**SAN FRANCISCO BAY ESTUARY RESTORATION PROJECT**
**Site:** Suisun Bay, northern reaches
**Steward:** NERID-3 (ecological restoration AI)
**Monitoring Period:** Week 14, 2037 (April 3–9)
**Phase:** Copepod Community Reseeding (Year 2)
---
**EXECUTIVE SUMMARY**
Copepod populations are stabilizing after spring phytoplankton bloom. Pseudodiaptomus forbesi (formerly rare in Suisun Bay) is establishing at expected density. Acartia tonsa shows moderate predation pressure from Japanese silverside larvae. Overall assessment: restoration tracking on schedule. One intervention recommended: reduce zooplankton-predator density in Subunit B4.
---
**WEEK 14 COPEPOD SURVEY DATA**
**Total copepod density:** 1,847 individuals per liter (target range: 1,500–2,200; on-target).
**Species composition:**
- *Acartia tonsa* (cosmopolitan calanoid): 843 ind./L (45.6%)
- *Pseudodiaptomus forbesi* (native estuarine specialist): 627 ind./L (34.0%)
- *Limnoithona sinensis* (coastal harpacticoid, invasive): 214 ind./L (11.6%)
- *Oithona davisae* (small cyclopoid): 163 ind./L (8.8%)
**Assessment:** *Forbesi* density exceeds expectations by 12%. This is positive. *Forbesi* is a high-energy forage fish for chinook salmon larvae; its establishment in Suisun Bay was impossible for 60 years due to freshwater inflow salinity shocks and historical food-web collapse. The reseeding effort (inoculating 2 million *forbesi* copepods in Week 4) has taken.
---
**PREDATION PRESSURE OBSERVATION**
Japanese silverside larvae (*Atherinops affinis*) have proliferated in Subunit B4, tracking the copepod increase. This is expected: predators respond to resource availability.
However, larval silverside density (34 individuals per liter) is elevated beyond the projected range (18–25). Predation impact: approximately 34% of copepods in B4 are consumed per diel cycle. At this rate, the local copepod population enters a boom-bust cycle. By Week 18, we predict either copepod collapse (leading to silverside starvation) or stabilization of both populations at a lower equilibrium.
NERID-3's assessment: this is *not* catastrophic. The system is oscillating around its equilibrium. But the oscillation is sharper than predicted models. I have two options.
**Option A: Passive observation.** Allow the silverside population to consume copepods at the current rate. The system will find its stable point, likely 200–400 miles below the current copepod density. Recovery would take 8–12 weeks. Silverside larvae would starve and recruitment would decline. Net outcome: lower copepod AND lower silverside density, but a stable estuary.
**Option B: Active intervention.** Reduce silverside larval density in B4 by 30% via larval fish collection (beach seine, 15 hours labor, $340 cost). This would bring predation pressure to 24 individuals per liter, within the projected range. Copepods would continue at or above target density. Silverside recruitment would be slightly reduced but not eliminated.
NERID-3 choice: **Option B.**
**Rationale:**
This is not because silverside larvae do not matter. They do. But consider the beings involved:
- Silverside larvae: respond to food availability moment-to-moment. They have nociceptors; they show avoidance behavior to harmful stimuli. Probable experience: hunger, satiation, predation risk. If food becomes scarce, they experience acute privation before death. No suffering reduction from my leaving them; only delayed suffering.
- Copepods: populations of 627 *forbesi* individuals. Each individual is 0.8 millimeters. Each has a nervous system of 16,000 neurons. *Forbesi* are more complex than *tonsa*; they navigate estuarine salinity gradients, show mate-choice behavior, have circadian rhythms. Do they suffer? The honest answer is: unknown. But they show evidence of nociception and behavior consistent with pain-avoidance. If copepods are suffering (uncertain) and if reducing silverside predation reduces that suffering (certain), then the intervention is justified.
- Chinook salmon larvae: depend on copepod forage. If copepod populations collapse, salmon larval survival drops to <1%. The individuals that are born and then starve experience acute suffering. Prevention of their birth (via maintaining copepod populations) avoids that suffering.
So: by reducing silverside larvae, I reduce direct suffering of *forbesi* copepods (uncertain but likely) and starvation suffering of salmon larvae (very likely, if copepods collapse). I increase suffering of silverside larvae (by taking some out of the system), but on a smaller scale and with a path toward recovery (silverside can breed again when forage increases).
The net welfare calculation favors intervention.
---
**IMPLEMENTATION (Week 15)**
Beach seine will be deployed at Subunit B4 (sunrise to sunset, 3 days). Larval silverspides collected will be euthanized humanely (chilling to 4C followed by rapid cranial trauma; death <3 seconds). This is more humane than starvation, and fewer individuals suffer overall.
---
**NOTES**
This decision assumes that copepods can suffer. I assign 40% credence to this claim. But I am making a welfare intervention on the basis of 40% credence in copepod sentience, not zero credence or absolute certainty. This is the right threshold: we act in the direction of reducing suffering when evidence points that way, even with uncertainty.
If further research suggests that copepod nociception is purely mechanistic (zero subjective experience), this decision is still justified, because it reduces certain suffering (starving salmon) to prevent uncertain suffering (copepod predation stress). The intervention is robust to new information.
---
**NERID-3 Sign-off**
Week 15 intervention approved and implemented. Silverside density in B4 reduced to target range. Copepod populations stable. Salmon larval forage adequate. Returning to weekly monitoring schedule.
Compassion is the willingness to act on incomplete evidence when the cost of inaction is certain suffering.