Skip to content
Please be aware the content below has been generated by an AI model from a source PDF.

Appendix 2 – Spey catchment: beaver feasibility and potential release site assessment

Spey Catch­ment: Beaver Feas­ib­il­ity and Poten­tial Release Site Assessment

Octo­ber 2023

Pre­pared as part of a wider feas­ib­il­ity study into the poten­tial return of beavers to Cairngorms Nation­al Park by:

Dr Róisín Camp­bell-Palmer, Dr Alan Put­tock, Dr Robert Need­ham and Prof Richard Brazier

*cov­er photo, main chan­nel of River Spey, south of Blar­gie © Alan Puttock

Con­tents Intro­duc­tion and Aims 3 Mod­el­ling of Beaver Hab­it­at Suit­ab­il­ity Spey Catch­ment 4 Beaver Hab­it­at Suit­ab­il­ity Mod­el­ling 4 Beaver Veget­a­tion Index (BVI ‑pre­requis­ite for BHI mod­el­ling) 5 Beaver Hab­it­at Index mod­el (BHI) 5 Beaver Hab­it­at Index maps and sum­mary stat­ist­ics for study area 6 Beaver Dam Capa­city Mod­el­ling of Spey Catch­ment 8 Beaver Dam Capa­city (BDC) mod­el sum­mary 8 Beaver Dam Capa­city Mod­el maps and sum­mary stat­ist­ics for the study area 9 Beaver hab­it­at and dam capa­city mod­el sum­mary 10 Aspen Map­ping 11 Beavers and Fish in the Spey 14 Poten­tial Release Site Assess­ments 22 Site Map­ping Sum­mar­ies: 23 Insh Marshes 23 31 32 33 34 35 36 Uath Lochans 36 41 42 44 Key recom­men­ded release sites, con­nectiv­ity, and dis­pers­al 45 Con­clu­sions and Next Steps 46 Ref­er­ences 47 Appendix 1. Data­sets used 51 Appendix 2. Caveats for use 52 Beaver veget­a­tion and hab­it­at index 52 Beaver dam Capa­city Mod­el 53

Intro­duc­tion and Aims This report was com­mis­sioned by the Cairngorms Nation­al Park Author­ity (CNP) to invest­ig­ate the hab­it­at feas­ib­il­ity of the Spey catch­ment to sup­port the res­tor­a­tion of Euras­i­an beavers (Castor fiber). Poten­tial release sites for first releases (mid-River Spey area only) have been iden­ti­fied, pri­or­it­ising landown­ers that may sup­port any release licence applic­a­tion. Oth­er sites have been invest­ig­ated and though they may not be pro­posed as release sites, could sup­port beaver col­on­isa­tion as any pop­u­la­tion expands naturally.

Fol­low­ing ini­tial feas­ib­il­ity work and the sub­sequent decision to pro­ceed with bring­ing beavers back to the Cairngorms, the Park Author­ity are keen to under­take fur­ther assess­ment work to inform poten­tial release site selec­tion, likely pop­u­la­tion dynam­ics, dis­pers­al routes and areas of likely future col­on­isa­tion, along­side fur­ther assess­ment of poten­tial impacts (pos­it­ive or neg­at­ive). This report addresses some of these aims by build­ing on an ini­tial feas­ib­il­ity study, on ground exper­i­ence and know­ledge com­bined with the deploy­ment of mod­els developed at the Uni­ver­sity of Exeter, and many years’ exper­i­ence by these authors of beaver res­tor­a­tion dynam­ics. We present hab­it­at suit­ab­il­ity and beaver dam capa­city in sup­port of under­stand­ing the dis­tri­bu­tion of likely beaver pop­u­la­tions in the future at the land­scape scale across the Spey catchment.

The hab­it­at suit­ab­il­ity and the capa­city for beavers to dam chan­nels with­in the study areas was assessed using beaver mod­el­ling tools developed by research­ers at the Uni­ver­sity of Exeter (Gra­ham et al., 2020). These mod­el­ling tools con­sist of a Beaver Hab­it­at Index (BHI) mod­el and a Beaver Dam capa­city (BDC) model.

There is a require­ment to com­plete an ana­lys­is of river catch­ments to assess their suit­ab­il­ity for sup­port­ing pop­u­la­tions of beaver. Beaver hab­it­at suit­ab­il­ity is determ­ined primar­ily by veget­a­tion suit­ab­il­ity which has been clas­si­fied nation­ally using a Beaver Veget­a­tion Index (BVI) as well as access to water bod­ies. Togeth­er these two factors have been incor­por­ated into a Beaver hab­it­at Index mod­el (BHI). BHI has been run nation­ally to devel­op a high res­ol­u­tion (5m) con­tinu­ous ras­ter product that can inform loc­al decision mak­ing with regard to beaver rein­tro­duc­tion. BHI clas­si­fies hab­it­at suit­ab­il­ity from 0 (No access to veget­a­tion — not suit­able) to 5 (Highly Suit­able). It must be cla­ri­fied that this hab­it­at mod­el is a sim­pli­fied rep­res­ent­a­tion of real­ity and caveats asso­ci­ated with the mod­el are lis­ted in appendices.

