If you could afford one more field day, where should it be—and what decision could it change?

Imagine a project team reviewing a rehabilitation map before commissioning its next survey. There are treatment boundaries, vegetation layers, photographs and several areas marked for investigation. Each specialist can identify something worth examining. The budget will cover one more day.

Choosing where to send the team requires more than finding the largest gap in coverage. It requires agreement about the decision ahead, the uncertainty affecting it and what a further observation could realistically resolve.

That is a useful next ambition for environmental intelligence: helping evidence guide its own follow-up.

Many maps legitimately serve other purposes, including baseline documentation, communication and reporting. But when a map supports an active management program, there is an opportunity to ask more of the thinking around it. Alongside what is known, it could help explain what should be investigated next, why that investigation matters and what would follow from different findings.

Begin with the choice facing the team

A management objective gives uncertainty its practical significance.

A team deciding whether to undertake further weed control needs different evidence from a team estimating native vegetation cover across an entire project. The same landscape can therefore warrant different survey priorities.

Research on monitoring within structured decision-making makes this connection explicit: monitoring design should follow the decision context and its associated uncertainties. Monitoring can inform the next action, assess performance or support learning about management responses. These purposes need deliberate consideration. [4]

A useful opening sentence for a survey brief is:

We need to choose between these actions, and this unresolved question could change that choice.

That sentence gives the imagery, records and proposed fieldwork something concrete to serve. It also exposes situations where the team has agreed to collect information without agreeing how it will be used.

An unusual place is not necessarily the most useful place

Three objectives are easily blurred: finding something unusual, reducing uncertainty and obtaining information that could improve a decision.

Consider a hypothetical map containing a conspicuous vegetation anomaly, a poorly documented corner and a modest patch close to a planned treatment area.

The anomaly deserves attention if understanding it could affect management. The undocumented corner may matter if its condition changes a project priority. The modest patch could be the most useful destination if identifying its vegetation would determine whether treatment proceeds.

None has an automatic claim on the next field day.

The practical test is to describe plausible findings before selecting the observation. Would each finding lead to a different action, timing, treatment method or allocation of effort? Would the proposed survey distinguish those findings well enough to help?

If every plausible result leaves the immediate action unchanged, the investigation needs another justification. It may still be important for compliance, baseline knowledge or longer-term surveillance. Its purpose should be explicit.

Use the existing record to build a testable question

Existing evidence can help turn a broad concern into a specific investigation.

For the hypothetical rehabilitation team, imagery might identify where vegetation patterns differ. Works records could establish whether those areas received the same treatment. Earlier field observations could help identify competing explanations.

The next step would be to ask what observation could distinguish those explanations. That might require species identification, a check of planting survival or an examination of ground conditions. Another image capture would be useful only if it could resolve something relevant.

Henrys, Mondain-Monval and Jarvis’s 2024 review describes adaptive sampling as changing subsequent sampling in response to information already collected. It considers both observation-led and model-based approaches, including using remote sensing to inform field sampling. [1]

The distinction matters: selecting a sensible follow-up visit is a practical planning exercise. A formal adaptive design requires explicit sampling rules and suitable analysis.

A hypothetical Northern NSW rehabilitation decision

Consider a riparian rehabilitation project in Northern NSW. This is an illustrative scenario, not a TALONYX case study.

The team must decide whether to commission further weed treatment in a recovering section or leave it untreated while continuing established monitoring.

Its records show earlier weed control and planting. Recent imagery shows dense vegetation in part of the section. Existing field observations do not establish whether that cover is predominantly target weeds, desired regeneration or a mixture.

The uncertainty that could change the choice is the vegetation’s composition and the implications for treatment. The team needs to distinguish a section requiring follow-up control from one where the proposed operation could unnecessarily affect desired plants.

A proposed field visit would examine the ambiguous section, record species and their distribution, and assess whether the treatment approach remains appropriate. The ecologist and works supervisor would agree the observations needed before mobilisation.

