Monday, May 3, 2021

May 2021 science summary

Leeta sitting


Unfortunately this is the second month in a row with a memorial dedication. Our dog (and recent office-mate) Leeta recently passed away. She was an extremely sweet 'muttweiler' who lived with us almost 13 years. She provided incredible emotional support during the pandemic, when I was dealing with harsh peer reviews on scientific papers, etc. We're struggling to adjust, so if my email replies are slow or disappointing, that's likely why.

I've just got reviews of three miscellaneous articles I read before she died (on long-distance mammal migration, forest fragmentation & degradation, and remote sensing forest canopy height).

If you know someone who wants to sign up to receive these summaries, they can do so at

Teitelbaum et al. 2015 is an analysis of how far big herbivorous mammals migrate. They looked at 94 populations of 25 species migrating from between 10-1600 km (~6-1000 mi). The longest migrations were from caribou, onagers (a kind of Asian donkeys), and saiga antelope (figure 1). The biggest driver of migration distance was how much green vegetation was present - animals go much farther when food is sparser (table 1, figure 3). There's a nice discussion of other relevant variables, but overall they struggled to accurately model migration distance.

Grantham et al. 2020 estimates how much forests have been fragmented and modified around the world. They look at proximity to infrastructure, agriculture, and tree cover loss, along with lost forest connectivity, to estimate forest modification. The way they defined modification means that only forests in the most remote and sparsely populated areas are scored as having high landscape integrity (see figure 2 and figure 4), although this was still ~40% of global forest area. They find 56% of protected forests have high landscape-level integrity (table 2). I agree with the authors that forest modification and degradation is important, but I don't think the authors made a good case that a) their findings are new / surprising, or that b) just mapping proximity to people is a great way to estimate ecological degradation let alone prioritize conservation action. It's true that being farther from people is generally helpful to forests, but the flip side is that this paper heavily devalues the natural areas that people most appreciate for recreation and ecosystem services, even though with high ecological function.

Li et al. 2020 evaluates the new ICESat-2 LiDAR satellite's forest canopy height product in northeastern China, and uses machine learning applied to radar and optical satellite imagery (Sentinel-1, Sentinel-2, Landsat-8) to fill in the gaps between ICESat-2's sample points. They found ICESat-2 performed well (compared to aerial LiDAR) at moderate resolution (30-500m, with 250m performing best, and 10m and 1km performing relatively poorly). There are some issues (e.g., their aerial data was taken with deciduous leaves present, but ICESat data was taken in winter with leaves absent, and they note the correlations appear to vary by site so will always will require local calibration), but overall this confirms the utility of ICESat and offers some options for producing seamless forest canopy height maps.

Grantham, H. S., Duncan, A., Evans, T. D., Jones, K. R., Beyer, H. L., Schuster, R., Walston, J., Ray, J. C., Robinson, J. G., Callow, M., Clements, T., Costa, H. M., DeGemmis, A., Elsen, P. R., Ervin, J., Franco, P., Goldman, E., Goetz, S., Hansen, A., … Watson, J. E. M. (2020). Anthropogenic modification of forests means only 40% of remaining forests have high ecosystem integrity. Nature Communications, 11(1), 1–10.

Li, W., Niu, Z., Shang, R., Qin, Y., Wang, L., & Chen, H. (2020). High-resolution mapping of forest canopy height using machine learning by coupling ICESat-2 LiDAR with Sentinel-1, Sentinel-2 and Landsat-8 data. International Journal of Applied Earth Observation and Geoinformation, 92(April), 102163.

Teitelbaum, C. S., Fagan, W. F., Fleming, C. H., Dressler, G., Calabrese, J. M., Leimgruber, P., & Mueller, T. (2015). How far to go? Determinants of migration distance in land mammals. Ecology Letters, 18(6), 545–552.

p.s. If you'd like to keep track of what I write as well as what I read, I always link to both my informal blog posts and my formal publications (plus these summaries) at