%%{init: {"flowchart": {"htmlLabels": true, "nodeSpacing": 60, "rankSpacing": 80, "curve": "basis"}, "themeVariables": {"fontSize": "18px", "fontFamily": "Arial, sans-serif"}} }%%
flowchart TB
ICE["Ice retreat<br/>timing"] --> BLOOM["Bloom timing<br/>and location"]
WIND["Summer storms<br/>and wind mixing"] --> NUT["Surface nutrient<br/>flux"]
NUT --> PP["Summer primary<br/>production"]
BLOOM --> LZOO["Large copepods<br/>and euphausiids"]
PP --> LZOO
LZOO --> DIET["Age-0 diet<br/>lipid"]
DIET --> ED["Age-0 energy<br/>density"]
GROW["Age-0 growth<br/>and size"] --> TE["Age-0 total<br/>energy"]
ED --> TE
TE --> WINS["Winter survival<br/>to age 1"]
WINS --> A12["Age-1 and age-2<br/>abundance"]
A12 --> R3["Age-3<br/>recruitment"]
classDef forcing fill:#e8f2f7,stroke:#26739b,stroke-width:2.4px,color:#1f2933;
classDef production fill:#ecf5eb,stroke:#6b8f24,stroke-width:2.4px,color:#1f2933;
classDef condition fill:#fff3df,stroke:#c68519,stroke-width:2.4px,color:#1f2933;
classDef survival fill:#e8f6f2,stroke:#13856f,stroke-width:2.4px,color:#1f2933;
class ICE,WIND,NUT forcing;
class BLOOM,PP,LZOO,DIET production;
class GROW,ED,TE condition;
class WINS,A12,R3 survival;
DAG Alternatives
Directed acyclic graph alternatives for hypotheses affecting prerecruit survival.
Each DAG is deliberately simpler than the integrated map. The purpose is to define testable alternatives, not to draw every plausible link.
Data Availability And Gaps By Hypothesis
The DAGs are useful only if each causal path can be tied to observable indicators. The table below separates data that are already available or routinely derivable from data that are partial, missing, or most needed for a recruitment model.
| Hypothesis | Available or routinely derivable data | Partial, missing, or needed data |
|---|---|---|
| A. Bottom-up energy storage | Age-0 length, weight, energy density, percent lipid, and diet composition from BASIS and related sampling; ice retreat, SST, wind mixing, stratification, nutrient, and bloom indicators; copepod/euphausiid observations from zooplankton surveys and targeted studies (Heintz et al. 2013; Moss et al. 2009; Siddon et al. 2013; Sigler et al. 2016). | A consistent annual age-0 total-energy index across warm and cold regimes; spatially matched prey-energy fields at juvenile sampling locations; fall-to-winter condition change and overwinter prey availability. Highest-value product: cohort-year age-0 total energy and diet-lipid index aligned with age-1 and age-3 recruitment. |
| B. Top-down predation and spatial overlap | Adult pollock biomass from assessments and surveys; predator abundance and distribution from shelf surveys; candidate overlap metrics; published predation and spatial-overlap models for adult pollock and arrowtooth flounder (Mueter et al. 2006; Spencer et al. 2016). | Annual juvenile age-1 and age-2 spatial distributions with uncertainty; year-specific predator diet, ration, and size-selective mortality estimates; separation of cannibalism from other predator mortality. Highest-value product: annual juvenile-predator overlap indices by predator group and juvenile age. |
| C. Transport and spawning-location match-mismatch | Egg/larval survey records, adult spawning distribution information, wind and circulation fields, temperature-dependent development rates, and biophysical transport model examples (Petrik et al. 2015; Smart et al. 2012). | Annual spawning-location products with uncertainty; coupled particle-tracking outputs for all candidate years; observed prey and predator fields collocated with modeled larval paths; direct estimates of early-stage mortality versus redistribution. Highest-value product: cohort-specific larval delivery index to prey-rich and predator-safe nursery habitat. |
| D. Time-indexed switching control | Time series for recruitment, SSB, adult biomass, broad climate state, ice/SST, predator biomass, and some prey and juvenile indicators; published evidence that bottom-up and top-down controls interact (Mueter et al. 2006, 2011; Sigler et al. 2016). | Long, consistent time series for age-0 condition, age-1/age-2 survival, predator overlap, and prey production on the same cohort calendar; sufficient contrast across consecutive climate states; explicit uncertainty propagation across lagged links. Highest-value product: cohort table with synchronized annual covariates from spawning year through age 3. |
Across all alternatives, the largest gap is not a complete absence of observations. It is inconsistent annual coverage across the full cohort pathway: eggs and larvae, fall age-0 condition, overwinter survival to age 1, age-1 and age-2 spatial overlap with predators, and final age-3 recruitment are rarely observed as one linked cohort record.
