Multi-species recovery strategy for Garry Oaks ecosystems: chapter 2

• Previous page
• Table of contents
• Next page

2. Recovery

• 2.1 Common limitations and threats
  • 2.1.1 Habitat Conversion
  • 2.1.2 Environmental stochasticity and demographic failure
  • 2.1.3 Alteration of hydrologic regimes
  • 2.1.4 Recreational activities
  • 2.1.5 Secondary succession due to fire suppression
  • 2.1.6 Eutrophication
  • 2.1.7 Marine Pollution
• 2.2 Recovery feasibility
  • 2.2.1 Recovery feasibility of bog bird's-foot trefoil
  • 2.2.2 Recovery feasibility of tall woolly-heads (Pacific population)
  • 2.2.3 Recovery feasibility of Kellogg's rush
  • 2.2.4 Recovery feasibility of water plantain buttercup
  • 2.2.5 Recovery feasibility of rosy owl clover
  • 2.2.6 Recovery feasibility of dwarf sandwort
• 2.3 Recovery goals and objectives for species at risk
• 2.4 General description of research and management activities needed to meet the objectives
• 2.5 Critical habitat
  • 2.5.1 Occupied habitat
  • 2.5.2 Potential habitat
• 2.6 Examples of activities likely to result in the destruction of any critical habitat identified in the future
• 2.7 Existing and recommended approaches to habitat protection
• 2.8 Schedule of studies to identify critical habitat
• 2.9 Anticipated impacts on non-target species
• 2.10 Social and economic considerations
• 2.11 Knowledge gaps
• 2.12 Evaluation and means of success
• 2.13 Examples of actions completed or underway
  • 2.13.1 Other recovery strategies
  • 2.13.2 Public outreach and habitat protection
  • 2.13.3 Habitat mapping, surveys, and inventory
  • 2.13.4 Management/stewardship plans
  • 2.13.5 Habitat Restoration Projects
• 2.14 Suggested timeline for completion of recovery action plan (RAP)

2.1 Common limitations and threats

Numerous factors have been identified as posing threats to plants at risk in vernal pools and other ephemeral wetlands on southeastern Vancouver Island (Table 3). In only some cases is there strong evidence linking a specific threat with recent population declines. The most obvious example is the loss of habitat and populations to urban development. However, at Harewood Plains, direct damage to bog bird's-foot trefoil plants from unauthorized all-terrain vehicle traffic has been well documented (Donovan 2004). At Somenos Marsh, a creek dredging operation has recently resulted in the destruction of a substantial segment of the tall woolly-heads population.

It is important to recognize the inherent complexity of vernal pool and related ecosystems and that, in most cases, population declines are likely a result of the cumulative impacts of many interacting anthropogenic stresses. Aside from direct habitat conversion, few of the identified threat factors have been diagnosed and tested empirically as driving declines for the individual species addressed in this strategy. However, many of these factors are cited in the literature as being key impacts for ephemeral wetlands in general (e.g., Witham et al. 1998).

2.1.1 Habitat conversion

Widespread loss of natural wetland habitat has occurred on southeastern Vancouver Island, with the greatest habitat losses occurring in the past 150 years. Conversion to agricultural and pasture lands, and urban and residential development, have been the main factors responsible for the declining extent of wetlands (McPhee et al. 2000). Habitat conversion has been implicated in the loss of populations of bog bird's-foot trefoil, tall woolly-heads, rosy-owl clover and possibly water plantain-buttercup during the past century (Illingworth and Douglas 1998, Douglas et al. 2001a, Donovan 2004, Fairbarns 2004).

2.1.2 Environmental stochasticity and demographic failure

Most of the species addressed in this Recovery Strategy are represented in Canada by just a few individuals inhabiting a very small (<300 m² ) spatial area (Table 2). These small populations may be highly vulnerable to stochastic environmental events and processes (e.g., drought, disease, disturbance from waterfowl, etc.). Even under stable environmental conditions, random variation in demographic parameters (a phenomenon referred to as “demographic stochasticity”) could propel these populations through fluctuations that include zero individuals, resulting in local extirpation (Menges 1998).

Some taxa may also be experiencing deleterious genetic effects such as inbreeding depression (i.e, a loss of vigour or fitness due to inbreeding), which is another hazard sometimes associated with small isolated populations (Huenneke 1991). Although genetic constraints are likely to be secondary to the demographic difficulties of small population size in a varying environment, inbreeding depression could theoretically contribute to decreased population growth rate and increased extinction probability (Schemske et al. 1994).

2.1.3 Alteration of hydrologic regimes

Both the timing and duration of the winter inundation phase are believed to have a marked effect on the floristic composition of vernal pools (Keeley and Zedler 1998). For example, some species initiate germination during the wetting phase (Zedler 1987; Bauder 1987a) whereas others require inundation (Keeley 1988). It is not uncommon that, as a result of annual variation in weather patterns, different species will dominate in different years within the same pool. Likewise, dryland species that normally would be excluded from pool basins due to their inability to tolerate inundation are sometimes able to colonize basins in low rainfall years (Zedler 1984, 1987). In California, pools deprived of sufficient moisture tend to be dominated by nonnative dry land plants, most notably grassland annuals (Bauder 1987b).

The timing of pool dry-down (e.g., early vs. late spring) may be an important factor constraining the ability of some pool species to persist locally, since rapid onset of drought conditions in shallow or otherwise drought-prone pools can result in truncated growing seasons and reduced seed production (M. Fairbarns, pers. comm. 2005). In such instances, the surrounding micro-catchment can play a crucial role in extending the growing/flowering season, by delivering late rains to the vernal pool, swale, or seep.

The timing of pool dry-down may also contribute to the exclusion of more typical wetland plants, as dry-down occurs just when air temperatures are rising rapidly. The transition period between inundation and drought appears to be too brief for establishment of many emergent wetland species (Keeley and Zedler 1998). Basin depth is likewise an important factor driving community composition, as deeper basins tend to retain water longer and hence favour different assemblages (Zedler 1987). Vernal pool plant species can tolerate lengthy periods of inundation, but they are not truly aquatic, so mortality increases with length of inundation and is often 100 percent when ponding exceeds six months (Bauder 1987b, 1992).

Consequently, hydrologic changes and erosion can cause significant changes in the pool flora (Bauder 1992). Irrigation common in urban and landscaped areas may encourage less stress-tolerant species, including exotic plant species. Trenching for public utilities, maintenance of fire access roads, fuel and chemical spills, and recreational activities can all cause damage to vernal pools and other ephemeral wetlands, particularly during the aquatic or drying phases when soils are most vulnerable and the organisms are growing or reproducing (U.S. Fish and Wildlife Service 1998a). Furthermore, although watershed contributions to pool filling may be minor relative to those from direct precipitation, water exchange between the vernal pool or seep and surrounding upland can play a major role in controlling water level relationships (Hanes and Stromberg 1998), thus maintaining overall landscape context may be vital to the long-term health of pools and seeps.

Ironically, ruderal species such as tall woolly-heads and bog bird's-foot trefoil have now begun to colonize some of the older tire ruts at Uplands Park and Harewood Plains. It remains to be seen whether these artificial “pools” are capable of sustaining colonies of endangered plants over the long term, or whether they will eventually come to act as “sink” habitats for species at risk. Regardless, considerable effort will now be required to recover the landscape matrix and restore the natural flow of water through these areas.

For each species, threats in each of the two categories are ranked as Nil, Low, Moderate (Mod.), High, or ambiguous (?).

