Page 3 - FATTORINI2011

TABLE 1. Geographical variables for the circum-Sicilian islands. S – number of species; A – Area (km2); Ds – Distance to Sicily
(km); Da – Distance to North Africa (km); Di – Distance to the nearest island (km); SDs – Sea depth to Sicily (m below sea level);
SDa – Sea depth to Africa (m below sea level); SDi – Sea depth to the nearest island (m below sea level). X1, X2 and X3 are the three
dimensions extracted by Non-metric Multidimensional Scaling from a matrix of inter-island faunal dissimilarity (Kulczinski 2
index).

Island SA Ds Da Di SDs SDa SDi X1 X2 X3
1500 10 –0.306 –0.049 0.075
1. Lipari 32 37.29 28 370 1 780 1500 366 –0.392 0.169 –0.178
1500 10 –0.181 0.248 0.073
2. Salina 24 26.38 39 365 4 780 1500 1281 –0.255 0.326 –0.330
1500 1281 –0.629 –0.038 0.416
3. Vulcano 22 20.87 21 375 1 780 1500 1281 –0.562 0.040 0.220
1500 600 –0.582 –0.033 –0.010
4. Stromboli 25 12.19 58 405 18 1200 1500 80 –1.004 0.603 0.339
1500 50 –0.917 0.290 –0.152
5.Filicudi 15 9.49 46 345 15 1500 1500 50 –0.831 –0.032 –0.102
1500 10 –0.737 –0.036 –0.628
6. Alicudi 18 5.1 53 330 15 1500 1500 20 –0.283 –0.089 –0.596
1500 1500 –0.334 –0.416 0.326
7. Panarea 22 3.34 42 390 14 780 400 42 –0.108 –0.468 0.484
400 42 0.001 –0.410 0.326
8. Basiluzzo 3 0.29 43.5 394 3.5 780 400 145 0.314 –0.322 1.046
500 500 0.292 –1.255 0.066
9. Lisca Bianca 4 0.0413 42 390 3 780 400 400 0.247 –1.021 –0.675
100 100 0.234 1.762 –0.285
10. Bottaro 7 0.0073 42 390 2.75 780 100 400 0.646 –0.527 –0.879
500 100 0.871 0.018 0.144
11. Scoglio Faraglione 4 0.0049 39 365 0.27 780 500 100 1.121 0.015 0.033
500 100 0.961 0.417 –0.441
12. Pietra del Bagno 3 0.0021 28 370 0.45 780 500 100 1.389 0.378 –0.413
500 100 1.047 0.432 1.141
13. Ustica 26 8.6 53 255 100 1500

14. Levanzo 18 5.61 12 152 4 42

15. Favignana 28 19.7 8 140 4 42

16. Marettimo 16 12.06 35 130 15 145

17. Pantelleria 23 86 95 67 115 500

18. Linosa 19 5.34 162 163 40 500

19. Lampione 4 0.025 220 130 18 500

20. Lampedusa 28 20.2 195 120 18 500

21. Malta 43 246 93 292 6 200

22. Gozo 26 67 82 285 6 200

23. Comino 10 3.5 85 295 2 200

24. Cominotto 2 0.25 85 295 0.1 200

25. Filfla 3 0.06 100 292 5 200

urban fabric, Discontinuous urban fabric, Industrial or commer- There are indications that various insular populations of tene-
cial units, Port areas, Airports, Mineral extraction sites, Dump brionid beetles described as subspecies differ profoundly geneti-
Sites, Green urban areas, and Sports and Leisure facilities), Cul- cally (cf. Chatzimanolis et al., 2003; Ferrer, 2008). The current
tivation (including Vineyards, Non-irrigated arable land, Natural taxonomic dividing line between species and subspecies, as
grassland, Annual crops associated with permanent crops, Com- applied to the tenebrionids of the circum-Sicilian islands, is
plex cultivation patterns, and Land principally occupied by agri- arguably arbitrary and the exclusion of subspecies could result
culture, with significant areas of natural vegetation), Coniferous in a significant underestimate of endemic island tenebrionid
forest, Broad-leaved and Mixed forests, Sclerophyllous vegeta- diversity. Thus, I considered both species and subspecies as
tion, Bare rock and Sparsely vegetated areas, and Wet areas “evolutionarily significant units” (Ryder, 1986) and included
(including Salt marshes, Salines, and Water bodies). Even if both categories in all analyses. The validity of the endemic sub-
these broad landscape units are coarse in comparison to the species Heliopathes avarus donatellae (Canzoneri, 1970) is dis-
scale at which insects perceive small-scale environmental het- puted and the identity of the endemic species Opatrum
erogeneity, they correspond well to distinct keystone structures melitense Küster, 1849 is uncertain. Both taxa have been provi-
(Tews et al., 2004) for tenebrionid species. In particular, each of sionally considered as valid. On the basis of the original
the seven classes used here corresponded to different microcli- description, the differentiation of Heliopathes avarus dwejrensis
mate conditions, food resources, and soil characteristics, which Scupola & Mifsud, 2001 from Gozo seems very subtle. The
are among the most important factors shaping tenebrionid com- same is the case for Probaticus cossyrensis Sparacio, 2007,
munities in the Mediterranean (Fattorini, 2008b). which was recently separated from Probaticus anthrax Seidlitz,
1896. All analyses were performed including and excluding
Geographical distances and land cover categories were these two taxa. Only results with all taxa included are presented,
obtained from the GIS of the Italian Ministry of the Environ- because exclusion of these two taxa produced virtually identical
ment and Protection of the Territory and Sea (2009) and from results.
Malta Environment and Planning Authority (2009). Sea depths
were obtained from bathymetric maps mainly from the Istituto Finally, cosmopolitan species strictly associated with human
Idrografico della Marina (1997). food, such as Palorus subdepressus (Wollaston, 1864), Gnato-
cerus cornutus (Fabricius, 1798), Alphitobius diaperinus (Pan-
Biological data zer, 1796), Tenebrio spp. and Tribolium spp. were excluded
from all analyses to reduce the risk of confounding natural pat-
Presence/absence data of individual species on each island terns with effects of man. A list of 45 references used to com-
(Appendix 1) are based on extensive field surveys made by spe- pile presence/absence data is given in Appendix 1.
cialists with similar sampling efforts among islands; the faunal
inventories are fairly complete. A total of 107 native taxa are
known from these islands.

661