The Gold Standard of Ecology | Sentinel-Site Network Drift, Site Loss and Comparable Replacement Design

1. Define the function of every sentinel

Back to contents

A sentinel may warn early, integrate an exposure pathway, represent a habitat or anchor a reference condition. Coordinates are not its whole identity. Record mechanism, target population, indicators, footprint, season and lead time so replacement means comparable function rather than mere proximity.

First-principles check: Name the decision, the people or population affected, the evidence needed and the limit that remains. Write the rule before looking at the preferred outcome. Then identify the authority, observation or calculation that could change it. This keeps the method tied to evidence rather than hindsight.

Worked practice: Write function cards for four estuary sites. Keep every intermediate step visible. A reader should be able to locate the input, apply the same rule and understand why a different result would follow if one fact changed.

Changed-context transfer: A reference site becomes exposed after land-use change. Preserve the underlying principle but replace the setting-specific details. State which parts travel, which do not, and which new authority or measurement would be needed.

Evidence record: Mechanism, target, footprint, indicators, thresholds and limits. Retain uncertainty and disagreement instead of compressing them into a clean label. The record should let another person reconstruct the decision without relying on the memory or status of the original decision-maker.

Decision-quality drill: Run the practice once under the most favourable plausible assumptions and again under the least favourable plausible assumptions. Compare which conclusion changes, which merely becomes less precise and which survives. If the result depends on an unstated convention, make that convention visible before acting. This is a transfer test: the method should guide a new case, not only explain the example after the answer is known.

Practical verification: Give the case file to a reviewer who did not build the analysis. Ask that person to reproduce the critical step, identify the stopping rule and name the strongest live alternative explanation. Resolve differences by returning to the governing evidence, not by averaging opinions. Record corrections, version the decision and keep a short list of conditions that would require reopening it.

2. Detect network drift before loss

Back to contents

Track access, habitat, land use, hydrology, species composition and instrument surroundings. Use maps, photos and maintenance notes. Decide whether observed change is the ecological process of interest or evidence that a stable monitoring function needs another anchor.

First-principles check: Name the decision, the people or population affected, the evidence needed and the limit that remains. Write the rule before looking at the preferred outcome. Then identify the authority, observation or calculation that could change it. This keeps the method tied to evidence rather than hindsight.

Worked practice: Design an annual context review with triggers. Keep every intermediate step visible. A reader should be able to locate the input, apply the same rule and understand why a different result would follow if one fact changed.

Changed-context transfer: Vegetation growth changes sensor microclimate. Preserve the underlying principle but replace the setting-specific details. State which parts travel, which do not, and which new authority or measurement would be needed.

Evidence record: Context observations, imagery, triggers, classification and action. Retain uncertainty and disagreement instead of compressing them into a clean label. The record should let another person reconstruct the decision without relying on the memory or status of the original decision-maker.

Decision-quality drill: Run the practice once under the most favourable plausible assumptions and again under the least favourable plausible assumptions. Compare which conclusion changes, which merely becomes less precise and which survives. If the result depends on an unstated convention, make that convention visible before acting. This is a transfer test: the method should guide a new case, not only explain the example after the answer is known.

Practical verification: Give the case file to a reviewer who did not build the analysis. Ask that person to reproduce the critical step, identify the stopping rule and name the strongest live alternative explanation. Resolve differences by returning to the governing evidence, not by averaging opinions. Record corrections, version the decision and keep a short list of conditions that would require reopening it.

3. Classify the reason for site loss

Back to contents

Loss from safety, access, failure, physical destruction, habitat transformation, budget or redesign has different bias. Record the last valid observation and whether loss relates to condition; storms removing exposed sites make attrition informative, not random.

First-principles check: Name the decision, the people or population affected, the evidence needed and the limit that remains. Write the rule before looking at the preferred outcome. Then identify the authority, observation or calculation that could change it. This keeps the method tied to evidence rather than hindsight.

Worked practice: Classify ten historic closures and bias direction. Keep every intermediate step visible. A reader should be able to locate the input, apply the same rule and understand why a different result would follow if one fact changed.

