:#6b8f24,stroke-width:2.4px,color:#1f2933;
classDef early fill:#fff3df,stroke:#c68519,stroke-width:2.4px,color:#1f2933;
classDef predator fill:#fdf2e8,stroke:#c8523f,stroke-width:2.4px,color:#1f2933;
classDef survival fill:#e8f6f2,stroke:#13856f,stroke-width:2.4px,color:#1f2933;
class SSB,EG spawner;
class TEMP,ICE,WIND,SST,COLD,DEV,SPAWN,TRANS,MIX physical;
class BLOOM,PP,ZOO,PREYMATCH,A0DIET prey;
class ELDIST,LARV,A0SIZE,A0ENERGY,JUVLOC early;
class ADULT,CANN,ARROW,PRED,PREDIST,OVERLAP predator;
class WINTER,A12SURV,AGE3 survival;
Eastern Bering Sea Walleye Pollock Early Survival Pathways
This site summarizes causal hypotheses for eastern Bering Sea walleye pollock survival from eggs and larvae through age-0, age-1, and age-2 fish reaching age 3. The synthesis is based on the local paper set in the parent reprints folder, with web-visible citation links supplied through DOI and source URLs.
The papers often quantify recruitment at age 1 or age 2. Here, age-3 recruitment means the same cohort passing through the prerecruit filters and remaining alive and available at age 3.
Main Conclusion
Two mechanisms have the clearest support:
- Age-0 energy gate. Late-summer and fall condition, especially total energy, lipid reserves, and access to lipid-rich copepods and euphausiids, strongly affects overwinter survival to age 1 (Heintz et al. 2013; Siddon et al. 2013; Sogard and Olla 2000; Sigler et al. 2016).
- Age-1 predation gate. Juvenile survival after the first winter is strongly affected by spatial overlap with adult pollock, arrowtooth flounder, and other predators, with age-1 mortality particularly important (Mueter et al. 2006; Spencer et al. 2016).
Temperature, sea ice, stratification, summer wind mixing, and transport matter because they modify these gates. Average temperature alone is a useful proxy only when the active pathway is unresolved (Smart et al. 2012; Petrik et al. 2015; Gann et al. 2016).
Integrated Causal Map
Site Contents
- Evidence: stage-specific synthesis and candidate indicators.
- DAGs: alternative directed acyclic graphs for competing hypotheses.
- PCMCI+: implementation plan for time-series causal discovery.
- Infographics: compact visual summaries for a science audience.
- References: cited references with DOI and source links.
Practical Models
A practical model does not need to include every process in the causal map. It should preserve the main stage structure: spawner output sets the starting cohort size, age-0 condition filters overwinter survival, and age-1 to age-2 predator overlap filters survival to age 3.
The models below are useful as a progression from a minimum assessment-linked model to a fuller causal model.
| Model | Form | Best use |
|---|---|---|
| 1. Baseline stock-recruit model | Age3_recruitment = f(spawner output, cohort density, assessment uncertainty) |
Establishes the baseline recruitment expectation and residual pattern before adding environmental or ecological covariates. |
| 2. Age-0 condition model | Age3_recruitment = f(spawner output, age0 total energy or energy density, cohort density, uncertainty) |
Tests whether fall juvenile size, lipid, or total energy explains later recruitment after accounting for spawner output. This is the most direct practical expression of the age-0 energy gate. |
| 3. Predator-overlap model | Age3_recruitment = f(spawner output, juvenile-predator overlap, adult pollock biomass, predator biomass, uncertainty) |
Tests whether survival from age 1 to age 3 is limited by cannibalism or other predators. Adult pollock biomass and arrowtooth flounder distribution are treated as drivers of spatially explicit predation risk. |
| 4. Transport and nursery-delivery model | Age3_recruitment = f(spawner output, spawning location, larval transport, prey match, early survival, uncertainty) |
Tests whether cohorts succeed because eggs and larvae are delivered to prey-rich and predator-safe nursery habitat. This model is most useful when particle tracking, egg/larval surveys, or spawning-location products are available. |
| 5. Integrated gate model | Age3_recruitment = f(spawner output, age0 condition, age1-age2 predator overlap, cohort density, uncertainty) |
Combines the two best-supported survival gates. Age-0 condition mediates ice timing, bloom timing, summer mixing, prey production, and SST. Predator overlap mediates cold pool area, predator biomass, juvenile distribution, and adult pollock biomass. |
For annual prediction, model 5 is the minimum useful causal model. For diagnosis, models 2 through 4 should also be fit separately so that a strong condition signal is not hidden by a predation signal, or vice versa. The transport model is especially useful as a bridge between climate forcing and the two survival gates because it determines where larvae and juveniles encounter prey and predators.