Bioges@aaphied History of the Neotropied adNeantmctic Simuliidae (Diptera)
'Institute Argontino de Investigaciones de las Zonas Aridas (MIZA),Casilla Cai-reo 507,5500 Mendoza, Ar- gentina. zDepartamento Cientifico de Entomologia, Musea de La Plats, Paseo del Bosquc sin, 1900 1,a Plata, Buenos Aires, Argentina
Ahtrack Using the lineages of taxa and their distribution on different areas of endcmism, and hy cladistic methodology appiiod to biogeography, we try to define the biutic history of the areas of endemism based on Neotropical and Neantarctic Simuliidae. Using tho information of the eladograms of nine Central and South American monophyletic supraspecific taxa of Simuliini and 16 areas of endeminm, we perform a Component Analysis with wmponent 1.5, using the assumption 2. A sccond analysis was mark using Biogeographic Parsi- mony Analysis; the data matrix was ilnalyzed with the prowam NONA. The ciadograms obtained show the pos- sible sequence of historic separation of areas of endemism, evidencing the presence of two large biota: tho aus- tral (Noantaretic) and the tropical (Neotropical),that were maintained in partial isolation. The areas of ende- inism of the austral biota are: Subantarctic. Central Chile. Pataeonian. Puna. Monte and Pamosm: those of the vicariant events that affocted the Simuiiiciae are sea inbvessions,iheemergence of the ~ndesand elimati~han&s.
Key words: Neotropical Region, Neant,arctic Rcgion, blackflies, cladistic biogeography, areas of endernism, biogeographic history
The similitude between hiotas from different This contribution uses the information avail- areas has called the attention of naturalists, who able on Simuliidae (Diptera). The Simuliidae is a have tried to explain their prohahie relations with well represented group in the Central and South a numher of hypotheses. Many of the hiogeo- America, with 12 genera and approximately 340 graphic schemes are based on the comparison species (Crosskey & Howard, 1997), most of which between different biotas of the world, proposing are endemic to this region (Coscarbn & Coscarbn dispersion hypotheses to explain the observed Arias, 1995). The Simuliidae occur in Central and similitude (Wallace, 1876; Simpson, 1964; South America from southern Mexico to Tierra Darlington, 1965; Raven & Axelrod, 1975). dei hego. They are distrihuted in a large num- The use of cladistic methodology in system- ber of environments rangingfrom sea level to 4700 atics (Hennig, 1965) and its application in hioge- meters of altitude. ography (Brundin, 1966; I-lennig & Wydozinsky, Numerous phylogenetic studies of the Cen- 1966; Rosen, 1978) have provided an operational tral and South American Simuliidae have been tool to analyze the distribution patterns (Platnick made in the last years (Coscarbn & Coscarbn- & Nelson, 1978; Nelson & Platnick, 1981; Arias, 1996,1997;Coscarbn & Miranda-Esquivel, I-Iumphries & Parenti, 1986). 1998; Coscarbn et al., 1996,19991, where possible Cladistic biogeography assumes a correspon- biogeographic infercnces are discussed for each dence between the phylogenetic relations of the one oithe groups analyzed. Coscarbn & Coscarbn- taxa and the relations of the areas they occupy, Arias (1995) identified the areas of endemism for and that organisms have evolved together with the Neotropical and Neantarctic Simuliidae the Earth. This would he why the "history of the which are used in this work, together with the areas of endemism" can he inferred startingfkom Simuliidae phylogenetic analyses, to obtain hy- the cladistic analyses of the several taxa com- potheses on the biogeographic history ofthis fam- posing a biota. ily in Central and South America. 120 Revista del Museo Argentina de Ci.encias Naturales, n. s. 3 12), 2001
MATERW AND METHODS in the humid galleries and piedmonts. It corre- sponds partly to the biogeographic provinces Selection of the areas named Sabana, Guajira, and Venezolana by The area under study comprises the Neotro- Cabrera and Willink (1980). Its their Pduna is pica1 Region and the Neantarctic Region of Cen- related to that of the Guiana and Uungas areas. tral and South America. These two regions are 4. Cerrado (CE):it is a large, diagonally dis- considered by many authors (Ringuelet, 1961; posed area in South America, located between Fittkau, 1969; Mkller, 1973) as only one region, Amazonia and the Mountains region of the Bra- the Neotrnpical. Nevertheless there are two bio- zilian southeast. It corresponds to what Cabrera tas with very dissimilar affinities (Jeannel, 1912; and Willink (1980) called Caatinga, Chaco, 1967; Monros, 1958; Kuschel, 1969; Lfalfter, Cerrado, and part of the Espinal. It is a tropical 1974; Porter, 1991; Roig-Juiient, 1992) which arid area with shrubby vegetation and xeric for- occupy areas that have been considered as dif- ests of Prosopis ('algarrobos') and Aspidosperma ferent biogeopraphic regions. One of them is the ('quehracho'). It constitutes a rich area of ende- tropical biota called Guiano-Brazilian (or rnism for insects (Roig-Jufient, 1998) related to Jeannei's Tnahresic, 1942, or Kuschel's Brasilic, the tropical fauna of Sgi~rthfunerica. 1.969) tha taxa of which are phylogenetically re- 5. Central Chile <%:II):iL includes the Chil- lated with other taxa from the tropical African can area from Coquimbc to Conception (between regions. Thc other biota in South America is the 30" and 37" S latitude), In its septentrional part Patagonian (or Jeannei's Paleantarctic, 2942, or the aridity conditions are higher (Coquimbo re- Monros' Neantarctic, 1958) with taxa more gion of Roig-Juiient, 1994), with small forests closely related to the taxa from other austra: along the creeks close to the coast. The areas, such as Australia, New Zealand, South meridionai part is characLerized by shrubby veg- Mrica, and numerous subantarctic islands. The etation (Cabrera & Wiliinh, 1980). Within this reson occupied by this austral biota of Soutlr Central Chilean area there are relicts of snban- America difSers greatly according to the differ- tarctic forests lilt. Campana) showing that the ent authors. Monriis (1958) considered only the Antardandean forest was much more extended. area occupied by Nolhofagus forests. Most of the 6. Desert (DE): this area extends along the authors considered that it constituted all the area Pacific coast fiom 5" S latitude in Ecuador to south of 30" S iatitude (Jeannel, 1967; Porter, 30" S latitude in Chile. The precipitation is 1991; Crisci el al., 1991). A third position scarce, getting to zero in Arica and Iquique (Morrone, 1996) regards it as