DAG A: Bottom-Up Energy Storage
Main claim: recruitment is set primarily by whether seasonal production creates lipid-rich prey and age-0 fish accumulate enough total energy before winter (Heintz et al. 2013; Siddon et al. 2013; Sigler et al. 2016).
Strong predictions:
- Fall total energy should mediate most environment-recruitment relationships.
- Cold years with low summer nutrient flux can fail despite favorable ice timing.
- Age-0 diet lipid, Calanus/euphausiid biomass, and age-0 energy density should be leading indicators.
DAG B: Top-Down Predation And Spatial Overlap
Main claim: recruitment is set primarily by whether juveniles overlap spatially with adult pollock, arrowtooth flounder, and other predators during age-0 to age-2 (Mueter et al. 2006; Spencer et al. 2016).
%%{init: {"flowchart": {"htmlLabels": true, "nodeSpacing": 58, "rankSpacing": 78, "curve": "basis"}, "themeVariables": {"fontSize": "18px", "fontFamily": "Arial, sans-serif"}} }%%
flowchart TB
ADULT["Adult pollock<br/>biomass"] --> CANN["Cannibalism<br/>pressure"]
ARROW["Arrowtooth<br/>biomass"] --> ATP["Arrowtooth predation<br/>pressure"]
COLD["Cold pool<br/>extent"] --> ATDIST["Arrowtooth<br/>distribution"]
TEMP["Temperature"] --> ATDIST
TRAN["Larval transport<br/>and juvenile distribution"] --> JLOC["Juvenile<br/>location"]
SPAWN["Spawning<br/>location"] --> JLOC
ATDIST --> OVER["Juvenile-predator<br/>overlap"]
JLOC --> OVER
CANN --> OVER
ATP --> MORT["Age-1 and age-2<br/>mortality"]
OVER --> MORT
MORT --> SURV["Survival<br/>to age 3"]
SURV --> R3["Age-3<br/>recruitment"]
classDef predator fill:#fdf2e8,stroke:#c8523f,stroke-width:2.4px,color:#1f2933;
classDef physical fill:#e8f2f7,stroke:#26739b,stroke-width:2.4px,color:#1f2933;
classDef distribution fill:#ecf5eb,stroke:#6b8f24,stroke-width:2.4px,color:#1f2933;
classDef survival fill:#e8f6f2,stroke:#13856f,stroke-width:2.4px,color:#1f2933;
class ADULT,CANN,ARROW,ATP predator;
class COLD,TEMP,TRAN,SPAWN physical;
class ATDIST,JLOC,OVER distribution;
class MORT,SURV,R3 survival;
Strong predictions:
- Adult/juvenile overlap and arrowtooth overlap should improve predictions after accounting for age-0 condition.
- Warm years can increase risk if predators move into juvenile pollock habitat.
- Spatial metrics should outperform basin-average temperature when predation is the active filter.
DAG C: Transport And Spawning-Location Match-Mismatch
Main claim: early distribution, set by adult spawning location and transport, controls whether larvae and juveniles encounter prey, nursery habitat, and predators (Petrik et al. 2015; Smart et al. 2012).
