VARIABILITY IN SOUTH AMERICAN
Psni (PmMoNAT., BAS0NMAT0PH0RA.)
Mara L1cia Ferreira Dias
Thesis presented for the, degree of Master of Philosophy
University of Edinburgh 1984
B L.
0 In accordance with the regulations of the University of Edinburgh, I hereby declare that this thesis has been composed entirely by myself and that all the work described herein was carried out by myself alone, except where stated in the acknowledgements. This thesis is dedicated to my parents, my husband, and my son for all their love and encouragement. CONTENTS
Page Number
ACKNOWLEDGEMENTS ...... vi
ABSTRACT ...... X
CHAPTER 1. INTRODUCTION...... 1
1.1 Objectives ...... 4
1.2 Literature review of Physidae
supraspecific taxonomy ...... 5
1.3 Literature review of South
American Physidae species 7
1.4 Summary of the present state of South American Physidae systematics., 19
CHAPTER 2. MATERIALS AND METHODS ...... 27
2.1 Samples of Physidae ...... 27
2.2 Shell measurements ...... 34
2.2.1 Fundamental orientation of
the shell ...... 37
2.2.2 Characters definition 40
2.2.3 Discarded characters 52
2.3 Shell statistical analysis 54
2.3.1 Regression analysis 54 Page Number
2.3.2 Principal Component
Analysis(PCA) ...... 56
2.4. Anatomical analysis ...... 57
2.4.1 Characters definition 67
2.4.2 Discarded characters 84
CHAPTER 3. INTRA AND INTERSPECIFIC VARIABILITY OF SHELL CHARACTERS IN STENOPHYSA
AND PHYSELLA POPULATIONS ...... 86
3.1 Regression analysis ...... 86
3.2 Principal component analysis (PCA) 87
CHAPTER 4. INTRASPECIFIC VARIATION OF ANATOMICAL CHARACTERS IN STENOPHYSA AND
PHYSELLA POPULATIONS ...... 163
4.1 Fundo Creek Stenophysa population 163
4.1.1 Variability ...... 163
4.1.2 Identification ...... 182
4.1.3 Assessment of characters 192
4.2 Idenfication and variability of
other Stenophysa samples ...... 197
4.3 Identification and variability of
Physella samples ...... 205 Page Number
CHAPTER 5. GENERAL DISCUSSION AND
CONCLUSIONS ...... 220
CHAPTER 6. REFERENCES ...... 238 vi..
ACKNOWLEDGEMENTS
I express my sincere thanks to my supervisors Dr. Tom Warwick and Professor Aubrey Manning of the University of Edinburgh and Dr. Warton Monteiro of Universidade de Brasilia for giving me the opportunity to develop this thesis partially in Edinburgh, Scotland and partially in Brasilia, Brazil. For their criticism, guidance, encouragement, helpful discussions, and for providing facilities I am grateful too. To Dr. W. Monteiro my sincere thanks also for the special help with the syntax of this thesis, valuable criticism and suggestions on shell methodology, and for critically reading this manuscript during the preparation of this thesis. rnough his guidance, I have learned a lot, specially about scientific logic. I am deeply indebted to my husband, Dr. Braulio F. S. Dias, for giving me encouragement and support to study South American Physidae, for critically reading this manuscript, for helpful discussion on several topics of this work, specially shell statistical analysis, and also for his help to obtain several references. My sincere thanks to Dr. George A. Te of Ann Arbor, Michigan, U.S.A., for lending me representative shells of South American Physidae (Analogs), his Physidae manuscripts, several vii.
transparencies of paratypes and holotypes, and, also for the discussion of some aspects of this project. I am much greateful to Dr. W. L. Paraense of Instituto Oswaldo Cruz, Rio de Janeiro, for the gift of preserved snails and empty shells of Physidae, collected by himself in different South American and West Indies Countries, and also for his help in obtaining references.
My appreciation also to the following museum curators for giving me the opportunity to study their Physidae Collections: Dr. A. H. Clarke (National Museum of Natural History, Washington, U.S.A.); Dr. W. K. Emerson (American Museum of Natural History, New York, U.S.A.); Dr. J. B. Burch (Museum of Zoology, University of Michigan, Ann Arbor, U.S.A.); Dr. D. Heppel (Royal Scottish Museum, Edinburgh, Scotland); and Dr. L. P. Neme (Museu de Zoologia, Universidade de São Paulo, Sao Paulo, Brazil), with special thanks for lending the paratypes of Physa (Physella) papaveroi Leme (MZ 16618) and specimens of "Pysa (Physella) cubensis" sensu Leme (MZ 17997).
I wish to thank Dr. Tom Warwick for helping to gather Physidae species from several places of Europe and for collecting some of them around Edinburgh.
My sincere thanks to the following scientists for the obtainment of Physidae material: Dr. J. B. Burch, for several alcohol lots of North American viii
Physidae; Dr. D. Heppel, for a shell of Stenophysa maugeriae and alcohol lots of Aplexa hypnorum; and to Dr. A. Norris, Department of Leisure Services, Leeds, England, for collecting Physella acuta from Leeds; Mrs. M. Fogan, Conchological Society of Great Britain and Ireland, Manchester, England, for collecting Physella gyrina from Chester, Liverpool, England; Dr. Shelagh Smith, Royal Scottish Museum, for collecting P. acuta from Hazeihead Park, Aberdeen, Scotland; Dr. Barry Colville, Leeds, for collecting A. hypnorurn from Leeds; Dr. Plummer, Royal Veterinary College, London, for collecting Physella acuta from Richmond Park, London, England; Dr. J. F. Vaz, from Superintend&ncia de Controle de Endemias (SUCEM), Säo Paulo City, for preserved material and shells of Stenophysa marmorata from So Paulo. State; Dr. W. Monteiro for collecting S. marmorata from the following Brazilian localities: Joo Pessoa, Paraiba State; Rio de Janeiro, Rio de Janeiro State; Corumb, Mato Grosso do Sul State; and Canabrava, Goiás; Dr. M. B. Monteiro of Universidade de Brasilia for S. marmorata specimens from Diamantina, Minas Gerais State; and Dr. B. F. Dias for S. marmorata from Belém, Pars State.
I am indebted to Maria Cristina Pons da Silva and to Dr. Inga L. V. Mendes from the Museu de Cincias Naturais (MCN), Porto Alegre, Rio Grande do Sul, Brazil; Dr. Arnaldo C. dos Santos Coelho, Museu Nacional, ix
Rio de Janeiro City; Dr. J. L. de Barros Araüjo from the Universidade Federal Rural do Rio de Janeiro, Itagual, Rio de Janeiro; Dr. L. P. Neme; Dr. D. Heppell; and, finally, to Mrs. Hext the librarian of Zoology Department, University of Edinburgh, for the obtainment of references.
For their helpful discussion on topics of this work 'I am grateful to the molluscan specialists, Dr. G. M. Davis, from the Academy of Natural Sciences, Philadelphia, U.S.A.; Dr. D. Heppell; and Dr. J. L. M. Leme, Museu de Zoologia, Universidade de São Paulo, Sao Paulo City.
My thanks also to Mrs. Maria Inez M. T. Walter and Mrs. Tnia Mara de A. Campos, Centro de Processamento de Dados (CPD - SAU), Universidade de Brasilia, for the help with computer analysis. Finally I would like to acknowledge the financial support of the Brazilian National Research Council (Conselho Nacional de Desenvolvimento Cientifico e Tecnolôgico - CNPq). x
ABSTRACT
The objectives of this study were to assess the intraspecific variability of shell and anatomical chracters used in Physidae systematics, determining the size independent ones, and to assess their taxonomical significance; to identify specimens from South America, specially Brazil, adding further distribution records, and evaluate the specific status of "Physa (Physella) papaveroi Leme, 1966", and "Physa (Physella) cubensis sensu Leme, 1966".
Twenty one shell measurements of 158 specimens and 26 anatomical characters of 90 specimens were scored mainly in Stenophysa marmorata (Guilding, 1828) and Physella (Costatella) acuta (Draparnaud, 1805) from South and Central America, West Indies and Great Britain. A new conchometric method was developed and several characters redefined. Eight shell characters were found to be size independent. Regression coefficient tests were calculated for S. marmorata and P. acuta populations without any significant difference. Two shell ratios were constant within the size range of the analysed material, granting them taxonomical use. Bursa Copulatrix Axis, Digestive Tract Pigmentation, Shape of Gizzard, Preputial Gland Presence and Retractor Muscle were constant for S. marmorata, xi
having generic diagnostic value. Tubular Nature of the Kidney, Penial Complex Type and Number of Segments in the Penial Sheath were constant for S. marmorata, having specific diagnostic value. Tendency of Preputial Gland to Flatten and Swelling of Penial Sheath Terminus were constant for P. acuta, having specific diagnostic value. The remaining anatomical characters were variable and of doubtful use for taxonomical purposes. States of Mantle Pigment Pattern and Mantle Lappet Type were redefined.
Nineteen shell characters of S. marmorata and P. acuta from 39 localities were studied by Principal Component Analysis (PCA). The three Principal Components were 75% of the total variation in S. marmorata, 89% in
P. acuta, and 74% in both . The first Principal Component (shell size) alone represented 53%, 58% and 51% of the total variation in S. marmorata, P. acuta, and both,
respectively. The second and third Principal Components (angles) separated S. marmorata from P. acuta, and disclosed distinct clusters within each species. No anatomical difference was found between these clusters within each species, confirming the intraspecific nature of such clusters.
This study discloses the first record of P. acuta (Draparnaud) for South America, and of Physella for Central Brazil. Physa (Physella) papaveroi Leme, 1966 was found to belong to the Physella acuta complex, and a new combination is given: Physella (Costatella) papaveroi Xli
(Leme). "Physa (Physella) cubensis sensu Leme, 1966belongs to the Physella acuta:complex, and should be named Physella (Costatella) acuta (Draparnaud). The presence of Physella cubensis (Pfeiffer) - is not confirmed, and doubts are raised on the correct identification of Physella lots identified as such in South America. The following new records of S. marmorata were found: Panama, Peru, and the Brazilian States Amazonas, Bahia, Espirito Santo, Minas Gerais, Federal District, Mato Grosso do Sul and Rio Grande do Sul. 1
CHAPTER 1. INTRODUCTION
South America has a rich, though poorly studied, fauna of freshwater Gastropoda. Paraense (1981a) wrote an introduction to the subject. There are ten families of freshwater snails in South America: five. Prosobranchia (Neritidae, Ampullariidae (= Pilidae), Hydrobiidae, Melaniidae and Pomatiopsidae); and five Pulmonata (Planorbidae, Ancylidae, Lymnaeidae, Chilinidae and Physidae). Planorbidae is the largest family of pulmonates in South America with several species belonging to seven genera (Paraense, 1975; Paraense,
1981a; Paraense, 1983a). Ancylidae is the second largest family with about five genera (Wurtz, 1951; Hubendick, 1967; Paraense, 1981a). The Lymnaeidae has at least four species belonging to Lymnaea sensu lato, and is apparently absent from Northeastern Brazil (Hubendick, 1951; Jaeckel, 1952; Paraense, 1976, 1982a, 1982b and 1983b ;and Ueta, 1976). The Chilinidae is endemic to Southern South America and has only one genus Chilina (Pilsbry, 1911 and Paraense, 1981a). The Physidae has few species belonging to the genera Stenophysa and Physella (Te, 1978).
Planorbidae is the best studied group of freshwater snails in South America due to its medical importance: some species of Biomphalaria are intermediate hosts of Schistosoma mansoni (Paraense, 1972; Pan American 2
Health organization, 1968). Some species of Lymnaeidae (Gonzales et al., 1974; Rezende et al., 1973; Tantalan et al., 1974) and Physidae (Prez Vigueiras and Moreno, 1938) have been reported as intermediate hosts of Fasciola hepatica in the Neotropics. However, as other workers have fai1 'ed to confirm natural and experimental infection for F. hepatica in several Physidae species (Lêon-Dancel et al., 1970; Tanta1an et al., 1974), this subject remains open for further conclusive studies. Studies on the morphology of Physidae are scarce in the literature. Te (1978) made an extensive review of the systematics on the world Physidae. He examined 20153 lots and 403 type lots from North American museums. He scored 433 specimens for soft anatomy characters and 1300 for shell characters, stucng 71 characters, being 37 shell characters, and 34 soft anatomy characters. The anatomical characters include tentacle, mantle edge, kidney, bursa copulatrix, digestive tract and penial complex, etc. He also made supplementary studies of jaw, radula (electron microscopy), foot terminus, egg capsules, electrophoresis (esterases in foot) and immunological tests. By cluster analysis Te calculated the similarity coefficients for shell and soft parts for all Physidae species and genera. His comprehensive study is now the basic work on the taxonomy of the family Physidae of the whole world, both at the specific and supraspecific level. He also studied the interspecific variability among the four existing 3
genera of Physidae. However, he did not have opportunity to study a large sample of South American material. Actually he dissected and scored. .35 specimens for soft anatomy characters, and 115 specimens for::'shell characters among the six species he recognized from South America, though most of that material was collected in the West Indies and Central America.. Though much progress has being made on Physidae systematics, there still are some difficulties on the identification of South American Physidae which may be overcome by detailed shell and anatomical studies at the intrapopulational level. It is well known that freshwater pulmonates may show a wide range of intrapopulational variation on their shell and anatomical characters (Hubendick, 1951 and Wurtz, 1949). Recent publication) have been emphasizing the analysis of intraspecific variability of different characters within snail populations (Janson, 1982; Paraense, 1980; Rao & Bhavanarayana, 1976; Ueta, 1976). Consequently, the number of accepted valid congeneric species is being reduced in several genera and families. The 392 specific names proposed in the family Physidae have been reduced to 48 valid species by Te (1978, 1980). However, Te (1978) has not evaluated the specific status of Physa (Physella) papaveroi Leme, 1966 and Physa (Physella) cubensis Pfeiffer, 1839 sensu Leme (1966). These are specific names which are currently applied to Physidae collections from Southeastern Brazil. 4
1.1 OBJECTIVES
The general purpose of this work is to study the shell biometry and internal anatomy of several South American Physidae populations, in order to improve the understanding of their taxonomy. The specific objectives are as follows: Revise the literature on supraspecific Physidae taxonomy and on South American Physidae species; re4i(fine and describe the anatomical and shell characters studied by Te (1978); produce anew methodology for shell measurements; assess the intraspecific variability of the anatomical and shell characters as used by Te (1978); determine those shell characters which are size independent; understand the growth rate of some shell characters by regression analysis; determine the least number of shell characters which are necessary to distinguish Stenophysa and Physella species by Principal Component Analysis (PCA); evaluate the significance of shell and anatomical characters used by Te (1978); assess differences in shell growth rate of six different Physidae populations by the Regression Coefficient test; separate clusters of Physidae populations and species by PCA; identify Physidae specimens collected from South American localities, specially from Brazil; evaluate the specific status of "Physa (Physella) papaveroi Leme, 1966 and "Physa (Physella) cubensis Pfeiffer, 1839 sensu Leme 1966"; contribute to a better understanding of the distribution of South America 5
Physidae species, specially within Brazil.
1.2. LITERATURE REVIEW OF PHYSIDAE SUPBASPECIFIC TAXONOMY
The first Physidae genus, Physa, was described by Draparnaud (1801) (genotype: Bulla fontinalis Linn, 1758). Fleming (1820) described the genus Aplexa (genotype: Bulla hypnorum Linné, 1758). Later on Fleming (1828) proposed Aplecta as an emendation to Apiexa. For most of the nineteenth and early twentieth century most species of Physidae were ascribed to the genus Physa if they had a short spire and to the genus 1exa or Aplecta if they had a high spire. In 1842 Haldeman proposed two new genera: Physella (genotype: Physa globosa Haldeman, 1841) and Physodon (genotype: Physa microstoma Haldeman, 1840), but these names were not utilized by other authors until recently. Dali (1670) created the genus Costatella (genotype: Physa costata New comb, 1861) and Martens (1898) erected the genus Alampetis (genotype: Physa osculans Haldeman, 1841). Pilsbry (1925) created the genus Petrophysa for the unique Physa zionis Pilsbry, 1925. The first generic name proposed for a South American Physidae was Stenophysa Martens (1898) as a subgenus of Physa (genotype: Physa sowerbyana d'Orbigny, 1841). Clench & Aguayo (1932.) proposed Haitia, as a subgenus of Physa, for the unique Physa elegans Clench & Aguayo, 1932, of Hispaniola (West Indies). Zilch (1956) proposed the genus Alampetista as a new name for Alampetis Martens, 1898 which was preoccupied by Alampetis Thompson, 1878. Starobogatov (1967) created the genus Afrophysa for the West African Physa waterlotti Germain, 1911; and Starobogatov & Streletskja (1967) proposed the genus Sibirenauta for Physa kultukiana. F. C. Baker (1926, 1928) was the first to note that the American short spired Physidae were different, from the European Physa Draparnaud and proposed to use the name Physeila Haldeman for them. Starobogotov (1967) proposed to divide the family Physidae into two
subfamilies: Physinae (Physa, Physelia, Petrophysa and Afrophysa) and Aplexinae (Aplexa, Stenophysa and Sibirenauta). Te (1975) synomymized Physodon Haldeman under Physella Haldeman. Te (1978) also synomymized Afrophysa Starobogatov under Stenophysa Martens, Sibirenauta Starobogatov and Streletzkja under Aplexa Fleming, Haitia Clench & Aguayo and Alampetista Zilch under Costatella Dali. Te (1978, 1980) proposed to divide the Physidae into two subfamilies: Physinae with Physa and Physella; and Aplexinae with Aplexa and Stenophysa. He proposed to divide Physella into three subgenera: Physeila sensu stricto, Petrophysa and Costatella, the latter divided in two sections: Alampetista and Costatella. Physa and Aplexa are small genera restricted to the Holarctic Region and Stenophysa is a small genus 7
restricted to the Neotropical Region (Te, 1978). Physella is a large genus restricted mainly to the Americas, and specially rich in North America, except for one species, P. acuta Draparnaud, which is native to the Mediterranean Region and introduced into all the other continents (Beetle, 1973; Bruggen, 1966; Clench, 1934; Hamilton- Attwell et a1., 1970; Jenkins, 1890; Te, 1978).
1.3 LITERATURE REVIEW OF SOUTH AMERICAN PHYSIDAE SPECIES
The Central American Region, including Mexico, has a rich fauna with 15 species of Physella, mostly belonging to subgenus Costatella, and five species of StenoFhysa (Te, 1978). The West Indies has a relatively poor fauna with seven Physella species, all in subgenus Costatella, and three species of Stenophysa (Te, 1978).
According to Te (1978), six of the species of these two regions (three Physella and three Stenophysa) also occur in South America. Compared with South America, the Physidae of Central America and West Indies are relatively well known and have been the subject of several careful studies (see, for example, for Central America the works of: Fischer & Crosse, 1886; Martens, 1898; Bequaert & Clench 1933 and 1936; and for the West Indies: Clench, 1936 and 1939; Aguayo, 1938; Richards, 1964; Harry & Hubendick, 1964 and Pointier, 1974 and 1976). FI
South America, on the other hand, has been very little collected and lacks review works. Therefore I shall concentrate my studies on the South American fauna. Following is a historical review of the papers published on the South American Physidae. The first record of Physidae for South America was made by Gray (1828) who described Physa peruviana Gray from Peru (type locality: swamps between Lima and Callao). Alcide d'Orbigny (1835 - 1846) travelled during eight years, between 1826 and 1833, throughout several countries in South America passing by: Rio de Janeiro in Brazil; Montevideo and Maldonado in Uruguay; Buenos Aires, the mouth of the Negro River around Carmen de Patagones and Viedma, and the Province of Corrientes in Argentina; Valparaiso, Cobija and Arica in Chile; most Provinces of Bolivia; Islay, Callao and Lima in Peru (Papavero, 1971). He stayed for a longer time in the Province of Corrientes (little over a year), around the mouth of the Negro River (eight months) and Bolivia (about three years). He collected extensively throughout these places giving much emphasis to freshwater molluscs. In 1835 d'Orbigny recorded the occurrence of Physidae in several parts of South America: Montevideo in Uruguay, Patagonia Province ( = Carmen dePatagones, Buenos Aires Province, according to Papavero, 1971) in Argentina; Valparaiso in Chile; and Lima Province in Peru. Besides, he indicated that physids were also S
present in some other places he visited in South America. He considered all these specimens as belonging to just one species: Physa rivalis Sowerby 1822 (type locality: Guadeloupe, West Indies).
In 1837, d'Orbigny again recorded only one species of Physidae for South America, Physa rivalis Sowerby, for which he gave a description and illustration. He also recorded its occurrence in Bolivia, and divided the species into two varieties he described as new: Phya rivalis var. major d'Orbigny (type locality: around and between Lima and Callao, Peru, in the swamps near the mouth of the Rirnac River) and Physa rivalis var. minor. d'Orbigny (type localities: Rio de Janeiro, Brazil, in small rivers specially near So Cristôvo and Botafogo Bay; Montevideo, Uruguay, in a small river near the Cerro in the Bay of Montevideo; in Batel River, Corrientes Province, Argentina; and Patagonia Province near the Negro River ( = vicinities of Carmen de Patagones, Buenos Aires Province), Argentina. He pointed out that major variety is the same species described by Gray (1828) as Physa peruviana Gray. Thus P. major d'Orbigny 1837 is apparently a junior synonym of P. peruviana Gray. Spix and Martius travelled extensively throughout Central and Amazonian Brazil from 1817 to 1820 collecting plants and animals, giving much emphasis to terrestrial and freshwater molluscs. A description of the molluscs they collected was published by Spix and Wagner (1827), where numerous 10 species were described. But curiously no Physidae was reported. Moricand (1833-1839) also recorded and described several species of freshwater and terrestrial molluscs collected in the neighbourhood of Salvador, Bahia, Brazil, by Mr. M. S. Blanchet, but again no Physidae was reported. Beck (1837) published a new name for South American Physidae: Bulinus (Aplexa) brasiliana Beck from Rio de Jañeiro, Brazil. However he gave no description or illustration for this species and consequently it is not a valid name (nomen nudum). Later, Anton (1839) recorded a new Phy.sidae from Peru, Physa peruviensisMtihlfeldt, and also recorded Physa panamensis Mfihlfeldt from Panama. Although he attributed these species to Mühlfeldt, Anton is in fact the author as Mhlfeldt never published these names. However since he gave no description or illustration for these species, they are not valid names (nomen nudum). D'Orbigny (1841) proposed Physa sowerbyana as a new name for Physa rivalis Sowerby 1822 (type locality: Guadeloupe, West Indies, preoccupied by P. rivalis (Maton & Rackett, 1807) of England. Küster (1844) described and illustrated Physa antonii KUster from Peru (new name for P. peruviensis "Mühlfeldt" Anton, 1839, which is a nomen nudum), Physa panamensis "Mühlfeldt" from Panama and Physa brasiliensis "Koch" from Brazil, thus validating these species. 11
Consequently KUster is in fact the autor of these species. Gould (1848) described and illustrated Physa venustula Gould from Peru. Martens (1859) described Physa venezuelensis Martens from Lagunhlla, Venezuela, and later on Martens (1873) recorded Physa rivalis (Maton and Rackett) from Caracas, Venezuela. Sowerby (1873) recorded Physa peruviana Gray from Guayaquil, Ecuador. Jousseaume (1887) described Aplecta carolita Jousseaume from Ecuador. Cousin (1887) described Aplecta gualbertoi Cousin from Mapasinga, Ecuador, and recorded Aplecta martinidella Jousseaume from San Nicolas, Canton de Megia, Ecuador (referring to a figure of Jousseaume (1887) which actually is an illustration of A. carolita Jousseaume). Jousseaume (1889) described Physa simoni Jousseaume from Laguna de Espino, near Caracas, Venezuela, and recorded Aplecta rivalis (Naton and Rackett) from Petare, near Caracas, Venezuela. Pilsbry & Rush (1896) recorded Physa sowerbyana d'Orbigny from Arroyo Miguelete, Prado de Montevideo, Uruguay. Martens (1898) recorded a Physidae similar to Physa (Stenophysa) panamensis "Mühlfeldt" Anton from Bahia (probably Salvador, Bahia State), Brazil; also Physa (Aplecta) peruviana Gray from Peru and Physa sowerbyana d'Orbigny from Continental South America. Corsi (1900) recorded Physa rivalis Sowerby var. minor d'Orbigny from Arroyo Pantanoso, near Montevideo, Uruguay. Preston (1907) described Pjysa cornea 12
Preston from Mérida, in Western Venezuela. Holmberg (1909) described Physa loosi Holmberg from Santa Lucia al pie del Cerro Pie de Palo, Provincia de San Juan, Argentina; and also Physa aspii Holmberg
from Laguna de los Murciélagos, Formosa, Argentina. Pilsbry (1911) in his monograph on the molluscs of Patagonia did not record any Physidae for that region. Fred Baker (1914) travelled extensively through Northeastern and Amazonian Brazil and recorded Physa rivalis (Maton and Rackett) from Cear-Mirim and Papary Lake in the State of Rio Grande do Norte and also from Belém, Pars (in an artificial lake in front of the Cathedral), and also Physa sp (young specimens) from the State of Maranho. He-ad not reporany Physidae in V the natural lakes and rivers which he examined in the Amazonian region. Lutz (1914) recorded an unidentified
Physidae (cited as Ehysa sp) from Rio de Janeiro, Brazil. H. B. Baker (1930) collected extensively in coastal Venezuela and recorded Aplexa (Stenophysa) panamensis "MQhlfeldt" KUster from Estaci6n Tachira, in Western Venezuela; Aplexa (Stenophysa) rival is rivalis (Maton and Rackett) from Bejuma, Palma Sola, Boquer6n, Tucacas and Caracas, all in central coastal Venezuela, and Physa (Physella?) cubensis Pfeiffer which is a form similar to jamaicensis C. B. Adams from Venezuela (specific locality not mentioned). He also proposed Bulinus (Aplexa) brasiliana Beck as a subspecies of 13
Aplexa rivalis (Maton and Rackett). Baker (1930) presented the first description of. soft parts of a South American Physidae (A. r. rivalis): Clench (1930) pointed, out that Physa rivalis Sowerby, 1822 from West Indies and South America was preoccupied by the homonym Physa rivalis (Turton, 1807) ( = rivalis Maton and Rackett, 1807) (type locality: Hampshire, England) and should thus be replaced by the next available name which is Physa marmorata Guilding, 1828 (type locality: Saint Vincent, West Indies). He also put Physa brasiliensis "Koch" Phil. 1845 (sic) in the synonymy of P. marmorata, and recorded P. marmorata for Trinidad. He expressed his opinion that P. sowerbyana d'Orbigny is different from P. marmorata Guilding. He manifested that the species described by H. B. Baker (1930), as Aplexa rivalis rivalis (Maton & Rackett) was apparently Physa peruviana Gray. He gave the following synonyms for Aplexa peruviana Gray, 1628: Physa rivalis Pot. & Mich., 1838; Physa peruviensis "Mtlhlfeldt" Anton, 1839; Physa antonhi Küster; Aplecta carolita Jousseaume, 1887; Aplecta martinidella Cousin, 1887 (nomen riudum referring to the figure of A. carolita Jousseaume); Aplexa rivalis (Maton & Rackett) H. B. Baker, 1930. Clench (1936) concluded that P. sowerbyana d'Orbigny is a synonym of Aplexa marmorata, contrary to his own opinion in his 1930 paper. He gave the following further synonyms for Aplexa marmorata Guilding: Lymnaea (Physa) rivalis Sowerby, 1822; Physa brasiliensis "Koch" 14
KUster, 1845; P. salleana Dunker, 1853 (type locality: Santo Domingo), P. salleanae "Dkr" Sowerby, 1873 (misspelling of salleana); P. acuminata "Gray" Sowerby,
1873 (type locality: St. Thomas); P. ven2tr.icosa.
"Guilding" Sowerby, 1873 (type locality: -St. Vincent); P. margaritacea "Shuttleworth" Paetel, 1889 (type
locality: Antigua) (nomen nudum). He recorded A. rnarmorata from: Botanic Gardens, Georgetown, British Guiana; Laranjeira, near Rio de Janeiro, Brazil; Trinidad, Barbados, Guadeloupe, St. Kitts, St. Thomas, and Hispaniola.
Clench (1936) also dealt with Physa cubensis Pfeiffer, 1839 (type locality: Cuba), recording it from Florida, U.S.A.; Bermuda, Bahamas, Cuba, Isle of Pines, Hispaniola, Jamaica, Puerto Rico, and St. Croix. He did not record this species for South America. He also gave the synonyms for this species.
Aguayo (1938) recorded Physa cubensis Pfeiffer from Trinidad and Northern South America. He considered Physa jamaicensis C. B. Adams, 1851 as a synonym of Aplexa (Stenophysa) marmorata (Guilding). Besides he presented information on the ecology of West Indian Physidae. Prez Vigueiras & Moreno (1938) recorded Physa cubensis as a new intermediate host for Fasciola hepatica in Cuba. Clench (1939) recorded the following new occurrences of Aplexa marmorata (Guilding): Jamaica, Neves, Antigua, and Marie Galante. Haas (1939) recorded Aplexa rivalis (Maton & - ;. .. 15
Rackett) from the, following localities in Northeastern Brazil: Guaramiranga, Ceara State;between Patos and Pombal, Paraiba 'State; Areias, near. Recife, Lago do Carro, Rio Branco and Rio Mandalim, all in Pernambuco State Benthem-Jutting (1943) further recorded Aplexarlvalis from Northeastern Brazil. Biese (1948) described Physa nodulosa Biese from Coquimbo Province, Chile (type locality: Rio Illapel, Illapel). He also described an albino form of P. nodulosa (P. nodulosa form albina) from Coquimbo, Chile (type locality: Rio Elqui, Albarrobito near Serena). Morretes (1949) recorded Aplexa rivalis (Maton & Rackett) from the following Brazilian localities: Curitiba, Parana State; Ponta do Ip& Arcado, Gois State; and Ceara State. He also recorded Physa janeirensis Beck (sic) from Rio de Janeiro, Brazil (apparently reffering to Bulinus (Aplexa) brasiliana Beck, 1837) .Baratini (1951) recorded P hysa rivalis (Maton & Rackett) from several localities in Uruguay. Parodiz (1965) recognized the occurrence of two species of Physidae in Northern South America: Aplexa peruviana (Gray) in the West and Physa marmorata Guilding in the East. To his judgement, the Physidae around the La Plata estuary ("Uruguay and Buenos Aires Province") were different from Aplexa peruviana (Gray) and could either be P. marmorata or represent a different species: probably P. sowerbyana d'Orbigny. He found Physidae specimens around the city of Buenos Aires 16 at Vila Pueyrredôn, Buenos Aires Province, Argentina. Barth (1957) recorded Aplexa brasiliensis (Kaelp.) (sic) (probably ref(erin to Koch) from the Parque Nacional do Itatiaia, Resende, Rio de Janeiro State, Brazil. Figueiras (1964) recorded Aplexa (Stencphysa) rivalis (Maton & Rackett) as common occurrence throughout Uruguay.
Richards (1964) described and illustrated the soft anatomy of Puerto Rican Physa cubensis and Aplexa marmorata. Harry and Hubendick (1964) also described and illustrated the soft anatomy of Physa cubensis and Physa marmorata -. According to Te (1978), the specimens treated by Harry & Hubendick as P. marmorata are in fact Stenophysa peruviana (Gray) Leme (1966) recorded Physa (Physella) cubensis Pfeiffer from a lagoon at Manguinhos, Rio de Janeiro City, Brazil, and described Physa (Physella) paveroi Leme (type locality: lagoon at Parque D. Pedro II, So Paulo City, Brazil). He studied the soft anatomy and shell morphology for both species in comparison with Physa (Physella) bermudezel Aguayo from West Indies. His identification of cubensis was confirmed by Clench. Leentvaar (1967) recorded Aplexa marmorata (Guilding) in the artificial Lake Brokopondo in Suriname, and pointed out that this species was absent in the River Suriname before the Brokopondo dam was built. Léon-Dancel, Ritchie & Chiriboga (1971) studied the refractiveness of Physa cubensis (Pfeiffer) 17 and Aplexa marmorata (Guilding) to infection by Fasciola hepatica in Puerto Rico. Tanta1en, Huiza & Capuiiay (1974) recorded Physa venustula Gould from Peru, and studied its role as a vector of Fasciola hepatica.
Mello & Ueta (1973) studied the morphology of the radula of Aplexa inarmorata (Guilding) (determined by
Lenie) from Campinas, So Paulo State, Brazil.
Abercrombie & Berg (1975) found that Aplexa
(Stenophysa) marmorata (Guilding) (identified by Parodiz) from coastal Southeastern Brazil, between Paraná and Rio de Janeiro States, were fed upon by larvae of the
Sciomyzidae fly Tecomia limbata (Wiedemann) in laboratory conditions. Ayala, Castleton & Malek (1976) recorded an unidentified Physidae from the Rio Grande Valley, Bahia, Brazil.
Freitas (1978) recorded an unidentified
Physidae (cited as Physa sp) from the following localities in Minas Gerais State, Brazil: Serra Verde in
Belo Horizonte Country; Piçarro in Nova Era Country; and
Itabira. Bredt & Mello (1978) also recorded an unidentified
Physidae (Physa sp) for the Distrito Federal and Formosa
Country, in Goiãs State, Brazil.
Vaz (1979) studied the nervous system of
Aplexa marmorata (Guilding) (identified by Leme). His samples were from the following localities in São Paulo
State, Brazil: Rio Pinheiros, in São Paulo City, Cãndido
Mota, Cubatao, Bebedouros, and Presidente Bernardes. Monteiro & Dias (1980) recorded an unidentified Physidae (Physa sp) for Lake Paranbá basin, in Brasilia, Distrito Federal, Brazil and emphasized its occurrence together with Biomphalaria species (P1anarbidae). Santos (1981) recorded Physa af. marmorata Gulicling as a holocene subfossil in Três Riachos in Umbuzeiro Country, Paraiba State, Brazil. Fernández (1981) recorded Physa loosi Holmberg from Chaco, Argentina; Physa aspii Holmberg from Salta, Argentina, and Aplexa (Stenophysa) marmorata Guilding from Northearn, Central and Eastern Argentina, Uruguay and Paraguay. The prece /ding literature survey indicates that Physidae snails are present in most parts of South America, though usually not abundant. They seem to be especially common in man made water bodies such as ditches, ponds and reservoirs (Leentvaar, 1967). However, the Physidae are, apparently, totally absent from the following regions: higher places in the Andes (d'Orbigny, 1837; Fernández, 1981); Patagonia (Pilsbry, 1911; Fernández, 1981); Central and Southern Chile (Biese, 1948) and the Amazonian Tertiary Valley (Spix & Wagner, 1827; F. Baker, 1914; Sioli, 1951, 1953a, 1953b, 1956a and 1956b). Sioli (op. cit.) observed that the acid waters of the Amazonian Tertiary Valley had a very poor gastropod fauna comprised only by Ancylidae and Ampuilariidae. He also examined neutral waters crossing areas where the 19
carboniferous layers surfaced. In those areas he found a very rich molluscan fauna, lacking Physidae snails though.
