Sunday, September 13, 2020

How scientists can improve their impact

Dog resting her head on her paws

This May a paper we've been working hard on for about 2.5 years finally came out! The basic idea is to provide tips for scientists to improve the chances that their research will have its desired impact. Essentially it's the paper my co-authors and I wish we had when we were starting as scientists. The dog picture above is 100% unrelated, sorry.

We have talked about this paper with well over a hundred people, and they all liked different things, and had different requests for accompaniments to it! Some wanted more context, some wanted a super-short version of it, some wanted video, etc. So we put together a whole package of resources (listed below and all available at; please take a look at whatever appeals to you.

  1. The full paper (~6,000 words, but we use simple language so it’s a fairly quick and easy read). It has context for why this matters, specific recommendations, and examples of what each recommendation looks like in practice.
  2. The need for this paper is covered in a Science brief on Cool Green Science (~500 words, 2.5 min reading time) –  it briefly explains the idea of the paper and not much else.
  3. The gist of the paper (a summary of the recommendations and brief examples) is available in a high level overview which also links to all of the products listed in this blog: (~900 words, ~4 min reading time). We also have a downloadable version of this overview to print and share (requested by a professor who wanted a short handout for her students).
  4. We talk about how we wrote the paper and what surprised us when writing it in an interview with OCTO (Open Communications for the Ocean) (~1,100 words, ~5.5 min reading time).
  5. There's more on why we wrote the paper and how scientists can start using it in a Cool Green Science interview (~2,500 words, ~12 min reading time).
  6. Finally, if you’d prefer video to text, we have a recording of a webinar about our paper which focuses on summarizing our recommendations and how they can help scientists avoid ‘wasting’ their research (22 minute presentation plus 35 minutes of discussion)
  7. Wondering what other people think is most important to improve scientific impact? This video recording of a panel discussion with four different research impact experts offers additional insights, and you can also read their advice as a blog. Or here's a recording of a second panel discussion focused on Latin America, the Caribbean, and African contexts.
  8. Finally, I shared some of the challenges in writing this paper (finding co-authors, dealing with critique, etc.) in an interview at Wildhub.

Tuesday, September 1, 2020

September 2020 science article summary

Millipede with witches butter fungus 


This is another short summary with just four articles on biodiversity (bugs in the US, global indicators, tropical moist forest quality, and bias in conservation textbooks in terms of which taxa etc. get featured).

If you know someone who wants to sign up to receive these summaries, they can do so at (no need to email me).

There have been a lot of papers documenting declines in invertebrate populations, from bees to flies, sometimes called the "insect apocalypse." But Crossley et al. 2020 use a large data set (from the Long-Term Ecological Research sites) to show that in much of the U.S., there's no clear trend (up or down). For abundance, some species are declining in some places, others are increasing, and overall the trend is pretty stable on net (See Fig 2 for details, including the exceptions to that pattern). Diversity is similarly flat on net (see Fig 3). The discussion (on the page w/ Fig 3) of possible explanations for why this paper had different results from others is interesting. They include: 4/5 sites this paper included that another seminal paper omitted showed positive trends, total abundance trends across spp. heavily weight the most numerous spp. and dwarf other changes, and this paper relied on more recent data (where others have found a decline is slowing).

Hansen et al. 2020 is a global analysis of moist tropical forest ecological quality and a great read. They use forests with high structural condition (meaning tall forests with several layers of understory trees and other plants, and high variation in plant size) and low human pressures as a proxy for overall ecological integrity (which typically also includes composition and function). The argument is that these forests have more habitat niches and can support more species, and that degraded structure is often due to stresses like logging which can have broad impacts (although they note limits of their approach up front). Fig 1 is a map w/ their results (& Fig 2 is a more helpful chart): they found 47% of remaining tropical moist forests had high integrity (both high structural condition and low human pressure, mapped as dark green), 33% had low structural condition (mapped as brown), and 20% had high structural condition but substantial human pressures (mapped as light green). 76% of the intact forest is in the Americas. In good news, forest w/ the best structure is being lost more slowly than more degraded forest (likely due to their remoteness, see fig 3). They have an ambitious suite of spatial recommendations in fig 4: extending protection to all remaining high integrity forests, plus restoration and working to reduce human pressure on the other forests.

Hoban et al. 2020 argue that new indicators are needed for a post-2020 CBD global framework for biodiversity. They recommend three new indicators: 1) # populations with effective population size above 500, 2) # current populations / # historic baseline of populations, 3) # species & populations w/ DNA-based genetic diversity monitoring, as well as keeping two existing CBD indicators (comprehensiveness of conservation of all species; and # of resilient, representative, and replicated plant genetic resources secured in medium or long-term conservation facilities). It's a fairly simple approach (albeit hard to empirically measure) for genetic biodiversity indicators.
Stahl et al. 2020 looked at 7 recent conservation textbooks and bias in what they focus on relative to natural prevalence (Fig 5 has a good summary). Some bias comes from underlying factors (research doesn't focus on species in proportion to their prevalence, more funding goes to charismatic species and richer countries), but regardless of the source they compared the proportion of examples to their prevalence on Earth. As you'd expect, the books favor examples using mammals over amphibians, North America over other continents, forests & coral reefs over other ecosystems, and tropical over temperate regions. It's an interesting topic, but there is at least one error (they claim only 3 of the textbooks mention ecoregions, but one of the other 4 discusses them at some length including an ecoregional map I created) which makes me wonder what else they could have gotten wrong (the author is looking into it and will get back to me). Ironically, the authors don't comment on potential bias in how they selected textbooks (e.g. only English language) or the methods they used (a focus on proportion of examples regardless of their value in explaining concepts). 

Crossley, M. S., Meier, A. R., Baldwin, E. M., Berry, L. L., Crenshaw, L. C., Hartman, G. L., … Moran, M. D. (2020). No net insect abundance and diversity declines across US Long Term Ecological Research sites. Nature Ecology & Evolution, (Table 1).

Hansen, A. J., Burns, P., Ervin, J., Goetz, S. J., Hansen, M., Venter, O., … Armenteras, D. (2020). A policy-driven framework for conserving the best of Earth’s remaining moist tropical forests. Nature Ecology & Evolution.

Hoban, S., Bruford, M., D’Urban Jackson, J., Lopes-Fernandes, M., Heuertz, M., Hohenlohe, P. A., … Laikre, L. (2020). Genetic diversity targets and indicators in the CBD post-2020 Global Biodiversity Framework must be improved. Biological Conservation, 248, 108654.

Stahl, K., Lepczyk, C. A., & Christoffel, R. A. (2020). Evaluating conservation biology texts for bias in biodiversity representation. PLoS ONE, 15(7), 1–11.