The Somenos Marsh site near Duncan differs from most other habitats described in this strategy in that it lies adjacent to, and is sustained by, a permanent wetland. Best be described as a “seasonally wetted wetland margin” as opposed to a vernal swale or seep, this site hosts the largest confirmed tall woolly-heads population in Canada (Table 2). The basins draining into Somenos Marsh together occupy over 7000 ha. This watershed has undergone extensive land development and clearing, including dyking, forest harvesting, road construction, residential and commercial development, and agriculture (Williams and Radcliffe 2001). Over the past 20 years, in-filling of the lake and eutrophication have led to gradually increasing summer lake levels, raising concerns over impacts on agricultural production (mainly hay) as well as on fish and waterfowl habitat (Williams and Radcliffe 2001). As a way of enhancing wildlife values while accommodating agricultural objectives along the lake shore, a proposal has been made to reduce summer lake levels by improving drainage out of the lake (Williams and Radcliffe 2001). It is unclear at present how such a management action would affect the tall woolly-heads site. However, if spring lake levels were permanently lowered to the point where seasonal flooding of the upper shore no longer occurred, the present habitat would likely be eliminated along with the population (Douglas et al. 2001a). Recently, large piles of spoil from a dredging operation in Somenos Creek were deposited on the shore next to the Somenos Garry Oak Protected Area, resulting in the inadvertent destruction of an estimated 1/3 of the tall woolly-heads population (M. Fairbarns, pers. comm. 2004).

Individual and population-level responses to disruptions in hydrology per se have not been well documented for any of the six taxa covered by this strategy, either for BC or in the literature in general. However, germination rates in Callitriche marginata, a rare vernal pool plant that co-occurs with Kellogg's rush and water plantain-buttercup at Uplands Park, are known to be sensitive to the timing of the onset of fall flooding, as well as to the length of inundation (Bliss and Zedler 1998). In general, species restricted to open, seasonally ephemeral wetlands are highly specialized ecologically and may not be adapted to respond to major structural alterations to habitat (Hanes and Stromberg 1998). Successful management of these areas may require research into the way individuals and populations respond to changes to the hydrologic regime or to naturally-occurring hydrologic processes.

2.1.4 Recreational activities

Recreational activities such as all-terrain vehicle (ATV) off-roading, mountain biking, and dog-walking have variously been cited as posing a threat to the integrity of vernal pool systems (Clark et al. 1998, U.S. Fish and Wildlife Service 1998b, 2000, Douglas and Illingworth 1998, Donovan 2004). Traffic from off-road vehicles, which compacts the soil, creates deep and near-permanent ruts, dislodges and crushes vegetation, and generally fragments the habitat, appears to represent the most serious threat. However, dogs also churn up moist loose soil and trample plants where they are exercised, while traffic from bicycles and people can result in the creation of new trails, change the micro-topography of pool bottoms, and crush pool vegetation. The long-term impacts of these activities have not been quantified, although light trampling, along with mowing and grazing (Barry 1998), may actually benefit some low-growing plants by suppressing taller exotic herbs and reducing competition. Some species (e.g., bog bird's-foot trefoil) may also require a certain amount of soil disturbance for germination.

The impact of off-road vehicles at Harewood Plains has been noted above (Section 2.1.2). Despite efforts by the former landholder (Weyerhaeuser Canada) to block trail access with boulder placements and ditches, recreational users of ATV's, 4x4's and (to a lesser extent) dirt bikes continue to access the area on a regular basis (C. Thirkill, pers. comm. 2004). In addition to altering the local hydrologic regime, off-roading has disturbed and compacted the soil, facilitated the spread of invasive species, and directly endangered the survival of plants such as bog bird's-foot trefoil due to crushing (Donovan 2004).

Foot, bicycle, and dog traffic represents an ongoing impact at the heavily visited Uplands Park (including Cattle Point), particularly in spring when the vernal pool/swale plants are flowering and moist soil is easily compacted. Bicycling is prohibited within the park, but this rule is rarely enforced. Where trails come close to pools, soil compaction and heavy traffic have in some instances eliminated vegetation cover (Collier et al. 2004). In the spring of 2004, local mountain bikers built an unauthorized bike jump within metres of one of the water plantain-buttercup subpopulations. To construct the jumps, dirt and mud was excavated from a nearby depression, creating a crater sized hole in the meadow. Such activities threaten the habitat of species at risk both through direct mechanical damage and by permanently altering the local hydrologic regime.

Although dogs in Uplands Park are required to be on leashes throughout the spring months (April-June), this rule is also rarely enforced, and pets can be observed at all times of year running and chasing balls in the wet meadow containing known populations of Kellogg's rush, tall woolly-heads, and water plantain-buttercup as well as numerous other red-listed plants. A recent park use survey conducted during the spring months found that almost 50 percent of visitors to the main meadow were accompanied by dogs, whereas only 2 percent of the dogs were observed to be on leash at a given time (Collier et al. 2004). As most of the plants under consideration are small and fragile (fully mature individuals of Kellogg's rush and tall woolly-heads are often less than 4 cm tall), they are easily trampled or ripped up by dogs, and may also be vulnerable to being buried or crushed by dog scat. A recently completed draft stewardship plan for Uplands Park notes: “The frequent presence of commercial dog walking activity in the Park will be highly damaging to soil and vegetation under most conditions. Particularly because plant species at risk are concentrated in open meadows where dogs are brought for exercise there is an obvious conflict between the preservation of park values and the uncontrolled use of the Park by some dogs” (Collier et al. 2004).

At Trial Island, the rosy owl-clover site attracts some foot traffic from recreational boaters, as does the dwarf sandwort site at Rocky Point (Fairbarns 2004, Penny and Costanzo 2004).

Population-level impacts from these various disturbances have not been quantified, but merit further investigation.

2.1.5 Secondary succession due to fire suppression

Prior to European settlement on Vancouver Island, fire--both natural and human-initiated--is thought to have played an important role in maintaining Garry Oak savannas (Hebda and Aitkens 1993, MacDougall et al. 2004). Aboriginal peoples used regular burning to maintain the open stand structure favourable to camas (Camassia spp.) and other root crops (Turner 1999, Fuchs 2001). The frequent, low intensity fires scorched the grass and shrub layer but left the oak canopy largely intact. The process slowed the succession of shrubs (e.g., Symphoricarpos albus, Rosa nutkana) and conifers such as Douglas-fir (Pseudotsuga menziesii), allowing herbaceous meadow plants to flourish. Over the last 150 years, fire suppression has led to encroachment of woody shrubs and Douglas-fir into areas once dominated by Garry oak, dramatically altering local community structure (Fuchs 2001, MacDougall 2004).

Like the larger Garry oak ecosystems in which they are found, many vernal pool systems are thought to be fire-maintained and thus characteristic of early- to mid-seral stages (Witham et al. 1998). Fire suppression has been acknowledged as a threat to vernal pools and wetland prairies in Oregon (Wilson 1999, U.S. Fish and Wildlife Service 2000, Kaye et al. 2001). It is unclear what historical role fire played in creating and maintaining vernal pool habitat on Vancouver Island. However, most of the vernal pools under consideration are small enough (less than 20 m²) to be easily affected by factors such as shading associated with woody plant encroachment.

At Uplands Park and other sites, fire suppression has already led to a decline in the vernal pool habitat, due to tree encroachment and a dramatic increase in the cover of shrubs such as native snowberry (Symphoricarpos albus) and nonnative Scotch broom (Cytisus scoparius), which now shade many vernally-moist microsites. The incursion of these species into adjacent habitats may have already begun to affect local hydrologic and light regimes, through alteration of drainage patterns, increased competition for water, increased shading, and thatch buildup. If unchecked, this process could result in feedback loops that accelerate the overall rate of secondary succession (Pollak and Kan 1998).

Few studies have addressed the problem of eutrophication in vernal pools. However, it is thought that extra nutrient loading in pools from agricultural or residential runoff favours the growth of other (non-specialist) aquatics that thrive in nutrient-rich conditions, and causes algal blooms that deplete the water of oxygen (U.S. Fish and Wildlife Service 1988b). Algal blooms sometimes form mats that cover plants and stunt their growth or prevent flowering (U.S. Fish and Wildlife Service 1998b).

