Uncovering the Cause of Yellow-Cedar Decline

Early research on cedar decline

Research on yellow-cedar decline outlined below began in the 1980s and was focused on evaluating biotic factors that were suspected as the primary cause of yellow-cedar decline.  The timing of death for yellow-cedar trees was coupled with ground surveys to determine when the problem may have been initiated and to establish patterns of subsequent spread. 

Symptomlogy and Biotic factors

Trees in varying stages of dying were examined for symptoms in their roots, bole, and crown and we developed a sequence of symptoms (Hennon et al., 1990).  Initially, fine roots die, then small diameter roots die, followed by formation of necrotic lesions on coarse roots, and finally necrotic lesions spread from dead roots vertically from the root collar up the side of the bole.  Crown symptoms occur after the early root symptoms.  Crowns generally died as a unit with proximal foliage dying first, then as trees finally died, distal foliage died.  Generally, the study of symptoms suggested a below-ground problem for affected trees.  The following groups of organisms were evaluated as potential pathogens involved in the decline syndrome, but each was ruled out by inoculation studies or by the lack of association with symptomatic tissue or dying areas of the forest.  

  • Higher fungi (Hennon, 1990; Hennon et al.; 1990)
  • Oomycetes (Hansen et al., 1988;  Hamm et al.; 1988)
  • Insects (Shaw et al., 1985)
  • Nematodes (Hennon et al., 1986)
  • Viruses and mycoplasmas (Hennon and McWilliams, 1999)
  • Bears (Hennon et al., 1990)

The general conclusion from these evaluations of symptoms and possible biotic factors was that no organism was the primary cause of the decline problem (Hennon et al., 1990; Hennon and Shaw, 1997).

Formation of the freezing Injury Theory

dying cedar tree

Abiotic risk factors

Several years ago, we evaluated the leading abiotic factors potentially associated with yellow-cedar decline (D’Amore and Hennon, 2006) and provide a summary here.  Soil saturation was associated with dead trees, particularly in the central patches of decline.  The soil saturation-dead tree relationship was not consistent; however, as some areas of decline occurred on well drained soils, and saturated bogs sites at a higher elevation had little cedar tree death.  Soil chemistry was examined, with a focus on aluminum, calcium, and acidity and their correlation with tree death.  Aluminum toxicity (Lawrence et al., 1995; Lawrence et al., 1997) or high acidity (Klinger, 1990) could potentially reach levels that would damage roots as soils become saturated, but no relationship of either factor with dead trees was found.  Depleted soil calcium is known to be involved in the cold tolerance of forest trees (Schaberg et al., 2001; Schaberg et al., 2002).  Rather than occurring at low levels indicating susceptibility to freezing, calcium was found in high concentrations where yellow-cedar trees had died, and is attributed to enrichment of soils from senescence and decomposition of cedar tissues.  Soil and air temperatures emerged as the abiotic factors most highly correlated with the death in the cedar forests (see below).  Thus, this study and the collective research that preceded it were the foundation for a new hypothesis that we elevated to explain yellow-cedar decline (Fig 5).  Each interaction in this complex hypothesis is explained below.   

hypothetic scenario for cedar decline - schematic

Our explanation for yellow-cedar decline involves cascading factors leading to freexing injury. The protective role of snow is illustrated.

Landscape position, soil properties create wet soils

The association of yellow-cedar decline with bog plant communities can be seen from aircraft or on aerial photography. Dead trees frequently occur around the edges of bogs, or on hillsides with mosaics of bogs and forests supported by moderate drainage (Hennon et al., 1990). The yellow-cedar mortality problem is not known to occur where yellow-cedar grows with western hemlock in productive forests that are not in proximity to bogs.

