2656

A practical framework for selecting among singlespecies, community-, and ecosystem-based recovery plans Mark S. Poos, Nicholas E. Mandrak, and Robert L. McLaughlin

Abstract: Science-based approaches for selecting among single-species, community-, and ecosystem-based recovery plans are needed to conserve imperilled species. Selection of recovery plans has often been based on past success rates with other taxa and systems or on economic cost, but less on the ecology of the system in question. We developed a framework for selecting a recovery plan based on the distributions and ecology of imperilled and nonimperilled species across available habitat types and applied it to fishes in the Sydenham River, Ontario, Canada. We first tested whether distributions of fishes were adequately predicted by habitat features hypothesized to limit the distributions of imperilled fishes versus a broader set of habitat features known to predict fish distributions. We then tested whether imperilled species occurred in similar or disparate habitat types. For the Sydenham River, an ecosystem-based recovery plan was deemed most appropriate because imperilled species occur in disparate habitat types. We lastly provide decision criteria to facilitate applications of our framework to the selection of recovery plans for other species and systems. Re´sume´ : Afin d’assurer la conservation des espe`ces en pe´ril, il est ne´cessaire de posse´der des me´thodes base´es sur la science pour choisir entre les plans de re´cupe´ration centre´s sur une seule espe`ce, sur une communaute´ ou sur un e´cosyste`me. Le choix d’un plan de re´cupe´ration se base souvent plus sur les taux de succe`s obtenus avec d’autres taxons et dans d’autres syste`mes ou sur le couˆt e´conomique, mais moins sur l’e´cologie du syste`me en question. Nous avons mis au point un cadre de se´lection d’un plan de re´cupe´ration base´ sur la re´partition et l’e´cologie des espe`ces menace´es et non menace´es dans les diffe´rents types d’habitats disponibles et nous l’avons utilise´ dans le cas des poissons de la rivie`re Sydenham, Ontario, Canada. En premier lieu, nous avons ve´rifie´ si les re´partitions des poissons pouvaient eˆtre ade´quatement pre´dites d’apre`s les caracte´ristiques de l’habitat que nous soupc¸onnions limiter les re´partitions des poissons menace´s, par comparaison a` celles obtenues a` partir d’un e´ventail connu et plus e´tendu de caracte´ristiques de l’habitat capable de pre´dire les re´partitions de poissons. Nous avons ensuite ve´rifie´ si les espe`ces menace´es se retrouvaient dans des types d’habitat similaires ou disparates. Dans la Sydenham, un plan de re´cupe´ration a` l’e´chelle de l’e´cosyste`me nous semble plus approprie´ car les espe`ces menace´es vivent dans des types d’habitat disparates. Nous fournissons enfin des crite`res de choix pour faciliter l’utilisation de notre cadre dans la se´lection des plans de re´cupe´ration pour d’autres espe`ces et syste`mes. [Traduit par la Re´daction]

Introduction Science-based approaches for selecting between recovery plans are needed to protect the growing number of imperilled species listed by conservation agencies. Federal agencies in Canada, the United States, Europe, Australia, and elsewhere are mandated by legislation to develop recovery plans for species listed as endangered or threatened (Tear et al. 1995). With the rate of species imperilment increasing (Miller et al. 1989; Ricciardi and Rasmussen 1999; Abbitt and Scott 2001), so is the need for management strategies necessary to facilitate recovery (Abell 2002; Lundquist et al. 2002). To date, decisions regarding the selection of recovery plan type have been based on comparisons of past

success rates (Boersma et al. 2001; Clark et al. 2002) or economics (Tear et al. 1995), while decisions based on the given biotic community are rarely assessed and are sorely needed (Clark and Harvey 2002). Recovery of imperilled species is typically implemented by conservation agencies using three recovery planning types: single-species, community-, and ecosystem-based plans. Single-species plans focus on the threats and recovery of an individual imperilled species (Franklin 1993; Caughley 1994). Although single-species plans are relatively successful, they are often expensive to implement (Campbell et al. 2002). Plans that focus on the recovery of more than one imperilled species or on the ecosystem in which those species live have been proposed as alternatives to single-species

Received 30 January 2008. Accepted 21 July 2008. Published on the NRC Research Press Web site at cjfas.nrc.ca on 26 November 2008. J20392 M.S. Poos1,2 and R.L. McLaughlin. Department of Integrative Biology, University of Guelph, Guelph, ON N1G 2W1, Canada. N.E. Mandrak. Great Lakes Laboratory for Fisheries and Aquatic Sciences, Fisheries and Oceans Canada, 867 Lakeshore Road, Burlington, ON L7R 4A6, Canada. 1Corresponding 2Present

author (e-mail: [email protected]). address: Department of Ecology and Evolutionary Biology, University of Toronto, Toronto, ON M5S 3G5, Canada.

Can. J. Fish. Aquat. Sci. 65: 2656–2666 (2008)

doi:10.1139/F08-166

#

2008 NRC Canada

Poos et al.

recovery plans (Grumbine 1997; Yaffee 1999; Clark and Harvey 2002). Community recovery plans facilitate the recovery of groups of imperilled species that share the same ecosystem or groups of taxonomically related species facing similar influences (Clark and Harvey 2002). Ecosystembased recovery plans focus more broadly on imperilled and nonimperilled species and their habitats (Franklin 1993). Each type of plan has been used in the management of imperilled species recovery, with varying levels of success (Boersma et al. 2001; Pikitch et al. 2004). For any recovery planning type to be effective, it must consider the biology of the imperilled species, the factors affecting its distribution and abundance within a given community, and the biology of the other nonimperilled species (Boersma et al. 2001; Clark and Harvey 2002). The development of a science-based framework for the selection of recovery plans intuitively begins with the description of the relationship between imperilled species and their habitat. Loss of critical habitat is the leading cause of species imperilment; consequently, greater than 99% of recovery plans target habitat protection as an objective for species recovery (Tear et al. 1995; Abbitt and Scott 2001; Dextrase and Mandrak 2006). However, as is often the case with imperilled species, biological information may be lacking and decisions for species recovery are needed regardless of data paucity (Foin et al. 1998). We use the identification of habitat-related factors (e.g., reduction, fragmentation, or modification) as a key starting point to discriminate between factors influencing imperilled species. We define imperilled species as those species designated as either threatened or endangered or those species given a conservation designation as suggested by the Committee for the Status of Endangered Wildlife in Canada. Although there are notable exceptions in factors that affect imperilled species (e.g., invasive species, disease, and over harvesting), our framework begins with the assumption that habitat-related factors are important first considerations when developing a recovery plan and that other factors can be included as they become known (see Dextrase and Mandrak 2006). Alternatively if other factors are known to influence the species or system in question, they should be included in the scenarios that we describe. Three scenarios regarding the distribution of imperilled and nonimperilled species across available habitat types can be initially used to select a suitable type of recovery plan (Fig. 1). The first scenario considers an assemblage or community where imperilled species occur in a subset of habitats that are unique relative to those used by nonimperilled species (Fig. 1a). This scenario might be expected in systems where imperilled species are indicative of changes to specific habitats (Landres et al. 1988; Lawler et al. 2002). For example, declines in imperilled or sensitive species are often an indicator of broader habitat changes (e.g., habitat fragmentation in birds; pH in invertebrates and fish), where the larger suite of more common species has yet to be impacted (Jackson and Harvey 1993; Herkert 1994). For this scenario, single-species and community recovery plans could be implemented easily because mitigating the same set of habitat variables could be proposed to facilitate the recovery of all of the imperilled species (Fig. 1a). However, if the habitat variables are altered, the risk of nonimperilled species being negatively affected remains.