%%{init: {"flowchart": {"htmlLabels": true, "nodeSpacing": 58, "rankSpacing": 78, "curve": "basis"}, "themeVariables": {"fontSize": "18px", "fontFamily": "Arial, sans-serif"}} }%%
flowchart TB
TEMP["Temperature"] --> SPAT["Spawning<br/>time"]
TEMP --> SPAL["Spawning<br/>location"]
WIND["Wind<br/>forcing"] --> CURR["Cross-shelf and<br/>along-shelf transport"]
SPAT --> INIT["Initial egg<br/>distribution"]
SPAL --> INIT
INIT --> ELD["Egg-larval<br/>distribution"]
CURR --> ELD
ELD --> PREY["Spatial match<br/>with copepod prey"]
ELD --> SAFE["Spatial separation<br/>from adult predators"]
PREY --> LGROW["Larval<br/>growth"]
SAFE --> LMORT["Lower larval and<br/>juvenile mortality"]
LGROW --> A0["Age-0 abundance<br/>and size"]
LMORT --> A0
A0 --> R3["Age-3<br/>recruitment"]
classDef physical fill:#e8f2f7,stroke:#26739b,stroke-width:2.4px,color:#1f2933;
classDef distribution fill:#ecf5eb,stroke:#6b8f24,stroke-width:2.4px,color:#1f2933;
classDef match fill:#fff3df,stroke:#c68519,stroke-width:2.4px,color:#1f2933;
classDef survival fill:#e8f6f2,stroke:#13856f,stroke-width:2.4px,color:#1f2933;
class TEMP,WIND,CURR,SPAT,SPAL physical;
class INIT,ELD distribution;
class PREY,SAFE,LGROW,LMORT match;
class A0,R3 survival;
Strong predictions:
- Particle tracking plus observed spawning distributions should explain more than advection alone.
- Warm/cold distribution shifts should predict prey match and predator overlap.
- Recruitment effects should be strongest before or around age 0, unless juvenile redistribution later overwrites the early pattern.
DAG D: Time-Indexed Switching Control
Main claim: climate state changes which control dominates. Bottom-up control dominates when prey limits juvenile condition; top-down control becomes stronger after productive years increase predator or adult pollock biomass (Mueter et al. 2006; Sigler et al. 2016).
%%{init: {"flowchart": {"htmlLabels": true, "nodeSpacing": 60, "rankSpacing": 82, "curve": "basis"}, "themeVariables": {"fontSize": "18px", "fontFamily": "Arial, sans-serif"}} }%%
flowchart TB
CLIM["Climate state<br/>in year t"] --> ICE["Ice retreat and SST<br/>in year t"]
ICE --> PREY["Prey field<br/>for cohort t"]
PREY --> COND["Age-0 condition<br/>for cohort t"]
COND --> AGE1["Age-1 abundance<br/>of cohort t+1"]
PCOHORT["Prior cohort<br/>strength t-1"] --> PREDAD["Adult and juvenile<br/>predator field t+1"]
AGE1 --> OVER["Predator-prey overlap<br/>t+1 to t+2"]
PREDAD --> OVER
OVER --> AGE3["Age-3 recruitment<br/>of cohort t+3"]
classDef climate fill:#e8f2f7,stroke:#26739b,stroke-width:2.4px,color:#1f2933;
classDef bottomup fill:#ecf5eb,stroke:#6b8f24,stroke-width:2.4px,color:#1f2933;
classDef topdown fill:#fdf2e8,stroke:#c8523f,stroke-width:2.4px,color:#1f2933;
classDef survival fill:#e8f6f2,stroke:#13856f,stroke-width:2.4px,color:#1f2933;
class CLIM,ICE climate;
class PREY,COND bottomup;
class PCOHORT,PREDAD,OVER topdown;
class AGE1,AGE3 survival;
Strong predictions:
- Ice timing effects should interact with adult biomass or predator biomass.
