Thursday, June 28, 2018


Since 2014 we have been observing and attempting to understand a dramatic die-off of the dominant understory sword fern species, Polystichum munitum, in the 120-acre old growth "Magnificent Forest" in Seattle's Seward Park

Please add your own observations at our iNaturalist project.

Extensive lab and field work, consultations with fern experts throughout the US, and recent reports of die-off from elsewhere in the Puget Lowlands of Washington state (see annotated regional map), lead us to conclude with considerable confidence: 

  • The die-off is caused by an (as yet unidentified) pathogen 
  • Climate change may be a contributing factor, but until the proximal cause of the die-off is determined, the role of climate change can not be known.  
  • Drought is probably not a cause (see rainfall records below).
  • Is unprecedented, not part of the P.munitum life cycle, in which ferns colonize open ground, do not reproduce under a closed canopy, with individual plants living for hundreds of years.
  • Holds implications for the forest at Seward Park, and for other forests in the region.

Some selected blog entries

Contact: Paul Shannon, Forest Steward

Monday, May 7, 2018

Fort Worden Recovery

Darrell Howe reports on a recent visit to Fort Worden that the ferns we thought were dying in 2017 are in fact recovering, sprouting new fiddleheads.   See his iNaturalist report and photos.   Last year's blog post of what we then interpreted as die-off is here.

This suggests a few different possibilities:

  1. The putative pathogen which we think causes the die-off is not always fatal: it can also cause a die-back from which ferns can recover.  We may see this phenomenon at Ground Zero in Seward Park, where two ferns (out of ~100),  which I previously judged to be dead, are now recovering.
  2. Some other phenomenon lies behind last year's apparently dead ferns at Ford Worden.  
Whatever the explanation - and ongoing observation may clarify - this recovery is very good news.

This recovery at least partially fits the pattern reported by the Pitterman Lab describing sword fern die-back due to drought in the redwood forests of California.   See the blog post here.  In brief:  fern response to drought includes stomatal closure followed by xylem embolisms followed by die back.  The ferns return in successive years, but continued cycles of drought + embolism lead to die-off.

Note that the California drought was more severe and lasted longer than what we have seen in the Pacific Northwest.  See precipitation records and California comparion here.

Thursday, April 5, 2018

What's Next? A proposal to plant Oregon Grape from native seed

My initial shock has diminished - at the broad loss of sword ferns to the north of Ground Zero, between the Hatchery enclosure and the sqebeqsed trail.  Some of that loss can be seen in this photo, looking northeast from near the sqebeqsed/Hatchery trail junction.  The original Ground Zero is out of frame to the right (south and a bit east).

Tim and Kramer have put in some observations plots, and Suzanne put in a 12-fern experimental planting,  north of (and to the left in) this photo.   Matt & co. planted perhaps 50 ferns just down-slope from the scene pictured here as part of (I hope I have this right) an eco-physiological study.

About 30 yards to the north of this site the forest understory gets healthy again.   Last week I went looking in that area for healthy ferns to monitor.  We do not yet have a detailed record of how fern fronds are affected, from health to first symptom to full death.   I figure that a photographic record of a few individual fronds on different plants may be useful.

While looking around for plants to monitor, I came across a few large areas dominated by healthy Oregon Grapemahonia nervosa. At Seward, healthy ferns often appear in mixed communities with this species.  Salal can be  intermixed as well.  All three species appear in various ratios, with complete dominance by a single species sometimes occurring,  but often in mixtures on a sliding scale, of the three species.

I now feel that complete sword fern recovery at Seward Park is a distant prospect - at best.  There is little to no evidence of natural regeneration  - of ferns, or any other species.  The only exception to this which I have seen thus far at Seward is a large bloom of the invasive weed Herb-Robert - in an area (not quite visible here) at  the top left of the above photo.

This all leads me to suggest that
  1. The next urgent task at Seward is restoration of native understory where the sword ferns have died off.
  2.  Oregon Grape and Salal are the best candidates for planting
  3.  Growing these from native seed, collected from this forest,  means that restoration will be inexpensive and low risk.
I bet there is a lot of practical knowledge and advice around on how to grow mahonia from seed.  Some casual web searching suggests that seeds are best collected in early September and then planted right away so that 'stratification' happens naturally over the winter.  Growth is slow in early years.  Maybe extra watering will not be needed.   Perhaps salal is similar?