2657 Fig. 1. Three scenarios (a–c) representing a framework of the habitat associations of imperilled and nonimperilled species. Solid circles represent individual imperilled species, Xs represent nonimperilled species, and the larger, open circles represent variation in all other taxa in the assemblage. Habitat gradients (axes) represent the factors of concern for the species or system in question.

The second scenario considers an assemblage or community where imperilled species occur together in a subset of the habitats used by nonimperilled species (Fig. 1b). This scenario might be expected where a system-wide habitat change negatively affects the entire suite of species or affects a subset of sensitive species. For example, changes in water temperatures through deforestation or loss of riparian zones are thought to affect both imperilled and nonimperilled species (France 1997; Casselman 2002). In this situation, single-species and community recovery plans would be well suited to facilitate the recovery of imperilled species because habitats could be modified to benefit all imperilled species (or a subset of sensitive species); accordingly, alterations to assist the recovery of imperilled species would have smaller effects on nonimperilled species because of the overlap in habitat preferences (Fig. 1). The third scenario considers a system where different imperilled species occur in subsets of habitats and their geographic distributions overlap with nonimperilled species (Fig. 1c). This scenario is likely to occur where many environmental factors contribute to species declines. Under this scenario, single-species and community recovery plans may impede recovery of other imperilled species, as well as negatively affect nonimperilled species. For such a scenario, ecosystem-based recovery plans would be better suited than single-species or community recovery plans because improvements at a watershed–ecosystem level would presumably cause improvements to the myriad of species found in that watershed–ecosystem. Single-species or community recovery plans would only be feasible if the distributions of imperilled species were restricted to different geographical units (e.g., subwatersheds), where each could be managed separately for the threats occurring there. These three scenarios represent key examples from a continuum of scenarios, and specific scenarios and variables (e.g., other threats) can always be implemented as they occur. We used four steps to explicitly apply our framework to imperilled fish species in the Sydenham River, Ontario, Canada (Fig. 1). We applied the framework to the Sydenham River watershed because this 2725 km2 watershed is being used as a model system to develop scientific tools needed to support recovery planning for imperilled fish species, as required by the Canadian Species at Risk Act (Dextrase et al. 2003; Staton et al. 2003). The Sydenham River is species-rich for a north temperate system, with over 80 na#

2008 NRC Canada

2658

Can. J. Fish. Aquat. Sci. Vol. 65, 2008

Fig. 2. Sites sampled (n = 75) in the Sydenham River, Ontario, in 2002 and 2003. Open circles represent locations sampled in 2002, solid circles are locations sampled in 2003, and open circles with black centers are locations sampled in 2002 and 2003. Numbers next to circles represent multiple sites in close proximity to one another. The inset shows the location of the Sydenham River relative to the Laurentian Great Lakes. Shetland CA is the Shetland Conservation Area.

tive and introduced fish species. First, we identified habitat features hypothesized in the literature to influence imperilled fish species in the Sydenham River. These species include the endangered northern madtom (Noturus stigmosus), the threatened eastern sand darter (Ammocrypta pellucida) and spotted gar (Lepisosteus oculatus), and species of special concern: greenside darter (Etheostoma blennioides), pugnose minnow (Opsopoedus emiliae), bigmouth buffalo (Ictiobus cyprinellus), and blackstripe topminnow (Fundulus notatus). These species are all currently encompassed within one recovery plan and are therefore considered collectively (Dextrase et al. 2003; Staton et al. 2003). Second, we tested whether those habitat features identified in the literature were good predictors of the types of habitats where the imperilled species were found. We conducted this test because recovery plans typically rely on species accounts from the literature without explicitly examining the adequacy and efficacy of the habitat features hypothesized to be influencing the distributions of imperilled species in the system they are managing. Third, we examined the habitat associations of imperilled and nonimperilled species to test among the three scenarios depicted in Fig. 1. Finally, we examined whether imperilled species were restricted to different subwatersheds that could be managed separately, as posed under scenario 3 (Fig. 1c).

Materials and methods Site selection Seventy-five sites were sampled across the Sydenham River watershed. Fifty sites were sampled in 2002, 25 of these were resampled in 2003 along with 25 new sites (Fig. 2). Sixty-two sites, at least 2 km apart, were selected to achieve uniform coverage across the watershed. Nonwadeable sections of the river near Shetland Conservation Area and in the lower portions of the watershed were not sampled. We defined each site as either a pool–riffle sequence or a stream reach approximately 60 m in length. Because much of the watershed has been impacted by agricultural drainage (Dextrase et al. 2003), many stream reaches have become channelized, lacking pool–riffle sequences. When a sample site was located within a channelized reach, sites were defined as a 60 m stream reach, following Bohlin et al.’s (1989) recommendation to use sites of roughly equal length. Fish sampling Fishes were collected with an electrofisher, seine nets, gill net and overnight sets of Windermere traps. Each site was sampled using single-pass electrofishing with a backpack Smith-Root model 12 electrofisher and using a seine net. #

2008 NRC Canada

Poos et al.