2.1.6 Eutrophication

On Vancouver Island, eutrophication associated with guano contamination by waterfowl is a growing concern at several vernal pool sites. This process is clearly evident at Mitlenatch Island, a provincially-designated seabird preserve that hosts one of the largest nesting populations of glaucus-winged gulls (Lárus glaucéscens) in the Georgia Strait (Miller 2005). Records from bird counts indicate that the seagull population on Mitlenatch Island has been increasing at an exponential rate since the turn of the century, possibly coinciding with the appearance of garbage dumps, shellfish processing plants and sport fishing cleaning stations in the nearby coastal communities of Campbell River, Powell River, and Comox (Merilees 1992).

Recent signs of guano contamination have also been observed in vernal pools around Victoria and on adjacent Gulf Islands (Miller 2005), where resident Canada geese have been undergoing a similar exponential population increase since the 1960's (Campbell et al. 1990, Carsen 2000) and may now be having a similar negative impact on local water quality. Through their grazing, Canada geese also strip vegetation and churn soils. To date, none of the populations covered by this strategy appear to have been directly affected by Canada geese activity, although nutrient inputs from this source may have already compromised the integrity of potential habitats elsewhere.

Presumably, decomposing dog feces could also pose a contamination threat at heavily visited vernal pool/swale sites such as Uplands Park. Nutrient enrichment from this source might be sufficient to encourage algal blooms that in turn affect water oxygen/CO₂ balance, light penetration, and other aspects of water quality, with potentially negative consequences for plant flowering, germination, and seedling survival rates (C. Björk, pers. comm. 2004). Unfortunately, this topic has received little attention to date in the vernal pool literature, and requires further study.

2.1.7 Marine pollution

The Strait of Juan de Fuca is one of the most active shipping lanes in the Pacific Northwest. Petroleum tankers, freighters, cruise ships, and other marine traffic regularly pass within a few km of Vancouver Island. Trial Island, which is close to Cattle Point, hosts the lone Canadian occurrence of rosy owl-clover as well as numerous other federally listed plant species at risk (Plants at Risk Recovery Implementation Group 2005b). Many of these populations are located in low-lying pools or meadows adjacent to the shore.

Much of the vernal pool habitat on Vancouver Island and Gulf Islands occurs just above the intertidal zone and is affected by salt spray during storm events. Because of their proximity to the ocean, vernal pool species may be particularly vulnerable to marine pollution. More research is needed to determine the impacts of both diffuse marine pollution and/or a catastrophic point source spill on species at risk (Plants at Risk Recovery Implementation Group 2005b).

2.2 Recovery feasibility

Recovery is defined alternatively as “restoring a species to a viable self-sustaining population level, able to withstand stochastic events and other environmental variables of a non-catastrophic nature” (National Recovery Working Group 2004); or as “any improvement in a species' probability of long-term persistence in the wild” (Environment Canada et al. 2004). Full recovery in the first sense may not be strictly feasible if a species is by nature very rare; if it has a naturally highly fragmented distribution; or if its habitat has undergone such major alterations that it is precluded from achieving a distribution and population size consistent with the above definition. In such cases, the goal of recovery may simply be to maintain the current population size and distribution while reducing or eliminating threats (National Recovery Working Group 2004).

The following sections define recovery, and its feasibility, for individual species in this strategy. In most case, further studies and trials will be needed to determine whether there are insurmountable barriers to the restoration of existing populations, the re-establishment of extirpated populations, and the establishment of new populations. For this reason, the ecological and technical feasibility of recovery may have to be re-evaluated once further research is conducted. However, following the precautionary nature of SARA, and to prevent undue extinctions or extirpations, the premise of this strategy is that recovery is technically and biologically feasible for all six species (Table 4).

“Feasibility criteria” from Environment Canada et al. (2004).

2.2.1 Recovery feasibility of bog bird's-foot trefoil

As with many other rare Garry oak-associated plant species on Vancouver Island, we lack adequate detailed information about the historical distribution of bog bird's-foot trefoil. However, the existence of herbarium records from as early as 1939 indicate that the species has been established in and around Nanaimo for at least half a century (Table 2). Presumably, these populations predate European settlement, but this is difficult to confirm. At least three occurrences, all near Nanaimo, are believed to have been extirpated in recent decades due to human factors (Donovan 2004, A. Ceska, pers. comm. 2004). There is no evidence to indicate that bog bird's-foot trefoil was ever abundant or widespread on Vancouver Island, although it is possible these extant populations represent historical remnants of a more contiguous distribution along the southeastern portion of Vancouver Island and the Gulf Islands. The current population viability of bog bird's-foot trefoil is rated as “moderate” (Table 4).

Although bog bird's-foot trefoil biology and ecology are poorly understood, field observations suggest that regular recruitment is occurring (Donovan 2004). At least some populations appear to large enough to be self-sustaining. There also appears to be sufficient favourable habitat remaining at Harewood Plains and elsewhere to support one or more additional viable populations. Therefore, if the threats described above (and any other threats identified in the future) can be removed or minimized, and if techniques can be developed to augment and re-establish bog bird's-foot trefoil populations where required, it is expected that the species can be recovered to a fully viable population level (Table 4). Successful recovery will depend on a combination of habitat protection and management, demographic intervention, and long-term population monitoring. The level of effort required for recovery is expected to be moderate.

2.2.2 Recovery feasibility of tall woolly-heads (Pacific population)

Currently, within British Columbia, the known range of tall woolly-heads is restricted to sites around Victoria and Duncan on Vancouver Island. As with many other rare Garry oak-associated plant species, we lack adequate detailed information about the historical distribution of tall woolly-heads. Early records (Macoun 1909, 1913) exist from as far north as Sidney on the Saanich Peninsula and as far west as Ucluelet, although the Ucluelet record is now thought to be erroneous (M. Fairbarns, pers. comm. 2004). Of the nine populations documented on the Saanich Peninsula, five no longer exist, while a sixth has not been observed since 1962 and is also presumed extirpated (Douglas et al. 2001a, Table 2).

In the absence of evidence to the contrary, therefore, it can be assumed this species was once considerably more common on the Saanich Peninsula than it is now and that its current restricted distribution is due, at least in part, to anthropogenic factors. Although the three remaining extant populations are robust and appear to be self-sustaining, recent events at the Somenos site, which resulted in the partial destruction of the population there (see Section 2.1.2), serve to underscore their continuing vulnerability to disturbance. The current population viability of tall woolly-heads is rated as “moderate to high” (Table 4).

For tall woolly-heads, we define “recovery” as restoration to a level of abundance and distribution on Saanich Peninsula comparable to historical levels (as far as these are known), with the aim of ensuring a high probability of persistence for this species. Recovery to this extent is currently deemed to be ecologically and technically feasible (Table 4). This assessment is predicated on the assumption that there is sufficient suitable habitat available for translocations (Table 4), an assumption that has not been tested. Beyond this requirement, recovery success will be closely tied to activities such as habitat protection and management, demographic intervention, and long-term population monitoring. It is anticipated such activities will also benefit other rare vernal pool species in BC. The level of effort required for recovery is expected to be moderate.

2.2.3 Recovery feasibility of Kellogg's rush

Kellogg's rush is known from one site in Canada, Uplands Park, where it was first documented in 1985 (CDC HERB database 2004). There is no evidence to indicate either that Kellogg's rush was ever abundant or widespread in Canada, or that it has declined substantially from historical levels. It is possible the extant population is simply a disjunct, northern outlier of its wider North American range. On the other hand, much of the potential Kellogg's rush habitat on southeastern Vancouver Island has been lost during the past century as a result of land-use practices (Fuchs 2001). Kellogg's rush is a minute plant with a highly inconspicuous growth form, and is easily overlooked in the field. The fact that it has not been collected elsewhere is thus not necessarily reliable proof of its true historical (or current) range, and the possibility that the Uplands population represents the last vestige of a formerly more widespread distribution must also be considered (Douglas and Illingworth 1998). The current population viability of Kellogg's rush is rated as “low” (Table 4).