1.4. SUMMARY OF THE PRESENT STATE OF SOUTH AMERICAN PHYSIDAE SYSTEMATICS
According to Te (1978) six species of Physidae are known to occur in South America: Stenophysa marmorata (Guilding, 1828), S. peruviana peruviana (Gray, 1828), S. peruviana spiculata (Morelet, 1849), S. panamensis (Küster, 1844), Physella (Costatella) venustula (Gould, 1848), P. (C.) squalida (Morelet, 1851) and P. (C.) cubensis (Pfeiffer, 1839). Te did not mention Physella papaveroi described by Leme (1966), neither "Physa loosi" andPhysa aspii"described by Holmberg (1909).
A brief summary of the synonyms used in the South American literature and distribution of the species is presented below:
1. Physella (Costatella) venustula (Gould, 1847) -. -
1848 Physa venustula Gould (p. 215). type locality: Lima, Peru.
1948 Physa nodulosa Biese (p. 237-8, Fig. 13). type locality: Rio Illapel, Illapel, Coquimbo, Chile.
Synonymy proposed by Te (1978).
1948 Physa nodulosa albina Biese (p. 239). type locality: Rio Elqui, Algarrobito, cerca.de Serena, Coquimbo, Thile. 20
Synonymy proposed. by Te (1978)
Distribution: Central and South America (Te,. 1978); Coquimbo Province in Chile (Biese, 1948); Peru (Tanta1en et. al., 1974; Gould, 1848).
Physella (Costatella) squalida (Morelet,1851)
1851 Physa squalida Morelet (p. 16) type locality: Swamps of the Rio Usumacinta, near Balacan, Tabasco, Mexico.
Distribution: Southwestern North America, Mexico, Central America, West Indies, Northern South America (Te, 1978)
Physelia (Costatella) cubensis (Pfeiffer, 1839)
1839 Physa cubensis Pfeiffer (p. 354) type locality: Cuba.
Distribution: Southeastern United States, Bermuda, West Indies, Central America and Northern South America (Te, 1978); Venezuela (H. B. Baker, 1930); Rio de Janeiro, Brazil (Leme, 1966); São Paulo, Brazil (Vaz, 1979).
Physella papaveroi (Leme, 1966) St 1966 Physa (Physella) papaveroi Leme (p. 270-5, 1 p1. 2nd figs. 2, 5; p1. figs. 1-5; 3rd p1. figs. 6, 8)
type locality: Parque D. Pedro II, Sao Paulo, São Paulo,
State, Brazil. 21
Distribution: recorded only for the type locality.
5. •Stenophysa marmorata (Guilding, 1828)
1822 Limnea (Physa) rivalis Sowerby (p1. a79, fig. 9)
type locality: Guadeloupe (West Indies). Synonymy proposed by: Smith (1895); Clench (1930); and Te (1978). Name preoccupied by Bulla rivalis Turton, 1807
(= rivalis Maton & Rackett, 1807), type locality: Hampshire, England (according to Clench, 1930).
1828 Physa marmorata Guilding (p. 534)
type locality: Sti. Vincentii (West Indies).
1837 Bulinus (Aplexa) brasiliana Beck (p. 116)
type locality: Rio de Janeiro, Brazil (nomem nudum). Synonymy proposed by: H. B. Baker (1930) (as a synonym of rivalis Maton & Rackett, 1807 (meaning rivalis Sowerby)).
1837 Physa rivalis minor d'Orbigny (p. 342)
type localities: "environs de Rio de Janeiro, princIpalement prs de Saint- Christophe et de la baie de Botafogo; petite riviere voisine du Cerro, dans la baie de Montevideo; prs du Rio-Batel, province de Corrientes; Patagonie, non loin du Rio Negro". 22
Synonymy inferred from d'Orbigny (1837).
1841 Physa sowerbyana d'Orbigny (p. 190, p1. 13, figs. 11- 13) New name for Physa rivalis (Sowerby).
Synonymy proposed by d'Orbigny (1841), Gray (1854), Smith (1895), Martens (1898), F. Baker (1914), H. B. Baker (1930) (these authors proposed this synonymy for rivalis Sowerby or rivalis Maton & Rackett meaning rivalis Sowerby); Clench (1936), and Te (1978)
1844 Physa brasiliensis "Koch" Küster (p. 10, p 1 . 1, figs. 18-20)
type locality: Brazil. Synonymy proposed by H. B. Baker (1930) (as a synonymj c/ / of Aplexa rivalis brasiliana (Beck)); Clench (1930); Te (1978).
1859 Physa venezuelensis Martens (p. 66)
type locality: Lagunilla, Venezuela. Synonymy proposed by H. B. Baker (1930) (as a synonym of rivalis Maton & Rackett, meaning rivalis Sowerby), and Te (1978). 23
1907 Phy,a cornea Preston (P. 497-8, fig. 20)
type locality: Mérida, Venezuela. Synonymy proposed by Te (1978).
Distribution: West Indies, Central and South America (Te, 1978); Venezuela (Martens, 1859, 1873; Jousseaume, 1889; Preston, 1907; H. B. Baker, 1930); Guiana (Clench, 1936); Surinam (Leentvaar, 1967); Para State, Brazil (F. Baker, 1914); Rio Grande do Norte State, Brazil (F. Baker, 1914); Ceará State, Brazil (Haass, 1939; Morretes, 1949); Paraiba State, Brazil (Haas, 1939); Pernambuco State, Brazil (Haas, 1939); Gois State, Brazil (Morretes, 1949); Rio de Janeiro State, Brazil (Beck, 1837; d'Orbigny, 1837; Clench, 1936; Barth, 1957); São Paulo State, Brazil (Mello & Ueta, 1973; Vaz, 1979); Parana State, Brazil (Morretes, 1949); Uruguay (d'Orbigny, 1837; Pilsbry & Rusch, 1896; Corsi, 1900; Baratini, 1951; Figueiras, 1964); Paraguay .(Fernändez, 1981); Buenos Aires Province, Argentina (Parodiz, 1956); Corrientes Province, Argentina (d'Orbigny, 1837); Mouth of the Negro River, Argentina (d'Orbigny, 1837); Bolivia (d'Orbigny, 1837); Valparaiso Province, Chile (d'Orbigny, 1835). 24
Stenophysa peruviana peruviana (Gray, 1828)
1828 Physa peruviana Gray (p.-5, p1. 6, fig. 10)
type locality: Swamps between Lirna and Callao, Peru.
1837 Physa rivalis major d'Orbigny (p. 342)
type locality: Callao, Peru. Synonymy proposed by d'Orbigny (1837).
Distribution: Mexico, Central America, West Indies and South America (Te, 1978); Peru (Gray, 1828; d'Orbigny, 1837; Martens, 1898); Ecuador (Sowerby, 1873)
Stenophysa peruviana spiculata (Morelet, 1849)
1839 Physa pruviensis "Mfihlfeldt" Anton (p. 48)
type locality: Peru (nomen nudum) (according to H. B. Baker, 1930) Synonymy proposed by Martens (1898) and Te (1978).
1844 Physa antonii Küster (p. 12, p1. 2, figs. 6-8) type locality: Peru (new name for P. peruviensis Anton, 1839) Synonymy proposed by Te (1978).
1849 Physa spicul,ata Morelet (p. 18) type locality: Campeche, Yucatan, Mxico.
1887 Aplecta carolita Jousseaume (p. 184, p1. 3, fig. 5)
type locality: San Nicolas, Canton de Megia (Ecuador) Synonymy proposed by Te (1978). - 25
1887 Aplecta marti-nidella Jousseaume" Cousin (p. 262, p1. 3, fig. 5)
type locality: San Nicolas, Canton deMegia (Ecuador) (nomern nudum referring to carolita JOusseäume, 1887) (according to Clench, 1930).
Synonymy proposed by Te (1978).
Distribution: Mexico, Central America, West Indies and South America (Te, 1978); Ecuador (Jousseauine, 1887; Cousin, 1887); Peru (Anton, 1839; Küster, 1844)
8. Stenophysa panamensis (Küster, 1844)
1839 Physa panamensis "Mt%hlfeldt" Anton (p. 49)
type locality: Panama (nomem nudum) (according H. B. Baker, 1930) Synonymy proposed by H. B. Baker (1930).
1844 Physa panamensis "Mtthlfeldt" Mister (p. 11, p1. 2, figs. 3-5)
type locality: Panama.
1887 Aplecta gualbertol Cousin (p. 261-2, p1. 4, fig. 3)
type locality: Mapasing.a, Equateur. Synonymy proposed by Te (1978).
1889 Physa simoni Jousseaume (p. 253, p1. 9, figs. 3-4)
type locality: Laguna de Espino, near Caracas, Venezuela Synonymy proposed by H. B. Baker (1930). 26
Distribution: Mexico (including Baja California), Central America and Northearn South America (Te, 1978); Venezuela (Jousseaume, 1889; H. B. Baker, 1930); Ecuador (Cousin, 1887).
loosi"Holmberg, 1909.
1909 Physa loosi Holmberg (p. 28)
type locality: Santa Lücia,al pie del Cerro Pie de Palo, Provincia de San Juan, Argentina.
(status unknown)
Distribution: San Juan, Argentina (Holmberg, 1909), Chaco, Argentina (Fernández, 1981)
10. "Physa àspii"Holmberg, 1909
1909 Pysa asajiJ Holmberg (p. 28)
type locality: Laguna de los Murcilagos, Formosa Argentina.
(status unknown)
Distribution: Formosa, Argentina (Holmberg, 1909), Salta, Argentina (Fernández, 1981).
It should be emphasized here that these synonyms and distribution records have been taken directly from the literature without any evaluation my part. Te (1978) examined critically most of the South American species deposited in North American Museums. But there are several records in the literature such as those of d'Orbigny (1835 and 1837) which have not been verified by recent authors and may therefore contain errors of identification. - 27
CHAPTER 2. MATERIALS AND METHODS
2.1 SAMPLES OF PHYSIDAE
Material from the following malacological collections was examined. UnB: Laboratôrio de Malacologia, Departameñto de Biologia Animal, Universidade de Brasilia, Brasilia,'Distrito Federal, Brazil, curators: Warton Monteiro and Mara Lucia Ferreira Dias; MZ: Museu de Zoologia, Universidade de São Paulo, São Paulo City, São Paulo State, Brazil, curators: Jose Luiz Moreira Leme and Licia Pena-Nemme; WLP: Dr. Wiadimir Lobato Paraense, Malacological collection, Rio de Janeiro City, Rio de Janeiro State, Brazil, curator: W. L. Paraense; GT: George A. Te, private collection, Ann Arbor, Michigan, U.S.A. The material I studied comprises: 24 samples of UnB collection, 13 of them from seven Brazilian States, and the remaining 11 lots are snails from England and Scotland; 10 samples of MZ collection are from nine Brazilian States, including paratypes of Physa (Physella) paveroi Leme, 1966 (MZ 16618),and 'Physa (Physella) cubensis" sensu Leme (1966) (MZ 17997); 14 samples of WLP collection, where four are from Brazil, four from other South American countries, and six from the West Indies ; edh South American species representative shells of Dr. Te collection were also examined: five shells of Stenophysa species, one of P. cubensis, one of P. acuta, and one of P. venustula (Table 2.1). 28
Table 2.1. List of Physidae lots used in shell and anatomical analysis.
Code Species S A Locality .Voüdher N9
L 1 Sm 02 01 Carrpa?ia, Buenos Aires, UnB-159 and Argentina WLP-16
L 2 Sm 01 01 Tigre, Buenos, UnB-160 and Argentina WLP-18 L 3 Sm 05 01 Belo Horizonte, Minas UnB-169 and Gerais, Brazil WLP-1633
L 4 Sm 02 01 Rio San Miguel, thui, UnB-170 and Uruguay WLP-1921
L 5 Sm 01 01 D. Pedrito, Rio Gram- UnB-171 and de do Sul, Brazil WLP-1931 L 6 Sm 02 01 Usine-Lairntin, Marti- UnB-175 and nique WLP-2601
L 7 Sm 01 01 Saint Mary, Jamaica UnB-179 and WLP-2697
L 8 Sm 02 01 La Isleta, Panama UnB-180 and WLP-2745
L 9 Sm 06 01 Urucuca, Bahia, Bra- MZ-23357 zil
L 10 Sm 04 00 Curitiba, Paran, MZ-17112 Brazil
L 11 Sm 02 01 Corurnb, Mato Grosso UNB-111 do Sul, Brazil
L 12 Sm 01 00 Fortaleza, Cear, MZ-1453 Brazil
L 13 Sm 11 01 Recife, Pernambuco, UnB-126 Brazil 29
Table 2.1. (Continued):
* Code Species S A Locality : Voucher N9 L 14 Sm 03 00 Ponta Ipg Arcado, NZ-636 Cois, Brazil L 15 Sm 08 02 Bananal, Brasilia, DF, UnB-19 and UnE- Brazil 47 L 16 Sm 04 01 Taixca, Brasilia, DF, UnB-106 Brazil L 17 Sm 06 03 Museu Nacional, Rio UnB-107 de Janeiro, PJ, Bra- zil L 18 Sm 01 01 Itagual, Rio de Janei- UnB-80 ro, Brazil L 19 Sm 01 00 Joo Pessoa, Paraiba, UnB-105 Brazil L 20 Sm 01 01 Canabrava, Gois, UnB-124 Brazil L 21 Sm 01 01 Diamantina, Minas Ge UnB-211 rais, Brazil L 22 Sm 01 00 São Jose do Rio Pre- NZ-16498 to, Sao Paulo, Brazil L 23 Sm 01 00 Itü, São Paulo, MZ-17116 Brazil L 24 Sm 01 00 São Vicente, São Pau- MZ-17123 lo, Brazil L 25 Sm 01 00 Nioaque, Nato Grosso 1,4Z-17117 do Sul, Brazil L 26 Sm 01 00 Espirito Santo, Bra- MZ-23352 zil 30.
Table 2.1. (Continued).
Code Species S A Locality. ..•. .:• Voucher L 27 Sm 01 01 Narimbondo, Minas Ge- MZ-16499 rais, Brazil . L 28 Sin 01 00 Peru MZ-7235 L 29 Sm. 30 33 Fundo Creek Basin, UnB-006 Brasilia, DF, Brazil L 30 Pac 05 05 . Zoo, Brasilia, DF UnB-35 Brazil
L 31 Pac 07 02 Alto da Boa Vista, UnB-85 Rio de Janeiro, Bra- zil * .L 32 Pac 01 01 Sarco, Cochabamba, Bo UnB-157 and WLP livia
L33 Pac 01 02 Lima, Peru UnB-164 and WIP- 1338 L 34 Pac 01 01 Canal 6, Santos, Säo UnB-173 and WLP- Paulo, Brazil 1958 L 35 Pac 01 02 Saint George, UnB-174 and WLP- Trinidad 2573 L 36 Pac 04 01 San Huan, Puerto UnB-177 and wr.. Rico 2650 L 37 Pac 04 01 Rio Piedras, Puerto UnB-178 and WLP- Rico 2652 L 38 Pac 30 03 University of Brasi- UnB-l94 lia, Brasilia, DF Brazil L 39 Pac 16 03 Nanguinhos, Rio de UnB-103; UnB WLP- Janeiro, RT, Brazil 1381
L40 Pf 01 01 Mid Lothian, UnB-133 Scotland 31
Table 2.1. (Continued)
Code Species S A Locality Voucher L 41 Pf 01 01 Water of Leith, UnB-134 Edinburgh, Scotland
L 42 Pf 00 01 Union Canal, Edinburgh, UnB-137 Scotland
L 43 'Pac 00 01 River Aire, Leeds, UnB-138 England
L 44 Pg 00 01 Chester, Liverpool, UnB-139 England
L 45 Pac 00 01 Hazeihead Park, UnB-141 Aberdeen, Scotland
L 46 Pac 03 03 Richmond Park, UnB-142 London, England
L 47 Pac 00 01 Sussex, England UnB-143 L 48 Ah 00 01 Sussex, England UriB-144
L 49 Pf 00 01 Union Canal, UnB-148 Edinburgh, Scotland
L 50 Pf 00 01 East Lothian, UnB-149 Scotland
L51 Sm 01 00 - cr-M63
L52 Si 01 00 - GT-613
L 53 Spa 01 00. - . Cr - T149
L 54 Spe 01 00 - Cr - 612 FCr
L 55 Ss 01 00 - Cr - .611 EY'T
L56 Pa 01 00 - GT-643
L57 Psi 01 00 - Cr-T127
L 58 Phc 01 00 - CT - M230
59 . L Pp 03 02 Pargue D. Pedro II, MZ - 16618 Säo Paulo City, Säo Paulo State, Brazil 32
Table 2.1. (Continued)
Code Species S A Locality Voucher N9 L 60 Pac 01 02 •Instituto Oswaldo MZ-17997 Cruz, Rio de Janei ro City, Rio de janeiro State, Brazil
L 61 ism 00 01 Ponte Sobradinho, UnB-18 Brasilia, DF, Brazil
L 62 Sm 00 01 Represa Parano, Bra- UnB-20 silia, DF, Brazil L 63 Sm 00 02 L 2 Norte, Brasilia, UnB-43 DF, Brazil L 64 Sm. 00 02 Manaus, Amazonas, thB-167 and WLP Brazil 1451 L 65 Sm 00 02 Bel&n, Pars, Brazil UnB-220 L 66 Pac 00 01 Jacarepaguä, Rio de UnB Janeiro, Brazil *No voucher number quoted, A = number of analysed snails for internal anatomy, S = number of analysed shells. The abbreviations for the species are: Ah = Aplexa hypnorum, Pa = Physella acuta, Pac = Physella acuta complex, Pg = Physella gyrina, Pc = Physella cubensis, Pf = Physa fontinalis, Pp = Physella papaveroi, Pv = Physella venustula, Si = Stenophysa irrpluviata, Sm = S. marmDrata, Spa = S. panarinsis, Spe = S. peruviana, Ss = S. spiculata. The abbreviations for the malacological collections are: GI' = George A. Te, private Collection, Ann Arbor, Michigan, U.S.A.; MZ = Museu de Zoologia da Universidade de São Paulo, São Paulo City, São Paulo State, Brazil; UnB = Laboratôrio de Malacologia, Departamento de Biologia Animal, UnJ.versic1ade de Brasilia, Brasilia, DF, Brazil; WLP = Wladimir Lobato Paraense, Malacological Collection, Instituto Gewaldo Cruz, Rio de Janeiro City, Rio de Janeiro State, Brazil. 33
This study was mainly performed on snails deposited in museums, with the exception of the ones from Brasilia, Federal District and Rio de Janeiro City, Rio do Janeiro State, which were collected from March to November 1978 (Monteiro & Dias, 1980), and the ones from England and Scotland, which I have gathered in the year of 1979. They were deposited in the UnB malacological collection. A metal net scoop having 14 cm in diameter, 8 cm deep, with a meter long wooden handle, was utilized for collecting snails in water and on aquatic plants. The shell and soft anatomy of Physa fontinalis, Physella acuta complex, Physella gyrina and Aplexa hypnorum, comprising six lots from Scotland and five from England (Table 2.1), were studied to provide 'a better understanding of several character states, since Te (1978) only cites typical species for the definition of the great majority of the character states. These samples have also provided a background on the anatomical and shell character patterns of some Physidae groups not present in South America. The representative shells of Dr. Te collection were utilized in the identification of the material studied. The examined museum collections do not represent random samples of the surveyed population. Besides, mature individuals could not be differentiated from irnznatures based on shell morphology. 34
2.2. SHELL MEASUREMENTS
This study was based on 24 shell characters which are related to those studied by Te'(i978) (Table
2.11). A diagram showing the main parts of A Physidae shell is given in Figure 2.1. They were grouped and enumerated in a sequence different from that-given by Te (1978). Five of these characters (S-1 to S-5) relate to shell size and shape; nine (S-6 to S-14) are related to spire size and shape; seven (S-15 to S-23, except S-18 and S-22) refer to aperture size and shape; after initial analysis, characters S-18, S-22, and S-24 were excluded from further shell analysis due to redundancy with characters S-is, S-16, and S-3, respectively. Not all characters presented by Te (1978) were utilized in this work: 29 of them were discarded, and those six that correspond to the ones Iworked with are listed in Table 2.11. See section 2.2.3 for comments on discarded characters. To represent,A best my observations, all utilized characters have undergone modifications in their definition, character states, and/ or on the way to be analysed. To count the number of whorls (Character S-i) the shell must be oriented in a way to be observed from the apex. On the other hand, since characters S-2 to S-24 involve straight length measurements, it requires an observation of the shell in its fundamental position. To measure those characters I followed a technique similar 35
Table 2.11. List of shell characters used for Physidae taxonanic assessment, indicating the correspondence between the characters as used in this work and those used by Te (1978).
Shell characters
This work TO (1978) CODE SYMBOL N A M E CODE S- 1 NW Number of whorls SC-25 s-.2 L1S2 Shell length SC-26 5- 3 SW Shell width - 5- 4 MIM2 Band of constant width - S- 5 Llr1 Apex to shell midpoint - S- 6 L1D1 Spire length - S- 7 DD2 Spire width - 5- 8 C2D1 Basal spire length - S- 9 C1LD Spire angle SC-31 S-lO BEE1 Sutural angle SC-32 S-li CA2 Point C to tangent A2 - S-12 BC Extremes of 3rd and 4th sutures - S-13 TU Maximum distance CD to 4th whorl - S-14 CD End 4th whorl to end basal suture - S-15 SS3 aperture width - S-16 D/S2 Aperture length - S-17 H1H2 Apical aperture width - * S-18 1112 Basal aperture width - S-19 D3DD4 Aperture insertion angle SC-33 S-20 D6D5D7 Adapical aperture angle - S-21 XY Distance to DZ - * S-22 DZ Basal suture to lower outer lip - S-23 F3G3 Parietal wash width SC-14 * S-24 FG Shell maximum width -
* Characters S-18, S-22 and S-24 were excluded due to redundancy with characters S-15, S-16 and S-3, respectively. 36
IT
L
OUTER LIP
APER
FIGURE 2.1. Shell of Physella acuta complex, from University of Lrasilia, Brasilia, DF, UnB-194, showing the main parLs of a hysidae shell. - . 37 to Te's 1975 and .1978. The main difference is that I filled the shell completely with water up to the peristome. Thus, the shells were leveled and'-steady, precluding slight differences in orientation due to.tilt, that, ultimately, would affect shell measurements. This position is so naturally stable that drawing ten times a same shell, that was removed from its original position, emptied and filled with water again, for ten successive times, produced ten superimposable drawings.
2.2.1. FUNDAMENTAL ORIENTATION OF THE SHELL
The determination of the shell main axis is fundamental for almost every measurement of shell characters. A leveled and steady shell is oriented with the opening on the observer's left side, in a way that the edge of the peristome is frontal to the observer. This is called fundamental position of the shell (Figure 2.2). The next fundamental step is the determination of the spire angle (character S-9). To do so I have looked for an angle that would represent best the shell shape as a whole. Searching the conchological literature I found that spire angle, as defined by Parodiz (1951), is the one that seems to contain most points of the spire development. However, his definition sets the shell apex as the vertex of the spire angle. Since the shell apex FIGURE 2.2. Shell of Physella acuta complex, UnB-194, in its fundamental orientation. A1C1 (rigth side) and AD (left side) are points determining spire angle sides; C1LD, spire angle; LS1, spire angle bisector; L1S1, shell axis; two parallels and two perpendicular lines to the shell axis enclose the shell within a parallelogram frame. L1S2, shell length; SW, shell width; F1G1 and F2G2, upper and lower limits, respectively, of a band where the largest segment of the shell width is constant; M, midpoint of the width of the band; F3G3, parietal wash width; L1D1, spire length; DD2, spire width; C2D1, basal spire length. 38
L
1
7 ..
1mm
C cl
ZBlCl D2
//N ND1
8 Fl Ml Is 1 \ ---- GG
2 M 2 G2 KF
7 S1
LEFT SIDE rS2 RIGHT SIDE
1mm - 39 is often blunt, the vertex is not referred to a point but to a curved, surface. ConsequentlyI modified Parodiz's definition to draw the spire angle as close as possible to the spire development. Besides, I have chosen: angle sides determined by well defined crosslinés, i.e., points where sutures meet shell contour in a plane, A1C1 and AD (Figure 2.2). These are the reasons that led to the determination of the shell axis based on a consistent position of the shell. This new method is close to many others found in the literature which are ultimately arbitrary and variable (Gaillard, 1973; Parodiz, 1951; Te, 1978). Thus, the spire angle C1LD is defined by the straight lines passing by the points A1C1 and AD (Figure 2.2), determined by the ends of the first suture (Al and A) and the ends of the basal suture(Cl and D), on each side of the shell in its fundamental position. These two straight lines cross at the vertex of the angle (L) that is close to the shell apex. According to this definition, the vertex of the spire angle is rarely located on the shell apex. In fact, any definition that necessarily include/that coincidence is operationa disadvantageous, since the shell apex is not referable to a unique point. The bisector of the spire angle defines two important marks of the shell: the shell apex and the shell axis. The shell apex is the point where the bisector crosses the upper line of the 40 protoconch (Li, Figure 2.2). The shell axis is the segment (L1S1, Figure 2.2) of the bisector from the protoconch to the collumelar lip. The determination of two parallel and two perpendicular. lines to the shell axis is very convenient as references to direct all the shell measurements. One parallel is tangent to the outer lip and the other tangent. to the body whorl. One perpendicular is tangent to the shell apex and the other to the aperture base. These four lines will enclose the shell within a parallelogram (Figure 2.2).
2.2.2. CHARACTERS DEFINITION
The code number of the character, as used by Te (1978), is given in parentheses after the name of each character, only for characters equivalent to those used by Te. Character S-i. Number of Whorls (SC-25) - The number of whorls of a shell is a very important character. However, the way it is counted, according to the literature, is far from uniform. The first step to count whorls is to determine the first whorl. To do so, it is necessary to follow the way the shell is built. In fact, the shell material begins to be gathered at a certain point during the embryonic life. This point is indicated by an arrow (0) in a stylized drawing (Figure 2.3). This arrow is fixed and shows the origin of the 41
W'U
FIGURE 2.3. Diagram to illustrate a method of counting whorls (1 to 3) in a stylized drawing of stages of shell formation (1 to 6). 0, origin of the protoconch; P,
direction of the growing zone. 42 first whorl. The growing zone direction, may be indicated by another arrow (p) that coils until it touches the first arrow (o). When the growing zone reaches the arrow that indicates the origin of the whorl, it closes a cycle, which is the first whorl. The following coilings always close further cycles (whorls) at the fixed arrow (Figure 2.3). Dealing with shells to be examined, they are fixed to a substract of plasticine in a vertical position so that its apical end can be observed under a stereomicroscope. It is necessary to choose a line to set the beginning and the end of each whorl. Thus, the curve of the tip of the protoconch was considered the origin of that reference line. Though this origin is not precise, it is a natural origin of the shell, and a reference line must begin at that point and follow the direction opposite to the growing pathway. The last whorl is measured to the nearest quarter of a complete 3600 turn (Figure 2.4).
Character S-2. Shell length (SC-26) - The length or height of the shell, L1S2 (Figure 2.2),is the distance from its apex to the point where the shell axis crosses the perpendicular that is tangent to the aperture base.
Character S-3 Shell width -SW (Figure 2.2), is the distance between the parallels to the shell axis that are tangent to the outer lip and to the bodywhorl. This 1. character indludes character S-24. / 43
FIGURE 2.4. Top view of a shell of Physella acuta complex, UnB-194, showing method of counting whorls from 1 to 5 and 3/4. The numbers are in the end of the respective whorls. 44
Character S-4. Band of constant width - The largest segment, perpendicular to the shell axis, lying between the outer lip and the body whorl..'periphery is determined empirically in a plane. Some physids have;:a band where the length of that segment is constant. The distance between the upper and the lower limits of the band (F1G1 and F2G2) is a band of constant width (M1M2) (Figure 2.2).
Character S-5. Apex to shell midpoint - The midpoint, M (Figure 2.2),of the width of the band defined in character S-4, is located along the shell axis. The distance from the shell apex to the midpoint of the band, L1M, defines character S-5.
Character S-6. Spire length - A perpendicular to the shell axis passing by the end of the basal suture, D (Figure 2.2),crosses the shell axis at Dl. The spire length L1D1, is the distance from the shell apex to Dl.
Character S-7. Spire width - The perpendicular to the shell axis passing by the end of the basal suture (D) determines D2 at the body whorl. The distance between the end of the basal suture to D2, DD2, is the spire width (Figure 2.2). '
Character S-8. Basal spire length - Two parallels, perpendicular to the shell axis, were drawn, respectively, by point D (as in character S-7) and by point C (Figure 2.2). The distance between the parallels, C2D1,is called basal spire length. 45
Character S-9. Spire angle (SC-31) -As defined in the fundamental orientation of the shell, spire angleClLD, is determined by straight lines passing by the points Aid and AD that pass by the ends of the first suture (Al and A) and the ends of the basal suture (Cl and D), on each side of the shell in its fundamental position (Figure 2.2).
Character S-b. Sutural angle (SC-32) - The shell axis crosses the third suture at E (Figure 2.5). This point and the suture extremity (B,Figure 2.5) on the left side of the shell in its fundamental position determines the straight line BE. The angle formed on the same side by BE and a perpendicular to the shell axis, at E, is the sutural angle, BEE1 (Figure 2.5). The conchological literature refers to sutural slope as equivalent to sutural angle (Moore, 1960; Te, 1975).
Characters S-il and S-12. Point C to tangent A2, and extremes of 3rd and 4th sutures, respectively - A line, on the left side of the shell, tangent to the fourth whorl and to the body whorl was drawn (Figure 2.6). B and C are extreme points of the third and fourth sutures, respectively, also on the left side of the shell. The distances from C to the tangent (A2) and from B to C were measured.
Characters S-13 and S-14. Maximum distance CD to 4th whorl contour, and end of 4th whorl to end of basal suture, respectively - The end of the fourth whorl (C) and the 46
FIGURE 2.5. Shell of Physella acuta complex; UnB-194, showing angular measurements. BELl, sutural angle; D3DD4; aperture insertion angle D4E'D5, adapical aperture angle. 47 end of the basal suture (D) (Figure 2.7) on the left side of the shell were linked by a straight line. The distance from points C to D, character S-13, was measured. The most distant point from CD to the fourth whorl was empirically determined. This distance measured as TU (Figure 2.7) defines character S-14.
Character S-15. Aperture width - Aperture width (Figure 2.8) is here defined as the distance between two parallels to the shell axis. One is tangent to the outermost point of the outerlip. The other is tangent to the outermost point of the aperture on the internal columellar lip edge. Aperture width is the distance SS3 (Figure 2.8). This definition differs from Te's suggestion in a drawing (Te, 1975) that includes the columellar lip width in the aperture width.
Character S-16. Aperture length - Is the distance between two parallels perpendicular to the shell axis: one is tangent to the aperture base; the other passes by the point where the outer lip meets the inner lip (D) and crosses the shell axis at Dl (Figure 2.8). The aperture length is the distance from Dl to S2.
Character S-17. Apical aperture width - The aperture length was divided into four equal bands perpendicular to the shell axis. H1H2 (Figure 2.8) in the inferior limit of the upper band is the apical aperture width.
Character S-18. Basal aperture width - Ii 12 (Figure 2.8) in the superior limit of the lowest band, as mentioned in FIGURES 2.6 (above) and 2.7 (below). Shells of Physella acuta complex, UnB-194. CA2, distance from point C to tangent A2; BC, extreme points of the third and fourth sutures; CD, distance from the end of the fourth whorl and the end of the basal suture; TU, most distant point from CD to the fourth whorl contour on the left side of the shell drawing. 48
1mm 49
D D1
12 ,,1
S3
9 1mm fIGURE 2.8. Shell of Physella acuta canpiex, UnB-194. SS3, aperture width D1S2, aperture length; 111H2, apical aperture width, 1112, basal aperture width. 50 character S-17) is the basal aperture width.
Character S7-19. Aperture insertion 1angle (SC-33) - The vertex of this angle is the end of' the basal suture on the left side of the shell in its fundamëntal'position (D, Figure 2.5). The sides of the angle were empirically found over the whorls adjacent to the end of the basal suture, by plotting points D3 and D4 on the left margin of the body whorl and outer lip, respectively. The distance D3D or D4D is equal to 5% of the length of the examined shell. Thus, the lines DD3 and DD4 define the sides of the aperture insertion angle D3DD4 (Figure 2.5). This seems to be the angle that aproximates better to the angle that Te (1975) presents in a drawing.
Character S-20. Adapical aperture angle - The definition of this character is analogous to the one of character S-19. The vertex of the angle is the point D (Figure 2.5) where the outer lip meets the parietal lip. The sides of the angle were found by plotting points D4 and D5 on the outer and parietal lips, respectively. The distance DD4 'or DD5 is equal to 5% of the length of the examined shell. Thus, the lines DD4 and DD5 define the sides of the adapical aperture angle D4DD5 (Figure 2.5).
Characters S-21 and S-22. Distance to DZ and basal suture to lower outer 42 respectively - The shell is placed in its fundamental position (Figure 2.2), it is then rotated around its axis until the external part of the outer lip covers completely the columellar lip (Figure 2.9). A 51
z. tmm
FIGURE 2.9. Shell of Physella acuta complex, UnB-194. DZ, distance from the end of the basal suture to the lowest point of the outer lip; XY, largest length perpendicular to DZ. 52 straight line link.ing the end of the basal suture (D) and the lowest point of the outer lip (Z) is drawn (Figure 2.9). The distance from point D toZ, character S-22, was measured. The largest length perpendicular to DZ; limited by this line and the outer lip contour, is found. This length, XY, defines character S-21.
Character 5-23. Parietal wash width (SC-14) - The character definition was used as proposed by Te (1975), however, the states as suggested by him were discarded due to difficulties of categorizing parietal wash according to those subjective states. It is difficult to any taxonomist to understand differences between a narrow, regular, wide, very wide and very very wide parietal wash. Thus, I suggest to quantify parietal wash width as follows. The perpendicular to the shell axis which contains the shell midpoint (Character S-5) crosses the parietal wash limits at F3G3 (Figure 2.2). This segment F3G3 is the measurement
of the parietal wash width.
2.2.3. DISCARDED CHARACTERS
For the reasons to be mentioned in the following paragraphs there were some characters studied by Te (1978) which were not considered in this work. The characters that refer to general and specific states - as shell shape, spire shape, aperture shape - are a single character split into two by Te (1978). 1 understand that 53
the separation of the character under general and specific implies granting double weight for a single character. This is not a proper procedure in numerical taxonomy, which presupposes equal weight for all characters (Sneath & Sokal, 1973). The characters specific shell shape, specific spire shape and specific aperture shape were discarded for their character states were innacurately defined. They are more precisely evaluated by shell measurements. All the variables which Te (1978) expressed as ratios were also discarded. In fact, to represent a variable as a ratio ens us dis,antages (Pearse, 1965; Atchley et al., 1976). Since a ratio is a two component variable the variation of one component may not follow the same ratepf the other. Besides the ratio's distribution often sard unusual, perhaps far from normal (Pimentel, 1979). The character aperture outer lip callus colour was lumped with character aperture outer lip callus since not all shells have outer lip callus. The characters translucency, glossine, shell Sk thickness, subsutural line colour and protoconch colour are known to be variable within fresjater snail species and may be correlated with certain ecological features of the environment, such as the amount of calcium in the water, water acidity, water turbulence (Baker, 1911; Wurtz, 1949; Hubendick, 1951; Pennak, 1953; Paraense, 1970). Also the limits of their character states are 54 difficult to determine precisely since they may show varying states of intergradation. Consequently, they were discarded. The character protoconch shape, was also discarded. The term protoconch is applied to initial whorls of the spire, specially to those cases where the initial whorls differ from the rest of the spire in terms of orientation, spire angle and sculpture (Moore, 1960). However, in shells like those of the physids where initial whorls do not differentiate from the rest of the spire, it is very difficult to delimit where the protoconch ends. Te (1978) has not explained his criteria to delimit protoconch, and for this reason this character has been discarded. The characters suture appression and aperture truncation point were not utilized in this study due to insufficiency of information on their methodology.