Johnson and Wilcock (1998), the first to measure hydrology directly, investigated three locations in southeast Alaska having steep slopes (25 to 39º) with and without decline.  Although dying portions of forests had smaller hydrologic contributing areas, they remained saturated longer than surrounding healthy forests (Johnson and Wilcock, 2002).  Overall, peak saturation in areas of cedar decline was not found to be greater than soil saturation found in healthy cedar or hemlock/spruce forests (Johnson and Wilcock, 1998; Johnson and Wilcock, 2002).  Sampling in two watersheds, D’Amore and Hennon (2006) reported higher water table levels in some portions of declining forests, but not where recent mortality had occurred near the perimeter of dead and dying trees.  These studies suggest the need for a better understanding of the links among hydrology, topography, and vegetation.  Greater insights among these parameters could be gained through the development and testing of a hydrologic model.   

Wet soils govern canopy cover

Soil drainage drives the stature and productivity of forests in southeast Alaska, which in turn controls canopy cover.  Saturated soils lead to scattered stunted trees that produce almost no cover (e.g., on bogs).  Regardless of tree death, overall tree productivity was highly correlated with mean water table values measured in wells at our study sites (D’Amore and Hennon 2006).  Currently, we are using LiDAR-derived digital elevation models (resolution is sub-meter) of the ground surfaces to map drainage patterns relative to tree death and canopy cover.  Two hydrologic models will be developed with detailed LiDAR terrain maps to determine possible associations between tree death and ground saturation.  One model will determine ground saturation solely on the basis of upslope contributing area, hillslope gradient, and forest characteristics.  The other hydrologic model will incorporate hillslope features found immediately downslope.  Development of these models will help elucidate parameters most influential in the decline of yellow-cedar.

Open canopy conditions influences exposure

We are using two methods to evaluate canopy cover and exposure at two watersheds:  hemispherical (i.e., “fisheye) photographs and LiDAR canopy interception.  Results on canopy cover from these bottom-up and top-down views are highly correlated.  LiDAR measures have the advantage of modeling canopy cover across the entire watershed, rather than rely on separate locations where hemispherical photographs are taken. Our tree plot information is providing values of both live and dead trees to compare to these two measures of canopy cover.  Canopy cover is largely driven by measures of live trees (e.g., basal area values driven by hydrology), but dead trees also make a smaller, but significant contribution.  These evaluations suggest that exposure (i.e., open canopy conditions) was controlled by hydrology through the suppression of forest productivity by wet soils in the initial phases of yellow-cedar decline.  These wet areas are occupied by the oldest snags that died decades ago (D’Amore and Hennon, 2006).  The exposed, open canopy condition has developed more recently at the perimeters of decline patches on soils with better drainage, through a feedback with tree death itself. 

aerial oblique view of dead cedar at Poison Cove

(A) Low Elevation Bog and Scrub

open canopy hemispherical photo
  1. Open Canopy, wet soil Conditions
  2. Roots are Shallow
  3. Soils and roots warm due to lack of insulating snow
  4. Cedar deharden prematurely before last extream freezing events
  5. The lack of snow causes the roots to freeze

(B) High Volume Dead

high volume dead canopy hemispherical photo
  1. Adjacent mortality creates canopy openings increasing exposure
  2. Roots are somewhat shallow
  3. Soils and roots warm considerably due to lack of insulating snow and canopy
  4. Cedar deharden prematurely before last extream freezing events
  5. The lack of snow and canopy cover cause the roots to freeze

(C) High Elevation Bog and Scrub

open canopy hemispherical photo
  1. Open Canopy, wet soil Conditions
  2. Roots are Shallow
  3. Soils and roots stay cool and insulated due to persistant snow
  4. Cedar dehardening is delayed past extream freezing events
  5. Persistance of Snow is the Key factor

(D) High Volume Live

closed live canopy hemispherical photo
  1. Closed canopy and shaded forest floor
  2. Roots are deeper
  3. Soils and roots stay cool and insulated due to canopy and shade
  4. Cedar dehardening is delayed past freezing event
  5. Canopy cover, rather than snow is the Key factor

Exposure affects dehardening and freezing temperatures

In regions of yellow-cedar decline, soil and air temperature near the ground are primarily controlled by canopy cover (D’Amore and Hennon, 2006).  Sites with less canopy cover have greater daily maxima, lower daily minima, and greater daily ranges of air temperature.  The accumulation of soil temperature, expressed as soil degree days, is greater in areas of less canopy cover.  These influences of canopy cover on air and soil temperature fluctuations are most pronounced in spring. We are continuing to associate air and soil temperature with canopy cover values derived from hemispherical photography and LiDAR canopy modeling (described above).    

