Article alert: #Environmental filters & #biotic interactions shape #riparian vegetation guild distributions @ESA_org #Ecosphere

Article alert: Multi-scale environmental filters and niche partitioning govern the distributions of riparian vegetation guilds



How are riparian plant species with shared life-history strategies, or "riparian vegetation guilds" distributed across landscapes? What environmental resources and processes influence these distributions? How does competition influence guild distributions and coexistence?

These are but a few fundamental questions in riparian plant community. They also happen to be the questions that I set out to answer somewhere in the summer of 2013, using vegetation and stream data collected across the U.S. Pacific Northwest. After doing some community analyses (see Hough-Snee et al 2014a) and work to see how riparian vegetation influenced instream large wood (Hough-Snee et al 2014b) in the same region, I turned my attention to identifying groups of species that occur along streams and that have clearly shared life history strategies, or riparian vegetation guilds (sensu Merritt et al 2009Merritt et al. 2010). My idea was that these life history strategies have been most commonly studied in response to a single resource or disturbance gradient, often hydrology.

When creating riparian flow-response guilds, species are grouped based on their traits that relate to water balance and tolerance of fluvial disturbance. By identifying which guilds have more hydrophytic or xeric strategies, guilds can be modeled to predict how vegetation changes as flows are modified by dams or diversions. I am a firm believer in this concept as a decision support tool. I also realize that not all streams' vegetation will respond to hydrology in a uniform fashion. For example, large, alluvial rivers will often show distinct vegetation on surfaces with different flood recurrence intervals. Bars and low terraces that are flooded frequently often exhibit more mesic and disturbance tolerant species (guilds) than infrequently inundated high terraces, hill slopes or canyon walls that have less obligate fluvial species and more drought and upland disturbance (e.g. wildfire) tolerant vegetation.

Guilds were determined using clustering of woody species by their component functional and morphological attributes. 
The dataset we used was small, low-order, wadeable streams, many of which are headwater streams where hill slopes and channels connect. This convergence between uplands and streams allows many environmental filters to shape what species occur at a given location. I set out with some friends and colleagues to put together a database of traits that describe each species' life form, persistence and growth, reproduction, and resource use in the riparian environment. By taking an exhaustive approach, we hoped to capture guilds that responded to multiple disturbance, resource and competition axes.

Guilds were differentiated based on numerous differences in above- and belowground architecture, stress responses, resource use, growth rate, etcc.
Immediately we ran into a problem: of the hundreds of species in the study region, very few had sufficient trait data to use exclusively quantitative traits that describe water and energy balance, and resource use. For many herbaceous species, simple categorical descriptions of rooting strategy and leaf type were often hard to come by. Based on this limitation, we reduced our trait-species search to those woody species in the dataset that occurred in 5% of sample reaches.

After creating the species-trait database with collaborators, Lloyd Nackley, Lexine Long, and Brian Laub, we quantitatively identified five guilds: (1) a tall, deeply rooted, long-lived, evergreen tree guild, (2) a xeric, disturbance tolerant shrub guild, (3) a hydrophytic, thicket-forming shrub guild, (4) a low-statured, shade-tolerant, understory shrub guild, and (5) a flood tolerant, mesoriparian shrub guild. Next, I worked with Dave Merritt of the USFS, and my trusty committee member, Brett Roper and committee chair, Joe Wheaton to model each guild's distributions. We modeled these guilds presence and absence across the Columbia and Missouri River Basins, particularly their responses to hydrogeomorphic setting, upland disturbances, landscape forest cover, and biotic competition with other guilds.
Guilds, unlike communities, are not always sets of co-occurring species, but sets of life history strategies that may or may no occur on a landscape. Here is an ordination of the combination of co-occuring guilds plotted alongside correlations with environmental filters (A - C) and guild occurrence (D).
The punchline: each guild corresponded uniquely to multiple environmental filters that would likely select for each guild's set of functional and morphological attributes (traits). Where guilds coexisted, their traits often shaped each guild's niches so that these guilds could co-occur. For example, at sites where large, long-lived, canopy-forming evergreen trees occurred, smaller statured species with shade tolerant strategies or strategies that tolerate edge effects (disturbance) often occurred as well. From this work, we find that riparian guilds' distributions are shaped by both biotic interactions and pressures from different sets of environmental filters.

As of Monday, I am pleased to announce that you can read more on this work directly at Ecosphere, the open-access journal of the Ecological Society of America. The article is also hosted at Researchgate, if that's your preferred archive. Supplemental Materials are available at Ecological Archives.

Big thanks to the team of folks who provided feedback on this work, read early drafts, and provided encouragement. It's all greatly appreciated!