2.3. SHELL STATISTICAL ANALYSIS
2.3.1. REGRESSION ANALYSIS
The regression of pairs of shell characters was calculated for three samples of snails, identified beforehand as S. marmorata by the examination of the internal anatomy, from Recife (Pernambuco State) (Voucher n9 UnB-126), Uruçuca (Bahia State) (MZ-23357), and Belo- 55
Horizonte (Minas Gerais State) (UnB-169) and one sample of Physella acuta complex, also identified by the internal anatomy, from Manguinhos in Rio de Janeiro City (Rio de Janeiro State) (UnB-103, 166) (Table 2.I)1. These: samples were compared with the conspecific samples from Brasilia, i.e., with S. marmorata from Fundo Creek Basin (UnB-006), and P. acuta complex from the University of Brasilia aquaria (UnB-194), by a regression coefficient test (Allen, 1976). The same test was used to compare the S. marmorata Fundo Creek population with the P. acuta complex, University of Brasilia population. The correlation coefficients and the respective tests of significance were calculated for twenty-one shell characters (Table 2.11), associated in 210 pairs, in a correlation matrix for both Fundo Creek and University of Brasilia samples. At first, these correlation matrices were arranged considering 24 characters. However as three of them were considered redundant to others, they were not taken into account in this work: character 15 is redundant to 18, 24 to 3, and 22 to 16. As characters 15, 3 and 16 were easier to be measured they were chosen instead of 18, 24 and 22 respectively. The correlation coefficients, the respective significance tests and the regression analysis were considered only for those pairs of characters whose correlation coefficients proved to be significant, and above 0.7, for the Fundo Creek (Stenophysa)and the University of Brasilia (Physella) samples, simultaneously. 56
Consequently, the coefficient of determination (r 2 ) should be always higher than 0.49. There was no clear basis for designating in each pair of variables one of them as the independent variable, and the other as the dependent variable. Therefore, the standard error of estimate was calculated reversing: the roles of the variables, first considering Y as a dependent variable and X as the independent one, and then considering X as the dependent variable and Y as the independent one. The choice of the dependent and independent variables was made on the basis of the X, Y or Y, X pair which presented the smallest standard error of estimate. The regression equations were determined by the method of least squares. All these statistical procedures were performed on a SPSS program (Statistical Package for the Social Sciences) (Nieetal, 1975) by a Burroughs 6700 computer, except for the regression coefficient test for which a special algol program was prepared based on Allen (1976).
2.3.2. PRINCIPAL COMPONENT ANALYSIS (PCA)
The 19 shell characters used are shown in Table 2.11, which also includes characters S-14, S-lB, S-22, S-23, and S-24 that were eliminated from the analysis njdata were lacking for two populations, or were excluded due to redundance with other characters. 57
At first, S. marmorata from 29 localities (Li to L29), amounting 102 shells, and P. acuta complex from 10 localities (L30 to L39), amounting 70 shells (Table
2.1) were independently considered for the PCA. Later, the 39 localities belonging to S. marmorata and P. acuta complex were analysed together. - The correlation of the shell characters with the five principal components were only considered for those characters which had correlation coefficients higher than half the highest correlation coefficient for each principal component.
Principal component analysis was carried out in a Burroughs 6700 computer, using a Fortran language programme of Bouzereau and Blank (personal communication),. based on Calliez and Pagis (1976). Characters were standardized by expressing each state as a deviation from the character state mean in standard deviation units.
2.4. ANATOMICAL ANALYSIS
The anatomical analysis was made on S. marmorata from 32 localities and P. acuta complex from 14 localities. Forty eight and 28 snails were used in the S. marmorata and P. acuta complex analysis, respectively, varying from one to thirty specimens per locality (Table 2.1). 58
Snails from UnB and WLP collections were all relaxed in a 0.05% nembutal solution, for four to six hours, according to the size of the. specimens. They were gradually immersed in water at about 70°C,. for. approximately 45 seconds and, then, drawn-out of the shell with a forceps. They were preserved in Railliet-Henry's fluid for at least 24 hours before dissection, and the larger specimen of each sample was dissected under stereomicroscope (Paraense, 1981). This study was only carried out with preserved material. Snails from MZ collection were preserved in 70% alcohol. This study was performed over 26 out of the 34 anatomical characters presented by Te (1978). These characters and their states are listed in Table 2.111. They were enumerated in a different sequence from that given by Te (1978). See section 2.4.2 for comments on discarded characters. Of the 26 anatomical characters,
(Table 2.111), 0i is related to the tentacle pigmentation (A-l); another to the mantle surface pigment pattern (A-2); two to the columellar side mantle lappet, which are related to the type of mantle lappet and to the columellar lappet number (A-3 and A-4); two to the kidney features, related to the kidney shape (A-5) and tubular nature of the kidney (A-6); three to the bursa copulatrix, related to the bursa copulatrix shape (A-7), bursa copulatrix axis (A-8) and bursa copulatrix duct connection (A-9); two to the digestive tract, related to the pigmentation pattern (A-b) and shape of gizzard 59
Table 2.111. Anatomical characters and their states used in this ulork for Physidae taxonomic assessment (based on Te, 1978).
Characters States Code Name A-i. Tentacle type entire and sircoth (as in (Figures 2.10 p1exa elongate) translucent, with narrow black core (as in Physella integra)
• translucent edged with broad black core (as in Physella ancillaria) white (as in Physella parkerii)
A-2. Mantle pignnt obscure circles (as in A. pattern elongate) (Figures 2.14 - small circles (as in Physella 2.20) osculans) smoothly pigmented (as in Physella zionis) unpigirnted (as in Physeila spelunca) large circles (as in Physella gyrina) very large irregular areas (as in Physella lordi) medium circles (as in P. integra)
A-3. Mantle lappet type A. wavy, non digitate (as in
(Figures 2.21 - Stenophysa maugeriae) 2.28) Table 2.111. (Continued)
Characters States Code Name
is a shorter version of state A; and was lumped with state A serrated, non digitate (as in Stenopriysa marrrorata) straight-edged, non digitate (as in A. elongata)
blunt digitate (as in Physa fontinalis)
stubby digitate (as in P. zionis) pointed digitate (as in P. integra)
spade-shaped digitate (as in P. gyrina straight-edged and wide (as in Physella microstriata)
A-4. Columellar lappet A. none (as in A. elongata) number B. high (6 to 9) (as in P. fontinalis) medium (4 to 5) (as in P. pornilia)
low (2 to 3) (as in P. zionis) 61
Table 2.111 (Continued)
Characters States Code Nau A-5. Kidney shape (as in Stenophysa maugeriae) (Figures 2.29 - 2.39) (as in A. elongata) (as in Physella virgata) (as in P. ingra) omitted, lumped with state C (as in P. fontinalis) (as in P. pond (as in P. gyrina) (as in P. lordi) (as in P. zionis) (as in P. lunca)
(as in P. oostata)
A-6. Tubular nature of full tubular to semi-lamellar the kidney (Figures = A-5, states C, D, F, G, H, 2.40 - 2.42) I, J, K, L. lamellar = A-5, state A crinkled = A-5, state B
A-7. Bursa cxpulatrix (as in Stenophysa peruviana) shape (Figures (as in Stenophysa marmDrata) 2.43 - 2.50) (as in A. elongata) lumped with state A, omitted (as in P. fontinalis) (as in Physella cubensis) 62
Table 2.111. (Continued)
Characters States Code i"arre
(as in Physella heterostropha)
(as in Physella boucarth) L. (as in P. nuttallii) A-8. Bursa copulatrix vertical (as in A-7, all axis (Figures 2.51 - states except H)
2.52) horizontal (as in A-7, state H) '
A-9. Bursa copulatrix (as in S. rnarirrata)
duct connection (as in A. elongata) (Figures 2.54 - (as in P. 9yrina) 2.56)
A-10. Digestive tract lightly pigmented (as in S. pigmentation marrtorata)
2.57 - (Figures evenly pigmented (as in 2.60) A. elongata)
partially pigmented (as in P. gyrina)
unpigmented (as in P. spelunca)
A-li. Shape of gizzard cylindrical (as in S. marrrrata) (Figures 2.61 - laterally lobed (as in P. 2.62) gyrina)
A-12. Preputial gland absent (as in A. hypnorum)
presence present (as in P. fontinalis) 63
Table 2.111. (continued)
A-13. Penial complex type ((as in S. maugeriae) (Figures 2.63 - (as in Stenophysa irnpluviata) 2.81) (as in P. boucardi)
(as in CIrU-85 unnamed) (as in P. costata) (as in P. fontinalis)
(as in P. vir) (as in P. heterostropha) (as in P. ancillaria) (as in Physella propiva nuttallii) (as in P. gyrina, P. gyrina sayii)
(as in P. zionis) (as in P. cubensis) (as in P. integra)
0. (as in P. lordi)
(as in S. maniorata) (as in Stenophysa pariamensis)
(as in Aplexa hypnorum)
(as in A. elongata)
A-14. Length ratio of PS greater than 3 times Pr penial sheath to PS lesser than 3 times Pr the preputium (PS/Pr) 64
Table 2. 111. (Continued)
Characters States Code Name Narrowing of the narrowed (as in P. cubensis) non glandular regular (as in P. gyrina) segment of penial sheath (Figures 2.83 - 2.84)
Nature of the swollen and distended (as in P. glandular segment boucardi) of the penial sheath regularly swollen and smooth (Figures 2.85 - 2.87) (as in P. ina; P. ancillaria and liaria) short and rounded (as in P. cubensis)
Length ratio of the PS Ca. 2 times Pr penial sheath to the PS Ca. 2.5 times Pr preputiin (PS/Pr) PS ca. 3 times Pr D. PS Ca. 1.5 times or less than Pr
A-18. Length ratio of the G greater than NG (as in P. non glandular lordi) segment to the G Ca. equal NG (as in P. glandular segment of gyrina sayii) the penial sheath G lesser than NG (as in P (NG/G) gyrina) G much lesser than NG (as in P. cubensis) 65
Table. 2.111. (Continued)
Characters States Code Name A-19. Swelling of glandular A. very heavy (as in P. }x)ucardi) segment (G) of penial B. heavy (as in P. ancillaria sheath (Figures 2.88 ancillaria) - 2.91) slight (as in P. vinosa) no swelling (as in P. gyrina)
Size of preputial medium (as in P. heterostroph gland small (as in P. acuta) large (as in P. integra) -
Tendency of preputial oval and compact, no gland to flatten apparent flattening (as in P. (Figures 2.92 - 2.94) heterostropha halei) flattened (as in P. acuta rounded and loosely packed (as in P. cubensis)
Length ratio of the PS shorter than Pr penial sheath to PS Ca. equal Pr the preputiurn (PS/Pr) PS longer than Pr
Swelling of penial not bulbous (as in P. sheath terminus virgata) (Figures 2.95 - 2.97) bulbous (as in P. integra integra)
heavily bulbous (as in P. integra brevispira) ORE
Table 2.111 (Continued)
Characters States Code Name Retractor muscles fleshy (as in S. manirata) (Figures 2.98 - 2.99) fibrous-like (as in sane A. elongata)
Length ratio of penial PS Ca. 1.5 to 2 times Pr sheath to the PS less than 1.5 to equal Pr preputium (PS/Pr) PS less than Pr D. PS 2 to 3 times Pr
A-26. Number of segirents A. 2-parted (as in S. rnaugeriae) in the penial sheath B. single-parted (as in A. (Figures 2.63, 2.68 hypnorum) and 2.81) C. 3-parted (as in A. elongata) 67
(A-li); and the remaining 15 are related to the penial complex (A-12 to A-26).
2.4.1. CHARACTERS DEFINITION
Te (1978) did not present definition and description of the anatomical characters and their states. The full description of each character and its states is given, therefore, in this work, and is based on Te's manuscripts and sketch drawings (Te, personal communication). In fact, the great majority of anatomical sketch drawings presented in this chapter were made by Te. The detailled explanations of the characters and of their states will also help to better understand Te'S 1978 paper. The descriptions of the characters and/or their states are as follows. The code number of the character as used by Te (1978) is given in parenthesis after the name of each character:
Character A-i. Tentacle type (AC-3) - This character is related to the tentacle colour pattern. State A. entire and smooth-refers to tentacles which are opaque, mucous and, consequently, with smooth appearance (Figure 2.10); State B.translucent with narrow black core - the internal structure of the tentacle exhibits a pigmented core, whose width occupies 50% or less of the tentacle diameter (Figure 2.11); State C.translucent with broad black core - the pigmented tentacle core occupies
2.10 2.11 2.12 2.13 2.14 2.15 A A A--- A A-= 2.16 2.17 2.18 2.19 2.20
ç; 2.21 2.22 2.23
2.24 2.25 2.26
/E= 2.27 2.28
FIGURES 2.10 - 2.13. Tentacle type, states A to D. FIGURES 2.14 - 2.20. Mantle pigment pattern, states A to G. FIGURES 2.21 - 2.28. Mantle lappet type, states A, C, U, E, F, G, H, I. These drawings are after Te (personal communication) 69 more than 50% of the tentacle diameter, and a translucent sheath covers the tentacle (Figure2.12); State D. white - the tentacle core and the outer sheath are not pigmented. Due to density differences their limit m4y be easily seen (Figure 2.13).
Character A-2. Mantle pigment pattern (AC-4) - The character was divided in the following states (Table 2.111): State A. Obscure circles - the unpigmented circles are covered by a pigmented layer of variable density (Figure 2.14); State B. Small circcles - the mantle is most pigmented with small unpigmented circles (Figure 2.15); State C. Smoothly pigmented - the mantle is
totally pigmented (Figure 2.16); State D. tJnpigmented - the mantle shows no trace of pigment (Figure 2.17);
State E. Large circles - the mantle has more than 50% of its area with unpigmented circles (Figure 2,18); State F. Very large irregular areas - the mantle has wide unpigmented irregular areas covering most of the surface (Figure 2.19); State G. Medium circles - the mantle has approximately 50% of its area covered with unpigmented circles (Figure 2.20).
Character A-3. Mantle lappet type (AC-11) - In the shell aperture there are two mantle reflexions over the inner lip edge, called lappets or digits. To be classified the lappets were drawn in camera lucida from their dorsal view. The different states are given in Table 2.111 and illustrated in Figures 2.21 to 2.28. I %.F
Character A-4. Columellar lappet number (AC-13/14) - Columellar lappet number was recorded directly from specimens or from their drawings. Though lappet number varies within a population, columellar and parietal lappet number are highly correlated with each other. Thus only columellar lappet number was scored. The states are given in Table 2.111.
Character - A-5. Kidney shape (AC-15 and AC-16) - to study the kidney shape its contour was drawn in camera lucida, the different states are given in Table 2.111 and illustrated in Figures 2.29 to 2.39. This character was uti1i.zed only for the study of the Fundo Creek population and later discarded.
Character A-6. Tubular nature of the kidney (AC-17) - This character is related to the morphology of the middle section of the kidney and may be categorized in three states. State A. full tubular to semi-lamellar - is distinct when the whole tubule may be visible all along the kidney. In that case the tubule seems to be turgid (Figure 2.40); State B. lamellar - some segments of the tubule walls may agglutinate with its adjacent segment, giving to them the appearance of a lamellar structure (Figure 2.41); State C. crinckled - the middle section has a crumpled appearance (Figure 2.42).
Character A-7. Bursa copulatrix shape (AC-25 and AC-26) - To study the bursa copulatrix shape its contour was drawn in camera lucida. The different states are given in Table 2.111 and illustrated in Figures 2.43 to 2.50. This I!
-. 229 230 231.
2.32 2~33 2.54
2.36
237 .238c439
2.40
LI M E~~ 2.41.
'IGURLS 2.29 - 2.39. Kicney shape, states A 1 B, C, D,
- ', G, H, I. J, i, L. FIGURES 2.40 2.42. Tubular nature of the kidney, states A to C. These drawings are after Te (personal communication) FIGURES 2.43 - 2.50. Bursa copulatrix shape, states A, B, C, E, F, G, H, L. These drawings are after Te (personal
communication). 72
2.44 2.43 2.46
2.49
2.50 73
character was only utilized for the study of the Fundo Creek population and later discarded.
Character A-8. Bursa copulatrix axis (AC-27) - To study the characteristics of the bursa copulatrix, the mantle was removed and the female reproductive system was dissected. Then, bursa copulatrix body and its duct, vagina, and lower third of the oviduct were drawn in camera lucida. Measurements were made over these camera lucida tracings: length and width of the bursa's body, and the length of the bursa's duct. The measurement of the length of the bursa's body was obtained following the direction of the duct's entrance in the body, according to the longer dimension of the bursa's body. In cases of crooked bursa's body, the length followed a direction up to the inflexion region, then changed the ,direction to the tip of the bursa's body. The sum of these segments represent the length of the bursa's body. The width of the bursa's body was obtained following the longer direction crossing the length. The total length of the bursa's duct was obtained in an analogous manner as in the case of bursa's crooked body length. The location of the bursa's body and the relative orientation of the duct's entrance were observed from surface view after opening a small hole on the mantle tissue that covers the bursa. The state of this character may be either A. vertical (Figure 2.51) or B. horizontal (Figure 2.52). The orientations are based on the largest dimension of the bursa's body in relation to the duct. The largest FIGURES 2.51 - 2.52. Bursa copulatrix axis, states A and B. FIGURE 2.53. Width of the bursa copulatrix duct in relation to the uterus, Bcb, Bursa Copulatrix body; Bcd, Bursa copulatrix duct; dl diameter of the uterus; d2, diameter of the bursa's duct; RP, reference point; Ut, uterus.
FIGURES 2.54 - 2.56. Bursa copulatrix duct connection, states A to C. Figures 2.57 - 2.60. Digestive tract pigmentation, states A to D. These drawings, with the exception of Figure 2.53, are after Te (personal communication). 74
I (f . 0 2.5]. o 2.52 h R
2.55
2.53
2.56
2.57 2.58
2.59 2.60 75 dimension may be in the same direction of the duct, being vertical; or, horizontal, that is crossed in relation to the duct direction.
Character A-9. Bursa copulatrix duct connection (AC-32) - This character involves the comparison of the diameter of the bursa copulatrix duct and the diameter of the uterus; and includes also the extent of adhesion of the bursa's duct to the oviduct. The bursa's duct runs beside the oviduct and merges with the uterus near the genital opening. The uppermost site where the duct merges with the uterus is a reference point (RP). The diameters of the uterus and the one of the bursa's duct (dl and d2) contain the reference point (Figure 2.53). The extent of adhesion of the bursa's duct to the oviduct was obtained by measuring the length of the segment of the duct that is attached to the oviduct wall. The adhesion of the duct was also registered by simple observation of the distance that separates the bursa's duct and the oviduct. The association of these characteristics allowed the division of the character in the following states (Table 2.111): State A. (as in S. marmorata) - the bursa's duct is apart from the oviduct. It has a connective tissue which attaches it to the uterus. The diameter of the duct is smaller than that of the uterus (Figure 2.54); State B. (as in A. elongata) - the bursa's duct is not contiguous to the oviduct and does not have fibers connecting them. The diameter of the base of the duct is larger than that of the uterus (Figure 2.55); 76
State C. (as in P. gyrina) - the lower portion of the bursa's duct is contiguous to the oviduct. These two parts are attached to each other by connective fibers. The diameter of the duct is smaller than that of the uterus (Figure 2.56); State D - discarded. Not distinguished from State A.
Character A-lO. Digestive tract pigmentation (AC-33) - The pigmentation may occur on the oesophagus, gizzard and /c intestine. The distritubion and intensity of pigmentation in these structures was grouped in four states: State A. lightly pigmented - as in S. marmorata, showing scattered tiny black spots (Figure 2.57); State B. evenly pigmented- as in A. elongata showing uniform pigmentation along the digestive tract (Figure 2.58); State C. partially pigmented -as in P. gyrina, showing patches of pigmentation over the digestive tract (Figure 2.59); State D. unpigmented - as in P. spelunca a cave dweller, without any black pigment (Figure 2.60).,
Character A-il. Shape of gizzard (AC-34) - Musculature structure of the organ indicates the following states: State A. cylindrical - the muscles are evenly distributed, and not girdling the organ (Figure 2.61); State B. laterally lobed - strong muscles girdles the gizzard, yielding a lobed form to the organ (Figure 2.62).
Character A-12. Preputial gland presence (AC-35) - In resting stage the preputium's lumen of the gland is internally in contact with the penis and the gland's FIGURES 2.61 - 2.62. Diagram of the shape of gizzard, states A and B. FIGURES 2.63 - 2.81. Penial complex, states A to S; Pg, preputial gland; N, G and Pr as in Figure 2.82. FIGURE 2.82. Stenophysa marmorata from Fundo Creek basin, penial complex; G, glandular part penial sheath; N, non glandular part penial sheath; P, penis; Pr, preputiuin; PS-1, total length of penial sheath; PS-2, length of penial sheath from the vas deferens to the insertion of the retractor muscle; Rm, retractor muscle; Vd, vas deferens. These drawings with 'the exception of 2.82, are after Te (personal communication and 1978). 77
It..s. 2.61 2.62
2.63 2.64
2.65 2.66
2.67 NPN —'Trn a.70
2.7]. 2.72
2.73
2.75 1 2.76
2.77 2.78 2.82 1mm 2.79 2.80 2.81 body extrudes to the surface of the fixed material, as may be observed under stereomicroscope. The protusion of the penis evertes the preputium and the gland's lumen opens on the preputium outer surface (Duncan, 1958). The dissected penial complex allows a direct observation of the preputial gland's body under stereomicroscope.. The gland may be either absent, state A (Figure 2.63); present, state B (Figure 2.65 to Figure 2.77).
Character A-13. Penial complex type (AC-36 and AC-37)- The dissected penial complex shows the following taxonomically important parts: the penial sheath which may be partially glandular; the preputium, and the preputial gland. The aspects of the glandular nature of the penial sheath, the aspect of the preputium, and the preputium size were examined under stereomicroscope (40 X), without histological preparation (Te, 1975). The sheath's pigments usually enhance the contours of the gland cells. A needle was used to tear the gland cells of poorly pigmented specimens allowing visualization. As the penial sheath was frequently invaginated into the preputium, the measurement of the total penial sheath length required a longitudinal opening of the preputium. The association of these characteristics allowed the division of the character in 19 states (Table 2.111) (Figures 2.63 to 2.81). 79
Character A-14. Length ratio of penial sheath to the preputium (AC-38) - The penial sheath extends from the attachment of the retractor muscle to the vas deferens (PS-2, Figure 2.82). This distance and thelpreputium length were measured over camera lucida tracing. This character has two states (Table 2.111).
Character A-15. Narrowing of the non glandular segment of penial sheath (AC-42) - This character refers to snails having a two - parted penial sheath: a glandular part proximal to the preputium and a non glandular part distal to the preputiuln (Figure 2.82). The non glandular penial sheath segment which contacts the glandular segment may be narrower than the non glandular penial sheath end, state A (Figure 2.83, Table 2.111). The lack of the mentioned narrowing is defined as state B (Figure 2.84, Table 2.111).
Character A-16. Nature of the glandular segment of the penial sheath (AC-47) - As mentioned in character A-13, the penial sheath may have a glandular and a non glandular part. Once dissected the glands are easily distinguished, forming a rough surface. The non glandular part is glossy and smooth. According to the profusion of glands per area, and the length of the glandular segment, three states were recognized (Table 2.111) (Figures 2.85 to 2.87).
Character A-17 . Length ratio of the penial sheath to the preputium (AC-48) - The character is defined as in A-14. Character A-17 has the following states: State A - PS FIGURES 2.83 - 2.84. Narrowing of the non glandular segment of penial sheath, states A and B, respectively;
G, glandular part of penial sheath; N, non glandular;
Pg, preputial gland; Pr, preputium. FIGURES 2.85 - 2.87. Nature of.penial sheath glandular segment, states A to C.
FIGURES 2.88 - 2.91. Swelling of glandular segment of penial sheath, states A to D; a maximum diameter of penial sheath glandular part; b, maximum diameter of penial sheath non glandular part. FIGURES 2.92 - 2.94. Tendency of the preputial. gland to flatten, states A to C.
FIGURES 2.95 - 2.97. Swelling of penial sheath terminus, states A to C; C, maximum diameter of penial sheath terminus; d, maximum diameter of the preputium part that is adjacent to the beginning of the penial sheath. FIGURES 2. 98 and 2.99. Retractor muscle states A and B. These drawings are after Te (personal communication). 80
2.83 2.84
2.85 2.86 2.87
2.89
- c-: iIIIII 2.90 2.91
111 11111111 -j1iiiii IIIlIIIIIIiI 2.92 2.93 2.94
WII d 2.95 jei. 2.97
KAOA"7- 1] 2.99 1.8 to 2.2 times Pr; State B - PS 2.3 to 2.7 times Pr; State C - PS is 2.8 to 3.2 times Pr; and State D - PS is less than 1.8 times Pr (Table 2.111).
Character A-18. Length ratio of the non glandular segment to the glandular segment of the penial sheath (NG/G)
(AC-49) -:: The penial sheath glandular and non glandular lengths were measured over camera lucida tracings. This character' is divided into four states (Table 2.111).
Character A-19. Swelling of glandular segment of penial sheath (AC-50)- Swelling is linked to gland profusion. The degree of gland profusiOn is qualified according to the diameter of the glandular segment in relation to the non glandular segment. The measurements were taken over the maximum diameter of the organ, over camera lucida tracings (Figures 2.88 to 2.91). This character is divided in the following states: State A. very heavy - maximum diameter of glandular segment (G) is 2.5 times greater than maximum diameter of the non glandular segment (NG); State B. heavy - maximum diameter of G is 2.5 to 2.0 times greater than NG; State C. slight - maximum diameter of G is 2 to 1.5 times greater than NG; State D. no swelling - maximum diameter of G is 1.5 to 1 times greater than NG.
Character A-20. Size of preputial gland (AC-52) - Three states were recognized: State A. medium -preputial gland length approximatelly is one third of preputial length; State B. small-preputial gland length is approximately one fourth or less the preputial length; State C. large. - preputial gland length approximately one half the preputial length; having as typical species for each state P. heterostrqpha, P. acuta and P. Integra, respectively.
Character A-21. Tendency of the preputial gland to flatten (AC-53) - This character has three states: State A. oval and compact; State B. flattened; and State C. rounded and loosely packed. The only state I have identified in this study is B, which qualifies P. acuta. The other states do not occur in the species I was dealing with. Consequently, I could not compare the states of this character (Figures 2.92 to 2.94).
Character A-22. Length ratio of the penial sheath to the preputium (AC-55) - The character is defined in A-14. Character A-22 has three states (Table 2.111).
Character A-23. Swelling of penial sheath terminus (AC-56) - This character involves the comparison between the diameter of the penial sheath terminus (C, Figure 2.95) and the part of the preputium that is adjacent to the beginning of the penial sheath (d, figure 2.95). The measurements were taken over the maximum diameter of the respective regions. This character is divided into the following states: State A. not bulbous - the maximum diameter of the penial sheath (PS) terminus is smaller than the maximum diameter of the distal end of the preputium (Pr) (Figure 2.95); State B. bulbous - PS terminus - 83 approximately equal-. to Pr (Figure 2.96); State C. heavily. bulbous - PS terminus greater thanPr (Figure 2.97).
Character A-24. Retractor muscles (AC-61) - The retractor muscle arises from the columellar muscleãnd divides into two, usually pigmented branches, which attach to the penial sheath. When the fibers of these branches are fused, the branches are considered thick and fleshy, state A (Figure 2.98). The retractor muscle may also be divided into a few thin muscle bundles, which attach to the penial sheath, state B Figure 2.99).
Character A-25. Length ratio of penial sheath to the preputium (AC-63) - The penial sheath length was considered twice: the total length, PSi (Fig. 2.82), was measured after a longitudinal sectionning of the preputium to exhibit the penial sheath part invaginated into the preputium; the other measurement was from the attachment of the retractor muscle to the beginning of thew'a's V deferens, PS2 (Figure 2.82). The preputium and penial sheath lengths were measured over camera lucida tracings. The character is divided in four states (Table 2.111).
Character A-26. Number of segments in the penial sheath (AC-65) - This character was divided into three states. One is a two parted penial sheath, which has a glandular part near the preputium and an apparently non glandular part toward the vas deferens, state A (N and G,Figure 2.63); another has an entirely glandular or entirely non glandular penial sheath, state B (Figure 2.68 and Figure 2.76); and a three parted penial sheath that is apparently non glandular near the preputium, and apparently non glandular also towards the vas deferens, but glandular in between those parts, state C (Figure 2.81). To finalize these character descriptions there is a important observation about the range of application of these characters to the Physidae species. Out of the 26 anatomical characters here described 13 of them (A-i to A-13) are applied to the analysis of all species of Physidae. However, characters A-14, A-15 to A-19, A-20 to A-23, A-24 to A-26 are only applied to the analysis of Physa, Physelia (Physella) and Physella (Costatella) section çostatella, Physella (Costatella), and Aplexa and Stenophysa, respectively (Te, 1978).
2.4.2. DISCARDED CHARACTERS
The characters mantle border skirt length, mantle border pigmentation, general and specific kidney shape, specific and general bursa copulatrix shape, constriction of the non glandular segment of the penial sheath and type of preputial head were also studied by Te (1978). For the reasons mentioned in the following paragraphs, they were not considered in this work. The characters mantle border skirt length and mantle border pigmentation were not utilized in this study due to insufficiency of information, on their methodology. The character type of preputial head was discarded due to lack of clear definition of theirstates:jn Te(1978). The characters that refer togeneral'and specific states, imply on granting double weight for a single character. For this reason, at a first stage, they were simply reduced to kidney and bursa copulatrix shape. At a second stage, after the findings of their variability within the Fundo Creek population they were definitely discarded (see item 4.1.3 for discussion).
The characters constriction of the non glandular segment of the penial sheath was discarded due to difficulties on distinguishing "narrowing" and "constriction", as mentioned by Te (1978) for characters AC-42 and AC-45, respectively. In fact, I understand that these two characters overlap, being redundant. Actually, there is a conflict on the use of "narrowed" in character AC-42, state A, and "not constricted" and "constricted" in character AC-45, states A and B, respectively. The idea of "narrowed" could be associated to "constricted". However, Te (1978) on the study of the cubensis group, associated "narrowed" in AC-42 with "not constricted" in AC-45. This does not seem reasonable to me since "not constricted" suggests unchanged diameter of the non glandular part of the penial sheath. CHAPTER 3. INTRA AND INTERSPECIFIC VARIABILITY OF SHELL CHARACTERS IN STENOPHYSA AND PHYSELLA POPULATIONS
3.1. REGRESSION ANALYSIS
The 21 shell characters used in the Physidae regression analysis are listed in Table 2.11. The shell measurements' statistics for the four S. marmorata populations, Recife, Uruçuca, Belo-Horizonte and Fundo Creek, and for the two P. acuta complex populations, Rio de Janeiro and University of Brasilia are given in Tables 3.1 to 3.VI. The matrices of correlation coefficients between 21 shell variables of S. marmorata Fundo Creek population, and P. acuta complex University of Brasilia population are shown in Tables 3.VII and 3.VIII. Out of the 210 combinations of pairs of variables, 76 were found to be significantly correlated in the S. marmorata sample, but only 23 pairs in the P. acuta complex sample. Those 23 pairs of variables are, simultaneously, significantly correlated in both samples. On the other hand, six shell characters of S. marmorata, S-4, S-9, S-b, S-19, S-20 and S-23, and twelve of P. acuta complex, S-1, S-4, 5-9, S-lO, S-il, S-12, S-13, S-14, S-19, S-20, S-21 and S-23, did not show significant correlation with any others character of the respective matrices (Tables 3.VII and 3.VIII). Table 3.IX presents the standard error of estimates for 23 pairs of variables of S. marmorata from Fundo Creek population. The listed pairs of variables were reversed successively, and the choice of the dependent and independent variables was made on basis of either X, Y or Y, X pair which presented the smallest standard error of estimate. - 87
Table 3. 1. Shell dinensions1statistics for the Fundo Creek S. marnorata population2 (N=30).
CHARACTER SYMBOL UNIT X Sd MIN MAX NUMBER
S- I NW Whorl 4.50 0.299 4.00 5.25 S- 2 L1S2 mm 9.57 2.125 6.48 16.59 S- 3 SW mm 5,00 1.143 3.65 9.05 S- 4 M1M2 mm 0.56 0.285 0.10 1.06 S- 5 L]11 mm 5.25 1.271 3.11 9.05 S- 6 L1D1 rum 2.77 0.654 1.68 4.34 S- 7 DD2 mm 3.45 0.777 2,30 5.65 S- 8 C2D1 rum 1.46 0.386 0.81 2.51 S- 9 CuD Degree 58.60 3.962 50.00 67.00 5-10 BEE1 Degree 9.26 5.132 1,00 18.00 S-li CA2 mm 0.07 0.022 0.050 0.14 S-12 BC nun 0.77 0.216 0.53 1.44 S-13 'IU nun 0.12 0.040 0.05 0.22 S-14 CD nun 1.63 0.414 0.92 2.66 S-iS SS3 nun 3.09 0.796 2.26 5.65 S-16 D8S2 mm 6.80 1.546 4.83 12.17 S-17 H1H2 rum 0.99 0.240 0,73 1.66 S-19 D3DD4 Degree 160.00 6.001 152.00 182.00 S-20 D6D5D7 Degree 27.88 8.754 10.00 55.00 * S-21 XI nun 0.60 0.202 0.27 1.06 * S-23 F3G3 rum 0.62 0.296 0.13 1.15
The names of characters are given in Table 2.11.
2lthbreviations as follows: N = number of specimens, R = mean, Sd = standard deviation, Min ndnimum value, Max = maximum value.
* N = 28. 9-01
Table 3.11. Shell dimensions' statistics for the Recife population of S. mariirata 1 (N=1l).
lER SYMBOL UNIT X Sd MIN MAX
5- 1 NW Whorl 4.31 0.162 4.00 4.50 S- 2 L1S2 mm 8.88 0.385 8.16 9.30 S- 3 Sw 4.83 0.287 4.36 5.18 s- 4 M1M2 iran 0.66 0.284 0.32 1.36 S- 5 L1M Iran 4.81 0.265 4.27 5.08 S- 6 L1D1 iran 2.30 0.155 2.06 2.52 S- 7 DD2 ruin 3.12 0.205 2.80 3.32 S- 8 C2D1 nan 1.38 0.115 1.15 1.51 S- 9 C1LD Degree 63.72 1.902 60.00 66.00 S-10 BEE1 Degree 3.72 2.832 0.00 7.00 S-il CT2 nan 0.06 0.014 0.05 0.09 S-12 BC iran 0.65 0.056 0.55 0.75 S-13 TU iran 0.15 0.020 0.14 0.20 S-14 CD ram 1.62 0.148 1.38 1.81 S-15 SS3 iran 3.10 0.240 2.66 3.58 S-16 D8S2 mm 6.60 0.297 6.01 7.04 S-17 H1H2 nan 1.14 0.142 0.96 1.36 S-19 D3DD4 Degree 150 5.600 140 160 S-20 D6D5D7 Degree 30.09 5.412 21.00 36.00 S-21 XY ruin 0.89 0.165 0.65 1.24 S-23 F3G3 mm 0.04 0.018 0.03 0.08
'Abbreviations as in Table 3.1. T
Table 3. iii. Shell dimensions' statistics for the Uruçuca population of S. rnarrrrata1 (N=6).