the areas south of (Cabrera & Willink, 1980). It has a warm cli- 30" S latitude, including also the Andean regions mate and the vegetation on the coast is almost north of this latitude up to Colombia (PBramo). null, with some vegetation on the Andean slopes The 16 areas of endemism proposed (Fig. I) or bordering the watercourses descending from have been delimited by Coscardn & Coscarbn- the mountains. The region possesses numerous Arias (1995) hased on the distribution ofthe endernisms adapted to the extreme aridity con- species of Simuiiidae (Simulium and eight of the ditions. Its Pauna is quite related to that of Cen- genera of Prosimuliini). The acronyms of the tral Chile and of the Puna. The same as in the areas and their location are given in Figure I. Central Chile area, there are relicts of the I. Amazonia (AM): it is a large area repre- Antarctandean forest like that of Fray Jorge. senting the Amazon basin, It corresponds to the 7. Guiana (GU): this area includes the rain forest. It is constituted by a series of recent Guianan massif, with the Itoraima and Parima fluvial sediments and the remnants of the moun- Tepuis, together with the Orinoco plains tains of' the Brazilian and Guiana shields (CoscarOn & Coscarbn-Arias, 1995). It is an un- (Cabrera & Willink, 1980). dulated plain alternating with plateaus reach- 2. Northern Andes (NA): this is the bio- ing almost 1600 meters of altitude, with nearly geographic province named Piramo by Cahrera vertical walls. The vegetation is characterized by & Willink (1980), located in the hidean region savanna and gallery forests along the rivers. The of Venezuela, Per%, Colombia and Ecuador he- fauna of the region has been considered by some tween 2500 and 3500 meters. authors as very ancient (Jeannel, 1942,1967); it 3. Caribbean (Cia): this area occupies the was called by Ihering (1927) the Arcbiguyana plains and mountain ranges of northern Colom- Region. bia and of Venezuela north of the Orinoco, and 8. MesoamericmMomtains (lilM): it com- the Lesser Antilles. It comprises difrerent envi- prises the highland forests of the Mexican and ronments with xerophytic vegetation, and jungle Central American mountains as far south as Roig Jurie~zt& Cosca?.dn: Biogec graphical history of Simuliidae 121
J '3,, ,,P .P%, ( 11. Panrpean (Pill): this is a plain with a few hills, with steppe vegetation, and moderate precipitation diminishing toward the west. It corresponds to the biogeographic provinces of the Pampa and part or the Espinal of Cabrera & Willink (1980). Its Eauna is considered to be closely related to that of the Brazilian south- east (Rinselet, 1961). 12. Patagonlam (PA):it comprises the aus- tral part of Argentina, from the Rio Colorado to Tierra del Fuego. The Patagonian area as it is here considered includes the meridional part or the Monte of Cahrera & Willink (1980) and their biogeographic province of Patagonia (CoscarSn & Coscar6n-Arias, 1995). The steppe is the pre- vailing vegetation and the precipitation is di- minished by the Andes mountains. The climate is temperate cold. Several elements of the fauna are related, like those of the Subantarctic area, with taxa from Australia and New Zealand. Nevertheless, in contrast to what occurs in the Subantarctic area, the taxa present in Fig, I. Areas of endemism: AM, Amazonia; NA, Patagonian in many cases reach the Pampean, Northern Andes; CA, Caribbean; CE, Cerrado; Puna, and Monte areas. CH, Central Chile; DE, Desert; Gu, Guiana; MM, 13. Puna (PU): it is located in the Mesoamericari Mountains; MO, Monte; PC, Pa- Altoandean plateau of Argentina, Chile, Bolivia cific; PM, Pampean; PA, Patagonian; PU, Puna; and Peru, between 3000 and 5000 meters of al- SE, Mountainous Region or the Brazilian South- titude. It is a strip that extends from 10' S lati- east; SIJ, Subantarctic; YU, Yungas. tude (Perh) to 33 V in province of Mendoza (Ar- gentina). It is a terrain of xeric conditions, with a marked thermal amplitude. The vegetation is Panama. It coincides with the biogeographic that of the steppe. The area here considered in- province called Mesoamericans de Montafias by cludes the Punefia, Prepunefia and Altoandina Cabrera & Willink (1980). The fauna of'the area biogeographic provinces of Cabrera & Willink has three types of elements: endemic, neotropicai, (19801, since none of the species of Simuliidae and nearctic (Halffter, 1964), of which the most shows the division proposed by these authors ancient elements would be fiom the beginning (Coscar6n & Coscar6n-Arias, 1995). Tho Pnna ofthe Cenozoic. fauna has great affinity with the Patagonian 9, Monte (MO): this area is located in west- (Morrone, 1996). central Argentina, from the south of Mendoza 14. Mountainous Region of the Brazil- and northeast of La Pampa to the south of ian Southeast (SE): it is distributed in Argen- Catamarca. The area here considered does not tina (Misiones), east of Paraguay and southeast include the austral part of the Monte of Cabrera of Brazil. It comprises a large region of the & Willink (1980). It is a semiarid area with tem- Planalto and of the Sierras of the Braailian At- perate climate and shrubby vegetation. The lantic coast. The vegetation are savannas and fauna is closely related to that of the Chaco tropical forests. It coincides with the Paranense (Stange et ol., 1976). and Atlantica biogeographic provinces of 10. Pacific (PC): includes the peripheral Cabrera & Willink (1980). lowlands under 1000 meters of altitude in 15.Subantarctic (SU): it corresponds to the Mexico, Central America, Antilles and a west- area of distribution of the Nothofagus forests ern strip of' Colombia and Ecuador. It is a extending from Concepciirn (37" S) to Tierra del rainforest and coincides with the provinces Fuego, including the oriental flank of the Pacifica and Carihe of Cabrera & Willink (1980). Patagonian Andes of Argentina. The climate is It is one of the most humid regions of the planet cold temperate with great humidity. The taxa and the temperature is constant almost all year- rrom this area have the greatest affinity with round. those from Australia and New Zealand (Morrone 122 Reursta del Museo Argentzno de Clenczas Naturales, n. s. 3 123, 2001 et al., 1994; Roig-Jufient, 1994) and in America genus with 27 species, distributed from southern they are restricted only to this region (Roig- Patagonia to Ecuador through the Puna. Some Jufient, 1998). species also occur west of the southern Andes, as 16. Yungas (W):it lies on the oriental slope in the Subantarctic, Central Chile, and Desert of the Oriental Andes between 500 and 3000 areas. This suhgenus is largely sympatric with the meters of altitude. It is iimited by the Amazonia Prosimuliini. Cladogram of areas, Fig. 4. and Cerrado areas to the west, extending from 3. S. (Heemionelha) Enderlein (Coscar6n, Colombia to the province of Catamarca in Ar- unpublished). This suhgenus comprises 20 spe- gentina, it coincides with the area proposed by cies distrihuted mostly in Mesoamerica, the Cahrera & Willink (1980) for the homonymous Carihhean and tropical regions of South biogeographic province. It is characterized by a America, excepting Amazonia (Coscarhn & cloudy forest, with a temperate climate which Coscarhn-Arias, 1995). Cladogram of areas, Fig. turns cold in the highest zones. 6. 4. S. (Psaroniocompsa) Enderlein (Coscarbn Selection of the taxa & Coscar6n-Arias, 1997). It includes 15 species, Nine monophyletic taxa of Simuliidae for distrihuted in disjunct areas: one austral area from
