133845.pdf

(Tuis.) #1
TRILOBITE BIOGEOGRAPHY OF GONDWANA 59

(14) Anterior furrow (furrow running from
anterior margin to distal portion of
preglabellar furrow, forming lateral
boundary to preglabellar area, sensu
Whittard 1960; see Fig. 2): 0 = absent, 1 =
incipient, only expressed anterolaterally
('fixigenal inflation' or 'fixigenal overhang'
condition), 2 = extends posteriorly to
preglabellar furrow (may superficially
appear to be a posteriorly positioned
anterior border furrow if true anterior
border furrow is absent, e.g. Dean &
Martin 1978, pl. 4, figs 1, 4,7, pl. 5, fig. 1).
(15) Anterior border furrow: 0 = absent or
weakly defined, 1 = strongly defined,
(16) Anterior border (dorsal view): 0 = trans-
versely straight to gently anteriorly
convex, 1 = strongly convex and pointed,
possibly forming distinct anterior 'pre-
cranidiaP projection.
(17) Surface tubercles (size and density): 0 = no
tuberculation, 1 = light tuberculation, 2 =
heavy tuberculation.
(18) Number of pygidial axial rings: 0 = up to
and including five, 1 = six or seven, 2 =
eight or more (the majority of Neseuretus
species have five to seven axial rings, and
several are described as having 'six or
seven', although the number of axial rings
varies from two to ten between different
taxa included in the analysis).
(19) Dorsal outline of terminal piece: 0 = con-
tinuous with anterior part of pygidial axis,
1 = anterior region laterally expanded
('swollen').
(20) Lateral outline of terminal piece: 0 = con-
tinuous with anterior part of pygidial axis,
1 = inflated.
(21) Postaxial pygidial morphology: 0 = weakly
defined terminal piece, with narrower
postaxial ridge extending to margin
('Neseuretus condition'), 1 = posterior
margin of terminal piece defined by dis-
tinct posterior border furrow, with no
postaxial ridge present, 2 = posterior
border furrow absent, narrow terminal
piece fused with prominent postaxial ridge
to form a continuous elongate structure
which extends posteriorly to the pygidial
margin ('SarrabesialVietnamia condition').
(22) Pygidial axis anterior to terminal piece: 0 =
evenly tapered, 1 = funnel-shaped (taper-
ing to a point slightly anterior to the
terminal piece, posterior to which the axial
furrows become parallel), 2 = swollen
(axial furrows abaxially convex anterior to
a parallel-sided posterior section).
(23) Pygidial lateral pleural furrow/vincular


furrow ( = 'cincture' of Dean & Zhou
1988): 0 = absent, 1 = present.
(24) Pygidial interpleural furrows: 0 = absent or
only present very close to the pygidial
margin, 1 = clearly defined across much of
pleural fields.
(25) Pygidial rib morphology: 0 = flattened, 1 =
ridge-like.

Cladistic analysis was conducted on
PAUP*4.0b4a (Swofford 2001), using a heuristic
search with 50 random addition sequence repli-
cates. Of the eight multistate characters, four (5,
7,21 and 22) were treated as unordered and four
(1, 14, 17 and 18) as ordered. Multistate taxa
were coded as polymorphisms. All characters
were unweighted. Searching was performed
twice, once using accelerated transformation
optimization (ACCTRAN) and a second time
using delayed transformation optimization
(DELTRAN).
Within these parameters, four minimal length
cladograms of 106 steps, occupying a single
island, were recovered (Consistency Index =
0.4151, Retention Index = 0.5867). ACCTRAN
and DELTRAN optimization resulted in identi-
cal sets of cladograms, as did subsequent
reweighting of the dataset. The strict consensus
of these cladograms is shown in Figure 3a (a
semistrict consensus has the same topology).
Support for individual nodes within the con-
sensus cladogram was assessed with bootstrap
analysis (Felsenstein 1985) using PAUP*4.0b4a;
bootstrap values are displayed on the cladogram
in Figure 3a. Removing the three taxa with five
or more missing values (Neseuretus attenuatus,
N. chaschuilensis and N. sexangulus} from the
character matrix and rerunning the analysis does
not improve resolution in the Neseuretus sub-
clade, recovering five different cladograms of
103 steps (Consistency Index = 0.4272, Reten-
tion Index = 0.5845) (Fig. 3b), and so the
consensus cladogram in Figure 3a is here used as
the basis for further analysis.
The consensus cladogram will provide the
basis for a systematic revision of the Reedocaly-
meninae in a separate study, and it is used here
as the basis for biogeographic analysis. The
reliability of the consensus cladogram obviously
underpins both any taxonomic or biogeographic
assumptions based on it. As several of the nodes
are only weakly supported by bootstrap analysis,
some caution should be employed when
evaluating biogeographic ideas based on the
topology of the resultant area cladogram. A
qualitative but straightforward method of
investigating the validity of the cladogram is a
consideration of the degree of congruence
Free download pdf