CHARACIER SYMBOL UNIT X Sd MIN MAX
S- 1 NW Whorl 4.50 0.418 4.00 5.00 S- 2 L1S2 mm 9.36 0.452 8.71 10.05 S- 3 SW mm 4.67 0.257 4.22 4.97 s- 4 N1M2 mm 0.32 0.170 0.17 0.65 S- 5 L1M mm 5.21 0.359 4.86 5.86 S- 6 L1D1 mm 2.94 0.467 2.26 3.63 S- 7 DD2 mm 3.39 0.286 3.12 3.91 S- 8 C2D1 mm 1.64 0.164 1.24 2.12 S- 9 CuD Degree 53.00 1.414 51.00 55.00 S-10 BEE1 Degree 6.50 3.450 1.00 11.00 S-li C1\2 mm 0.11 0.034 0.08 0.15 S-12 BC mm 0.82 0.037 0.75 0.85 S-13 'IU mm 0.17 0.055 0.11 0.27 S-14 CD mm 1.72 0.327 1.42 •2.29 S-15 SS3 mm 2.86 0.302 2.61 3.46 S-16 D8S2 mm 6.36 0.192 6.01 6.54 S-17 H1H2 mm 1.01 0.263 0.78 1.50 S-19 D3DD4 Degree 153 8.300 149 170 S-20 D6D5D7 Degree 25.66 7.394 16.00 35.00 S-21 XY mm 0.59 0.089 0.50 0.72 S-23 F3G3 mm 0.08 0.025 0.05 0.11
'Abbreviations as in Table 3.1. 90
Table 3. jj• Shell dixiensions' statistics for the Belo-Horizonte S. manrata population1 (N5).
LRCTER SIMBOL UNIT Sd MIN MAX
S- 1 NW whorl 5.00 0.379 4.50 5.25 S- 2 L1S2 nm 14.27 2.300 11.12 16.61 S- 3 SW mm 7.41 1.032 5.86 8.30 S- 4 M1M2 mm 0.80 0.393 0.39 1.27 S- 5 L1M mm 7.89 1.879 5.86 9.83 S- 6 L1D1 mm 3.57 0.750 2.77 4.32 S- 7 DD2 mm 4.53 0.707 3.69 5.17 S- 8 C2D1 nun 1.87 0.432 1.38 2.35 S- 9 CUD Degree 2.20 3.114 58.00 65.00 S-10 BEE1 Degree 12.00 3.391 8.00 17.00 S-il C2\2 nm 0.09 0.031 0.06 0.14 S-12 BC rrun 1.13 0.211 089 1.38 S-13 TO mm 0.11 0.051 0.05 0.17 S-14 CD nun 2.18 0.473 1.70 2.69 S-15 SS3 nun 4.77 0.517 3.91 5.29 S-16 D8S2 nun 10.69 1.616 .8.27 12.29 S-17 H1H2 nun 1.56 0.186 1.34 1.76 S-19 D3DD4 Degree 169 7.600 159 179 S-20 D6D5D7 Degree 22.40 9.607 8.00 34.00 S-21 Xy nun 1.13 0.429 0.80 1.86 S-23 F3G3 nm 1.71 0.708 1.00 2.63
'Abbreviations as in Table 3.1. Ji
Table 3.v. Shell dinrnsions' statistics for the University of Brasi ha P. acuta canpiex population (N=30).
CHAPACTER SYMBOL UNIT X Sd MIN MAX NUvIBER S- 1 NW Whorl 5.41 0.296 4.75 6.00 S- 2 L1S2 nun 13.83 1.355 11.88 16.77 S- 3 SW iran 8.88 1.065 7.29 11.52 S- 4 M1M2 mm 0.90 0.422 0.21 1.59 S- 5 L1M iran 733 0.634 6.28 8.64 S- 6 L1D1 mm 3.63 0.368 2.89 4.71 S- 7 DD2 iran 5.78 0.642 4.71 6.94 S- 8 C2D1 mm 1.75 0.232 1.32 2.24 S- 9 CuD Degree 71.10 5.442 63.00 89.00 S-10 BEE1 Degree 8.00 4.720 0.00 15.00 S-il CA2 mm 0.16 0.036 0.10 0.25 8-12 BC Iran 1.16 0.109 1.00 1.44 S-13 TU mm 0.23 0.061 0.10 0.33 S-14 CD mm 2.23 0.296 1.69 2.95 S-15 SS3 mm 5.03 0.614 4.15 6.52 S-16 D8S2 mm 10.19 1.190 8.55 13.47 S-17 H1H2 iran 2.17 0.396 1.38 3.05 S-19 D3DD4 Degree 145.60 9.401 132.00 172.00 S-20 D6D5D7 Degree 37.93 9.428 19.00 55.00 S-21 XY mm 0.64 0.222 0.17 0.10 S-23 F3G3 iran 0.79 0.331 0.37 1.81
'Abbreviations as in Table 3.1. 92
Table 3.vI. Shell dimensions' statistics for the Rio de Janeiro City P. acuta ccruplex population - (N=16).
SYMBOL UNIT X Sd MIN MAX
S- 1 NW %1jorl 4.95 0.164 4.75 5.25 S- 2 L1S2 rum 8.40 0.612 7.22 9.55 S- 3 SW mm 4.82 0.355 4.27 5.38 S- 4 M1M2 rum 0.41 0.223 0.09 0.92 S- 5 L1M mm 4.70 0.311 4.19 5.33 S- 6 L1D1 mm 2.99 0.312 2.53 3.69, S- 7 DD2 nun 3.52 0.326 2.90 4.19 S- 8 C2D1 rum 1.42 0.181 1.16 1.79 S- 9 CUD Degree 53.75 2.978 50.00 61.00 S-b0 -BEEl Degree 13.75 6.758 0.00 26.00 S-li CP2 mm 0.14 0.032 0.08 0.20 S-12 BC rum 0.87 0.119 0.70 1.12 S-13 'RI mm 0.18 0.044 0.09 0.25 S-14 CD nun 1.58 0.219 1.28 2.01 S-15 SS3 nun 2.87 0.243 2.49 3.27 S-16 D8S2 mm 5.42 0.423 4.65 6.08 S-17 H1H2 nun 1.16 0.186 0.82 1.56 S-19 D3DD4 Degree 138 13.500 115 163 S-20 D6DD5 Degree 47.31 9.492 34.00 73.00 S-21 XY mm 0.47 0.137 0.23 0.73 S-23 F3G3 rum 0,32 0.169 0.14 0.87
1Abbreviations as in Table 3.1. 93
Table 3.VII. Matrix of correlatiOn coefficients between 21 shell variables 2 of Stenophysa marrnorata, Fundo Creek population.
S-1 S-2 S-3 S-4 S-S S-6 s-7
S- 1 1.00
S- 2 0.80 1.00
S-. 3 0.77 0.99 i.:oo S- 4 0.05 -0.13 -0.20 1.00 s- 5 0.79 0.98 0.96 -0.15 1.00 S- 6 0.74 0.92 0.87 -0.01 0.94 1.00 S- 7 0.70 0.95 0.93 -0.17 0.95 0.94 1.00 S- 8 0.62 0.84 0.80 -0.02 0.86 0.95 0.92 S- 9 -0.23 0.08 0.12 -0.30 0.05 0.05 0.10 S-1 0.05 0.13 0.15 0.03 0.13 0.13 0.06
S-11 0.69 0.71 0.70 -0.23 0.71 0.67 0.70 S-12 0.68 0.90 0.89 -0.19 0.92 0.89 0.87
S-13 0.51 0.62 0.56 -0.10 0.67 0.78 0.72 S-14 0.64 0.89 0.85 -0.10 0.90 0.96 0.98
S-15 0.75 0.96 0.97 -0.16 0.93 0.85 0.91
S-16 0.79 0.99 0.99 -0.18 0.96 0.84 0.91
S-17 0.75 0.84 0.87 -0.11 0.29 0.67 0.77
S-19 0.42 -0.22 -0.22 -0.10 -0.25 -0.24 -0.18
S-20 -0.06 0.03 0.10 -0.19 -0.04 0.00 0.16
S-21 0.59 0.71 0.71 -0.23 0.71 0.63 0.74 S-23 0.46 0.47 0.49 -0.45 0.49 0.34 0.49 94
Table 3.vII (Continued)
S- 8 S- 9 S-b S-il S-12 S-13 S-14 s- 1
S- 2
S- 3
S- 4
S- 5
S-6
S- 7
S- 8 1.00
S- 9 -0.62 1.00
S-lO -0.00 -0.07 1.00 s-li 0.51 -0.18 0.02 1.00
S-12 0.77 -0.07 0.29 0.73 1.00
S-13 0.79 -0.00 -0.14 0.57 0.59 1.00
S-14 0.98 0.11 0.05 0.59 0.81 0.80 1.00
S-15 0.78 0.06 0.18 0.67 0.88 0.41 0.83
S-16 0.76 0.12 0.13 0.70 0.87 0.53 0.81
S-17 0.63 0.11 -0.01 0.54 0.66 0.32 0.66
S-19 -0.14 0.33 0.17 -0.34 -0.28 -0.30 -0.13
S-20 0.09 0.22 0.09 0.05 0.07 -0.01 0.11
S-21 0.64 0.02 0.12 0.62 0.69 0.53 0.68
S-23 0.42 0.34 -0.40 0.30 0.32 0.38 0.46 Table 3.VII. (Continued)
S-15 S-16 S-17 S-19 S-20 S-21 S-23 s-i S-2
S-3 S-4
S- 5 S-6
S- 7 S-8 S-9
S-10 S-il
S-i2
S-13 S-i4
S-i5 1.00 S-16 0.96 1.00
S-17 0.87 0.86 1.00
S-19 -0.13 -0.20 -0.29 1.00
S-20 0.05 0.04 0.15 -0.09 1.00
S-21 0.73 0.71 0.61 0.04 0.21 1.00
S-23 0.46 0.50 0.60 -0.02 0.19 0.58 1.00
1A11 coefficients above 0.7 are significant at the 0.001 level of probability
2The names of these variables are given in Table 2.11. Table 3-VIII. Natrix of correlation coefficients 1 between 21 shell variables of Physella acuta corrlex, University of Brasilia population.
S- 1 S- 2 S- 3 S- 4 ; S- 5 S- S- 7 s-i 1.00 S- 2 0.36 1.00
S- 3 0.33 0.97 1.00 S- 4 0.16 0.48 0.50 1.00
S- 5 0.12 0.82 0.78 0.48 1.00
S- 6 0.19 0.40 0.22 0.29 0.59 1.00 S- 7 0.46 0.93 0.90 0.51 0.71 0.62 1.00 S- 8 0.26 0.53 0.46 0.39 0.45 0.80 0.65
S- 9 0.34 0.50 0.60 0.36 0.16 -2.35 0.47 S-10 0.17 -0.06 -0.10 0.01 -0.03 0.09 -0.01
S-11 0.06 0.41 0.35 0.15 0.49 0.58 0.43 S-12 0.31 0.37 0.23 0.11 0.35 0.60 0.41
S-13 0.29 0.31 0.27 0.22 0.11 0.45 0.41 S-14 0.32 0.72 0.67 0.20 0.52 0.62 0.78
S-15 0.31 0.87 0.91 0.42 0.67 0.39 0.81
S-16 0.34 0.96 0.97 0.48 0.75 0.27 0.86
S-17 0.18 0.81 0.87 0.43 0.66 0.35 0.73 S-19 0.02 -0.01 -0.04 -0.04 0.16 -0.03 -0.02 S-20 -0.12 -0.00 0.05 0.13 -0.03 -0.07 -0.02
S-21 -0.00 0.09 0.09 0.02 0.04 -0.12 -0.03
S-23 0.36 0.42 0.48 0.08 0.30 0.10 0.34 97
Table 3.VIIL (Continued)
F' S- 8 S- 9 S-b S-li S-12 S-13 S-14 s-i S-2
S-3 S-4 S-5
S-6
S- 7
S- 8 1.00
S- 9 -0.03 1.00 S-10 -0.12 -0.04 1.00 S-li 0.21 -0.24 0.18 1.00. S-12 0.31 -0.20 0.40 0.52 1.00 S-13 0.77 0.08 -0.09 0.11 0.22 1.00 S-14 0.68 0.25 0.06 1.33 0.44 0.52 1.00 S-15 0.39 0.46 -0.15 0.36 0.18 0.18 0.59
S-16 0.37 0.67 -0.11 0.28 0.21 0.22 0.60
S-17 0.42 0.45 -0.15 0.22 0.02 0.20 0.52
S-19 -0.35 -0.05 -0.02 0.15 0.08 -0.62 -0.23
S-20 -0.16 0.07 0.07 0.04 0.02 -0.29 0.17
S-21 0.01 0.12 0.04 0.15 -0.04 0.21 0.08
S-23 0.20 0.25 -0.41 0.06 -0.08 0.08 0.14 98
Table 3.VIII. (continued)
S-15 S-16 S-17 s-19 S-20 s-2I S-23
S- 1 S- 2
S-3 S- 4 s- 5 S- 6 S-7 S- 8 S-9 s-b s-il S-12
S-13 S-14 s-is 1.00 S-16 0.86 1.00
S-17 0.88 0.81 1.00 S-19 -0.06 0.01 -0.24 1.00
S-20 0.03 0.02 0.23 -0.01 1.00
S-21 0.00 0.14 0.04 0.36 -0.17 1.00 S-23 0.59 0.44 0.53 0.04 -0.09 0.14 1.00
Superscripts 1 and 2 - see Table 3.VII. Table 3. 1X. Standard error of estimate for 23 pairs of variables 1 of S. mantirata from the Fundo Creek population. Every listed pair is followed by a pair representing its reversed form.
• VARIABLES STANDARD ERROR VARIABLES STANDARD ERROR OF ESTIMATES x OF ESTIMATES
S- 2 S- 3 0.17 S- 3 5- 2 0.33 S- 2 S- 5 0.22 S- 5 S- 2 0.38 S-2 S-7 0.24 S-7 S-2 0.66 5- 2 S-14 0.19 S-14 S- 2 1.00 S- 2 S-15 0.21 S-15 S- 2 0.57 S- 2 S46 0.26 S-16 5- 2 0.35 5- 2 S-17 0.13 S-17 5-- 2 1.18 S-3 S-5 0.34 S-5 S-3 0.30 S-3 S-7 0.28 S-7 S-3 0.41 S-- 3 S-15 0.20 5-15 S- 3 0.29 S- 3 S-17 0.12 S-17 5- 3 0.57 S-S S-7 0.24 S-7 S-5 0.40 S- 6 S- 8 0.12 S- 8 S- 6 0.21 S-- 7 S-14 0.11 S-14 S-- 7 0.21 5- 7 S-15 0.33 S-15 5- 7 0.32 S- 7 S-17 0.15 S-17 5- 7 0.50 5- 8 S-13 0.03 S-13 S- 8 0.15 S-15 S-17 0.12 S-17 S-15 0.39 S-16 S- 3 0.17 S-- 3 S-16 0.23 S-16 S- 5 0.37 S- 5 S-16 0.45 S-16 5- 7 0.32 S- 7 S-16 0.65 S-16 S-15 0.21 5-15 S-16 0.41 S-16 S-17 0.12 S-17 S-16 0.79
'The names of these variables are given in Table ?..II. 100
The correlation coefficients and their significance between pairs of shell variables for three S. marmorata populations from Belo-Horizonte, Recife and Uruçuca, and for P. acuta complex from Rio de Janeiro, are given in Table 3.X. The parameters of the linear regression equations (Y = a + bX) were calculated only for those pairs of variables whose correlation coefficients' values were above 0.7. Therefore, the number of pairs of variables for which the linear regression equation was calculated, differed for each of the samples: 23 for Fundo Creek, 12 for Belo-Horizonte, 15 for Recife, 10 for Uruçuca, 23 for the University of Brasilia and 19 for Rio de Janeiro. These results are given in Tables 3.XI and 3.XII. The shell width as well as the distance of shell apex to the midpoint of the shell band (L1M) were computer plotted against shell length of two populations, University of Brasilia and Fundo Creek (Figs. 3.1 and 3.2). Similarly, shell aperture width as well as L1M were also computer plotted against shell width of those two populations (Figs. 3.3 and 3.4). Other four linear regressions were also computer plotted but not discussed in this work: the shell aperture width was computer plotted against shell length, against shell width, as well as against shell aperture length; and shell aperture length against shell width. Regressions shown from Figs. 3.1 to 3.4 were chosen to illustrate some shell parameters relationships. • For a given shell length, S. marmorata snails show a smaller shell width when compared to P. acuta complex. Actually, based on the relationships taken from Fig. 3.1, S. marmorata's shell width is approximately 0.50 its shell length; while P. acuta complex's shell width is approximately 0.65 its shell length. Generally speaking, based on Fig. 3.2, in both S. marmorata and P. acuta complex population samples Table 3.X. Correlation coefficients (r) and significance test (p) between shell variables in four different populations of the genera Stenophysa and Physella. Null Hypothesis: p = 0 is rejected at the 0.05 level of probability.
tenopysa Physella Variables - B. Horizonte (N=5) Recife (N=11) Uruçuca (N=6) Rio de Janeiro City (N=15)
X y r p r p r p r p
S- 2 S- 3 0.96 < 0.01 0.86 < 0.001 0.91 0.01 0.87 < 0.001 S- 2 S- 5 0.97 <0.01 0.90 <0.001 0.86 <0.05 0.87 <0.001 S- 2 S.- 7 0.98 < 0.05 0.82 < 0.005 0.93 < 0.01 0.95 < 0.001 s- 2 S-14 0.91 <0.05 - - 0.91 <0.02 0.86 <0.001 * S- 2 S-15 0.73 >0.10 - - 0.84 <0.05 0.78 <0.001 S.- 2 S-16 0.99 < 0 .005 0.91 < 0.001 - - 0.89 < 0.001 * S- 2 S-17 - - - - 0.71 > 0.10 - - * S- 3 S-S 0.88 <0.05 0.90 <0.001 0.72 >0.10 - - S-3 S-7 ------* S- 3 S-15 0.71 > 0.10 - - - - 0.90 < 0.001 S- 3 S-17 - - 0.85 < 0.001 - - 0.75 < 0.005 s- 5 S- 7 0.98 <0.05 0.83 <0.001 0.95 <0.05 0.86 <0.001 S- 6 S- 8 0.96 <0.01 0.89 <0.001 0.74 <0.10 0.95 <0.001 S- 7 S-14 0.91 <0.05 0.79 <0.005 0.99 <0.001 0.90 <0.001 Table 3.X. (Continued)
Stenophysa Physella Variables 1 ______B. Horizonte (N=5) Recife (N=11) Uruçuca (N=6) Rio de Janeiro City (N=15)
X Y r p r p r p r p
S- 7 S-15 - - - - 0.91 < 0.02 0.82 < 0.001 * S- 7 S-17 - - - - 0.77 > 0.10 - - S- 8 S-13 - - - - - - 0.78 < 0.001 S-iS S-17 - - - - 0.86 < 0.05 - - * S-16 S- 3 0.98 <0.005 0.81 <0.005 0.80 <0.10 0.95 <0.001 S-16 S- 5 0.93 < 0.05 0.77 < 0.01 - - - - S-16 S- 7 0.95 < 0.02 - - - - 0.82 < 0.001 * S-16 S-15 0.77 > 0.10 - - - - 0.84 < 0.001 S-16 S-17 - - 0.90 < 0.001 - - - -
1'the names of these variables are given in Table 2.11. - The dashes indicate values not calculated since the respective correlation coefficient values are not significant and between - 0.7 and 0.7.
* Null hypothesis is accepted. Q Table 3. XI. Regression equation parameters (a, bi for four 5, ynarnorata populations.
Variables1 Fundo Creek (N=30) B. Horizonte (N=5) Recife (N=11) Uruçuca (N=6) X Y a b a b a b a b s- 2 s- 3 - 0.092 0.531 1.283 0.429 - 0.886 0.643 - 0.207 .0.520 S- 2 S- 5 - 0.378 0.588 - 0.456 0.795 - 0.691 0.619 - 1.245 0.690 s- 2 s- 7 0.119 0.348 0.238 0.300 - 0.740 0.435 - 2.130 0.590 S- 2 S-14 - 0.015 0.172 - 0.497 0.187 - - - 4.477 0.662 5- 2 S-15 - 0.360 0.361 - - - - - 2.377 0.560 s- 2 S-16 - 0.066 0.717 0.759 0.695 0.345 0.704 - - 5- 2 S-17 0.086 0.094 - - - 1.449 0.292 - S-- 3 5- 5 - 0.098 1.071 - 4.099 1.617 0.759 0.837 - - 5- 3 S- 7 0.278 0.635 - - 0.326 0.578 - 0.434.: .:.0 820 S- 3 S-15 - 0.268 0.673 ------S- 3 S-17 0.076 0.182 - - - 0.902 . 0.424 - S- 5 S- 7 0.397 0.581 1.608 0.370 0.026 0.643 - 0.565 0.759 S- 6 S- 8 - 0.086 0.558 - 0.090 0.551 - 0.134 0.659 - - S- 7 S-14 - 0.132 0.512 - 0.565 0.606 - - - 2.126. 1.135 S- 7 S-15 - 0.132 0.935 - - 3.343 - 0.076 - 0.419 0.968 5- 7 S-17 0.173 0.236 . - - - - - . - 5- 8 S-13 0.000 0.082 . - - 0.047 0.075 - - 0 Table 3. XI. Continuation
Variables1 Fundo Creek (N=30) B. Horizonte (N=5) Recife (1,1=11) Uruçuca (N=6)
X Y a b a b a b a b
S-15 S-17 0.177 0.262 0.361 0.250 - - 1.129 0.748
S-16 S-.3 0.022 0.731 0.717 0.626 - 0.361 0.786 - -
S-16 S- 5 - 0.096 0.786 - 3.713 1.085 0.259 0.689 - -
S-16 S- 7 0.343 0.456 - - - - - -
S-16 S-iS - 0.278 0.496 - - - - - -
S-16 S-17 0.077 0.134 - - - 1.688 0.429 - -
'The names of these variables are given in Table 2.11. - The dashes indicate values not calculated since the respective correlation coefficient values are not significant or between - 0.7 and 0.7.
I-.. 0 105
Table 3. xii. Regression equation parameters for two populations of P. acuta complex.
VARIABLES1 UNIVERSITY OF BRASILIA RIO DE JANEIRO (N=30) (N=15)
X Y a b a b
S- 2 5- 3 - 1.759 0.768 0.566 0.507 S- 2 S- 5 1.549 0.418 0.974 0.443
S- 2 S- 7 - 0.358 0.444 0.722 0.504
S- 2 s-14 0.030 0.159 - 0.998 0.306 S- 2 S-15 0.360 0.361 0.264 0.310
S- 2 S-16 - 1.522 0.846 0.252 0.614
S- 2 S-17 - 1.107 0.237 0.609 0.210
S- 3 5- 5 2.859 0.504 - -
S- 3 S- 7 0.943 0.545 - 0.335 0.798
S- 3 S-15 0.339 0.528 - 0.104 0.616
5- 3 S-17 - 0.701 0.324 . - 0.745 0.394
S- 5 S- 7 0.867 0.670 - 0.533 0.862
S- 6 s- 8 - 0.086 0.558 - 0.235 0.553
S- 7 S-14 0.127 0.364 - 0.533 0.600 S- 7 s-15 0.514 0.780 0.733 0.606
S- 7 5-17 - 0.436 0.451 - -
5- 8 S-13 - 0.123 0.204 0.094 0.191
S-15 S-17 -.0.693 0.570 - 0.236 0.486 S-16 S- 3 0.033 0.874 0.512 0.797
S-16 5- 5 2.929 0.432 - - S-16 5- 7 1.049 0.464 0.074 0.636 S-16 S-15 0.471 0.497 0.250 0.483
S-16 S-17 - 0.574 0.269 - -
Superscript 1 see Table 3.XI.
- The dashes indicate values not calculated since the respective correlation coefficients values, are between - 0.7 and 0.7. 106
12
E E O,2 11
10 • STENOPHYSA MARMOPATA Y1-O.093+O.532x ri0.99
9
• = 0 PHYSELLA ACLJT.A COMPLEX y2•1.759+0.769 r2 0.98
-J =Ui 7- U,
6
5
4
6 7 8 9 10 31 12 13 14 1 1 17
SHELL LENGHT MM
Fig. 3l. Computer plotted regressions of shell width against shell length, for shells of two Physidae populations P. acuta complex from University of Brasilia, and S. marmorata from Fundo Creek Basin. 107
'C
E E 0 9 O.STENOPNYSA MtPM0PATA yi'-O.379+O.588x 0 0 0 2 ri.0.98 0 1 . 0 0 0 8 0.PHYSELI.t. ACUTA COMPLEX 0 y2.1.549+0.418X 0 r2•0.63 0 00 800 0 0 co 0 -J 6 0 • *2
5 0 0 • 00 S
4 5. S
1 0 3 , 6 7 8 9 10 11 12 13 14 15 16 17 SHELL LENGTH mm
Fig. 3.2. Computer plotted regressions of the length of the distance of shell apex to the midpoint of the band (L1M) against shell length, for shells of two Physidae populations, P. acuta complex from University of Brasilia, and S. marmorata from Fundo Creek Basin. 108
11.0.97
OPHVSELLA A CUTA COMPLEX w y20.340+0.529x Cr rl.O.92 D F- Cr w a &~o < 5
-J -J =Iii Cl) 4
3
2 1
1 1 2 3 4 5 6 7 8 9 11 SHELL WIDTH MM
Fig. 3.3. Computer plotted regressions of shell aperture width against shell width for shells of two Physidae populations, P. acuta complex from University of Brasilia and S. marmorata from Fundo Creek Basin. iU:3
10
E E 9
0 0 00 8 0 0
/o o 00 .1 00 2 o
e.STENOPHYSA MARMORATA • -0.098 p1.071 x ri • 0.99
0.PHYSELLA ACUTA COMPLEX y2.2.86+0.504X r2 • 0.79
2
1
0 2 3 4 5 6 7 8 9 10 11 12 13 SHELL WIDTH mm
Fig. 3.4. Computer plotted regressions of the length of the distance of shell apex to the midpoint of the band (L1M) against shell width, for shells of two Physidae populations, P. acuta complex from University of Brasilia and S. marmorata from Fundo Creek Basin. 110 there is approximately a 1:2 ratio between the distance from the shell apex to the midpoint of the band (L1M) and the respective shell length. Nevertheless, Fig. 3.2 also suggests that L1M/SL ratio varies for both populations when snail sizes are considered. Small S. marmorata snails tend to show a L1M/SL ratio close to 0.50, while larger snails tend to show a ratio higher than 0.50. Thus, it indicates that large S. marmorata snail shells are wider on the region situated above the middle of the shell's length. On the other hand, Fig. 3.2 suggests that small P. acuta complex shells tend to have a L1M/SL ratio higher than 0.50, while large shells tend to have this ratio close to 0.50, showing and apparent negative allometric relationship. Therefore, large P. acuta complex shells are wider in the region close to the middle of the shell's length. Based on Fig. 3.3, for a given shell width, S. marmorata snails have a wider shell aperture than P. acuta complex. S. marmorata's shell aperture width is approximately 0.61 its shell width, for individuals having shell widths above 5 mm, while P. acuta complex shell aperture width is approximately 0.55 its respective width. According to Fig. 3.4, S. marmorata population sample shows an approximately 1:1 ratio between shell apex to the midpoint of the band (L1M) and the respective shell width. This suggests the L1M/SW ratio is apparently isometric. On the other hand, L1M/SW ratio for the P. acuta complex population sample is close to 0.90. This ratio shows an apparently negative allometry. This means that P. acuta complex shell's width seems to grow faster than L1M. Though these apparent growth differences can be obtained from Figs. 3.1, 3.2, 3.3, and 3.4, for both species, the regression coefficient's tests have not shown significant differences among the two species for those shell parameters relationships. As only those pairs of characters which had 111 correlation coefficients above 0.7 were suitable for the regression coefficient test, the number of pairs of characters tested in each pair of samples varied: 14 pairs for Recife with Fundo Creek, 16 for Uruçuca with Fundo Creek, 17 for Belo-Horizonte with Fundo Creek (Table 3.XIII), 19 for the University of Brasilia with Rio de Janeiro, and 23 for Fundo Creek with University of Brasilia (Table 3.XIV). The regression coefficients of the chosen pairs of variables were not significantly different comparing the Fundo Creek population with those from Belo-Horizonte, or Recife, or Uruçuca (Table 3.XIII). The regression coefficients of P. acuta complex samples from University of Brasilia with Rio de Janeiro, and the University of Brasilia sample of P. acuta complex with S. rnarmorata sample from Fundo Creek Basin did not have significantly different regression coefficients either (Table 3.XIV). Regression analysis is a technique that allows elimination of the powerful effect of body size and age upon snails measurements (Gould, 1966 and 1981). Regression analysis is also important to understand the growth rate of some characters in relation to others. If the rate is known, any young or adult specimen can serve as a representative of its population or species. The test-of the regression analysis has not shown significant statistical differences among the populations or species here considered. Had I worked with much larger samples and sorted out the specimens for different microhabitats, for each locality, the regression coefficient could be different. It is known that snail populations living in different environments commonly show significant differences on the shape of the shell (Rao and Bhavanarayana, 1976; Janson, 1982). It is possible that parameter a of the regression equation (White and Gould,1965) could be different between the compared populations. A statistical test of difference of means Table 3. XIIL Test of the regression coefficients of the Stenophysa Inarm3rata populations: Furido Creek with Recife (d.f. = 39), and Fundo Creek with Uruçuca (d. f. = 32), Fundo Creek with Belo-Horizonte * (d. f. = 34). Null Hypothesis : - 2 = 0.
Fundo Creek with
Variables Belo Horizonte Recife Uruçuca 2 x y tc p tc p tc p
S- 2 S- 3 0.199 > 0.50 - 0.261 > 0.50 0.022 > 0.90 S- 2 S- 5 - 0.362 > 0.50 - 0.080 > 0.90 - 0.289 > 0.50 S- 2 S- 7 0.163 > 0.50 - 0.288 > 0.50 - 0.821 > 0.40 S- 2 S-14 - 0.096 > 0.90 - - - - S- 2 S-15 0.555 > 0.50 - - - 0.681 > 0.50 S- 2 S-16 0.026 > 0.90 0.02,6 > 0.90 - - * S- 2 S-17 - - - 1.618 > 0.10 - 2.286 < 0.05 S- 3 S- 5 0.503 > 0.50 0.336 > 0.50 0.105 > 0.90 S- 3 S- 7 - - 0.099 > 0.90 - 0.335 > 0.50 S- 3 s-is 0.475 > 0.50 - - - - * 5- 3 S-17 - - - 1.106 > 0.20 2.135 < 0.05 S- 5 S- 7 0.432 >0.50 0.123 >0.90 - 0.356 > 0.50 S- 6 5- 8 0.018 > 0.90 - 0.324 > 0.50 - 0.234 > 0.50 * S- 7 S-14 - 0.234 > 0.50 - - - 2.116 < 0.05 S- 7 S-15 - - 1.336 > 0.10 - 0.039 > 0.90 Nj Table 3.XIII. (Continued)
Fundo Creek with
Variables 1 Belo Horizonte Recife Uruçuca
x Y tc 2 P tc p tC
S- 7 S-17 - - - - - 1.212 > 0.20 S- 8 S-13 - - 1.356 > 0.10 - 0.107 > 0.90 S-iS S-17 0.034 >0.90 - - 1.343 > 0.10 S-16 S- 3 0.149 >0.50 - 0.094 > 0.90 - 0.535 > 0.50 S-16 S- 5 - 0.372 >0 50 0.184 >'0.50 - - S-16 S-7 ------S-16 S-15 0.513 > 0.50 - - - - S-16 S-17 0.920 > 0.20 - - -
1The names of these variables are given in Table 2.11.
= abscisse of Student's distribution - The dashes indicate values not calculated since the respective correlation coefficient values are between - 0.7 and 0.7. * Null hypothesis is rejected at the 0.05 level of probability. Table 3.xIv. Test of the regression coefficients of the P. acuta complex sample from the University of Brasilia with P. acuta complex from Rio de Janeiro (d. f = 41) and P. acuta complex (University of Brasilia) with S. * marrnorata from Fundo Creek Basin (d. f. = 56). Null hypothesis: = 0.
P. acuta complex (Univ. of Brasilia) with
Variables 1 P. acuta complex (Rio de Janeiro) S. marmorata (Fundo Creek)
X Y tc 2 p tc S- 2 s- 3 0.347 > 0.50 - 1.206 > 0.20 S- 2 S- S 0.443 > 0.50 0.802 > 0.40 S- 2 S- 7 0.088 > 0.90 - 0.435 > 0.50
V.UII U.V s- 2 s-15 0.103 > 0 90 0.189 > 0.50 S- 2 S-16 0.218 > 0.50 - 0.489 > 0.50 S- 2 S-17 0.109 > 0.90 - 1.153 > 0.20 S- 3 5- 5 - - 1.264 > 0.20 -: S- 3 S- 7 - 0.415 > 0.50 0.241 > 0.50 S- 3 S-15 - 0.142 > 0.50 1.201 > 0.20 S- 3 S-17 - 0.242 > 0.50 - 0.856 > 0.20 S- 5 S- 7 - 0.178 > 0.50 - 0.210 > 0.50 S- 6 S- 8 0.008. > 0.90 0.136 > 0.50 Table 3.XIV. (Continued)
P. acuta complex (Univ. of Brasilia). with
P. acuta complex (Rio de Janeiro) S. marmorata (Fundo Creek) Variables1 •tc 2 p tc
S- 7 S-14 - 0.500 > 0.50 0.616 > 0.50 S- 7 S-15 0.158 > 0.50 0.228 > 0.50 S- 7 S-17 - - - 0.654 > 0.50 S- 8 S-13 0.057 > 0.90 - 1.017 > 0.20 S-15 S-17 0.169 > 0.50 - 1.168 > 0.20 S-16 s-.3 0.090 > 0.90 - 0.614 > 0.50 S-16 S- 5 - - • 0.882 > 0.20 S-16 S- 7 . - 0.304 > 0.50 - 0.022 > 0.90 S-16 S-15 - 0.063 > 0.90 0.202 > 0.50 S-16 S-17 - - - 0.975 > 0.20
Superscripts 1,2 - see Table 3.XIII.
I-I I-i 01 116 would be the easiest way to verify this difference. However this test could not be applied to samples of this work since the range of the shell sizes varied for each sample (Table 3.1 to Table 3.VI). Actually it is impossible to separate young from adults through shell morphology differences. Although I have concentrated in collecting larger snails, most samples are relatively small and most probably do not adequately represent the true distribution curve of the adult population in those localities. Generally speaking the rate of growth of different shell dimensions is apparently uniform such that the relative proportions of the shell change little during ontogenesis. This supports the use of ratio characters as done by Te (1978) in his revision of world Physidae. Nevertheless, some apparent differences of the relative proportion of shells variables with increase in size could be observed between S. marmorata and P. acuta complex snails (Figs. 3.1 to 3.4). 117
3.2. PRINCIPAL COMPONENT ANALYSIS
The means for shell characters of S. marmorata and P. acuta complex populations are given in Table 3.XV. They were used to estimate the correlation matrices for S. marmorata (Table 3.XVI) and P. acuta complex (Table 3.XVII), respectively. The means of S. marmorata together with P. acuta complex were also used to estimate the composite correlation matrix of the variables (Table 3.XVIII). The correlation of the 19 shell characters with the five principal components are given in Tables 3.XIX, 3.XX and 3.XXI. The five principal components account for 85% of the variation in S. rnarmorata, the first component alone accounting for 53% (Table 3.XIX). The characters which are closely correlated with the first component of S. marmorata are, by decreasing order of importance, from 0.988 to 0.586 shell length (S-2), spire width (S-7), shell width (S-3), aperture width (S-15), aperture length (S-16), apex to shell midpoint (S-5), extremes of 3rd and 4th sutures (S-12), apical aperture width (S-17), spire length (S-6), basal spire length (S-8), band of constant width (S-4), number of shoris (S-l), distance to DZ (S-21), point C to tangent A2 (S-il) (Table 3.XIX). Thus, the first component is associated to variables that represent the overall dimensions of the shell. The second component of S. marmorata is correlated with the following characters ranked from 0.867 to 0.441: spire angle (S-9), aperture insertion angle (S-19), spire length (S-6), and adapical aperture angle (S-20) (Table 3.XIX). The second component is closely correlated to the spire angle (S-9). Actually this character is not correlated to any other component, - 118
Table 3.XV. Means for shell characters of 39 populations used in the principal component analysis (PCA)
Variables Localities1
Pimbers L2 L3 L4 L5 L6 L7 L8
S- 1 4.25 5.00 4.90 4.63 5.00 5.00 5.25 4.75 S- 2 9.13 13.04 14.27 11.90 14.49 12.19 10.95 08.66 S- 3 4.77 6.96 7.41 6.35 7.80 6.81 6.31 4.71
S- 4 0.25 1.09 0.80 0.73 0.76 0.87 0.45 0.50 S- 5 5.00 7.10 7.89 6.59 8.22 6.19 6.03 4.62 S- 6 • 2.34 3.48 3.57 3.23 4.07 3.07 3.07 2.42 S- 7 3.10 4.71 4.53 4.27 5.17 4.47 4.47 3.21
S- 8 1.22 1.74 1.88 1.81 2.03 1.61 1.62 1.17
S- 9 64° 65° 620 63° 580 680 670 61°
S-10 12° 08° 12° 01° 01° 08° 159 020 S-1.1 0.04 0.14 0.09 0.06 0.08 0.09 0.08 0.09
S-12 0.71 1.16 1.13 1.07 1.35 0.91' 0.73 0.66
S-13 0.12 0.14 0.11 0.15 0.17 0.17 0.22 0.14
S-14 1.44 2.03 2.18 2.14 2.20 1.94 1.90 1.42.
S-15 2.96 4.42 4.77 3.60 5.34 3.81 3.97 3.01
S-16 6.79 9.56 10.69 8.70 10.42 9.04 7.82 6.24 S-17 0.90 1.38 1.56 1.28 1.78 1.45 1.28 0.94
S-19 1630 162 0 169 0 159 0 163 ° 1619 1570 155 0
S-20 310 260 22 240 140 270 29° 32° S-21 0.75 0.94 1.13 0.87 0.93 0.69 0.88 0.74 119
Table 3. XV. (Continued)
Variables Localities 1 2,3 Numbers L 9 L 10 L 11 L 12 L 13 L 14 L 15 L 16
S- 1 4.50 4.38 4.75 4.75 4.32 4.84 4.22 5.18 S- 2 9.36 10.06 9.48 8.71 8.89 17.63 6.85 11.60 S- 3 4.67 5.16 5.23 4.36 4.84 9.87 3.61 5.20
S- 4 0.33 0.53 0.43 0.18 0.66 1.37 0.31 0.50 S- 5 5.21 5.49 5.71 4.86 4.81 9.22 3.66 6.87 S- 6 2.94 2.81 2.51 2.98 2.30 3.74 1.88 3.55 s- 7 3.39 3.23 3.37 3.26 3.12 5.82 2.49 4.05 S- 8 1.63 1.42 1.33 1.56 1.38 1.78 0.97 1.90 530 S- 9 550 63° 53° 64° 67° 58° 55°
S-10 07° 11° 06° 05° 04° 080 060 07° S-11 0.12 0.10 0.06 0.09 0.07 0.20 0.09 0.10 S-12 0.83 0.86 0.75 0.82 0.65 1.38 0.54 1.00
S-13 0.18 0.12 0.11 0.16 0.15 0.20 0.10 0.17
S-14 1.73 1.55 1.54 1.74 1.63 2.29 1.04 2.11
S-15 2.87 3.24 3.01 2.57 3.11 6.30 2.09 3.55 S-16 6.36 7.39 6.98 5.73 6.60 14.03 4.98 8.07
S-17 1.02 1.12 1.05 1.00 1.15 1.82 0.72 1.12 0 0 1510: S-19 153° 153 ° 155° 159 ° 150 156 158°
S-20 26° 32° 17° 23° 310 290 260 25°
S-21 0.60 0.86 0.63 0.69 0.90 1.31 0.46 0.94 120
Table 3.XV. (Continued)
Variables Localities1 Nthe L17 L18 L19 L20 L21 L22 L23 L24
S- 1 4.58 4.75 4.50 4.25 4.75 4.50 4.75 5.00 S- 2 08.48 09.19 08.94 08.94 10.39 10.39 12.01 15.08 s- 3 04.48 05.07 04.91 04.72 05.47 05.59 06.16 07.20 S- 4 0.40 0.55 0.55 0.37 0.17 0.50 0.50 0.59 S- 5 4.78 4.82 4.72 4.72 5.81 5.75 5.54 8.90
S- 6 2.38 2.51 2.20 2.29 2.62 2.79 3.14 5.51 S- 7 3.04 3.01 3.35 3.03 3.51 3.63 4.09 5.42
S- 8 1.48 1.05 1.19 1.28 1.34 1.45 1.89 2.63 590 S 9 560 69° 60° 61° 62° 60° 48° S-10 12° 20° 05° 09° 11° 06° 08° 110
S-11 0.06 0.12 0.07 0.09 0.08 0.11 0.06 0.13 S-12 0.70 0.90 0.69 0.73 0.84 0.89 0.88 1.78
S-13 0.28 0.10 0.14 0.14 0.14 0.11 0.16 0.17
S-14 1.59 1.26 1.47 1.47 1.62 1.45 2.20 2.80
S-15 2.62 3.22 3.03 2.93 3.46 3.47 3.90 4.58
S-16 6.09 6.73 6.74 6.70 7.76 9.00 8.87 8.81
S-17 0.95 1.21 1.01 0.96 1.17 1.62 1.26 1.35
S-19 1550 148° 159° 1550 1560 1620 160° 154°
S-20 26° 37° 28° 20° 14° 26° 22° 410
S-21 0.63 0.75 0.55 0.96 0.61 0.73 1.01 0.65 121
Table LXV. (Continued)
Variables Localities1 3 Numbers 2,L 25 L 26 L 27 L 28 L 29 L 30 L 31 L 32
S- 1 5.00 5.00 5.00 4.75 4.64 5.00 4.43 4.50
S- 2 11.26 13.69 10.39 12.14 9.57 9.23 7.39 7.30
S- 3 5.66 7.10 5.81 5.60 5.00 5.61 4.22 4.19
S- 4 0.12 0.65 0.61 1.07 0.56 0.63 0.34 0.46
S- 5 6.60 7.25 2.51 6.85 5.25 5.02 4.26 3.94
S- 6 4.28 3.40 2.51 2.95 2.77 3.01 2.65 2.45
s- 7 3.96 4.49 3.57 3.83 3.45 3.54 3.12 3.15
S- 8 1.95 1.59 1.23 1.57 1.46 1.16 1.30 1.29
S- 9 52° 610 66° 62° 60° 62° 54° 61°
S-10 110 09° 00? 14° 090 06° 07° 12°
S-11 0.12 0.11 0.11 0.09 0.07 0.13 0.10 0.08
S-12 1.07 1.23 0.73 0.94 0.77 0.81 0.79 0.75
S-13 0.19 0.14 0.14 0.12 0.12 0.12 0.16 0.08
S-14 2.14 1.81 1.51 1.89 1.63 1.37 1.48 .1.45
S-15 3.33 4.56 3.69 4.46 3.09 3.48 2.96 2.49
S-16 9.50 10.20 7.93 9.18 6.80 6.10 4.70 4.85
S-17 1.45 1.45 1.23 1.38 0.99 1.29 1.01 0.99
S-19 149° 1559 160° 151° 160° 144° 143° 159°
S-20 22° 24° 19° 25° .. 27° 41° 42° 450
S-21 0.82 1.18 0.67 1.26 1.23 0.48 0.41 1.58 Table 3. XV. (Continued) ,
Variables Localities 2,3 NUmberS .L 37: L 33 L 34 L 35 L 36 j 38.:L 39
S- 1 4.50 4.75 4.50 5.00 5.00 5.41 5.00
S- 2 08.16 09.44 08.21 09.79 11.93 13.83 08.41
S- 3 05.14 05.92 05.50 05.83 07.10 08.88 04.83 S- 4 1.19 0.22 0.27 0.69 0.72 0.91 0.41
5- 5 4.54 4.92 4.45 5.36 6.44 7.34 4.70 S- 6 2.38 2.62 1.93 2.98 3.40 3.64 2.99
5-- 7 3.58 4.08 3.71 4.25 4.88 5.79 3.52 S- 8 1.24 1.45 1.10 1.43 1.61 1.75 1.42 S- 9 66° 72° 82° 65° 66° 71° 54° 0 S-10 14° 00° OtD° 12 05° 08° 14°
s-11 0.04 0.11 0.04 0.14 0.13 0.16 0.14 S-12 0.64 0.78 0.55 0.88 1.02 1.16 0.87 5-13 1.09 0.22 0.09 0.18 0.14 C.24 0.18
S-14 1.38 1.73 1.42 1.68 1.96 2.23 1.58
S-15 2.98 4.08 3.21 3.15 4.12 5.03 2.87
S-16 5.78 6.81 6.28 6.81 8.53 10.19 5.42 S-17 1.28 1.51 1.38 1.18 1.59 2.17 1.16
S-19 1509 1350 170° 1399 1589 1460 138°
S-20 42° 410 53° 47° 410 380 47° 5-21 0.27 1.12 1.47 0.43 0.67 0.64 0.47
1 Localities' namesare given in Table 2.1. 2Variables' namesare given in Table 2.11. zn variables, except S-9, S-b, S-19 and S-20, are given in mm. 123
Table 3.XVI. Matrix of correlation coefficients between the means of 19 shell characters of S. marmorata, from 2 29 localities
S-i S-2 5-3 S-4 S-5 S-6 S-7
S- 1 1.00 S- 2 0.61 1.00
S- 3 0.58 0.96 1.00 S- 4 0.29 0.70 0.73 1.00 S- 5 0.50 0.89 0.81 0.55 1.00
S- 6 0.64 0.79 0.67 0.28 0.82 1.00 S- 7 0.69 0.96 0.95 0.64 0.88 0.83 1.00 S- 8 0.57 0.75 0.62 0.27 0.79 0.94 0.80
S- 9 0.00 0.07 0.24 0.46 -0.12 -0.43 0.09
S-1 0.07 0.04 -0.01 -0.06 0.17 0.08 -0.03
s-11 0.33 0.55 0.55 0.48 0.44 0.47 0.51 S-12 0.53 0.88 0.79 0.47 0.88 0.92 0.87
S-13 0.34 0.19 0.21 -0.01 0.20 0.29 0.31 S-15 0.57 0.96 0.96. 0.77 0.82 0.66 0.92
S-16 0.52 0.94 0.95 0.70 0.81 0.64 0.88
S-17 0.55 0.86 0.88 0.63 0.75 0.64 0.83
S-19 0.03 0.21 0.29 0.18 0.09 0.00 0.24 S-20 -0.12 -0.02 -0.05 0.14 0.06 0.12 0.01
S-21 0.25 0.59 0.55 0.61 0.53 0.30 0.48 124
Table 3.xvl. (Continued)
S- 8 S- 9 S-b S-li S-12 S-13 S-15
S-i
S-2
S-3
S-4
S-5
S-6
S- 7
5- 8 1.00
5- 9 -0.38 1.00
S-10 0.00 -0.18 1.00
S-il 0.29 -0.15 0.10 1.00
S-12 0.84 -0.31 0.07 0.56 1.00
S-13 0.40 -0.06 0.03 0.10 0.16 1.00
S-15 0.61 0.21 0.04 0.54 0.78 0.15 1.00
S-16 0.59 0.22 0.02 0.57 0.74 0.14 0.94
S-17 0.58 0.15 0.01 0.51 0.72 0.13 0.88
S-19 0.08 0.42 -0.29 -0.22 0.05 -0.17 0.22
S-20 0.05 -0.18 0.37 0.25 0.14 0.00 -0.05
S-21 0.31 0.17 0.18 0.25 0.37 -0.03 0.64 125
Table 3.XVI. (Continued)
S-16 S-17 S-19 S-20 S-21
s-i 5-2
S-3 S-4
S-5 S-6
S- 7 S- 8 S- 9 s-b
s-il S-12
S-13 S-15
S-i6 1.00
S-17 0.92 1.00 S-19 0.22 0.21 1.00 S-20 -0.14 -0.14 -0.31 1.00
S-21 0.64 0.51 0.05 -0.05 1.00
1 The names of the characters are given in Table 2.11. 2 The localities' names are given in Table 2.1. 126
Table 3.XVfl..Matrix of correlation coefficients between the means of 19 shell characters 1 of P. acuta complex, from ten localities 2 .
S-i S-2 S-3 S-4 S-5 S-6 S-7
S- 1 1.00
S- 2 0.86 1.00
S- 3 0.80 0.98 1.00
S- 4 0.31 0.41 0.41 1.00
S- 5 0.86 0.99 0.97 0.45 1.00
S- 6 0.89 0.82 0.70 0.37 0.84 1.00
S- 7 0.79 0.98 0.98 0.40 0.97 0.72 1.00 S- 8 0.73 0.83 0.75 0.27 0.83 0.83 0.84
S- 9 0.02 0.32 0.47 -0.01 0.26 -0.27 0.42 S-10 0.08 -0.17 -0.25 0.53 -0.12 0.18 -0.18
S-11 0.86 0.64 0.52 0.03 0.66 0.90 0.55
S-12 0.86 0.84 0.74 0.29 0.86 0.97 0.78 S-13 0.67 0.61 0.57 -0.08 0.61 0.64 0.62 s-15 0.71 0.92 0.94 0.23 0.90 0.66 0.91
S-16 0.78 0.98 0.99 0.40 0.97 0.70 0.99 S-17 0.70 0.91 0.96 0.35 0.89 0.58 0.93
S-19 -0.37 -0.09 -0.02 0.01 -0.10 -0.41 -0.04
S-20 -0.41 -0.52 -0.46 -0.44 -0.52 -0.65 -0.43
S-21 -0.15 0.04 0.16 -0.59 -0.03 -0.41 0.12 127
Table 3.XVII. (Continued)
S- 8 S- 9 S-b S-li S-12 S-13 S-15
S- 1
S- 2
5- 3
S-4
S-5
S-6
S- 7
S- 8 1.00
S- 9 -0.03 1.00
S-10 0.08 -0.64 1.00
S-il 0.72 -0.34 0.11 1.00
S-12 0.91 -0.21 0.14 0.88 1.00
S-13 0.76 -0.03 -0.13 0.76 0.70 1.00
S-15 0.73 0.44 -0.45 0.52 0.70 0.67 1.00
S-16 0.78 0.48 -0.25 0.52 0.74 0.56 0.93
S-17 0.69 0.54 -0.34 0.39 0.63 0.56 0.95
S-19 -0.32 0.49 -0.25 -0.60 -0.35 -0.70 -0.14
S-20 -0.56 0.26 -0.05 -0.46 -0.62 -0.44 -0.58
S-21 -0.15 0.80 -0.83 -0.30 -0.31 0.01 0.25 128
Table 3.XVII .(Continued)
S-16 S-17 S-19 S-20 S-21
s-i S-2 S-3 5-4 S-5 S-6 S- 7 S- 8 S- 9 s-b s-il S-12 S-13 S-is S-16 1.00 S-17 0.95 1.00 S-19 0.02 0.02 1.00 S-20 -0.43 -0.46 0.44 1.00 S-21 0.18 0.27 0.45 0.45 1.00
1The names of the characters are given in Table 2.11.
2 The localities' names are given in Table 2.1. Table 3.XVIII. Matrix of Correlation coefficients between the means of 19 shell variables of S. marmorata and P. acuta complex, from 39 localities 2 .
S-1 S-2 S-3 S-4 S-S S-6 S-7
S- 1 1.00
S- 2 0.59 1.00
S- 3 0.63 0.92 1.00
S- 4 0.30 0.59 0.64 1.00
S- 5 0.51 0.91 0.81 0.50 1.00
S- 6 0.64 0.79 0.66 0.28 0.82 1.00
5- 7 0.72 0.90 0.95 0.58 0.84 0.78 1.00
S- 8 0.52 0.77 0.61 0.25 0.81 0.93 0.74
S- 9 0.05 0.03 0.30 0.27 -0.10 -0.38 0.21
S-10 0.07 -0.00 -0.08 0.11 0.12 0.10 -0.08
S-li 0.52 0.48 0.53 0.33 0.41 0.51 0.53
S-12 0.55 0.87 0.75 0.42 0.88 0.93 0.81
S-13 0.44 0.29 0.33 -0.03 0.28 0.35 0.41
S-15 0.58 0.94 0.95 0.64 0.83 0.66 0.90
S-16 0.50 0.95 0.90 0.58 0.84 0.65 0.82
S-17 0.61 0.75 0.88 0.53 0.67 0.56 0.85
S-19 -0.18 0.21 0.13 0.07 0.14 -0.04 0.06
S-20 -0.01 -0.29 -0.08 0.04 -0.22 -0.10 0.02
S-21 0.06 0.45 0.39 0.16 0.40 0.14 0.31 - 130
Table 3.XVIIL(Continued)
S-8 s- 9 s-b s-il s-12 s-13 s-15
s- 1 S- 2 S- 3 S- 4 S- 5 S-6 S-7 S- 8 1.00 S- 9 0.33 1.00
S.- lo 0.03 -0.34 1.00 S-il 0.29 -0.15 0.10 1.00 S-12 0.85 -0.30 0.09 0.55 1.00 S-13 0.43 -0.05 -0.02 0.35 0.26 1.00 S-15 0.62 0.22 -0.07 0.50 0.77 0.29 1.00 S-16 0.64 0.14 -0.03 0.45 0.74 0.24 0.91 S-17 0.49 0.35 -0.11 0.49 0.62 0.29 0.84 S-19 0.09 0.21 -0.21 -0.42 0.63 -0.37 0.11 S-20 0.22 0.25 0.13 0.18 -0.12 -0.06 -0.16 S-21 0.25 0.31 -0.18 -0.03 0.22 -0.01 0.50 131
Table 3.XVIII. (Continued)
S-16 S-17 S-19 5-20 S-21
s-i S- 2 S- 3 S- 4 S-5 S-6 S- 7 S-8 S- 9 S-10 s-li S-12 S-13 S-15 S-16 1.00 S-17 0.77 1.00 S-19 0.27 -0.00 1.00 S-20 -0.38 0.04 -0.42 1.00 S-21 0.53 0.32 0.37 -0.20 1.00
The names of the variables are given in Table 2.11.
2The localities ' names are given in Table 2.1. 132
Table 3.XIX. Principal component analysis: correlation of the 19 shell characters' with the •five components for S. marmorata .
Principal Components Character 1 2 3 4 .5
S- 1 0.654 -0.125 -0.246 -0.342 0.092 S- 2 0.989 0.057 0.003 0.644 0.012 S- 3 0.953 0.213 0.022 -0.048 -0.055 S- 4 0.687 0.427 0.396 -0.049 -0.190 S- 5 0.906 -0.155 0.008 0.120 0.148 S- 6 0.829 -0.467 -0.222 0.139 -0.005 S- 7 0.973 0.016 -0.118 -0.055 -0.057 S- 8 0.780 -0.410 -0.344 0.079 0.072 S- 9 0.028 0.868 0.150 -0.334 -0.029 S-1 0.053 -0.385 0.556 -0.138 0.599 S-11 0.586 -0.191 0.389 -0.037 -0.476 S-12 0.893 -0.307 -0.057 0.237 -0.074 S-13 0.243 -0.306 -0.295 -0.780 -0.048 S-15 0.952 0.205 0.096 0.017 0.020 S-16 0.933 0.231 0.068 -0.010 0.031 S-17 0.885 0.179 -0.007 0.003 0.006 S-19 0.168 0.610 -0.405 0.297 0.090 S-20 0.006 -0.441 0.619 0.037 -0.257 S-21 0.589 0.253 0.349 0.030 0.406
1The names of the. characters are given in Table 2.11. Table 3.XX.. Principal component analysis: correlation of the 19 shell characters' with the five cononentsfor P. acuta complex.
Principal Components Character 1 2 3 4 5
S- 1 -0.893 -0.159 0.067 0.269 0.105 S- 2 -0.978 0.152 -0.114 0.056 -0.035 S- 3 -0.940 0.299 -0.146 0.006 0.047 S- 4 -0.378 -0.268 -0.838 -0.172 0.208 S- 5 -0.976 0.095 -0.139 0.077 -0.044 S- 6 -0.882 -0.400 0.025 0.118 -0.187 s- 7 -0.949 0.242 -0.130 0.075 0.073 S- 8 -0.888 -0.163 0.074 0.070 -0.034 S- 9 -0.171 0.931 -0.136 -0.036 0.251 s-1 0.097 -0.785 -0.446 0.254 0.255 s-11 -0.749 -0.438 0.388 0.238 -0.076 S-12 -0.906 -0.318 0.068 0.152 -0.199 S-13 -0.736 -0.152 0.553 -0.103 0.286 S-15 -0.911 0.325 0.077 -0.230 -0.036 S-16 -0.936 0.310 -0.151 0.041 0.020 S-17 -0.876 0.397 -0.115 -0.127 0.072 S-19 0.276 0.635 -0.509 0.299 -0.395 S-20 0.607 0.365 0.079 0.638 0.267 S-21 0.062 0.921 0.355 0.086 0.014
1The names of the characters are given in Table 2.11. 134
Table 3.XXI. Principal component analysis; correlation of the 19 shell characters' with the five components for S. marmorata together with P. acutacomp1ex.
Principal Components Character 1 2 3 4 5 S- 1 0.690 -0.273 -0.219 0.185 0.088 S- 2 0.974 0.112 0.132 -0.053 0.056 s- 3 0.945 0.175 -0.201 -0.010 0.033 S- 4 0.584 0.212 -0.296 -0.495 0.162 s- 5 0.913 -0.018 0.215 -0.114 -0.038 S- 6 0.839 -0.372 0.265 0.016 -0.252 S- 7 0.955 0.022 -0.178 0.055 -0.114 S- 8 0.796 -0.236 0.392 0.139 -0.214 S- 9 0.044 0.685 -0.656 0.114 0.029 S-1 -0.001 -0.446 0.094 -0.649 0.380 S-11 0.568 -0.449 -0.338 -0.049 0.083 S-12 0.889 -0.239 0.227 -0.109 -0.209 S-13 0.383 -0.399 -0.140 0,622 0.338 S-15 0.940 0.195 -0.111 -0.029 0.066 S-16 0.914 0.262 0.103 -0.042 0.194 S-17 0.831 0.145 -0.347 0.039 -0.025 S-19 0.080 0.720 0.475 -0.101 -0.134 S-20 -0.159 -'0.271 -0.695 -0.226 -0.507 S-'21 0.394 0.587 0.126 0.091 0.012
1The names of the characters are given in Table 2.11. 135 according to the criterion established to set the correlation limits. On the other hand, characters S-19 and S-20 have a weaker correlation with the second component, and are also correlated with the third component. Character 5-6 is also influenced by shell length (first component) and has a weaker correlation with the second component than character S-9 (Table 3.XIX). Thus, for S. marmorata snails, the second component represents mainly the spire angle of the shell (S-9). The third component of S. marmorata correlates with characters ranked from 0.618 to 0.344: adapical aperture angle (S-20), sutural angle (S-10), aperture insertion angle (S-19), band of constant width (S-4), point C to tangent A2 (S-il), distance to DZ (S-21), basal spire length (S-8) (Table 3.XIX). Characters S-il and S-21 are influenced by shell size because they are mainly correlated to the first component. Characters S-4 and S-8 are influenced by shell size and spire angle due to their correlation with the first and second components (Table 3.XXII). Characters S-19 and S-20 are also weakly correlated with the second component (Table 3.XIX). Character S-10 correlates only with the third component. Thus, the third component represents mainly an association of three angles of the shell for S. marmorata snails: S-20, S-10 and S-19. The fourth component of S. marmorata is only correlated (-0.779) with the character maximum distance CD to 4th whorl (S-13) (Table 3.XIX). The fifth component of S. marmorata is correlated with characters ranked from 0.598 to 0.405: sutural angle (S-b), point C to tangent A2 (S-li), and distance to DZ (S-21) (Table 3.XIX). The five principal components account for 97% of the variation in P. acuta complex. The first component alone accounting for 58% (Table 3.XX) . The characters which are more correlated with the first component of P. acuta complex are, by decreasing order of importance, from -0.978 to 0.607: shell length (S-2), Table 3.XXII. Correlation of the 19 shell characters 1 with the principal components for Stenophysa marmorata, Physella acuta complex, and S. marmorata together with P. acuta complex.
Principal Components 1 2 3 4 5 Character Species
S P S+P S P S+P S P S+P S P s-I-p S P S-I-P
S-i + + + - - - - - - - - - - - - S-2 + + + - - - - - - - - - - -
S-3 + + + - - - - - -. - - - - - - S-4 + - + + - - + + - - - + S-5 + + + - - - - - - - - - - - - S-6 + + + + - + - - - - - - - - - S-7 + + + - - - - - - - - - - - - S-8 + + + - - - + S-9 - - - + + + - - + - - - - + - S-b - - - - + + + + - - - + + + +
• S-1i + + + - - + + - - - - - + - - S-12 + + + - - - - - - - - - - + S-13 - + - - - + - + - + - + - + + Table lxxii. (Continued)
Principal Components 1 2 .3 4 5 Character Species S P S+P S P S+P S P S+P S P S+P S P S+P
S- 15 - + + - - - - - - - - - - - - S-16 + + + - - - - - - - - - - S-17 + + + S-19 - - + + + + S-20 - + - + - - + S-21 + - - - + +
S = Stenophysa marmorata P = Physella acuta complex S+P = S. marmoràta and P. acuta complex + = Correlated - = Not correlated
1The name of the characters are given in Table 2.11. 138
apex to shell midpoint (S-5), spire width (S-7), shell width (S-3), aperture length (S-16), aperture width (s-15), extremes of third and fourth sutures (S-12), number of whorls (S-i), basal spire length (S-8), spire length (S-6), apical aperture width (S-17), point C to tangent A2 (S-il), maximum distance CD to fourth whorl (S-13), and adapical aperture angle (S-20) (Table 3.XX )i• This component groups most variables that are linked to the size of the shell, having the same meaning as for S. marmorata. The second component of P. acuta complex is correlated with characters ranked from 0.930 to 0.634: spire angle (S-9), distance to DZ (S-21), sutural angle (S-lO), and aperture insertion angle (S-19) (Table 3.XX ). The second component is closely correlated to the spire angle (S-9) and to the distance to DZ (S-21) (Table 3.XX ). Besides, Table 3.)Qalshows that these two characters are only correlated to the second component for P. acuta complex snails. However, characters S-lO and S-19 are weakly correlated both to the second and third components. For these reasons, the second component represents mainly the spire angle of the shell and the distance to DZ for P. acuta complex snails. P. acuta complex differ from S. marmorata snails since the second component of the latter is represented mainly by the spire angle. Therefore, these species can not be compared directly on the basis of the second component. calculated separately for each taxon. The third component of P. acuta complex is correlated with characters ranked from -0.837 to 0.445: band of constant width (S-4), maximum distance CD to the fourth whorl (S-13), aperture insertion angle (S-19), and sutural angle (S-b) (Table 3.XX) . Component three is closely correlated to character S-4. This character has no correlation with the first nor with the second component (Table 3.XXII).On the other hand, characters S-19 and S-10 are weakly correlated to the third and second components. Thus, the third component represents 139 mainly the band of constant width for the shell of p acüta complex snails. This differs from S. marmorata snails where the third component is mainly related to angles. Therefore, these species cannot be compared directly on the basis of the third component calculated separately for each taxon. The fourth component of P. acuta complex is correlated only (0.638) with the character adapical aperture angle (S-20). The fifth component of P. acuta complex is more correlated with characters ranked from -0.395 to 0.208: aperture insertion angle (S-19), maximum distance CD to fourth whorl (S-13), adapical aperture angle (S-20), sutural angle (S-10), spire angle (S-9), band of constant width (S-4), and extremes of third and fourth sutures (S-12) (Table 3.XX .). Therefore, this component is not strongly correlated with any of the characters measured, and has, thus, no special meaning. The five principal components account fOr 84% of the total variation in S. marmorata together with P. acuta complex, the first component alone accounting for 51% (Table 3.XXI). The characters which are more correlated with the first component are, by decreasing order of importance from 0.974 to 0.690: shell length (S-2), spire width (S-7), shell width (S-3), aperture width (s-15), aperture length (S-16), apex to shell midpoint (S-5), extremes of third and fourth sutures (S-12), spire length (S-6)., apical aperture width (S-17), basal spire length (S-8), number of whorls (S-l), band of constant width (S-4) and point C to tangent A2 (S-li) (Table 3.XXI .). Thus, again, the first principal component represents size. The second component of S. marmorata plus P. acuta complex is correlated with characters ranked from 0.720 to 0.372: aperture insertion angle (S-19), spire angle (5-9), distance to DZ (S-21), point C to tangent A2 (S-li), sutural angle (S-b), maximum distance CD to fourth whorl (S-13), and spire length (S-6) (Table 3.XXI ). The second component is closely - 140 correlated to the, aperture insertion angle and spire angle. These two characters are also correlated to the third component. Although characters S-21, S-lO and S-13 have no correlation with the first 'and third components, they are relatively weakly correlated tOcthe second component.. Character S-li has no correlation 'with the first component, and S-6 is weakly correlated to the first and second components. If S. rnarmorata and P. acuta complex snails are analysed together, the second component represents mainly the aperture insertion angle and the spire angle. The third component of S. marmorata plus P. acuta complex is correlated with characters ranked from -0.694 to -0.295: adapical aperture angle (S-20), spire angle (S-9), aperture insertion angle (S-19), basal spire length (S-8), and apical aperture width (S-17) (Table 3.XXI) . This component is closely correlated to the adapical aperture angle and spire angle Character 5-20 has no correlation with the first and second components, while characters S-19 and S-9 are also correlated to 'the second component, and characters S-B and S-17 to the first component. Therefore, the third component represents mainly the adapical aperture angle and the spire angle, if S. marmorata and P. acuta complex snails are analysed together. The fourth component of S. marmorata plus P. acuta complex is correlated with characters ranked from -0.649 to -0.494: sutural angle (S-b), maximum distance CD to fourth whorl (S-13) and band of constant width (S-4) (Table 3.xxI). The fifth component of S. marmorata and P. acuta complex is correlated with characters ranked from -0.506 to 0.337: adapical aperture angle (S-20), sutural angle (S-b) and maximum distance CD to fourth whorl (S-13) (Table 3.xxI). The first component represents mainly a measurement of the overall dimension of the shell when S. marmorata, P. acuta complex, or S. marmorata together with P. acuta complex samples are considered. Indeed, 141 eight shell measurements may be considered redundant to the first component: S-i, S-2, S-3, S-S, S-7, S-12, S-15 and S-16 (Table 3.XXII). Since the, first component is most closely related to the length of :the shell (S-2) this measurement can be used to describe accurately the first component which represents the size of the shell. Thus, the other seven characters -above mentioned could be discarded due to their redundancy. However, as the shell apex is commonly subjected to erosion, I would suggest shell width as an alternative character to shell length, to represent the first component. As the first component represents the size of the shell, this axis was not considered in the principal component analysis. This is so because the samples utilized were not representative of the true distribution frequencies of the size classes of the populations surveyed. As mentioned elsewhere, it was not possible to separate young from adults through shell morphology alone. As the second and third components are related mainly to shell angle characters, but not correlated to any of the linear shell characters (Table 3.XXII), they were considered for the application of the Principal Component Analysis method. On the other hand, the fourth and fifth components were also ignored for the following reasons. They account for only a small percentage of the total variation (Tables 3.XIX, 3.XX, 3.XXI). When the data of these components were plotted against another component, the distribution did not reveal any possible differentiation among localities or species. They are not strongly correlated to any particular shell variable. The plot of the second versus the third component for the different -populations of S. marmorata snails is shown in Fig.3.5. The majority of the localities is clustered within the borders of quadrants I and IV. Fundo Creek population (L 29) is located near the center of this cluster. There are seven localities: 3 0 I
IC
13
/ 26 29 / 8 20 1 2PC 15 (7 22 19\ L 63) . / minor cluster 21 23
© main cluster
-2 lIt
- -' U 1 2 H. Fig. 3.5. Plot of the second and third principal components for Stenophysa marmorata snails. Numbers refer to localities which are listed in Table 2.1. 143
5, 12, 18, 24, 25, 27 and 28 from Rio Grande doSul, Cea- râ, Rio de Janeiro, So Paulo, Mato Grosso do Sul, Minas Gerais and Peru, respectively, which are located apart from this cluster, in quadrants I, II, III and IV (Fig. 3.5). Though they are apart from the main cluster, their position has a limited significance since they are represented by single specimens. These seven localities depart from the main cluster with respect to the second and/or third component. Since S-9 is the character most closely correlated to the second component, I compared the S-9 shell character value of each locality with the range of.variation.of the S-9 character from the Fundo Creek sample of 30 specimens (Table 3.1 ). Six of these localities, 5, 12, 18, 25, 27 and 28, have spire angle (S-9) values within the range of variation of the Fundo Creek population (Table 3.1.). With respect to locality L 24, the spire angle measurement is close to the lowest limit of the range of the variation found in the Fundo Creek population. For this reason all the seven localities. may be considered as belonging .to..S. marmorata species. As. S-b, S-19, S-20 are the characters most closely correlated to the third component, their values for these seven localities were compared with the range of variation of the respective characters of the Fundo Creek population (Table 3.1.). Though L 25 and L 28 are slightly below the range of S-19, and L 18 is slightly below the range of S-19 and slightly above the range of S-b, all the seven localities may be considered as belonging to S. marrndrata. There are three localities L 9, L 16 and L 17 which have sample size from four to six specimens, and are positioned in quadrant III, forming a minor cluster. These localities, are separated from the main cluster in quadrants I and IV only by the second component (Fig. 3.5). There are two localities, .L 14 (n = 3), which is distant from the main cluster in quadrants I and IV, with respect to the third component, 144
and L 10 (n = 4), which is apart from that cluster, with respect to the second. and third components. These five localities, specially those forming the minor cluster, might represent different taxa separated from that represented by the main cluster. A test for difference of means of characters S-9, S-19 and .S-20 was performed for four localities: one locality With small spire angle (L 9), two localities with large spie angle (L 3 and L 13), and another locality (L 29) with intermediate values of spire angle (Table 3.XxIII). No statistical difference was found between these localities with regard to the adapical aperture angle. However, the means of spire angle (S-9) and aperture insertion angle (S-19) was found to differ statistically between some of these localities. However, there are localities with intermediate mean values for these two angles (Tables 3.XV. and 3.XXIII), thus suggesting all the Stenophysa populations studied belong to the same taxon which is extremely variable but with normally distributed spire angle values (Fig. 3.8). In fact, there is no apparent geographic pattern that match the association of. the localities in Fig. 3.5. For example, L 6 1 L 3, and L 19 are closely associated though they correspond to localities geographically far apart. In addition L 29 and L 16, both from Federal District, are situated far apart in quadrants I and III, respectively (Fig. 3.5). It is possible, therefore, that the great majority of the variation, as observed in the different samples of S. marrnorata, in relation to the second and third components, may be the phenotypic expressions of adaptations to different local environmental conditions. The plot of the second versus the third component for different populations of P. acuta complex snails is shown in Fig. 3.6. The majority of the localities is clustered mainly within quadrants II and III (cluster A). There are three localities, L 33, L 34 and L 35 from Peru, So Paulo, and Trinidad, respectively, 145
Table 3.xxIII.Lbstfor difference of means (t) between S. marmorata samples.
Character Spire Aperture Adapical LC114-' angle Insertion Aperture (S-9) Angle (S-19) Angle (S-20)
Belo }krizonte (L-3) 62.2 ac 168.8 a 22.4 a Riacho Fundo (L-29) 58.6 a 160.3 b 27.8 a Uruçuca (L-9) 53.0 b 153.2 c 25.6 a Recife (tr-13) 63.7 c 149.9 c 30.1 a
* Means with same letter do not differ statistically at 0.05 level. I 0 cluster
(73767 2nd Pc
37 ter 4
nI
2 3 4
Fig. 3.6. Plot of the second and third principal components for Physella acuta complex snails. Numbers refer to localities which are listed in Table 2.1. .L.•Z I which are located apart from this cluster in quadrants III, I and IV, respectively (Fig. 3.6). Though they are apart from the remaining localities, their position has a limited significance since they are represented by single specimens. Localities L 33 and L 34 depart from the majority of the localities with respect to the third component. Since S-4 is -the character most closely correlated to the third component for P. acuta complex snails, the S-4 values of these localities were compared with the range of variation of the S-4 character from the University of Brasilia sample (L 38) of 30 specimens (Table 3.V ). Thus, localities L 33 and L 34, have band of constant width (S-4) values within the range of variation of the University of Brasilia population (Table 3.V.). Locality L 35 departs from the remaining localities with respect to the second component not only because it has a large spire angle (82 ° ) but specially for having the largest value of distance to DZ (1.47 mm). As S-9 and S-21 are the characters most closely correlated to the second component of P. acuta complex snails, their values for locality L 35 were compared with the range of variation of these characters in the sample from the University of Brasilia (Table 3.V ). They were found to be within the range of variation of the respective characters of the University of Brasilia population. Consequently it is possible that these three localities may belong to the same taxon of the University of Brasilia sample. There is no apparent geographic pattern that match the association of localities in Fig 3.6. For example, L 37 and L 38 are closely associated though they correspond to localities geographically far apart (Puerto Rico and Brasilia). In addition, L 36 and L 37, both from Puerto Rico, are situated far apart in quadrants II and IV, respectively (Fig. 3.6). As in the case of S. marrnorata these variations may be associated to local environmental conditions. 148
However, Fig. 3.6 suggests that the different samples of P. acuta complex may be grouped in two clusters, A and B, which are separated by both the second and third principal components. As shown earlier, the second principal component is represented mainly by the s'Ire angle (S-9) and distance to DZ (S-21), whereas the third principal component is represented mainly by the band of constant width (S-4). The means of characters S-9, S-21 and S-4 were found to differ statistically between the two clusters (Table 3.XXIV). The spire angle shows a bimodal pattern emphasizing the non normality of the distribution (Fig.' 3.9). The snails belonging to cluster A have large spire angles and to cluster B have small spire angles (Table 3.XXIV).The snails belonging to cluster B come all from Rio de Janeiro, whereas the ones belonging to cluster A have a wide range of' distribution. Thus, these findings suggest that these two clusters belong to different taxa. The plot of the second versus the third component for different populations of S. rnarmorata together with P. acuta complex is shown in Fig. 3.7 The majority of localities of S. marmorata is spread over quadrants I, II and Iv, but clustered in quadrant I, whereas, all P. acuta complex localities are restricted to quadrants III and IV, and clustered in quadrant III. The third component which is closely correlated to the adapical aperture angle (S-20) and the spire angle (S-9) is separating S. rnarmcrata localities from those of P. acuta complex. On the other hand, the second component which is closely correlated to the aperture insertion angle (S-19) and the spire angle (S-9), also improved the separation of the localities of these two species, although with -some superposition. Table 3.XXV shows that characters S-9, S-19 and S-20 differ statistically for the two genera. This study shows that PCA, focused on shell biometry alone, can be usefully applied to the identification of two species, S. marmorata and P. acuta Table 3.XXIV. Test for difference of means (t) of three, shell characters between two P. acuta complex groups. Null hypothesis: X 1 = X 2 .
Character Taxa N SD SE F Prob. T Prob. 2 Spire angle Physella A 44 68.75 6.35 0.96 (S-9) Physella B 23 53.69 3.27 0.68 3.78 0.001 12.81 0.000 Distance to Physella A 44 0.61 0.22 0.03 3.06 0.006 3.69 0.000 DZ (S-21) Physella B 23 0.45 0.12 0.03 sand of Physella A 44 0.83 0.39 0.06 Constant Physelia B 23 0.29 0.21 0.04 3.53 0.02 6.04 0.000 width (S-4)
SE = Standard error, other abbreviations as in Table 3.1. The probability of F is the ratio of the variances of the two groups._In pairs of groups with same variance ( P > 0.05), the variance values were pooled to calculate t values, otherwise the variances were treated separately.
2 Null hypothesis is rejected at the 0.05 level of probability.
'0 C r 24 0 tr 25 12 23 05 21 0 29 3• • 20 0
0 e • 17 10 15 ____ 2 2nd Is- PC o 13 ,18 32® • 00 0 8 0 31 6 17 14 0 3 • 39 Physella____ 0 030 -2 350
0 36 34
-o - -1 0 1 2 3 4
Fig. 3.7. Plot of the second and third principal components for Stenophysa marmorata ( 0 ) together with Physella acuta complex snails C®). Numbers refer to localities which are listed in Table 2.1. U' 20
18
16
14 z 12
LL 8
6
4 2
0 48 50 52 54 56 58 60 62 64 66 68 70 72 74
SPIRE ANGLE(OEGREES)
Fig. 8 Frequency distribution of spire angle of Stenophysa rnarrnorata from all localities (N = 102). 10
9
8
7
Cr U.
.o Ov 04 04 5b 55 60 62 64 66 68 70 72 74 76 78 80 82 84 86 88 90
SPIRE ANGLE (DEGREES)
Fig. 39 Frequency distribution of spire angle of Physella acuta complex of cluster B (N = 23) and A (N = 44). un
111) Table 3.XXV. Test for difference of means CO three shell characters between P. 'acuta complex and S.' marmorata samples 1 .
Character Taxa N SD SE F Prob. T Prob. Spire Physella A 44 68.75 6.35 0.96 1.70 0.031 8.51 0.000 angle Stenophysa 102 59.63 4.87 0.48 (S-9) Spire Physella B 23 53.70 3.27 0.68 2.22 0.034 -7.11 0.000 angle Stenophysa 102 59.63 4.87 0.48 (S-9) Aperture Physella A+B 70 144.00 11.34 1.36 2.50 0.000 8.38 0.00 Insertion Stenophysa 102 157.00 7.17 0.71 angle Adapical Physella A+B 70 41.79 9.56 1.14 1.45 0.090 -11.18 0.000 Aperture Stenophysa 102 26.79 7.95 0.79 (S-20)
1Abbreviations and explanations as in Table 3.XXIV.
I-J Ui 154
complex. Characters;S-20, S-9 and S-19 being of most importance to distinguish both speOies through shell morphology. It should be pointed out that linear characters related to the first component were not evaluated. The shells of P. papaveroi Leme:and P. cubensis sensu Leme were not included in the PCA, and are illustrated in Figures 3.10 and 3.11. Since 3shells of P. papaveroi had spire angle ranging from 850 to 910 and band of constant width ranging from 0.34 to 0.46 ram, they possibly differ from both clusters A and B of P. acuta complex (Table 3.XXIV). Besides, P. ppaveroi differs also for a strong shouldering in its spire. A single shell of P. cubensis sensu Leme had a spire angle of 670 and a band of constant width of 0.65 mm. These values agree with the material of P. acuta complex of cluster A (Figure 3.12), and differs from cluster B (Figure 3.13), from P. papaveroi. and from the analog of P. cubensis (lot GT-M230). This latter material is very similar to those of cluster B. The shells of P. acuta complex of clusters A and . B were also compared with shells of P. acuta complex from Britain and found to be similar, although the British material tended to have spire angles of intermediate values between the averages of the two clusters. However, material from London (L 46) (UrtB-142) exhibited exceptional variation in shell morphology, having specimens with high spire and small spire angle (52 0 ) and specimens with lower spire and larger spire angle (68° ) (Figures 3.14 and 3.15). This suggests that the differences found between clusters A and B might be intraspecific. For comparison a representative shell of S. marmorata of the main cluster is given in Figure 3.16. The shell of the extreme locality along the second principal component (Figure 3.5) is also given: L24, which has the narrowest spire angle (Figure 3.17). •1
FIGURE 3.10. Shell of Physella papaveroi from São Paulo city, São Paulo, MZ-16618 (paratype). The spire angle is 87.° 1 r J. -) 3
3.10 15b
3. 11
FIGURE 3.11. Shell of 'Physefla cubensis" sensu Leme from Rio de Janeiro City, Rio de Janeiro, MZ-17997. The spire angle is 67.° FIGURE 3.12. Shell of Physella acuta complex from the University of Brasilia, Brasilia, DF, UnB-194 (L-38), representative of cluster A (Figure 3,6). The spire angle is 71. lb 7
3.12 18
3.13 FIGtJE 3.13. Shell of Physeila acuta complex from Rio de Janeiro City, Rio de Janeiro, LlnB-103 (L-39), representative of cluster B (Figure 3.6). The spire angle is 54°. 159
3.14 FIGURE 3.14. Shell of Physella acuta coilex from London, ETigland, UnE-142, representing the narrcwst specimen in the sample. The spire angle is 5z.0 160
3. 15
FIGURE 3.15. Shell of Physella acuta complex from London, England, UnB-142, representing the widest specimen in the sample. The spire angle is 68.° FIGURE 3.16. Shell of Stenophysa rnarmorata from Fundo Creek basin, Brasilia, DF, UnB-006 (L-29), representative of the main cluster (Figure 3,5), The spire angle is 60,° 161
3. 16 162
3.17
FIGURE. 3.17. Shell of Stenophysa rnarrnorata from So Paulo, i'IZ-17123 (L-24) , the narrowest specimen in Figure 35. The spire angle is 48.° 163
CHAPTER 4. INTRASPECIFIC VARIATION OF ANATOMICAL CHARACTERS IN STENOPHYSA AND PHYSELLA POPULATIONS.
4.1. FUNDO CREEK STENOPHYSA POPULATION
4.1.1. VAPIABILITY
Thirty specimens of Stenophysa from the Fundo Creek Basin, Brasilia, Federal District, Brazil (L29) (Table 2.1) were utilized in this study. The sixteen anatomical characters, A-i to A-13, and A-24 to A-26 (Table 2.111), were separated in two categories, constant and variable characters, as they were found in the studied sample. The characters which were constant in all analysed specimens from Fundo Creek basin are: A-i, A-6, A-b, A-li, A-12, A-13, A-24 and A-26. Comments on these characters are presented below.
A-i. Tentacle type - With a translucent edge and a broad pigmented core, state C (Figure 4.1).
A-6. Tubular nature of the kidney - Fully tubular in 97% of the examined snails (Figure 4.8) and rarely half tubular and, half lamellar (3%). These two variations fit in state A.
A-lO. Digestive tract pigmentation - Lighly pigmented, showing scatered tiny black spots, state A 0 0.5mm 4.1.
Pi
Gz
0 1mm
4.2
FIGURES 4.1 and 4.2. Tentacle (top) and part of the digestive system of Stenophysa marmorata specimens from Fundo Creek basin, UnB-006. Tentacle type, broad solid core (State C): Ted, translucent edge without pigmentation; Bpc, broad pigmented core; Digestive system: C , caecum; Gz, gizzard; Int, intestine; M, muscle; Oe, oesophagus; Pi, pigments. 165
(Figure 4.2).
A-il. Shape of gizzard The gizzard, a non- bibbed organ, has its shape varying from;rcylindrical (26%) to round (74%). These variations ft in state A.
Preputial gland presence - The gland was absent in all analysed specimens, state A (Figure 4.3).
Penial complex type - The snails from this sample had a penial sheath that was glandular in its part proximal to the preputium, transparent and apparently non glandular in its distal part. As the sheath is transparent, the penis was seen within it, and was frequently twisted at the end of the non glandular part. Besides, the passage from the penial sheath glandular part to non glandular part is funneled and smooth, with no sign of juncture. Since the.penial sheath was frequently invaginated to a lesser or greater extent into the preputium, the length of the glandular part from the retractor muscle to the tip of the penial sheath inside the preputium was one third to one half the total glandular part (Figure 4.3). Despite the variation in length of the glandular part of the penial sheath,all specimens analysed belonged to state P.
A-24. Retractor muscles - Two thick and flesh unpigmented muscular branches which attach to the penial sheath, state A (Figure 4.3). 166
N
1 mm iIGURE 4.3. Penial complex type, State P, of Stenophysa marmorata from Funco creek basin, UnB-006. Nb, muscular branches, other abbreviations as in Figure 2.82. Retractor muscle state A. 167
A-26.. Number of s egments in the penial sheath - Two parted, a glandu1ar ; pärt near the preputium, and a non glandular part toward the vas. deferens, state A
(Figure 4.3). . S
The variable characters in the sample are: A-2, A-3, A-4, A-5, A-7, A-8, A-9, and A-25. Comments on these characters are presented below.
Mantle pigment.pattern. The mantle had circular unpigmented areas, usually medium sized (state G) in 69% of the specimens (Figure 4.4); less frequently large sized (15%), state E; very large (12%), state F; and small sized (0), state B.
Mantle lappet type - Mainly serrated (state C) in 58% of the specimens, and sometimes wavy (state A) (14%), blunt digated (state E) (14%), or spade digitated (state H) (14%) (Figures 4.5 to 4.7). Some specimens had more than one type of digitation.
Columellar lappet number - Most specimens (81%) had from 6 to 9 lappets (state B) in their mantle edge, a few specimens (4%) had four to five lappets (state C), but 15% had no lappets at all (state A).
- A-5. Kidney shape - The kidney is located in the ventral surface of the mantle and is divided in three regions: a narrow pericardial region, an expanded middle region with a transverse sinuous tubule ending in Ibb
Im
FIGURE 4.4. Dorsal view of the mantle with the pigmentation pattern of Stenophysa marmorata specimens from Fundo Creek basin, UnB-006. Mbs, mantle border skirt; Sgr, Secreting groove; Pi, pigmentation; Uc, unpigmented circles, approximately medium sized, State G.
169
Pi
Br
4.5
EVA 0.5mm -4
FIGURES 4.5 to 4.7. Mantle lappet type of Stenophysa marmorata specimens from Fundo Creek basin, UnB-006. By, blood vessels; Pi, pigmentation. Top, serrated (State C); middle, blunt (State E); bottom, spade digitated (State H). 170 a short ureter region, whose opening is near the base of the left margin of the pneumostomatic lobe. The kidney shape is determined according to the morphology of these three regions (Figure 4.8).
The pericardial region was in , a continuum with the middle part of the kidney, widening gradually in width in 63% of the analysed snails (Figures 4.9, A to D). The remaining individuals had a more elongate pericardial region with a narrowing abrupt transition to the middle region of the kidney (37%) (Figure 4.9, E to
G). The shape of the middle region also was variable within the sample from Fundo Creek population (Figure 4.10). In fact, in 40% of the snails the middle region was wider near the pericardial region, and in 60% near the center of the kidney. The ureter region had a gradual variation in its shape from tapering to blunt tip (Figure 4.11). According to the variability in the three regions of the kidney, most specimens belonged to three states, while others presented some variations that do not fit in any of the given states. The states are: C in 67% of the examined specimens, D in 3%, and H in 17% of them. The remaining 13% do not belong to any of the states presented.
A-7. Bursa copulatrix sh ape - I have found twelve variations of shape in the bursa copulatrix, described as (Figure 4.12): Al = oval, A2 = pyrifonn, A3 =pjriform 171
Tu. -
Pre
FIGURE 4.8. Kidney of Stenophysa marmorata specimens from Fundo Creek basin, UnB-006, showing pericardial, middle and ureter regions, and the full tubular nature of the kidney (State A). Mr, middle region; Pre, pericardial region ; Tu, tubules ; Ur, ureter region. 172
Prc
FIGURE 4.9. Outline of the kidney showing the passage between the middle (Mr) and pericardial (Pre) regions of S. marmorata specimens from Fundo Creek basin, UnB-006. A to D and E to G are examples of continuous and abrupt passages, respectively, between those two extreme regions. The arrow indicates the commonest kind of passage. 'VI' .LIJ
L e
FIGURE. 4.10. Outline of the kidney of Stenophysa marmorata specimens from Fundo Creek basin, UnB-006, showing the variation of the shape of the middle region
(Mr). Abbreviations as in Figure 4.8. FIGURE 4.11. Outline of part of the kidney of Stenophysa marmorata specimens from Fundo Creek basin, UnB-006, showing variations in ureter region (Ur) shape. Other abbreviations as in Figure 4.8. The arrow indicates the commonest ureter region shape. 174 mr
ur
'\ 175
4A2 -
FIGURE 4.12. Outline of the variation of the bursa copulatrix shape of Stenophysa marmorata specimens from Fundo Creek basin, UnB-006. Al to A3, oval to piriform shape; Al to B6, oval to B - shaped; Al to C3, oval to rectangular shape. 176 elongated, Bi = ellipsoidal waisted, B2 = oval waisted, B3 - slightly kidney shaped, B4 = elongated kidney shaped, B5 deeply kidney shaped, B6 = B - shaped, Cl = sub- rectangular shaped, C2 - round shaped, and C3 = flattened. The above variation were fit into six states: A in 26% of the examined specimens, B in 40%, C in 7%, E in 17%, and H in 3%. Though the bursa copulatrix dimensions and the length of the bursa copulatrix duct were not included in any character, they were examined and described as follows. The length of the body of the bursa copulatrix ranges from shorter to approximately 4 times longer than its width (Figure 4.13). In 80% of the examined specimens, the bursa copulatrix body is 1 to 2.5 times longer than wide, and 20% vary in length from 2.5 to 4 times the width of the bursa copulatrix body (Figure 4.13). The length of the duct of the bursa copulatrix ranges from smaller to approximately 2.2 times longer than its body length. Precisely, in 80% of the examined specimens the duct is 1 to 2.2 times longer than its body length (Figure 4.14).
A-8. The bursa copulatrix axis - In 97% of the specimens the bursa copulatrix body was aligned with the duct and the duct entered it centrally, state A (Figure 4.15). However, in 3% of the specimens the bursa copulatrix body length direction diverged from the duct axis orientation to form an angle close to 900, state. B (Figure 4.16). FIGURES 4.13 and 4.14. Frequency distribution of ratio of bursa copulatrix body length to its width, and length ratio of bursa copulatrix duct to the body, respectively, of Stenophysa marmorata specimens from Fundo Creek basin, UnB-006. .1-Il
6
T
6 >- 0 lii
0 4 Lu cr u- I
2
3.0 0.5 1.0 1.5 2.0 2.5 BODY LENGTH/ BODY WIDTH RATIO 4.13
>- 0 z Lu 0 Lu Of Li
10 1.3 1.6 19 Z.Z DUCT/ BODY LENGTH RATIO 4.14 178
415
FIGURES 4.15 and 4.16. Outline of the bursa copulatrix axis orientation, states A and B, respectively. Bursa copulatrix duct connection degree of adhesion, state C, represented by both Figures 4.15 and 4.16.Bcb, bursa copulatrix body; Bcd, bursa copulatrix duct; Ct, connective tissue; Fgo, female genital opening ; Ut, uterus. I-wJ
A-9. Bursa copulatrix duct connection - This character involves the ratio of the diameter of the bursa copulatrix duct (42) to the diameter of the uterus (dl). All Fundo Creek basin specimens, but one, showed a ratio that varied from 0.23 to 0.90 (Figure 4.17). The only exception had a d2: dl ratio of 1.07. This character also includes the extent of adhesion of the bursa's duct to the oviduct. Eighty percent of the specimens had the proximal portion of the bursa's duct attached to the oviduct wall by connective fibers. The remaining specimens had their duct totally adhered to the oviduct. The association of these two characteristics fit 77% of these specimens in state C (Figure 2.56). The remaining specimens had either the diameter of duct larger than that of the uterus or their duct totally adhered to the oviduct. Thus, these do not fit in any of the mentioned states. Though these variations are included in a single state, the character was considered among the variable ones for the 23% specimens that did not fit in state C.
A-25. Length ratio of penial sheath to the preputium (PSi/Pr or PS2/Pr) - If PS1 length is considered, the PSi/Pr ratio in the Fundo Creek specimens belonged to the following states: A-17% of the specimens; B-43%; C - 10%; and D - 30%. The ratio between these two organ lengths ranged from 0.86 to 2.85, mean 1.67, standard deviation 0.59 (Figure 4.18). On the other hand, if PS2 length was considered, the specimens belonged to the following states: A - 30%; B - 33%; C - 33% and 180
>- 0 z 1Li
0 Ui IL
d2/ dl RATIO
FIGURE 4.17. Frequency distribution of the ratio of the diameter of the bursa copulatrix duct (d2) to the diameter of the uterus (dl) of specimens of S. rnarmorata from Fundo Creek basin, UnB-006. .L$ J.
C-) z Li 0 Li Cr U.
0.8 1.3 1.8 2.3 2.8 5.5 PSI/Pr LENGTH RATIO
FIGURE 4.18. Frequency distribution of length ratio of penial sheath to the preputium (P51/Pr) of S. marrnorata specimens from Fundo Creek basin, UnB-006. 182
D - 4%. The ratio between these -two organ lengths ranged from 0.65 to 2.10, mean 1.24, standard deviation 0.44
(Figure 4.19).
4.1.2. IDNTIFICATION
There are eight anatomical characters, A-i, A-5, A-8,! A-9, A-10, A-li, A-12, and A-24 (Table 2.111) considered as constant within the genus Stenophysa (Te, 1978). In addition, in Fundo Creek sample A-1, A-6, A-10, A-li, A-12, A-13, and A-24 are constant. Comments on these characters are presented below.
A-l. Tentacle type - According to Te (1978) this character has four states in the family Physidae, and only one of them is presented in Stenophysa, state A. As all the specimens of this Stenopysa sample have tentacle type broad solid core (state C) they differ from all Stenophysa material examined by Te.
A-6. Tubular nature of the kidney - This character has three states in the family Physidae and two of them (states A and B) occur in the genus Stenophysa. State A is common to all species of Stenophysa, except for S. maugeriae and S. impluviata (Te, 1978). The Fundo Creek sample has state A in 100% of the specimens. Therefore this population differs from S. maugeriae and S. impluviata and agrees with the remaining species (Table 4.11). 183
>- 0, a LiJ 0 Lu
LL
2
C U.b 9.9 1.2 1.6 1.8 20 2.5 PS2/Pr LENGTH RATIO
FIGURE 4.19. Frequency distribution of length ratio of penial sheath to the preputium (PS2/Pr) of S. marmorata specimens from Fundo Creek basin, UnB-006. 184
Table 4.1. Ccparison of the ccmnonest states' of eight anatomical characters of the Fundo Creek Basin (N = 31) with those of the genus Stenophysa. Only those characters which are constant within this genus were considered (Te, 1978).
Fundo creek Basin Stenophysa2 Charadters 1 States Frequency -- (N = 35) A- 1. Tentacle type C 100% A A- 5. Kidney shape C - 67% A Bursa copulatrix axis A 97% A Bursa .copulatrix C 77% A duct connection A-10. Digestive tract A 100% A pigmentation A-li. Shape of gizzard A 100% A A-12. Preputial gland A 100% A presence A-24. Retractor muscles A 100% A
'The letters (states) have different meaning for each character (see Table 2.111).
2The data for the genus Stenophysa were taken from Te (1978). Table 4.11. Comparison of the commonest states of eight anatomical characters of the Fundo Creek Basin (N = 31) with those of the five Stenophysa species recognized by Te (1978). Only those characters which are variable within the genus and relatively constant within the species were considered Te (1978). The relative number of agreements with each species are given at the }ottom 2 . The data for the five species were taken from Te (1978).
Stenophysa species Spa Characters Fundo Creek Basin Sm Spe Srrau2 Si State Frequency (N=lO) (N=7) (N=lO) (N=4) (N=4)
A- 2. Mantle pigment pattern G 69% G G A A A A- 3. Mantle lappet type C 58% C A A C A A- 4. Columellar lappet number B 81% B A A B A A- 6. Tubular nature of the kidney A 100% A A B B A A- 7. Bursa copulatrix shape B 40% B A A A A A-13. Penial complex type p 100% p A A B. Q A-25. Length ratio of penial sheath A3 30% to the preputium PS2/pr) B 33% B A A B C C 33% A-26. Ni.nuber of segments in the A 100% A A A A B penial sheath
co U, Table 4.11. (Continued)
Stenophysa species Fundo Creek Basin Sm Spe Sniau2 Si Spa Characters State Frequency (N=liD) (N=7) (N=lO) (N=4) (N=4) N9 of agreements 8/8 4/8 2/8 4/8 2/8
'The letters (states) have different meaning for each character (see Table 2.111).
Stenophysa maugeriae; other abbreviations as in Table 2.1.
3These three states A, B, and C were considered equivalent in the Fundo Creek population. Thus, the three states were reckoned for the number of agreements of the five Stenophysa species.
p-J 00. 187
A-10. Digestive tract pigmentation - This character has four states in the family Physidae and only one of them occurs in Stenophysa: state A (Te, 1978). The Stenophysa Fundo Creek sample has all its specimens with diggestive tract pigment, state A, agreeing totally with Te (1978). A-li. Shape of gizzard - This character has two states in the family Physidae (Te,.1978) and only one of them is present in the genus Stenophysa, state A (Te, 1978). The analysed specimens of this - sample have gizzard shape, state A, thus,agreeing totally with Te (1978).
Preputial gland presence - This character has two states in the family Physidae, and only one of them occurs in the genus Stenophysa, state A (Te, 1978). Actually the presence of a discernible preputial gland is a character which distinguishes the members of the subfamily Physinae from those of the subfamily Aplexinae (Te, 1975 and 1978). The preputial gland is absent in all Fundo Creek specimens, state A.
Penial complex type - This character has 19 states in the family Physidae and four of them occur in the genus Stenophysa, states A, B, P and Q (Te, 1978). The specimens from the sample have the penial complex agreeing to Te's (1978) description for S. marmorata, state P, except for the length of penial sheath glandular part. Actually, state P is similar to state A in many aspects. The difference is that state P has a penial sheath funnelled and transparent, without juncture between glandular and non glandular part. Te (1974, 1975, 1978 and 1980) does not mention the occurrence of invagination of penial sheath into the preputium and neglects the measurement of the glandular part within the preputium. Though a segirent of the penial sheath is naird non glandular;, this description is actually restricted to the resolution level of the stereomicroscope. In fact, histological sections show that this part is also glandular' (Te, 1975)
A-24. Retractor muscles - This character has two states in the subfamily Aplexinae. Only one of them, states A, occurs in Stenophysa genus (Te, 1978). In fact, only state A is present in the population here analysed.
Though A-5, A-8 and A-9 were considered by Te (1978) as constant characters, Fundo Creek sample has presented variability in these characters, as discussed below.
A-5. Kidney shape - According to Te (1978), this character has eleven states in the family Physidae but only one of them occurs in Stenophysa, state A. The examined Stenophysa sample has three of the eleven states mentioned by Te, plus a new fourth variation. Thus, as this sample does not present state A, and do present states C (and/or E), D and H it differs from all the Stenophysa material described by Te (1978).
A-8. Bursa copulatrix axis - This character has two states in the family Physidae and only one of 189 them occurs in the:genus Stenophysa, state A. In fact, the great majority of Stenophysa:Fündo Creek specimens presented state A, agreeing with Te's results.
A-9. Bursa coju1atrix duct connection - This character has three states in the family Physidae and all of them occur in the su family Aplexinae: two in genus Aplexa and one in genus Stenophysa, state A (Te, 1978). As most specimens from Fundo Creek basin have state C and 23% do not fit in any of the given states, these specimens are different with respect to this character from all Stenophysa material described by Te.
In conclusion, in the sample, characters A-b, A-lb, A-12, and A-24 are constant and have the same states as described by Te (1978) for all Stenophysa species. Character A-8 shows 97% of constancy and also follows Te's description. Character A-6 is constant and have the same states as described by Te (1978) for all the Stenophysa species, with the exception of S. maugeriae and S. impluviata. Character A-13 is also constant and coincides with the one of S. marrnorata (Te, 1978) . However, characters A-i, A-5, and A-9 do not follow Te's description and are variable, with exception of A-i which is constant (Table 4.1). Characters A-2, A-3, A-4, A-7, A-25, and A-26 are variable anatomical characters within the genus Stenophysa. Comments on these characters are presented below. 190
Mantle pigment pattern - This character has seven states in the family Physidae and two of them occur in the genus Stenophysa: states A and G (Te, 1978). The specimens from Fundo Creek basin have four of the seven states: B, E, F, and G. As state G occurs in the great majority of specimens of this sample, this indicates that the population of Fundo Creek basin has greatest affinities with S. marmorata and S. peruviana (Table 4.11). Te (1978) and other authors have not found states B, E and F in this genus.
Mantle la2pet type - This character has eight states in the family Physidae and two of them occur in the genus Stenophysa. They are states B and ,C (Te, 1978). The sample of Fundo Creek basin has four of these states: B, E, and H. State C occurs in the great majority of the specimens of this population, indicating their greatest similarity with S. •rnarmorata and S. impluviata (Table 4.11). States E and H have not been previously mentioned for the genus Stenophysa by Te or by other authors.
Columellar lappet number - This character has four states in the family Physidae and two of them occur in the genus Stenophysa, states A and B (Te, 1978). The specimens of -the Fundo Creek population have three of them: states A, B and C. In fact, state B occurs in the great majority of the specimens. This is also a characteristic of S. rnarmorata and S. impluviata. State C has not been previously mentioned in the 191
literature for Stenophysa.
A-7. Bursa copulatrix shape - This character has 9 states in the family Physidae, and two of them, A and B occur in Stenophysa. As stateB isthe commonest state in this population, it suggests greater affinity of this population with S. marmorata. States C, E, H and K, which occur in three different genera of this family (Te, 1978), have not been mentioned before by Te, or by any other author, for Stenophysa.
A-25. Length ratio of penial sheath to the preputium (PS/Pr) - The great majority of the snails of the Fundo Creek population had the distal part of their phial sheath inserted into the preputium, probably due to relaxation; Although Te does not define penial sheath length, the drawings of penial complex in his 1975 and 1978 papers suggest it is from the insertion of the retractor muscle to the beginning of the vas deferens (PS2 length). Thus, this length was used for comparison with Te's results. Actually the character penial sheath/ preputium length ratio has four states in the family Physidae and three of them occur in the Stenophysa species: A, B, and C. However, considering PS2 length, the three states, A, B and C are equally common in this population, preventing any conclusion about the affinities of the Fundo Creek sample. Concerning PSi length, state B is the commonest in the population, indicating that this sample has highest similarity with 192
S. irnpluviata and. S. marmorata
A-26. Number of segments in penial sheath - This character has three states in'-the family Physidae and two of them occur in Stenophysa, states A and B (Te, 1978). State A, which is found in all species of Stenophysa (Te, 1978), except for S. panamensis, occurs in all the specimens of the Fundo Creek basin population. Searching anatomical characters variability in the genus Stenophysa, S. marmorata presents the highest percentage (100%) of agreement with specimens of Fundo Creek basin (Table 4.11). The similarity with the other species, S. impluviata, S. peruviana, S. panamensis and S. maugeriae, is 50%, 50%, 25% and 25%, respectively. Thus, based on this coincidence, the population of the Fundo Creek basin is here identified as S. marrnorata (Guilding, 1828).
4.1.3. ASSESSMENT OF CHARACTERS
Te (1978) examined several characters of both external and internal morphology of snails of the family Physidae, excluding those that were "redundant, invariant, too variant, difficult to score or appeared ill defined or in need of further study and clear definition" (Te, 1978, page 22). He found 71 characters suitable for use in Physidae taxonomy, 37 shell and 34 soft anatomy characters. Sometimes, however, these 193
numbers varied depending on the genus was dealing with, for example, in genus Stenophysa he analysed just 21 •anatomicalcharacters. It is not possible to draw conclusions about the range of variation of the different characters at intraspecific level in his work because he only shows the commonest state for each character in each species. Intraspecific study is so important that this section deals only with the range of variation of those characters within the sample from the Fundo Creek Basin population. Only 16 of these characters were analysed as the others were discarded (see Chapter 2.4.2). Half of the 16 characters were considered to be constant within the genus Stenophysa by Te (1978), and are discussed below (Table 4.1). Tentacle type (A-l) is constant in Fundo Creek population sample (state C) and differs from the commonest state of Stenophysa (state A) (Te, 1978). Thus, either this character is variable within. Stenophysa genus, and within S. marmorata species, or the fixative solution may interfere with the consistency of. the character. Studies on live material is wanted. Results regarding characters A-5 and A-9 also do not confirm Te (1978) because they are variables within the studied sample. Te's states for character A-5 lack precise definition and there is always a risk on choosing any state among a series of states separated by irrelevant differences. Characters A-5 and A--9 have a large range of variations which seem 194 sometimes to iritegrade the states and for this reason they seem to be unsuitable for systematical purposes, and should be reevaluated for their diagnostic value at generic level. Actually, the intergradation suggests that they can no longer be considered constant within the genus Stenophysa. Consequently, they will be considered from now onwards as characters of doubtful use. Five characters, A-8, A-lO, A-li, A-12 and A-24, are here considered characteristic for the genus Stenophysa because they have small or no variation among the specimens of Fundo Creek basin sample, and confirm Te's results (1978) (Table 4.1);. According to Te, characters A-5, A-9 and A-12 are good for separating Aplexinae from Physinae. However, with respect to A-5 and A-9, the analysis of the sample from Fundo Creek basin does not corroborate that conclusion for their variability. For generic level Te considers A-5, A-9 and A-lO important characters for the separation of Aplexa from Stenopysa (Te, 1978). Analysis of the Fundo Creek population sample only confirms the importance of character A-lO to do so, for the others are too variable. There are eight anatomical characters which are variable within the genus Stenophysa (Te, 1978): A-2, A-3, A-4, A-6, A-7, A-13, A-25 and A-26 (Table 4.11). Three of these, A-6, A-13, and A-26, were found to be constant within the studied population and so they are considered highly diagnostic at specific level, even 195 when dealing with-single specimens. Due to their exclusive states, character A-6 separates S. maugeriae and S. impluviata from other species, while character A-26 is highly diagnostic for S. panamensis (Table 4.11). Character A-13 has four states in Stenophysa genus,
A, B, P and Q (Table 4.11), and has consequently, a high diagnostic value. It allows the identification of three out of the five species of the genus: S. marmorata, S. irnpluviata and S. panamensis. Based on character A-13 alone, the studied population from Fundo Creek basin is identifiable as S. marmorata The other five characters, A-2, A-3, A-4, A-7, A-25, are variable within the sample of Fundo Creek basin population. They have a range of variation that intergrade with other three genera of this family, Physella, Aplexa and Physa. Such gradation has not been found before neither by Te nor by other authors in Stenophysa. As the characters have different degrees of expression in the same population (variable characters, Mayr, 1976) they have a low taxonomic significance, i.e., they are not reliable for analysis of single individuals. They might nevertheless be helpful in a population analysis in the absence of other alternatives. Character A-25 is a quantitative character with continuous variation and, according to Pankhurst (1978), such type of character tends to show a great deal of variation mostly due to environmental influences on the growth of organism, and are therefore 196 poor characters from the point of view of information content". On the penial corrLex study, the consequences of choosing either PSI or PS2 1enths may. lead to different conclusions about similarity among the species. I prefer to rely on PS2 measurements since they correspond to Te's method to measure the organ. Character A-4 is also quantitative and its range of variation is dependent upon the size of the specimens (Leme, 1966). Further studies are needed to specify the character's size/age dependence, to assess its role in systematics. Character A-7 is variable depending on sperm content. It becomes bibbed when full and spheroidal when normal (Duncan, 1958). Furthermore, the states are difficult to discriminate because they intergrade. Characters A-2 and A-3 have states difficult to differentiate, because they intergrade. Such states may be redefined as follows. Character A-2 should have four states, instead of seven: State A- obscure circles; State B - unpigmented circles; State C - smoothly pigmented; and State D - unpigmented. These are equivalent to Te's states A; B, E, F and G; C; and D, respectively. Character A-3 should have four states, instead of eight: State A - wavy; State B - digitated; State C - straight edged and narrow; State D - straight and wide. These are equivalent to Te's states A; C, E, F, G and H; D, and I, respectively. To minimize chances of mistakes on identification, it is important 197 to agglutinate B, E, F and G which are intergrading states of character A-2, as well as to group C, E, F,
C, And H, the intergrading states of character A-3. Another advantage is to limit the variability of these states which were initially intraspecifically variable, to - become of diagnostic value in the identification of species of the genus.
4.2. IDENTIFICATION AND VARIABILITY OF OTHER STENOPHYSA
SAMPLES
Only those anatomical characters which are variable within the genus Stenophysa (Table 4.11) (Te, 1978) were considered in the study of the Stenophysa material other than Fundo Creek population (Table 2.1). Character A-7, bursa copulatrix shape, however, was discarded due to its extreme variability in Fundo Creek population (Figure 4.12). Character A-1, tentacle type, though considered by Te (1978) as constant for the genus (state A), was also examined due to its difference in relation to Fundo Creek specimens (state C) (Table 4.1). The states of characters of 25 specimens from 22 localities are given in Table 4.111. I examined first the characters A-6, A-13 and A-26, tubular nature of the kidney, penial complex type, and number of segments in the penial sheath, respectively, Table 4.111. Comparison of states of eight anatomical characters of 25 Stenophysa specimens from 22 loia1ities 2 .
Characters Code Locality A-i A-2 A-3 A-4 A-6 A-13 A-25 A-26 L 1 Buenos Aires, Argentina D D A C A P B A
L 2 Buenos Aires, Argentina C D - A B P - A L 3 n.i Bela Horizonte, MG C G C B A P A A L 3 n.2 Bela Horizonte, M3 C B C B A P A A L 3 n.3 Belo Horizonte, MG C E C B A P B A
L 4 Chul, Uruguay •- - - - - P B A L 5 B. Pedrito, RS C B C B A P A A L 6 Martinique C B A A A P A A L 7 Jamaica C G C B A P A A L 8 La Isleta, Panama B D E B A P B A
L9 UruQuca, BA C D - - B P B A L 1 Coranbã,MS D B C C A P B A L 13 Recife, PE B D C B A P B A L 15 n.1 Brasilia, DF C F G B A P A A
L 15 n.2 Brasilia, DF - - - - - P B A
L 16 Brasilia, DF C B - - A P A A
L 61 Brasilia, DF C G D - B P A A 00 Table 4. III. (Continued)
Characters Code Locality A-i A-2 A-4 A-6 A-13 A-25 A-26 L 62 Brasilia, DF C G C B A P B A L 63 Brasilia, DF C B E B A P A A L 17 Rio de Janeiro, RJ C E E B A P B A
L 18 Itagual, R3 - - - - - P B A
L 21 Diarnantina, MG - - - C - P B A
L 27 4arirnLondo, MG - - - - - P B A
L 64 Manaus, AM - - - - - P B A L 65 Belên, PA C D - A A P B A
'The letters (states) have different meaning for each character (see Table 2.111,.
The sites of the localities are given in Table 2.1. - The dashes indicate missing data.
H
0 200
since they were the most reliable anatomical characters for discriminating Stenophysa species, as discussed in section 4.1.3. Consequently, allthe dissected specimens belong to S. marmorata (Guilding.), for the constancy of the states (Table 4.111). Only three specimens, one from Buenos Aires (L 2), one from Uruçuca (L 9), and a third from Brasilia (L 61) had character A-6, state B, instead of A. for this character, it is possible that the difference between a tubular (state A) and a lamellar (state B) kidney represents a physiological condition of the organ. Besides, these three specimens, L 2, L 9, and L 61, have the same penial complex type which is diagnostic to discriminate Stenophysa marmorata. In fact, this type is also common to the other specimens. This is a strong support to consider the specimens as S. marmorata. A necessary restriction is that the diagnostic value of - character A-6 can not be applied to single specimen identification due to its variability. Character A-i, tentacle type, was variable, state C being the commonest (80%) and states B (10%) and D (10%) less common. Actually, state C is also the commonest state in Fundo Creek population (100%). Since Te has found state A for all the Stenophysa species, and, being state A smooth and unpigmented tentacle, it is possible that a bleaching factor may be involved in the fixative fluid. Character A-2, mantle pigment pattern, is 201 highly variable, state D being the commonest (42%), while other specimens have states B (21), E (11%), F (5%) or G (21%). This variability agrees with the results of •Fundo Creek population for the absence of states A and C, and presence of states B, E, F and G. However, as discussed in section 4.1.3, states B, E, F and C intergrade and should be treated together as a single new state B. thus, the commonest state becomes state B (unpigmented circles) in 58% of specimens. The only difference between this group and Fundo Creek population is the occurrence of 42% state D (unpigmented) in the former. My own observation shows that pigmented mantle kept in Railliet-Henry fixative fluid loses pigmentation with time. Character A-3, mantle lappet type, is highly variable, state C being the commonest (53%), while other individuals have states A (13%), D (7%), E (20%) or G 7%). This variability is similar to that found in Fundo Creek population, except that Fundo Creek presented some specimens with state H and no individuals with states D and G. However, as discussed in section 4.1.3, states C, E, F, G and H intergrade and are best treated together as a single new state B. Following the new proposed states, state B (digitated) is the commonest (80%), plus states A (13%) and C (7%) in the remaining specimens. Actually, state B (86%) is the commonest state in Fundo Creek population. This high frequency on state B makes specimens from the studied localities similar 202 to those of Fundo Creek population, except for L 62 which belongs to state C. Character A-4, columellar lappet number, is also variable, state B being the commonest (69%), while other individuals have states A (19%) or C (12%). This prevalence of state B among these individuals agrees closely with the results of Fundo Creek population and those of Te (1978) for S. marmorata. Character A-25, length ratio of penial sheath to the preputium is variable. Most specimens have state B (63%) and the remaining state A. The prevalence of state B may distinguish these specimens from Fundo Creek population, which is variable, and may show closer similarity to Te's (1978) results for S. marmorata. Table 4.IV summarizes these results showing the close similarity of Fundo Creek and the remaining localities with Stenophysa marmorata of Te (1978). To identify the specimens, eight characters were selected for their variability within the Stenophysa species (Table 4.IV). A comparison between the states of Fundo Creek population, other localities, and Te's (1978) results for S. marmorata shows a striking coincidence among the three groups, except for character A-i. Actually, this coincidence allows the enclosure of Fundo Creek and other localities specimens in a single species, S. marmorata. These anatomical data support the irrelevancy of the variation pointed out by shell PCA. Figure 4.20 shows the distribution of the 203
Table 4. IV. Cornparisorx of the ccrnnonest states of eight anatomical characters of Stenophysa mannorata from the Fundo Creek Basin (N = 31) and 22 other. localities2 (N = 25) with those given by Te (1978) (N 10).
Fundo Creek Basin Other localities Te (1978) Characters State Frequency State Frequency State A- 1. Tentacle type C 100% C 80% A A- 2. Mantle pigment B 100% B 58% B pattern3 A- 3. Mantle lappet B 86% B 80% B type 4 A- 4. Columellar lappet B 81% B. 69% B number A- 6. 'Iübular nature of A 100% A 89% A the kidney A-13. Penial complex p 100% p 100% p type A-25. Length ratio of A 30% penial sheath to B 63% B the preputium B 33% (PS2/Pr) C 33% A-26. Number of segments A 100% A 100% A in the penial sheath
he letters (states) have different meaning for each character (see Table 2.111 for all characters except A-2 and A-3; for the new characters see superscripts 3 and 4).
2The localities are given in Tables 2.1 and 4.111.
3me states given here are the new states proposed in section 4.1.3. The equivalence with the older states are: new A = A; new B = B, E, F and G; new CC; new D=D.
4The states given here are the new states proposed in section 4.1.3. The equivalence with the older states are: new A = A and B; new B=C, E, F, G and H; new CD; hew D=I. FIGURE 4.20. Map of distribution of the Stenophysa marmorata lots analysed in this study. The numbers refer to the locality code as given in Table 2.1. Localities 14, 16 and 28 are not precisely located due to insufficient information about the collection sites. 204
200
100
00
00
00
0 0
0°
0°
90° 80° TO° 60° 500 400 205
Stenophysa specimens identified as S. marmorata in this study. The results are based on the study of dissected specimens and analysis of empty shells.
4.3. IDENTIFICATION AND VARIABILITY OF PHYSELLA SAMPLES
A preliminary identification of the material of Physella listed in Table 4.V was based on the study of the penial complex (character A-13). The material was considered as Physella (Costatella) section Alampetista, belonging, therefore, to Physella acuta complex. In order to approach the identification of this material at a specific level only those variable characters within section Alampetista, and constant for each species, were considered. There are 12 such characters (Te, 1978). However, three of them, mantle border skirt length, mantle border pigmentation, and type of preputial head, were discarded for the reasons given in section 2.4.2. Two other characters, A-S. Kidney shape and A-7. Bursa copulatrix shape, were also discarded due to their extreme variability within Fundo Creek population of S. marmorata (see section 4.1.3). The remaining seven characters, A-i, A-2, A-4, A-13, A-20, A-22 and A-23, were examined in at least one specimen from each of 16 localities (Table 4.V). At first, the specimens from localities L 30, Table 4.V. Caparison of the states of seven aiatcrncal characters of 28 Physella specimens fran 16 localities 2 .
Characters Code Voucher n9 Locality A-i A-2 A-4 A-13 A-20 A-22 A-23
L 30 UnB-35 n.1 Brasilia, DF B D B H* B A A L 30 tJnB-35 n.2 Brasilia, DP B D B H A A A L 30 tJnB-37 Brasilia, DF C D C H B A A
L 30 UnB-65 n.1 Brasilia, DF B D C H B - A L 30 UnB-65 n.2 Brasilia, DF B B C H A B A L 38 UnB-194 n.1 Brasilia, DF C C G H B A A L 38 UnB-194 n.2 Brasilia, DF C C E H A A A L 38 UnB-194 n.3 Brasilia, DF C C G H A A A L 31 UnB-85 n.l Rio de Janeiro, PJ B G G H B A A L 31 UnB-85 n. Rio de Janeiro, RI B G G -H B A A L 39 UnB-103 Rio de Janeiro, RI B G G H B A A L 39 UnB-156 Rio de Janeiro, RI C G G H B A A L 39 UnB-166 Rio de Janeiro, RI C C C H - B . A
MZ-17997 - L 60 Rio de Janeiro, RI D - H B B A L 66 UnB-221 Rio de Janeiro, RI C B B H B B A L 34 UnB-173 Santos, SP B B B H B C A
L 59 MZ-16618 - são Paulo, SP C - H B B A 0 0 Table 4.V (Continued)
Characters Code Voucher N9 Locality A-i A-2 A-4 A-13 A-20 A-22 A-23 L 32 UnB-157 Cochabamba, Bolivia A D D H A B A L 33 UnB-164 n.1 Lima, Peru A D D H C C A L 33 UnB-164 n.2 Lima, Peru A D D H A B A L 35 UnB-174 Trinidad A B B H* A B A L 36 UnB-177 Puerto Rico A B B H A B A L 37 UnB-178 Puerto Rico C G G H* A B A L 43 UnB-138 Leeds, England C E C H B A A L 45 UriB-141 Aberdeen, Scotland C E B H B A A L 46 UnB-142 n.1 London, England B E C H B A A 46 L UnB-142 n.2 London, England C E C H A A A L 46 UnB-142 n.3 London, England C B B H B A A
'The letters (states) have different meaning for each character (see Table 2.111).
2The localities are given in Table 2.1. - The dashes indicate missing data. * Penial complex variation within state H. 208
L 32, L 36, L 37 and L 38 (Table 2.1) grouped in cluster A in the principal component analysis of shell characters, were compared with those specimens of localities L 31 and L 39, grouped in cluster .B (see Figure 3.6). The little difference, as presented in Table 4.VI, was not considered relevant to distinguish the two clusters, since all characters, except A-13, play a relative role to discriminate species at population level (discussion: section 4.1.3). In fact, A-13 is among the most reliable characters to distinguish Physella species. The size of the preputial gland is one of the characters used to distinguish P. acuta from P. heterostropha. It is important that all individuals of cluster B presented small preputial glands, state B (Figure 4.21), while- most cluster A individuals presented medium size preputial gland, state A (Figure 4.22). Other minor differences are probably due to fixative properties, as in the cases of state C of character A-i, tentacle type, and state D of character A-2,rnantle pigment pattern, as discussed in section 4.1.3. Two specimens of cluster A presented penial complex slightly different from the others and were marked as H* in Tables 4.V and 4.VI. This difference consists on a narrowing on the preputial head. However, these variants can hardly be mixed with any other given state (Te, 1978), and are best treated as variations within state H (Figures 4.23 and 4.24). Principal component analysis of shell 209
Table 4 .VI. Frequencies of the statesof seven variable anatomical characters of Physella acuta complex specimens included in clusters A and B.
Number of specimens Characters States Cluster A Cluster B
A- 1. Tentac1e type A 2 0 B 4 3 C. 5 2
A- 2. Mantle pigment pattern 3 B 6 4 C 0 1 D 5 .0
A- 4. Columellar lappet number A 1 0 B 6 4 C 4 1
A-13. Penial complex type H 9 5 2 0
A-20. Size of preputial gland A 7 0 B 4 4 A-22. Length ratio of penial A 7 4 sheath to the preputium B 3 1 (PS2/Pr) A-23. Swelling of penial sheath A 8 5 terminus C 1 0
See Table 2.111 for the waning of the states. The letters have different meaning for each character.
2Cluster A comprises specimens from localities L 30, L 32, L 36, L 37 and L 38; and cluster B comprises specimens from localities L 31 and L 39 (Table 2.1).
3The states given here are the new states proposed in section 4.1.3. See Table 4. IV for equivalence with older states. PANUO
t;i
4. 22 4.21
FIGURE 4.21. Penial complex of Physella acuta complex, from Rio de Janeiro City, Rio de Janeiro State, UnB-103 (L 39), representative of cluster B (Figure 3.6). Pg, preputial gland; Pr, preputium; Ps, penial sheath; Rrn, retractor muscle; Vd, vas deferens.
FIGURE 4.22. Penial complex of Physella acuta complex from University of Brasilia, Brasilia, DF, UnB-194 (L 38), representative of cluster A (Figure 3.6), abbreviations as in Figure 4.21. 211
1mm 4.23 r4i
FIGURE 4.23. Penial complex of Physella acuta complex from Brasilia, DF, UnB-35 (L 30), varia rit of type 1.1(1.1*) Abbreviations as in Figure 4.21. FIGURE 4.24. Penial ccr.ipiox of Phys.ella acuta complex, from Rio Piedras, Puerto Rico, UnB-178 (L 37) , variant of type H(H*). P, penis. Other abbreviations as in Figure 4.21. 212
characters segregated localities L 33, L 34, L 35 from clusters A and B (Figure 3.6). These three localities are anatomically close to those included in clusters A and B except that L 35 has a penial complex type H* (Figure 4.25), and L 33 has preputial gland larger than that in Physella from localities included in clusters A and B (Figure 4.26). Asa results, the specimens from both clusters were regarded as belonging to the same taxa and were lumped together with all Neotropical localities of Physella acuta complex, except localities L 59 and L 60. P. acuta complex from 13 Neotropical localities were compared with the same species complex from Britain (L 43, L 45, L .46), and P. cubensis (Te, 1978), as well as the states of the only two P. acuta complex species (Te, 1978) already recorded for South America, P. venustula (Gould) and P. squalida (Morelet), and two other species which might have been introduced in this continent, as they have been introduced elsewhere, P. acuta (Draparnaud) and P. heterostropha (Say) (Te, 1978) (Table 4.VII). Only three characters were constant in both the Neotropical and British material (Characters A-13, A-21 and A-23) which indicates that these snails have greatest agreement with P. acuta and P. Heterostropha (Table 4.VII). Character A-4, though variable, even within a same locality (Table 4.V), shows great similarity between the Neotropical material and P. acuta sensu stricto. All the examined Physella specimens Presented flattened preputial gland mostly of medium or 213
m F
4.25
FIGURE 4.25. Penial complex of Physella acuta complex from Saint George, Trinidad, UnB-174 (L 35), variant of type H(H*). Abbreviations as in Figure 4.21. 214
Vd
[
.Lmm 4.26 4.27
FIGURE 4.26. Penial complex of Physella acuta corrlex from Lima, Peru, UnB-164 (L-33), with large preputial gland. Abbreviations as in Figure 4.21. FIGURE 4.27. Penial complex of Physella acuta complex from Aberdeen, Scotland, UnB-141 (L 45), with medium preputial gland. P, penis. Other abbreviations as in Figure 4.21.
Table 4. VII. Comparison of the canrronest states of seven anatomical characters of Physella acuta complex from 13 Neotropical localities (N = 23) with five species of Physella according to TO (1978).
Great Britain Neotropical Region 3 Characte PS Pa Ph Pv Pc State Frequency State Frequency A- 1. Tentacle type. C 80% B 36% C B B C B C 41% A- 4. Colurrl1ar lappet number B 40% B 67% C B C B C C 60% A-13. Penial complex type H 100% H 100% G H H G M Size of preputial gland B 80% A 48% B B A B C B 48% Tendency of the preputial B 100% B 100% B B B B C gland to flatten Length ratio of penial sheath A 100% A 53% A A B!: A D to the preputium (PS2/pr) B 42% Swelling of penial sheath A 100% A 100% A A 4 . B B terminus 'The letters (states) have different meaning for each character (see Table 2.111). 2The localities are given in Table 4 .V. 3lthbreviations as follows: Ps, Physella squalida; Pa, P. acuta; Pc, P. cubensis; Ph, P. heterostropha; U, Pv, P. venustula. 216 small size. The penial sheath is entirely non glandular. The ratio between the lenghts of penial sheath to the preputium ranged from shorter to approximately as long as the preputiwn. The preputial head has a whitish appearance, and the end of the penial sheath presented a smaller diameter than that of the preputial head. These characteristics fit the examined Physella specimens within penial complex type H (character A-13). Characters A-20 and A-22 were variable within the Neotropical material and presented approximately equal number of states A and B individuals, agreeing partially, therefore, with both P. acuta and P. heterostropha (Table 4.VII). Characters A-i and A-2 are equal for these two species (Table 4.VII). Based on the similarity of the states of characters. A-20 and A-22 the British material belongs to P. acuta sensu stricto (Table 4.V11). These samples from Britain agree in every detail with the South American analysed specimens especially in the penial complex type (Figure 4.27). The specimens of P. papaveroi (MZ 16618) and P. cubensis sensu Leme (MZ 17996) show affinities with both P. acuta and P. heterostropha (Tables 4.V and 4.VII). The only anatomical characters which could help to distinguish P. acuta from P. heterostropha (Characters A-4, A-20 and A-22) were found to be variable. However, the small size of P. papaveroi and P. "cubensis" preputial gland (Figures 4.28 and 4.29) suggest a close affinity between these and P. acuta sensu stricto.
217 d
PS
Rm Pg
we
Pr
1 mm WA 4.29 FIGURE 4.28. Penial complex of Ph , (Physella) papaverol Le, Sao Paulo City, So Paulo State, i4-166lu (L 39), paratype, with small preputial gland. P, penis. Other abbreviations as in Figure 4.21.
FIC-IJRE 4.29. Penial complex of Physa (Physella) cubensis sensu Lerne, Plo de Janeiro City, Rio Cc Jneiro State, MZ-17997 (L 60) with small preputial gland. Abbreviations as in Figure 4.21. 218
Figure:4.30 shows the distribution of Physella specimens identified. in this study as P. acuta sensu stricto and P. papaveroi. The - distribution is based on the study of dissected specimens and .arialysis• of empty shells. FIGURE 4.30. Map of distribution of the Physella (Costatella) acuta sensu stricto and Physella (Costatella) papaveroi lots analysed in this study. The numbers refer to the locality code as given in Table 2.1. All localities but L 59 belong to P. acuta. All the samples were collected in artificial lakes, thanks or ditches. 219
20°
JAMAICA UERTORICO
ARTINIQUE 0 PA AMA ((ç Tr,. INIDAD - -. 10° S V NEZUELA -
Surinam luliana ) COLOMBIA
\ •) ..\ ______0° (1 CUADOR";\ .1' ) ii c \PERU . 10° T . iB R AZ I:L
BOLIVIA C32 \ 20 ° -- • 1 ______I - V4\•t 1\ t 59 :30°
40°
500
900 800 700 600 500 400 220
CHAPTER 5. GENERAL DISCUSSION AND CONCLUSIONS
Many shell measurements that are found in the litterature (see for example Parodiz, 1951; Hubendick, 1951) are difficult to be repeated since they lack a well determined shell orientation. Since any tilt in the shell to be measured may modify all references for measurements, some comparisons may lead to nonsense. The new conchometric method developed in this work allows a precise shell orientation to be repeated by any researcher. To evaluate the significance of shell characters the following considerations on size dependence is given. Shell characters S-9, S-10 and S-19 were found to be size independent at intra and interpopulational level in S. marmorata and P. acuta complex. S-20 shell character has the above attributes just in S. marmorata, but is size independent only at intrapopulational level inp. acutacomplex. S-21 and S-4 were found to be size independent, at inter and intrapopulational level in P. acuta complex, but S-4 is also size independent, at intrapopulational level, in S. marmorata.S-13 was found to be size independent only at interpopulational level in S. marrriorata, while S-23 is size independent at intrapopulational level in both species. S-23 was not analysed at interpopulational level. 221
Size dependence, on the other hand, may lead to redundancy. PCA revealed that characters S-1, S-2, S-3, S-5, 5-7, S-12, S-15 and S-16 were all found to be redundant with shell size. Character S-14, though not analysed in PCA, was also found to be redundant with size based on intrapopulational correlation. Characters S-2 and S-3 should be chosen for morphornetric shell analysis due to the high correlation with size and easiness for measurement. Character S-3 should be selected in cases of shell apex erosion. Characters S-6, S-8, S-ll and S-17 were found to be redundant with size. In addition they were found to be correlated with the Adapical Aperture Angle, Sutural Angle, Aperture Insertion Angle and Spire Angle. Therefore these characters are influenced both by size and angles of the shell, being considered of doubtful use. Size alone accounted for about 50% of the total variation observed in shell principal component analysis of S. marrnorata and P. acuta complex. It is regrettable that mature individuals could not be distinguished from iminatures, based on shell morphology, and the analysed lots did not represent random samples of the surveyed populations. Consequently, the shell characters related to size could not be considered in an attempt to discriminate taxonomic units among the studied material. All these difficulties suggest that the use of characters which are size dependent, for taxonomical purposes, should be abandoned. Several workers have 222 documented how the maximum size in different populations within a single, Physidae species varies according to environmental characteristics such as size and duration of water bodies, chemical characteristics of thewater, level of polution, water turbulence, etc (Aboul-Ela & Beddiny, 1969; Baker, 1930; Wurtz, 1949). On the other hand, if the effect of size upon these characters could be canceled, then such characters should be profilable for taxonomical purposes. This can be achieved either through regression analysis with tests of difference of regression coefficients, b,and allometric coefficients, a, or by the use of ratios where the desired character is divided by 'another character which only reflects size, preferably the shell length or the maximum shell width. But the ratios can only be used with confidence where the ratio is constant throughout the size range of a given population. Data here presented demonstrate. that this constancy is not invariably found among Physidae species. As a result much of the literature on snail taxonomy based on ratios of shell dimensions should not be accepted readily (Atchley et al., 1976). Besides, the ratios distribution often are unusual, perhaps far from normal (Pimentel, 1979). Furthermore the use of parameter a of the regression equation has been much criticized in the past, but as put by White and Gould (1965) there is no reason not to utilize it, provided that the same scale is used for the material to be compared. In fact, statistical 223 tests of differences of parameter a have provided a powerful tool to separate similar: species or environmental morphs (eg. Baxter, :1983). The use of regression analysis provides a much powerful insight into the understanding of the taxa being studied, better than the use of ratios, and should, therefore, be preferred. Future studies on Physidae systematics should bebased on tests of differnce of the regression coefficient b and the allometric coefficient a. The relationship between some characters was investigated. For a given shell length, S. marmorata has a shell width smaller than that of P. acuta; and for a given shell width, S. marmorata has a shell aperture wider than P. acuta. Large P. acuta shells have LIM (character S-5) smaller than S. marmorata of the sam e— size. It means that the former has its widest region situated closer to the shell apex than the latter. This relationship does not hold for smaller shells. Some shell characters presented different allometric growth curves for S. marmorata and P. acuta, evidenciating the dangers of using ratios of characters for identification purposes. On the other hand, at least for some shell chararacters, SW/SL and SS31SW, the ratios of two shell measurements were found to be constant throughout the size range of the analysed material, giving support to the use of these ratios in taxonomic studies. This study indicates that ratios can only be used in Physidae systematics for those cases where the 224 taxa to be compárd have the same regression coefficients but different parameter a. Another approach to dit1nguish Physidae species was the application of PCA to the shell data. Stenophysa marmorata and Physella acuta could be distinguished through shell PCA based on characters S-20, S-9 and S-19 which are the major representative of the 2nd and 3rd principal components of the total shell variation measured. S. marmorata has wider Aperture Insertion Angle (S-19) and narrower Adapical Aperture Angle (S-20) than Physella acuta, and intermediate spire angle (S-9) values between those presented by the two clusters of Physella acuta. Besides the shell morphometry approach, anatomical data provided support to species diagnosis. Out of the eigth anatomical characters regarded by Te (1978) as constant for the genus Stenophysa, and therefore of generic diagnostic value, only five, A-8, A-lO, A-li, A-12 and A-24 proved to be constant within the analysed material in this work. These five characters were also regarded as constant for the section Allampetista of Physella (Costatella) by Te (1978), but only A-12 was analysed in the Physella acuta complex material in this work, and was found to coincide with Te's results. Character A-12 is highly diagnostic for separating Aplexinae from Physinae, while characters A-lO and A-li help to distinguish Stenophysa from the other genera. Character A-8 helps to distinguish only Physella 225
(Physella) from the.remaining genera and subgenera as defined by Te (1978). Three other anatomical characters, A-i, A-5 and A-.9, were found to be variable within the studied populations of Stenophysa marmorata, and character A-i was also found to vary among the analysed Physella acuta complex samples (characters A-5 and A-1 were not analysed for Physella in this study). Besides being variable, these three characters differed in their commonest states from those given by Te (1978), for Stenophysa, agreeing with the commonest states given by Te to Physella and Physa. All these three characters have poorly defined and intergrading states. Furthermore, character A-i is apparently affected by preservation techniques. These characters are, therefore, regarded of doubtful taxonomic value for the Physidae. Out of eight anatomical characters considered by Te (1978) as variable within genus Stenophysa, and therefore of specific diagnostic value, only three, A-6, A-13 and A-26, were found to be constant within the Stenophysa material analysed. Character A-13 was analysed also in Physella material and proved to be constant too. Character A-13 is probably the most valuable character to identify South American Physidae species, even for single specimens. States A and B of character A-6 were not always easily distinguished. The difficulty might be due to physiological differences at
fixation. - 226
The ot•her five characters,A-2, A-3, A-4, A-7 and A-25/A-22, which Te (1978) considered variable in genus Stenophysa ; proved to be variable within the Stenophysa material analysed. Additionally characters A-2, A-4 and A-25/A-22 were found to be variable also within the Physella acuta complex material studied. All these five characters have intergrading states which are arbitrarily separated. They present a range of variation which overlap with results of other Physidae genera. The variability of these characters is commonly found within populations of Physidae and other freshwater snails.It was already documented by Baker (1901) for characters A-3 and A-4, Duncan (1958) for character A-7, Richards (1964) for character A-3, Leme (1966) for character A-4, Hamilton-Attwell et al (1970) for characters A-2, A-4 and A-22/25 and Paraense (1980) for character A-7. Character A-4 is correlated with the size of the snail (Leme, 1966; Hamilton-Attwell et al,,1970). Character A-7 is influenced by physiological state of the snail at fixation, namely the quantity of sperm within the bursa copulatrix (Duncan, 1958). Character A-2 might be affected by fixation, with consequent loss of pigmentation. All these five characters, therefore, have poor diagnostic value. Three additional characters, A-20, A-21 and A-23, were analysed in Physella acuta complex material in this study. Characters A-21 and A-23 were found to be constant among the studied samples, while character A-20 227
exhibited wide variability. Characters A-20 and A-21 are important to distinguish Physella (Costatella) of section Allampetista from section Costatella, but should be treated as continuous quantitative variables, in order to avoid difficulties to distinguish intergrading arbitrary states. Considering internal anatomy, the literature presents evidence of variation in genus Stenophysa. Brazilian populations, as discussed in this work, is also variable. The descriptions of the soft anatomy of S. marmorata published by Baker (1930) and Richards (1964) based on Venezuelan and Puerto Rican material, respectively, agree closely with the results obtained in this work (Chapter 4). Richards (1964) described variation in mantle lappet type (Character A-3) with specimens having either serrated or wavy mantle edge. According to Te (1978), only S. marmorata and S. impluviata have serrated mantle, while the remaining species of Stenophysa genus have wavy mantle edge. The occurrence of both states might suggest that Richards was dealing with two species. On the other hand, occurrence of high variability of this character in Brazilian populations of S. marmorata weakens its diagnostic value. In addition, Richards' material exhibited PS/Pr ratio larger than those given by Te (1978). The conflict is unimportant since the present study points out a high variability of that character. Richards (1964) also mentioned the occurrence of two types of Stenophysa in 228
Puerto Rico, a large form he identified as S. marmorata, and a smaliform. The small form is. - characterized by digitated mantle edge and short bursa copulatrix duct, while the large form has serrated mantle edge and long duct. The Brazilian populations of S. marmorata here studied presented similar variation as the Puerto Rican Stenophysa, suggesting that these two forms are intraspecifics variations. Principal component analysis indicated that the shells of the Neotropical material analysed in this work might be separated in two clusters, main cluster and minor cluster, based on differences of spire angle. Tests of regression coefficients for 23 pairs of shell characters between the S. marmorata populations of Fundo Creek (L 29) with Recife (L 13) and Fundo Creek (L 29) with Belo Horizonte (L 3) belonging to the main cluster did not show any significant difference, but tests between samples of Fundo Creek (L 29) (main cluster) and Uruçuca (L 9) (minor cluster) disclosed significant differences between three pairs of characters, S-2 and S-17, S-3 and S-17, S-7 and S-14. The result of the regression involving localities of clusters A and B reinforces the differences between the two clusters as observed in PCA. Notwithstanding, there are populations presenting intermediate values of spire angle. Actually, the frequency distribution of spire angle for all studied S. marrnorata localities is normally distributed. 229
Besides, no difference was found among 16 anatomical characters studied in localities belonging to the main and minor clusters. Thus, it is concluded that the above mentioned shell differences between specimens from main and minor clusters only represent an expression of intraspecific variability. Stenophysa marmorata has apparently been correctly identified in the past by most authors based on shell characters. However, there were some doubts to distinguish this species from S. peruviana and S. .impluv.iata. South American specimens from East of Andes were usually ascribed to S. marmorata, while those from West of Andes were identified as S. peruviana and those from Central America as S. irnpluviata. However, Te (1978) reported S irnpluviata from South America (Uruguay, Te personal communication) and S. peruviana from Brazil and Venezuela. Therefore, great care should be taken while attempting to identify South American Stenophysa, since these three species are conchologically and anatomically very similar (Te, 1978). According to the results of the present study, the only reliable anatomical character for the separation of these species is the nature of the glandular part of the penial sheath. S. impluviata has its glandular part of penial sheath adenate to the preputium and occurs behind the junction of the. retractor muscles, while S. peruviana and S. marmorata have their glandular part not adenate to the preputium and the retractor muscles join the 230 penial sheath in the middle of the glandular part (Te, 1978). S. peruviana differs from S. marmorata. S. Euviana has a sharp junction separating the glandular from the non glandular part of the penial sheath, and the glandular part is short and rounded, while S. marmorata has a glandular part which is funnelled and joins smoothly the non glandular part (Te, 1978). It must be emphasized that further studies on the variability of this character are needed to prove that such differences do not integrade in populations of S. peruviãna and S. impluviata. If this character proves to be variable, then, the separation of these three species will not be tenable unless additional differences are found, and they will fall into the synonymy of S. marmorata or E. auviana, whichever proves to be the oldest name proposed since both were published in 1828. As a complementary comment, the two subspecies of S. peruviana proposed by Te (1978), S. p. peruviana (Gray) and S. p. spiculata (Morelet) can not be maintained as such becauce they do not differ anatomically, are very similar conchologically and have the same distribution range (Te, 1978). They can not be regarded as subspecies as they occur simpatrically. Therefore, either spiculata is considered a mere shell ecotype or it should be treated as a distinct species. The former alternative seems to be more appropriate at this stage, as no important anatomical difference was found between them (Te, 1978). 231
This study not only confirms the presence of S. rnarmorata in several Neotropical areas already mentioned in the literature, but also adds the following new records: Panama; Peru; Amazonas State, Brazil; Bahia State, Brazil; Espirito Santo State, Brazil; Minas Gerais State, Brazil; Federal District, Brazil; Mato Grosso do Sul State, Brazil; Rio Grande do Sul State, Brazil. I had the opportunity to study the shells of the specimens recorded by Morretes (1949) for Curitiba, Paranä State, Brazil; Ponta do Ipê Aràado, Goiás State, Brazil; and Fortaleza, Cearä State, Brazil. The morphology of all these specimens agrees with the other specimens of S. marmorata analysed in this work. Actually, all Stenophysa material examined in this work belong to S. marmorata (Guilding). This picture confirms the idea of earlier workers (Clench, 1936; Parodiz, 1956; Morretes, 1949; Fernandez, 1981) that this species is widespread and common throughout South America East of Andes. The Southernmost recorded for this species in the recent literature and in this work is Buenos Aires, Argentina and Montevideo, Uruguay (35 ° Latitude South). However, Orbigny (1835 and 1837) recorded a species of Physidae (as Physa rivaljs Sowerby var. minor = Physa sowerbyana Orbigny) from the mouth of Rio Negro in Northern Patagonia, Argentina (41 ° Latitude South). Physa sowerbyana is considered a junior synonym of S. marmorata, but this record should be confirmed by anatomical study. The Northernmost record for this 232 species is Hispaniola and Jamaica in the West Indies
(19 0 Latitude North) (Clench, 1939, and Pointier, 1976). The present study does not enlarge this latitudinal spread of the species but includes samples from both extremes. The finding of S. marmorata in.Brasilia, DF, Manaus, AM, Corumbá, MS, and Nioaque, MS, extends considerably the known distribution of the species but. fails to prove that it is native in the Central Brazilian Plateau, Amazonian Basin or Paraguay Basin.. In the Federal District, for example, this species was not found in undisturbed natural water bodies (Monteiro & Dias, 1980). As mentioned in the introduction, earlier collectors failed to find this species in the Amazon Valley (eg. Sioli, 1951). The literature registers a widespread occurrence of P. cubensis throughout the Americas. In fact, Physella cubensis (Pfeiffer) is a common West Indies Physidae and has also been recorded for Southeastern United States, and Central America (Te, 1978), and for South America thrice (Baker, 1930; Leme, 1966; Te, 1978). However, this species is very similar to P. acuta (Draparnaud) which has been introduced into several countries of different continents, possibly by aquarists (Beetle, 1973; Bruggen, 1966; Clench, 1934; Hamilton-Attwell et al., 1970; Te, 1978). P. cubensis and P. acuta are, apparently, being mistaken in the Neotropical Physidae literature. Te (1978) gave much emphasis to the penial sheath morphology to distinguish 233 these two species. Actually, the penial sheath morphology is an important tool to orient distinction between P. cuhensis and P. acuta. P. acuta has a totally non glandular penia.l sheath, while P. cubensis has a penial sheath divided into anon glandular and a glandular part. However, Te (1978) points out also that the morphology of the glandular penial sheath of P. cubensis is intermediate between that presented in Physella (Physella) specimens which have a well marked glandular part and that presented in Pysel1a (Costatella) section Alampetista which have a totally non glandular penial sheath. This suggests that the glandular part of the penial sheath of P. cubensis may not be so conspicuous, or is less evident than in Physella (Physella) spp. In fact, the presence of a glandular part in the penial sheath was not reported by earlier workers as Richards (1964) and Harry and Hubendick (1964) in so called P. cuhensis material found in Puerto Rico. The descriptions and illustrations in those authors papers do not allow identification of glandular penial sheath. Thus, it is 'not possible to make a safe judgement about the correct identification of the species, either P. cubensis or P. acuta. In fact, Te (1978) reported P. acuta for Puerto Rico, and the two Puerto Rican Physidae lots analysed in this work belong to P. acuta. The record of Baker (1930) is only based on shell morphology and could not be evaluated for lack of 234 anatomy studies. However Lem&s (1996) description is based on shell and internal anatomy, though lacking comment on presence or absence of-glandular part in penial sheath. Nevertheless, I have examined Leme's material and no glandular penial sheath was found. Thus, I identified as P. acuta sensu stricto those specimens studied and identified by Leme (1966) as P. cubensis. The descriptions of the soft anatomy of P. acuta published by Hamilton-Attwell et al. (1970) and Aboul-Ela et al. (1969) support the criteria used in this work to identify P. acuta. Of course, further studies should be made on the glandular morphology of the penial sheath of Neotropica.l Physeila species, specially on P. cubensis material, to evaluate the distinctiveness of its glandular part. To my understanding, based on this study, the occurrence of P. cubensis in South America is not confirmed, while the presence of P. acuta sensu stricto is recorded for the first time. In fact, the -finding of P. acuta in South America is not a surprise as this species has been introduced in all continents. Furthermore, the sporadic occurrence of this Physidae in South Artierica, usually associated to artificial and polluted water bodies, plus its total absence in earlier surveys in South America, suggest a recent introduction of P. acuta in the continent. The finding of P. acuta in the Federal District represents the first record for the genus Physella in Central Brazil. 235
Te (1978) emphasized the apparent absence of Pleistocene fossils of P. acuta in the Palaearctic region, and considered the possibility of P. acuta not - being native in Europe, but simply a primitive introduction from the New World within the recent historic period. The North American P. heterostropha (Say) is almost indistinguishable from P. acuta, and could in fact represent the same species (Te, 1978). The understanding of P. acuta morphological variation was obtained from shell and anatomical analysis. Though each of these studies was found to be enlightening, an association of these two approaches is necessary. The finding of bimodality of spire angle among the Neotropical material of Physella acuta sensu stricto in this work raises some questions. Should specimens belonging to PCA clusters A and B be treated as different species, or do they belong to a single variable species? The bimodal distribution can in part be discounted by the small size of the analysed samples. It is inferred that a larger sample specially with specimens from additional localities, would probably include individuals with intermediary spire angles. It is well known that P. acuta populations in the Old World and P. heterostropha populations in the New World are higly variable with regard to shell shape, specially spire height and spire angle (Hamilton-Attwell et al., 1970; Wurtz 1949),Wurtz (1949) indicated that P. heterostropha in North America is dimorphic with regard 236 to spire height. In fact, the two morphs occur either alone or together in a same population. Hamilton-Attwell et al. (1970) found a great spectrum of shell shape variation in
P . acuta of South. Africa, with populations either composed totally, or predominantly, by one morph, or presenting individuals of all the different shapes together. The population of P. acuta of London analysed in this study presented much variation in spire height just as the South African material. This general shell variability within single populations, and the lack of anatomical differences between individuals of clusters A and B, support the conclusion that the Neotropical material of both clusters (A and B) do belong to a single variable species: Physella (Costatella) acuta (Draparnaud). Thus, it is predictable that further collecting of P. acuta in the Neotropics will disclose the occurrence of specimens that represent clusters A and B together in a same
population. The first occurrence of the genus Physella in Brazil was registered by Leme (1966). He named the species P. papaveroi pointing out its differences from P. cubensis sensu Leme. P. venustula and P. squalida were the only two other Physella species known to occur in South America (Te, 1978) before this study. In the present study, P. papaveroi was found to be anatomically different from these two species, but it agrees closely with P. acuta,, based on anatomical characters. Thus; belongs to P. acuta complex. However, its distinct shell 237 with blunt spire and very large spire angle readily distinguishes P. papaveroi from P. acuta. Therefore, P. papaveroi is considered a valid species and the following new combination is proposed: Physella (Costatella) papaveroi (Leme). It has been collected only from the type locality, a man made lake in a green area in downtown. São Paulo City. The natural distribution and habitat of this species remain unknown. 238
CHAPTER 6. REFERENCES
(* Paper not read. Information obtained elsewhere.)
