A bit of metadata: As you may remember, I mentioned in my last post that I was super-pregnant. Well, in true devotion to my craft, on Darwin Day, I doubled my fitness, and now I have a new writing buddy. Concurrent to this1,2, I also submitted revisions to a manuscript which had come back from review the week before. The paper was accepted several weeks later. This paper has a lot to it, so I will be exploring a few aspects of it in a series of blog posts. This is the first.
In this post, I will share some opinions based on my interpretations of the results from this recent paper- it in no way represents the views of my employer, and is not intended to be used as a specific recommendation for pest management decisions on any individual farm.
So, I’m pretty excited about my latest paper (read about Shifts in dynamic regime of an invasive lady beetle are linked to the invasion and insecticidal management of its prey, now available as a preprint at Ecological Applications).
I tweeted about it as I submitted the final copy-edits:
Reasons I love this MS: 1) Ladybugs 2) Model Selection 3) 100% #otherpeoplesdata 4) Fully scripted analysis 5) Cool, topical results
— Christie Bahlai (@cbahlai) March 12, 2015
This paper hits almost all of the things I’m most excited about, and I’ll get to the other stuff in later posts.3 Today, though, we’re going to talk about what we found in this paper, and what it means for management of agricultural systems.
Neonicotinoids.4 Seems like everyone is talking about them lately- particularly their impacts on bees, birds, and their general efficacy at controlling target pests. It’s all pretty complicated, and I, like many others, believe the issue is being over-simplified. I am not going to take a position on whether we should continue to use neonics,5 but I do have a bit of a new perspective to add to this debate.
So, on to the study. First, I’ll give you a bit of background on the system we’re talking about.
Soybean is a major field crop in central North America. In 2000, a new invasive species, the soybean aphid, arrived in North America and swiftly became a really nasty pest in a crop that previously had very few troubles with insects. Our region had several major outbreaks immediately following their invasion. Soybean aphid is eaten by a broad suite of predators, including another notable invasive species, the Multicoloured Asian ladybeetle, Harmonia axyridis.
Since 1989, ladybeetle populations have been monitored at the Kellogg Biological Station in southwestern Michigan. This monitoring captured the 1993 invasion of Harmonia to the region, and the subsequent population dynamics of this species.
In a 2012 paper, my co-author and study PI Doug Landis remarked on an apparent pattern in the dynamics of Harmonia: 1) generally moderate populations of Harmonia prior to soybean aphid arrival (in 2000), 2) a distinct 2-year cycle in peaks of Harmonia during the acute phase of soybean aphid invasion (2001-6) and a return to moderate populations of Harmonia afterwards. In the Ecological Applications paper, my colleagues and I use quantitative techniques to determine if the patterns we observe fit certain theoretical patterns, and if so, if we could detect changes in the pattern, and figure out when they were.
In short, they do,6 we can,7 and we did. What our model tells us is that after 2000, corresponding with the invasion of soybean aphid, Harmonia‘s intrinsic rate of increase (a measure of how fast a population can grow) went up by 20%. Simultaneously, the carrying capacity for Harmonia, a measure of how large a population an environment can sustain, went up by 40%. This is not surprising- a new, highly abundant, and highly suitable aphid food source suddenly became available- this means good times for ladybugs. What is surprising, though, is what happens in 2006: our model tells us that at this time, Harmonia‘s dynamics return to values identical8 to those before soybean aphid’s arrival- essentially, from a quantitative standpoint and looking at it from the Harmonia perspective, it’s like soybean aphid stopped existing.
Now, soybean aphid did not disappear from North America. In fact, I started my PhD studying this insect in 2007, so I’m absolutely certain it didn’t. But the aphid did become a lot harder to find than it had been previously. Outbreak populations still happened, but they became rarer, less severe, and patchy. It seemed likely that, because Harmonia‘s first shift was clearly related to the invasion of soybean aphid, its second shift was related to this apparent decline in soybean aphid. However, we still didn’t have a clear explanation for what actually happened to the aphids.
We turned to scouting records and extension publications to get regional estimates of soybean aphid infestation, and then we compared these records, region by region, year by year, to estimates for pesticides used in soybean and we found an unexpected result: that soybean aphid outbreaks have a negative association with neonicotinoid use- that is, regions with more neonicotinoid seed treatment use have fewer or less severe aphid outbreaks.
Why was this result unexpected? Well, there is a growing body of literature that shows neonicotinoid seed treatments are largely ineffective at preventing aphid outbreaks in soybean, when you compare treated and untreated fields side-by-side. Neonics usually only last about 40 days after planting, and soybean aphid outbreaks usually occur mid-to-late summer, so there is no insecticidal activity at the time when aphids cause economic damage. However, for reasons we detail in the paper9, it appears that these insecticides can have dramatic effects on early season establishment of these aphids, and, as I’ve argued before, early season survivorship is extremely important in season-long aphid dynamics. This, in turn, affects Harmonia– fewer aphids to eat- both early and late in the season.
Although this result is correlative, the evidence suggests that widespread adoption of a prophylactic seed treatments are having indirect negative effects on Harmonia (which could be a good or bad thing depending on your view of Harmonia-as a valuable biocontrol organism or as a disruptive invasive species)
Further, our study suggests that if seed treatments were to be limited (EPA reviewing their use in America right now, and Ontario has already introduced legislation to dramatically reduce their use.) there could be a return to increased soybean aphid and more explosive Harmonia population dynamics. To mitigate this outcome in the event of changes to neonicotinoid use, we suggest alternate management strategies, including host plant resistance and biocontrol, which could potentially offer similar, landscape-level suppression of these two invasive species, but without the controversy that follows the use of this class of insecticides.
1. Well, not *exactly* concurrent to this. The manuscript revisions were delivered about 21 hours before the baby was. I’m not superwoman.
2. As such, I’m giving you fair warning that I’m likely to pepper my posts with yet even more sciencing-while-parenting and women-in-science type asides.
3. Wherein we will talk about the fun stuff- i.e. the ~*MATH*~<3<3
4. That got your attention, didn’t it!?
5. !! WARNING OPINION TO FOLLOW I believe judicious use of insecticides is a necessary part of maintaining productivity in our current agricultural systems, and for the most part are used safely within the agricultural systems with which I’m familiar (US and Canadian field and horticultural crops), HOWEVER I believe that we, in general, use pesticide more often than we need, and can dramatically reduce environmental risk using integrated pest management systems. END OPINION PLEASE DON’T FLAME ME BUT ALSO REMEMBER I HAVE A PHD IN THIS I REALLY JUST WANT TO TALK ABOUT COUNTING LADYBUGS ON THIS BLOG
6. Our Harmonia population time series fit the Ricker model very, very well.
7. Using model selection, we determined that our patterns were best explained when we computed our model using two discrete break points in the time series.
8. Within statistical error, but with nice, narrow error bars- so close that model selection says they’re the same model. That’s two statistical paradigms backing me. So, this:
9. And reasons I will go into detail with in an upcoming post- oh, the math we’ll enjoy together, friends.