The type of seine net (bag or straight) used and the deployment of the remaining sampling methods were dependent on the depth of each site. A bag seine (8.2 m  2 m  2 m, 7.5 mm mesh), gill nets (5 cm stretched mesh), and Windermere traps (1 m diameter) were used at sites when depths were over 1 m. A straight seine (9.8 m  2 m, 7.5 mm mesh) was used when depths were 0.5–1 m. A small straight seine (4.3 m  0.6 m, 7.5 mm mesh) was used when depths were <0.5 m. Captured fishes were identified to species, counted, and either returned to the river (as the case for all imperilled species) or preserved as voucher specimens and sent to the Royal Ontario Museum (Toronto, Ontario). Each site was systematically sampled by the electrofisher (pulsed DC current at 200–225 V, 60 Hz, pulse length = 3 ms) in an upstream direction at a rate of 5 electrofishing seconds per square metre (OMNR 2007). Total sampling effort depended on the area of the site; however, a minimum of 2000 s (mean 4257 ± 130 s) was shocked at each site, exceeding the recommendation of at least 1500 s proposed by Yoder and Smith (1999). The entire area of the site was sampled again by hauling a seine net in the downstream direction. The number of seine hauls varied from five to eight depending on the number of obstructions at the site. In total, the combination of electrofishing and seining captured over 95% of the imperilled species at sites where they were known to occur (Poos et al. 2007). Gill nets (5 cm stretched mesh) were used as block nets during sampling, unless a natural obstruction (e.g., shallow water) was present. Although absences were assumed as true absences because of the relatively large amount of sampling compared with others (e.g., Yoder and Smith 1999, OMNR 2007), we validated this assumption by modeling confidence intervals around each subsection and found all imperilled species had asymptotic relationships with a probability of detection >0.95 at the level we sampled (Poos et al. 2007). Habitat measurement Forty-two habitat variables were measured at each site. These represented habitat variables considered to specifically influence the distribution of imperilled fish species in the Sydenham River (Dextrase et al. 2003), as well as habitat variables considered to influence the distribution of stream fishes in general (Richter et al. 1997; Flather et al. 1998; OMNR 2007). Habitat variables were categorized as geomorphological, substrate, and chemical. Geomorphological variables measured attributes of the channel and stream hydrology and riparian buffers (OMNR 2007). Substrate variables measured the physical stream bottom where fishes were sampled (OMNR 2007), while chemical variables measured water quality at the sample sites (Table 1). Geomorphological and substrate variables were measured using the Ontario Stream Assessment Protocol (OMNR 2007). This protocol was selected because it has proven useful for testing hypotheses regarding how fish communities respond to habitat change and for developing conservation strategies (Stanfield and Jones 1998). In-stream measurements were recorded at six equally spaced points along 10 equally spaced transects, totaling 60 in-stream measurements per site. In-stream measurements included hydraulic head (HH) (±1 mm) as an index of velocity, water depth (±1 mm, AvgDepth), average bank undercut (±1 mm), per-

2659

centage of sample points with nonfilamentous algae (NFL), and percentages of sand (median particle size 0.1 mm; labeled as PrSand), cobble (median particle size 20 mm; labeled as PrCobble), and clay (median particle size 0.01 mm; labeled as PrClay). Riparian measurements included percent riparian buffer (PrVegSq), amount of aquatic grasses (PrGrass), average bank particle size (±1 mm), and average bank angle. Water quality variables were measured using a HydroLab DataSonde 4a multiprobed sensor. The sensor measured specific conductivity (±0.001 mScm–1), turbidity (±50 nephelometric turbidity units, NTU), pH (±0.2), dissolved oxygen (±0.2 mgL–1), and nitrate concentration (±2 mg(I-N)–1). These measurements were made approximately 20 m above the upstream end of the sample site continuously for the duration of sampling. Measurements were averaged (± standard error) over the sampling duration for each site sampled. Data analysis Redundancy analysis was used instead of canonical correspondence analysis because a linear response model was considered appropriate given that the extent of species turnover (beta diversity) along the longest gradient of a detrended correspondence analysis (DCA) was less than 3 (Leps and Smilauer 2004). Further, the imperilled species captured in this study were all found in greater than 10% of the sites sampled (using all sampling methods) and as such did not suffer from zero-inflated bias as one would assume with most imperilled species (Martin et al. 2005). Species scores were standardized by dividing by their standard deviation so that species with large variance did not unduly influence the analysis (ter Braak and Smilauer 2002). Scaling was focused on interspecies correlations to interpret spatial relationships among species (ter Braak and Smilauer 2002). The 42 habitat variables were reduced to a more interpretable number of variables in two steps. First, eight variables that were only suitable for measurement at a small number of samples sites (<0.1% of transect points) were removed. For example, the presence of watercress (Rorippa nasturtiumaquaticum), a groundwater indicator in temperate climates, revealed that the plant was found at only 0.02% of the transect points. Second, four derived variables were created by averaging multiple repeated measurements and using the average, instead of the individual measurements. These derived variables were average bank angle, average percent riparian buffer, average undercut, and average bank particle size. For example, the four riparian measurements of the slope of each bank were summarized as average bank angle. In total, 20 variables were included in the overall redundancy model (Table 1). A reduced redundancy model including only those habitat variables identified in the literature as influencing imperilled species found on the Sydenham River was also developed (Table 1). Statistical significance of individual axes was tested using Monte Carlo permutation tests (Leps and Smilauer 2004). After the first axis was tested, it was used as a co-variable in the test of the second axis, and so on, until all axes were tested for significance in explanatory power (Leps and Smilauer 2004). Only axes significant at p < 0.05 were included in subsequent sections. #

2008 NRC Canada

2660

Can. J. Fish. Aquat. Sci. Vol. 65, 2008 Table 1. Variables from the literature identified as influencing imperilled fish species in the Sydenham River, Ontario (reduced model) and stream fish assemblages in general (overall model). Predicted variable (Measured variable)

Eastern sand dartera

Greenside darterb

Blackstripe topminnowc

Bigmouth buffalod

Spotted suckere

Reduced model Depth (AvgDepth) Flow (HH) Nonfilamentous algae(NFL) Riparian buffer (PrVegSq) Aquatic grasses (PrGrass) Sand substrate (PrSand) Cobble substrate(PrCobble) Turbidity (AvgTur) Nitrate (AvgNO3) Dissolved oxygen (AvgDO)

0 (0) 0 (+) 0 (+) + (+) 0 (–) + (0) 0 (+) – (–) – (–) + (+)

– (–) + (+) + (+) + (+) 0 (–) 0 (+) + (+) – (–) 0 (–) + (+)

0 (+) 0 (–) 0 (–) 0 (–) + (+) 0 (–) 0 (–) + (+) 0 (+) 0 (–)

+ (+) – (–) 0 (–) – (–) 0 (+) 0 (–) 0 (–) + (+) 0 (+) 0 (–)

+ (+) – (–) 0 (–) 0 (–) + (+) 0 (–) 0 (–) – (+) 0 (+) 0 (–)

Overall model Clay substrate (PrClay) pH (pH) Average undercut Bank particle Average bank angle Average substrate size Gravel substrate Pebble substrate Boulder substrate Conductivity

0 (–) 0 (–) NA NA NA NA NA NA NA NA

0 (0) 0 (0) NA NA NA NA NA NA NA NA

0 (–) 0 (–) NA NA NA NA NA NA NA NA

0 (–) 0 (–) NA NA NA NA NA NA NA NA

0 (+) 0 (+) NA NA NA NA NA NA NA NA

Note: Predicted and observed signs of the associations between presence of an imperilled species and the habitat variable are shown (+, a positive correlation; –, a negative correlation; 0, no correlation). Observed associations appear in parentheses. NA, not applicable. a

