Vancouver Island COSEWIC assessment and status report: chapter 7

Population Sizes and Trends

Search effort

In 1972 local hunters became interested in Vancouver Island marmots, which until that date were known from only a handful of museum specimen records. Since then much of Vancouver Island has been searched for marmot colonies. Routledge and Merilees (1980) surveyed 97 mountains and classified each in terms of habitat suitability for marmots. The last “new” wild marmot colony was discovered in 1985. The probability of finding new occupied habitat patches is high, particularly given recent releases and use of radio-telemetry. However, the probability of finding large new populations of marmots or large habitat patches is low.

Systematic annual population counts began in 1979 (Munro et al. 1985). Marmots were classified as adults, yearlings or pups (young-of-the-year) based on size and pelage (Bryant 1996). Intensity and extent of surveys varied from year to year. The smallest effort occurred in 1975, when a single colony was visited on one day. The greatest effort occurred in 1997, with 242 visits to 37 colonies. Population counts over 34 years provide data on minimum numbers of adults, yearlings and pups at 49 colonies and 1569 site-year combinations.

Bryant and Janz (1996) estimated the accuracy of these surveys by comparing raw count results with known numbers of marmots at five intensively studied colonies where most individuals were ear-tagged (Bryant 1996). They concluded that count success is highly variable. On days with excellent weather, experienced observers and a known population of marked animals, detection varied from all individuals to none. Smaller colonies are easier to census than larger ones, and success also depends on time of day, with mornings generally being better than afternoons. Marmots become increasingly difficult to detect after August (both because of vegetation growth and changed activity patterns). Reproductive females with pups tend to be quite predicable, while males and two-year-olds make larger daily movements and are less so. Although > 9 visits are necessary to obtain a near-complete count of marmots in a colony, 2 to 4 counts usually detect 65-75% of the animals present (Figure 3).

Based on this analysis, Bryant and Janz (1996) concluded that for sites visited once, observers probably counted 40-60% of adults actually present, depending upon time of year. For most site-year combinations (2 or more visits in June and July), observers probably saw 66-78% of adults, and 75-89% of young. Confidence limits on these estimates are unreliable because of differences in coverage, visibility, observer experience, and count intensity.

Surveys have changed over time, both because of increasing reliance on telemetry and due to declining populations. It is easier to count small numbers of radio-tagged marmots at a given site than to count >30 untagged marmots, or to read ear-tags using a spotting scope (Bryant 1996). Thus, count success has increased over time and population estimates since 2000 likely approach a true census.

Figure 3. Accuracy of marmot counts. Transformed values for daily counts at colonies with known numbers of adults were randomly re-sampled to create 100 trials of 10 counts each (for clarity, results from only 25 trials are shown). The cumulative success curve (bold line) was fitted using linear regression (log-transformed x values, slope = 0.397 and constant = 0.540). On average, 2 counts resulted in detection of 66% of the adults actually present, but 9 counts were required to account for >90% (Bryant and Janz 1996).

Figure 3. Accuracy of marmot counts. Transformed values for daily counts at colonies with known numbers of adults were randomly re-sampled to create 100 trials of 10 counts each (for clarity, results from only 25 trials are shown).

Abundance

Given the variation in count intensity and coverage, raw annual counts are not particularly informative. Early efforts to “correct” these count results relied on calculation of observed/expected ratios and a correction factor based on count effort (Bryant 1997). Recently, locally weighted regression (Cleveland 1979) was used to interpolate missing values, using colony-specific data obtained from previous and subsequent years for the same site (Bryant 2000). This practice avoids the assumption that trends at sampled colonies reflect trends at colonies not visited in a given year. Approximately 29% of the resulting site-specific population values (a marmot-year is a marmot seen alive at one site in one year) needed to be interpolated. Sampling intensity and coverage was particularly good in 1980-1986 (n = 1285 marmot-years, 9% interpolated) and 1995-2006 (n = 565 marmot-years, 4% interpolated). Relatively poor coverage and intensity occurred before 1980 (n = 791 marmot-years, 73% interpolated) and from 1987-1994 (n = 1231 marmot-years, 36% interpolated).

