Warning This Web page has been archived on the Web.

Archived Content

Information identified as archived on the Web is for reference, research or recordkeeping purposes. It has not been altered or updated after the date of archiving. Web pages that are archived on the Web are not subject to the Government of Canada Web Standards, as per the Policy on Communications and Federal Identity.

Skip booklet index and go to page content

Recovery Strategy for the Pink Sand-verbena (Abronia umbellata) in Canada (Proposed)

1 Background

1.1 Description

1.1.1 Description of the species

The species

“Abronia umbellata is a perennial herb from a thick, deep taproot. Canadian plants have trailing stems up to 1.5 m long with short branches and opposite, thick, fleshy, densely glandular leaves. The leaves are lanceolate to narrowly egg-shaped, 2–6 cm long, 0.8–3.5 cm wide with stalks 2.5–7 cm long. The many-flowered, rounded heads occur on stout, 2–4 cm stalks. The flowers consist of a 6–8 mm long, greenish to pinkish perianth tubes that flare into pink, 5-lobed limbs 5–6 mm wide. The fruits consist of 10–12 mm long achenes that are prominently 3- or 4-winged. The wings of the achenes are wider than the achenes. Each achene has a single, brown, seed approximately 1.5 mm wide and 3 mm long” (COSEWIC 2004).

There are two subspecies (umbellata and breviflora) but only the latter occurs in Canada (Kaye 2002, Hitchcock 1964). Pink Sand-verbena is used as the English name for both subspecies. Throughout this recovery strategy Pink Sand-verbena is used to refer to Abronia umbellata ssp. breviflora.

Populations and distribution

Pink Sand-verbena is restricted to the Pacific shorelines of North America (Figure 1) ranging from central Vancouver Island south to central California (Kaye 2002).

Conservation ranks are provided in Table 1. Subspecies breviflora is ranked globally threatened. In Washington it has been reported from four sites (all in the northwest corner of the state) and has not been seen recently, so it is now ranked as extirpated for that state. Kaye (2003a) reports sixteen Oregon populations which have been observed during extensive surveys between 1998 and 2003 but thirteen of these populations were introduced as part of a recovery program. Between 1993 and 2003 three to five natural populations occurred in any given year. While there are more numerous populations in California, the subspecies has been listed as a Species of Concern by the US Fish and Wildlife Service (Kaye 2003a). According to this listing, additional information is needed to support a proposal to list under the United States Endangered Species Act (Oregon Natural Heritage Information Center 2004).

Figure 1.Distribution of Pink Sand-verbena (Abronia umbellata ssp. breviflora) in North America (from COSEWIC 2004).
Figure 1. Distribution of Pink Sand-verbena (Abronia umbellata ssp. breviflora) in North America (from COSEWIC 2004).

In Canada, populations of Pink Sand-verbena have been recorded at Clo-oose Bay, Ahousat and Pachena Bay (Figure 2, Table 2). Table 2 provides more accurate locational information than was contained in the COSEWIC status report (2004).

Table 1. Conservation ranks for Pink Sand-verbena (Abronia umbellata ssp. breviflora)Source: NatureServe 2005, BC Conservation Data Centre 2005, Washington Natural Heritage Program 2005, California Department of Fish and Game, Natural Diversity Database. July 2005.

1 California subdivides S2 taxa according to the degree of threat they face, with S2.1 taxa facing a greater degree of threat than all other S2 taxa.

The Ahousat and Pachena Bay populations are probably extirpated. The earliest record of the Clo-oose Bay population was a sighting in about 1940. Over the next few years as many as 10–12 plants were observed in any given year (Delcie Cox pers. comm. 2005, Jim Hamilton pers. comm. 2005), information which was not presented in the COSEWIC status report. Jim Hamilton reported the population to the BC Conservation Data Centre in 2000 and a specimen (Douglas #13339) was collected later that year. It is not clear whether the population was extant during the intervening years. Two plants were seen in 2000 and three plants were seen in 2001. It has not been seen since then despite careful surveys in 2002, 2003, 2004 and 2005 (Jim Hamilton pers. comm. 2005, Matt Fairbarns pers. obs.) but the species may remain in a local seedbank (see Section 1.1.2: Biological needs, ecological role and limiting factors).

Figure 2. Distribution of Pink Sand-verbena (Abronia umbellata ssp. breviflora) in Canada. Inset highlights populations south of Barkley Sound. Circle represents the recently observed (2001) population at Clo-oose Bay, triangles represent historic populations at Ahousat and Pachena Bay. This figure supersedes that in the COSEWIC status report (2004) which contained an error in the location of the Ahousat population.
Figure 2. Distribution of Pink Sand-verbena (Abronia umbellata ssp. breviflora) in Canada. Inset highlights populations south of Barkley Sound. Circle represents the recently observed (2001) population at Clo-oose Bay, triangles represent historic populations at Ahousat and Pachena Bay. This figure supersedes that in the COSEWIC status report (2004) which contained an error in the location of the Ahousat population.