Williams et al. 1989; Daniels 1993; Holm and Mandrak 1996; Facey 1998. Scott and Crossman 1973; Grossman and Freeman 1987; Greenberg 1991. c McKee and Parker 1982; McAllister 1987; Dextrase et al. 2003. d Scott and Crossman 1973; Stewart et al. 1985; Goodchild 1990. e Trautman 1957; Scott and Crossman 1973; Page and Burr 1991. b

The usefulness of habitat variables identified from the literature as influencing the imperilled species found on the Sydenham River was assessed by comparing the reduced redundancy model with the overall redundancy model. The comparison was made by using the variables isolated in the reduced model as co-variables and testing whether addition of the remaining variables in the overall model significantly increased the proportion of variance explained. Statistical significance of the improvement was tested using an unrestricted Monte Carlo permutation test (p < 0.01). The permutation test determined whether the variance in species’ occurrences explained by the additional habitat variables was greater than that expected by chance (Leps and Smilauer 2004). Evaluation of the scenarios Biplots of the first two habitat axes from the redundancy analysis were used to generate figures comparable to Fig. 1, visualize the habitat associations of imperilled and nonimperilled fishes, and distinguish among the scenarios. Locations where imperilled species were detected were plotted on maps using the geographic information system ArcView 9.1 (ESRI, Redlands, California) to determine whether the

imperilled species were aggregations occurring in certain subwatersheds.

Results We collected 67 fish species from the Sydenham River, including five imperilled species. In total 43 928 (2002 = 20 685; 2003 = 23 244) fishes were captured, with the imperilled species eastern sand darter, greenside darter, blackstripe topminnow, bigmouth buffalo and spotted sucker (Minytrema melanops) accounting for 0.19%, 9.20%, 1.10%, 0.07%, and 0.04% of the total abundance, respectively. We did not capture three imperilled fishes, the pugnose minnow, spotted gar, and northern madtom. Two of these species (northern madtom and spotted gar) are thought to be extirpated from this watershed, while the pugnose minnow was captured in nonwadeable sections not sampled in this study (N.E. Mandrak, unpublished data). Habitat associations of imperilled and nonimperilled species were most consistent with scenario 3 (Fig. 1c). The imperilled species were found in contrasting habitat types interspersed throughout the range of habitats used by fishes without conservation designations. Moreover, the eastern #

2008 NRC Canada

Poos et al.

2661

Fig. 3. Biplots from a redundancy analysis displaying the habitat associations of imperilled fishes (symbols) in the Sydenham River, Ontario, in 2002 and 2003 in relation to nonimperilled fishes (Xs) using habitat factors identified in literature as influencing imperilled fishes (reduced model) and subdivided as (a) geomorphological, (b) substrate, and (c) water quality variables and (d) additional variables thought to influence stream fish in general, but were shown to be good predictors of imperilled species (overall model). Percent variance explained by each axis is provided in parentheses. Species are shown by the following: circle, eastern sand darter; triangle, greenside darter; square, blackstripe topminnow; star, bigmouth buffalo; diamond, spotted sucker.

sand darter and greenside darter were found in habitats characterized by high hydraulic head (HH), and hence flow, large amounts of nonfilamentous algae (NFL), high proportions of cobble substrate (PrCobble), high amounts of riparian buffers (PrVegSq), and high concentrations of dissolved oxygen (AvgDO). Conversely, the blackstripe topminnow, bigmouth buffalo and spotted sucker were found in habitats characterized by deep water (AvgDepth), high turbidity (AvgTur), and high nitrate concentration (AvgNO3). The habitat variables hypothesized as important in literature accounts of the imperilled species were found to be good predictors of the occurrence of imperilled species in the Sydenham watershed. The redundancy analyses demonstrated that the model based on a broad suite of habitat variables known to influence stream fishes in general did not provide a significantly better fit then a model based solely on those habitat variables believed to influence imperilled species (14.5% for the reduced redundancy model versus 17.3% for the overall redundancy model, p > 0.05). In addition, there was a significant relationship between the habitat variables believed to influence imperilled species and the

occurrence of the fishes in general (axis 1: eigenvalue = 0.1; F = 5.3; p = 0.001; axis 2: eignenvalue = 0.04; F = 2.6, p = 0.002). Our consideration of variables not identified for imperilled species in the literature identified two additional variables that were helpful in predicting the occurrence of imperilled species in the Sydenham River. The variables were percent clay substrate (PrClay) and pH (Fig. 3d). Adding those variables to the reduced redundancy model explained an additional 2.6% of the variance in the occurrence of fish species (eigenvalue = 0.017, F = 2.07, p = 0.003; eigenvalue = 0.014, F = 1.75, p = 0.01, respectively). Only axis 1 was significant (axis 1: eigenvalue = 0.019; F = 1.9; p = 0.01; axis 2: eigenvalue = 0.012; F = 0.87, p = 0.32). The occurrence of the eastern sand darter, blackstripe topminnow, and bigmouth buffalo tended to be in habitats characterized by low amounts of clay and low pH. Spotted sucker also tended to be in habitats characterized by high amounts of clay and high pH. The occurrence of the greenside darter was unrelated to the amount of clay and water pH (Table 1; Fig. 3). #

2008 NRC Canada

2662

Discussion As rates of species imperilment and habitat degradation increase, meeting the challenges of species recovery has become increasingly difficult (Foin et al. 1998; Simberloff 1998). One strategy to circumvent this downward trend has been to reduce the emphasis on species-specific recovery plans (which require more effort per species) and use ecosystem-based recovery plans, where several imperilled species can be managed across entire systems, such as watersheds (Grumbine 1994; Brunner and Clark 1997). Although ecosystem-based recovery plans may be scientifically appropriate, the willingness of managers to adopt these strategies may be difficult given the past success of singlespecies recovery approaches, which have been four times more likely to exhibit recovering status trends for imperilled species (Boersma et al. 2001; Clark and Harvey 2002). Of the number of limitations addressed for shortcomings of ecosystem-based recovery plans, the lack of integration of ecologically relevant information into recovery plans are almost exclusively linked with lack of improved status (Tear et al. 1995; Clark et al. 2002). This may be more related to discrepancies between the type of recovery plan chosen and the specific conservation tasks identified for recovery in the biological system or species in question rather than a specific type of recovery plan (Lundquist et al. 2002). For example, the lumping of species into community plans is often not based on ecologically defensible criteria (e.g., similarity of threats) or represents inconsistent associations with preferred habitat ranges (Clark and Harvey 2002; Rahn et al. 2006), which is counterintuitive to the goals of that recovery plan. This study introduces a framework to improve the implementation of selecting an appropriate recovery plan type by providing a basis for incorporating species-specific associations directly into decision criteria (Fig. 4). The goal of using decision criteria is to incorporate an understanding of species–habitat relationships for the selection and implementation of an appropriate type of recovery plan. Classification and validation of habitat variables thought to be critical for the conservation of imperilled species is fundamental to the presented framework because habitat characteristics often become key components of species recovery plans (Tear et al. 1995), under the assumption that they are important, but often without appropriate validation (Gerber and Hatch 2002). Our application quantitatively demonstrated that habitat variables identified in the literature as influencing the distributions of imperilled fish species found in the Sydenham River were good predictors of the habitats where those species were found; however, this scenario may not always occur. Validation of habitat variables can be specifically useful for isolating uncertain predictors or for identifying new habitat predictors not previously recognized in the literature. We consider the classification and validation of habitat characteristics as preliminary steps in the decision framework (Fig. 4). These steps are necessary for the success of the framework because plans that incorporate species-specific biology are typically more successful than those that do not (Boersma et al. 2001; Lawler et al. 2002; Lundquist et al. 2002). Further, identification of new, important habitat predictors of imperilled species (through validation procedures) may lead to the development of better recovery plans by improving our knowledge