Kellogg's rush is an annual species and as such exhibits variable aboveground dynamics; recorded population sizes have ranged from as few as 3 to >200 individuals (Costanzo 2003). Given its apparent range marginality as well as high natural rarity, achieving a “high probability of persistence” may be an inappropriate goal to set for this species (National Recovery Working Group 2004). In this instance, population and distribution objectives that aim to maintain a minimum viable population with a moderate probability of persistence may be more appropriate (see Table 5, recovery goals). Recovery of Kellogg's rush to such a level is currently deemed to be ecologically and technically feasible (Table 4). The level of effort required for recovery is expected to be moderate to high.

2.2.4 Recovery feasibility of water plantain buttercup

The existence of several early herbarium records from Cadboro Bay and Oak Bay (e.g., Newcombe 1890, Anderson 1900) suggests that water plantain-buttercup was historically much more abundant in the Victoria area than at present. Urbanization and habitat loss are likely the main contributing factors in the decline in numbers over the past century. Although water plantain-buttercup was probably never widespread in the region, the existence of a Ballenas Island population, some 115 km to the north of Victoria, suggests the possibility that this species may have once occupied a more contiguous distribution along the southeastern portion of Vancouver Island and the Gulf Islands. However, there is no evidence to confirm this. The current population viability of water plantain-buttercup is rated as “low to moderate” (Table 4).

Since much of its original habitat has been permanently lost to development, full restoration of water plantain-buttercup to historical levels of occurrence is probably not feasible (even if these levels were known). For this species, an appropriate definition of “recovery” will thus likely fall somewhere on the spectrum between “survival” and “full recovery.” Nevertheless, there are currently no obvious biological impediments that would prevent water plantain-buttercup from attaining, with some intervention, a viable population level. Although small, the Uplands Park and Ballenas Island populations appear to have remained generally stable over the past decade (CDC HERB Database 2004). A proportion of plants at both sites flower each year, and regular juvenile recruitment has also been observed. Across its range, water plantain-buttercup exhibits a relatively broad range of habitat tolerances ranging from muddy ditches, pond margins, and stream banks to swales and moist alpine meadows (see Section III, Species Information), suggesting that neither demographic constraints nor habitat specificity will be critical limiting factors for recovery. Restoring water plantain-buttercup to a level of long-term viability will necessitate increasing both the sizes of extant populations and also the number of total occurrences, goals deemed to be both ecologically and technically feasible (Table 4). Given the species' current extreme rarity, the level of effort required for recovery is expected to be moderate to high.

2.2.5 Recovery feasibility of rosy owl clover

Nine historical records, dating from 1887 to 1954 and representing at least five distinct populations, exist for rosy-owl clover in British Columbia. All originate from the Saanich Peninsula on southeastern Vancouver Island, and all are now thought to be extirpated (Fairbarns 2004). The species is currently known from a single vernal seep on Trial Island, near Victoria. This population has fluctuated from 40 to about 1000 individuals between 1998 and 2004 (Fairbarns 2004, M. Fairbarns, unpubl. data). The site is located within the Trial Island Ecological Reserve and as such is effectively protected under the Ecological Reserves Act. However, the seepage area contains a number of alien invasive plant species that could eventually out-compete rosy-owl clover. Furthermore, the site's small size (approx. 300 m²) makes it highly vulnerable to stochastic environmental events (e.g., drought, recreational disturbances). If the rosy-owl clover population were eliminated, there would be no chance of a rescue effect from elsewhere. Consequently, the current population viability of dwarf sandwort is rated as “low” (Table 4).

Urban development in and around Victoria is believed to be the main factor responsible for the decline in the number of local occurrences. The widespread invasion of aggressive alien species may also have contributed to the decline, as rosy-owl clover does not easily tolerate competition (M. Fairbarns, unpubl. data). Although rosy-owl clover is not considered a vernal pool specialist--it also occurs in open, vernally wet habitats such as ditches, prairies and open fields--most other sites on Trial Island and nearby areas of Vancouver Island that might have been capable of supporting populations have been rendered unsuitable due to encroachment by highly competitive shrubs and grasses. At the same time, spring seepage conditions such as those at the Trial Island site may be extremely difficult to replicate artificially. Lack of available suitable habitat could thus be a serious impediment limiting the reintroduction potential of this species (Fairbarns 2004).

For rosy-owl clover, as for water plantain-buttercup, “recovery” will likely fall somewhere on the spectrum between “survival” and “full recovery.” Aside from habitat availability, there are currently no obvious biological impediments that would prevent rosy-owl clover from attaining, with some intervention, a viable population level. At the Trial Island site, seed production, although variable, is generally high (greater than 1000 seeds per year), and regular juvenile recruitment has been observed (M. Fairbarns, unpubl. data). Restoring rosy-owl clover to a level of long-term viability will necessitate (in addition to reducing or eliminating current threats) increasing both its abundance and total extent of occurrence. This is deemed to be both ecologically and technically feasible (Table 4). Given the species' current extreme rarity, the level of effort required for recovery is expected to be moderate to high.

2.2.6 Recovery feasibility of dwarf sandwort

Dwarf sandwort is known from a single vernal seep near Rocky Point. This small population was first documented in 1977 (CDC HERB database 2004). Intensive searches in adjacent areas have so far failed to yield any additional localities, and there is no evidence to indicate that dwarf sandwort was ever abundant or widespread in Canada, or that it has declined substantially from historical levels. It is possible the extant population is an isolated disjunct, the result of a chance long-distance dispersal event. On the other hand, dwarf sandwort is a minute plant with a highly inconspicuous growth form, and is easily overlooked in the field; the lack of collections elsewhere may thus not be reliable indication of its true historical (or current) range. Furthermore, much of the suitable dwarf sandwort habitat on southeastern Vancouver Island has been lost during the past century as a result of land-use practices (Fuchs 2001). The possibility that the Rocky Point population represents the last vestige of a formerly more widespread distribution must therefore also be considered.

The extant site is located on an uninhabited coastal bluff within the grounds of the Department of National Defence's Rocky Point Ammunition Depot, rendering it temporarily secure from development. Nevertheless, dwarf sandwort, like rosy-owl clover, exhibits highly variable aboveground dynamics: recorded population size has ranged from as few as nine to several hundred individuals (Penny and Costanzo 2004, M. Fairbarns, unpubl. data). Random fluctuations in demographic performance could be sufficient to drive this population to extinction. Furthermore, the site's small size (approx. 20 m²) makes it highly vulnerable to stochastic environmental events (e.g., drought, disturbances from waterfowl, trampling by trespassing boaters). If the dwarf sandwort population were eliminated, there would be no chance of a rescue effect from elsewhere. Consequently, its current viability is rated as “low” (Table 4).

Given its apparent range marginality as well as high natural rarity, achieving a “high probability of persistence” may be an inappropriate goal to set for this species (National Recovery Working Group 2004). In this instance, population and distribution objectives that aim to maintain a minimum viable population with a moderate probability of persistence may be more appropriate (see Table 5, Recovery goals). Recovery of dwarf sandwort to such a level is currently deemed to be ecologically and technically feasible (Table 4). The level of effort required for recovery is expected to be moderate to high.

2.3 Recovery goals and objectives for species at risk

Recovery goals for the six species in this strategy are presented in Table 5. Preventing further losses or declines (or an increase in rarity) is, in all cases, an important first step in recovery. This is reflected in Goal 1: “maintain extant localities at current levels of abundance or greater” (Table 5).

