Wrapping up the 2013 #SWSAnnualMeeting

Last week*, the Society of Wetland Scientists annual meeting took over the Duluth Entertainment Convention Center, in lovely Duluth, MN. There were several highlights and a few lowlights, but overall it was a great time.

Location: First, Duluth is not where I would throw my conference, unless it was for the local chapter of the Minnesota association of high school hockey coaches or Walleyes Unlimited, etc. I like flying directly into a big city, hopping a train and proceeding to do business at the conference. Duluth just screams, "we got a smokin' deal on this venue" and "In light of this deal, your airfare will be $200 more than flying into Minneapolis or Madison." I didn't help to organize the conference, so I'll opt out of the Henny Penny critiques. If I didn't hustle up the conference, I don't have much right to hate on the locale, etc.

Venue: The DECC was sufficient, and had pretty nice rooms for talks, socials, lunch, plenaries, etc. It was also within walking distance of all major lodging - good call there.

Plenaries: Of the four plenaries, Joy Zedler and Gene Turner gave total rockers. Joy discussed plant invasions across her ecoregion, using case studies from the UW-Madison Arboretum to make her points. Gene discussed the loss of marsh habitats in the Mississippi River Delta and how these losses might synergistically interact with anthropogenic global change.

Highlights: I was very fortunate to participate in a day of wetland restoration talks (my full slides here) put on by Andy Herb, Owner of Alpine Eco. It was a greats session, with everything from mapping restoration objectives in the Upper Midwest to looking at the restoration of alpine fens in Colorado and desert rivers in Utah. A big thanks to Andy for having everyone out to give talks. The sessions on geomorphology and wetland mitigation were bangers, and lots of folks delivered great talks over the course of the week.

I had a great time engaging with the undergraduate mentoring program students and mentors. The students were exceptional and all of them put out great posters on some impressive research. I was particularly impressed with the posters of Denzell Cross, Dayvis Blasini, Anita Arenas and Mitchell Hinton. Look out ecology world, because these aspiring young scientists are waaaaaaaaay ahead of the game. We're talking big questions and advanced methods including AFLPs and well-done ordinations coming from young students. Every student had really learned from his or her research experiences and was already thinking about how he or she could improve the work they had done in the future. Ambitious, thoughtful and impressive are the primary adjectives that come to mind when discussing this year's student cohort. A hearty thanks goes out to Frank Day at ODU for his hard work to make the mentoring program a reality and to all of the sponsoring SWS Chapters and ESA SEEDS.

Four extremely talented women being awarded one of SWS's highest distinctions, being named a fellow of the Society. These awards come at a time when the twitter hashtag #miserablesexratio is used to discuss many science meetings and achievement panels. These individuals have been doing incredible research for decades, so it was a pleasure to see Beth Middleton, Joy Zedler, Katherine Ewel,  and Carol Johnston lauded by the Society for their long and important careers. I should also note, that these awards have historically been very gender balanced, reflecting that SWS is not the old white dude party that other societies may be. The full list of awardees is here.

The SWS Women in Wetlands section threw a great breakfast (yes, it's open to everyone) and discussed some very real and important issues of gender in ecology, research and private-sector consulting.

Overall, it was a great conference. Like all of the SWS meetings, I had a wonderful time, saw some exceptional talks, met some great people and got even further behind on real work. I'm already looking forward to the Joint Aquatics Meeting next year in Portland. After all of the fun, there were some looming scientific problems that many polite individuals didn't point out on the spot...

Lowlights: The biggest lowlight was seeing several talks that had very clearly pseudo-replicated designs. If you put your left hand on your head, you have one head. If you put your left and right hands on your head, you still have one head. If you count two heads just because both hands are touching it, well, you're wrong. These pseudo-replication errors are especially egregious in the greenhouse or lab, where replicates can be easily made by allocating treatments within smaller units rather than allocating multiple plants of the same treatment to larger mesocosms, etc.

The second lowlight is people not accounting for spatial autocorrelation within their field sampling schemes. If you collect plots along a transect, you technically need to account for this transect in your analyses. Better yet, one should probably take many transects with fewer quadrats and then pool their data within each transect in a given locale. For example, if you have ten watersheds, why take ten quadrats on four transects (and treat each quadrat as individual, independent replicates) within each watershed? You could just as easily take forty spatially balanced, random quadrats in each watershed. Transects can be good for gradients, but man, people need to look at what their experimental units are. Generally it will be transect level data comprised of pooled subsampled units.

The same goes for spatial effects within treatments. If you have thirty field sites, and ten are treatment A, ten are treatment B and ten are treatment C, then try to make sure that the ten A sites don't all cluster together and the ten B sites don't all cluster together, etc. This spatial effect confounds the treatment effect, especially in natural experiments where collinear gradients almost certainly exist.

Lastly - please, please, please test for statistical power before deciding that you've sampled enough. Just. Do It.

I really suggest everyone read up on Hurlbert 1984 before systematically throwing quadrats out onto the landscape and assuming statistical independence just because they wish it so.

*Whirlwind travel to a wedding in the Northwest leaves me wrapping this up a week after the fact. Sorry 'bout that.