Spawning Site Selection in Interior Fraser River Coho Salmon Oncorhynchus Kisutch: an Imperiled Population of Anadromous Salmon from a Snow-Dominated Watershed
Total Page:16
File Type:pdf, Size:1020Kb
Vol. 16: 249–260, 2012 ENDANGERED SPECIES RESEARCH Published online March 22 doi: 10.3354/esr00401 Endang Species Res Contribution to the Theme Section ‘Endangered river fish: threats and conservation options’ OPENPEN ACCESSCCESS Spawning site selection in interior Fraser River coho salmon Oncorhynchus kisutch: an imperiled population of anadromous salmon from a snow-dominated watershed Crystal J. McRae1,2,*, Kyla D. Warren1,3, J. Mark Shrimpton1 1Ecosystem Science and Management (Biology) Program, University of Northern British Columbia, Prince George, British Columbia V2N 4Z9, Canada 2Present Address: Committee for General Education, National Dong Hwa University, Shoufeng, Hualien, 97401 Taiwan 3Present address: Rescan Environmental Services Ltd., Smithers, British Columbia V0J 2N0, Canada ABSTRACT: Understanding habitat requirements for spawning is important for species at risk of extirpation to define areas for protection of this critical life stage. Interior Fraser coho salmon (IFC) Oncorhynchus kisutch spawn in tributaries of the upper Fraser River, British Columbia, Canada. They are regionally isolated, genetically distinct, and have been listed as endangered. Recent enumeration efforts have shown declining escapement and range contraction. IFC spawn in snow-dominated watersheds, yet much of our understanding of spawner habitat requirements is based on information for coho from coastal rain-dominated systems. Consequently, we examined the relative influence of habitat features on spawning site selection. Our sampling focused on why fish chose particular locations and not others in what seemed to be suitable habitat. We defined several logistic regression models, including various combinations of physical, chemical and hydrological microhabitat features, and used an information-theoretic approach to evaluate the relative plausibility of these models. Models combining intragravel measures of electrical conduc- tance, dissolved oxygen, temperature and specific discharge were the best models associated with spawning microhabitat selection, and jackknife validations showed that these models had good predictive abilities. Intragravel variables, therefore, appear to play a prominent role in spawning site selection for IFC. Understanding the influence of the intragravel environment on spawning site selection will help identify habitat requirements as well as potential threats to fish populations and will contribute to the development of comprehensive conservation initiatives. KEY WORDS: Interior Fraser coho · Spawning · Site selection · Intragravel habitat Resale or republication not permitted without written consent of the publisher INTRODUCTION freshwater habitat is frequently cited as a reason for declines in salmonid populations (Nehlsen et al. Selection of a suitable site for spawning is crucial 1991, COSEWIC 2002). As such, the evaluation of because the highest rates of mortality in salmonids specific spawning site features may provide vital generally occur during the incubation period (Quinn information for conservation and management initia- 2005), and this mortality is closely related to the fea- tives and add to the understanding of species- tures of the spawning/incubation site. Changes in specific habitat needs. habitat features may also be deleterious to larval Studies of spawning habitat requirements in development and survival. In fact, perturbation of salmonids have tended to focus on stream depth, ve- *Email: [email protected] © Inter-Research 2012 · www.int-res.com 250 Endang Species Res 16: 249–260, 2012 locity and physical properties of the channel, such as ter & Hauer 2000). Describing spawning site substrate size (Bovee 1982, Louhi et al. 2008). Others selection solely in terms of upwelling and down- have found that not all seemingly suitable habitat for welling may provide some insight into habitat selec- spawning is used (Mull & Wilzbach 2007). In rivers tion cues, but this likely summarizes the driving with altered hydrographs because of hydroelectric forces behind site selection too simplistically. Addi- power generation, the inclusion of stream flow fluctu- tional physical and chemical features of hypor heic ation has increased the predictive ability of models water are important for understanding the cues af- (Geist et al. 2008), but apparently suitable sites are fecting spawning site selection, as well as the source still not used for spawning. Traditional habitat fea- of upwelling water (Alexander & Caissie 2003). tures, therefore, may not provide the most effective Our study examined habitat features that influence means of assessing spawning habitat suitability (Geist spawning site