The species under consideration differ with respect both to life history (e.g., annual vs. perennial) and specific habitat preferences (e.g., vernal pools vs. seeps). Although currently unknown, key demographic parameters (e.g., fecundity, intrinsic growth rate, population structure, dispersal, rescue effects) likely also differ (Miller 2004). The number, size, and distribution of populations needed for long-term persistence are ultimately determined by these intrinsic factors in combination with genetics (e.g., inbreeding effects) and various extrinsic factors including the rate of habitat change, stochastic events such as fire and drought, and interactions with competitors.

Population viability analysis (PVA) is one tool now commonly employed to set specific population and distribution targets for recovery (Shaffer 1981, Menges 1986, Nantel et al. 1996, Menges 2000, Caswell 2001). PVAs provide assessment of population persistence (or extinction risk) based on a combination of empirical data and modeling scenarios. Spatially explicit, metapopulation models are useful for predicting the importance of local extinction and colonization in overall dynamics and to predict how species might be affected by anthropogenic changes to landscapes (Schemske et al. 1994, Menges and Dolan 1998). Unfortunately, the data required to construct a PVA are difficult and expensive to collect (Beissinger and Westphal 1998), and therefore not often available to recovery planners. Moreover, many aspects of plant life history can present challenges when obtaining data for PVAs. These include seed and bulb dormancy (Kalisz and McPeek 1992, Miller et al. 2004), periodic recruitment (Menges and Dolan 1998), and clonal growth (Dammon and Cain 1998, Hawryzki 2002).

Population viability analyses have not yet been completed for any of the species in this strategy. Instead, Goal 2 proposes specific population and distribution targets based on best available knowledge and the combined expertise of recovery team members. The targets reflect either the number of known historical populations, the estimated minimum number of populations required to redistribute the species across its former range, and/or the estimated number of populations required to achieve a level of viability consistent with Goal 3 (Table 5).

Future population dynamics can be predicted only through demographic models because these represent the only framework that can integrate the birth and death rates that determine changes in population size (Caswell 2001). The most critical demographic parameter for endangered species management is λ, the discrete population growth rate. When λ is less than 1, the population is projected to decline; when λ = 1, the population is stable; and when λ is greater than 1, the population is projected to grow. However, because λ can vary greatly from year to year, a single estimate of λ is usually not a reliable predictor of long-term population dynamics. In most instances, stochastic simulations that take into account year-to-year variability in λ (indicated by λs) provide a more accurate indication of long-term trends (Caswell 2001, Caswell and Kaye 2001). Goal 3 (Table 5) sets out a quantitative, measurable target for λs that, if met, will ensure species persistence over the short term (10 years). It also provides a measurable standard against which to evaluate the efficacy of management actions, thus serving as a useful target to guide future adaptive management.

Table 6 lists the short-term (5-10 year) common objectives that will contribute to achieving the recovery goals by addressing known threats and COSEWIC assessment criteria.

2.4 General description of research and management activities needed to meet the objectives

Recovery activities have been grouped under seven broad strategies designed to address the threats and meet the recovery objectives (Table 7). These are outlined below in general order of priority, although the exact order may vary with species.

1. Habitat protection and stewardship: A primary focus of this recovery strategy is to prevent further loss and fragmentation of vernal pool habitats. Habitat with known occurrences of species at risk should be protected and any new occurrences as they are discovered should become priorities for protection.
   
   Protecting and securing occurrences will involve the development of arrangements that ensure the land supporting the occurrence is perpetually managed for the benefit of species at risk. Practical methods for ensuring this protection include a) direct acquisition of land (from willing sellers) by public agencies or private organizations with formal commitments to plant conservation, b) development of stewardship agreements or conservation covenants and easements with implicated landowners, and c) legislative protection.
   
2. Landholder contact: Involving landowners/managers in effective management of vernal pool habitat will be key to the recovery of species at risk. This will include developing proactive communication with different landowners/managers and involving them in the recovery planning process. It is also necessary to determine the legislation, regulations and policy that apply to different public lands. Implicated landowners/managers should be encouraged to collaborate with researchers, participate in restoration projects, and support species at risk monitoring. Other landowners may need to be identified and contacted for critical habitat studies at historical locations and for potential reintroduction sites.
   
3. Ecological research: Further habitat surveys, along with research on habitat attributes, are needed to completely delineate survival habitat (see Knowledge gaps).
   
   The community ecology and population biology of most species are poorly known. Effective long-term management will require species-specific information on habitat preferences, demographic rates, genetics, germination requirements, dispersal patterns, pollination, impacts of competition from native and nonnative invasive species, and responses to management practices such as mowing and invasive species removal. Annual censuses should be undertaken to determine whether populations are stable, increasing, or declining over time (evaluating success in meeting one of the recovery goals is contingent upon long-term monitoring of population trends at known occurrences). Where possible, annual census data should be analyzed using stage-based demographic models to determine population growth rate (λ) as well as the sensitivity of λ to different life-stage transitions (Caswell 2001).
   
   Permanent study plots should be established in at least one population of each species, and graduate students and/or trained ecologists enlisted to design and undertake careful, quantitative field studies that investigate these and other questions relevant to conservation and management of species at risk.
   
   Prior to initiating field studies, recovery planners should contact landholders to establish protocols for conducting on-site research.
   
4. Habitat restoration and site management: Appropriate management and restoration of habitats and ecosystem processes, both at protected sites and elsewhere, is crucial to long-term species recovery. Where possible, overgrown, degraded, or otherwise non-functioning wetland sites should be identified and restored to a point where they can serve as future recovery habitat for vernal pool species at risk. Threats to currently occupied sites should be assessed, and site-specific management plans developed and implemented to address threats and restore habitat. Priority habitats for restoration and management include Uplands Park, Somenos Marsh, and Harewood Plains.
   
   Information gained during monitoring and research should be used periodically to update and modify the site management plans, as needed.
   
5. Population augmentation and establishment: Meeting the long-term recovery goals will in most cases necessitate increasing the size of extant populations and/or establishing new ones, either at formerly occupied sites or at new sites. The Action Plan will thus include an augmentation/translocation plan for each of the six species in the Recovery Strategy. Translocation plans should (a) specify the conditions warranting, and (b) guide the introduction of, off-site plant material (seeds or plants) into areas currently supporting populations or for establishing new wild-land occurrences. In this, translocation plans will follow the recommendations outlined in the document “Guidelines for Translocation of Plant Species at Risk in British Columbia” (BC Ministry of Environment, in prep.). Note: it is essential that threats be addressed prior to undertaking population enhancements or reintroductions.
   
   Where the objective is to enhance population size at existing locations, information gained through demographic research (e.g., sensitivity analysis) should be used to target the life history stages most important to population growth (Schemske et al. 1994). Prior to establishment of new populations, it is recommended that one experimental population of each species be established at a suitable site for the purpose of testing and refining management techniques, and to serve as a baseline reference. In some circumstances, creation of an ex situ seed bank may be desirable to improve options for re-establishment of species in the event of catastrophic population loss or destruction of existing habitat.
   
   Specialists in native plant propagation should be consulted for advice on plant propagation techniques and protocols for long-term seed storage. Included with the Action Plan should be a determination of appropriate numbers of seeds to be collected from each known occurrence and the length of collection intervals needed to maintain seed viability with adequate genetic diversity in perpetuity.
   
6. Inventory and monitoring: A detailed inventory (and mapping) of extant vernal pools, seeps, and swales on southern Vancouver Island and adjacent Gulf Islands is crucial to overall conservation of these habitats and will also aid in the delineation of recovery habitat for individual species.
   
   Another important step in the recovery process will be to develop and implement a monitoring plan for all occurrences to identify new and ongoing threats over a period of at least 10 years. Biologists and land managers should participate in development of this plan, which should focus on determining habitat condition, anthropogenic threats, and gross population responses to these factors, as well as detailed assessments of population size and demography. The plan should also prescribe periodic reporting procedures to ensure monitoring results are regularly forwarded to relevant stakeholders and agencies. Information gained through inventory and monitoring should be adaptively incorporated into site management plans.
   
