Considerations on Middle and Upper Devonian Thamnophyllidae soshkina in Poland

A C T A Vo l. II

P A L A E O N T 0 L_0 G I C A 19 5 7

MARIA ROZKOWSKA

P O L O N I CA No ••2-3

CONSIDERATIONS ON MIDDLE AND UPPER DEVONIAN THAMNOPHYLLIDAE SOSHKINA IN POLAND

PART 11*

Abstract.- Cora ls withhor se-shoe di ssepirnerrtsand resting on themtrabecular fans, formerly inc luded 'by the writer in [the subfamily Pachyphyllinae Stumm, are now referred to the family Tha m n ophyllidae "Sos hkina. The systematic divis io n of this family is given. An analysi s of correlation between the number of septa and the diameter of calice or of tabularium (n/d or nit) in gene r a Th am no p h yH u m Penecke, Macgeea We bste r, PexiphyHum Wa lther an d PachyphyHum Edw a r d & Haime has mad e possible the determin at ion of the range of in dividual variabl lity and the elucidation of evolut ionary tendencies within the quoted genera. The coefficient of correlation for particu lar spe cies is of diagnostic value an d ma y constitute their stratigraphical index. It has been establish ed that sp ecies prev ious ly de scr ib ed and referred (R ozkowska, 1953, 1956) to the gen er a-Pseudoacerv ularia Schlu ter and Pach y phy l lum Edwards & Haim e should be assigned to genus Pac h yp h y llu m. Withi n this genus four groups of species , with pecul iar, similar coefficients of correlation, shou ld be differe n tiated. Conclusions concerning speciation have 'been ded uced fr om biometrical treatment and from on togenic and morph ological st udies.

A descrip tion is given of the phyletic and chro no logicalevolution of Tha m n ophy lli-

~ae. In addition to material from the Holy Cross Mts. sp eci m ens collected at , Debn jk and in the Sud e ten'w e r e used by thewrtter for the present work.

INTRODUCTION

The taxonomy of Pal eozoic cor als has not, thus far, been adequately establis he d. Differ ent sys te matics are adap te d by various aut h ors , based on their own taxonomic criteria. The morphology of corals, however, ought not to be regarded as the only adequate taxonomic criterion, since forms belongin g to dis tant phyl etic lines may , in consequence of convergen ce, pr odu ce similarities-of st ru ctu r e· Microstructure sh ould be mad e the .main basis her e as it is closel y connected with the mech an ism

*-Par t I: "Pachyp hyllina e from the' Middle Dev onian of,the Holy Cross Mts .':

was pu blish edjn 1956, :y.o!:..I,No. 4 _of4ct~ palaeont .Pol. -: '_. . .. .

82 MARIA RO:l:KOWSKA

of excretion by the polyp of exoskele ta l parts (A. Sch c up pe , 1956). Cor al- lites com m only know n under the name of Pachyp hy ll inae hav e a most peculiar microstructure, by E. D. Soshk in a (1949) called hex acoralloid, since in these forms, as in Hexa cor all a , the trabeculae have a fan-like arrangement. Hors e shoe dis sep im en ts bein g the bas e of th eir line of divergence, the distal edge of septa is arched, with strongest con v exit y above the horse-shoe line. Th e arrangem ent of trabecular granu le s on septa resembles that in Hex a cor a ll a as the granules are parallel to the distaledge of septu m.Septa protrude bey on d the pseu dotheca as "costae " . From the outside they are mask ed by a thi n epitheca, which, howev e r , does not reach to th e edge of calyx, but ter minates somewh a t lower.

The body of the pol yp mus t have, therefore, been over hanging the cal y x on the outside. Simple an d colo n ia l forms are always pr ovided with a talon for attach me n t. This is a feature common in the Te tracorals. In Protomacgeea dobruchnensi s Rozk. which is a small and primitive form, the wr iter has nev erth el ess obser ve d the presence of a pedicil so com mon in He x aco r alla . And yet Thamnoph yllidae are typ ica l Tetraco r als whose ontogeny is chara cte r is ed by numerical increase of septa and by bilater al sym metry pecu liar to this gr oup . This is ma nkedly strong in some genera, particula rly so in youthfu l sta ges when the cardina l septum is .rema r ka b ly long or short, while the fossula has a pr om in en t outlin e .

