**Background:** Global production of Penaeus vannamei (white-leg shrimp) increased by nearly 53% in five years, from 3,803.6 thousand tonnes in 2015 to 5,812.2 thousand tonnes in 2020, with an estimated 167 billion individual shrimp of this species farmed annually. Despite growing scientific evidence and legislative recognition of sentience in decapod crustaceans—including the UK amending its Animal Welfare (Sentience) Act 2022 to include decapods, and Switzerland requiring electrical stunning or mechanical brain destruction for crustaceans—no standardized, non-invasive welfare assessment protocols exist for penaeid shrimp. The authors aimed to develop such protocols covering the entire production cycle: reproduction, larval rearing, postlarval transport, and grow-out in earthen ponds.
**Methods:** Welfare indicators were grouped under four domains (nutrition, environment, health, behavior), with the fifth domain (psychology) assessed indirectly. Reference values were derived from a literature search covering 1976–2023 using Google Scholar, with specific search terms including "vannamei AND shrimp AND pond -biofloc" combined with domain-specific indicators. Three scores were assigned: Score 1 (ideal variation limits), Score 2 (sublethal tolerable variations), and Score 3 (unacceptable levels risking welfare and survival). Field validation was conducted through technical visits to one farm and one larval rearing laboratory in Rio Grande do Norte state, and three farms and two larval rearing laboratories in Ceará state, northeastern Brazil. Indicators that could not be collected on-site were discarded.
**Key Results:** The final protocols include detailed tables for each production stage:
- **Reproduction (breeders):** 11 environmental indicators (e.g., temperature score 1: 24.5–32.5°C, pH score 1: 7.5–8.5, dissolved oxygen score 1: ≥62% saturation, non-ionised ammonia score 1: 0.00–0.10 mg/L, stocking density score 1: ≤150 g/m²); 14 health indicators including anatomical assessment of antennae, rostrum, eyes, gills, hepatopancreas, motor appendages, exoskeleton, musculature, gastrointestinal tract, luminescence, sexual maturation, invasive procedures (eyestalk ablation), mortality (score 1: ≤10%), and genetic selection; 5 nutritional indicators (digestive tract filling, feeding frequency ≥3 times/day for score 1, crude protein ≥35% for score 1); and 3 behavioural indicators (swimming, feeding response, anaesthesia).
- **Larval rearing:** 9 environmental indicators with stricter thresholds (e.g., non-ionised ammonia score 1: 0.00–0.01 mg/L, nitrite score 1: 0.0–0.1 mg/L, stocking density score 1: ≤250 larvae/L or ≤100 postlarvae/L); 10 health indicators including health certification (SPF/SPT/SPR), luminescence, larval stage uniformity (≥75% for score 1), malformations (≤5% for score 1), hepatopancreas staining and lipid vacuole condition, epibiont encrustation, necrosis, melanisation, and mortality (score 1: ≤30%); 5 nutritional indicators with size-specific feed requirements; and 2 behavioural indicators (phototaxis ≥95% for score 1, swimming activity ≥95% for score 1).
- **Postlarvae transport:** 12 indicators across all domains, including transport density (score 1: ≤750 postlarvae/L), dissolved oxygen (score 1: ≥80% saturation), mortality (score 1: ≤5%), and swimming behaviour after transport (≥95% active for score 1).
- **Grow-out (juveniles/adults):** 11 environmental indicators (e.g., temperature score 1: 25.5–32.4°C, pH score 1: 6.5–8.5, dissolved oxygen score 1: ≥65% saturation, salinity score 1: 10.0–40.9 psu, stocking density score 1: ≤40 shrimp/m², transparency 35–50 cm); 10 health indicators; 8 nutritional indicators with weight-class-specific values (e.g., feed crude protein ≥35% for shrimp ≤3.9 g, apparent FCR ≤1.5 for 4.0–8.9 g shrimp); and 3 behavioural indicators covering routine swimming, escape behaviour during harvest, and clinical reflexes during stunning/slaughter.
**Clinical Implications:** The protocols represent the first comprehensive attempt to operationalize shrimp welfare assessment across all production stages using only non-invasive methods. The authors explicitly excluded physiological biomarkers requiring haemolymph collection or other invasive procedures. They note that many proposed indicators are already integrated into routine good aquaculture practices (GAP), suggesting feasibility. The scoring system (1–3) is designed to be less subjective than protocols with more categories. The authors acknowledge challenges including knowledge gaps in penaeid biology, uncertainty about larval sentience (applying the precautionary principle), and the interactive nature of welfare indicators (e.g., salinity affects oxygen solubility and shrimp oxygen consumption simultaneously). They predict convergence with precision aquaculture technologies (biosensors, computer vision, IoT) but emphasize that current protocols can be implemented immediately by any shrimp farmer without specialized equipment.