Can. J. Fish. Aquat. Sci. Vol. 65, 2008

of habitat associations for imperilled and nonimperilled species. The discovery of previously unimportant habitat predictors will increase knowledge of the biology of species inhabiting the management area, thereby increasing rigor to the assessment of scenarios 1–3. The framework and decision criteria presented here represent spatially explicit criteria from which to base management decisions for communities with imperilled species. By considering the effect of species threats with geographic location, there is a high likelihood that the success rate of community- and ecosystem-based plans can be improved. For example, we found that imperilled species in the Sydenham River were found in different habitat types and were not restricted to different subwatersheds, which suggests considerable challenges to recovery planning using community recovery plans. Such situations can be expected to pose constraints whereby actions taken to assist the recovery on one imperilled species will potentially hamper the recovery of another. In addition, strict calculation of threat similarity (Clark and Harvey 2002) may not encompass the geographic extent from which those threats should be mitigated. For example, the Sydenham River recovery strategy has focused on reducing turbidity to mitigate threats to critical habitat (Dextrase et al. 2003). Reducing turbidity throughout the watershed is predicted to benefit the eastern sand darter and greenside darter, because both species occur at sites characterized by low turbidity. However, reducing turbidity could negatively affect the blackstripe topminnow, bigmouth buffalo, and spotted sucker because these species occur at sites characterized by high turbidity. Single-species or community recovery plans could be effective if the eastern sand darter and greenside darter were primarily found in the less turbid eastern subwatershed and the blackstripe topminnow, bigmouth buffalo, and spotted sucker were primarily found in the turbid northern subwatershed. Such a scenario would require partitioning of recovery actions for each watershed so that each subwatershed could be managed differently. However, this is not the case in the Sydenham River, thereby, supporting selection of an ecosystem-based recovery plan and not a community plan (Fig. 4). Recovery actions focusing on single factor remediation, such as the reduction of turbidity, need to be tempered with the trade-offs in species-specific responses, specifically that any shift may be at the detriment of at least some imperilled species. Although this type of situation may ultimately doom ecosystem-based approaches to lower success rates (by reducing the viability of some imperilled species), presumably there would be a net benefit to the other imperilled and nonimperilled species by improving habitat quantity and quality through an ecosystem approach. A stronger, science-based approach using decisions tailored to the organisms and system of concern can help with the selection of a recovery plan type. With the rate of species imperilment increasing (Miller et al. 1989; Ricciardi and Rasmussen 1999; Abbitt and Scott 2001), so is the need for recovery plans defining management strategies necessary to facilitate recovery (Abell 2002; Lundquist et al. 2002). To date, recommendations regarding the choice of recovery plan have been based on either the past success rates with other taxa and systems (Boersma et al. 2001; Clark et al. 2002) or economic cost of the different recovery plans #

2008 NRC Canada

Poos et al.

2663

Fig. 4. Decision criteria for determining the appropriate recovery planning type for imperilled species based on species–habitat associations identified in Fig. 1.

(Tear et al. 1995) and not necessarily on the ecology of the system in question. While past success and economics are important considerations for choosing a type of recovery plan, they do not explicitly recognize that the success rate and cost of each type of recovery plan may vary according to the ecological situation in which the recovery plan is applied. For example, the higher success rates observed for

single-species recovery plans relative to community plans found by Boersma et al. (2001) may be a consequence of the former being implemented in simple systems and of the latter implemented in more complex, challenging systems. Alternatively, the cost effectiveness of community recovery plans should be improved when the recovery plan incorporates knowledge of local conditions, and this should also im#

2008 NRC Canada

2664

prove the rate of recovery (Grumbine 1994). In addition, both single-species and community approaches may be more cost effective in situations where factors contribute similarly to communities with imperilled species, such as an increase in water temperature through the loss of riparian buffers, and where recovery can be achieved in a relatively short time frame (Lawler et al. 2002). Of course, the framework developed has limitations and understanding these can help identify where they can be applied most successfully and how it might be improved. First, implementing the framework requires quantifying the species–habitat associations, but it does not demonstrate cause and effect relationships. Therefore, habitat modification as a recovery action, such as reducing turbidity, may not necessarily lead to corresponding changes in the abundances of imperilled species due to the complex ways in which species interact with the environment and the effects of other factors not measured, such as connectedness among suitable habitats. However, cause and effect could be tested subsequently either in formal field experiments or through adaptive management. Second, the variance explained from multivariate analyses (e.g., redundancy analysis, canonical correspondence analysis) will tend to be lower than their univariate counterparts (e.g., logistic regression, least squares regression, analysis of variance) and should not be taken as a surrogate for poor fit of the model per se (Legendre and Legendre 1998). In this case, the low variance on the multivariate axes represent the overall ability of the environmental correlates in describing the relatively large species assemblages (n = 67) across the relatively large number of sites sampled (n = 75) and as such is not surprisingly low. Although we should always strive for rigorous species assessments, in many cases extensive sampling of the kind preformed here may not be feasible. As is often the case with endangered species, biological information may be sparse for rare species and decisions for species recovery are needed regardless (Foin et al. 1998). In general, the models represent the integration of the best available knowledge across the study system and species in question. As the decision criteria presented here (Fig. 4) does not incorporate explicitly statistical methods, care should be used to ensure appropriate models are applied. Finally, we did not consider other imperilled taxonomic groups, such as mussels (Metcalfe-Smith et al. 2003) or benthic macroinvertebrates or the imperilled fishes not captured in our sampling. The addition of other imperilled species may provide more insight as to which habitat characteristics would be best mitigated in the situation where several imperilled species need to be alleviated with contrasting habitat types. Indeed, the majority of the declines attributed to imperilled mussels are thought to be due to turbidity (Dextrase et al. 2003). The inclusion of other imperilled species would, therefore, improve the implementation of an ecosystem-based recovery strategy, and such data should be included when implementing the most appropriate type of recovery plan. However, in our situation, adding the imperilled species not sampled would not have altered our results or the choice of which recovery plan type would be most appropriate. The reason for this is that as imperilled species are added to the multi-