The hexacoralloi d structure aut h orises the establish ment for this tetracoral grou p of a separa te systema ti c unit disti n guished by ex ce ptional progressiveness. Mos t like ly it is not an acci dental occu r r e nc e, but ma y possibly suggest a latent evol utionary potentiali ty in this grou p of Te traco r als. The wr it er believes this fact to su pp ort the hypothesi s that Hexacor al s have a direct ge netic connect ion with Tetr acoral s. .

This morphology , quite unusual for Tet racor als, has attr a ct ed the attention of a nu mb e r of stu de n ts who have stressed its peculiar features without , however, bringing them into the systematics of this group . It is only during th e last fe w yea rs that some authors have expressed th e opinion that this grou p ought to be plac ed in a separate systemati c uni t- Sosh ki n a (1941) united forms ha v in g a hexacor all oid struct ure an d horse-s h oe dissepiment s into the famil y of Thamno phyllidae which included the genera Thamnophyllum Penec k e, Macgeea Webst er, Synaptophyllum Simps on and Pachyphyllum E. & H. In 1954 this wr it er took a step furth er in establishing a "grou p" of Tha m n op h y lli da Soshkin a with one family, Tha mnop h y llidae Soshki n a, only. The exceptional position occupied by this group of species is also stressed by Sch ou ppe¹

1 Prof.Dr A. v.Schouppe of Munste r wrote to the aut hor in 1956that his studies 'In the genotype of Phillip sastraea hen nahi Lonsdale enabled him to ascertain the

DEVONIAN THAMNOPHYLLIDAE IN POLAND

(1956) who st ro ngly advises its sep a r ation as a distinct sys t em a ti c unit in oppos it ion to other Tetracoral gr ou ps .

Two groups may be differen ti ated am ong for ms having more or less sym met r ical trab ecular fan s:'

1) forms without hors e-shoe disse pime n ts, e.g. Cera tophy llum Gtirich , Haplotheci a Frech , and

2) forms wit h hor se-shoe diss epiments, e.g. Thamnophyllum Penecke, Synapt ophy llu m Simpson , Macg eea Webster , Pachyphyllum Edwards &

Haim e, Pexiphyllum Walther . In this grou p are, thus, assemb le d the ge- nera which the pres ent writer has so far in clud ed into the subf am ily Pachy phyll inae Stumm, wh ile Sos hkina an d Schouppe have referred them to the family Thamnophyllidae Soshkina. The name of the sub- family Pachyphyllinae Stumm, 1949, used by the present writer in her papers published in 1953 an d 1956, th us becomes obsolete. The family Tha m n ophylli dae Soshkina intowhich the writer now includes all genera wit h sym me t r ically arranged tra becula r fans an d rows of horse-shoe disse pime nts, is here divided into two subfa m ilies, viz. Thamnophyllina e mih i an d Mac geeinae mihi. The sub fa m ily 'I'hamnophyllinae'essentially inclu des colonial forms displaying the pattern of phacelloid or massive colon ies with indistinct bilateral sy m metr y. The subfamily Macgeeinae repr esen ts sim ple forms. Budding is here ex t re mely rare and there is no "m assi ve colony" stage, while bilateral sy m metry is distinct through out the ontogen y. Each of thes e subfamilies links several mutu all y rel ated species. The classification of the family Thamnophyllidae is the r efore as follows:

Family Thamnophyllidae Soshkina , 1941

Su bfam iily Thamnophyllinae mihi

Genus Thamnophyllum Penecke,1894

" Synaptophyllum Simpson,1900, emend. Rozkowska,1953

" Pachyphyllum Edwards & Haime, 1850.

Subfamily Macgeeinae mihi

Genus Protomacgeea Rozkowska,1956

" MacgeeaWebster,1889

" Pexiphyllum Walther, 1928, em end . mihi.

Out of the above quoted genera, Trapezophyllum Etheridge, 1899, is the only one thus far never recorded in Poland. It has been described

presence in this species of horse-shoe dissepiments, Hence, PhiHipsastraea d'Orbi- gny and PachyphyHum 'Edw ar ds & Haime 'Would apparentily be synonymous.