   Because some occurrences and sub-occurrences are located on private land, monitoring protocols will need to be developed in consultation with willing landowners. For this reason, any conservation agreements or easements developed as per item 1 should include provisions for regularly scheduled monitoring.
   
7. Public outreach and education: Both habitat protection and population monitoring will be facilitated by improved public awareness and appreciation of vernal pool ecosystems. Public awareness of the rarity and conservation of vernal pool plants may also foster interest in the unique character of the regional flora and its natural history. Public outreach and information programmes should therefore be set up and maintained.

See Tables table3 & table6 for numbered list of threats and objectives.

2.5 Critical habitat

No critical habitat, as defined under the federal Species at Risk Act [s2], is proposed for identification at this time in the ‘Vernal Pools Recovery Strategy.'

While much is known about the habitat needs for survival and recovery of the species included within this recovery strategy, more definitive work must be completed before any specific sites can be proposed for protection as critical habitat. It is expected that critical habitat will be proposed within one or more recovery action plans following: (1) consultation and development of stewardship options with affected landowners and organizations and (2) completion of outstanding work required to quantify specific habitat and area requirements for these species.

A schedule of studies outlining work necessary to identify critical habitat is found below (Section 2.8).

Notwithstanding the above, information on the current state of knowledge on the habitat needs and sites of occupation of the species included in this recovery strategy are provided below.

2.5.1 Occupied habitat

For plants covered in this strategy, occupied habitat is known to exist on several federal municipal, and private lands. Adjacent upland areas are believed to contribute directly to sustaining hydrologic functions within certain areas of occupied habitat. As such, it will be important to evaluate the need for protection and management of these uplands as it relates to the maintenance of quality habitat.

For example, the security of occupied habitat for tall woolly-heads, Kellogg's rush, and water plantain-buttercup should probably include concern for the entire complex of vernal pools and swales at Uplands Park, together with their associated micro-catchments. Similarly, for bog bird's-foot trefoil, this might include the wetland complex, including riparian zones, at Woodley Range Ecological Reserve.

Sites occupied by bog bird's-foot trefoil sites at Harewood Plains, Extension Rd., and White Rapids Rd., support the greatest concentration of remaining bog bird's-foot trefoil plants in Canada.

In order to accurately define the known occupied habitat (and ultimately critical habitat) additional surveys and research will be necessary as outlined below (Section 2.8).

2.5.2 Potential habitat

Research, including field experiments, is required to determine the feasibility of translocation (e.g., seeding or planting out) as a method to expand populations into what is currently unoccupied habitat. Such potential habitat may be required to improve population viability and achieve recovery of certain species.

Potential habitat may need to undergo extensive restoration and/or threat mitigation before it becomes suitable for use. It is recommended that, for the species covered by this strategy areas that should be assessed as potential habitat include:

• Naturally-occurring vernal pool, seep, or other ephemeral wet area greater than1 m² on southeastern Vancouver Island and the Gulf Islands having the necessary ecological characteristics , along with assessing a 20 m buffer zone around said feature.
• The associated watershed and hydrologic features, including upland habitat, that contribute to the filling and drying of the above vernal pool or ephemeral wet area, and that maintain suitable periods of inundation, water quality, and soil moisture for species at risk germination, growth and reproduction, and dispersal.

There are several reasons for proposing this broad working definition for potential habitat:

1. Vernal pools are essentially small habitat islands within other communities (Jain 1976), hence island biogeographic theory predicts high rates of extinction in particular pools over time (MacArthur and Wilson 1967). The conservation implication of this is that preserves must contain many vernal pools to serve as sources of propagules for colonization. At present, we have very little concept of the number and sizes of populations required for metapopulation^{Note de bas de page 3} persistence. Until we can gather more information about the spatial dynamics of these species, the most prudent approach is to assume that many discrete habitat patches are needed for long-term persistence.
   
2. Coastal bluffs, where many vernal pools are found, comprise one of the rarest components of the Garry oak ecosystem, representing only 0.3% (1,043 ha) of the land area of southeastern Vancouver Island and the Gulf Islands (Ward et al. 1998). The occurrence and distribution of coastal bluff ecosystems are influenced by the presence of exposed bedrock geology and by proximity to the shoreline, exposure to salt spray, and prevailing winds (Ward et al. 1998). Vernal pools, in turn, make up only a tiny fraction of the area within coastal bluffs, making them an extremely rare habitat commodity. Therefore, until evidence suggests otherwise (e.g., a pool already contains rare species that would be negatively impacted by the introduction of another rare species), most intact vernal pools should be tentatively regarded as potential habitat.
   
3. A number of other vernal pool plants have either been approved, or proposed, for COSEWIC assessment (Table 8) and may be incorporated into this strategy at a later date. Defining occupied habitat broadly at the outset to include all intact vernal pools reduces the risk of prematurely excluding from the strategy any sites critical to the recovery of other species at risk.

2.6 Examples of activities likely to result in the destruction of any critical habitat identified in the future

Examples of types of activities that would be expected to result in the destruction of any critical habitat identified in the future include:

• residential development
• recreational off-road vehicle use
• garbage dumping
• logging road construction
• utility corridor maintenance
• wetland draw-down
• draining, ditching and dredging
• bicycle jump construction

2.7 Existing and recommended approaches to habitat protection

Current levels of protection for sites in this strategy range from “none” to “effectively protected.” Two sites (Ballenas Island and Rocky Point) are on uninhabited land managed by the Department of National Defence (DND). These properties are officially off-limits to the public and thus are not presently threatened by development. Current levels of protection for sites in this strategy range from “none” to “effectively protected.” Two sites (Ballenas Island and Rocky Point) are on uninhabited land managed by the Department of National Defence (DND). These properties are officially off-limits to the public and thus are not presently threatened by development.

Two sites (Woodley Range and part of Trial Island) are located within provincial ecological reserves. Ecological reserves offer protection in principle to all plant species found within their boundaries through the Ecological Reserves Act, which requires that a valid park use permit be obtained before a plant can be destroyed, damaged or disturbed. However, both species in question (bog bird's-foot trefoil and rosy owl-clover) occur near the boundaries of their respective reserves and thus remain vulnerable to influences from adjacent land use practices. Specifically, Woodley Range Ecological Reserve borders a residential area. Trial Island Ecological Reserve is bounded on one side by provincial land leased for a commercial radio broadcast station, and on the other side by a lighthouse station operated by The Canada Coast Guard.

Three populations of tall woolly-heads, a water plantain-buttercup population, and the single Kellogg's rush occurrence are afforded some measure of protection from urban development by virtue of their location in municipal parks or nature sanctuaries. There is a single historical (unconfirmed) record for tall woolly-heads at Francis/King Regional Park, which is managed by the Capital Regional District (CRD) Parks. Most other confirmed populations occur on privately-held land (Table 2).

A number of alternative approaches to protecting habitat exist. These potentially include: stewardship agreements such as conservation covenants (a legal agreement by which a landowner voluntarily restricts or limits the types and amounts of development that may take place on the land to protect its natural features), direct land acquisition. Sites covered by this strategy that would benefit immediately from some form of conservation covenant include Harewood Plains and other private properties around Nanaimo that support populations of bog bird's-foot trefoil. In consideration of its high conservation value, ecological sensitivity, and relatively undeveloped condition, Harewood Plains is also identified as a priority habitat area for land acquisition.

Somenos Marsh, Uplands Park, Christmas Hill, and Francis/King Park already have, or soon will have, habitat management plans in place. Specific protection for vernal pools and other ephemeral wet areas at these sites should be developed within the framework of existing plans.