Can. J. Fish. Aquat. Sci. Vol. 65, 2008

variate data, the more likely that those species will be found with some aspect of habitat that is contrasting with another imperilled species without geographic isolation (scenario 3). This framework has a number of strengths that may allow it be transferred to other species and systems. First, the framework for the selection of recovery plan type can be standardized and is straightforward and transparent. Second, the framework is rigorous by using strong science to validate proposed habitat predictors and to test among competing scenarios and is comprehensive by explicitly considering imperilled and nonimperilled species. Third, the framework presents opportunities to learn, for example, through identifying new habitat predictors to include in the community model (e.g., pH or percent clay in our model). Fourth, the framework lends itself to an adaptive approach as biological information becomes more available, limiting factors become validated, and extinction risks become better quantified. Finally, the framework is flexible in the use of most multivariate statistical methodologies, insofar that they are used appropriately and can distinguish between the scenarios we present. These advantages may lead to improved success with species recoveries by lessening the use of inappropriate recovery plans that are based solely on past success and tight fiscal resources (Grumbine 1994, 1997; Boersma et al. 2001). As the recovery of imperilled species will be hindered if the objectives of recovery plans cannot be met (Boersma et al. 2001; Lundquist et al. 2002) and as loss of imperilled species rises, it is becoming increasingly important that conservation actions, such as the selection of recovery plans, be transparent, sound, and scientifically defensible.

Acknowledgements Funding was provided by Fisheries and Oceans Canada Species at Risk Program (SARCEP) and the federal Interdepartmental Recovery Fund (IRF) to NEM and a Natural Sciences and Engineering Research Council of Canada (NSERC) discovery grant to RLM. We thank Environment Canada and the Royal Ontario Museum for contributing to this study. We also thank M. Finch, J. Keenliside, M. Parslow, and S. Foley for their field assistance; A. Dextrase, S. Staton, J. Barnucz, T. Heiman, E. Holm, J. Clark, F. Neave, and T. den Haas for technical assistance; D.A. Jackson for statistical advice; and D.L.G. Noakes and anonymous reviewers for providing suggestions on earlier drafts of this manuscript.

References Abbitt, R.J.F., and Scott, J.M. 2001. Examining differences between recovered and declining endangered species. Conserv. Biol. 15: 1274–1284. doi:10.1111/j.1523-1739.2001.00430.x. Abell, R. 2002. Conservation biology for the biodiversity crisis: a freshwater follow-up. Conserv. Biol. 16: 1435–1437. doi:10. 1046/j.1523-1739.2002.01532.x. Boersma, P.D., Kareiva, P., Fagan, W.F., Clark, J.A., and Hoekstra, J.M. 2001. How good are endangered species recovery plans? Bioscience, 51: 643–649. doi:10.1641/0006-3568(2001) 051[0643:HGAESR]2.0.CO;2. Bohlin, T., Hamrin, S., Heggberget, T.G., Rasmussen, G., and Salt#

2008 NRC Canada

Poos et al. veit, S.J. 1989. Electrofishing — theory and practice with special emphasis on salmonids. Hydrobiologia, 173: 9–43. doi:10. 1007/BF00008596. Brunner, R.D., and Clark, T.W. 1997. A practice-based approach to ecosystem management. Conserv. Biol. 11: 48–58. doi:10.1046/ j.1523-1739.1997.96005.x. Campbell, S.P., Clark, J.A., Crampton, L.H., Guerry, A.D., Hatch, L.T., Hosseini, P.R., Lawler, J.J., and O’Connor, R.J. 2002. An assessment of monitoring efforts in endangered species recovery plans. Ecol. Appl. 12: 674–681. doi:10.1890/1051-0761(2002) 012[0674:AAOMEI]2.0.CO;2. Casselman, J.M. 2002. Effects of temperature, global extremes, and climate change on year-class production of warmwater, coolwater, and coldwater fishes in the Great Lakes basin. Am. Fish. Soc. Symp. 32: 39–60. Caughley, G. 1994. Directions in conservation biology. J. Anim. Ecol. 63: 215–244. doi:10.2307/5542. Clark, J.A., and Harvey, E. 2002. Assessing multi-species recovery plans under the Endangered Species Act. Ecol. Appl. 12: 655– 662. doi:10.1890/1051-0761(2002)012[0655:AMSRPU]2.0. CO;2. Clark, J.A., Hoekstra, J.M., Boersma, P.D., and Kareiva, P. 2002. Improving U.S. endangered species act recovery plans: key findings and recommendations of the SCB recovery plan project. Conserv. Biol. 16: 1510–1519. doi:10.1046/j.1523-1739.2002. 01376.x. Daniels, R.A. 1993. Habitat of the eastern sand darter, Ammocrypta pellucida. J. Freshw. Ecol. 8: 287–295. Dextrase, A.J., Staton, S.K., and Metcalfe-Smith, J.L. 2003. Recovery strategy for species at risk in the Sydenham River: an ecosystem approach. National Recovery Plan No. 25. Recovery of Nationally Endangered Wildlife (RENEW), Ottawa, Ont. Facey, D.E. 1998. The status of the eastern sand darter, Ammocrypta pellucida, in Vermont. Can. Field-Nat. 112: 596–601. Flather, C.H., Knowles, M.S., and Kendall, I.A. 1998. Threatened and endangered species geography. Bioscience, 48: 365–376. doi:10.2307/1313375. Foin, T.C., Riley, S.P.D., Pawley, A.L., Ayres, D.R., Carlsen, T.M., Hodum, P.J., and Switzer, P.V. 1998. Improving recovery planning for threatened and endangered species. Bioscience, 48: 177–184. doi:10.2307/1313263. France, R. 1997. Land–water linkages: influences of riparian deforestation on lake thermocline depth and possible consequences for cold stenotherms. Can. J. Fish. Aquat. Sci. 54: 1299–1305. doi:10.1139/cjfas-54-6-1299. Franklin, J.F. 1993. Preserving biodiversity: species, ecosystems, or landscapes? Ecol. Appl. 3: 202–205. doi:10.2307/1941820. Gerber, L.R., and Hatch, L.T. 2002. Are we recovering? An evaluation of recovery criteria under the U.S. Endangered Species Act. Ecol. Appl. 12: 668–673. doi:10.1890/1051-0761(2002) 012[0668:AWRAEO]2.0.CO;2. Goodchild, C.D. 1990. Status of the bigmouth buffalo, Ictiobus cyprinellus, in Canada. Can. Field-Nat. 104: 87–97. Greenberg, L.A. 1991. Habitat use and feeding behavior of thirteen species of benthic stream fishes. Environ. Biol. Fishes, 31: 389– 402. doi:10.1007/BF00002364. Grossman, G.D., and Freeman, M.C. 1987. Microhabitat use in a stream fish assemblage. J. Zool. (London), 212: 151–176. Grumbine, R.E. 1994. What is ecosystem management? Conserv. Biol. 8: 27–38. doi:10.1046/j.1523-1739.1994.08010027.x. Grumbine, R.E. 1997. Reflections on ‘‘What is ecosystem management?’’. Conserv. Biol. 11: 41–47. doi:10.1046/j.1523-1739. 1997.95479.x. Herkert, J.R. 1994. The effects of habitat fragmentation on mid-