(Coscarhn & IMiranda-Esquivel, 1998) and the the Pacific region, from Panama to Ecuador. Cia- remaining eight are suhgenera of Simulium dogram of areas, Fig. 8. (Coscarhn & Coscar6n-Arias, 1996,1997; Coscav6n 5. S. i.Ina,qureEium) Cosar6n & Wygodeinsky et al., 1996, 1999). The ciadograms have been se- (Goscar611 & Coscar6n-Arias, 1997). This suhgenus lected according to the criteria proposed by has 13species. Its distribution area has the shape Humphries (1981). Therefore the cladograms of of an arc, from Panama and Venezuela to the Bra- the chosen areas have to comply with the follow- zilian southeast and northeast of Argentina, very ing: a) at least three should have areas of ende- similar tothat of S. iPsaroniocompsa), but with- mism from the studied area; h) at least one of the out reaching Patagonia. Cladogram of' areas, Fig. terminal taxa in the area should he endemic of 10. one area. This is hecause if all the taxa occupy 6. S. (Chirostilbia) Enderlein (Coscarh & two or more areas of endemism, the information Goscar6n-Arias, 1997). There are 12 species in provided by the ciadogram is irrelevant. this suhgenus, most of them in the Brazilian For each lineage a ciadogram of the funda- southeast, although its distribution reaches out mental areas was obtained, the terminal taxa of this area and is similar to that of S. being replaced by the area(s) of endemism they ilnaequalium). Cladogram of areas, Fig. 12. occupy (Figs. 2, 4, 6, 8, 10, 12, 14, 16, and 18). 7. S. (Eclemnaspis) Enderlein (Coscar6n, In some cases the area cladograms have been unpublished). It comprises 39 species (Coscarh simplified, as for example in Simulium & Coscar6n-Arias, 1995) which are wideiy dis- iPsilopelmiai, which has entire clades occupy- tributed in South America, and on a smaller ing only one area and provide the same infor- scale in Central America, from the Pacific and mation. the Caribbean to Patagonian. Cladogram of ar- I. Prosimutiini (Coscarhn & Miranda- eas, Fig. 14. Esquivel, 1998). Three genera endemic to South 8. S. (PsiZopeZmia) Enderlein (Coscar6n et and Central America, comprising 74 species, are al., 1996). It comprises 21 species distributed considered here (Coscarbn & Coscar6n-Arias, from Mexico to the north of Colombia, Venezu- 1995). The tribe includes Cnesia Enderlein (re- ela, Guyana, Suriname and French Guiana stricted to theNothofagus fbrests),Pedrowygomya (Coscar6n & Coscar6n-Arias, 1995). One species Coscar6n & Miranda-Esquivel (distributed in the (S. iP.) escomeli Rouhaud) is distrihutcd in the Puna), and Gigantodm Enderlein (distributed from coastai desert of Peru and north of Chile. Cia- southern USA to the south of Patagonia). The ter- dograrn of areas, Fig. 16. minal units for the genus Gigantodm used by 9. S. (Simulium) Latreille (Coscar6n et al., Coscar6n and Miranda-Esquivel(1998) correspond 1999). This cosmopolitan suhgenus has 25 to the four species of thecigantodm cortesii group neotropical species that constitute a monophyl- and one species from each of the remaining groups etic group (Aspathia Enderlein), and are distrih- of species. Cladogram of areas, Fig. 2. uted in Central America, the Carihhean and the 2. SimuEium (Pternaspalha) Enderlein septentrional region of South America (Coscar6n (Coscarhn & Coscar6n-Arias, 1996). This is a suh- & Coscar6n-Arias, 1995), with a species endemic Rozg Jurient & Coscardn Bzogeographrcal hrstory of Szmulzzdae 123
CIl IlE SU
CHNAbMPASUW
1- 1- CII NA PA PU SU
Prosimuliini
Simulium (Pternaspathaj
-
M PC - ?--- CACEOUSEW
CE SE SE PC NA AMCEGUNASE AM CA CE GU SE GU AM Simulium flemicnetha) Simulium (Psaroniocompsai
Figs. 2-9. Figs. 2-3, Prosimuliini: 2, fundamental area cladogram; 3, Nelson consensus tree from the 138 resolved area cladograms. Figs. 4-5, Simulium (Pternaspathai: 4, fundamental area cladogram; 5, Nelson consensus tree from the 151 resolved area cladograms. Figs. 6-7, Simulium (Heminecta): 6, fundamental area ciadogram; 7, Nelson consensus tree from the 91 resolved area cladograms. Figs 8-9, Simulium (Psaroniocompsa): 8, fundamontal area cladogram; 9, Nelson consensus tree from the 153 resolved area cladograms. 124 Revzsta del Museo Argentzno de Czenczas Naturales, n. s 3 121, 2001
GU AM Cf GU SE SE CE PM Sl? CE SE CA CE GUPC FYI PU SE CB GUSE
CACEGUPASEYU
Simulium (lnaequalizimj Simulium (Chirostilbia)
AMCEGUPMSE CADEMONAPCPMYU 1 NA PC CAGUMONAPAPCPMPUSE NA - AM CA CE GU SE NA PC CA GU W CE DE NA DENAPCPUYU flC~GUNAPCPUYUMONAPCPUYU CANAPCSSCYU P(:
14 ___ CAGUPCW psm
MM PC Simulium (Psilopelmiai Simulium (Eetemnaspisj
Figs. 10-17. Figs. 10-11, Simulium (Inaequaliumi: 10, fundamental area cladogram; 11, Nelson consensus tree from the 459 resolved area cladograms. Figs. 12-13, Simulium 1Chirostilbai: 12, fundamental area cladogram; 13, Nelson consensus tree from the 4 resolved area cladograms. Figs. 14-15, Simulium (Ectemnaspis): 14, fundamental area cladogram; 15, Nelson consensus tree from tho 267 resolved area cladograms. Figs. 16-17, Simulium 1Psilopelmiai: 16, fundamental area cla- dogram; 17, Nelson consensus tree from the 15 resolved area cladograms. jgrnl~hicaihistory of Simuliidne 125
cladograms. For this, each component is turned into binary numbers. The absent areas have been neutrally treated, as non-comparable and not as absent, as proposed by Wiley (1988). For each essential area cladogram each component has been codified only once, although it may he present more than once. The basal component is not considered in the matrix when it replicates a component already codified in the cladogram. These matrices have heen analyzed with the Wagner algorithm using the program NONA (Goloboff, 1993).