Sincere thanks are due to Professor Schouppe for this interesting information.

84 MARIA RozKOWSK A

from the Lower Devonian and Eifelian of Australia and recently from the Eifelian of the Rhine Province (A. Glinski, 1955).

In the text and explanationsof tables an dfigures the following symb ols ar e used :c - class,n - number of sep ta,d' - diamet er of calyx , Mn - arithmetic mean of septa in classes, Md - ar ith me tic mean of diameters in classes, Mc - correlation coefficien t of classes, Ms - correlation coefficient of species_,

*

The author wishes to ex p ress her heartiest thanks to Pr ofessor Dr.

Roman Kozlowski of the Warsaw University for cons t ructive crit icism an d advice given on the manuscript and for helpful discu ssio n of th e general problems involved.Speci al thanks are tendered to Mrs. J. Gr usz- czynska for the executi on of the gr aphs and dr awin gs from photog r aph s of the thin sect ions, an d to Mrs.J. Humnic k a for the pains taken in doing th e English translation of the pres ent paper.

VARIATIONS OF SP E C I F I C CHA R A CTERS WI T H IN GENUS THAMNOPHYLL U M PENE C K E

The vari ability of corals belonging to the Thamnophyllidae is str ong but fits in to the stru ct ural sche me ch ara cteristic of the parti cul ar

B·

Fig. 1. - Thamn ophy llu m kunthi (Dames), topotyp e.

Mok r zesz6w, Upper Fr asnian. Longitu d inal se cti on: A of specimen with closely spaced complete. tabula e, B of specimen with wide ly sp aced an d incomplete tabula e.

syste ma tic unit. Stron ges t var ia bi li ty is obse rved in the pattern of tabulae as is clearly sh own in two spe cime ns of Tham n oph yUtim kun thi (Dam es) from..theUpperF rasn ian in Mokrzeszow, figured in fig. 1 A, B. Closel y

DEVONIAN THAM NOPHY LL I D AE IN POL AND 85

alig ned, parallel, complete tabulaeoccur her e alo ng with dist an t, irregula r an d incomplet e ones. Dissepimen ts and septa ar e less variable, while microstruc tur e is the most cons ta n t feature. Nu m ber of septa in relation to the corresponding diameter of calyx (n/d ra tio) is a diagnos tic specific an d subspe cific featu r e. Ev e r y cor r el ation, how eve r , is subject to.

flu ctuation and changes with the age'of the individual. This can be distinc tly see n in va r iatio n cu rves plotted on measurements of individuals belo ng ing to one population. It is difficu lt to asce r t ai n correlation chan ges dependent on ont og e-nicst ages by stu dy in g them on a single fossil corallite on ly.Hence the n/dcorrelation da ta must be based on individuals of whole populations in diffe r ent on tog enic stage s. The mean values here will approach th ose of increas e cor re la t ion ofone in dividual.The term"increase corr elation" was introduced by H. Klahn (1920, p. 25-2 6), whos e methods wer e'given the pref er ence by the present writer in her statistical treatment Thevariability of foss il corals by mea ns,of biometrical analysis has sofa r beenstudie dby R. Richter (1916) in thecas e of C,alceolasan dalina L. and by K. G. Voynovsk iy-Kr iger (1956). The latter author, when in quiring int o the ontoge n y.of Pal eozoic corals, has also 't a k en account of both the n/d and aid cor relations, viz. the rela tion ~f ^{the dist}ance:'between septa to the diameter of cal yx .

Variations of quantit ative characters in ~hamnophyllum kozlow skii R6zk.

In corals the sou n de st basis for biometr ical analysis'of va riability is provi ded by the corr el ation between number of sep t a 'a n d diameter of calyx (n/d) sin ce thes e values are most susceptib le tQ.!lu m erica l represen- tation . For this purpose aux ili a ry tables (table 1) have been drawn up, su b div id in g the diameters of calices in to classes of 1 mm distance ea ch, measu red with accuracy of 0.1 mm. Such a table enables the Mc correla tion coeffici ent of classes to be calculated on the relation of the arith m e t ical mean of the number of sep ta an d the diameter of calyx with res pect to every class .The mean class coefficient of Mc1-Mcx correlation constit utes a constant and diagnostic value. It is the correlation coefficient of species, viz.Ms.