I am eager to hear advice, suggestions, and comments on this plan.

Wednesday, March 28, 2018

Invasive Earthworms?

Astute reader Betsy Dowling, a Seattle native now studying urban horticulture at Mesa Community College in Arizona, noticed the similarity between the Seward Park 2011-2017 before and after photos which introduce this blog, and this pair, from the 5th edition of Soil Science and Management, Edward Plaster, 2008 - copied here (alas) without permission:

On a quick trip to Seward Park last night I performed a ridiculously ad hoc assay, digging and examining a cubic foot of soil from three spots:  ground zero (GZ: 1 worm),  active die-off zone (ADZ: 4 worms) and a healthy dense sword fern community off the Erratic Trail (ET: 4 worms).   This establishes nothing other than that earthworms (no species ID) are present.  

Here is a hodgepodge of academic papers.   2002 research by Michael Gundale apparently introduced the topic, describing on the effects of invasive earthworms on a rare, mostly underground fern:

Influence of Exotic Earthworms on the Soil Organic Horizon and the Rare Fern Botrychium mormo

The research is summarized in the  Conservation Magazine.  The original paper is listed here, with this abstract:

Forests north of the last glacial extent have no native earthworms. Exotic earthworms are now colonizing forests that are naturally free of earthworms. It is currently unknown how these exotic earthworms might affect rare plants. To determine whether there is an association between the presence of an exotic earthworm species and extirpation of the rare fern Botrychium mormo, I surveyed 28 populations documented and counted previously. I estimated current population sizes of B. mormo, soil horizon thickness, earthworm species present, and carbon content, nitrogen content, and pH of the A soil horizon. Two earthworm species were abundant, Lumbricus rubellus and Dendrobaena octaedra. Dendrobaena octaedra had no significant association with any soil variable or with B. mormo extirpation. Lumbricus rubellus was significantly associated with B. mormo extirpation and a mull humus type. Where L. rubellus was present, O1 and O2 horizons were significantly thinner. I conducted a laboratory microcosm experiment to determine whether L. rubellus could create the conditions it was associated with in the field. Microcosms with L. rubellus resulted in a significant reduction in the thickness of the O1 and O2 horizon and a significant increase in the thickness of the A horizon. This experiment suggests that L. rubellus created the conditions with which it was associated in the field. The intrinsic rate of increase (r) of B. mormo was best explained by the O2 depth, which implies that this soil layer supplies a critical resource. My results support the idea that exotic earthworms alter the forest floor, leading to negative changes in native vegetation.

A 2016 review from the German Centre for Integrative Biodiversity Research:  

European earthworms decrease species diversity in North America

2009:  Distribution and impacts of invasive earthworms in Canadian forest ecosystems.  A short section titled Coast and Columbian forests of British Columbia will be of some interest.

Thursday, March 15, 2018

A Lazarus Fern?

When I returned to Ground Zero a week after the two lines of twelve ferns were planted, in order to take the first set of time-lapse photos, I was shocked and delighted to discover a healthy fern which had not been planted by us.    Since sword ferns, we are told, will not reproduce under a closed forest canopy, I assume that this fern has come back from the dead:  enough life lingered in the rhizome for it to regrow after three years of inactivity.   Hence: the Lazarus Fern.   Definitive evidence for this claim - of regrowth rather than new growth - could come from removing soil and locating (or not locating) a large rhizome; young ferns do not acquire a rhizome for at least five years.    I do not want to disturb this rare, resuscitated, old survivor - or is it a new young plant? - not yet.

Based upon the size of the seven fronds, and the time of year, these fronds must have been 2017 growth.   I did not notice this frond last year.  This is puzzling, but not completely unlikely:  inasmuch as  I did not expect to see any fern regrowth,  I may well have overlooked this one instance.