RESULTS
From the analysis af the different sets of re- solved area cladograms it was not possible to obtain a cladogram shared by all the cladograms. Simulium (Simulium) Likewise, it was not possible to obtain a general cladogram of areas shared by more than two or three lineages. Nevertheless, eight of the treated lineages (exception made of S. (Ectemnaspis)) shared a reduced area cladogram (Fig. 20), with only eight of the 16 trcated areas, Two reduced ciadograms of 11 areas shared by six lineages were a!so'found (Fig. 21). Tbe impossibility of finding shared area cladograms is mainly due to the different position presented by certain areas of endemism in the resolved area cla- dograms, as for example Desert (DE), Yungas Figs. 18-19, Sinziliium iSinzuliirm): 18, funda- (Ili), Cerrado (CE), Caribbean (CAJ, and North- mental area cladogram; 19, Nelson consensus ern Andes (NA). tree from the 32 resolved area cladograms. The codification of the components of the essential area cladograms permitted the obtention ofa basic data matrix (17 areas by 115 components) (Table 1). Its analysis resulted in a to the Desert area Ciadogram of areas, Rg 18 cladogram (Fig. 22) 233 steps long, with a con- sistency index of 48 and a retention index of 56. Data analysis This cladogram is compatible with that ohtained Two different techniques were used to ohtain through the intersection of eight of the nine lin- the general ciadogram of areas. All the resolved eages used (Fig. 20). area cladograms were obtained under assump- The three general area cladograms (Figs. 20- tion 2 from the essential area cladograms (Figs. 22) show the possible sequence of the historical 2, 4, 6, 8, 10, 12, 14, 16 and 18) using the pro- separation of the areas of endemism in South gram Component 1.5 (Page, 1989). The sets of America. The discussion of the relations between resolved area cladograms are represented areas will be mainly based on the general cla- through Nelson consensus trees (Figs. 3, 5, 7, 9, dogram of areas obtained through parsimony 11,13,15,17, and 19; the number of cladograms (Fig. 221, hut will also take into account the two obtained is indicated above each figure). The re- cladograms obtained through intersection (Figs. solved area cladograms were compared to ohtain 20-21). The separation sequence of the areas is general cladograms of areas through intersection, represented graphically in Figure 23. that is to say shared cladograms of areas (Crisci In the cladogram obtained by parsimony the et al., 1991). first area to he separated is tho Desert. The sec- A second method used was Wiley 's (1988) Bio- ond separation clearly marks a division between geographic Parsimony Analysis, where a data the Neotropical and the Neantarctic Regions (Fig. matrix is built based on all the essential area 23b), which is also shown by the general cla- Xeaistu del Museo Argentina de Ciencias Natarules, n. s. 3 (21, 2001
2 dograms of areas ohtained through intersection -f (Figs. 20-21). The cladograms ohtained through oj 2 - iuterso~!ionand that obtained by parsimony are 5 totally conpent regarding the separation ofaus- trai areas (Neantarctic). The sequence shows a first sstiparation of the clade (PM, MO) (Fig. 23cj 7 ...,: and later the separation of the remaining areas ") .$ in two clades (PA, PU) and (CE-I,SU) (Fig. 23d). . ... The first Neotropical area to separate in the r< three general cladograms of areas (Figs. 20-22) 2 is Amazonia (Fig. 23e). The subsequent separa- -,.* tion sequence in the cladogram obtained by par- i! simony is the following: first the two more aus- 'I 'Iiij tral areas (CE, SEj separate (Fig. 230 from the 0 other septentrional areas, the Guiana area hi- .-"a, .. lows (Fig. 23g), then the Caribbean area, and a::oj ,-i finally the Yungas area separates from the
M 5 septentrional areas (fig. 23h). For the Neotm- - pica1 areas, one of the cladograms ohtained by +g 6 intcrsection (Fig. 21) does not present the same separation sequence, as in this ciadogram the t: Guiana area may he the sister area of Amazonia, -0 0 *0 or be the second to separate. m ._-. ";"" ... DISCUSSION oj ;- +" ...3 The close relations of some groups Si?nuliam species and of Prosimuliini with taxa from other continents show that these Diptera were already present and widely distributed in Gondwana before its fragmentation. The exist- ence of fossils gives evidence that some genera already existed at least in the Middle Jurassic in Asia and at the beginning of the Cretacic in Australia (Crosskey, 1991). This wide distribu- tion has been also stated for other insects like Ichneumonidae (Porter, 1991) and Carabidae (Roig-Jufient, 1998). During this period of the Mesozoic the climate was warm and not sea- sonai; this climatic unifbrmity would have al- lowed a wide distribution of organisms in the Gondwana (Porter, 1991). The general cladogram of areas shown in fig- ure 22 shows the possible sequence of vicariant events undergone by this Gondwanic biota. Nev- ertheless, it is very difficult to try to relate the known geological events that may have had vicariant effects with the sequence of the cla- dogram (Fig. 22). In the case of insular isola- tions (for example, New Zealand) it is easier to assign a vicariance, as the possibilities for dis- persion are reduced. In continental areas it be- comes much more difficult. The vicariant eflects within continents may finish when harriers dis- Rozg JuRent & Goscaron: Biogeographical history of Simuliidae 127
. 1313
PU
22
€ M?? Figs. 20-22. General area cladograms. 20, reduced general area cladogram of eigth areas of endemism shared by eigth lineages. 21, reduced gen- eral area cladogram of 11 areas of endemism shared hy six lineages. 22, general area cladogram obtained by parsimony.
lowed by the establishment of new endemic spe- tures. Plate tectonics connected South America cies in more restricted areas; should barriers dis- with North America and separated it fi-om Ant- appear, the dispersion of the descendant species arctica. The rising of mountain chains, mainly would occur to produce forms again widely dis- the Andes, caused lower precipitation regimes tributed, that would he affected hy other (Shmida, 1985). Finally, this period was charac- vicariances (Roig-Jufient & Flores, in press). terized by numerous marine regressions and This is what has happened in South America, transgressions (Volkheimer, 1971) that in many hecause since the Cretacic the continental area cases occupied most of the austral region of South has undergone large transformations (Tarling, America. 1980; Uliana & Biddle, 1988). During this pe- The analysis of the distrihution of the spe- riod there were great climatic changes. Suhtropi- cies of the treated groups shows that redundant cal fbrests existed in the Paleocene in Santa Cruz distributions exist in many cases. They might (Argentina), with a climate that remained warm he owed to the installation of harriers that later temperate in the Eocene, temperate during the on disappeared, thus allowing for a new Oligocene and Miocene, and turned to cold in the cosmopolitism of the newly formed groups. This Pliocene, especially when a cold water current would not be strange for South America, as most started in the circumantarctic sea, producing the of the harriers from the end of the Mesozoic and cooling of South America and an expansion of from the Cenozoic were continental seas that open environments such as savannas and pas- underwent great oscillations (ingressions and 128 Revista. del Museo Argentino de Ciencias Naturales, la. s. 3 121, 2001 regressions) through time (Volkheimer, 1971), barriers, like epicontinental seas, had disappeared. causing a new cosmopolitism of the species when It has been postulated that for the Eocene there they disappeared. existed in austral South America two biogeo- Considering the above, we try to correlate a graphic units that cvoived under different climatic geological event with a vicariant event shown conditions according to their difkrent latitudinal by the three general arca ciadograms (Figs. 20- locations (Pascual & Ortiz-Jaureguizar, 1990; 22). This separation involves the two large Pascual et al., 19961, one located north and the Gondwanic biotas of South America, the Neo- other Palagonic, connected to Australia. The lat- tropical and the Neantarctic. The division of the ter union lasted probably until the Middle Eocene Gondwanic biota has been proposed for the Up- when Antarctica underwent a great cooling pro- per Cretacic (Jeannel, 1942, 1967; Halffter, cess with the formation ofthe ice cap, and Drake's 19741, when a marine ingression occurred dur- passage opened (Pascual, 1984). Later on, during ing the Maastrichtian and lasted until the Pa- the Middle Tertiary the rising of the Andes and leogene, separating South America in two parts the consequent desertification, plus the marine (Uliana & Biddie, 1988) (Fig. 23b). This sea made Paranense ingression of the Middle Miocene kept its entrance through eastern Argentina and the Neantarctic region isolated fiom the Brazii- reached Bolivia. Another sea tongue was a nar- ian region along the east and the north (Porter, row Patagonic sea connecting the Atlantic and 1991). the Pacific, thus isolating the Austral and Prin- Even though the separation mentioned above cipal Cordilleras until the Miocene (Carnacho, is clearly shown by the three cladograms (Figs. 1967). South America was thus divided in two, 25-22), the first area to he separated in the cla- one austral region connected with Australia hy dogram obtained by parsimony is the Desert way of Antarctica, and a tropical region inter- (Figs. 22 and 23a). There are no geologic data mittently connected with North America, and that permit the supposition of a very early sepa- connected to a lesser extent with the Afiican ration of this area with respect to the rest of South tropical region. This vicariant event between the ~merica.'koig-Jufient & Flores (in press), in their two regions of South America explains that aus- ana1y;is of the eremic areas of austral South tral groups like the Prosin~uliiniare closely re- America, postulate that the area of the Atacama lated to the taxa of the remaining austral areas desert is one of the first to separate from the re- (Wygodzinsky & Coscar611, 1962; Coscar6n, maining arid areas, even though its age would 1985) and not to thc other South American he between the end of the Eocene and the begin- Simuliidae. The Maastrichtian vicariant event ning of the Oligocene (37 Mya). This would con- also affected the genus Simulium, widely dis- tradict what has been obtained through the par- tributed in thc Gondwana, separating Simulizim simony analysis (Fig. 22), according to which the 1Pternaspatha) and the Sirnuliurn "blanca,si Desert would have to be earlier than the group", with a neantarctandean distrihution, Maastrichtian sea from the Cretacic. Even from the remaining subgenera with a Ncotropi- though Shmida (1.985)proposes that the Atacama cal distrihution. Most of the other subgenera of is an ancient desert because of its high diversity Silnulium are Neotropical and have a closer re- of unique forms, he does not consider it to be of lation Lo Ethiopic taxa Iike S. (Tlzyrsopelmaj, S. Mesozoic origin but from the Cenozoic. The ori- (Anasolenj, S, (Ecteternnaspisj and S. gin of this desertic region constitutes an enigma. (Ponzeroyellumi (Crosskey, 1991). There also In some groups like the Ichncumonidae there exists a relation to some subgenera of Sirnuliurn occur Nearctic elements and there is no ende- from the Nearctic region, like S. ifiarleaj, S. mism of Neantarctic groups (Porter, 1991). This (Psilopelmiai, S. iSimulium), and the S. is not the case for the Simuliidae, where in the (Ectemnaspis) dinellii group, all of them largely Atacama region there appear Neantarctic developed in Mexico and Central America, and endemisms as well as Neotropical or Nearctic which through transitional bridges passed to elements that could have arrived before North South America and/or viceversa, as it is known America separated from South America in the that has occurred in other groups of insects Cretacic, coinciding with the rupture of (Halffter, 1976). Gondwana and the separation of Africa. These two faunal groups, the Neotropical and It is possihle, nevertheless, that the basal lo- the Neantarctic, were not only separated by geo- cation of the Desert (DE) in the parsimony analy- graphic barriers but, Iike vertehrates and other sis could be owed to a methodological error. Six organisms, continued isolated oven though the from the nine analyzed taxa of Simuliidae have Roig Jurient & Coscardn: Biogeographical history of Simuliidae 129