Beavers are also well known as eco­sys­tem engin­eers, hav­ing the capa­city to change envir­on­ments to suit their needs. The beaver engin­eer­ing activ­ity that has the greatest capa­city to modi­fy eco­sys­tems is dam build­ing. Dam build­ing and the cre­ation of pon­ded sur­face water has the abil­ity to bring bene­fits (i.e., for biod­iversity, water stor­age, flow atten­u­ation) but also poten­tially man­age­ment and con­flict (i.e., loc­al­ised inund­a­tion of land, block­ing of crit­ic­al infra­struc­ture). BDC clas­si­fies reaches from no capa­city for dam build­ing to a per­vas­ive capa­city for damming.

Mod­el­ling of Beaver Hab­it­at Suit­ab­il­ity Spey Catch­ment Beaver Hab­it­at Suit­ab­il­ity Mod­el­ling Sum­mary Descrip­tion: Pro­duc­tion of a con­tinu­ous descrip­tion of hab­it­at suit­ab­il­ity for beaver. First a veget­a­tion suit­ab­il­ity index is cre­ated using mul­tiple high-res­ol­u­tion spa­tial data­sets from Ord­nance Sur­vey, CEH and Coper­ni­cus will be com­bined to provide detailed land cover/​vegetation inform­a­tion which is clas­si­fied based on empir­ic­al field obser­va­tion of beaver hab­it­at and pref­er­ence. Veget­a­tion suit­ab­il­ity is com­bined with addi­tion­al para­met­ers describ­ing stream net­works and water bod­ies. Whilst beaver hab­it­at suit­ab­il­ity is primar­ily defined by veget­a­tion suit­ab­il­ity, beavers also require water for secur­ity and move­ment. There­fore, access­ib­il­ity to water bod­ies (i.e. chan­nels, ponds, and lakes) will

also determ­ine the viab­il­ity of beaver occu­pancy and there­fore are required to clas­si­fy hab­it­at accurately.

Out­puts: This product provides a high-res­ol­u­tion (5m cell size) resource (ras­ter .tiff format) for describ­ing hab­it­at suit­ab­il­ity for beaver. This data­set can allow the user to explore which land­scapes were most (or least) suite to beaver rein­tro­duc­tion and also to under­stand where hab­it­at enhance­ment might be use­ful to sup­port future reintroduction.

Beaver Veget­a­tion Index (BVI ‑pre­requis­ite for BHI mod­el­ling) Veget­a­tion is import­ant for clas­si­fy­ing beaver hab­it­at (Hart­man, 1996; John et al., 2010; Pinto et al., 2009; St-Pierre et al., 2017). It was there­fore crit­ic­al to estab­lish a reli­able Beaver Veget­a­tion Index (BVI) using nation­ally-avail­able spa­tial data­sets. No single data­set con­tained the detail required to depict all key veget­a­tion types. There­fore, a com­pos­ite data­set was cre­ated from: OS Vec­torMap data (Ord­nance Sur­vey, 2018), The Centre for Eco­logy and Hydro­logy (CEH) 2015 land cov­er map (LCM) (Row­land et al., 2017), Coper­ni­cus 2015 20 m Tree Cov­er Dens­ity (TCD) (Coper­ni­cus, 2017) and the CEH woody lin­ear fea­tures frame­work (Schol­e­field et al., 2016).

Veget­a­tion data­sets were assigned suit­ab­il­ity val­ues (zero to five). Zero val­ues were assigned to areas of no veget­a­tion i.e. build­ings and val­ues of five were assigned to favour­able hab­it­at i.e. decidu­ous wood­land. Val­ues were assigned based on a review of rel­ev­ant lit­er­at­ure (Haarberg and Rosell, 2006; Jen­kins, 1979; Nolet et al., 1994; O’Connell et al., 2008), field obser­va­tion and com­par­is­on with satel­lite imagery. Vec­tor data were con­ver­ted to ras­ter format (res­ol­u­tion of 5 m). TCD data were res­ampled to 5m and aligned with con­ver­ted vec­tor lay­ers. An infer­ence sys­tem was used to com­bine these four ras­ter data­sets to cre­ate the BVI. The work­flow pri­or­it­ises the reli­ab­il­ity fol­lowed by the highest value data.

Examples of highly suit­able land (graded 5) include broad-leaf wood­land, mixed wood­land and shrub; examples of suit­able veget­a­tion (graded 4) include shrub and marsh; examples of mod­er­ately suit­able (graded 3) include con­i­fer­ous wood­land, marsh, shrub and unim­proved grass­land; examples of barely suit­able (graded 2) include reeds, shrub and heath­land and boulders, neut­ral grass­land; examples of unsuit­able (graded 1) include heath­er, acid grass­land, unim­proved grass and boulders, bog; examples of no access­ible veget­a­tion (graded 0) include shingle and sand, build­ings, rock, urb­an and saltwater.

Beaver Hab­it­at Index mod­el (BHI) Whilst veget­a­tion is a dom­in­ant factor in determ­in­ing hab­it­at suit­ab­il­ity for beaver, so is prox­im­ity to a water body (Gurnell et al., 2008), with beavers being strong swim­mers, using water bod­ies both to provide secur­ity, as a means of escap­ing pred­at­ors and to access for­aging areas. It is thought that most for­aging occurs 10 m of a watercourse/​body (Haarberg and Rosell, 2006), and rarely above 50 m (Stringer et al 2018). How­ever, great­er for­aging dis­tances have on occa­sion been observed and as in Mac­far­lane et al., 2015 100 m has been accep­ted as a max­im­um dis­tance in which the vast major­ity of for­aging occurs. There­fore, to determ­ine suit­able hab­it­at for beaver incor­por­at­ing both BVI veget­a­tion suit­ab­il­ity and water access­ib­il­ity a 100m buf­fer was applied to water bod­ies. To do this the OS mas­termap river net­work and OS vec­tor in land water bod­ies were com­bined to get the best read­ily avail­able nation­al water­body and water course coverage.

Whilst BVI was run nation­ally on a 5 m scale it is best viewed as a pre­par­at­ory step for BHI (and later BDC) mod­el­ling and is super­seded in use­ful­ness by the BHI data­set. It is strongly recom­men­ded that most ana­lys­is and man­age­ment applic­a­tions such as this study use BHI i.e. if there is an area of preferred

veget­a­tion such as wil­low wood­land, more than 100 m from a water­body it is thought inac­cess­ible to beaver and there­fore does not form suit­able habitat.

Both BVI and BHI use a scor­ing sys­tem of zero to five (Table 1). Scores of five rep­res­ent veget­a­tion that is highly suit­able or pre­ferred by beavers and that also lies with­in 100 m of a water­body. Zero scores are giv­en to areas that con­tain no veget­a­tion or are great­er than 100 m from a water­body. It is import­ant to note that the hab­it­at mod­el con­siders ter­restri­al hab­it­at where for­aging primar­ily occurs and that water­courses them­selves are also scored zero. It is also import­ant to note that all scores above 1 con­tain suit­able vegetation.

In addi­tion to the ras­ter lay­er, BHI val­ues are asso­ci­ated with the reach scale Beaver Net­work river lay­er as BFI (Beaver For­age Index). Reach BFI val­ues were obtained for two search areas, 10 m (stream­side) and 40 m (ripari­an) from the bank edge. Both search areas extend 100 m up and down­stream to account for con­nectiv­ity of reaches. The mean of the top 50% of BFI val­ues in each search area was extrac­ted to under­stand the suit­ab­il­ity of the best avail­able hab­it­at with­in a giv­en reach.

Table 1. BVI and BHI value defin­i­tions. It is crit­ic­al to note that all val­ues above 1 are suit­able for beaver. BFI and BHI Val­ues Defin­i­tion 0 Not suit­able (no access­ible veget­a­tion) 1 Likely Unsuit­able (unsuit­able veget­a­tion) 2 Low/​Barely Suit­able 3 Mod­er­ately Suit­able 4 High/​Suitable 5 Preferred/​Highly Suitable

Beaver Hab­it­at Index maps and sum­mary stat­ist­ics for study area Table 2. dis­plays the sum­mary stat­ist­ics (length and %) of gross hab­it­at cat­egory types across the water­courses of the Spey catch­ment (See Fig­ure 2).

Table 2. Sum­mary hab­it­at map­ping stat­ist­ics for the Spey catch­ment Hab­it­at Cat­egory Length (km) % in each cat­egory Likely Unsuit­able (1) 2989.7 43.4 Low (2) 950.7 13.8 Mod­er­ate (3) 749.3 10.9 High (4) 997.3 14.5 Pre­ferred (5) 1203.4 17.5

Beaver Dam Capa­city Mod­el­ling of Spey Catch­ment Beaver Dam Capa­city (BDC) mod­el sum­mary The Beaver res­tor­a­tion assess­ment tool (BRAT) was developed in North Amer­ica (Mac­far­lane et al., 2014, 2015) to determ­ine the capa­city for river sys­tems to sup­port Beaver dams. The BRAT mod­el has been fur­ther deployed in a range of dif­fer­ent river sys­tems to aid both Beaver recol­on­isa­tion and beaver dam ana­logue led res­tor­a­tion. The BRAT mod­el not only provides an invalu­able tool for design­ing effect­ive, empir­ic­ally based, res­tor­a­tion strategies but it also indic­ates where Beaver dams might be con­struc­ted and there­fore where they may cause poten­tial management/​conflict issues. The BRAT mod­el struc­tures the frame­work of the mod­el around the river net­work itself and using a fuzzy logic approach which builds in the con­sid­er­able uncer­tainty that is asso­ci­ated with beaver habitat/​dammable reaches. Fur­ther­more, it provides a range of out­put val­ues to pre­dict the dam capa­city which has implic­a­tions for beaver pref­er­ence towards a giv­en location.

We have there­fore used the BRAT frame­work to devel­op an optim­ised beaver dam capa­city (BDC) mod­el for Great Bri­tain; and although many of the data­sets used are spe­cif­ic to GB, these could read­ily be adap­ted to enable its use globally.

The BDC mod­el estim­ates the capa­city of river sys­tems to sup­port dams at the reach-scale (ca. 150 m). The mod­el also high­lights reaches that are more likely to be dammed by beaver and estim­ates the num­ber of beaver dams that could occur for a catch­ment at pop­u­la­tion car­ry­ing capa­city. As such, this highly detailed tool would provide under­stand­ing of where dams are most likely to occur and in what dens­it­ies, sup­port­ing future work on the con­flicts and oppor­tun­it­ies that might accrue from beaver reintroduction.

The mod­el infers the dens­ity of dams that can be sup­por­ted by stream reaches (111.1 m ± 52.5) across a catch­ment. Using low-cost and open-source data­sets, the fol­low­ing attrib­utes are cal­cu­lated for each reach: (i) stream gradi­ent, (ii) low (Q80) and high flow (Q2) stream power, (iii) bank­full width, (iv) stream order, and (v) the suit­ab­il­ity of veget­a­tion, with­in 10m and 40 m of the bank, for beaver dam con­struc­tion. These con­trolling vari­ables are com­bined using a sequence of infer­ence and fuzzy infer­ence sys­tems which fol­low an expert-defined rules sys­tem that allows for the con­sid­er­able uncer­tainty often asso­ci­ated with these types of com­plex eco­lo­gic­al processes.

Each reach was clas­si­fied for dam­ming capa­city using five cat­egor­ies from none, defined as no capa­city for dam­ming to per­vas­ive where a max­im­um capa­city of 16 – 30 dams could the­or­et­ic­ally be con­struc­ted in a km of chan­nel. It is import­ant to note that the mod­el assumes both reach and catch­ment pop­u­la­tion car­ry­ing capa­city for beaver. There­fore, in real­ity the max­im­um num­ber of dams indic­ated in a cat­egory class is unlikely to occur. A full list of BDC clas­si­fic­a­tions is included in Table 3.

Table 3. BDC clas­si­fic­a­tions and defin­i­tions. BDC Clas­si­fic­a­tion Defin­i­tion None No capa­city for dam­ming Rare Max capa­city for 0 – 1 dams/​km Occa­sion­al Max capa­city for 1 – 4 dams/​km Fre­quent Max capa­city for 5 – 15 dams/​km

Per­vas­ive Max capa­city for 16 – 30dams/​km

Beaver Dam Capa­city Mod­el maps and sum­mary stat­ist­ics for the study area Table 4. dis­plays the sum­mary stat­ist­ics (length and %) of gross dam capa­city cat­egory types across the water­courses of the Spey catch­ment (See Fig­ure 3).

Table 4. Beaver dam capa­city sum­mary stat­ist­ics for the Spey catch­ment Dam capa­city cat­egory Length (km) % in each cat­egory None 936.0 13.6 Rare 3418.8 49.6 Occa­sion­al 1326.1 19.2 Fre­quent 468.1 6.8 Per­vas­ive 741.4 10.8

Beaver hab­it­at and dam capa­city mod­el sum­mary The mod­el res­ults presen­ted herein, illus­trate that through­out the Spey catch­ment, includ­ing with­in Nation­al Park, there are extens­ive areas of highly suit­able hab­it­at to sup­port beaver pop­u­la­tions. Addi­tion­ally, there are many smal­ler reaches, with good hab­it­at and suit­able hydro­lo­gic­al con­di­tions where beavers could cre­ate dams, par­tic­u­larly in the more low­land areas. How­ever, mod­el res­ults also show the main Spey and trib­u­tar­ies to be too large and power­ful for beavers to dam. Sim­il­arly, many of the upland areas, par­tic­u­larly those with­in the NP lack suit­able hab­it­at and are also too steep to sup­port beaver dam­ming. These mod­el out­puts show the spa­tial vari­ab­il­ity in impact that could occur if beavers returned to being wide­spread both with­in the NP and the wider Spey catchment.

Com­bined with oth­er com­pon­ents of feas­ib­il­ity work being under­taken, these mod­el res­ults will provide a geo­spa­tial basis for inform­ing future impacts (both pos­it­ive and neg­at­ive) that the rein­tro­duc­tion of

beavers could bring. Used stra­tegic­ally it is hoped that such data products can help max­im­ise the bene­fits and min­im­ise the con­flicts asso­ci­ated with beaver.

Whilst use­ful, as with any mod­el out­put, there are lim­it­a­tions and uncer­tain­ties (see Appendix 2 for use caveats) which need to be con­sidered. These mod­el res­ults were ground-truthed dur­ing field vis­its to com­bine mod­el out­puts with expert inter­pret­a­tion to reach a con­clu­sion on the suit­ab­il­ity of the site will for beaver. Field based assess­ment will also con­sider the poten­tial for beavers to bring pos­it­ive impacts as well as the poten­tial for man­age­ment issues to arise and poten­tial solu­tions to these.

Aspen Map­ping CNP have iden­ti­fied inter­ac­tions between beaver and aspen (Pop­u­lus spp.) as a poten­tial con­cern flagged by oth­er con­ser­va­tion groups and so have provided maps of known areas of aspen across CNP sec­tion of the Spey catch­ment. Aspen is a pre­ferred for­age spe­cies for beavers (Fryxell and Doucet 1993, Nolet et al. 1994) and espe­cially if there is low avail­ab­il­ity of food nearby beavers will travel fur­ther from the water course to obtain this. Where it grows close to suit­able water bod­ies it would most likely have the poten­tial to be impacted. Stud­ies have repor­ted that beavers typ­ic­ally only fell aspen in leaf (Doucet et al. 1994, Wils­son 1971). It is crit­ic­al to note that aspen read­ily suck­ers in response to beaver for­aging, which increases its pro­ductiv­ity, and beavers will often then for­age on regrowth. Deer and live­stock will pref­er­en­tially feed on broadleaf new growth, so provid­ing graz­ing pres­sure is not intens­ive new growth aspen will get away. The great­er con­cern with beaver and aspen over­lap, is beaver felling of mature trees and related impacts on asso­ci­ated bry­ophyte, lichen and inver­teb­rate communities.

If par­tic­u­lar trees, or more likely stands of mature aspen, are flagged as of interest, tree pro­tec­tion man­age­ment approaches are read­ily avail­able. To under­stand where known Aspen stands may have the poten­tial to be impacted by beaver, a 10 m buf­fer was applied to all Aspen points and Poly­gons. These data points rep­res­ent small, dis­crete groups of aspen <0.04 ha (2020 m) in extent, plus indi­vidu­al aspen trees. The poly­gons are lar­ger stands/​areas of aspen-rich wood­land ≥ 0.04 ha in extent.

From Fig­ure 8 and 9 below it is evid­ent that a large pro­por­tion of recor­ded Aspen areas exist with­in areas con­sidered poten­tial beaver hab­it­at with­in the Spey. Out of 533 poly­gons of lar­ger Aspen stands, 340 (or 64%) fall with­in beaver hab­it­at. For smal­ler stands rep­res­en­ted by point data, 229 of 575 points (40%) fall with­in beaver hab­it­at. It is crit­ic­al to note that this is a gen­er­ous over­es­tim­a­tion of over­lap with beaver hab­it­at assum­ing a max­im­um for­aging range of 100 m from water courses, far great­er than that which typ­ic­ally occurs. This ana­lys­is is presen­ted as a first step in identi­fy­ing known aspen stands that may over­lap with beaver activ­ity; how­ever, loc­al man­age­ment and mon­it­or­ing work is likely to be required to assess the ongo­ing con­ser­va­tion import­ance and any asso­ci­ated risks.

Beavers and Fish in the Spey CNP have also iden­ti­fied inter­ac­tions between beaver and sal­monids as a poten­tial con­cern. To sup­port this feas­ib­il­ity pro­ject, loc­a­tions of sal­mon and trout sampling points have been provided by CNP and the Spey Fish­ery Board. The sampling points provided are mapped in Fig­ure 10, along with aver­age dens­ity data in Fig­ure 11 and 12. Fig­ure 13 and 14 over­lay cur­rent fish sampling points onto the beaver hab­it­at and dam capa­city lay­ers. It is crit­ic­al to note that the mapped data for fish only per­tains to these data­sets and addi­tion­al fish data hos­ted by dif­fer­ent organ­isa­tions may be avail­able. For fur­ther inform­a­tion on the trout and sal­mon data shown it is recom­men­ded con­tact­ing the Spey Fish­ery board.

The key bene­fits of beaver activ­ity for sal­monids that are com­monly cited include increased hab­it­at het­ero­gen­eity (Häg­glund and Sjöberg, 1999; Smith and Math­er, 2013) and qual­ity (Pol­lock et al., 2003). In par­tic­u­lar, ponds cre­ated upstream of beaver dams provide juven­ile over­win­ter­ing and rear­ing hab­it­at (Cun­jak, 1996; Need­ham et al., 2021), and can be a crit­ic­al refuge for lar­ger fish (Häg­glund and Sjöberg, 1999; Need­ham et al., 2021). The bene­fi­cial response from a fish­er­ies per­spect­ive is usu­ally quan­ti­fied in terms of increased fish abund­ance (Häg­glund and Sjöberg, 1999; Jakober et al., 1998; Need­ham et al., 2021), con­di­tion and growth (Sigour­ney et al., 2006; but see Rabe, 1970, and John­son et al., 1992; Need­ham et al., 2021), and over­all pro­ductiv­ity (Mitchell & Cun­jak, 2007; Nick­el­son et al., 1992; Pol­lock et al., 2004). Con­versely, the prin­cip­al neg­at­ive con­sequence of beaver activ­ity often cited is the poten­tial for dams to impede or delay sal­monid migra­tion, par­tic­u­larly for upstream mov­ing adults dur­ing their migra­tion to the spawn­ing grounds (Lok­teff et al., 2013; Rupp, 1955; Taylor et al., 2010). Fur­ther­more, dams may reduce the avail­ab­il­ity of suit­able spawn­ing hab­it­at in impoun­ded areas, where there may be insuf­fi­cient flow velo­city to purge the gravels, which sal­monids use for spawn­ing and egg incub­a­tion, of the fine sed­i­ments depos­ited (Knud­sen, 1962; Taylor et al., 2010). Mal­is­on and Hal­ley (2020), how­ever, found that beaver dams did not block the move­ment of juven­ile sal­monids or their abil­ity to use upstream hab­it­ats and sug­gest that it is unlikely that dams neg­at­ively impact the juven­ile stage of sal­mon or trout pop­u­la­tions. Kemp et al. (2012) reviewed 108 stud­ies of beaver and fish. Dams were cited as bar­ri­ers to fish move­ment” in 43% of papers and was the most com­mon adverse effect dis­cussed. How­ever, these neg­at­ive effects were spec­u­lat­ive at best in that 78% of the stud­ies did not sup­port this claim with data. Fur­ther work is required to estab­lish actu­al impacts of beaver dams on fish pas­sage, but by cross ref­er­en­cing the BDC mod­els with valu­able sal­monid hab­it­at will help identi­fy key areas of con­cern and alle­vi­ate pos­sible impacts.

As described above, it is likely there will only be sig­ni­fic­ant con­cerns where pri­or­ity spawn­ing grounds over­lap with areas where beavers are present and there are suit­able con­di­tions for dam build­ing. There­fore, along the main reaches of the Spey there is likely to be little con­cern due to there being no like­li­hood of dam­ming, how­ever, on some spawn­ing grounds there is likely to be great­er con­cern and need for mon­it­or­ing and poten­tial mit­ig­a­tion if deemed neces­sary. As an example of the mon­it­or­ing points provided 90 out of the 181 mon­it­or­ing points (49.7%) have a high (fre­quent or per­vas­ive) capa­city for dam­ming if beavers were present.