The first correlation stu died by the present writer was that of mentioned features in Thamnophyllum kozlowskii. The correlation for th e population from Sitk6wka and that for the Wietrznia specimens - both fr om the Lower Frasnian - were caLculated independently.

Table 1 specif ied measurements for 97 Wietrznia specimens.Th_e range of diameters varies from 5 to 15 mm. There are from 16 to 29 major septa.

The gr ea test number of variants is found in class Md6·The mean number

86 MARIA ROzKOWSKA

of septa (Mn6) is 23. The mean values of septa are in extrabold. The line which joins these mean Mn1-l0 is oblique and nearly rectilinear thus resembling the line which represents correlation of ontogenic increase (Klahn, 1920).

Ta bl e I

Correlation between number of septa and diameter of calyx in Thamnophyllum kozlowskii Rozk, from the Lower Frasnian in Wietrznia

~

Numhcr 01 5.5 6.5 7.5 8.5 9.5 10.5 11.5 12.5 13.5 14.5

major septa

16 2 ³

17 1 2

18 1 2 1 2

19 2 ^{4 1}

20 2 2 1 1

21 1 6 4 3,

22 5 6

23 3 3 ³

24 1 1 5 4 1

25 4 3 3 ²

26 2 1 3 1

27 1 1

28 2

29 1

- - - - - -^{- - - -- -- -- -- -}

Total: 4 7 6 15 15 24 11 7 6 2

97 specimens

Numerical data resulting from table 1 are as follows (table 3): row 1 - specifies data for classes c1-clO; row 2 - those of the relation between the mean number of septa and the mean diameter of calyx for every class;

row 3 - results of calculated correlation constituting the class coefficient of correlation Mel-MclO. Finally the correlation coefficient of species Ms for the Wietrznia Thamnophyllum kozlowskii is given.

In the case of Th. kozlowskii from Sitk6wka, represented by 148 spe- cimens, correlation table 2 shows similar values. Range of diameters is from 5 to 16,while that of the number of septa is slightly higher being 14 to 32. The maximum number of frequents is noted in class 6, where the mean number of septa is 24. This figure exceeds by one the mean number

DEVONI AN THAMNOPHYLLIDAE IN POLAND 87

Tab l e 2

Cor r e la t ion betw een number of septa and diam eter of calyx in Tham n ophy Uu m koz- low sk ii Rozk . from th e Low er Fra sni an in Sitk6wka

~

Num be rof 5.5 6.5 7.5 8.5 9.5 10.5 11.5 12.5 13.5 14.5 15.5 majorsepta

14

I

I I

15 1

16

17 1 1 1

18 1

19 2 4 1

20 1 5 3 1

21 1 5 1

22 1 5 7 5 1

23 1 3 3 7 1

24 4 4 9 7

25 1 3 4 2

26 5 7 6 4 1

27 2 1 3 1

28 3 3 1 1 2

29 1 1 1

30 1 1 1

31 1

32 1

I I

Total: I ³ I ⁷ 1¹³ I²¹ I²¹ I ³⁷ I ²² I ¹⁰

148 specimens

5 5 4

Table 3

Numerical data of class correlation coefficients for Thamnophyllum kozl ow sk ii Rozk , fr om the Lower Fra snian in Wietrznia

Classes I c1 I c2 I c3 I c4 I c5 I c6 I c7 I c8 I c9 I c10

I

~I~

Mn/Md ^{- - - -} - ^{- -} - - - -

5.5 6.5 7,5 8.5 9.5 10.5 11.5 12.5 13.5 14.5 - - - - - - - - -^{- - -}- --^{- - - -} - - - - - _\

Mel-MclO 2.9 2.6 2.5 2.4 2.3 2.2 2 2 1.9 1.8

Ms kozl owskii, Wie trznia 2.28

88 ^{MARIA R6zKOWSKA}

of major septa in Wiet r zn ia specimens. Correl a tion coefficients calculated for the different classes on table 2 are as follows:

Table 4

Numerical data of class correlation coefficients for Thamnophyllum kozlowskii Rozk. from the Lower Frasnian in Sitkowka

Classes ._1 c l

I ^;

15 17 20 22 23 24 25 26 27 ~ ³⁰

MnjMd - - - - - - - - - - - - ^{- -}

- -

5.5 6.5 7.5 8.5 9.5 10.5 11.5 12.5 13.5 14.5 15.5

- -^{- - - -} - - ^{- - - - - -- - - -}

Mc1-M cll 2.7 2.6 2.6 2.6 2.4 2.3 2.2 2.1 ^! 2 1.9, 1.9

Ms kozlowskii, Sitkowka = 2.30-

Th es e da ta sh ow th e correlation coefficient of species for specimens of

( )' .