A correction factor based on count intensity was applied to the aggregate actual and interpolated colony-specific values to arrive at probable numbers of marmots (Figure 4). The correction factor varied from 1.19 to 1.66 (mean = 1.40, similar to alpine marmots: 1.25: Cortot et al. 1996). Because young-of-the-year typically emerge in July there is little time for repeated counts and the same approach could not be employed. Instead a constant multiplier (1.20) was used, based on average litter size at five intensively studied colonies divided by the average litter size at other colonies for which numbers of repeated counts were smaller, and at which some pups were likely missed (Bryant 1998).

This approach does not substantively change previous estimates of population sizes or recent dynamics (Bryant and Janz 1996). The population increased from 100-150 during the 1970s to 300-350 during the mid-1980s, in large part due to colonization of at least ten clearcut sites (Bryant 1998). Marmots inhabiting natural habitats declined while numbers in clearcuts were still increasing.

Figure 4. Annual count results and probable numbers of wild marmots, 1972-2007. Data for pups and adults are pooled. The solid line represents estimated annual abundance after colony-specific interpolation and application of a correction factor based on numbers of repeated counts per site-year combination. Note that actual and interpolated estimates merge after 2000 due to increased reliance on telemetry.(A. Bryant unpublished data)

Figure 4. Annual count results and probable numbers of wild marmots, 1972-2007. Data for pups and adults are pooled. The solid line represents estimated annual abundance after colony-specific interpolation and application of a correction factor based on numbers of repeated counts per site-year combination. Note that actual and interpolated estimates merge after 2000 due to increased reliance on telemetry.

The number of wild individuals has been affected by ongoing recovery efforts, as 56 wild-born marmots were removed from 1997 through 2004 to establish a captive population (Bryant 2005). Most animals were captured as pups (n = 31) or yearlings (n = 8), because it was reasoned that younger marmots would more readily habituate to captivity. In addition the intent was to minimize disruption of wild colonies by not removing reproductive adults. Marmots inhabiting clearcut habitats were preferentially targeted (n = 30). Thus management efforts were partially responsible for the population decline from 1997 though 2004. Wild populations were augmented by releasing both captive-born marmots and wild-born marmots transplanted from other colonies in order to provide a mate for solitary individuals (Table 2). As of 2007 at least 4 captive-born marmots and 3 transplanted marmots have produced litters in the wild.

Table 2. Annual numbers of marmots taken into captivity, transplanted, and released. Adapted from Bryant (2007).
Year Male
Into captivityTable notea
Female
Into captivityTable notea
Total
Into captivityTable notea
N captured as pups Male
TransplantedTable noteb
Female
TransplantedTable noteb
Male
ReleasedTable notec
Female
ReleasedTable notec
Total
ReleasedTable notec
N subsequently recaptured
1996     -   3 3     6  
1997 2 4 6 2Females         -  
1998 6 2 8 3Males         -  
1999 9 10 19 6Males, 4Females         -  
2000 3 2 5 2Males, 2Females   2     2 1Female
2001 3 4 7 2Males, 3Females 1       1  
2002 4 2 6 2Males, 2Females         -  
2003 4   4 2Males 2   2 2 6 1Female
2004   1 1 1Female     5 4 9  
2005     -       11 4 15  
2006     -       20 11 31 1Male
Totals 31 25 56 17Males, 14Females 6 5 38 21 70 1Male, 2Females

Fluctuations and trends

Vancouver Island marmot colonies fluctuate in size from year to year. While the overall population trend is generally negative since the mid-1980s, individual colonies exhibited divergent population trajectories (Figure 5). For example, numbers of marmots at the Green Mountain summit colony remained relatively stable until the early 2000s, while larger (Heard 1977, Milko 1984) colonies at Haley Lake and Bell Creek declined dramatically after 1994. The population at the Butler Peak “west roads” clearcut increased from 2 in 1982 to over 30 by 1988 followed by a rapid decline and near-extinction by 2000, when the last survivor was transplanted.