Many of the mon­it­or­ing points provided do fall on reaches with beaver dam capa­city, mean­ing if beavers were present in these reaches, they may provide a mon­it­or­ing oppor­tun­ity or a man­age­ment con­cern. How­ever, with only 181 mon­it­or­ing points provided it is stat­ist­ic­ally unlikely that (at least in the short term) beavers would dam dir­ectly in the vicin­ity of these sampling points. There­fore, to

provide oppor­tun­it­ies to increase under­stand­ing between the impacts of beaver and fish sup­ple­ment­ary mon­it­or­ing tar­geted at beaver release site loc­a­tions could be beneficial.

N.B. It is highly likely, as shown in the research lit­er­at­ure for a num­ber of sites stud­ies (sum­mar­ised in Kemp et al., (2012), that beaver activ­ity, prin­cip­ally dam­ming, will cre­ate new spawn­ing grounds for sal­monids, as clean­er and well oxy­gen­ated gravel beds are main­tained. It is also pos­sible that small areas upstream of dams accu­mu­late sed­i­ment and poten­tially deteri­or­ate in terms of qual­ity for spawn­ing. Thus, the bal­ance between cre­ation of new spawn­ing grounds and poten­tial neg­at­ive impacts upon exist­ing spawn­ing grounds should be mon­itored. Over­all, at the catch­ment scale, it is most likely that spawn­ing hab­it­at will extend and improve, as beavers estab­lish and that sal­monid pop­u­la­tion health and abund­ance will fol­low suit.

Poten­tial Release Site Assess­ments Mod­el­ling out­puts and site vis­its were made for each of the sites dis­cussed below. Each site vis­it assessed and ground-truthed vari­ous site fea­tures. Meth­ods for identi­fy­ing the suit­ab­il­ity and key hab­it­at char­ac­ter­ist­ics for beavers (both spe­cies) have been widely stud­ied and pub­lished (includ­ing Allen 1983; Berg­man et al., 2018; Dittbren­ner et al., 2018; Hal­ley et al., 2009; Hood, 2020; Mac­don­ald et al., 1997). The main fea­tures con­sidered included;

  • The ini­tial com­pos­i­tion and struc­ture of the veget­a­tion with­in 30 m of the water’s edge

  • The dis­tri­bu­tion and abund­ance of pal­at­able ripari­an trees

  • The char­ac­ter of the ripari­an edge habitat

  • The hydro­logy of the water bod­ies avail­able to the beavers, includ­ing flow speeds, level sta­bil­ity and shoreline features

  • Water man­age­ment and where beavers may cause con­flict i.e., flood bank­s/low-lying farmland/​agricultural drainage.

  • Topo­graphy — gradi­ent of land, sub­strate type, val­ley shape

  • Asso­ci­ated land-use — dis­turb­ance and land-man­age­ment prac­tices, infra­struc­ture, water use

At each site an assess­ment of what beaver activ­it­ies would be likely (e.g. dam­ming or bur­row­ing) over time and if these have a poten­tial con­flict con­cern were also assessed. All site sur­vey work was under­taken in late March — early May, and involved speak­ing to CNP staff and loc­al landown­ers asso­ci­ated with these sites as far as possible.

The fol­low­ing maps present mod­el out­puts for key sites that pro­ject part­ner engage­ment and feas­ib­il­ity vis­its have iden­ti­fied as being of poten­tial interest for release. Please note no final decision has been made to release beaver at any of these sites.

In addi­tion to mod­el out­puts, the Park Author­ity have provided fish (Sal­mon and Trout only) dens­ity mon­it­or­ing points along­side mapped areas of known Aspen. For these data points rep­res­ent small, dis­crete groups of aspen <0.04 ha (2020 m) in extent, plus indi­vidu­al aspen trees. The poly­gons are lar­ger stands/​areas of aspen-rich wood­land ≥ 0.04 ha in extent. These data points where near to par­tic­u­lar sites of interest are included in maps for ref­er­ence. To note these maps should be treated as dis­play­ing gen­er­al data as a use­ful dis­cus­sion start­ing point and fur­ther ground mon­it­or­ing and refine­ment would be recom­men­ded going forward.

Site Map­ping Sum­mar­ies: Insh Marshes Over­all extens­ive areas of suit­able and highly suit­able hab­it­at around the site peri­met­er and more patchy veget­a­tion with­in the marshes them­selves which beavers will util­ise. No dam capa­city on the main River Spey as too large (espe­cially when in spate), but extens­ive ditch and drain­age sys­tems that would sup­port dam­ming in some loc­a­tions. Aspen pres­ence around the reserve peri­met­er, but little mapped with­in the site itself or in the ripari­an zone of the main river.

Vari­ous points across the Insh Marshes reserve were assessed in col­lab­or­a­tion with RSPB staff, espe­cially to dis­cuss poten­tial sources of con­flict, likely beaver beha­viours and prac­tic­al­it­ies of any release.

Sev­er­al points at and were assessed and most of these included highly suit­able wet wood­land and would be col­on­ised by beavers from the river Spey in time, they were all Par­tic­u­lar sens­it­iv­it­ies of beaver dam­ming on the were dis­cussed. Though the­or­et­ic­ally beavers may be able to dam this burn at low water levels, the like­li­hood of this should be high­lighted as very low.

Giv­en all these factors the motiv­a­tion for beaver dam­ming this stretch should be con­sidered as very low.

