Hello,
A pro tip from my boss - if you've been frustrated by the 8,000 character limit for instructions for agents in Copilot, you can break up additional instructions into text files and upload them as 'knowledge'. For example, those extra files could provide detailed criteria about how to make a judgment, a template to use for the format of outputs, etc.
As always - If you know someone who wants to sign up to receive these summaries, they can do so at http://subscribe.sciencejon.com (no need to email me).
GLOBAL TREE COVER:
Neumann et al. 2026 is an announcement of a new global tree cover map split into five types: 1) primary forest (no visible human disturbance, although recovery from wildfire and pests is ok, and may include low-intensity fragmentation, 2) forest naturally regenerating after human disturbance (like abandoned ag lands, formerly clearcut forest no longer managed for timber, etc.), 3) planted forests (basically plantations with >40 year rotation but may be less intensively managed), 4) plantation forest (intense management and ≤40 year rotation), 5) tree crops / agroforestry (palm, cocoa, etc.). The distinctions are useful but they also note several limits like inability to detect shade-grown cacao or coffee under native tree canopy, challenges in splitting human vs. natural disturbance, etc. Read a blog about it at https://globalnaturewatch.org/blog/data-and-tools/new-forest... or a more technical blog at https://deepmind.google/blog/beyond-tree-cover-mapping-the-worlds-forest-types/
CITIZEN SCIENCE:
Steen et al. 2026 looked at the accuracy of citizen scientists in FL, and one finding really struck me. They found accuracy varied a lot (<50% for many mammals and some snakes, high nineties for some birds) but that even for the species ordinary people usually got wrong, the total NUMBER of correct identification often exceeded the number from professional biologists. So as long as there is appropriate quality assurance in not assuming the accuracy of citizen science records, it's an important tool to boost the total number of wildlife observations we have available. Be careful reading the tables which are confusing, e.g., in Table 4 while 98% of species REPORTED AS American mink were correctly identified, the 37.41% number means that only 37% of submitted images which were determined by researchers to ACTUALLY be mink were reported as mink. So in that case almost all reported minks were correct, but almost as many actual minks were falsely reported to be river otters as correctly identified. Also - they included red fox twice on different rows and counted both in the total which is an error - there could be other errors lurking.
DAM REMOVAL:
Goodman et al. 2026 used sonar to track Chinook salmon recolonizing rivers after the removal of the four hydro dams on the Klamath in 2024. The fish passages in the remaining dams are working pretty well! In 2024 they saw >7,700 salmon migrate upstream, increasing to ~13,000 in 2025 and occupying 88% of their historic range. ~300 made it 580 km to the uppermost tributaries, and some juveniles successfully hatched. https://www.nytimes.com/2026/09/24/climate/klamath-salmon-re...
REFERENCES:
Goodman, D. H., Denton, K., Pess, G. R., Whelan, J., Stefankiv, O., Soto, T., Corum, A., Ortega, J., O’Rourke, O., Hereford, M. E., Giudice, D., & Som, N. A. (2026). Rewilding the upper Klamath River Basin: rapid recolonization of Chinook salmon following the world’s largest dam removal. Scientific Reports, 16(1), 25322. https://doi.org/10.1038/s41598-026-65437-0
Neumann, M., Raichuk, A., Potapov, P., Lesiv, M., Overlan, M., Rey, M., Rajakumar, R., Conserva, M., Stanimirova, R., Sims, M. J., Carter, S., Goldman, E., Jiang, Y., Scheibenreif, L., Georgieva, I., Shchepashchenko, M., Fritz, S., Clinton, N., Stanton, C., … Purves, D. (2026). Global forest typology at 10-meter resolution for forest and land-use monitoring. Scientific Data, 13(1), 1434. https://doi.org/10.1038/s41597-026-08299-5
Steen, D. A., Sylvia, A., Enge, K., Miller, K. E., Smith, L., Poli, C., & Rizkalla, C. (2026). Accuracy of wildlife observations identified by citizen scientists in Florida. Wildlife Society Bulletin, July, 1–14. https://doi.org/10.1002/wsb.70056
P.s. This bumblebee foraging from snakeroot in my garden is a reminder that having late-blooming flowers as part of your garden mix is important to ensure that their population isn't limited by the sparse flowers this time of year. In my garden snakeroot, obedient plant, lavender, and rosemary are the best performers in late October into November.
