Mammals along the Egnatia Highway in Northern Greece. Internship - PowerPoint PPT Presentation

Master 2 “Ecologie & Biodiversité” research project Speciality: BIODIV “Biodiversity conservation” 2011-2012 “ Assessment of the use of crossing structures by bears and other Mammals along the Egnatia Highway in Northern Greece.” Internship supervisors: Author: Theodorou Kostas Dimitras Elisabeth Legakis Anastasios

1. Introduction One cause of habitat fragmentation: Transport infrastructure (Forman, 2000)  Variable ecological risks : • Fragmentation & Isolation of populations  Loss of genetic diversity • Wildlife -Vehicle c ollisions

New field of study: Road ecology Achieving the maintenance of opportunities for animals to move across broad landscapes. Design and incorporation of wildlife mitigation crossing structures into road construction and improvement projects. (Clevenger & Waltho, 2005) .

For most species, limited knowledge regarding : - Effective & affordable crossing structure design (Spellerberg, 1998) . - The characteristics that promote or discourage their use (Clevenger & Waltho, 2000) . “Location of a crossing structure in relation to habitat quality” or “structure design” ? (Cain et al., 2003, Clevenger & Waltho, 2000, 2005; Ng et al., 2004).

Evaluation of the effectiveness of crossing structures  monitoring their use from wildlife species : • Camera surveillance • Sand or marble dust track -pads. In Northern Greece :

Egnatia Highway, 2009 – present : 16 bear- vehicle collisions, 13 of which were fatal. Reasons: • Highway situated in heart of bear’s habitat ( estimated population size 33-51 individuals ) (Scouras & Drosopoulou, 2005, Karamanlidis et al., 2011) • Inadequate means of protection (regarding fences)

5 Different types of mitigation crossing structures on the Egnatia Highway: Box Culvert Local Road Viaduct Wildlife Underpass Oversized Streamed Culvert

Aim of this study: Whether and how much the mitigation crossing structures are used by the surrounding wildlife.

Particular focus on the Brown Bear (Ursus Arctos): • Study area: part of the distributional range of the larger of 2 independent populations ( 150 individuals out of a national population of 190-260 individuals ). • Species considered Endangered in Greece. (Greek Red Data Book) • “Priority Species” according to the European Legislation (EU Directive 92/43, Annex II).

2. Materials & Methods Study area: NW Greece, Grevena, Pindos mountain range • 9,14 km highway section: 10 mitigation crossing structures of 5 ≠ types. • Monitoring period: Sept. – Dec. 2011

• 4 photographic cameras Checked every 12-15 days • 6 video cameras Data analysis • Analysis carried out on Data from 8 out of 10 sample locations • Focused on: 6 key crossing characteristics

Analysis of key crossing characteristics the dimensions of the structures (height, length, 1. Structu width), re’s the Openness Index [(width*height)/ (length)], charac 2. teristic the mean slope of their entrances/exits, 3. s the surrounding vegetation and the % of forest cover 4. at 100m. & 500m. as well as the distances from water sources (streams and rivers), Landsc ape’s the distances from villages, livestock pens, forests, 5. charact paved and unpaved roads, eristics the amount of human activity in the area. 6.  Pearson’s rank correlations coefficients  Stepwise multinomial logistic regression

3. Results i) Use by animal species Table 1: Nature and frequency of use of all types of mitigation structures . Total Verified and records probable crossings Crossings (as % of total records) Species Verified Probable Total No use Bear 14 0 14 0 14 100 Wolf 18 7 25 3 28 89,2 32 7 39 3 42 92,8 Total large mammals Wildcat 3 1 4 4 8 50 Fox 8 13 21 11 32 65,6 Mustelid 4 0 4 1 5 80 Badger 1 1 2 0 2 100 Hare 0 0 0 2 2 0 Total small 16 15 31 18 49 63,2 mammals Cat 5 2 7 3 10 70 Dog 56 20 76 8 84 90,4 61 22 83 11 94 88,2 Total domestic animals

Table 2: Detections and frequencies (by total number and %) of each species. Species Frequency Percent . Bear 14 7,6 Wolf 28 15,1 Native large mammals 42 22,7 •185 Detections of 255 Individuals (some of the records captured Cat 10 5,4 grouped individuals). Dog 84 45,4 Domestic animals 94 50,8 Budger 2 1,1 Fox 32 17,3 Hare 2 1,1 Mustelid 5 2,7 Wildcat 8 4,3 Small mammals 49 26,5

Table 3: Human presence in all mitigation structures Frequency Type of human presence Pedestrians 46 Bicycles 5 Shepherds with Livestock 100 Motocross 71 Vehicles 247 Total 469

Figure 1: Histogram showing the number of detections by mitigation structure, clustered by Pass Through.