Th. kozlowskiifrom theloca litiesof Wieti'znia'and Sitk6w ka tobe virtually ide ntical th ough the specim ens bel on g to somewhat different facies, namely:

Ms Th. kozlowskii (Wietrzn ia) from a near-reef facies , - 2.23 Th. kozlowskii (Sitk6wka) from a reef built by Ta b u la t a anti Tetracoralla" = 2.3(\

Graphs of correlation coefficients for specimens of Thamnoph y lLum kozlowskii collected.from these two localities are plotted in fig. 2. The cu r ves}·.tm very ~close, are of similar chara ct e r and intersect in young stages. The difference in distance of both lines, equivale nt to approx. one majorsep tum, suggests that the Wietrznia specime ns may be of a sligh tly you nger geological age as will be seen from fig . 3. Th e vari an ts converge in the proximity of th e line of correlation , but the ir var iation s range is different. The course of graphs and the arrangement of-variants indicate th at both assemblages are conspecific, but that most like ly they ar e referable to somewhat different stratigraphical zones.

Variations.of qiuuituative characters in othe r species of genus Thamnophyllum Penecke

The f()llowing.se:ven species and subspecies of. genus Thamnophyllum have been recorded (1953, 1956) by the writer from Devonian beds of Poland:

DEVONIAN TH AM NO P H YLL IDAE IN POLAND 89

...

16.5 a

c

14.5

o.

c

12.5

/1.koZr~wskii, ^Wietrznia l1.koiJowshii . SithOwka

o 0 co

10.5

00

o

17 ^.0

0

15

0 00

!

6.5 8.5

23

'9 27

25

2f 29 .11

Fig. 2. ThamnophyHum kozlowskii Rozk , Gr aph of cor r ela ti on cu rves n/d (re lation of number of ma jorsep ta (1) to diameter of calices, in mm). Continu ous line - for Lower Fra snian specimens fro m Sitk6wka; va r ian ts marked by empty circles. Interrupted line - for Wietrzn ia speci mens:

variants mark ed by dots. Numerical data given in tabl es 3 and 4.

"

"

Up per Fr asn ian:

Midd le Lower Giv etian :

Th. kunthi (Dames)

Th.kozlowskiisuperiusR6zk.² Th. kozlowskiiRazk.

Th. caespitosum (Goldfuss) Th. trigemme pajchelae Rozk.

Th. trigemme (Quenstedt) Th. skalense Rozk,

2 The Middle Frasnian form from th e Holy Cross Mts., in 1953 referred by this writer.to Th. monozon at u m Soshkina, is not identical with the Ural species, since it does not displa y lateral budding; its variation curve is practically the same as tha t of the ancestr al form Th. kozl owski i. The pres ent writer regards it, therefore, as a mutati on of the latter species .

90 MARIA RozKOWSKA

Results of calculations and measurements ar e specif ie d in table 5, separately for everyone of thes e forms, as above for Th. kozlowskii.

The following conclusions may be drawn on these numerical data (fig. 3):

nl

Th.•r.a/en. e ~

In.triqemme - - - - -- Ttitr.pajchelae ~---­

Th.caespitosum 000000000000

Ib.k.ozTowshii

Iti.k.supe riu s _~---­

Th.hunth i -rT,.,.,. ...