Changed-context transfer: Access is withdrawn after poor results. Preserve the underlying principle but replace the setting-specific details. State which parts travel, which do not, and which new authority or measurement would be needed.

Evidence record: Reason, timing, ecological relation, last data, bias and response. Retain uncertainty and disagreement instead of compressing them into a clean label. The record should let another person reconstruct the decision without relying on the memory or status of the original decision-maker.

Decision-quality drill: Run the practice once under the most favourable plausible assumptions and again under the least favourable plausible assumptions. Compare which conclusion changes, which merely becomes less precise and which survives. If the result depends on an unstated convention, make that convention visible before acting. This is a transfer test: the method should guide a new case, not only explain the example after the answer is known.

Practical verification: Give the case file to a reviewer who did not build the analysis. Ask that person to reproduce the critical step, identify the stopping rule and name the strongest live alternative explanation. Resolve differences by returning to the governing evidence, not by averaging opinions. Record corrections, version the decision and keep a short list of conditions that would require reopening it.

4. Build a replacement frame from function

Back to contents

Generate candidates serving the same mechanism and stratum, then compare habitat, exposure, elevation, hydrology, detectability, access and viability. Preserve rejected candidates. If no comparable replacement exists, redesign the claim instead of forcing a match.

First-principles check: Name the decision, the people or population affected, the evidence needed and the limit that remains. Write the rule before looking at the preferred outcome. Then identify the authority, observation or calculation that could change it. This keeps the method tied to evidence rather than hindsight.

Worked practice: Score candidates for a lost downstream integrator. Keep every intermediate step visible. A reader should be able to locate the input, apply the same rule and understand why a different result would follow if one fact changed.

Changed-context transfer: All candidates differ in tidal range. Preserve the underlying principle but replace the setting-specific details. State which parts travel, which do not, and which new authority or measurement would be needed.

Evidence record: Candidate frame, criteria, scores, field checks and unresolved differences. Retain uncertainty and disagreement instead of compressing them into a clean label. The record should let another person reconstruct the decision without relying on the memory or status of the original decision-maker.

Decision-quality drill: Run the practice once under the most favourable plausible assumptions and again under the least favourable plausible assumptions. Compare which conclusion changes, which merely becomes less precise and which survives. If the result depends on an unstated convention, make that convention visible before acting. This is a transfer test: the method should guide a new case, not only explain the example after the answer is known.

Practical verification: Give the case file to a reviewer who did not build the analysis. Ask that person to reproduce the critical step, identify the stopping rule and name the strongest live alternative explanation. Resolve differences by returning to the governing evidence, not by averaging opinions. Record corrections, version the decision and keep a short list of conditions that would require reopening it.

5. Use overlap to estimate site effects

Back to contents

Run old and new sites together across relevant seasons and events when possible. Pair observations, use the same protocol and measure covariates. Estimate level, variance, timing and response-shape differences; high correlation can coexist with a threshold-changing offset.

First-principles check: Name the decision, the people or population affected, the evidence needed and the limit that remains. Write the rule before looking at the preferred outcome. Then identify the authority, observation or calculation that could change it. This keeps the method tied to evidence rather than hindsight.

Worked practice: Plan a one-year dry, wet and event overlap. Keep every intermediate step visible. A reader should be able to locate the input, apply the same rule and understand why a different result would follow if one fact changed.

Changed-context transfer: The old site will close in six weeks. Preserve the underlying principle but replace the setting-specific details. State which parts travel, which do not, and which new authority or measurement would be needed.

Evidence record: Design, paired data, events, covariates, models and adequacy. Retain uncertainty and disagreement instead of compressing them into a clean label. The record should let another person reconstruct the decision without relying on the memory or status of the original decision-maker.

Decision-quality drill: Run the practice once under the most favourable plausible assumptions and again under the least favourable plausible assumptions. Compare which conclusion changes, which merely becomes less precise and which survives. If the result depends on an unstated convention, make that convention visible before acting. This is a transfer test: the method should guide a new case, not only explain the example after the answer is known.