2.8 Schedule of studies to identify critical habitat

Potentially important habitat attributes for each species, as far as these are known, have been summarized in Table 9. Further study in the following areas is required to define critical habitat for all species:

1. Identify critical habitat attributes for each species (e.g., moisture regime, length of inundation, soil and chemical properties, plant cover, etc.;). Suggested completion date: 2009.
   
2. Using established survey and mapping techniques (applied during phenologically appropriate periods), delimit the boundaries of all occupied habitats designated as survival habitat. Property boundaries at Somenos Marsh may require additional surveying in order to confirm legal tenure of the tall woolly-heads site. Additional surveys are also required to determine locations and tenure of all remaining bog bird's-foot trefoil sites south of Nanaimo. Suggested completion date: 2009.
   
3. For each occupied habitat, delimit the boundaries and condition of the associated water micro-catchment (that part of the catchment on which the hydrology of the site directly depends). Suggested completion date: 2009.
   
4. Identify, map, and describe all intact, functional vernal pools seeps, swales and other ephemeral wetlands on southeastern Vancouver Island and adjacent Gulf Islands that are currently unoccupied by species at risk. Rate these habitats for their potential to support the six identified species, as well as other species at risk. Suggested completion date: 2009.
   
5. Identify, map, and rate for restoration potential any significant, naturally-occurring vernal pools on southeastern Vancouver Island and adjacent Gulf Islands whose structure and/or function has been lost or compromised as a result of secondary succession related to fire suppression and/or alien plant invasion, eutrophication, or intentional draining. Suggested completion date: 2009.
   
6. Through experimental trials, test the suitability of high-ranking sites for plant translocations/reintroductions. Suggested completion date: 2009 and ongoing.
   
7. Using a scientifically valid and defensible method, identify the proportion, percentage, or some quantitative/qualitative measure and distribution characteristics of the habitat areas identified in steps (2-6) that must be protected to ensure the achievement of the recovery target of each species at risk. Suggested completion date: 2009.
   
8. Identify land ownership/jurisdiction for these proposed critical habitat areas, as well as landholder attitudes regarding species recovery and critical habitat designations. Suggested completion date: 2009.
   
9. Identify anticipated threats to these proposed critical habitat areas and recommend general or specific measures that can be employed to protect them from such threats. Suggested completion date: 2009.
   
10. Obtain appropriate peer review of findings resulting from steps (1-9). Suggested completion date: 2009.
    
11. Use the information gathered in steps (1-10) to provide advice to the Competent Minister responsible for finalizing Critical Habitat designations for species at risk. Suggested completion date: 2009.

2.9 Anticipated impacts on non-target species

The six species covered in this strategy represent just a small fraction of the total number of rare and/or endangered plant taxa that occur in vernal pools and other ephemeral wetlands on southeastern Vancouver Island and Gulf Islands (Table 8). Moreover, a number of COSEWIC-listed and provincial red- and blue-listed plant species co-occur in association with one or more of the species covered in this strategy and could be affected (positively or negatively) by management activities undertaken to recover the latter. Some of these taxa include: bearded owl-clover (Triphysaria versicolor ssp. versicolor) and Vancouver Island beggarticks (Bidens amplissima) (found with tall woolly-heads); Macoun's meadowfoam (Limnanthes macounii), winged water-starwort (Callitriche marginata), and Muhlenberg's centaury (Centaurium muhlenbergii) (found with tall woolly-heads, Kellogg's rush, and water plantain-buttercup); and dense spike-primrose (Epilobium densiflorum) (found with bog bird's-foot trefoil).

Because of the large number of plant taxa at risk and the high concentrations of rare species at some locations, it is not possible to describe all of the possible positive and negative impacts associated with recovery. Many of these plants are threatened by the same primary factors (e.g., development activities, introduced species, trampling) that threaten the six focal species, and therefore should benefit in general from actions taken to mitigate those threats. However, not all species should be expected to respond in the same way to specific activities such as weed removal, burning, or even protection from disturbances like trampling. Intentional introductions (translocations) of species at risk into sites already inhabited by other species at risk could also have unforeseen consequences for the latter.

Vernal pools and swales could serve as essential habitat for other rare organisms besides plants, including rare invertebrates (cf. fairy shrimps in the U.S.; U.S. Fish and Wildlife Service 1994). Currently there are no documented instances of rare animals (either vertebrate or invertebrate) associating closely with vernal pools in Garry oak and associated ecosystems on Vancouver Island (R. Cannings, pers. comm. 2005, T. Chatwin, pers. comm. 2005). However, this could simply reflect the fact that relatively little effort has been made to date to identify such associations. It is possible that rare aquatic organisms (e.g., insects or mollusks) do occur within pools, seeps, and swales, but have either been overlooked or else have not been recognized in the past as being rare (R. Cannings, pers. comm. 2005). Thus, although at present there are no anticipated negative impacts to “non-plant” species at risk associated with recovery of vernal pool plants, a similar precautionary approach is called for in this case (J. Heron, pers. comm. 2005).

This strategy recognizes the importance of the entire vernal pool community and also that of associated Garry oak ecosystems. By focusing on permanent habitat protection, maintenance of hydrologic regimes, habitat restoration, and public outreach, it is expected that the approaches recommended here will benefit not only individual species at risk but the wider ecological community as well. A program of research to identify specific impacts on associated species at risk will be provided in the Recovery Action Plan (RAP).

To ensure that recovery actions do not conflict with other actions planned or underway, open communication should be maintained with the following GOERT Recovery Implementation Groups (RIGs) and Steering Committees:

• Inventory, Mapping & Plant Communities RIG
• Conservation Planning & Site Protection RIG
• Restoration and Management RIG
  • Invasive Species Steering Committee
  • Native Plant Propagation Steering Committee
  • Fire & Stand Dynamics Steering Committee
• Invertebrates at Risk RIG
• Vertebrates at Risk RIG
• Research RIG
• Communication, Coordination & Public Involvement RIG

2.10 Social and economic considerations

Recovery of species at risk and restoration of imperiled habitats associated with Garry oak ecosystems will contribute to biodiversity, health and functioning of the environment and enhance opportunities for appreciation of such special places and species thereby contributing to overall social value in southwestern British Columbia. The natural beauty of Garry oak ecosystems in the lower mainland, Gulf Islands and Vancouver Island are an important resource for British Columbians that provide for a robust tourism and recreation industry. Protecting these natural spaces, biodiversity and recreation values has enormous value to the local economy.

Some activities occurring in and around vernal pools and other ephemeral wet areas can impact sensitive species at risk. Deleterious impacts on species at risk and the integrity of these spaces may occur through activities that:

• modify or damage hydrologic processes important for maintenance of these sites,
• directly or indirectly introduce species, native or non-native, that alter the biotic or abiotic environment in a manner detrimental to processes important for the perpetuation of the vernal pool complex,
• directly damage or destroy an individual species at risk (such as through trampling or wheeled activities), or
• modify or destroy vernal pools or other ephemeral wet areas (such as through in-filling).

Vernal pools and other ephemeral wet areas are rare on the landscape and the overall land area required for physical protection of these sites is relatively small. Effective mitigation of potentially detrimental activities can be accomplished through careful planning and environmental assessment of proposed developments and site activities and sensitive routing of travel corridors and recreational activities.

Recovery actions could potentially affect the following socioeconomic sectors: recreation; private land development; forestry; operations and maintenance activities. The expected magnitude of these effects is expected to be low in almost all cases.

2.11 Knowledge gaps

Successful decisions about conservation, restoration, and management of vernal pools and other ephemeral wet areas would be enhanced by several types of information currently unavailable.

• Species inventories. Not all historical locations have been recently inventoried to determine if populations still persist at these sites (Table 2). Information updates on extant populations are also essential to confirm current species distribution and population status, especially in the case of bog bird's-foot trefoil (M. Donovan, pers. comm. 2005).
  