Figure 5. Population trends at six well-studied colonies, 1972-2006. Data represent either the highest daily count of marmots in a year, or the number known to be alive based on mark-recapture or radio-telemetry. No correction factors were applied. Data updated from Bryant and Janz (1996) and Bryant (1998). (A. Bryant unpublished data)

haley lake bowl Bell creek Green Mtn. sommit
Pat Lake clearout Green Road K44A clearout butler West Road clearout

There are a variety of reasons for colony-specific population trends. Birth “pulses” occur because most colonies are small (Bryant and Janz 1996) and adult females do not produce litters every year (Bryant 2005). Fluctuations also occur because of changes in survival and dispersal rates. For example, the Haley Lake colony declined rapidly from 1994 to 1996. Simultaneously the populations at Green Mountain K44A and Sherk Lake clearcuts grew after new immigration occurred. There is no common year after which declines occurred at most colonies. However, colony-specific survival rates were spatially correlated, i.e., colonies in close proximity tended to show similar patterns (Bryant 2000).

The colonization of clearcut habitats during the 1980s led to dramatic changes in local marmot densities. Most colonization events occurred within 1-2 km of previously existing natural colonies (Bryant 1998), and most new colonies became much larger than those in adjacent natural habitats. Local population densities changed, with numbers of marmots per km² increasing to more than 20 in the centre of the range (Figure 6).

By the mid-1980s more than half of the known marmot populations inhabited 4 adjacent mountains, with most animals living in clearcuts on Butler Peak, Haley Lake, Gemini Peak and Green Mountain. Bryant (1998) suggested that high local densities attracted predators. Recent monitoring of radio-tagged cougars and wolves supports this idea, suggesting that some individuals return repeatedly to hunt in the same meadows (D. Doyle pers. com.).

When the marmot population collapsed during the 1990s, often it was areas with the highest density which declined first. The Haley Lake colony (Heard 1977, Bryant 1996) declined from 25 to 10 in 1994-95. The largest colony ever recorded (39 animals in the Butler Peak “west roads” clearcut in 1994) was reduced to 15 individuals in 1995. Conversely, colonies in natural meadows with low densities and without adjacent clearcut colonies were more likely to persist. By the early 2000s the total number of marmots in the Nanaimo Lakes area had fallen to ~30 individuals, with most living in low-density natural colonies on the periphery of their geographic range.

Over the last 16 years (1991-2007) or 3 marmot generations, the estimated total population has declined from about 195 to about 85 (Figure 4), a 56% decline. Because released marmots are excluded from this assessment, however, the 3-generation estimated decline is nearly 80%. Over the last 3 years, numbers have increased (Figure 4). There are currently about 150 marmots in captivity, but they are not included in this assessment.

Figure 6. Changes in marmot density and distribution in theNanaimo Lakes region. The area depicted is 1127 km². Shown are mean densities of adults over two five-year periods, with the lightest to darkest levels of shading corresponding to low (0.1-5 adults/km²), moderate (5.1-10 adults/km²) and high densities (10.1-20 adults/km²). The dots reflect all known breeding or hibernation burrows verified during 1972-2006.(A. Bryant unpublished data)

A) 1982-1986

Figure 6.     Changes in marmot density and distribution in theNanaimoLakesregion. The area depicted is 1127 km². Shown are mean densities of adults over two five-year periods, with the lightest to darkest levels of shading corresponding to low (0.1-5 adults/km²), moderate (5.1-10 adults/km²) and high densities (10.1-20 adults/km²). A) 1982-1986

B) 2002-2006

Figure 6.     Changes in marmot density and distribution in theNanaimoLakesregion. The area depicted is 1127 km². Shown are mean densities of adults over two five-year periods, with the lightest to darkest levels of shading corresponding to low (0.1-5 adults/km²), moderate (5.1-10 adults/km²) and high densities (10.1-20 adults/km²). B) 2002-2006

Rescue effect

There is no possibility of rescue effect as the Vancouver Island marmot does not exist anywhere else.

Page details

Date modified: