A co‐development approach to conservation leads to informed habitat design and rapid establishment of amphibian communities

1. Effective wildlife restoration is a critical requirement of many conservation actions. The outcome interventions can be optimized through knowledge species’ habitat requirements, but few studies consider the impact using explicit evidence from dedicated local research to inform design phase management. Furthermore, administered externally top down, whilst simpler than those developed in discussion with multiple stakeholders including land managers (i.e. co-development), run risk failing engage people. 2. In this study, we focus on Scottish Highlands improve availability and suitability breeding ponds for amphibian assemblages. We collected analysed data based 129 ecological variables across 88 reference quantify preferences. used findings these analyses construction or 25 intervention co-developed partnership (landowners, foresters, citizen scientists government agencies). Following interventions, monitored communities at sites over 4 years. assessed presence abundance all five native amphibians (the anurans Rana temporaria Bufo bufo, salamanders Lissotriton helveticus, L. vulgaris Triturus cristatus) egg searching, dip-netting, torching trapping. 3. new habitats were overall characterized by conditions more favourable ponds. recorded total 51 colonization events. Within two seasons after restoration, hosted full complement species, mirroring diversity patterns found 4. Our study shows that requirements working commercial landmanagers ensure equitable benefits prior designing actions promote rapid efficient recovery wildlife. L'eficàcia en la restauració de fauna salvatge és un requisit bàsic moltes accions conservació. El resultat les intervencions conservació es pot optimitzar mitjançant el coneixement dels requeriments d'hàbitat espècies, però pocs estudis han examinat utilitat evidències generades investigacions dutes terme per dissenyar-ne l'hàbitat. Tanmateix, gestionades centralitzadament, tot i que són d'execució més senzilla desenvolupades col·laboració amb diferents parts interessades, inclosos els propietaris (gestió cooperativa), corren risc no aconseguir implicació població local. En aquest estudi, ens centrem regió Escocesa incrementar disponibilitat d'estanyols millorar seva idoneïtat reproducció comunitat autòctona d'amfibis. Vam recollir analitzar dades sobre ecològiques estanys estanyols referència, quantificar preferències locals espècies utilitzar conclusions d'aquest estudi informar disseny l'hàbitat construcció o feta conjuntament altres interessades (propietaris, gestors, científics, naturalistes organismes governamentals). Posteriorment, vam dur seguiment comunitats d'amfibis durant anys. Es va avaluar presència l'abundància cinc amfibis autòctons (els anurs bufo urodels cerca d'ous, utilització salabres, trampes torxes (cerca nocturna). Els nous hàbitats caracteritzaven general unes condicions favorables als referència. registrar noves colonitzacions. Dins dos anys següents restauració, trobaven algun alguns estanyols, quals mostraven patrons diversitat equivalents nostre confirma recerca ecològica treball cooperatiu garantir l'equitat beneficis abans dissenyar conservació, promoure una recuperació ràpida eficient autòctona. Effiziente Renaturierungsmaßnahmen spielen vielen Naturschutzprojekten eine zentrale Rolle. Der Erfolg von Artenschutzinterventionen kann durch Wissen über lokale Habitatsanforderungen optimiert werden, aber detaillierte tierökologische Lokalstudien werden nur selten der Planungsphase Naturschutzmaßnahmen explizit berücksichtigt. Von oben diktierte sind oft leichter zu implementieren Interventionen, die unter Einbezug Diskussionen mit lokalen Landnutzern und Interessensgruppen „Ko-Entwicklung“ geplant durchgeführt werden. Solche Projekte jedoch einem hohen Risiko unterworfen, Lokalbevölkerung nicht aktiv engagieren. vorliegenden Untersuchung beschreiben wir ko-entwickelte Naturschutzintervention im schottischen Hochland, wurde, um das Vorhandensein Qualität Laichgewässern für Amphibiengemeinschaften verbessern. Um Habitatpräferenzen quantifizieren, sammelten Daten ökologische Variablen Referenzgewässer. Die Ergebnisse dieser Analysen wurden verwendet, neue oder restaurierte Laichgewässer Zusammenarbeit wie Landbesitzer, Förster Lokalregierungen schaffen. Nach Errichtung Gewässer ihre Besiedelung Amphibien vier Jahre untersucht. Das Häufigkeiten fünf lokal heimischen Arten (die Anuren Grasfrosch, Erdkröte, sowie Schwanzlurche Teichmolch, vulgaris, Fadenmolch, Kammmolch, wurde mittels standardisierter Feldarbeit Keschern, Reusenfallen visuellen Methoden (Nachweis Eiern, Nachterhebungen Suchlampen) quantifiziert. neu geschaffenen Laichhabitate boten statistisch nachweislich bessere Bedingungen Wir wiesen insgesamt Neubesiedlungen bestimmten nach. Bereits zwei nach Restaurierung enthielten Artengemeinschaften, vergleichbaren Referenzgewässern entsprachen. Unsere hat gezeigt, dass Datenerhebung Analyse lokaler vor Durchführung gemeinsam einer Einbeziehung Bedürfnisse raschen Verbesserung ökologischen Situation Tierpopulationen führen kann. Increased recognition global biodiversity crisis need have 21st century coincided higher economic uncertainties (Díaz et al., 2019; IMF, 2020), leading reduced investment (Hirsch 2020). As worsen, there an increasing effective are well-executed supported empirical (Addison 2013). Co-development projects managers, citizens, practitioners, policy makers researchers has been proposed as means implementing successful (Vercammen & Burgman, Wauters Mathijs, evidence-based approach appears good fit co-development it recommends systematic evaluation peer-reviewed grey literature along practitioners’ people's experience (Sutherland, Pullin, Dolman, Knight, 2004). Conservation both practical focuses finding solutions likely appeal broad range stakeholders, includes social economic, well ecological, considerations (Gilby decline major concern linked anthropogenic drivers loss, use change, pollution, over-exploitation, disease invasive species (Beebee Griffiths, 2005; Scroggie 2019), often acting combination. Remediation such threats particularly challenging outside protected areas places primarily managed purposes (Denoël Hartel, Scheele, Rozylowicz, Horcea-Milcu, Cogălniceanu, Indeed, forestry agriculture, rather se, comprises greater proportion rural globally provides vital (e.g. Irwin 2014; Pywell 2012). Amphibians particular limited powers dispersal, gain may thus only connection between nature reserves other (Pickett Thompson, 1978). situations, proactive management aids population (Sterrett 2019). One widely promoted, effective, form agricultural areas, where traditional too prescriptive, creation ‘pondscapes’ (ecologically meaningful clusters ponds) designed support wildlife, aquatic-breeding (Hill 2018; Peterman, Anderson, Drake, Ousterhout, Semlitsch, 2013; Rannap, Lõhmus, Briggs, 2009). Pondscapes funded agri-environmental incentive schemes, part mitigation measures following developments. However, appear piecemeal, amphibians, which led calls landscape scale planning connectivity needs target (Brown, Street, Nairn, Forstner, 2012; Rannap Interventions must generally take account aquatic terrestrial provide breeding, feeding dispersal opportunities (Schmidt, Arlettaz, Schaub, Lüscher, Kröpfli, pond had mixed success, relatively evaluated their effectiveness (Smith, Meredith, Sutherland, Ecological features constructed restored significantly influence richness viability residing (Peterman Shulse, Trauth, Gardner, 2012), important because markedly differ established newly created (Drayer Richter, 2016; Korfel, Mitsch, Hetherington, Mack, 2010). Studies document temporal persistence indeed show fail reach typical community composition might not attain index values (Lesbarrères, Fowler, Pagano, Lodé, Even when available, information rarely considered This summarizes two-step effort five-species assemblage area Northern Scotland. sequential objectives (i) preferences (ii) comparison existing sites, determine whether assessment-based promotes colonization. was (five 20 constructed), situated Moray Firth-Aviemore Highlands, UK (4°35′–3°35′W, 57°38′–57°11′N; Figure 1, Table 1). Reference included glacial ponds, latter ranging earliest detailed maps (surveyed c. 1870) approximately 15 years ago. A previous differences origin age (Miró, O'Brien, Hall, Jehle, 2017). They consisted 40 post-1990 records European Protected Species cristatus (see below), 48 selected surveying nearest randomly generated grid within (3°35′-4°35W, 57°11’-57°38′N; see Miró 2017, details). practitioners (including authors, DOB KOB), farmers, golf club, governmental organizations regional NGO (Highland Amphibian Reptile Project – HARP). Lengthy discussions held each potential stakeholder explain rationale understand constraints pose businesses. took place four nodes area, criteria connectivity, soil/geology compatibility (Figure 1; A1 Supporting Information). None source populations reserves, although (on National Forest Estate) partly conservation. Pond known high diversity. Construction partners (Forest Land Scotland NatureScot) directly hired contractors. Landowners received financial incentives part. served basis preferences, before applying develop siting below). surface ranged 2 168,500 m2 (median 975 m2), altitudes 10 248 m a.s.l. 91.5 m). surveyed least three, four, times annually 2014 2018 during season (March–June). All region (McInerny Minting, 2016): anurans, common frog (R. temporaria) toad (Bufo bufo), together three salamanders, smooth newt (Lissotriton vulgaris), palmate (L. helveticus) great crested (T. cristatus). Survey techniques British Recording scheme (NARRS) protocol, carried out experienced licensed surveyors suitable weather (night water temperature ≥6°C, low wind disturbance surface). NARRS protocol standardizes survey perimeter encompassed dip netting, (flashlighting) minnow traps (ARG-UK, Griffiths Langton, 2003; Beckett, Foster, 2001). Sewell, Beebee, (2010) combination methods frequency minimize imperfect detection. Data surveys combined given year absence followed NatureScot guidance welfare biosecurity non-native control advised field workers 2008). To investigate most composition, derived fieldwork 41 desk (for further details descriptors, Appendix Information Topographical obtained GIS 1:25,000 mapping agency Ordnance Survey. Given context proximity occupied non-occupied did include spatial autocorrelation avoid unnecessarily complexity collinearity models. former identified historic maps) Autumn–Winter 2014–2015 (except one Winter 2017–2018), 13-t excavator low-pressure tracks. Based attention paid use, soil type, bank slope, constructing restoring (Table There planting introduction species: naturally colonized. 