2665 western grassland bird communities. Ecol. Appl. 4: 461–471. doi:10.2307/1941950. Holm, E., and Mandrak, N.E. 1996. The status of the eastern sand darter, Ammocrypta pellucida, in Canada. Can. Field-Nat. 110: 462–469. Jackson, D.A., and Harvey, H.H. 1993. Fish and benthic invertebrates: community concordance and community–environment relationships. Can. J. Fish. Aquat. Sci. 50: 2641–2651. doi:10. 1139/f93-260. Landres, P.B., Verner, J., and Thomas, J.W. 1988. Ecological uses of vertebrate indicator species: a critique. Conserv. Biol. 2: 316– 328. doi:10.1111/j.1523-1739.1988.tb00195.x. Lawler, J.J., Campbell, S.P., Guerry, A.D., Kolozsvary, M.B., O’Connor, R.J., and Seward, L.C.N. 2002. The scope and treatment of threats in endangered species recovery plans. Ecol. Appl. 12: 663–667. doi:10.1890/1051-0761(2002)012[0663:TSATOT]2.0. CO;2. Legendre, P., and Legendre, L. 1998. Numerical ecology. 2nd English ed. Elsevier Science BV, Amsterdam, the Netherlands. Leps, J., and Smilauer, P. 2004. Multivariate analysis of ecological data using CANOCO. Cambridge University Press, London, England. Lundquist, C.J., Diehl, J.M., Harvey, E., and Botsford, L.W. 2002. Factors affecting implementation of recovery plans. Ecol. Appl. 12: 713–718. doi:10.1890/1051-0761(2002)012[0713:FAIORP]2. 0.CO;2. Dextrase, A.J., and Mandrak, N.E. 2006. Impacts of alien invasive species on freshwater fauna at risk in Canada. Biol. Invasions, 8: 13–24. doi:10.1007/s10530-005-0232-2. Martin, T.G., Wintle, B.A., Rhodes, J.R., Kuhnert, P.M., Field, S.A., Low-Choy, S.J., Tyre, A.J., and Possingham, H.P. 2005. Zero tolerance ecology: improving ecological inference by modeling the source of zero observations. Ecol. Lett. 8: 1235–1246. doi:10.1111/j.1461-0248.2005.00826.x. McAllister, D.E. 1987. Status of the blackstripe topminnow, Fundulus notatus in Canada. Can. Field-Nat. 101: 219–225. McKee, P.M., and Parker, B.J. 1982. The distribution, biology, and status of the fishes Campostoma anomalum, Clinostomus elongates, Notropis photogenis (Cyprinidae), and Fundulus notatus (Cyprinodontidae) in Canada. Can. J. Zool. 60: 1347–1358. doi:10.1139/z82-182. Metcalfe-Smith, J.L., DiMaio, J., Station, S.K., and DeSolla, S.R. 2003. Status of the freshwater mussel communities of the Sydenham River, Ontario, Canada. Am. Midl. Nat. 150: 37–50. doi:10.1674/0003-0031(2003)150[0037:SOTFMC]2.0.CO;2. Miller, R.I., Williams, J.D., and Williams, J.E. 1989. Extinctions of North American fishes during the past century. Fisheries, 14: 22– 38. doi:10.1577/1548-8446(1989)014<0022:EONAFD>2.0.CO;2. OMNR. 2007. Ontario stream assessment protocol. Ontario Ministry of Natural Resources, Picton, Ontario. Available from www. mnr.gov.on.ca/226871.pdf [accessed 1 August 2007]. Page, L.M., and Burr, B.M. 1991. A field guide to freshwater fishes of North America north of Mexico. Houghton Mifflin Company, Boston, Mass. Pikitch, E.K., Santora, C., Babcock, E.A., Bakun, A., Bonfil, R., Conover, D.O., Dayton, P., Doukakis, P., Fluharty, D., Heneman, B., Houde, E.D., Link, J., Livingston, P.A., Mangel, M., McAllister, M.K., Pope, J., and Sainsbury, K.J. 2004. Ecosystembased fishery management. Science (Washington, D.C.), 305: 346–347. doi:10.1126/science.1098222. PMID:15256658. Poos, M.S., Mandrak, N.E., and McLaughlin, R.L. 2007. The effectiveness of two common sampling methods for assessing imperilled freshwater fishes. J. Fish Biol. 70: 691–708. doi:10.1111/j. 1095-8649.2007.01349.x. #

2008 NRC Canada

2666 Rahn, M.E., Doremus, H., and Diffendorfer, J. 2006. Species coverage in multispecies habitat conservation plans: Where’s the science? Bioscience, 56: 613–619. doi:10.1641/0006-3568(2006) 56[613:SCIMHC]2.0.CO;2. Ricciardi, A., and Rasmussen, J.B. 1999. Extinction rates of North American freshwater fauna. Conserv. Biol. 13: 1220–1222. doi:10.1046/j.1523-1739.1999.98380.x. Richter, B.D., Braun, D.P., Mendelson, M.A., and Master, L.L. 1997. Threats to imperilled freshwater fauna. Conserv. Biol. 11: 1081–1093. doi:10.1046/j.1523-1739.1997.96236.x. Scott, W.B., and Crossman, E.J. 1973. Freshwater fishes of Canada. Fish. Res. Board Can. Bull. 184. Simberloff, D. 1998. Flagships, umbrellas, and keystones: is singlespecies management passe´ in the landscape era? Biol. Conserv. 83: 247–257. doi:10.1016/S0006-3207(97)00081-5. Stanfield, L.W., and Jones, M.L. 1998. A comparison of full-station visual and transect-based methods of conducting habitat surveys in support of habitat suitability index models for southern Ontario. N. Am. J. Fish. Manage. 18: 657–675. doi:10.1577/15488675(1998)018<0657:ACOFSV>2.0.CO;2. Staton, S.K., Dextrase, A., Metcalfe-Smith, J.L., DiMaio, J., Nelson, M., Parish, J., and Holm, E. 2003. Status and trends of Ontario’s Sydenham River ecosystem in relation to aquatic species at risk. Environ. Monit. Assess. 88: 283–310. doi:10.1023/ A:1025529409422. PMID:14570419. Stewart, K.W., Suthers, I.M., and Leavesley, K. 1985. New fish