Fig. 23. Sequency of vicariant events, a) separation of Desert; b) separation of Neantarctic and Neotropical biotas; c) separation of Pampean and Monte areas; d) separation of Patagonian and Puna from Subantarctic and Central Chile areas; e) separation of Amazonia; f) separation of Cerrado and Mountainous region of the Brazilian Southeast; g) soparation of Guiana; h) separation of Yungas and Caribbean from the rcmainig Neotropical areas. 130 Revzsta del Museo Argentino de CLC species distributed in this area. None of the es- much alike Wolkheimer, 19711, thus showingthat sential cladograms show the Desert area as a the other Neantarctic areas were still connected. hasal one (Figs. 2, 4, 6, 14, 16, and 18). The re- This early separation of areas east of the Andes solved area ciadograms of these taxa show that would be in agreement with that proposed by in most cases the location is apicd (Figs. 3, 5, 7, Axelrod et al. (1991) for the Monte, which would 17, and 19). Differences are in that the have originated in the Miocene. In spite of this Neantarctic taxa from the Desert are more closely origin in the Middle Miocene, the existence of a related to austral taxa (Figs. 3 and 5), while the semiarid hiota has been mentioned for west cen- Neotropical taxa fiom the Desert (Figs. 7,17, and tral Argentina for the end of the Eocene 19) are more closely related to taxa with higher (Volkheimer, 1971; Axelrodet al., 1991). The sepa- diversity in septentrional areas from the ration of the areas belonging to this first clade, Neotropic. When codifying the components he- Monte and Pampean, could he owed to the rising tween Neantarctic and Neotropical taxa they of the Andes (reaching 2000-3000 meters of alti- would he contradictory, therefore the simplest tude) that created the arid conditions that origi- position is the hasal. nated the Monte (Axelrod et al., 1991). The rising of the Andes also isolated the clade EvoBution ofthe A~uskrialAreas (Neisntxctie) formed by the Puna nnd Patagonian areas from The area considered as Neantarctic coincides the clade formed by the Central Chile and Suh- greatly with that proposed hy Jeannel (1967) and antarctic areas. As mentioned hefore, in the hy Crisci et al. (1991). Many of the areas of en- Lower Miocene the fauna of Patagonia and Chile demism presently considered as Neantarctic, was the same (Pascuai et al., 1996). In the especially those in southemtern Argentina, were Middle Miocene the most important Quechua occupied by the Neotropical hiota until the phase occurs (1411.5 Mya); it had already begun Eocene (55-50 Mya) with a subtropical warm at the end of the Lower Miocene with the rising temperate climate (Volkheimer, 1971). Until that of the Puna, reaching an altitude of 3000 meters. period the Neantarctic hiota was restricted to the This rising of the Andes together with the de- austral region of Patagonia. It was only in the sertification of the oriental areas was the Middle Eocene (40-45 Mya) that the expansion vicariant event that separated geographically of the Neantarctic elements occured, up to lati- and climatically the Chilean areas from the tude 25"-30" S, in Sierra do Mar in Brazil, and Puna and Patagonian areas. According to ascending through the Andes up to Ecuador Jeannel (1967) the suhantarctic forest was dis- Woikheimer, 1971). This was probahly caused tributed to the east over most of Patagonia un- by a climatic change, with a lowering of the tem- til the Miocene, and suffered a great regression perature (Axelrod et al., 1991; Pascual, 1984) because of the conditions of aridity. because of the opening of a passage south of Tas- The separation between the Patagonian and mania that produced a glaciation, thus causing Puna areas was caused by the progressive rising in South America the displacement of the suh- of the latter. In the Upper Miocene, Puna and tropical forests to the north, dismemhering the the Sierras Pampeanas, the mountains presently Protochaco. Therefore, the vicariant events that limiting the occidental border of the Chacoplain, separated the austral areas must have occured began to rise (Harrington, 1968). The rising pro- after the Eocene, since when they have occupied duced and arid region west of the Sierras all the area where they are presently found. Pampeanas and sierras subandinas, region which The first separation (Figs. 20-22, 23cj is the at that time showed conditions similar to the clade formed by the Monte and the Pampean present Chaco forests (Pascual et al., 1996). areas. This would not be strange, as the central With regard to the separation of the Central part of Argentina has been isolated repeatedly Chile area from the Nothofagus forests (Suban- hy marine ingressions. During the last ingres- tarctic), a very recent division (Quaternary) has sion in the Lower Miocene (20 Mya) the south- been proposed, caused by the Bio Bio Barrier ern Bnenos Aires province and the central part (Solervicens, 1987). There exist relictual areas of Argentina constituted large islands sur- in Central Chile with Nothofagus forests (La rounded by an epicontinentai sea (Uliana & Campana) and other suhantarctic species in the Biddle, 1988). This ingression totally isolated Coquimho region (Fray Jorgej, proving that the these two areas from