Conservation gains hinge on what happens after the agreement

Agreements set direction, but delivery is the test

At COP17, countries agreed the first global review of progress under the Kunming–Montreal Global Biodiversity Framework. Its warning is plain: faster action is needed if the 2030 goals are to be met. Yet the design of a future financial mechanism was left to COP18, and new reporting describes an unresolved funding gap for conservation in Africa. A shared assessment can sharpen accountability, but it does not itself close that gap or demonstrate progress on the ground.

The conference also agreed operating rules for the CBD’s permanent body on Article 8(j), creating a formal route for Indigenous peoples and local communities to participate and advise. The body remains Party-led, and final decision-making authority stays with Parties. The practical question is whether the new arrangements will bring meaningful influence and adequate support, not simply whether they exist.

The limits of agreement were also evident at CCAMLR, which ended without new Antarctic protected areas or precautionary measures for the krill fishery for the tenth consecutive year. Meanwhile, a proposed Northern Territory bill would remove the statutory requirement for water plans to set aside environmental water. The bill is still before the Legislative Assembly, and its ecological effects are unknown. In both cases, the protection of nature depends on decisions that remain unsettled.

Protection on paper meets pressure on the ground

Angola’s law establishing Serra do Pingano as a Partial Nature Reserve is a concrete legal step in a forest ecosystem reported to have lost substantial cover. But the area covers less than the wider region researchers had proposed for protection, and management and inspection arrangements are still to be set out separately. The designation creates a basis for action; it does not yet show that forest loss has slowed.

At Rothbury, wildlife trusts have completed the purchase of the 3,800-hectare estate after a £30m fundraising campaign. Nature recovery, nature-friendly farming and a long nature corridor are stated ambitions, not ecological results already achieved. The scrutiny of public funding and concerns raised about tenant farmers underline that turning ownership into conservation will also require decisions about land management and local relationships.

A Brazilian federal judgment sets different requirements for occupants and cattle in Piripkura territory, while requiring implementation to avoid contact with people living in voluntary isolation. Whether the orders have been carried out is not yet known. At the Denniston Plateau, a resubmitted mining application has not yet passed the EPA’s completeness check. And a proposed PFAS treatment for farmland remains a modelled approach, not a field-tested clean-up. In each case, a plan or procedural step is only the beginning of the test.

Climate pressure complicates resilience and planning

Aircraft measurements from four Amazon sites indicate that the sampled region was, on average, a net carbon source from 2010 to 2024. Fires were the largest measured source, while deforestation and degradation were also important drivers, and drought and heat amplified emissions in some years. The finding is significant, but the sampling does not describe every part of the basin or every year.

A report on the cryosphere warns that warming above 1.5°C could cause long-lasting losses across ice and snow systems. A swift return to that threshold or lower could limit further damage and allow some ice to recover, but would not reverse every loss. At the same time, a global rain-gauge analysis found rainfall concentrated into fewer wet days at many stations, though trends were not uniform and the study does not show a worldwide increase in event intensity.

Other research points to possible ways to plan for stress, while keeping the limits of the evidence in view. Across US forest plots, greater tree-species richness was associated with stronger resistance and recovery from climate extremes, but richness explained only a limited share of variation. A Southern African model identified October as generally favourable for planting bambara groundnut, while highlighting frost risk and the need for field validation. Neither study alone establishes which interventions will work locally.

Recovery needs more than counts

The Elwha River study offers a clear example of why monitoring needs several measures. After a sharp decline during dam removal, invertebrate abundance returned close to baseline within three years of peak sediment transport subsiding. But richness and community composition remained altered at the end of monitoring in 2017. A numerical rebound and the return of a community’s former make-up are not the same outcome.

The identification of Shortnose Cisco in Lake Superior updates what is known about the species’ occurrence, supported by genetic and morphological evidence and a recent trawl record. Two fish in that catch do not establish a viable population. In the Wairau Aquifer, modelling suggests lower riverbed levels could account for much of the groundwater decline in the recharge area, but climate, flows and abstraction may also contribute. In both cases, better evidence is needed to guide decisions and assess future change.

Marine evidence raises questions about risk and response

In the north-western Mediterranean, reported analysis found extensive high-speed vessel traffic despite a voluntary recommendation for some vessels to slow when cetaceans are detected or reported. The figures raise concern about collision risk, but do not show that the recommendation caused a change in speeds or reduced whale strikes. The disagreement over whether slower travel should be mandatory is therefore also a question about what evidence will be used to judge effectiveness.

At Pulau Redang Marine Park, microplastics were detected in 64% of 200 sampled coral fragments. The finding establishes contamination in those samples, not harm to coral health or the particles’ source. Meanwhile, observations of crown-of-thorns seastar feeding at two low-density sites found living coral remnants after most recorded feeding events. The researchers’ suggestion that this may be compatible with recovery is a hypothesis, not a measured reef-wide benefit, and should not be extended to outbreak conditions.

Presence, movement and local trends need follow-through

Several species stories add valuable records without resolving population status. Rescue records confirm that Bangka slow lorises persist on the island, but cannot establish how many remain. A documented Ghana cuckooshrike in Taï National Park confirms a recent occurrence after a long gap in records, not recovery. In San Diego County, cactus wren counts were broadly steady in consistently monitored plots, with differences between local clusters. Each finding strengthens the evidence base while leaving trend and cause uncertain.

GPS maps of South Sudan’s antelope migration could help planners consider the effects of infrastructure and land use on movement, but the reporting does not show that a specific decision has changed. Similarly, tracking of Manx shearwaters found flight-direction deviations consistent with a mismatch between internal clocks and local solar time, but did not establish harm to the birds. For conservation practice, the next step is to connect records and models to sustained monitoring and decisions, then test what those decisions achieve.