# Optimizing Fishery Survey Design in Guangdong’s Restricted Coastal Waters

## TL;DR

The study finds optimal ways to design fishery surveys in Guangdong's coastal waters to maximize species detection and accuracy within budget limits.

## Contribution

The study introduces a comparison of survey design methods to optimize biodiversity detection and abundance estimation in restricted coastal waters.

## Key findings

- Systematic sampling detected the most species but was unstable in some cases.
- Stratified random sampling by depth provided the most reliable fish abundance estimates.
- Sampling in autumn or spring and autumn together minimized errors when surveys were limited.

## Abstract

Coastal waters in Guangdong are rich in marine life but face pressure from overfishing, pollution, and climate change. Managers need surveys that are both accurate and affordable. We carried out four seasonal bottom-trawl surveys at 186 sites inside the restricted fishing area and tested how different ways of choosing survey sites and survey timing affect results. We recorded 563 species and found clear seasonal and spatial changes, with the highest catches in summer. The more sites and the more seasons we sampled, the more species we found. A regularly spaced grid of sites detected the most species but could be unstable in some situations. Choosing sites at random within depth bands (shallow to deeper water) gave the most reliable estimates of how much fish was present. Sampling all four seasons required only 88 sites to find about four fifths of all species; if fewer surveys are possible, autumn alone, or spring and autumn together, performed best. These findings show how to design lean, effective surveys that stretch limited budgets while improving protection and management of key spawning and nursery habitats.

The coastal restricted fishing area of Guangdong contains key spawning and nursery habitats with high biodiversity but growing ecological pressure, yet the influence of survey design and sampling frequency on biodiversity detection and abundance estimates remains unclear. We conducted four seasonal bottom-trawl surveys in 2023–2024 at 186 stations and compared fixed-site sampling (FS), simple random sampling (SRS), stratified random sampling by depth (StRS), and systematic sampling (SS). We recorded 563 species (446 fishes, 101 crustaceans, 16 cephalopods), observed seasonal shifts in dominant taxa, and found catch rates varied seasonally and spatially, peaking in summer. Species detection rose with station number and sampling frequency. For species richness, SS produced the highest detection and the lowest error and bias but showed volatility; StRS and SRS were more stable. For abundance, StRS had the lowest error, whereas SRS had the smallest absolute bias. Across all four seasons, 88 stations achieved an 80% richness detection rate; among reduced-frequency designs, autumn-only, spring–autumn, and autumn–spring–summer minimized errors. These results guide cost–precision trade-offs: SS (with random starts and interval rotation) for richness-oriented aims, and depth-based StRS for abundance, supporting optimized long-term monitoring and management in the northern South China Sea.