FLEA BEETLES AND BIOGEOGRAPHY OF THE MACARONESIA (COLEOPTERA, CHRYSOMELIDAE, ALTICINAE) (*") The subject of this communication consists of a biogeographical study on flea-beetles of the Macaronesia. Flea beetles are a group of small phvtophagous Coleoptera, belonging to the Chrysomelidae family and Alticinae subfamily, with jumping characteristics. For this study I considered the following insular groups and islands: - the Madeiran islands, with Porto Santo, Ihlas Desertas and Madeira, itself; - the Canary islands, with Fuerteventura, Lanzarote, Gran Canaria, Tenerife, Gomera, Hierro and La Palma; - the Cape Verde islands, with Santo Antao, San Vincente, Santa Luzia, San Nicolau, Sal, Boa Vista, Maio. San Tiago, Fogo and Brava; from Santa Luzia, Boa Vista and Maio no flea beetle species is known till now. Moreover, I did not consider the islands of the Azores and Salvages because they have not enough information as yet for this Coleoptera group. In three insular groups examined the following number of flea-beetle taxa are present: - 25 species in 9 genera in the Madeiran islands; - 45 species in 10 genera in the Canary islands: - 11 species in 6 genera in the Cape Verde islands In Fig. 1, we can point out how the percentage of endemic taxa is higher in Cape Verde and Canaries (54.5% and 62.2% respectively), while it is relatively low in Madeira (38%). On the contrary the percentage of faunistic elements of wide distribution is higher in Madeira (58%) and lower in the Cape Verde and Canary islands (respectively 35.6% and 36.4%). The endemic ctaxa of the Macaronesia common to the insular groups considered are very scarce. Three macaronesic species are common to the (') Parco Nazionale del Circeo - 04016 Saearrola, Latina (Italia). (*') Ricerche efiettuate con Fondi rrogati dal Minis~ero della Pubblica Istruzione (40 %).
Fig. I - Number of taxa and percentage of the huuistic elements in each insular group. Fig. 2. - Number of taxa and percentage of the faunistic elemeuts in the Canary Islands.
Madeiran and Cana~y Islands (Longitarsus nubigena Wollaston, Longitarsus isoplexidis Wollaston and Psylliodes vehemens Wollaston). However with regard to Longitarsus isoplexidis and Psylliodes vehemens in the Canary islands, the former is present with endemic subspecies persimilis Wollaston and the latter is present with endemic subspecies normandi Heikertinger. One macaronesic species (Longitarsus messer.~chmidtiae Wollaston) is also common to the Canary and Cape Verde islands. Possibly this species was introduced accidentally in the CapeVerde islands. To point out the similarity existing among the different islands considered, I used the Baroni Urbani and Buser index and cluster analysis according to the average linkage method. I Fig. 3. - Dendrogram of the similarity among the islands. The result of this analysis is shown in Fig. 3. In this dendrogram we can point out three main clusters: - the first containing the Ihlas Desertas, Porto Santo, Madeira and the Palearctic region: - the second containing all the Canary islands; - the third, the most isolated, containing all the Cape Verde islands and the Ethiopian region.
TABLE 1 Flea beetle species in ever?; island considered Taxa: Phyllorreto cheironlii... Phyllorre~n consobrinn... Ph~llorrero dohrninnn... Phyllorreto laevicepr... Phyllorrero rufilarsir... Phyllorrero sublrrrn... Phyllotrero variipennis... Apkfhono crassipes... Aphlhono lnrvisrirna... Aphrhorro.sig,zolifrons... Heyrovrkyn orrovbidecerzs.. Heyrovskyo convexior... He~rovrkyo gornere,~sir... Hevrovskyo,norairri... Hryrovrkya occidrn~alir... Heyrovskyn oromii... Heyrovskyo pait,nnn... Heyrovsksa plrnifrons... Heyrovskvo Irisrrs... Heyrovskyo wollnsrorii... Longiforrus aeneus... Longilorsus irzcospicuus... Longilarsus ;soplexidis s. srr. Lungi~arsus isoplexidis persi-,nilis... Lon~itarsus jnndienslr... Lungi~orsus kleiniiperdo... Longilarsui lnxicornic... Lungi~nrrus madrrensis... Longirnr.su.s mrs,sersck,nidliae. s. sa.... Longirar.~~~ mer.ssrrch,nidlioe gornerensis... Longi1nrs~r.s ~nirri... Longilorsus nen,orus... Longitorsus nigroforciolus... Lungitarsr,.~ nubigenn... Longiroriu.~ ochrolrucus,. 51r.... Longirarsur parvu1u.v..... Lungi1arsu.s prolen.vis...
