Note: For an update on this issue, see this new post.
This post documents a research project I have been heading, with the help of University of Washington undergraduates, and a group of concerned citizens from the Seward Park neighborhood.
Seward Park is one of Seattle’s flagship parks, and one of only 2 parks I know of in Seattle that has any substantial remaining old-growth forest. In 2013, Catherine Alexander, a local citizen naturalist and daily user of Seward Park, noticed that Sword Ferns (Polystichum munitum) were mysteriously dying in the understory of Seward Park’s old-growth. A frequent park user myself, having recently moved into the neighborhood, Catherine p0inted out the problem to me in the winter of 2014/15. I was convinced the ferns would make a comeback in the spring, but I was surprised when nature proved me wrong. While several other local citizen naturalists sensed the problem was worsening, no one had actually taken the time to quantify the extent of the damage and spread of damage, let alone address the causes. In 2015, Paul Shannon, a leader of a citizen group called “Friends of Seward Park”, enlisted me to help get a handle on the problem.
With the help of two excellent undergraduates from the University of Washington, Kramer Canup and Tristan O’Mara, we began to consider experimental designs for monitoring the current extent of damage and change in damage over time. While we considered a gridded plot system which would cover areas inside and well outside the die-off zone, as well as a protocol for fine scale transects crossing the perimeter of the die-off area (to detect expansion of die-off), limited time, funding, and human power caused us to settle on more minimalist protocols. To map the perimeter of the damaged area (interestingly, the damage seems to be worst at central point, and is spreading radially from that point, rather than patchily throughout the forest), we used GPS and GIS to mark and connect points at which 50% of the individual ferns in a given area exhibited at least 30% dead fronds (Figure 1). To better understand the dynamics of the die-off over time, we randomly placed 20 plots inside and just outside the perimeter of the damage zone (Figure 1). Within each of these plots, we mapped every fern and counted fern stems categorizing them as living (if they had any amount of green coloration at all) or dead (if they had no amount of green coloration). Adjacent to each plot, we also sampled for fungal pathogens on fern roots in collaboration with the Washington State University extension program. All plots were surveyed in November 2015 and again in May 2016 (see results below). We hope to repeat all of these protocols again in November 2016 and again in 2017.
Figure 1. We randomly placed 20 5m x 5m plots throughout the region of fern damage in Seward Park. The red line polygon encloses the area of worst damage, nearly 100% death of all sword ferns, the area in which die-off was first noted in 2013. The black polygon marks the outer edge of fern die-off according to a protocol we developed. Blue plots are plots that gained fern stems between November 2015 and May 2016, and red plots are plots that lost fern stems over the same period.
Tristan O’Mara and Kramer Canup (UW undergrads) surveying ferns in November 2015, in a healthy sword fern understory.
Tristan O’Mara and Kramer Canup surveying ferns in a plot at the center of the die-off zone. Nearly 100% death of the ferns is observed here. Prior to 2013, the understory in this area looked much like the understory in the image above. The density of dead fern fronds in this area rivals the density of live fronds in the area pictured above.
Laying out the plots, and staking the corners. November 2015.
After the resurvey of the 20 plots in May 2016, some initial results are in. The punch-line is that some plots/ferns gained green stems (not surprising in the spring-time), but most plots/ferns lost green stems (surprising given that it was spring). On average, there was an 18% loss of green stems (p<.002) across all of the ferns across all of the plots (Figure 2).
Figure 2. The average number of live stems per fern declined by 18% from November 2015 to May 2016. This is especially surprising given that our 2016 resurvey was in the spring. One would expect a brief uptick in green stems in spring, following typical summer die-back in the previous growing season.
Figure 3. Sword ferns neatly snipped and stacked at the entrance to a mountain beaver burrow in Seward Park.
While documenting the status of the die-off has been an important first step, there are many possible reasons for the die-off. Unfortunately, we have very little insight into the reasons. 2015 was an unusually hot and dry year, but the problem began under normal weather conditions in 2013. Still, we may see some recovery following this summer (yet another reason to continue monitoring). Testing for pathogens in the soil and root systems has yielded negative results. There are some pathogens there in small quantities, but not in quantities expected if they were the sole cause of the die-off. Die-off seems to be worst on an east-facing slope that is riddled with mountain beaver (Aplodontia rufa) burrows. Sword fern is one of their favorite foods, as evidenced by the photo above (taken in an area of healthy fronds neighboring the die-off zone)(Figure 3). I hypothesize that mountain beavers (largely unchecked by predators in Seward Park) may be part of the issue here. I have found other areas in the park that show evidence of historical sword fern die-off, and my sense is that predators such as coyotes, which come and go from the park over time, may cause ups and downs in the beaver populations. Currently, there are no coyotes known to be using the park.
Future work: With the help of Patrick Tobin (UW College of Forest Resources) and graduate students, we will expand the plot system we have already established, monitoring and mapping ferns and overstory trees. We plan to take a detailed look at soil arthropod communities inside and outside the die-off zone. We will monitor soil nutrients and soil moisture in a methodical way inside and outside the die-off zone. We will continue sampling for pathogens. And most importantly, we will begin a sword fern planting experiment, using a paired design (in which each treatment is paired with a control), with possible treatments including mountain beaver exclosures and fungicide application.
I believe that systematic monitoring and experimentation are an important step forward for understanding the ecology of Seattle parks. Amazingly, there is very little of this going on in Seattle’s parks, given the value of these natural spaces to the city. Volunteers donate countless hours to restoration projects and general stewardship, but we need to keep track in a quantifiable way, of how successful these efforts are, in order to get a better handle on the health of our natural spaces. With respect to the sword fern die-off, at this point in time we simply don’t know enough to justify gathering a restoration crew and re-planting the area with sword ferns. Due to lack of systematic monitoring across Seattle parks, we don’t even know if this sort of die-off is occurring in other Seattle parks (although as far as we know it is not).
Going forward, we will need help in the form of human power (undergrads, grad students, citizens?) and funding. Contact me at timbillo (at) uw.edu if you’d like to help