Temporal Variations in Seed Crops and Bear Use


Abundance of whitebark pine in the northern Rocky Mountains has varied going back to the Pleistocene, but perhaps at no time so dramatically as during the last 4-5 decades, since 1970. Subsidiary to abundance of the trees themselves, seed crops have also varied on a per tree basis, but typically at the scale of years rather than decades or centuries. When put together, changes in tree abundance coupled with inter-annual variation in per tree cone crops has resulted in enormous changes in landscape-level availability of seeds to grizzly bears, with implications for diets and use of habitats and landscapes by bears.

The graphs at right show some of the ways that Yellowstone grizzly bears have responded to differences in availability of pine seeds at an aggregate population level. The top-most graph (A) shows how observed levels of pine seed feeding activity, as a proportion of total aggregated over years, varied with the percent of a home range comprised of whitebark pine forests. Increases in use tended to accelerate as availability of mature trees increased. The Cooke City environs are emblematic of dynamics that organize around abundant whitebark pine; notably, heavy use of whitebark pine seeds by bears in this area has persisted from at least the 1980s up to the present despite major losses of whitebark pine elsewhere in the ecosystem.

Graphs B and C show how annual exploitation of pine seeds by bears varied with annual sizes of cone crops. Feeding activity (B) increased at a decelerating rate as cone crops increased, emblematic of a Type II functional response. Remains of pine seeds in feces (C) increased in a sigmoidal manner, emblematic of a Type III functional response. In both cases, though, use approached a plateau (i.e., asymptote), reflective of a potential saturation of demand by bears at very large crop sizes. This shift seemed to occur when the numbers of cone counted on individual trees at fixed transects exceeded 20.

Inter-Annual Variation in Crops & Bear Feeding

Data collected in the Yellowstone ecosystem between 1978 and 1990 provide fascinating insights into temporal and spatial patterns of cone production on whitebark pine trees and related reflective inter-annual variation in the levels and locations of bear foraging on pine seeds.

The figure at left immediately above shows how the elevational distribution of feeding on pine seeds by grizzly bears varied among four successive large crops, set against the elevational distribution of whitebark pine trees themselves. Each diagram shows elevational patterns for exploitation of a given crop by bears, typically spanning two years. Elevations range from highest, top, to lowest, bottom. Bear activity is shown by the black lines with green dots. The elevational distribution of mature whitebark pine trees, which is fixed, is shown in light green, including a range of variation denoted by darker green. The elevational distribution of all whitebark pine, including seedlings and saplings, is shown in gray. Interestingly, seedlings and saplings are consistently more abundant at lower elevations compared to mature cone-producing trees, which speaks to a potential to grow at lower-elevations precluded by competition from more thrifty and warmth-tolerant trees such as lodgepole pine and Douglas-fir.

The basic takeaway from the figure at left above is that the elevational distribution of bear feeding--and cone production--varied from one large crop to the next. Cones and related bear feeding were concentrated at lower elevations during 1978-1979, but even moreso during 1987-1988, at mid-elevations during 1985-1986, and in a way that most closely mirrored the full distribution of whitebark pine during 1989-1990. More to the point, the elevational patterns were non-repeating.

This same point is born home by the diagrams in A of the figure at top right. These diagrams show the joint elevational and aspectual distribution of bear feeding for three of the same crop years as shown in the figure at top left. During 1985-1986, feeding was concentrated at mid-elevations on east aspects. During 1987-1988, concentrations were at mid- to lower-elevations on north and west aspects. And during 1989-1990, feeding was concentrated at high elevations on southeast aspects. The point here being that abundant cones were produced on different aspects and at different elevations during successive large crops. As a consequence, grizzly bears ended up exploiting the full elevational and aspectual distirbution of whitebark pine when aggregated over multiple years. In other words, no part of the whitebark pine zone was dispensable.

Parenthetically, this non-repetitive pattern of cone production reflects the exigencies of reproduction by individual trees. Intervals between large cone crops on any given tree vary between 4 and 9 years, which is reflective of the long time it takes a high-elevation tree such as whitebark pine to replenish carbohydrate reserves after a bumper crop. Given that large crops at a landscape level occur at 2-4 year intervals, cone production would thus have to shift among trees occupying different site types. Generally speaking, though, poor crops typically follow good crops, modified by a tendency for crops to be larger following a series of warmer growing seasons, which predictably accelerate replenishment of carbohydrate reserves in an otherwise cold environment.