Follows: TABLE 1 Flea beetle species in every island considered Longilnr.vur ~lragulolus pollii l l s... Longirarsus ~irigicollis... Longiiorsus vilis... Epirhrir allordi... Pudogrico fuscicomir (3)... Monruro cht).ranrhrmi... Choerucnr,na f horlrn- Chortocuema sp. (5)....Mniophiloso,no loeve... Arroromur epilochnoides... Argo.w,nu.s obscuripennis... Diboliv rnognifica... Dibulio obrusa... Psylliodrs orrnulonr... P.sylliudrs amplicollis... Psyllioder ongurricrpr... Psylliuder chrrsocrpholo... * P.~ylliode.s rnoroccona... Psylliuder rorrota... P,?ylliode.~ urnbrotilis... MA: Madeira: PS: Porto Santo: ID: Ihlar Dewrtrs; PA: La Palma: HI: Hierro: GO: Gomera; TE: Tenerlfe; GC: Gran Canaria; LA: Lanzarote: FU: Fuerrevenrura: ST: San Tiago; SV: San Vincenle: SA: Santu Antao: SN: San Nicolau: SL: Sal; BR: Brava; FO: Fogo: PR: Paleartic region: ER: Erhiopic regnon. (I) Madar & Madar (1965) quoted thir rpecier in Porro Santo, while in Erber (1986) Longirnrrus codinni Madar & Madar is no1 reported. (2) Leonardi (1973) quoted lh~s species in Madeira. Surely the previous cimtlons for Madeira of Lunyiianur iycopi (Foudrarj are also to refer to L. ordirrolur (Foudns). In fact Wollasron noted that the specimens alttibuled by h~rn to L. iycupi had a stronger punc~ation: instead this features is charactenstic of L. urdinorur. This larc rpcctcr is not reported in Erber (1986). (3) Lundblad (1958) and Erber (1986) quoted also Pudogrico /u.scicomir (Linnaeur) in the Canar). Islands. I think they misinterpreted the abbreviation.ca. reponed in Winkler (1924-1932) understanding ir ar -1nrulae Canariae~ (=Can..) and not as *Caucarur- 14) Erber (1986) quoted a specimen of Clzor,ounrmri ch-. horrm.sis (Geofh-oy) collected in Madeira. This identification being doubrful. I will not eonrider this tara in the statirrical analysis. (5) S~herer (1986) cited the presence in the Cape Verde Archipelago of one species of Churrocnernn Stephenr roba ably new. In thir care as well. I will not consider thir rara in the s~acistical analysis.
In Fig. 4, the result of the principal component analysis, is reported. It is very similar to the result of the preceding cluster analysis. Looking at this graphic, we can still observe: - the flea-beetle fauna of the Madeiran islands has an evident paleartic origin; - the flea-beetle fauna of the Canary islands has a clear paleartic origin too, and actually it is sufficiently defined. Further confirmation of the paleartic origin of present flea-beetle fauna of the Madeiran and Canary islands, by systematic affinities of respective endemic taxa, is also, given: - on the contrary the flea-beetle fauna of the Cape Verde islands has a very clear Ethiopian origin. Fig. 4. - Principal component analysis anlong the islands At this point I will consider only the Canary islands. Looking at the dendrogram in Fig. 5, we can observe: - Lanzarote and Fuerteventura islands are very similar. These two islands are both characterized by: 1) shorter distance from the African coast; 2) lower altitude: 3) lesser pluviometrical values; 4) scarce vegetation and complete lack of forest;
- the western islands, as La Palma, Hierro, Gomera and Gran Canaria, are similar. They are characterized by: 1) longer distance from the African coast and a greater degree of ~oceanity*; 2) higher altitude; 3) higher pluviometrical values: 4) richer vegetation and presence of many woods, - finally Tenerife is more similar to the continent than to the other Canary islands. Regarding Tenerife, it has a higher percentage of faunistic elements of wide distribution (50%) (Fig. 2), of which one part was probably introduced accidentally by man and the other part caught. directly by the island itself during the dispersion phase of insects. With its wide area and its high altitude, Tenerife most probably is like an efficient -trap. especially for the flying insects. Moreover the presence of a relatively higher number of exclusive endemic species (6) in Tenerife, may Fig. 5. - Dendrogram of the similarity among the Canary Isands. CO = African continent. TABLE 2 Number and percentage of the farrnistic elements in each insular group Madeiran Is. Canary Is. Cape Verde Is Number of tala............... Macaronesic endemic elements Exclusive endemic elements... Wide distribution elemcnu....