Table 2. Summary of populations of Pink Sand-verbena (Abronia umbellata ssp. breviflora) in Canada
Population (Tenure)Location*Status and Description
Clo-oose Bay (Parks Canada)10U3668875390570±10m
NAD 83
Possibly extirpated, last seen in 2001 (3 plants), may persist in seedbank. In 1940s it occurred in scattered locations south of the Cheewhat River along Clo-oose Bay beach.
    Apparently extirpated, not recorded since 1915. No population estimates available.
Pachena Bay
(Parks Canada)
10U3442805406778± 500m
NAD 83
Potentially extirpated, not recorded since 1927. No population estimates available.

* Locations are in Universal Transverse Mercator (UTM) coordinates. Coordinates are only provided for federal land.

The COSEWIC status report (2004) did not provide estimates of the extent of occurrence or area of occupancy. The three populations form a triangle measuring about 112 km² but about 30% of the triangle covers open ocean, leaving a corrected historical extent of occurrence of about 80 km². If one assumes that the Clo-oose Bay population is extant while the other two populations have become extirpated, then the contemporary extent of occurrence is only a few square metres, defined by the contemporary area of occupancy. The historic area of occupancy is unknown. The extent of occurrence in Canada represents far less than 5% of the species global range. The Canadian area of occupancy and population size in 2001 represent far less than 1% of the global values for ssp. breviflora.

1.1.2 Description of the species' needs

Biological needs, ecological role and limiting factors

Pink Sand-verbena does not reproduce from cuttings or pieces of the plants (Kaye 1998); thus dispersal of the species is dependent on seeds. While Pink Sand-verbena is capable of perennating, the exposed habitat of most sites results in the loss of established plants each year during winter storms. The Canadian populations are replenished by recruitment from a local seedbank and/or by long-distance dispersal from southern populations. Seedbanks may be particularly important to the persistence of populations. The seeds remain viable for prolonged periods under laboratory conditions and rates of germination are very low until the fruit enclosing the seed has been completely abraded. Low germination rates have been observed among untreated fruits scattered along natural beaches but contrary to what is stated in the COSEWIC status report (2004), there have been no similar studies using cleaned seed (Kaye 1999a, Kaye et al. 1998, Kaye pers comm. 2005). The distribution of banked seeds is unknown but they may be buried on site (some seeds may be deeply buried and unable to germinate until sand is reworked, bringing them up to shallow depths) or in habitats unsuited to germination, growth, and maturation, e.g., in nearby off-shore sand deposits or above the storm-tide line (COSEWIC 2004, Tom Kaye pers. comm. 2005).

Populations may be quite small and periodically disappear and re-appear at a site (Kaye 2004), as has occurred at Clo-oose Bay. While this may be due to periodic 'rescue' by seeds transported north from Oregon populations, the appearance at the same locations on the same beach at Clo-oose Bay suggests that populations are probably recruited from a local seedbank rather than by long-distance immigration.

Habitat Needs

Suitable habitat in British Columbia is restricted to upper sand beaches, just below the driftwood zone, along the outer coast in the Coastal Western Hemlock Biogeoclimatic Zone (very wet hypermaritime subzone, southern variant). These gently-sloping, west-facing sites are scoured by high tides and winter storms. Few plants are adapted to this drought-prone, ephemeral habitat and only scattered specimens of European searocket (Cakile maritime), a European introduction, occur along this area of the beach (COSEWIC 2004, Matt Fairbarns pers. obs.).

Imper (1987) describes Pink Sand-verbena habitat in California north of Big Lagoon: fine sand beaches with no soil development, unstabilized free flowing sand, low, broad beach with 15–23 meters between normal high water and the foredune, an elevational difference of less than 1.5–3 meters between normal high water and the base of the foredune, and a foredune less than 2 meters with a rolling rather than abrupt face. It is also noted that Pink Sand-verbena seems to do best around the mouth of small creeks. Though it does exist in sand dunes in the southern portion of its range, Pink Sand-verbena appears restricted to sand beaches (below the foredune) from northern California to British Columbia (Tillet 1967, Wilson 1972, Imper 1987, Kaye 2004).

Suitable sand beach habitat is uncommon on the outer coast of British Columbia. Three beaches have records of former occurrences of Pink Sand-verbena and a number of other beaches are potential sites for the latter species since they support, or have supported, Yellow Sand-verbena (Abronia latifolia), a species with similar habitat requirements. While habitat is regionally uncommon it is not a limiting factor for recovery of this species.