The graphic immediately above highlights an important but often overlooked phenomenon: seasonal variation in consumption of pine seeds by grizzly bears, with the potential for that feeding to continue during the spring and summer of the year following when cones were produced. As I've noted before, almost all of the seeds eaten by Yellowstone grizzlies are obtained from cones excavated from red squirrel middens. By design, middens are an excellent medium for preserving over-wintered cones and seeds, and, because of that, a critical contingency of bear feeding.

More specifically, consumption of seeds that were produced in disproportional abundance at the highest elevations during 1989 were consumed latest of the three featured crops, and over-wintered in the greatest amounts to be exploited the following year. By contrast, the cones disproportionately produced at low elevations during 1987 were exploited by bears early, and the seeds therein mostly consumed by the end of that same year. Few cones over-wintered to be eaten to following year.

Parenthetically, the record-breaking consumption of pine seeds by Yellowstone grizzly bears during 1979, which is not shown above, consisted almost entirely of seeds obtained from cones over-wintered from the prodigious crop produced during 1978.

Catastrophic Losses During the Last 45 Years

Whitebark pine has been under assault during the last 40-50 years primarily from forces unleashed by humans. The result has been catastrophic declines severe enough to lead the US Fish & Wildlife Service to conclude that whitebark pine warranted being listed as Threatened under the US Endangered Species Act--although listing was precluded by the lack of Service resources.

White Pine Blister Rust:

The Northern Continental Divide

The first agent of destruction to take the stage was a non-native fungal pathogen called white pine blister rust (Cronartium ribicola). This disease affects all five-needled pines, and had spread east and south from an original introduction in coastal British Columbia around 1900. All North American white pines are highly vulnerable to blister rust, but whitebark pine is the most vulnerable of all. Roughly 85-99% of all infected trees die. The life history of blister rust is complex, but infection is clearly signaled at an advanced stage by the eruption of cankers, as per what you see at right. 

Blister rust had started to take a major toll on whitebark pine by the late 1960s and early 1970s in northwestern Montana. One of the first comprehensive surveys of blister rust infection and resulting mortality among whitebark pines was undertaken by Bob Keane in the early 1990s, bench-marked at the year 1991. The map immediately above summarizes the results of his survey. At that time, most whitebark pine west of the Continental Divide had already been killed by blister rust. Mortality dropped off to low levels the farther south and east he went. But a recent survey undertaken during 2004-2006 by Carl Fiedler and Shawn McKinney found that roughly 75% of all whitebark pine were dead and 90% of the remainder were infected and fated to die areas that Keane had surveyed 25 years before.


Whitebark pine had been functionally extirpated in this region by blister rust, certainly as a bear food--in only 45 years. Although pine seeds had once been a major food of bears along the North Fork of the Flathead and throughout the main range of the Rocky Mountains, they are no longer. A once important food is essentially gone from the Northern Continental Divide Ecosystem. Probably not by coincidence, the concluding demise of whitebark pine along the East Front coincided with a marked increased in fires as well as increased dispersal of grizzly bears east into the Great Plains (see Conservation).

Wildfire and Mountain Pine Beetles: Yellowstone

Blister rust was detected in the Yellowstone ecosystem as early as the 1940s, but exhibited a very slow rate of spread that was attributed to drier colder conditions in this region. Most people were relatively blithe about prospects for whitebark pine. But, as it turned out, the major threat to emerge in this region was not blister rust, but rather an endemic insect called mountain pine beetle (Dendroctonus ponderosae; a bark bettle), unleashed by climate warming, exacerbated by some perverse effects of wildfire (see at right).

Prior to the early 2000s, bark beetles had been largely excluded from the high elevations occupied by whitebark pine largely by bitter cold winter temperatures. Mountain pine beetles largely subsisted on lodgepole pine at elevations 500-3000 feet lower. Although there had been limited outbreaks of beetles among whitebark pine during earlier warm-dry periods (e.g., the 1950s),there was little indication that the effects had been devastating.