If target weeds dominate accessible treatment patches, the team could refine and proceed with follow-up control. If desired regeneration predominates, it could defer that operation there. If weeds and desired plants are closely intermingled, it could reconsider the method or seek specialist advice before issuing instructions.

An inconclusive visit would remain inconclusive. It would not automatically justify treatment or establish that treatment was unnecessary.

Elsewhere on the same project, recent field evidence may already confirm target weeds and an agreed, suitable control method. If another survey would leave that instruction unchanged, the team could proceed with the authorised work while retaining required checks.

Before visiting the ambiguous section, the team would agree a review rule: once the observations adequately distinguish the operational choices, issue a documented decision. If identification or treatment risk remains unresolved, escalate the specific question to the appropriate specialist. Avoid repeating a general survey without explaining what the repetition could resolve.

This example concerns the next operational choice. It does not establish that earlier treatment caused the observed vegetation pattern.

Information has value through what it enables

Formal value-of-information analysis compares expected management performance with and without additional information, accounting for observation costs and consequences such as delay.

Bennett and colleagues’ 2018 study extended that analysis across multiple management units under limited budgets. Its demonstrations used simulated conservation cases; they were not measured outcomes from Australian rehabilitation projects. [2]

The useful connection for everyday planning is to work through the possible results before commissioning the survey.

What would the team do if its current interpretation were confirmed? What if it were contradicted? Could the observation arrive too late to influence the choice?

A discussion can expose these dependencies without calculating probabilities or claiming formal optimisation. Where the stakes justify quantitative analysis, suitable ecological and statistical expertise should shape it.

Preserve the evidence needed for broader conclusions

Targeting a problem area and estimating condition across a whole landscape are different jobs.

If a team increasingly visits suspected failures, a simple summary of those visits may make overall condition appear worse. If it concentrates on successful patches, the reverse can occur. Changing locations can also complicate comparisons between monitoring rounds.

Henrys and colleagues identify preferential sampling, imperfect data and model assumptions as important considerations. An adaptive design needs analysis appropriate to how observations were selected. [1]

One option is to retain an established sampling component while adding targeted investigations. Williams and colleagues discuss hybrid designs combining conventional and dynamic survey approaches; their study developed a model-based design for sea otters in Alaska. That supports a design option, rather than a universal prescription for rehabilitation monitoring. [3]

For a local project, agree with the survey designer how targeted observations, reference sites and any probability-based sample will contribute to the conclusions. Record why each added site was selected. Do not quietly pool observations collected for different purposes.

Know when to act, continue or seek advice

A follow-up observation must fit the scale and timing of the question. A visit to one patch cannot establish condition across a catchment. An observation made outside suitable detection conditions may leave the central uncertainty unresolved.

NSW field-survey guidance explains that non-detection does not necessarily establish absence and identifies survey method, conditions and observer experience as influences on detectability. [6]

Before redirecting effort, check applicable approval conditions, mandatory surveys, funding requirements and the established monitoring design with the responsible practitioner.

Adaptive management is broader than adaptive sampling: it connects objectives, action, monitoring and subsequent adjustment. Australian water-quality guidance recognises that management often proceeds before uncertainty is fully resolved. [5]

Sometimes the appropriate next step is therefore action, continuation of an established survey, or specialist advice.

A practical brief for the next investigation

Before commissioning additional work, write down:

  1. The choice: What decision is due, who owns it and when?
  2. The evidence: What is already known, and from which observations?
  3. The uncertainty: Which unresolved issue could change the choice?
  4. The investigation: What would be measured, where and under what conditions?
  5. The consequences: What would different findings mean for management?
  6. The review rule: When will the team act, stop investigating or seek advice?

This is a proposed planning framework, not a new scientific method.

Its purpose is to make the connection between evidence and action visible. A useful environmental map could carry that connection alongside its layers: the question, the candidate investigation and the decision waiting on it.

If a real environmental decision is waiting on your next survey, TALONYX welcomes a discussion about how the existing evidence could help frame that investigation.

Discuss a project

Related TALONYX capability

Environmental intelligence GIS, remote sensing & spatial analysis Rehabilitation & restoration monitoring

Sources and further reading