Fig. 6. - Linear relationship between number of species and distance h-om the continent (Y = 0.005~X + 12.100). *.=a cnw) Fig. 7. - Linear relationship between number of species and area (Y = 0.007.X + 5.232)
::::I 33.3 be due to its environmental complexity, its wide area as well as its high altitude. Still regarding to the Canary islands. I considered the number of flea-beetle species present in eve[? single island, and the physical parameters of the island itself. For every island I considered the following measures: 1) its distance From the African coast; 2) its distance from Tenerife; 3) its area; 4) its maximum altitude. I applied the regression analysis to these data, looking for a possible relationship between the number of flea-beetle species present in every island (dependent variable) and four physical parameters considered (indipendent variables). The results of this analysis are the following (see Figs. 6-9): - the distance from the African coast seems to have no influence on the number of species present in each island (correlation coefficient is equal to 0.099); for the Canary islands the distance from the Africa coast is important only For the flea-beetle faunistic composition, ad we observed above; - the altitude seems a little more significant than the area; (the correlation coefficient for altitude is equal to 0.859, while the correlation coeficient for area is equal to 0.723); if these two last independent variables are considered together in the multiple regression analysis they show a very Number of raxa, percentage of the fnunisric elements and physical paramefers of the Catiary islands Number of taxa..................... Macaronesic endemic elements (C).. Exclusive endemic elements (%)..... Widedistfibutionekmnts,,. - 16.7 25.0 16.7 50.0 13.3 335 7 12 14.3 28.6 16.7 33.3 437 450 390 339 231 141 115 123 150 62 110 300 255 (1) The d~stances are measured from the center of evely island. (2) The values of the area of each tsland are those reported in Den~enn (1963)
39 RECRESSION T(InIYSIS - Canary Islands - TE Fig. 8. - Linear relationship between number of species and altitude (Y = 0.006.X + 2.483). Fig. 9. - Linear relationship between number of species and distance From Tenerife (Y = -0.053.X + 21.227).
TABLE 4 Matrix of the Baro~zi Urbani and B~rser index arnong the islands
Correluiion n~atrix utnong the is1u~zd.r T I D P S A PA H1 GO TE GC LA FU ST SV SA SN SL BR FO PR ER
significant relationship with the number of flea-beetle species present in each island (in this case the multiple correlation coeficient is equal to 0.954); - the distance of every Canary islands from Tenerife is vet important too (correlation coefficient is equal to -0.723). On the grounds of the above statistical information, I believe that TeneriFe represents, for the other Canay islands, a sort of csubcontinent. where a great part OF animal and vegetable specles arrive from surrounding regions and from where a part OF them can reach the other islands of this insular groups. This is possible if the wide area, high altitude and central position of Tenerife, is considered. Bloxol M. (1987) - I Chpsomelidoe Alticinne delle isul<, Cnnurie (CaleopleraJ. efragm. Entomol.n, Roma. 19. 339-362. DER\,ENK C. (1963) - Ler Canaries. Horizun~ de Frnnce. Paris, 166 pp. ERBER D. (1986) - Tile Chrv.so,nelidue of Madeira rbol. Mus. Mun. Funchals, 38 (174). 43-69. LEoNARDl C. (1973) -Note curologiclie e tossortorniche su nlcrsni Longilarsus can ~.irazione di due rpecie nuove per I'llalia (Coleoptera Chrycom~lidue),,,Atti Soc, ital. Sci. nar. Museo Civ. Stor, nat. Milanon, 114 (I), 5-42. LUNDBL4U 0. (1958) - Die Arthropodenfauna van Mndezm aoch den Erx.rbni.csen der Reise von Prof. Dr. 0. Lundblod Juli-Augur1 19.35. XXXV. Die Kirferfnunn der inrein Madeira, aarkiv for Zoologin. 11 (30). 461-524. MADAR J, and Maom J. (1965) - Britrag zur Systerno~ik und Zoogrographie eini~erpnloenrkrirchen Verrelrer der Subfamilie Holficinne (Cul., Chrvrotnelidne). 'Acra ent. bohemoslov.., 65 (2). 138-151. SCHERER C. (1 986) Die Halliciden der Kapverden icoleoplern: Chry.sotnelidne: Alticinne). scour, Firsch. Insf. Senckenberg,,, 81. 65-86. WlNKLEn A. (1924-1932) - Co~olagus Cclleop~erorurn regionis polornrc~icae. Wien, 1317-1352 (par pars Halticinae).