- Consecutive favorable bottom-up years can later increase top-down mortality.
- Recruitment models should allow state-dependent effects rather than a single stationary environment coefficient.
References
Heintz, Ron A., Elizabeth C. Siddon, Edward V. Farley, and Jeffrey M. Napp. 2013. “Correlation Between Recruitment and Fall Condition of Age-0 Pollock (Theragra chalcogramma) from the Eastern Bering Sea Under Varying Climate Conditions.” Deep-Sea Research Part II: Topical Studies in Oceanography 94: 150–56. https://doi.org/10.1016/j.dsr2.2013.04.006.
Moss, Jamal H., Edward V. Farley, Angela M. Feldmann, and James N. Ianelli. 2009. “Spatial Distribution, Energetic Status, and Food Habits of Eastern Bering Sea Age-0 Walleye Pollock.” Transactions of the American Fisheries Society 138 (3): 497–505. https://doi.org/10.1577/T08-126.1.
Mueter, Franz J., Nicholas A. Bond, James N. Ianelli, and Anne B. Hollowed. 2011. “Expected Declines in Recruitment of Walleye Pollock (Theragra chalcogramma) in the Eastern Bering Sea Under Future Climate Change.” ICES Journal of Marine Science, ahead of print. https://doi.org/10.1093/icesjms/fsr022.
Mueter, Franz J., Carol Ladd, Michael C. Palmer, and Brenda L. Norcross. 2006. “Bottom-up and Top-down Controls of Walleye Pollock (Theragra chalcogramma) on the Eastern Bering Sea Shelf.” Progress in Oceanography 68 (2–4): 152–83. https://doi.org/10.1016/j.pocean.2006.02.012.
Petrik, Colleen M., Janet T. Duffy-Anderson, Franz J. Mueter, Katherine Hedstrom, and Enrique N. Curchitser. 2015. “Biophysical Transport Model Suggests Climate Variability Determines Distribution of Walleye Pollock Early Life Stages in the Eastern Bering Sea Through Effects on Spawning.” Progress in Oceanography 138: 459–74. https://doi.org/10.1016/j.pocean.2014.06.004.
Siddon, Elizabeth C., Ron A. Heintz, and Franz J. Mueter. 2013. “Conceptual Model of Energy Allocation in Walleye Pollock (Theragra chalcogramma) from Age-0 to Age-1 in the Southeastern Bering Sea.” Deep-Sea Research Part II: Topical Studies in Oceanography 94: 140–49. https://doi.org/10.1016/j.dsr2.2012.12.007.
Sigler, Michael F., Jeffrey M. Napp, Phyllis J. Stabeno, Ronald A. Heintz, Michael W. Lomas, and George L. Hunt. 2016. “Variation in Annual Production of Copepods, Euphausiids, and Juvenile Walleye Pollock in the Southeastern Bering Sea.” Deep-Sea Research Part II: Topical Studies in Oceanography 134: 223–34. https://doi.org/10.1016/j.dsr2.2016.01.003.
Smart, Tracey I., Janet T. Duffy-Anderson, John K. Horne, Edward V. Farley, Christopher D. Wilson, and Jeffrey M. Napp. 2012. “Influence of Environment on Walleye Pollock Eggs, Larvae, and Juveniles in the Southeastern Bering Sea.” Deep-Sea Research Part II: Topical Studies in Oceanography 65–70: 196–207. https://doi.org/10.1016/j.dsr2.2012.02.018.
Spencer, Paul D., Kirstin K. Holsman, Stephani Zador, et al. 2016. “Modelling Spatially Dependent Predation Mortality of Eastern Bering Sea Walleye Pollock, and Its Implications for Stock Dynamics Under Future Climate Scenarios.” ICES Journal of Marine Science 73 (5): 1330–42. https://doi.org/10.1093/icesjms/fsw040.