• Habitat inventories. No comprehensive inventory exists of the number, size, location, type, land jurisdiction, and current land use of remaining vernal pool complexes on southeastern Vancouver Island and the Gulf Islands. Such information is crucial for identifying potential habitat, and also for identifying opportunities for conservation easements and land acquisition. Geographic information system (GIS) techniques could be used in conjunction with established survey techniques to map the distribution and size of vernal pool sites and to provide an accessible record of the ecological, landscape, and political contexts of each occurrence.
  
• Community composition. In many cases, the historical species composition of vernal pools is poorly known, making it difficult to know what the most appropriate restoration target should be for a given pool. We also need better information on the patterns and causes of variation in species composition from one vernal pool site to another. Such information would allow managers to tailor restoration and management decisions for particular sites and to improve decisions about what pool complexes to protect.
  
• Demography and population dynamics.Detailed information on demography and population dynamics, including metapopulation dynamics, is critical for establishing realistic recovery goals, and also for designing plans to meet those goals (Massey and Whitson 1980, Schemske et al.1994). In the absence of such information, the task of setting quantitative recovery criteria (e.g., the number of geographically distinct, self-sustaining populations required for viability) remains an exercise in guesswork at best. Unfortunately, the ecological dynamics of most vernal pool species in southern British Columbia remain virtually unknown. Effective management for conservation of these species will ultimately require additional data on plant vital rates (i.e., stage-specific birth, growth and death rates), population growth rate, seed bank dynamics, pollination requirements, population genetic structure, dispersal, and rescue effects.
  
• Microsite attributes. More information is needed on the individual requirements and tolerances of species with respect to microsite attributes such as soil texture and chemistry, soil depth, moisture levels, nutrient content, pH levels, and hydroperiod (length and timing of inundation) (Table 9).
  
• Threats. To help guide management of vernal pools in areas frequented by the public (such as municipal parks), site-specific studies are needed to quantify the short- and long-term impacts of trampling (by people, bicycles, and dogs), contamination from dog feces, and maintenance activities such as mowing and fire suppression. It is possible that some types of disturbance, such as trampling and mowing, actually benefit vernal pool plants by reducing competition. However, this hypothesis has not been tested.
  
• Restoration. Another important gap in our knowledge is how vernal pool ecosystems (and species) will respond to restoration efforts. Ephemeral wetlands can differ greatly in composition, and some evidence shows that different wetland sites also respond individualistically to disturbance. If this variability in response is widespread, then management recommendations cannot be applied uniformly over sites. Management planners would also benefit from a more detailed understanding of the successional dynamics of these systems.
  
• Seed storage and propagation techniques. Preferred seed storage methods are unknown for most species, as are the techniques and conditions necessary for successful ex situ and in situ cultivation (in the event this is required).
  
• “Other” rare vernal pool species.More information is needed on the ecological utilization of ephemeral pools by rare organisms from other broad taxonomic groups such as dragonflies and damselflies, butterflies, amphibians, mollusks, crustaceans and birds.
  
• Impacts of climate change. Ephemeral pool species and ecosystems are tied directly to temperature and precipitation patterns, and thus at least have the potential to be greatly affected by climate change. Any large shift in wetting and drying patterns will likely result in a different system, probably with retention of some species, but with significant changes to community and ecosystem properties also likely (Graham 2004). Pollen analyses have revealed that Garry oak savannas and associated ecosystems were historically more widespread on Vancouver Island than at present, and are likely to expand in range again if climate warming proceeds as predicted (Hebda et al. 2000). Monitoring ephemeral pool ecosystems could thus provide early indications of biological impacts of shifting climate (Graham 2004). Research might also indicate potential habitat that could develop as a result of small or large scale shifts in the ranges of different ecosystem types.

2.12 Evaluation and means of success

Evaluation of the overall approaches to recovery set out in this strategy will be largely accomplished through routine monitoring of the status of species at risk, hydrologic regimes, and within-pool habitat trends through time. Where possible, target levels have been established for plants in terms of abundance, viability, and occupied range. These targets will be used to assess progress. The Recovery Strategy will be reviewed in five years to evaluate the progress on stated objectives and to identify additional approaches and changes that may be required.

Additional performance measures that can be used to evaluate the progress of recovery include:

• Formalization of critical habitat designations through a Recovery Action Plan
• Level of protection achieved for proposed critical habitat
• Knowledge gaps addressed
• Successful prioritization of vernal pool sites for acquisition and protection under conservation covenants
• Creation of economic or other incentives for private landholders to protect vernal pools
• Number of high priority sites protected by acquisition or conservation covenants
• Designation of the six species under the provincial Wildlife Act as Species at Risk
• Number of education and outreach materials developed for vernal pools (e.g., Insert Sheets for inclusion in the GOERT field manual of species at risk in Garry oak and associated ecosystems)
• Number of sites with appropriate management plans implemented
• Creation of an ex situ seed storage programme (if deemed necessary)
• Creation of protocols, best management practices, and a decision support tool or equivalent to guide reintroductions and translocations
• Number of vernal pool sites improved through invasive species control and other restoration activities

2.13 Examples of actions completed or underway

This section provides a partial list of recovery strategies, plans, and actions already completed or underway that could affect species recovery and that link to the broad strategies for recovery identified in Section 2.5. Numerous additional examples exist that have not been listed. Further details can be obtained through the Garry Oak Ecosystems Recovery Team (GOERT).

2.13.1 Other recovery strategies

• Recovery Strategy for Garry oak and Associated Ecosystems and Their Associated Species at Risk in Canada: 2001-2006 (GOERT 2002)
• National Recovery Strategy for Bartramia stricta (rigid apple moss) (Plants at Risk Recovery Implementation Group 2005a)
• National Multi-species Recovery Strategy for Species at Risk in Maritime Meadows Associated with Garry Oak Ecosystems (Plants at Risk Recovery Implementation Group 2005b)
• National Recovery Strategy for Species at Risk in Garry Oak Woodlands (Plants at Risk Recovery Implementation Group 2005c)

2.13.2 Public outreach and habitat protection

• Preparation and delivery of education and extension materials to various audiences, including the hosting of workshops relating to rare plant conservation and preparation of “Field Manual: Species at Risk in Garry Oak and Associated Ecosystems in BC” (GOERT)
• Harewood Plains Environmentally Sensitive Areas Project (City of Nanaimo)
• Symposiums on Rare Plants Occurring on Federal Lands (Interdepartmental Recovery Fund and Canadian Forest Service)

2.13.3 Habitat mapping, surveys, and inventory

• Numerous surveys of vernal pool habitats and species completed on behalf of the BC CDC, Interdepartmental Recovery Fund, GOERT, and other agencies.
• Rare plant inventories and mapping for Uplands Park/Cattle Point (BC Conservation Data Centre 2003) and Somenos Garry Oak Protected Area (Roemer and Fairbarns 2003, Maslovat 2004, Roemer 2004)
• Ongoing studies of demographic and phenologic patterns in rosy-owl clover and dwarf sandwort are currently being conducted by M. Fairbarns.

2.13.4 Management/stewardship plans

• Uplands Park Stewardship Plan (Collier et al.2004)
• Somenos Management Plan (Williams and Radcliffe 2001)
• Uplands Park Rare Species Management Plan (District of Oak Bay, in prep.)
• Range and Training Areas Plan (Dept. of National Defence)
• Management Plan for Christmas Hill Nature Sanctuary (District of Saanich, in prep.)

2.13.5 Habitat restoration projects

• Uplands Park invasive species removal, ongoing (District of Oak Bay)
• Invasive species control on DND properties (Interdepartmental Recovery Fund, Dept. of National Defence, and Canadian Forest Service)
• Garry oak meadow restoration at Somenos Garry Oak Protected Area (GOERT).

2.14 Suggested timeline for completion of recovery action plan (RAP)

It is recommended that a draft Recovery Action Plan be completed by October 2009.