There are a few other rather straggly ferns surviving at ground zero.  They are all - except for this fern, and one other I will describe in a moment - found very close to the trail or to large trees.   All of these survivors are failing, and fewer of them are found with each passing year.   Perhaps some sort of micro-site effects, or interspecies cooperativity, is responsible for their imperfect survival. 

The other healthy fern found at ground zero is one of the two nursery ferns we planted in November 2014, at the suggestion of Jillian Weed, parks plant ecologist.   I watered both ferns through the first two summers; one has survived:

This survivor, and the surprising regrowth of the Lazarus Fern suggests a possibility: that the putative pathogen killing the ferns,  which spreads at the rate of about 30 meters a year, may now be absent from ground zero.    Perhaps something in the genetic makeup of the Lazarus Fern, or its micro-site, of some unknown factor - some apparently unique factor/s allowed it to regrow.   The survival of one of the two 2014 plantings is consistent with this hypothesis:  whatever kills the ferns now appears to be gone.    This may be a cause for optimism - or it may be that the killing agent will return and once again attack ferns at ground zero if the ferns are re-established.  

For now, I will take comfort in the possibility that the scourge has moved on, and that at least one native plant is resistant.   We will collect spores from the Lazarus Fern in July and raise up a bunch of baby ferns,  which will also possibly be resistant, for further experimental planting.   If this is a genetic trait, and if it has high penetrance, we may have stumbled upon a partial remedy for the sword fern die-off.

Experimental planting at Seward Park

In February we planted three lines of young sword ferns, twelve ferns in each line.  These ferns were part of our annual generous allotment of three hundred restoration plants from Seattle Parks and the Green Seattle Partnership.  They come from regional native plant nurseries; the plants are about two years old. 

Two of these lines are in the original ground zero at Seward Park, an area that has been barren now for three years, with no natural regeneration of any species.    The third line is about fifty yards north and just outside of the current boundaries of active die-off.   Every fern was "watered-in".  Half of them - every other fern in each line - were infused with mycorrhizal spores.

I hypothesize (based on evidence presented in the next blog post "A Lazarus Fern?") that the likely pathogen responsible for die-off has swept through ground zero and is no longer present.  Thus I predict that the two lines of twelve ferns in ground zero, if we water them through the first summer or two, will mostly survive.  As the die-off zone spreads it will soon encompass the northern line of twelve ferns.  I predict that most of these twelve ferns will die.    More specifically, and as a nod to statistical rigor, I predict a p-value < 0.05 when, after five years, we compare the survival rates at the two contrasting sizes, ground zero (24 ferns) and the northern line (12 ferns, aka AD for "active die-off).  Here this is mocked up in a little R code:

GZ <- as.integer(runif(24) + 0.7)       #   75% survival: 0 1 1 1 1 1 0 0 1 0 1 0 1 1  
                                                           #                           1 1 1 0 1 1 1 1 1 1 
AD <- as.integer(runif(12) + 0.25)    #   25% survival: 1 0 0 0 1 0 0 0 1 0 0 0
t.test(GZ, AD)$p.value                      #   0.0047

I photograph all thirty-six ferns every week and will create a time-lapse video for each fern as these weekly photos accumulate.  The two GZ lines are shown in the picture above.

Thursday, January 11, 2018

Report of Die-off in Snohomish County: Lord Hill Regional Park

Amy Lucas, Senior Park Planner with Snohomish County Parks, reports:

We recently received a citizen report of a similar die-off at our Lord Hill Regional Park. I went in the field yesterday with one of our Senior Rangers and found 3 suspect sites. Two sites were small, and surrounding salal had spots, much like black spot on a rose plant.

We hiked to a much larger site that showed signs of complete sword fern mortality that looked like it was spreading radially. The ground at that site was barren, and the brown fronds were spreading to surrounding plants. 

These photos seem convincing to me, though I have not yet visited the site for a close inspection, and we as yet do not have measurements of extent or rate of spread.  Leaves in the bottom-most photo perhaps obscure dead fern crowns?  It usually takes two years to progress from first symptoms to crown stubble.  

This report, if confirmed, indicates a new direction of spread, to the north and east of Seattle, about 30 miles from Seward Park, and 30 miles west of Baring, a small and apparently isolated residential die-off   - Paul