Practical verification: Give the case file to a reviewer who did not build the analysis. Ask that person to reproduce the critical step, identify the stopping rule and name the strongest live alternative explanation. Resolve differences by returning to the governing evidence, not by averaging opinions. Record corrections, version the decision and keep a short list of conditions that would require reopening it.

6. Worked example: a marsh sentinel moves inland

Back to contents

Erosion forces a salinity station two hundred metres inland. The new site floods less and sits higher. Six months of overlap shows similar direction but lower peaks and a two-hour delay. The programme creates a new era, keeps raw series, reports bridge uncertainty and does not adjust vegetation without a stable relation.

First-principles check: Name the decision, the people or population affected, the evidence needed and the limit that remains. Write the rule before looking at the preferred outcome. Then identify the authority, observation or calculation that could change it. This keeps the method tied to evidence rather than hindsight.

Worked practice: Decide which indicators can be bridged. Keep every intermediate step visible. A reader should be able to locate the input, apply the same rule and understand why a different result would follow if one fact changed.

Changed-context transfer: A major storm occurs only after closure. Preserve the underlying principle but replace the setting-specific details. State which parts travel, which do not, and which new authority or measurement would be needed.

Evidence record: Site histories, overlap, elevation, model, limits and thresholds. Retain uncertainty and disagreement instead of compressing them into a clean label. The record should let another person reconstruct the decision without relying on the memory or status of the original decision-maker.

Decision-quality drill: Run the practice once under the most favourable plausible assumptions and again under the least favourable plausible assumptions. Compare which conclusion changes, which merely becomes less precise and which survives. If the result depends on an unstated convention, make that convention visible before acting. This is a transfer test: the method should guide a new case, not only explain the example after the answer is known.

Practical verification: Give the case file to a reviewer who did not build the analysis. Ask that person to reproduce the critical step, identify the stopping rule and name the strongest live alternative explanation. Resolve differences by returning to the governing evidence, not by averaging opinions. Record corrections, version the decision and keep a short list of conditions that would require reopening it.

7. Separate relocation from ecological change

Back to contents

Use unaffected sites, covariates, before-after plots and mechanism to distinguish a site step from regional change. Do not subtract a difference automatically: overlap may be unusual and interactions can make one correction invalid. Preserve competing explanations and sensitivity.

First-principles check: Name the decision, the people or population affected, the evidence needed and the limit that remains. Write the rule before looking at the preferred outcome. Then identify the authority, observation or calculation that could change it. This keeps the method tied to evidence rather than hindsight.

Worked practice: Analyse one relocation with three stable sites. Keep every intermediate step visible. A reader should be able to locate the input, apply the same rule and understand why a different result would follow if one fact changed.

Changed-context transfer: A heatwave changes all sites in the relocation month. Preserve the underlying principle but replace the setting-specific details. State which parts travel, which do not, and which new authority or measurement would be needed.

Evidence record: Comparison sites, covariates, models, alternatives and sensitivity. Retain uncertainty and disagreement instead of compressing them into a clean label. The record should let another person reconstruct the decision without relying on the memory or status of the original decision-maker.

Decision-quality drill: Run the practice once under the most favourable plausible assumptions and again under the least favourable plausible assumptions. Compare which conclusion changes, which merely becomes less precise and which survives. If the result depends on an unstated convention, make that convention visible before acting. This is a transfer test: the method should guide a new case, not only explain the example after the answer is known.

Practical verification: Give the case file to a reviewer who did not build the analysis. Ask that person to reproduce the critical step, identify the stopping rule and name the strongest live alternative explanation. Resolve differences by returning to the governing evidence, not by averaging opinions. Record corrections, version the decision and keep a short list of conditions that would require reopening it.

8. Maintain identifiers, eras and coordinates

Back to contents

A replacement gets its own identifier and start date even when inheriting a role. Preserve original coordinates, elevation, habitat and method. Record replaces, overlaps, supplements or changes function; harmonised series must reference a transformation version and raw values.