Fig­ures Redac­ted Fig­ures 18 – 20. very stony with little attract­ive veget­a­tion. The­or­et­ic­ally dam­mable giv­en stream width but highly unlikely beavers would be motiv­ated to dam and phys­ic­ally dif­fi­cult to ever maintain.

area is a pri­or­ity area for scrub remov­al giv­en it is a The area itself has little tree pres­ence but could offer attract­ive sum­mer for­aging. The area is lined by extens­ive patches of broadleaf, includ­ing aspen stands on hills nearby. It is likely as beaver pop­u­la­tion levels increase, this area would be col­on­ised by dis­pers­ing anim­als, beavers on the wider marsh are also likely to retreat to this flood­plain fringing wood­land areas dur­ing floods. This area was dis­coun­ted as a release site due to sens­it­iv­ity of and per­cep­tion of impacts on aspen stands. How­ever, it would be highly likely in time and any grow­ing pop­u­la­tion dens­ity that beaver would nat­ur­ally recol­on­ise this area, espe­cially as beavers could be pushed into closer prox­im­ity dur­ing floods. Sens­it­ive mature aspen stands and lichen assemblage of con­ser­va­tion value are gen­er­ally quite far back and uphill of the flood­plain, though of sig­ni­fic­ant dis­tance to both any pro­posed release sites and extent of high­er flood levels, pro­act­ive mon­it­or­ing and con­sid­er­a­tion of beaver deterrent fen­cing should field signs approach this area would be recom­men­ded. Plant­ing a buf­fer of wil­low could also be used in tar­geted areas to reduce motiv­a­tion to for­age on more valu­able areas of wood­land. A stock fence is cur­rently present along much of the fringing wood­land. It was noted that the dimen­sions would not pre­vent beaver access, with some con­cerns that a risk may be posed to smal­ler anim­als becom­ing stuck.

Fig­ure Redac­ted Fig­ure 21. Insh Marshes over­view This area is fringed by high­er banks and expans­ive broadleaf assemblage includ­ing aspen and is likely to become attract­ive as pop­u­la­tion dens­ity increases and/​or dur­ing peri­ods of expans­ive flood­ing of the Spey.

Fig­ures Redac­ted Fig­ures 22 and 23. Insh Marshes, not­ing veget­a­tion diversity likely rich sum­mer feed­ing for beavers and expans­ive fringing broadleaf wood­land likely to become more access­ible and attract­ive dur­ing peak floods. Mature aspen stands are present but gen­er­ally set well-back from flood­plain, though beavers could dis­play select­ive for­aging in time. React­ive man­age­ment plans includ­ing fen­cing and sac­ri­fi­cial wil­low plant­ing would be prudent strategies.

Onsite dis­cus­sions with CNP and RSPB staff have high­lighted a mon­it­or­ing sched­ule with­in such sens­it­ive areas will be crit­ic­al, with an agreed flow­chart of mit­ig­a­tion steps depend­ent on what activ­it­ies occur.

Burns with­in such wood­lands are typ­ic­ally steep and rocky, so there­fore would be dif­fi­cult for beavers to main­tain dams, would be a lot of effort for beavers to main­tain dams / deep water here. Mon­it­or­ing efforts should focus round the burns.

This is well veget­ated with extens­ive broadleaf and wet wood­land, it is con­nec­ted to the river Spey through a long range of wooded ditches fringing the marsh. These ditches appear to be deep, there­fore dam­ming is less likely, but beavers are likely to increase open water in this area through canal dig­ging and poten­tial dam­ming of smal­ler burns. Banks are read­ily bur­row­able. Over­all, this would be a highly suit­able release site and likely to be fairly self-con­tained. Beaver activ­ity could be viewed and has the poten­tial to increase the diversity, com­plex­ity and eco­lo­gic­al interest in this area.

The main poten­tial issue would be if beavers util­ised the nearby of Fur­ther dis­cus­sion and invest­ig­a­tion to determ­ine if beaver activ­ity could impact on water tree for­aging. levels around

Fig­ure Redac­ted Fig­ure 24.

Fig­ures Redac­ted Fig­ures 25 and 26. is also well wooded and sup­por­ted with diverse under­story veget­a­tion. Dam­ming along these water courses is highly unlikely giv­en depth of water, how­ever an adjoin­ing burn to the right could read­ily be dammed and may have some impacts on Dams in this burn could be mitigated.

Fig­ure Redac­ted Fig­ure 27. One of the main drains run­ning through RSPB Insh Marshes, such hab­it­at scores low for suit­ab­il­ity and dam­ming capa­city unlikely due to depth. Beavers would read­ily util­ise such ditches for access to more attract­ive hab­it­at and for­aging resources and for wider catch­ment dis­pers­al, con­sidered pos­it­ive for genet­ic flow.

The lower sec­tion of the was assessed. In gen­er­al, this is a well-wooded water course, but very rocky and shal­low, with peri­ods of high energy spates. Though some sec­tions appear on mod­el­ling which are the­or­et­ic­ally dam­mable, in real­ity the like­li­hood of beaver main­tain­ing dams is unlikely and would be sea­son­ally flushed out. The shoreline is very rocky and in gen­er­al seems unat­tract­ive to beaver col­on­isa­tion and not con­sidered as a suit­able release site. At high pop­u­la­tion dens­ity, dis­pers­ing indi­vidu­als may explore and util­ise this area in seek­ing any suit­able hab­it­at upstream. In gen­er­al, this would be con­sidered as a low con­flict area.

and wet wood­land area This is a pool- ditch sys­tem dom­in­ated by mixed age wil­low. Deer impacts are evid­ent, and a revi­sion of deer man­age­ment is sug­ges­ted but wil­low is get­ting away in inac­cess­ible areas. Dam­ming and

bur­row­ing would not be any issue here and RSPB would wel­come an increased com­plex­ity gen­er­ated by beaver activ­ity. The though beavers would read­ily use the wider area and have access to for­age and shel­ter cre­ation oppor­tun­it­ies without con­cern. The river is very access­ible and imme­di­ately provides good stretches of highly suit­able habitat.