Areas above the storm-tide line and off-shore sand deposits may play an important role in storing and releasing seed during storms. Winter storms may play a necessary role by exposing deeply-buried seed, restricting the establishment of competing vegetation, and establishing suitable germination conditions (Kaye 2002, COSEWIC 2004, Matt Fairbarns pers. obs.).


SARA defines residence as:

a dwelling-place, such as a den, nest or other similar area or place, that is occupied or habitually occupied by one or more individuals during all or part of their life cycles, including breeding, rearing, staging, wintering, feeding or hibernating[SARA S2(1)].

Individual Pink Sand-verbena plants do not use a dwelling place similar to a nest or den, and therefore do not qualify under SARA for having a residence. There would be no additional legal protection not already afforded by protection of the individual and its critical habitat.

1.2 Threats

1.2.1 Threat 1: Demographic collapse

The greatest apparent threat comes from the small size of the population, which disposes it to stochastic events and demographic collapse (populations may become too small to sustain themselves). It is not clear that the seedbank contains sufficient seeds to perpetuate the population, especially since there has been no seed production for several years.

1.2.2 Threat 2: Recreation

The Clo-oose Bay population occurs along the edge of the West Coast Trail and large numbers of hikers visit the beach during the summer months (Matt Fairbarns pers. obs.). They may cause direct damage to Pink Sand-verbena by trampling on or brushing against plants while they are hiking, camping, or collecting firewood. Hikers may also damage the plants by gathering the attractive flowers. The threat from recreational activity is ranked more highly here than it was in the original COSEWIC status report (2004) based on a review of the literature and discussions with Jim Hamilton (Matt Fairbarns pers. comm. 2005).

1.2.3 Threat 3: Winter storms

Winter storms are both a threat to individuals and a necessary process which maintains habitat for the population. The sand beach habitats of Pink Sand-verbena are usually scoured by high tides and winter storms thus few, if any, plants in this zone ever persist over winter. These storms also alter beach and sand dune morphology, altering the availability of suitable germination sites and perhaps bringing buried seeds close to the surface (Tom Kaye pers. comm. 2005). Natural shoreline geomorphic processes associated with winter storms are likely to restore suitable substrates for germination and growth every year.

1.2.4 Threat 4: Increased log debris

The impacts of winter storms may be exacerbated by elevated inputs of driftwood (due to timber harvesting and log booming along the coast) that may serve to increase beach scour and occupy potential growing sites. Driftwood also modifies the flow of sand in the ecosystem and increased deposition of wood from anthropogenic sources may eliminate the Pink Sand-verbena habitat. Some anthropogenic log mediated changes have already occurred at the Clo-oose Bay site. Further research is required to determine the ultimate effect of increased log deposition on Pink Sand-verbena habitat.

1.2.5 Threat 5: Invasive species – Beachgrass

The COSEWIC status report (2004) states that the greatest threat to the persistence of populations of Pink Sand-verbena is posed by invasive grasses such as Beachgrass (Ammophila arenariaand A. breviligulata). These robust grasses alter sand dune dynamics and, according to the status report, are capable of intruding onto upper sand beaches (the natural habitat of Pink Sand-verbena at Clo-oose Bay; COSEWIC 2004).  In Oregon, Beachgrass abundance is significantly correlated with decreases in Pink Sand-verbena size and reproduction (Kaye 2004). These grasses do not occur at Clo-oose Bay at present but may represent a serious long-term threat because they are present 'nearby' along the west coast and capable of long-distance dispersal. However, it has been proposed that the habitat at Clo-oose and Pachena Bays is not suitable for Beachgrass and that these species do not pose a threat at these locations (Nick Page pers. comm. 2005). Beachgrasses also threaten Pink Sand-verbena survival because it is likely that they presently occupy and make unavailable potential dispersal habitat. The preceding threats are more immediate.

1.3 Critical Habitat

1.3.1 Identification of the species' critical habitat

Critical habitat is proposed in accordance with the recovery goals presented in Section 2. Only “survival habitat” required by the single presumed extant population is identified in this recovery strategy. The upcoming action plan is expected to propose additional critical habitat necessary for the re-establishment or replacement of two extirpated populations along the west coast of Vancouver Island (“recovery habitat”).

Plants belonging to the Clo-oose Bay population have been observed along the upper beach from south of the Cheewhat Indian Reserve 4a southwards to the end of the beach (Jim Hamilton pers. comm. 2005); this constitutes the core area critical to the survival of the population. The storm tide zone above the upper beach, as well as the lower beach and shallow off-shore zone, are also critical to the species' survival because of the role these areas play in determining beach dynamics. Further habitat needs are detailed in Section 1.1.2. Based on the above species needs and expert consultations, the beach from the southern edge of the Cheewhat Indian Reserve 4a southwards to the end of the sand beach including a buffer of 30 metres on either side of the mean high tide mark should be considered critical habitat at Clo-oose Bay (see Table 2 for UTM locality). The 30 metre buffer is based on expert advice to protect the upper beach as well as the lower beach and shallow off-shore zone (Matt Fairbarns pers. comm. 2005).