the rest. During this same distribution of the Neantarctic forests has under- period, Lower Miocene (25-15 Mya), the gone a large retraction. It is very probable that Patagonian vertebrate fauna was connected with the conformation of the biota of central Chile can that from Central Chile and the two were very be explained by arid conditions produced by the Rozg Ju~ient& Coscardn: Biogeographical history of Simuliidoe 131 installation of Humholdt's cold current, which occupicd all the Amazonas and Orinoco basins impedes precipitation in the continent (Porter, and persisted until the Pleistocene (Frailey el al., 1991). This would he coetaneons with the 1988). Quechua phase (14115-11 Mya) and would not be The areas of Cerrado and the Brazilian as recent as proposed by Solervicens (1987). Southeast have kept in contact for a longer time and in many cases there exist large transition Evrsllalion of the septentricnal arcas of en- areas among them, owed in many cases to the demism (Neolropicl riparian forest entering fiom the Brazilian The sequence of the separation of arcas in Southeast to the Cerrado. It is very likely that the Neotropical region as per the cladograms is these two areas have undergone repeated 8ize very different from that posed hy Cracraft & fluctuations, changing their boundaries because Prum (1988). For these authors the first separa- of climatic changes. Coscar6n & CoscarCin-Arias tion is constituted by the Mesoamerican region (1 997) proposed avicariant event for the Cerrado- (Choco and Central America) from the South Sierras of the Braeilian southcast because of the American Nentropical areas. in our case this change in sea level in the Lower Oligocene, when separation is among the last ones. Other infor- the sea level rose more than 200 m, as well as in mation also points toward a later formation of' the Miocene when it rose more than 100 m. the Mesoamerican fauna. Haffer (1967a, b, 1969, The separation of the remaining areas must and 1975) poses that a biota exchange occurred have been cai~sedhy the progressive rise of the around the no~.thernmostregions of the Colom- Andes and the lowland areas east of the Andes bian Andes, where during the Quaternary gla- and north of the Bolivar Geos.yncline. The Carib- ciations the lowering of'the sea level gave way to bean (Fig. 23h) was the first area;o separate. This an exchange. HalFter (1978) considers that an area was occupied during long periods hy marine earlier exchange, that permitted the entry of ficies, from the Eocene to the Lower Oligocene, South American fauna to Central America, oc- and increased in surface and altitude during the curred during the Miocene. Later, but in the ~pperPligocencand Miocene, hut remained iso- same period, a most generalized faunal exchange lated from the rest of Sonth America (Raven & between Central and South America occured, Axelrod, 1975). The rising of the Sierra de Perija, especially through the lowlands (PC), after the the Oriental Cordillera in Colomhia and the Cor- (brmation of the Tehuantepec isthmus, and the dillera de Morida in Venezuela, occurring at the more recent Ibrmation of the Panama isthmus. end of the Oligocene and the Miocene, must have According to our study the first area to sepa- had a vicariant role between Guiana and the Car- rate is Amazonia (Figs. 20-22,23e). It is not pos- ibbean (Cracraf't & Prnm, 1988). sible to assign a vicariant event for the isolation From the remaining areas the next to sepa- of this area from the others. The scparation of rate was the Yungas area. This was largely owed Guiana from the other areas is conflictive. This to the rise of the Andes and to the appearance area could have separated as a clade together of environmental conditions favorablc to the with Arnazonia (Fig. 211, and would be the sis- cloudy forest. Hence, an ecological factor has ter area to the remaining Neotropical areas (Fig. been the vicariant element. This oriontal region of the Andes had a humid, yungas type forest already during the Tertiary from Ecuador to Jeanne1 (1942) and Halffter (1974) proposed that Chile (Porter, 1991). This area surely has had Guiana is one of the three biotic units of the an important role in the dispersion of the Cen- Upper Cretacic. This assumption would he con- tral American faunas toward the south. The rise sistent if Guiana were a basal area like of the Andes also created the separation of the Amazonia (Fig. 21). It is very probable that this other areas, one with extreme conditions due to area has a very ancient origin, as there were altitude, the Northen Andes (NA), from the marine ingressions from the Lower Eocene plains of the Pacific area and from the (Senonian,,Marajo-Blen trench) to the Lower Mesoamerican Mountains. Miocene (Paramaiba Basin) that totally isolated the Guiana area from the Brazilian northeast CONCLUSIONS (Harringtnn, 1968). The vicariant role could he assigned to the hypothetical Amazonas lake that The entire analysis of the phylogenies of the formed when the connection with the Pacific different groups of Simuliidae clearly shows that through Ecuador was closed at the time of the the fauna existing in South America during tho iise ofthe Andes in the Middle Miocene. This lake Cretacic suffered a main vicariant event. This 132 Reulsta del Museo Argentno de CL~?ncias Naturales, n. s. 3 (21, 2001 event can he correlated with a large marine Bmndin, L. 1966. Transanttaretierelationships and their ingression.This event split the ancestral biota simificance. as evidenced bv chironomid midges. in two large biotas with different affinities, the >\I,). \I>>:3 LT.sp.: :,.h~.t!~.!~~.: !: !tt<.I,b: Neotropical related to tropical African fauna, .to.l .\i,~tc::~.mILL .i:xi r.~.,