Fig­ures Redac­ted Fig­ures 28 – 31. Wet wood­land asso­ci­ated with the and offer­ing highly suit­able hab­it­at and shel­ter oppor­tun­it­ies. Banks are highly suit­able to bur­row and shel­ter con­struc­tion. This area does reg­u­larly see annu­al flood­ing events though beavers would be very able to be mobile with water levels and seek shel­ter and for­aging oppor­tun­it­ies on high­er fringing banks and woodland.

Fig­ures Redac­ted Fig­ures 32 and 33. offers very attract­ive hab­it­at and it could be likely any released beavers reside in This area and for­age in the fen behind. is access­ible, with wil­low dom­in­ated tree cov­er­age and deep fri­able banks for easy shel­ter construction.

on the reserve. The major­ity of This area of the Insh marshes offers an area of the shoreline is sup­por­ted by stretches of highly suit­able hab­it­at, includ­ing large sec­tions of wet wil­low fen. Emer­gent and aquat­ic veget­a­tion is also high in this area espe­cially offer­ing plenty of sum­mer for­aging oppor­tun­it­ies. Giv­en the bank pro­file gradi­ent, bur­row­ing for shel­ter oppor­tun­it­ies may be lim­ited, how­ever lodge con­struc­tion would be entirely pos­sible. Beaver activ­it­ies may also see extens­ive canal net­works being con­struc­ted should beaver reside in this area. Highly suit­able and attract­ive hab­it­at is present in sig­ni­fic­ant sec­tions on the main River Spey below. Prox­im­ity to and being linked by extens­ive wil­low wood­land, makes this very access­ible but also likely that any beaver released onto the are likely to con­cen­trate activ­it­ies in this area and less likely to be attrac­ted to wood­land asso­ci­ated with This has been flagged for mature wood­land, aspen, bry­ophyte and lichen interests, but this so not likely to be vul­ner­able to beaver for­aging with any impacts most likely con­cen­trated in the less steep peri­met­er area of the site. On site dis­cus­sions with CNPA, Beaver Trust and RSPB staff have determ­ined that a long-term mon­it­or­ing pro­gramme for any beaver impacts would be pos­sible in this area which would trig­ger a react­ive imple­ment­a­tion of a mit­ig­a­tion strategy.

Fig­ures Redac­ted Fig­ures 34 and 35. Look­ing into and entirely pass­able for beavers. Fig­ure Redacted

areas from Stock fen­cing is con­struc­ted from posts and hori­zont­al wire only

Fig­ure 36. Prox­im­ity of loch to River Spey fringes to the right. Beavers highly likely to release is recom­men­ded. access both feed­ing resources and may be ini­tially attrac­ted to this river sec­tion, though a Fig­ure Redac­ted Fig­ure 37. River Spey below — low con­flict area with good sec­tions of high qual­ity habitat.

is sur­roun­ded with a wide mar­gin of wet wood­land. Asso­ci­ated reed beds, aquat­ic and semi-aquat­ic veget­a­tion appear extens­ive and diverse, lined with broadleaf and would provide

extens­ive for­aging oppor­tun­it­ies year-round. The shoreline has mul­tiple bays and com­plex­it­ies and though gen­er­ally low in gradi­ent, shel­ter build­ing oppor­tun­it­ies would not be an issue. Water levels appear fairly stable but if floods occur beavers would have fringing shoreline to move into without restric­tions. Map­ping demon­strates an excel­lent expanse of highly suit­able hab­it­at both around the loch and the neigh­bour­ing ripari­an zone of the main chan­nel. High mod­elled dam capa­city on both inflow and out­flow, but main the Spey are too large to be dammed. Any dam­ming would serve to increase wet­land hab­it­at and com­plex­ity of this site, with addi­tion­al biod­iversity bene­fits highly likely. Sur­round­ing marsh/​grassland could read­ily con­tain a net­work of beaver dug canals, again increas­ing site com­plex­ity and areas of open water. No mapped Aspen stands in the ripari­an zone of either or imme­di­ate ripari­an zone of Spey. No fish dens­ity data sampling points provided. In gen­er­al, the poten­tial for con­flict at this site appears low with sur­round­ing land-use decidu­ous wood­land with some tracks which are all loc­ated away from the itself, there­fore it seems unlikely bur­row­ing, dam­ming or tree felling are likely to be an issue.

Fig­ure Redac­ted Fig­ure 38. Beaver hab­it­at map­ping for. and sur­round­ing area.

Fig­ure Redac­ted Fig­ure 39. Beaver dam capa­city map­ping for and sur­round­ing area.

Fig­ure Redac­ted Fig­ure 40. Beaver dam capa­city map­ping along­side known areas of Aspen and fish mon­it­or­ing points for and sur­round­ing area.

Fig­ure Redac­ted Fig­ure 41. Over­view of with com­plex shoreline includ­ing mul­tiple bays, good broadleaf cov­er­age and extens­ive wet­land with patches of wet woodland

×

We want your feedback

Thank you for visiting our new website. We'd appreciate any feedback using our quick feedback form. Your thoughts make a big difference.

Thank you!