Temperature appears to have a particularly strong influence on the fall hardening and spring dehardening processes for yellow-cedar (Puttonen and Arnott, 1994; Hawkins et al., 2001). Thus, we propose that yellow-cedar trees growing in microsites with less canopy cover, controlled by either hydrology or mortality feedback, are triggered to deharden prematurely by warming in late winter and early spring

Shallow roots, dehardening, and freezing cause tree death

We evaluated the seasonal cold tolerance of yellow-cedar and co-existing western hemlock in open- and closed-canopy forests and at several elevations at one of our study sites (Schaberg et al., 2005).  In fall, yellow-cedar in open canopy settings were more cold tolerant than cedar in closed-canopy, whereas western hemlock appeared unresponsive to canopy conditions.  In winter, yellow-cedar had cold tolerance to about -40°C, colder than any recorded temperature for the region.  Susceptibility of yellow-cedar to cold temperatures develops in later winter and spring.  In our testing of tree tissues (Schaberg et al., 2005), yellow-cedar dehardened almost 13°C more than western hemlock between winter and spring, so that yellow-cedar trees were more vulnerable to freezing injury in spring than western hemlock.  Also, trees that we tested growing above 130m elevation were more cold hardy than those growing below 130m.       

seedlings in pots some showing foliar symptoms

Blocks of seedlings on the left and middle were protected by perlite as a mimic for snow. These seedlings remained healthy as soil temperatures never reached the -5C (23F) lethal temperture for yellow-cedar fine roots. By contrast seedlings on the right were unprotected, their soil dropped below this temperature theshold, and all roots were killed. See the publication Schaberg et al. 2008 for more information.

We have observed severe freezing injury to yellow-cedar seedlings growing in Juneau across several years, each time foliar symptoms appear at the end of March or early April.  Based on these observations and our cold tolerance testing of mature trees, we initiated a study on seedlings to more intensively evaluate the spring dehardening and cold tolerance of root and foliage tissue in late winter and early spring.  This cooperature project with scientists from Vermont created a miniature version of yellow-cedar decline with seedlings (Schaberg et al. 2008). We found that the critical theshold for yellow-cedar fine roots is -5C (23F), below which roots are killed. Perlite was used to mimic the insulating qualities of snow, resulting in full protection of fine roots and the seedlings because these lethal soil temperatures were not reached as they were in the soils of unprotected seedlings (no perlite).

Protection of snow

Snow appears to protect yellow-cedar from this presumed freezing injury.  Our measurements of snowpack at the Poison Cove study site indicates that yellow-cedar growing around an open-canopy bog at 150m, a setting without the decline problem, has snowpack through April or May during some years.  Snow appears to offer some form of protection for yellow-cedar, perhaps by 1) delaying the dehardening process, and/or 2) protecting fine shallow roots from freezing.  In either case, the presence of snow through March and April apparently allows yellow-cedar to pass a period of potential vulnerability that affects trees growing without snow.  At our mid-scale analysis, the lack of spring snow may explain why yellow-cedar decline is limited to lower elevations and why it reaches higher elevations on warm aspects compared to cold aspects.  At the broad scale, the distribution of yellow-cedar decline aligns very closely with the lowest snow zone.  Some change in the environment must have initiated yellow-cedar decline. It is possible that reduced late winter and spring snowpack as the climate emerged from the Little Ice Age represents that environmental trigger.   

A case study on Mount Edgecumbe, modeling snow from the past and into the future can be seen in an animation on the detection page. A poster we created further explores the relationship between terrain, snow accumulation and the presence of cedar decline.