29

27

25

23

21

19

17

/5

13

6.5 8.5 10.5 12.5 14.5 a

Fig. 3. - Graph of corre la tio n curve s nJd of some Tham 1Wphyllu m species from the Middle an d Upper Dev onian of Poland. Number s of major sep ta - on ordinate, diameters of ca lices - on abscissa , in

mm. Numerical dat a given in table 5.

a) The range of variation in number of septa is limited, there bein g 12 to 30 major septa. Sp ecies with greatest range of variation are most plastic, namely the Givetia n Th. trigemme and the Frasnian Th . kozlowskii. In dwarfed species this range is very small (Th. tri gemme pajchelae, Th. kunthi);

DEVONIAN THAMNOPHYLLIDAE IN POLAND 91

b) Class coeff icients of cor relation are higher in Givetian for ms, being markedly lower in the Frasni an specimens as shown here below:

Fr asn ian: Th . kozlowsk ii Givetian : Th. trigem m e

Th. ska lense

2.8 - 1.9 5.0-2.7 4.8- 3.0

Upon comparing the various n/d correlations it is to be noted that during ontogenic evolution an d phylogeny of Thamnophyllum the number of septa increases at a smaller ra te th an the diameter of calices (negative allometry). In consequence, th e proportions in classes ar e reduced. Negative growth all om et ry in rela tion to number of septa is thus an evolutionary trend in gen us Thamnophyllum and a characteristic generic feature;

c) The above conclusion is also confirmed by data of correlation coefficients for the different species accor di ng to their stratigraphical age. Sp ecim ens of these species hav e namely been collected from a sequ en ce of layers:

Ms Frasnian : Th. kozlows k ii superiu s 2.20

Th.kozlo w sk ii(fromSitk6wka) 2.30 Givetian : Th. caespitosum 3.17

Th. trigemm e 3.42

Th. skale n se 3.64

The above quoted values may , the re fore, be a useful stra tig r aph ical index since the correlation coefficient of Givetian species has a value exceeding three, while that of the Frasnian species is only slightly higher than two. It is easy to note that the correlation value decreases in species of Thamnophyllum occurring in successiv e Givetian and Frasnian layers. Hence the inference that the value of correlation n/d is foremost a function of time;

d) The dwarfed forms Th-trigemme pajchelae and Th. kunthi have not been included among those specified ab ove in paragraph c). The subspecies Th. trig emme pajch elae , alth ough one of the youn gest Givetian representatives of genus Thamnophyllum and pene- contemporaneous with Th. tri gem m e, shows a high correlation coefficient (3.60) nearest to th at of Th. sk alense, the oldest Middle Givetian form (Ms = 3.64). Similarly , Th. kunthi, the you ng est species from the Upper Frasnian of Mokrzesz6w, also displays a high correlation coefficient (Ms = 3.70). Both these dwarfed forms which hav e proba bly lived under unfavourable envi r on men tal conditions thus show a certain regression in relation to Th. sk alen se, a more primitive form;

Ta bl e 5

Constant biometrie al data in respect to va r iousspeci es of ThamnophyUum a. Thamnophyllum sk a lense- 67 measured specimens

<!:J

""

Classes e1 e2 e3 e4 e5 e6 e7

12 14 16.7 20 21.6 23.4 24.6

Mn/Md - - - - ^{- - - - - - - -} - -

2.5 3.5 4.5 5.5 6.5 7.5 8.5

Mel-Me7 4 3:7 ^', ,

3.6 3.3

4.8 3.1 3

Number of specimens

in per cent figures 2 8 18 34 16 10 12

, , , , ,

16 18 19.6 31.4 23.2 24.3 26.5 28

Mn/Md - - - - - - ^{- - - -} - - ^{- - - -}

3.5 4.5 5.5 6.5 7.5 8.5 9.5 10.5

Me1-Me8 5 4 3.6 3.3 3 3 2.8 2.7

Number of specimens

in per cent figures 1 3 6 16 36 21 9 7

Msskalense= 3.64 b. , Thamnophyllum tr i gem me - 109 measu red specime ns

Class es e1 e2 e3 c4 e5 c6 e7 c8

::.::»

;0...

:»

;0o

N

o~

~:s

:»

e. Thamnophyllum eaespitosum

Ms tri gemme= 3.42 15 measur ed specimens

Classe s e1 c2 e3 e4 e5 e6

I 14 I 17.3 20 I _22.5

Mn/Md - - - ^{- - - - - - -}

3.5 5.5 6.5 8.5

Mel-Me6 4

I (3.5) 3.1 3 (2.8) 2.6

Ms caespi tosum=^3.17