First-principles check: Name the decision, the people or population affected, the evidence needed and the limit that remains. Write the rule before looking at the preferred outcome. Then identify the authority, observation or calculation that could change it. This keeps the method tied to evidence rather than hindsight.

Worked practice: Design metadata for three successive sites serving one role. Keep every intermediate step visible. A reader should be able to locate the input, apply the same rule and understand why a different result would follow if one fact changed.

Changed-context transfer: Coordinates stay while sensor height changes. Preserve the underlying principle but replace the setting-specific details. State which parts travel, which do not, and which new authority or measurement would be needed.

Evidence record: Identifiers, relations, eras, methods, transformations and provenance. Retain uncertainty and disagreement instead of compressing them into a clean label. The record should let another person reconstruct the decision without relying on the memory or status of the original decision-maker.

Decision-quality drill: Run the practice once under the most favourable plausible assumptions and again under the least favourable plausible assumptions. Compare which conclusion changes, which merely becomes less precise and which survives. If the result depends on an unstated convention, make that convention visible before acting. This is a transfer test: the method should guide a new case, not only explain the example after the answer is known.

Practical verification: Give the case file to a reviewer who did not build the analysis. Ask that person to reproduce the critical step, identify the stopping rule and name the strongest live alternative explanation. Resolve differences by returning to the governing evidence, not by averaging opinions. Record corrections, version the decision and keep a short list of conditions that would require reopening it.

9. Decide which trends can continue

Back to contents

Assess level, direction, threshold, variance and phenology separately for each indicator. Where evidence is weak, analyse eras, include a site indicator, widen uncertainty or restart the baseline. A visible break is better than an invented smooth trend.

First-principles check: Name the decision, the people or population affected, the evidence needed and the limit that remains. Write the rule before looking at the preferred outcome. Then identify the authority, observation or calculation that could change it. This keeps the method tied to evidence rather than hindsight.

Worked practice: Create an indicator-by-claim continuity matrix. Keep every intermediate step visible. A reader should be able to locate the input, apply the same rule and understand why a different result would follow if one fact changed.

Changed-context transfer: Chemistry bridges while detectability changes. Preserve the underlying principle but replace the setting-specific details. State which parts travel, which do not, and which new authority or measurement would be needed.

Evidence record: Claim matrix, treatment, uncertainty, exclusions and review. Retain uncertainty and disagreement instead of compressing them into a clean label. The record should let another person reconstruct the decision without relying on the memory or status of the original decision-maker.

Decision-quality drill: Run the practice once under the most favourable plausible assumptions and again under the least favourable plausible assumptions. Compare which conclusion changes, which merely becomes less precise and which survives. If the result depends on an unstated convention, make that convention visible before acting. This is a transfer test: the method should guide a new case, not only explain the example after the answer is known.

Practical verification: Give the case file to a reviewer who did not build the analysis. Ask that person to reproduce the critical step, identify the stopping rule and name the strongest live alternative explanation. Resolve differences by returning to the governing evidence, not by averaging opinions. Record corrections, version the decision and keep a short list of conditions that would require reopening it.

10. Protect coverage during replacement

Back to contents

Use temporary sites, mobile sampling, remote sensing or intensified neighbours with defined limits. Prioritise by consequence and lead time. Emergency methods do not become permanent selection automatically and should not be silently merged with the long-term protocol.

First-principles check: Name the decision, the people or population affected, the evidence needed and the limit that remains. Write the rule before looking at the preferred outcome. Then identify the authority, observation or calculation that could change it. This keeps the method tied to evidence rather than hindsight.

Worked practice: Build a six-month plan after sudden failure. Keep every intermediate step visible. A reader should be able to locate the input, apply the same rule and understand why a different result would follow if one fact changed.

Changed-context transfer: Two sentinels fail in the same storm. Preserve the underlying principle but replace the setting-specific details. State which parts travel, which do not, and which new authority or measurement would be needed.

Evidence record: Gap map, temporary methods, limits, triggers, capacity and transition. Retain uncertainty and disagreement instead of compressing them into a clean label. The record should let another person reconstruct the decision without relying on the memory or status of the original decision-maker.