end of the wodd. Distribution and history ofthe fir 143-153. southern lifelendland with assesment ofcontinenla1 Morrane, JJ. 1996. The biagoographical Andean subre- drifl, Cambridge, Mass., Harvard University press. gion: A proposal exemplilied by arthropod taxa 236 pp. (Aracl~nida,Crustacoa, and Hexapoda). Neolropica Fittkau, E.J. 1969. The fauna of South America. Bio- 42: 103-114 geography and Ecology in South America. E. J. Morrone, J.J., S. Roig-Juiient & J.V. Crisci,. 1994. Cla- Fittkau, H. Kinge, 6.H. Schawake and H. Sioli. 2: distic biogeography of terrestrial subantarctic 6324-658. beetles (Insects: Coleopterai from South America. Frailry, C.D., E.L. Lavina, A.Rancey, & J. Pereira de Nalional Geopra~hicResearch and Ezoloration Souza (Pho). 1988. A proposed PloistoceneII-lo- lO(1): 104-115 ' iocene lake in the Amazon Bassin and its signiti- Miiller, P. 1913. Tihe dispersal centres ofterrestrial urr- cvnce to Amazonian Geology and Biogeography. lebrntes in the Neotro~icalRealrr~. A sturlv in lhe Acla Ammnicu 18(3-4): 119-143. evolution of the Neotropical biota and ils r~aliue Goloboff, P. 1993. NONA: a tree-searching program. landscape. The Hague, Dr. W, Junk publisher. Published by the Author. Nelson, G. & N.I. Platnick. 1981. Systematics and bio- IIaKer, J. 1967s. Sornc allopatric species pairs of birds geography: Cladistics and uieariance. New York, in northwestern Colombia. Auk 84: 343-365. Columbia University Pross. - 196% Speciation in Colombian fbrest birds west Page, R.D.M. 1989. COMPONENT user's manual. Re- of the Andes. Am. M~~seuniNouilates 2294: 1-57 lease 1.5. Published by the Autor. Aockland, New - 1969. Speciation in Amazonian forest birds. Sci- Zealand. ence 165(3889): 131-1.37. Pamial, R. 1984. Late tertiary mammals of southern -- 1975. Avifauna of northwestern Colombia, South South America as indicators of climatic deteriora- America. Bonn. Zool. Moitogr. 7:1-182. tion. Vol2: 1-30. EIL:Quaternary of South America fialner, G. 1964. La entomofauna Americana. Ideas and Anlarelic Peninsula (Rahassa, J., edi. A.A. acerca de su origen y distrihuciiin. Ehlia enlonrol. Balkcna, Rotterdam-Boston. Mez. 6: 1-108. Pajcual, R. & P. Bondesio. 1982. Un roodor - 1974. Elements anciens de l'entomofaune Cadiatheriinae (Hydrochoeridaei de la edad ncotropicale: scs implicttions biog6ographiques. Huayquerienso (Mioceno tardio) de la Pampa. Quaest. Enlomol. 10: 223-262. Sumaria de 10s Ambientes terrostres on la Argen- - 1.976. 1)istribuciiin do 10s insoctos en la zona de tina durante el Mioceno. Ameghiniana 19: 19-35. trasicidn mexicana, relaciones con la entornofauna Pascual, K. & E. Ortiz-Jaureguizar. 1990. Evolving cli- de Norteam4riea. Folia entomol. Mex. 35: 1-64. mates and rnarnmal IBunas in Cenozoic South - 1978. Un nuevo patr6n de dispersiiin en la Zona de America. J. Hum, euol. 19: 29-60. Transieiiin Mexicans: el Mesoamerieano de Mantaiia, hlia cnlornol. Mez. 39-40: 219-222. Bennig, W. 1965. Pl~ylogeneticsyslernatics. Ann. Rev. Entomol. 10: 97-116. Abh. 30: 2675-319. liennig, W. & P. Wygodzinsky. 1966. The Dipterafauna Platnick, N. & G. Nelson. 1978. A method of analysis of New Zealand as a urobiein in svstematics and for historical Bioileopra~Ilv.--..Syst. . 2001.27: 1-16, Porter, C.C. 1991. Biogeografia do 10s Ichneumonidae Chilenos (Ilymenoptera: Ichneumonidaei. Acta . " eriloniol. Chil. 16: 31-68. uanccs in elad~sties:Proceedings oflhc first meet- Itappoport, E. 1968. Algunos problemas hiogeogralicos ingofthe Willi Heiinig Society. V.A. Funk and D.R. del Nuevo Mundo con especial referencia a la Hrooks. New York, New York Botanical Garden: Regi6n Neotropicd. En: Biologie de 16 Amirique 177-207. Australe (Delarnare Debouteville, C.D. & E. FIumphrics, C.J. & Llt. Parenti. 1986. Cladistic bio- Rappoport, edsi. CNRS et CNICT, Paris. geography. Oxford, Ciarendon press. Rappoport. Vol 4: 54-110. ihering, H. Von 1927. Die Geschichfe der Raven, P.M. & D.I. Axelrod. 1975. History of the flora Atlaritisclzen Oreans. G. Fiscl~er,Jena and ijuna of Latin America. Am. Sci. 63: 420-429. aJeannel, R. 1942. La Grr~esedrs faunra terrestres. Rinmelot. R.A. 1961. Rasilos fundamentales de la Elimeizts de Biogdogrdpiiie. Paris, P1,ess. Univers. iaogeo'pa~iade la ~rgentina.I'liysis, Buenos Aims I'rance. 513 pp. 22(6'3): 151-170. - 1967. Bioeeoma~hiede i'Ameriaue australe. PAils. Roiduiient. S. 1992. Insectos deAm6rica del Sur Aus- : ! . ! .. 1 .i.. 1 : .i 3 '2. iral a tiavt-s del enibque de la biogeografia I.., % . I...Il I . t I'; ., I- Histiirica. Mulleyuina 1: 107-114. ,'\I<. ,, .'\.\'I, ,'>!,.\' . , ; !.$
Housquets eds Introduccibn a la b~oeeoarafia- -. en nurn Press, Nev York, London. Pp 1-4:l pp. Latinoarndrica: Conceptus, leorias, mdtodos y Uliana, M.A. & K.T. Biddle. 1988. Mesoaoic-Cenozoic aplieaciones. UNPuM. paleogeographic and goodynamic evolution of Rosen. D.E. 1978. Vicariant ~alternsand historical ex- Southern South America. Rev. Brasil. Geociencias p1;wation in Hiogeogaphy. Syst. Zool. 27: 159.188, 18(2):172-190. Shmida, A. 1985. Biogcagraphy of the desert flora.En: Volkheimer, W. 1971. Aspoetas paieoclimatol6gicos do1 l3venari.A. M, I. Nov-Meir & 11. V. Goodall ods. Terciario Argentinn. Rev. Mus. Ax. Cs. Nut. l(8) Hot deserts and arid Shrublands. New York, (Paleontologia): 213-262. Elsevier. 12A: 23-77. Wallace, A.R. 1876. The geographical dist~ibutioizof Sirnoson, G.G. 1964. Euolucidn y GeozrafZa: Hisloria animal% with a slwdy of the relatior~shipsof liuing ke la Fauna de Arniricn ~aiina.EU~~BA, Buenos and extinct faurios as elucidating the pas1 ciianges Airas. of the earlh's surfaee. Vols. 1-11, Reprinted ed., Solorviaens,.I. 1987. Fiiogenia y hiageopafia drl genero Haher Publishing Co., New York, 1962. Eurymetopum Blanchard, 1844 (Coleoptera: Wiley, EO 1988. Parsimony analysis and vicariance Cloridae: Phyllohaeninae). Acta enlornal. Chil. 14: biogeapaphy. Syst. Zool. 37: 271-290. 127-154. Wygodzinsky, 1'. & $. Goscar6n, 1962. On the relat,ion- Stanre.-, L.A.. A.L. Teran. A. Willink. 1976. Entomo- ships and zoogeographical significance of fauna de ia provincia ~iogeogr6ticadel Monte. Acla Aush.osin~uliumanthracir~urn (Bigot), a blackfly Zoo(. Lill. 32: 73-120. from southern South America (Iliptera: l'ariini.. I>.tT. 1980. The Geoloeic Evolution of South Simuliidael. Pacific Insects 4(1): 235-244. &&a with Special Reference to the 200 ~illioil Years. En: Ciochon l3.L. &A. B. Chiarelli eds., Euo- Recibido: 4-XI-2000 lutionary biology of the Neu Wold Morzheys, Pie- Aceptado: 31-X-2001