Interview with Dr. Zach Portman
Dr. Zach Portman conducting fieldwork
In our first blog post, we speak with Dr. Zach Portman, a bee taxonomist working in Dan Cariveau's group at the University of Minnesota.
Hi Zach, thank you for taking time to answer a few questions. First, can you give a brief description of what your current position is and what an average day on the job looks like for you?
I am the bee taxonomist in the Cariveau native bee lab at the University of Minnesota. That means I am the one who identifies all the bees for many different projects. In Minnesota alone, there are over 400 species of bees, many of which are poorly known and difficult to identify, so that’s what I spend most of my time doing. In terms of an average day, I spend a lot of time at a microscope looking at specimens. Though I also do a lot of work helping out with various projects, especially writing up results for publication in scientific journals.
What group of bees do you focus on in your research?
I did my PhD research on the genus Perdita, which is an extremely diverse group of bees that are mostly found in the arid areas of the western US. Since then, I’ve broadened my focus and become more of a general bee taxonomist. For example, recently I’ve done research projects looking at a recently introduced bee in the genus Pseudoanthidium as well as uncovering a cryptic species in the genus Andrena.
Earlier this year, you published a forum article about bee monitoring in Annals of the Entomological Society of America. In it, you outline a major conundrum when it comes to monitoring native bees: that each of the primary methods has its own inherent flaws and biases. Can you briefly describe this conundrum?
The fatal flaw with all the common methods (bowl trapping, vane trapping, and netting) is that they don’t actually measure bee abundance. The reason for this is that we don’t know what proportion of bees from the surrounding community are caught. For example, if you catch 100 bees in a bowl trap, did you catch 1%, 10%, or 50% of the bees in that area? The truth is we have no idea. Further, we don’t know how catch rates differ among the various species and how they are influenced by the environment. For example, do you catch more bees when there are more flowers because the flowers draw bees in, or do you catch fewer bees because bowl traps compete with flowers? We just don’t know.
As a result, these methods don’t allow you to detect changes in bee abundance, which is one of the primary goals of most monitoring programs.
This is further compounded by the fact that each method has taxonomic biases. For example, bowl traps catch many more sweat bees (family Halictidae) than other methods. And this is something that has received a lot of study, but only in relation to other methods. As a result, we know bowl trapping catches way more sweat bees than netting, but we don’t know which ones best reflect the underlying bee community.
“The fatal flaw with all the common methods (bowl trapping, vane trapping, and netting) is that they don’t actually measure bee abundance. The reason for this is that we don’t know what proportion of bees from the surrounding community are caught.”
Can you tell me a little bit about what inspired you to write this article?
It actually started out as a mini-rant I tweeted out due to the frustration caused by a project I was working on where I was identifying 10,000 or so Dialictus. Bowl traps catch tons of Dialictus, and for those who aren’t familiar, these little sweat bees are some of the most difficult to identity, to the point where only a handful of people in the US can identify them reliably. These specimens were from a monitoring program that was based largely on bowl traps, And I was just sitting there thinking to myself, “how in the world does this contribute towards monitoring?”
My tweet ended up generating lot of discussion, but I think my main point was misunderstood by a lot of people, since the response from many people seemed to be “I agree bowl traps are biased so we should be using other methods too!”
So, I ended up writing a long-form rant explaining why all the common methods are flawed and this grew into a full-fledged paper. Luckily, I was able to work with my coauthors, Dr. Dan Cariveau and Dr. Bethanne Bruninga-Socolar, to tone it down a bit from simply a rant about bowl traps and also try and suggest some ways forward.
There was also another strong motivating factor for me, which is that I felt like there really needed to be a voice representing the scientists who are unhappy with current bee monitoring methods. I think most bee scientists recognize that bowl traps and other passive methods are deeply flawed, but you wouldn’t know that from reading the scientific literature. There’s been a vocal minority of scientists who have been pushing bowl traps as the go-to methods for years and years. As a result, you see tons of well-meaning people and scientists who want to monitor bees who look up the best way to do it, decide to bowl trap, and then end up with tons of specimens with no idea how to identify them, no place to store them, and no idea how to analyze the data.
So this paper was my attempt to introduce a dissenting voice into the discussion and question the prevailing wisdom that bowl trapping is the best way to monitor bees.
In your article, you state that “we need to use methods that allow for more targeted collection of data that inform specific monitoring goals”. Can you give a few examples of what this type of strategy might look like?
One of the reasons I focus on trying to do more targeted collecting is because the bottleneck in bee monitoring generally occurs in the identification, storage, and analysis phases rather than the collecting phase. Especially with passive traps, it’s relatively easy to go out and collect thousands of specimens. It’s much harder to get good identifications on those specimens because there are so few people with that expertise.
So, we really need to step back and think about how to effectively monitor bees without overwhelming ourselves with a flood of specimens. And that means recognizing that we can’t monitor all the bee species. For example, it doesn’t really matter how many Dialictus or Nomada you catch if no one can identify them.
In terms of scaled-down goals, I think that there are a lot of good possibilities that include monitoring habitat specialists, monitoring focal plants to detect changes in ecosystems, or monitoring a smaller subset of species such as the bumble bees. The overall goal here is to avoid the bottlenecks that we currently face that lead to thousands and thousands of unidentified or poorly-identified specimens that don’t actually inform monitoring efforts and conservation decision-making.
Why is it important to monitor native bees at large regional (or national) scales? How does this relate to bee conservation?
Large-scale monitoring data are important because they inform other scientists, the public, and policymakers about how bees are doing. I think that the lack of good monitoring data is really hurting bee conservation efforts because it prevents us from making informed decisions about how to best allocate resources for bee conservation, especially in terms of which threats to prioritize and which habitats to protect.
If you were tasked with designing a national native bee monitoring program for the US, what would it look like? Would you target specific focal bee groups, focal plant taxa, and/or focal regions? Would community scientists be involved? What other components would your program have?
In terms of designing a national monitoring program, I think that the first step needs to be to step back and really consider what the goals are and what management decisions will change in response to new monitoring information.
I think a big problem we have right now is that there is a lot of indiscriminate collecting that falls under the umbrella of “monitoring” but doesn’t actually contribute to monitoring goals.
For a national monitoring program, I would definitely want to focus on a narrower set of bees in order to get good, actionable information, rather than trying to do too much and end up with a lot of poor-quality data. I’m hesitant to recommend specific ways to do this because I think there are a lot of potential solutions that are still being developed. Two methods that I think show a lot of promise are new genetic methods for estimating population size as well as the growing number of community scientists. Especially for bees that can be identified from photographs, community data can provide rapid preliminary data that can help detect declines as they occur, rather than waiting years to figure out that a species has declined after-the-fact, like what happened to some of our declining bumble bee species.
Lastly, I think there needs to be serious focus and funding for basic taxonomy and natural history work. For so many bee species, they are either impossible or extremely difficult to identify because their classification and identification resources haven’t been updated since the 1950’s or 60’s. Not to mention all the undescribed and poorly known species. And even for the species we can identify well, we know so little about the biology of most of them. That means we can end up in a situation where we can detect declines, but don’t know enough about the biology of a bee to be able to help it recover. Without serious funding and resources dedicated to basic taxonomy and natural history, we’re just going to end up spinning our wheels when it comes to effective monitoring.