Can. J. Fish. Aquat. Sci. Vol. 65, 2008 distribution records in Manitoba, Canada and the role of a manmade interconnection between two drainages as an avenue of dispersal. Can. Field-Nat. 99: 317–326. Tear, T.H., Scott, J.M., Hayward, P.H., and Griffith, B. 1995. Recovery plans and the Endangered Species Act: are criticisms supported by data? Conserv. Biol. 9: 182–192. doi:10.1046/j. 1523-1739.1995.09010182.x. ter Braak, C.J.F., and Smilauer, P. 2002. CANOCO reference manual and CANODRAW for Windows user’s guide: software for canonical community ordination (Version 4.5). Microcomputer Power, Ithaca, New York. Trautman, M.B. 1957. The fishes of Ohio. Ohio State University Press, Columbus, Ohio. Williams, J.E., Johnson, J.E., Hendrickson, D.A., Contreras-Balderas, S., Williams, J.D., Navarro-Mendoza, M., McAllister, D.E., and Deacon, J.E. 1989. Fishes of North America endangered, threatened, or of special concern: 1989. Fisheries, 14: 2–20. doi:10.1577/1548-8446(1989)014<0002:FONAET>2.0.CO;2. Yaffee, S.L. 1999. Three faces of ecosystem management. Conserv. Biol. 13: 713–725. doi:10.1046/j.1523-1739.1999.98127.x. Yoder, C.O., and Smith, M.A. 1999. Using fish assemblages in a state of biological assessment and criteria program: essential concepts and considerations. In Assessing the sustainability and biological integrity of water resources using fish communities. Edited by T.P. Smith. CRC Press, Boca Raton, Fla. pp. 17–63.

#

2008 NRC Canada

species, community-, and ecosystem-based recovery ...

For the Sydenham River, an ecosystem-based recovery plan was deemed most appropri- .... decisions for species recovery are needed regardless of data.

819KB Sizes 0 Downloads 136 Views

Recommend Documents

Community Recovery in Hypergraphs - Changho Suh
significant attention due to its wide applicability to social network ... Model (CBM) is one of the most popular models in the .... whose capacity is 1−H(θ) [10].

Community Recovery in Hypergraphs - Changho Suh
arises in many fields of science and engineering. Among ... of this section, we relate our problem to a d-right-degree linear code, and ... community recovery in hypergraphs, and to the best of our ...... Theoretical Computer Science, vol. 411, no.

A practical framework for selecting among single- species, community ...
Abstract: Science-based approaches for selecting among single-species, community-, and ecosystem-based recovery plans are needed to conserve imperilled ...

Using bird species community occurrence to ... - Wiley Online Library
point counts in a 1560 km2 study area, remote-sensed data and models incorporating imperfect ... restore old-growth conditions and communities (Jönsson et al.

Estimating fishing mortality of major target species and species ... - frdc
Background. The volume of shark ..... channels (newspapers, fishing websites and newsletters) and word-of-mouth. Incentives including ...... images these should be outlined in this section outline and attach them where possible. Manuscript ...

Estimating fishing mortality of major target species and species ... - frdc
improve data quality; and the designation of a number species as either no-take or .... management of shark mortality needs to consider all interactions.

Course: E0210 Recovery from Disaster: A Local Community Role
Jul 12, 2016 - information on how to apply for EMI courses: http://training.fema.gov/Apply/. Tribal and voluntary organization representatives can submit their application (with supervisor's signature) directly to NETC Admissions Office. Your applica

Community and virtual community
arts and recreation; and the popular music group the Grateful Dead. The. WELL was created and .... hoods-to defining it in terms of social networks” (Wellman & Gulia,. 1999, p. 169) and pose a .... Page 10 ...... bombing campaign. Significant ...

Community and virtual community
Wellman and Gulia suggest that critics of virtual community often take as their starting point a .... has been open to some question” (Burnett, 2000, p. 1). ..... Yet, these systems in many ways replace the coffee pot and the water cooler as the ..

Reactive oxygen species, Aging and Antioxidative Nutraceuticals.pdf ...
There was a problem previewing this document. Retrying... Download. Connect more apps... Try one of the apps below to open or edit this item. Reactive oxygen ...

Reactive oxygen species, Aging and Antioxidative Nutraceuticals.pdf
Reactive oxygen species, Aging and Antioxidative Nutraceuticals.pdf. Reactive oxygen species, Aging and Antioxidative Nutraceuticals.pdf. Open. Extract.

Automatic Bird Species Identification for Large Number of Species
is important to obtain reliable information about the popu- lation of wild animals. .... In our digital era, the analog signal is sampled, several times per second, and ...

WRA Species Report
zones] "It can tolerate cold winter temperatures but will die off to the ground after the first frost of the season. ... It is typically grown in U. S. gardens as a cool weather annual or biennial." 204. 2011. Starr, F./Starr, K.. .... America, espec

The Companion Species Manifesto:
the tropic work required for ontological choreography .... gious activity in the newspaper business. ..... are the masters or the duped can hang on the outcome.

Species pubs.pdf
“A New Flora of Fiji” Albert C. Smith,. Austrobalia 3 (1991). Variation in Hoya ... Species pubs.pdf. Species pubs.pdf. Open. Extract. Open with. Sign In. Details.

WRA Species Report
e+armeria&o=plants. [Naturalized beyond native range? Potentially Maui. Apparently able to colonize higher elevation sites in tropical and subtropical regions] "Olinda, Hawea Pl., along unmaintained portion of gravel road. One plant. Adventive. Uplan

WRA Species Report
CAB International. Forestry Compendium. CAB International, Wallingford, UK. [Quality of climate match data? 2-high] "Natural distribution of I. edulis ranges from Colombia and Venezuela to north-western Argentina, and from the Andean foothills to Atl

Species pubs.pdf
Page 1 of 50. 1. Species Publications. Hoya acuminata Bentham ex Hooker f.. Decaisne DeCandolle Prodromus 8 (1844) 633. Robert Wight. Handbook of Indian Flora 2 (1866) 240. R. Wight. Flora of British India 4 (1883) 53. Sir Joseph Dalton Hooker. Hoya