ABERCROMBIE, J. & C. 0. BERG, 1975. Natural. History of Thecomya limbata (Diptera:. Sciomizidae) from Brash. Proc. entomol. Soc. Wash., 77 (3): 335-368. ABOUL-ELA, I. A. & E. A. M. BEDDINY, 1969. Anatomical and functional studies on the reproductive system of Physa acuta Draparnaud, 1805.Proc. Zool. U. A. R., 3. 131-162. AGUAYO, C. G., 1938. Los Moluscos Fluviatiles Cubanos. Parte II. Sistematica. Mem. Soc. cubana Hist. nat., Havana, 12: 253-276. ALLEN, A. L., 1976. An introduction to linear regression and correlation. W. H. Freeman and Company, San Francisco, xi + 213 p. ANTON, H. E., 1839. Verzeichniss der Conchylien welch sich in der Samxnlung von Hermann Edward Anton - befinden. Halle, e Anton, XVI + 110 p. ATCHLEY,W. R. et al., 1976. Statistical properties of ratio. I. Empirical results. Systematic Zoology, 25: 137-148. AYALA, C. A., C. W. CASTLETON & E. A. MALEK, 1976. Survey of freshwater snails in the Rio Grande Valley (Bahia, Brazil). Ciência e Cultura, São Paulo, 28 (7), suplemento: 439-440. BAKER, F., 1914. The Land and Fresh-water mollusks of the Stanford expedition to Brazil. Proceedings of the Academy of natural Sciences of Philadelphia, 65: 618-672, pls. 21-27. 23.9
BAKER, F. C., 1901. The digitations of the mantle in Physa. Bull. Chicago Acad. Sci., 2: 225-228. BAKER, F. C., 1911. Lymnaeidae of North and Middle America. Chicago Academy of Sciences, Special publication n9 3, XVI + 539 p. BAKER, F. C., 1926. Nomenclatural notes on American freshwater Mollusca. Trans. Wisconsin Acad. Sci., Arts & Lett., 22: 193-205. BAKER, F.. C., 1928. Freshwater Mollusca of Wisconsin. Part 1. Gastropoda. Bull. Wisconsin geol. natur. Hist. Surv., 70: 494 p. BAKER, H. B., 1930. The mollusca collected by the University of Michigan-Williamson expedition in Venezuela. 0cc. Paps. Mus. Zool. Univ. Michigan, 210: 92 p. BAKER, H. B., 1931. (Review of Clench, W. J., 1930: Notes on Physidae with descriptions on new species) . Nautilus, 44: 106. BARATINI, L. P., 1951. Malacologia Uruguaya. S. 0. Y. P., Montevideo, 293 p. BARTH, R., 1957. A fauna do Parque Nacional do Itatiaia. Boletim do Parque Nacional do Itatiaia, Rio de Janeiro, 6, II + 149 P., 39 figs., 1 map. BAXTER, J. M., 1983. Allometric relationships of Patella vulgata 1. Shell characters at three adjacent sites at Sandwick Bay in Orkney. Journal of natural History, 17: 743-755. *BECK, H., 1837. Index Molluscorum praesentis aevi. Musei Principis Augustissimi Christiani Frederici. Hafniae, 1/2: 1-124. BEETLE, D. E., 1973. Physa acuta in Virginia. Nautilus, 87: 28. 'qv
*BENTIIEM_JUTTING, W. S. S. van, 1943. Uber eine Sarnrnlung nichtmariner Mollusken aus den mederschlagsarmen Gebiete Nordost-Brasiliens. Arch. far Hydrobiol., 39:. 458-489.
*BEQUAERT, J. C. & W. J. CLENCH, 1933. The non-marine mollusks of Yucatan pp. 525-545, P1. 68. Chapter 2u. In: Shattuck, G. C. (Ed.). The Peninsula of Yucatan. Carnegie Institute of Washington, Publication no. 431, XVII + 576 p.
BEQUAERT, J. C. & W. J. CLENCH, 1936. A second contribution to the molisucan fauna of Yucatan, pp. 61 - 75, pis. 1, .2. Chapter VIII. In: A. S. Pearse, E. P. Creaser & F. G. Hall (Eds.), The Cenotes of Yucatan, A Zoological and Hydrographic survey. Carnegie Institution of Washington. Publ. no 457, 304 p.
BIESE N., W. A., 1948. Revision de los moluscos terrestres y de agua dulce provistos de concha de Chile. III Parte. Boletim del Museo nacional de Historia Natural, Santiago, Chile, 24: 217-239.
BREDT, A. & D. A. MELLO, 1978. Notas sobre o ciclo bioligico de duas espécies de dipteros da familia Sciomyzidae. Rev, bras. Biol.,, Rio de Janeiro, 38 (4) : 767-770.
BRUGGEN, A. C. van, 1966. Physa acuta, an introduced species new to South Africa. J. Conchol., 26: 49-50.
CAILLIEZ, F. & J. P. PAGES, 1976. Introduction a l'analyse des donnes. SMASH, Paris.
CLENCH, W. J., 1930. Notes on Physidae with descriptions of new sp. 0cc. Paps. Soc. natur. Hist. Boston, 5: 301-315.
CLENCH, W. J., 1934. Physa acuta Drap. in Massachusetts. Nautilus, 48: 33. 241
CLENCH, W. J., 1936. The Physidae of the West Indies. Mem. Soc. cubana Hist. nat., Havana, 10: 335-342. CLENCH, W. J., 1939. New records for West Indian Physidae. Men. Soc. cubana Hist. nat., Havana, 13: 77-78. CLENCH, W. J. & C. G. AGUAYO, 1932. New Haitian mollusks. Proc. New Eng. Zool. Club, 13: 35-38. CORSI, A. F., 1900. Moluscos de la Republica Oriental deL Uruguay. An. Mus. Hist. nat. Montevideo, 2: 2917525, 44 figs. COUSIN, A., 1887. Faune Malacologique de la Rpub1ique de 1l'Equateur. Bull. Soc. zool. France, 12: 1871- 287. *DALL, W. H., 1870. On the genus Pompholix and its allies, with arevision of the Limnaeidae of authors. Ann. Lyc. nat. Hist. New York, 9: 333-361. DUNCAN, C. J., 1958. The anatomy and physiology of the reproductive system of the freshwater snail Physa fontinalis (1.). Proc. zool. Soc. Lond., 131: 55-84. DRAPARNAUD, J. P. R., 1801. Tableau des Mollusques terrestres et fluviatiles de la France. Chex. Bassange, Masson et Besson Libraires, Paris, 116 p. FERNANDEZ, D., 1981. Mollusca GASTEROPODA - Physidae, In: Fauna de agua Dulce de la Republica Argentina; Fundaciôn para la Educaciôn la Cincia y la Cultura (FECIC). Buenos Aires, 15(6): 84-96. FIGUEIRAS, A., 1964. La malacofauna dulceaquicola del Uruguay. Parte 1 . Comun. Soc. Malacol. Uruguay, 1 (7): 101-202. 242
FISCHER, P. & H. CROSSE; 1886. Etudes sur les mollusques terrestres et flutiatiles du Mexique et du Guatemala; In: Mission Scientifique au Mexique et dans 1' Amerique Centrale, Recherches Zoologiques, partie 7me, 2, 731 p. *FLEMING, J.,1820. Brewster's Edinburgh Encyclopaedia. XIV, p. 617. *FLEMING, J.,1828. A history of British animals. Edinburgh, XXII + 565 p. FREITAS, J. R. de, 1978. Descriço do Habitat de Biomphalaria glabrata. UFMG/ Dept9 de Biologia Geral - CNPq, Miniplan-SIP-08/107, Belo Horizonte, 64 Pi 48 tables, 34 figs. GAILLARD, M. C., 1973. Contribuci6n al conocimiento del g€nero Littoridina Souleyet, 1852 en Argentina. Ph.D. Thesis, Facultat de Ciencias naturales y museo de La Plata, La Plata, 140 p. GERMAIN, L., 1911. Contribution a la faune malacologique de 1'Afrique Equatorial. Bull. Mus. nat. d'Hist. nat., 17: 319-324. GONZALES, J. C., W. M. SANCHEZ, J. W. THOME, P. C. GONALVES & C. M. B. OLIVEIRA, 1974. Lymnaea columella, hospedeiro interrnediärio de Fasciola hepatica (Lin. 1758) no Rio Grande do Sul, Brash. Arq. Fac. Vet. Univ. Fed. Rio Grande do Sul, Porto Alegre, 2 (1): 37-40. GOULD, A. A., 1848. (Shells collected by the United States Exploring Expedition under the command of Charles Wilkes) . Proc. Boston Soc. nat. Hist., 2: 214-215. GOULD, S. J., 1966. Allometry in pleistocene land snails from Bermuda: The influence of size upon shape. Journal of Paleontology, 40(5): 1131-1141. 243
GOULD, S. J., 1981. The mismeasure of man. W. W. Norton & Company, New York, N. Y., 352 p. *GRAY, J. E., 1828. Spicilegia Zoologica, Part 1 & 2, London, 12 p., 11 pis. GUILDING, L., 1828. Observations on the zoology of the Caribbean Islands. Zool. Journ., 3: 537-544 (Plates in vol. 5). HAAS, F.,: 1939. Zur kenntnis der Binnen-Mollusken N :-Brasiliens. Senckenbergiana, 21(3-4): 254- 27 8!. *HALDEMAN, S. S. 1842. A monograph of the freshwater univalve mollusca of the United States. Physidae. E. G. Dorsey, Philad. 40 p. HAMILTON-ATTWELL, V. I., K. N. de KOCK & J. A. van EEDEN, 1970. The occurrence and distribution of P. acuta Drap. in the Rep. S. Africa. Wetensk. Bydraes Potchefstroom Univ., 26:1-11. HARRY, H. W. & B. HUBENDICK, 1964. The freshwater pulmonate mollusca of Puerto Rico. Göteborgs K. Vet. 0. Vitterh. Samh. Handi., ser. B, 9 (5): 1--77, 16 plates. HOLMBERG, E. L., 1909. Mollusca Argentina Nonnulla. Apuntes de Historia Natural, Buenos Aires, 1 (2): 27-28. HUBENDICK, B., 1951. Recent Lymnaeidae: their variation, morphology, taxonomy, nomenclature and distribution. Kungi. SvenskaVetenskapsakad. Handi., 4th ser., 3 (1): 1-223, 5 pis., 369 figs. HUBENDICK, B., 1967. Studies on Ancylidae: the Australian, Pacific and Neotropical formgroups. Acta Reg. Soc. Sci. Lit. Gothoburgensis, Zool., 1: 1-52. JAECKEL, S., 1952. Short review of the land and Freshwater molluscs of the North-East states of Brasil. Dusenia, Curitiba, 3 (1): 1-10. 244
JANSON, K., 1982. Phenotypic differentiation in Littorina saxatilis Olivi (Mollusca, prosobranchia) in a small area on the Swedish West Coast. J. moll. Stud., 43: 167-173. JENKINS, A. J., 1890. Physa acuta (Drap) in Scotland. J. Conchol., 6: 270-271. JOUSSEAUME, F., 1887. Mollusques nouveaux de la Rpub1ique de l'Equateur. Bull. Soc. zoo'l. France, Paris, 12:! 164-186. JOUSSEAUME, F., 1889. Voyage de M. Eugene Simon on Venezuela, mollusques. Mem. Soc. zool. France, 2:232-259. KUSTER, H. C., 1844. Part 17: Limnaeaceen, pp. 1-Z8, tab. 1-4, In: Martini & Chemnitz (Eds..), Systematisches Conchylien Cabinet,ed. 2, vol. 1, Nfirnberg. LEENTVAAR, P., 1967. The artificial Brokopondo Lake of the Suriname River. Its biological implications. Atas Simpôsio Biota Amazônica, CNPq, Rio de Janeiro, 3 (Limnologia): 127-140. LEME, J. L. M., 1966. Sobre a ocorr&ncia do sub-gnero Physella no Brasil, e descrição de uma nova espécie (Mollusca, Gastropoda). Papis Avulvos do Departamento de Zoologia, São Paulo, 19 (24): 269-278. *LON....DCEL, D., 1970. Life history of Lymnaea columella (Say) and its experimental infection with Fasciola hepatica (L.): J. Dep. Agric. P. Rico, 54 (2): 297-305. *LON_DANCEL, D., L. RITCHIE & J. CHIRIBOGA, 1971. Refractiveness of PIysa cubensis (Pfeiffer) and p1exa (sic) marmorata (Guilding) to Fasciola hepatica (L.). J. Agric. Univ. P. Rico, 55: 267- 270, 1 fig. -
S 245
LINNAEUS, C., 1758. Systema Naturae. Tomus I. Regnum Animale. 10th Ed., British Museum (Natural History), London. Facsirnilé.reprint (1956). 824p. *LUTZ, A., 1914. Notas dipterolôgicas. Contribuiço para o conhecimento dos primeiros estudOs-de Tabnidos brasileiros. Mem. Inst. Oswaldo Cruz, Rio de Janeiro, 6: 43-49. MARTENS, E. von, 1859. ueber einige Land - und Sfisswasser- Schnecken aus Venezuela. Malakozool. B1It., Cassel & Berlin, 6: 59-66. *MARTENS, E. von, 1873. Die Binnenmollusken Venezuela's. pp. 157-225, In: Festsch. Feier Hundertj. Best. Gessellsch. Naturforsch. Freunde Berlin. Ferd. Dürnmlers Verlagsbuchhand., Berlin. MARTENS, E. von, 1898. Land and Freshwater Mollusca. In: Godman and Salvin (Eds.), Biologia Centrali- Americana. Vol. 9, XXVIII + 706 p., Taylor & Francis, London. MATON, W. G. & T. Rackett, 1807. A descriptive catalogue to the British Testacea. Trans. Linn. Soc. (1), 8: 17-250, 6 p 1 . MAYR, E., 1976. Principles of Systematic zoology. Tata McGraw-Hill Publishing Company , New Delhi. (Reprinted). MELLO, M. L. S. & M. T. UETA, 1973. Observations by polarised light on the radula of Aplexa rnarmorata. Nautilus, 87 (3): 79-82. MONTEIRO, W. & M. L. F. DIAS, 1980. Distribuiço de moluscos pulmonados aquticos e identificação dos planorbideos da bacia do Lago Paranoá, Brasilia (Mollusca, Gastropoda). Rev, bras. Biol., Rio de Janeiro, 40 (1): 67-74. MOORE, R. C., 1964 (Ed.). Treatise on Invertebrate Paleontology. Part I. Mollusca 1. Geological Society of America and University of Kansas Press, Lawrence, XII + 1351 p. - .... . 246
*MORELET, . 1849.. Test. Noviss. I.
*MORELET, . 1851. Test. Noviss.II. MORICAND, S., 1833. Note sur queiques espces nouvelles de coquilles terrestres. Mem.'Soc.Phys. Hist. nat., Genve, 6: 537-544, p1. I. MORICAND, S., 1836. Mémoire sur. les coquilles terrestres et .fluviatiles, envoyées de Bahia par M. S. Blanchet. Mémoires de la Société de Physique et d'Histoire naturelle de.Genve, 7: 415-466, p1.11. MORICAND, S., 1837. Premier supplement au Memoire sur les coquilles terrestres et fluviatiles de la Province de Bahia, envoyees par M. Blanchet. Mem. Soc. Phys. Hist. nat. Genve, 8: 139-148, p1. III. MORICAND, S., 1838. Second supplement au Memoire sur les coquilles terrestres et fluviatiles de la Province de Bahia, envoyées par M. BLANCHFI'.. Mem. Soc. Phys. Hist. nat. Genve, 9: 43-50, p1. IV. MORICAND, S., 1839. Troisiême Supplement au Memoire sur les coquilles terrestres et fluviatiles de la Province de Bahia. Mem. Soc. Phys. Hist. nat. Genve, 10: 53-66, p1. V). MORRETES, F. L. de, 1949. Ensaio de Cat].ogo dos rnoluscos do Brasil. Arquivos do Museu Paranaense, Curitiba, 7: 5-216. NIE, N. H., HULL C. H, JENKINS J. G., STEINBRENNER K., BENT D. H.,1975. SPSS statistical package for the social sciences. Ed. .2, McGraw-Hill, 675 p. ORBIGNY, A. D.,-1835. Synopsis terrestrium et fluviatilium molluscorum, in suo per American Meridionalem itinere collectorum. Magazin de Zoologie, Paris, 5, Classe 5(61-62): 1-44. 247
ORBIGNY, A. D., 1835-1846. Voyage dans 1'Am6rique Mridiona1e. P. Bertrand, Paris & Ve. Lerrautt, Strasbourg, vol. 5 (3): Mo1iisques, XLII +758 P. (Physidae on pages 34l-342(1837)). : ORBIGNY, A. D., 1841-1853. Mollusques, pp. 1-264., 1-380, p1. 1-28 + 11 bis, In: La Sagra, R. -- de (Ed.), Histoire Physique, Politique et Naturelle de l'le de Cuba. Vols,1 and 2 plus Atlas. P. Bertrand, Paris. PAN AMERICAN HEALTH ORGANIZATION, 1968. A guide for the identification of the snail intermediate hosts of schistosomiasis in the Americas. Pubi. no. 168: ix + 122 P. PANKHURST, R. J., 1978. Biological identification. The principles and practice of identification methods in Biology. Edward Arnold, London, vi + 104 p. PAPAVERO, N.., 1971. Essays on the history of Neotropical Dipterology, with special reference to collectors (1750-1905). vol. I. Museu de Zoologia, Universidade de São Paulo, São Paulo, 216 p. PARAENSE, W. L., 1970. Planorbideos hospedeiros interme- dirios do Schistosoma mansoni,pp. 13-30, In: Cunha, A. S. (Ed.), Esquistossomose mansoni. Universidade de São Paulo, São Paulo. PARAENSE, W. L., 1972. Fauna planorbidica do Brasil, p. 213-239, In: C. S. LACAZ, R. G. BARUZZI & W. SIQUEIRA Jr. (Eds.), Introdução a geografia médica do Brasil. E. Blttcher & Univ. São Paulo, São Paulo. PARAENSE, W. L., 1975. Estado atual da sistemática dos planorbideos brasileiros. Arq. Mus. Nac., Rio de Janeiro, 55: 105-128. 248
PARAENSE, W. L., 1976. Lymnaea viatrix: a study of topotypic specimens. Rev, bras. Biol., Rio de Janeiro, 36: 419-428. PARAENSE, W. L., 1980. Drepanotrema cimex: Synonymy, variation and geographical distribution (Mollusca, Planorbidae). • Rev. brasil. Biol., 40 (1): 101-113. PARAENSE,W. L., 1981a. Gastropoda pp. 200-207, In: S. H. HURLBERT, G. RODRIQUEZ & N. D. SANTOS (Eds.), Aquatic Biota of Tropical South America, Part 2. San Diego State Univ., San Diego, California, xi + 298. p. PARAENSE,W. L., 1981b. Biomphalaria occidentalis sp. n. from South America. Mem. Inst. Oswaldo Cruz, 76(2) : 199-211. PARAENSE, W. L., 1982a. Lymnaea viatrix and Lymnaea columella in the Neotropical region: a distributional outline. Mem. Inst. Oswaldo Cruz, 77 (2): 181-188. PARENSE, W. L., 1982b. Lymnaea rupestris sp. n. from southern Brazil (Pulmonata: Lymnaeidae). Mem. Inst. Oswaldo Cruz, 77 (4): 437-443. PARAENSE, W. L., 1983a. A survey of planorbid molluscs in the amazonian region of Brazil. Mem. Inst. Oswaldo Cruz, Rio de Janeiro, 78 (3): 343-361. PAR1.ENSE, W. L., 1983b. Lymnaea columella in northern Brazil.Mem. Inst. Oswaldo Cruz, 78 (4): 477-482. PARODIZ, J. J., 1951. Métodos de conquiliometria. Physis, 20 (58): 241-248. PARODIZ, J. J., 1956. Notas sobre "Physa" (Gastr. Puim. Basommat.).Neotropica, La Plata, 2 (17): 19-21. PEARSE, S. C. 1965. The measurements of a living organism. Biometr. Prachim., 6:143-152. 249
PENNAK, R. W. 1953. Fresh-water Invertebrates of the United States. Ronald Press,.' . New York. PEREZ VIGUEIRAS, I. & A. MORENO, 1938. "Physa cubensis" (Mollusca), Un nuevo hospedeiro intermedirio de "Fasciola hepatica" (Trematoda). Mem.Soc. cubana Hist. nat., Havana, 12: 74. PFEIFFER, L., 1839. Bericht fiber die Ergebnisse meiner Reise nach Cuba im Winter 1838-1839. Archiv. ffir Naturgeschichte, Berlin, 5 (1): 346-358. *PILSBRY, H. A., 1911. Non-marine Mollusca of Patagonia. Reports Princeton Univ. Expeditions to Patagonia, 1896-1899. Princeton & Stuttgart, 3 (5): 513-687. PILSBRY, H. A., 1925. A freshwater snail, Physa zionis, living under unusual conditions. Proc. Acad. nat. Sci. Philad., 77: 325-329. PILSBRY, H. A., & W. H. RUSH, 1896. List, with notes of Land and Freswater shells collected by Dr. WM. H. RUSH in Uruguay and Argentina. Nautilus, 10: 76-81. PIMENTEL, R. A., 1979. Morphometrics, the multivariate analysis of biological data. - Kendall/Hunt. Publishing Co. Dubuque, Iowa, x + 276 p. POINTIER, J. P., 1974. Faune malacologique dulçaquicole d'fle de la Guadeloupe (Antilles Françaises). Bull. Mus. nat. Hist. flat. Paris, Zool. 159: 905-933, 16 figs., 3 pis. POINTIER, J. P., 1976. Repartition locale et biogeographie des mollusques dulçaquicoles de la Guadeloupe (Antilles françaises). Malacological Review, 9: 85-103. PRESTON, H. B., 1907. Description of new species of Land and Freshwater shells from Central and South America. Ann. Mag. natur. Hist., ser. 7, 20: 490-498. 250
RAO, M. B. and P.. V. BHAVANARAYANA, 1976. Environment and shell variation in relation to distribution of a tropical marine snail,.:Drupa tuberculata (Blainville). J. moll. Stud., 42: 235-242.
REZENDE, H. E. B., J. L. ARAUJO de B., P.A. C. GOMES, S. NUERNBERG, M.. P. NETO, G. P. OLIVEIRPJ & R. P. MELLO, 1973. Notas sobre duas espécies de Lyrnnaea Lamark, 1799, hospedeiras intermedirias de Fasciola hepatica L. no Estado do Rio de Janeiro (Mollusca, Gastropoda, Basornmatophora, Lymnaeidae). Arq. Univ. Fed. Rural, Rio de Janeiro, 3 (1): 21-23.
RICHARI5S, C. S., 1964. Studies on Puerto Rican Physidae. Pubi. Hlth. Rep.., 79: 1025-1029.
SANTOS., S. B. dos, 1981. Restos de Physidae (Mollusca, Gastropoda) da turfa de Tr&s Riachos, Municrpio de Umbuzeiro, Paraiba (Resumo). Programa e Resumos do 79 Encontro Brasileiro de Malacologia, Rio de Janeiro, p. 25. SIOLI, H., 1951. Alguns resultados e problemas da limnologia amazônica. Bol. Teen. Inst. Agr. Norte, Belém, 24: 1-4.4.
SIOLI, H., 1953a. Limnologische Untersuchungen und Betrachtungen zur erstmaligen Endtdeckung endemischer schistosomiasis (Sch. rnansoni) im Amazonasgebiet. Arch. Hydrobiologie, 48: 1-23.
SIOLI, H., 1953b. Schistosomiasis and limnology in the Amazon Region. Amer. J. Trop. Med. & Hyg., 2 (4): 700-7.07.
SIOLI, H., 1956a. 0 Rio Arapiuns. Estudo limnológico de urn corpo d'água da regiäo do terciário, plioceno, série das barreiras, do Baixo Amazonas. Bol. Teen. Inst. Agr. Norte, Belem, 32: 1-115.
SIOLI, H., 1956b. As 5guas da regio do Alto Rio Negro. Bol. Teen. Inst. Agr. Norte, Belem, 32: 117-164. 251
SNEATH, P. H. A. &'R. R. SOKAL., 1973. Numerical Taxonomy. Freeman, San Francisco, 573 p.
*SOWERBY, G. B 1 1822. The genera of recent and fossil shells. London, 2 vols., 267 pis.
SOWERBY, G. B., 1873-4. Monograph of the genus Physa. In: L. A. Reeve (Ed.), Conchologia Iconica, Vol. 19 (318, 319): 26 p., 12 pls., Reeve, Benham & Reeve, London.
SPIX, J. B. de & J. A. WAGNER, 1827. Testacea fluviatilia quae in itinere per Brasiliam annis 1817-1820 jussu et auspiciis Maximiliani Josephi I. Bavariae Regis Aug ustissimi. F. P. de Schrank & C. F. P. de Martius (Eds.), Monachii, Typis C. Wolf, Munich et Leipzig, V + 36 p, 29 plates.
STAROBOGATOV, Y. I., 1967. on the systematization of freshwater pulmonate molluscs. Trudy zool. Inst. Leningr.., 42: 280-304. (In Russian)
STAROBOGATOV, Y. I. & E. A. STRELETZKAJA, 1967. Composition and zoogeographical characteristics of freshwater malacofauna of Northern part of Far East. Trudy zool. Inst. Leningr., 42: 221-268. (In Russian )
TANTALEAN, W. M., F. A. HUIZA & G. R. CAPU1AY, 1974. Los hospedeiros intermedirios de Fasciola hepatica en el Peru. 1. Estudio de la infeccion natural y experimental de Lymnaea viator, Lymnaea diaphana y Physa venustula. Biota, Lima, 10 (81): 243-250.
TE, G. A., 1973. A brief review of the systematics of the family Physidae. Malacol. Rev. 6 (1): 61.
TE, G. A., 1974. Studies on Physidae (Pulmonata) Basornmatophora). 1. Penial complex morphological groupings. Malacological Rev., 7 (1): 43-44. 252
TE, G. A., 1975. Michigan Physidae, with systematic notes on Physella and Physodon (Basommatophora: Pulmonata). Malacol. Rev., 8 (1-2): 7-30. TE, G. A., 1978. A systematic study of the family Physidae (Basommatophora: Pulmonata). Ph.D. Thesis, The University of Michigan, xii + 324 p.,University Microfilms International, Ann Arbor. TE, G. A., 1980. New classification system for the family Physidae (Pulmonata: Basommatophora).Arch. Moll., 4/6: 179-184. UETA, M. T., 1976. Aspectos morfomtricos e biolôgicos de Lyrnnaea columella Say, 1817 (Gastropoda, Pulmonata). Ph.D. Thesis, Universidade Estadual de Campinas, Campinas, 190 P. VAZ, J. F., 1979. Apontamentos sobre a anatomia macroscô pica do sistema nervoso de Aplexa marrnorata (Guilding, 1828) (Basommatophora: Physidae). Anais do 59 Encontro dos Malacologistas brasileiros, Mossoró, RN, 14-17/julho/1977., Publicaç5es Avulsas da Fundação Zoobotânica do Rio Grande do Sul, Porto Alegre, 4: 55-67. WHITE, J. S. and S. J. GOULD, 1965. Interpretation of the coefficient in the allometric equation. American naturalist, 99: 5-18. WURTZ, C. B., 1949. Physa heterostropha (Say). Nautilus, 63: 20-33. *WURTZ, C. B., 1951. Catalogue of Ancylidae of South and Central America and the West Indies, with description of a new species. Nautilus, 64 (4): 123-131, p1. 4, figs. 8-8a. *ZILCH, A.,1956. Nornenklatorische Bemerkungen. Arch. Molluskenk., 85: 85. 253
ZILCH, A., 1959-1960. Gastropoda. Part 2. Euthyneura. In: Schindeiwoif, Handbuch der Paleozoologie, vol. 6, Borntraeger, Berlin. Lief. 1-4, 834 p.