500 150 46 163 141 2015 protocols outlined above. Habitat characteristics also same 2016 season, second except 2017—2018 Summer 2019. relative importance shaping redundancy analysis (RDA; Wollenberg, 1977), response variable, explanative variables. (and events later analyses), presence/absence, coded 1/0. Calculations Euclidean distance Hellinger transformed data, allows computing indices beta comparisons occurrence (1/0) (Legendre De Cáceres, Legendre Legendre, First, performed individual RDAs retaining significant (p < 0.05). Second, linear dependencies among retained variables, since could render RDA regression coefficients unstable sequentially computed variance inflation factor (VIF) rejected ones highest value until none them above Seventeen VIF below 3, threshold indicative worrisome (Zuur, Ieno, Walker, Saveliev, Smith, variable selection procedure Information. Since tested similarity groups Kruskal–Wallis non-parametric tests group homogeneity, post hoc pairwise Bonferroni adjustment. B. favoured different prefers larger, deeper commonly containing fish (Hartel 2007; Knapp, Miró, Sabás, Ventura, Winandy, Darnet, Denoë, 2015), split into groups: present, without only. illustrated violin plots (Hintze Nelson, 1998). investigated changes 3 yearly measures. number pond, year. pooled defined pool Finally, dissimilarity (or diversity) site-by-species (rows columns, respectively) table sum matrix squared deviations column 2013), again Gallagher, bias caused sample sizes 999 random equally sized subsamples (n = 15) allow category ‘reference’ randomizations build years, kept dataset. statistical test comparisons. Due large size, applied significance level α 0.001 reduce probability type I errors (Cohen, 1988). performing permutational multivariate (PERMANOVA; Anderson Gorley, 2008) against factor. divided only, correction. Computations 1/0 normalized transformation PERMANOVA Principal Component Analysis (PCA) space built described RDAs. then projected (2014–2018) centroids (mean SE) PCA space. visual comparisons, presence/absence categories marked differentially graphics. PERMANOVAs PERMANOVA+ PRIMER software (Anderson 2008), 9999 permutations 0.05. R (R Core Team, 2018) basic functions packages adespatial (Dray vegetarian (Charney Record, 2012) vegan (Oksanen vioplot (Adler Kelly, 2018). Analysing associated taxonomic structure (F 4.681, p 0.001), explained 35.1% 2a). nine forward process significant, prominently macrophyte coverage, adjacent woodland, organic mud substrate slightly sloping banks From single-species perspective, showed contrasting compared 2a; A2 Occurrence positively very bank, negatively coverage A2). R. thickets scrub, Among helveticus presence, tree shore, scrub T. insufficient clear Once assessed, line findings. targeted locations 2b). By already similar better 2b, excluding only). No contained fish. For 76.6% 50% fishless (χ2 16.85, 0.001). Intervention indiscernible whereas lower proportions shore unfavorable characteristic helveticus. Organic richness, desirable maturation, less 2). Forty-three colonizations (84%) (21 2015, 22 2016). responsible frequent colonized 13 first (2015). 17 seven 2016. Fourteen period: cristatus. exceeded (2016; median 1.93 1.66 continued increase 2.3 (Kruskal–Wallis chi-squared 5144.7, 0.001; 3a). Pooled increased time 4873, 3b), maximum possible area) Beta 3454.6, 3c) reaching stable works (2017), however still levels 0.69 1.05 deviations, respectively). calculated works), due lack variation matrix. confirmed reached compositions (pseudo-F 9.6228, 0.0001; 3e 3d). occupancies 2b; A3 43 works, 2018. addition, comparable intervention, Lesbarrères 2010; Petranka, Harp, Holbrook, Hamel, propose mostly our application acquired quantitative intervention. made project terms, cost under £20,000, estimate volunteer monitoring being around 400–500 worker-hours, project) land-use terms. suggest project's success bringing approaches. Three landowners previously undertaken work land, state forest FLS) done so formal basis. Participants motivations. wished seen stewards, protecting whose survival represents legacy land-management systems handed down ancestors; families owned 17th century. farm desire integrate profitable businesses, doing peers. participants joined convinced another manager. Whilst important, neither participated schemes bureaucracy involved. attraction current form-filling, any