selection for interior Fraser coho (IFC) & Dauble 1998, McHugh & Budy 2004); rather, the salmon Oncorhynchus kisutch. IFC are reproduc- complexity of spawning site selection may encompass tively isolated from lower Fraser River coho (Irvine a wider range of habitat features. Surface features 2004) and because of this are genetically distinct such as slope are associated with the potential for from coho in the lower Fraser River (Small et al. streambed scouring (DeVries 1997). Habitat features 1998). Few studies have examined spawning habitat such as overhanging vegetation, woody debris and features of interior watersheds with snow-dominated undercut banks have also been found to influence hydrological regimes, yet many populations that spawning site selection (McMahon et al. 1996). Cover spawn in these regions are of conservation concern. features have the potential to provide protection from The Committee on the Status of Endangered Wildlife predators as well as adverse stream conditions, such in Canada (COSEWIC) concluded ‘that there is a as high stream velocity. Additionally, there may be a serious risk of extinction of Interior Fraser coho’ behavioral component to spawning site selection, as a (COSEWIC 2002, p. iii), and thus designated IFC as behavioral preference for spawning on existing redds an endangered species in 2002. With this designa- even when suitable spawning habitat was not limited tion, IFC were considered for legal protection under has been shown (Essington et al. 1998). Conversely, the Species at Risk Act (SARA); however, based on spawning clusters and redd superimposition may be uncertainties in the marine environment and poten- more effectively explained by interactions between tial socio-economic impacts, listing under SARA was groundwater (phreatic) and surface water (stream) denied (Canada Gazette 2006). For IFC, it has been that occur in the hyporheic zone (Valett et al. 1993, suggested that loss of freshwater habitat is a primary Geist et al. 2002). factor driving current population declines (Bradford Phreatic groundwater originates deep in the sub- & Irvine 2000). Freshwater habitat requirements for surface and, because of long residence times, con- IFC are unknown. Studies on the habitat require- tains a significant component of dissolved solutes ments of lower Fraser River and coastal coho popula- (Freeze & Cherry 1979), is devoid of oxygen (Freeze tions are comparatively abundant, but may not be & Cherry 1979) and shows little seasonal change in applicable to IFC populations. It is likely that IFC life temperature (Baxter & Hauer 2000). Similarity of wa- history strategies differ from those of coastal coho ter in the hyporheic zone to either phreatic water or populations, as interior populations evolved with stream water is mainly a function of the penetration substantially different hydrological regimes: interior depth and residence time of the hyporheic flow path- watershed runoff is primarily snow-dominated, while way. Stream water infiltrates the streambed (down- coastal watershed runoff is rain-dominated. Numer- welling), flows through the subsurface and later ous features have the potential to influence where emerges (upwelling) back into the stream (Boulton fish spawn and it is likely that IFC select spawning 1993). On a small scale, upwelling and downwelling sites based on particular habitat features. If known, are primarily governed by stream discharge, stream - these habitat features could be used to predict the bed permeability and channel morphology (Boulton probability of site use and such locations could be 1993). Upwelling of hyporheic water has been linked protected. To characterize habitat features important to salmonid spawning sites in chinook salmon Oncor - for spawning, we measured the physical and chemi- hynchus tschawytscha (Geist 2000) and rainbow cal features of IFC spawning sites and areas not used trout O. mykiss (Sowden & Power 1985), but down- for spawning, fit the data to a set of plausible candi- welling has also been linked to spawning sites in in- date models by logistic regression analysis, and iden- terior snow-dominated watersheds for bull trout tified the most parsimonious model(s) using an infor- Salvelinus confluentus (Baxter & McPhail 1999, Bax- mation-theoretic approach. Work was conducted in a McRae et al.: Spawning site selection in anadromous salmon 251 watershed used as an indicator population for IFC cific spawning sites (used sites) were defined by the spawner escapement. presence of a redd. For sample sites where no spawn- ing occurred (unused sites), we chose locations where available habitats were within the published MATERIALS AND METHODS limits of spawning for depth, substrate and velocity. Unused sites, therefore, appeared appropriate for Study site spawning based on habitat features of coho salmon redds in McKinley Creek, but no