Temporal Patterns

of Fruit-Production & Consumption

It is a fact of nature that abundance of foods vary with time, whether annually, seasonally or daily. This variation is the driver of what ecologists euphemistically call the "bottom-up effects" on behaviors and birth and death rates of animals that consume the involved foods. Alternatively, "top-down effects" are typically attributable to predation. And, in fact, abundance of fruits in the northern US Rocky Mountains varies with predictable rigor seasonally, and with vagarious unpredictability annually. As I describe under Demography, this variation has had consequences for growth of Northern Continental Divide grizzly bear populations, but here I confine myself to describing the nature and drivers of temporal variations.  

Annual Variation in Fruit Production

Of the all the researchers who have studied grizzly bears in the northern US Rockies and adjacent Canada, Wayne Kasworm and Bruce McLellan have been the most dedicated in their collection of data on annual variation in fruit availability and, of the two, Wayne has been the most rigorous. Wayne's efforts have focused on huckleberry (Vaccinium), buffaloberry (Shepherdia), Serviceberry (Amelanchier), and, more recently, mountain ash (Sorbus sp.), all in the Cabinet-Yaak ecosystem of far northwestern Montana and adjacent Idaho. Bruce focused solely on monitoring huckleberry along the North Fork of the Flathead River, but did so for an exceptionally long period of time. Thank you to both of them. Parenthetically, Kate Kendall started some promising work on monitoring fruit production in Glacier National Park during the early 1980s, but then dropped it to pursue a traditional focus of counting bears.

The figures at left summarize data on annual abundance of huckleberry (top), buffaloberry (middle), and serviceberry (bottom) from the Cabinet-Yaak and North Fork. The dark gray dots at top are estimates generated by Bruce McLellan, and the light gray dots estimates by Wayne for his study area. The white and gray dots for buffaloberry are the result of applying two different methods, one of which was dropped (represented by the white dots). I've also fitted 3-year-running averages to each time series as a means of smoothing the considerable annual variation. These trend lines are represented by the solid red(dish) lines.

The main result to note from these trend data is what I call a "berry famine" during roughly 2000-2008 (highlighted in yellow), consistent across all three species, and consistent for huckleberry between the North Fork and Cabinet-Yaak. As I elaborate under Demography, this famine had dramatic consequences for grizzlies in this part of the world. 

Another axiom of ecology is that any given event is caused by something, often a complex constellation of factors. In this case, the question is what caused the famine? A graduate student named Zac Holden provided a tentative answer for huckleberry and serviceberry using Wayne's monitoring data, logically focusing on annual variation in weather. He found that both huckleberry and serviceberry production were highest following cool springs (April-June), as was huckleberry production during following summers with wide-amplitude day-night temperatures during July. The physiological mechanisms behind these correlations are somewhat mysterious, but the modeled relationships were predictive.

In the graph at left I've plotted July diel temperature range at top and April-June temperature at bottom for northwestern Montana going back to 1959. As with the graphs above, I've also fitted a 3-year-running average to these data. Series of years with high-amplitude July temperatures are highlighted in yellow, as are series with warm springs below.

The "berry famine" seemed to have been kicked off by a preceding series of comparatively warm springs coupled with years of little day-night temperature variation during July which lasted into the middle of the famine. However, even with amelioration of these weather-related drivers, it appears as if recovery of fruit crops was not instantaneous, but rather lagged by several years. Put together, the seminal lesson seems to be that there are, indeed, probable lags of several years in fruit production in response to climatic drivers, which perhaps compounds the lags in response of bear populations to availability of foods such as huckleberry (see this entry on NCDE Demography).

Seasonal Variation in Consumption

Not surprisingly, consumption of fruit by grizzly bears reflects fruit availability, both annually and seasonally. Unfortunately, evidence for annual variation in consumption comes almost wholly from anecdote rather than from data, which is a predictable consequence of logistical difficulties in collecting enough field evidence to support annual estimates of consumption. By contrast, there is ample direct evidence of more predictable seasonal variation in fruit consumption by bears, some of which I presented in the introduction to this section on fruit, albeit without differentiating which species were being consumed when. Consumption invariably peaks after fruits produced during the current year have reached full ripeness, which typically occurs around mid-summer.

In the figures at right I provide a little more detail on this seasonal progression for two study area: the North Fork, emblematic of west-side environments, and the Alberta foothills, emblematic of east-side environments. Notably, thanks to Bruce McLellan (top; again) and Robin Munro (bottom), the progression provides details on which types of fruit were being eaten, when, with blue denoting huckleberry, dark red, buffaloberry, and lighter red, kinnikinnick (Arctostaphylos uva-ursi).

In addition to the pattern illustrated in the Introduction of greater consumption of buffaloberry in east-side continental climates compared to the more maritime west-side climates, these figures also illustrate the more prolonged period of fruit consumption in more productive west-side environments (roughly 4 weeks compared to 2) along with an onset roughly 3-weeks earlier. Consumption of lower-quality kinnikinnick berries is also largely confined to early or late in the year after fall frosts or over-winter weather have produced higher sugar content.  

Without intending to belabor the obvious, seasonal patterns of fruit consumption reflect the culmination of fruit properties that enhance nutritional benefits for bears, including size, sugar content, and digestibility, all of which tend to be temporally correlated. The figures at left illustrate such patterns for two higher-quality fruits: serviceberry (top) and huckleberry (bottom).

The top graph shows size of serviceberry fruits as a function of "growing-degree-days," which is a rough proxy for progression of the summer season. Increase in size is clearly not a linear function of cumulative warmth. Rather, size skyrockets once the season has accumulated 600-700 growing degree days. In part, this reflects the comparative affinity of serviceberry for warmth, but nonetheless with broader implications for all fruit-producing species.

The figure at bottom shows sugar concentration for huckleberries based on data collected by Kate Kendall in Glacier National Park. The basic point is obvious: sugar content peaks at peak ripeness, coincident with peak consumption by grizzly bears (see immediately above), but then drops as berries weather. By implication, declining consumption of huckleberries during mid-September is probably as much a function of declining nutritional benefit from consumed berries as it is from diminished abundance of fruits resulting from widespread consumption by frugivores. 

Seasonal Variation in Fruit Quality