Figure 2: Line chart showing the use of mitigation structures by each species clustered by day and night. “At night many large predators species move along roads that have little vehicular or people traffic” Forman & Alexander, 1998

ii) Correlation with mitigation structures attributes and the surrounding landscape Table 4 : Pearson’s rank correlation coefficients for bear crossings (verified and probable crossings combined) Species Height Openness Length Width Distance Distance Distance Vehicles Index from from from passed villages Livestock's water through pens resource the mitigation structures Bears 0,794* 0,824* NS NS 0,737* NS -0,843** 0,741* Statistically significant relationships are indicated with asterisks (*P.value<0.05, **P.value<0.01) NS=Non Statistical Significance

Table 5: Pearson’s rank correlation coefficients for other species crossings (probable and verified crossings combined). Species Distance from forests from the right entrances Mean slope at 500m. Wolf 0,882** -0,709* Fox 0,901** NS Wildcat 0,834** NS Dog 0,847** -0,735* p < 0. 01 p < 0.05 NS = Non Statistical Significance

iii) Multinomial Logistic Regression Table 6: Information regarding the Multinomial Logistic Regression. Model fitting Pseudo Predictor Goodness Of Fit Parameter (GOF) Estimates variables information R-square influencing the model Concerning Sig.=0,001<0,05 Pearson 41,362 Sig.=0,003<0,05 Openness Cox and Snell the 0,152 Index Deviance 39,766 (Bears) structure’s Nagelkerke 0,160 characterist df = 35 McFadden 0,054 ics Concerning Sig.=0,000<0,05 Pearson 44,438 Sig.=0,001<0,05 Distance Cox and Snell 0,157 the Deviance 46,582 (Bears) from landscape’s Nagelkerke 0,165 water sources characterist df = 42 McFadden 0,056 ics

4. Discussion Concerning the Brown Bear: • The brown bear prefers large and open structures (O.I = 1.68). - In accordance with other studies : Cavallaro et al., 2005, Putnam, 1997, Kusak et al., 2009, Clevenger & Waltho, 2000. ! In contrast with previous study : Tritsis, 2010: Bears also used structures of low openness (0.075 ≤ O.I ≤ 0.5). - Studies conducted in Reasons which can different seasons (Summer vs. Autumn). explain this discrepancy - Small sample size of bear detections in both studies.

• Bears avoid mitigation crossing structures near rivers and streams !!! Result of human activity (Smith, T.S., 2001) / Turney & Roberts, 2004 : Opposite conclusion on Grizzly Bears • Significant positive correlation: Distance from villages ~ use of the crossings by Bears. (ease of finding food within agricultural areas nearby?) • Positive correlation: Use of the crossings by Bears ~ presence of vehicles  Bears not disturbed by daily high use of crossings by vehicles  Local Roads efficient corridor for Bear movements  Mitigation crossing structures constructed near local roads maybe an overage…

Concerning species other than Bears: • Negative correlation: Wolves’ & Dogs’ movements ~ mean slope at 500m from mitigation crossing structure’s entrance <=> Wolves are using for their movements areas with smoother terrain (Iliopoulos, 2010) . • Foxes, wolves, dogs and wildcats prefer to use structures with cover nearby. - Similar results for wildcats & foxes by Rodriguez et al. on 1997. - For medium mammals, generally, in accordance with many previous studies : Ng et al., 2004, Clevenger et al., 2001, Clevenger et al., 2003. - In contrast (foxes) with Mata et al., 2005

5. Conclusion In NW Greece :  Openness Index most important for Bears.  Structure’s location & Vegetation Cover more important for wolves, foxes & wildcats. Further work needed in order to reveal the influence of seasonality in Bear movements!

6. Overview Field work : Elisabeth Dimitras & Y. Lazarou & Y. Mertzanis (from Callisto team) GIS Database : Distances from villages, water sources, mean slopes etc. Alexios Giannakopoulos (from Callisto team) Statistical analysis : Elisabeth Dimitras with the guidance from an expert on SPSS

Thank you for your attention !!!!