That all changed around 2000. Unprecedented warmth coupled with drought unleashed a massive outbreak of mountain pine beetles among whitebark pine in the Yellowstone ecosystem and, as with blister rust, whitebark pine turned out to have little intrinsic resistance. By the late-2000s, many of the mature whitebarks in this region were dead. A comprehensive survey undertaken by Wally Macfarlane and Jesse Logan in 2009 documented this catastrophe. The map in C, above, shows the results. Pretty much all of the overstory whitebark pine were dead in areas shaded gray, headed there in areas shaded red, and exhibiting high levels of mortality everywhere shaded orange. "Severe" mortality was evident in 46% of the region. The 5% of the area exhibiting only "trace" mortality, shaded green, was almost wholly restricted to the Beartooth Plateau, Wind River Range, and Grand Tetons. Beetle-caused mortality has continued, resulting in the near-extirpation of whitebark pine seeds as a bear food in most of the Yellowstone ecosystem--in a mere 15 years.


But the decline of whitebark pine as a source of bear food was nuanced. The graphs at left above in A and B provide the essential details. Figure B shows annual production of cones on trees at transects monitored each year since 1980. Despite the fact that per tree cone production is quite variable from one year to the next, there is a curious trend. Multiyear average cone production was roughly two-times higher during 2006-2014 compared to earlier periods, coincident with the most dramatic and culminating declines of mature trees from bark beetle-caused mortality. This decline is denoted by the dark red solid line in A. Putting per tree cone production together with relative numbers of surviving mature trees gives an index of landscape level cone and seed availability each year, which is shown by the dotted line in A. As you can see, the uptick in per tree cone production mitigated for loss of trees to beetles between 2004 and 2009. But, after that, losses to beetles swamped increases in per tree cone production to yield a terminal decline.


So, why the increase in cone production on surviving trees? As I noted above, all else equal, cone production seems to be favored by warmer growing seasons and resulting accelerated accumulation of carbohydrate reserves among mature whitebark pine. It is possible that, at the same time that warmth was driving a massive outbreak of pine beetles, it was also fostering large cone crops on surviving trees, which temporarily masked what is almost certainly the beginning of a terminal loss of this bear food. 

Common knowledge would have us believe that wildfire is almost invariably good for whitebark pine, at least when reckoned over decades. Whitebark pine is relatively intolerant of shade and, barring its very highest-elevation haunts, is progressively replaced over several centuries by more shade-tolerant subalpine fir (Abies lasiocarpa) and Engelmann spruce (Picea engelmannii). Wildfire kills these competitors and prepares the site for emergence of whitebark pine seedlings from seed caches made by Clark's nutcrackers (Nucifraga columbiana).

But this simplistic rendering is nullified by climate change. Moreover, from the perspective of bears, a wildfire eliminates whitebark pine as a source of seeds for up to a century afterwards.


Relevant to this last point, extensive historically unprecedented wildfires burned orughly 1,688,000 acres in the Yellowstone ecosystem during 1988. And areas supporting mature whitebark pines were not exempt. When the fires burned out, roughly 17% of the forests supporting significant amounts of cone-producing whitebark pine had been burned. The map in C, above, shows the fires in red relative to the distribution of whitebark pine, denoted by various colors explained at left.

This loss of whitebark pine negatively affected Yellowstone's grizzly bears. The bars in D, above at right, provide a before and after summary of ecosystem-wide use of whitebark pine seeds by grizzlies. Early-season consumption (before July 15th) declined by over two-thirds. Late-season use declined by roughly 25%. These declines were mostly the result of losing mature whitebark pine, but also partly a result of smaller average squirrel middens after the fires caused squirrels to squeeze into smaller territories. I explain the importance of midden size to grizzlies in the section on Spatial patterns.

Concluding Details about the Mountain Pine Beetle Outbreak

Some apologists for the status quo have argued that the outbreak of mountain pine beetles in whitebark pine during 2000-2010 was not unprecedented, and merely an interruption of some sort of long-term "normal." A corpus of research puts the lie to such claims, most definitively including a recent paper by Polly Buotte and co-authors. She presents a detailed reconstruction of the outbreak, along with a powerful model that predicts past and projects future climate suitability for beetles. Her reconstruction of climate suitabiliity goes back to 1950 and is shown in light tan in the graph above left for the period 1985-2010. This graph also shows the observed extent of beetle-caused mortality of whitebark pine as red dots and predicted extent as a black line plus bounding variability in dark brown. The sustained period of conducive climatic conditions that occurred between 2000 and 2009 was unlike anything that had occurred before going back to 1950. And, as I elaborate under Future prospects, beetle outbreaks are going to become more common, not less, among whitebark pine as the climate continues to warm.