Decision-quality drill: Run the practice once under the most favourable plausible assumptions and again under the least favourable plausible assumptions. Compare which conclusion changes, which merely becomes less precise and which survives. If the result depends on an unstated convention, make that convention visible before acting. This is a transfer test: the method should guide a new case, not only explain the example after the answer is known.

Practical verification: Give the case file to a reviewer who did not build the analysis. Ask that person to reproduce the critical step, identify the stopping rule and name the strongest live alternative explanation. Resolve differences by returning to the governing evidence, not by averaging opinions. Record corrections, version the decision and keep a short list of conditions that would require reopening it.

11. Audit attrition and replacement bias

Back to contents

Review which sites are lost, how candidates are selected and whether accessibility, habitat or exposure distribution shifts. A constant station count can hide disappearance of harsh or remote conditions. Audit metadata, overlap and public trend claims by site era.

First-principles check: Name the decision, the people or population affected, the evidence needed and the limit that remains. Write the rule before looking at the preferred outcome. Then identify the authority, observation or calculation that could change it. This keeps the method tied to evidence rather than hindsight.

Worked practice: Audit a decade of closures and replacements. Keep every intermediate step visible. A reader should be able to locate the input, apply the same rule and understand why a different result would follow if one fact changed.

Changed-context transfer: The count stays constant while harsh sites vanish. Preserve the underlying principle but replace the setting-specific details. State which parts travel, which do not, and which new authority or measurement would be needed.

Evidence record: Attrition, candidate decisions, overlaps, coverage and corrective action. Retain uncertainty and disagreement instead of compressing them into a clean label. The record should let another person reconstruct the decision without relying on the memory or status of the original decision-maker.

Decision-quality drill: Run the practice once under the most favourable plausible assumptions and again under the least favourable plausible assumptions. Compare which conclusion changes, which merely becomes less precise and which survives. If the result depends on an unstated convention, make that convention visible before acting. This is a transfer test: the method should guide a new case, not only explain the example after the answer is known.

Practical verification: Give the case file to a reviewer who did not build the analysis. Ask that person to reproduce the critical step, identify the stopping rule and name the strongest live alternative explanation. Resolve differences by returning to the governing evidence, not by averaging opinions. Record corrections, version the decision and keep a short list of conditions that would require reopening it.

12. The gold standard preserves function before continuity

Back to contents

A long record is useful only if its meaning survives. Strong networks define function, detect drift, classify loss, select by mechanism and estimate site effects. When no honest bridge exists, they preserve the break and restore coverage without inventing history.

First-principles check: Name the decision, the people or population affected, the evidence needed and the limit that remains. Write the rule before looking at the preferred outcome. Then identify the authority, observation or calculation that could change it. This keeps the method tied to evidence rather than hindsight.

Worked practice: Review one replacement from early drift to claim update. Keep every intermediate step visible. A reader should be able to locate the input, apply the same rule and understand why a different result would follow if one fact changed.

Changed-context transfer: Repeat after restoration changes the site function. Preserve the underlying principle but replace the setting-specific details. State which parts travel, which do not, and which new authority or measurement would be needed.

Evidence record: Function cards, context, loss, candidates, overlaps, eras and audits. Retain uncertainty and disagreement instead of compressing them into a clean label. The record should let another person reconstruct the decision without relying on the memory or status of the original decision-maker.

Decision-quality drill: Run the practice once under the most favourable plausible assumptions and again under the least favourable plausible assumptions. Compare which conclusion changes, which merely becomes less precise and which survives. If the result depends on an unstated convention, make that convention visible before acting. This is a transfer test: the method should guide a new case, not only explain the example after the answer is known.

Practical verification: Give the case file to a reviewer who did not build the analysis. Ask that person to reproduce the critical step, identify the stopping rule and name the strongest live alternative explanation. Resolve differences by returning to the governing evidence, not by averaging opinions. Record corrections, version the decision and keep a short list of conditions that would require reopening it.