At this time proposed critical habitat is NOT sufficient for the recovery of this species. Habitat critical to the recovery of populations elsewhere cannot be defined until suitable sites have been identified. Further work is required to identify critical habitat comprehensively, and is outlined in the schedule of studies, below. Future action plans will likely identify additional parcels of critical habitat deemed necessary and sufficient to support recovery of the species.

1.3.2 Examples of activities that are likely to result in the destruction of the critical habitat

Critical habitat may be destroyed by beach developments or introduction of non-native invasive plant species (e.g., grasses), both of which can occupy habitat directly and/or alter shoreline dynamics and sand flow. Alteration in the flow of sand through the habitat can have a dramatic effect on beach morphology to the point that the specific attributes required by Pink Sand-verbena (see Section 1.1.2) are eliminated. Specific activities which threaten recovery habitat cannot be determined until that habitat has been identified.

1.3.3 Existing and recommended approaches to habitat protection

The existing habitat at Clo-oose Bay lies within Pacific Rim National Park Reserve and is protected by provisions of the Canada National Parks Act and the Species at Risk Act.

1.3.4 Schedule of studies to identify critical habitat

Habitat suitability mapping

Sand beaches along the west coast of Vancouver Island, between Port Renfrew and Estevan Point should be mapped, field-assessed, and evaluated for suitability as sites for reintroduction. This should be completed by 2008.


Apparently suitable sites should be surveyed during late August or early September to search for unreported populations. A single year of survey of a given site is insufficient because the plant may fail to germinate and grow for several years in a row. Accordingly, all potential sites should be surveyed in at least three separate years between 2006 and 2010. If overlooked populations are found, critical habitat should be defined on a site-by-site basis.


The Clo-oose Bay site should be monitored annually (August or early September) to to determine species presence.

Any of the above actions could become incorporated into an upcoming action plan for the species.

1.4 Actions already completed or underway

The Institute for Applied Ecology (Corvallis, Oregon) has been conducting a series of studies into the biology of Pink Sand-verbena in coastal Oregon for several years. Work has been completed on several aspects of the plant's conservation biology including the genetic diversity of natural populations, dormancy mechanisms, propagation techniques, and re-introductions (Braun 1991; Karoly 2001; Kaye 1995, 1996, 1998, 1999a, 1999b, 2000, 2001, 2002, 2003a, 2003b, 2004; Kaye et al. 1999; McGlaughlin 1999; McGlaughlin et al. 2002).

Seeds were collected from plants at Clo-oose Bay in 2001 and set aside for germplasm conservation and as a source for experimental study. Some seeds are in the possession of Jim Hamilton, a resident of Clo-oose. Others, collected by George Douglas, are stored at the Pacific Forestry Centre in Victoria, BC. Seeds collected at Clo-oose Bay were used locally to test propagation techniques--all plants died without flowering: In one study, seeds were germinated in sand-filled 20-gallon drums but the plants failed to flower. In a second experiment, plants were seeded into a shallow trench but the test was aborted when hikers collected the wooden markers for kindling and trampled the test site (Jim Hamilton, pers. comm. 2005). A small number of seeds were sent to researchers in Oregon and grown in 2 gallon pots filled with coarse sand as part of a common garden experiment with local Oregon plants; these grew well and flowered (Tom Kaye, pers. comm. 2006).

The Clo-oose Bay site has been monitored annually since 1999.

1.5 Knowledge gaps

It is unclear whether there is a buried seedbank at Clo-oose Bay.

There is inadequate information on how the small quantity of seeds collected at Clo-oose Bay can be used to restore or augment the local population. There needs to be more information on number of seeds required to propagate new plants for reintroduction. Furthermore, there is little applied experience in the development of in-situ 'increase gardens' where seed reserves can be 'bulked up'.

Re-introduction techniques developed by the Institute for Applied Ecology in Oregon have not been tested on the west coast of Vancouver Island.

The demographic and phenological characteristics of Canadian Pink Sand-verbena populations, and underlying factors, have not been described. For instance the relative contributions of seedbank inputs (on site seed production, transportation to site from other populations) and outputs (transport away, and viability decline over time in the wild) are not known.

It is not clear whether populations persist at Pachena and Ahousat, or whether there are populations elsewhere on Vancouver Island.  Past survey efforts have not been sufficient, considering that populations of this taxon are often small and may remain dormant in the seedbank for several years. If buried seedbanks are present at one or both of these sites then they should be recovered using local germplasm rather than seed from elsewhere.

Further research is required to elucidate effects of log debris and invasive beachgrass on Pink Sand-verbena and its habitat.