Higher Systematics of the Pentatomoidea 139
2.3.1 Size of Project
Within the Pentatomidae, we have recognized a total of ten subfamilies, two of which have two tribes,
another has four, and another has five. Within the subfamily Pentatominae, we have recognized 42 tribes,
another twelve Gross genus groups and eight Linnavuori genus groups that do not fit cleanly within
already established tribes. This roughly adds up to 80-85 groups that will need to be studied. A compre-
hensive analysis would include at least two or more taxa from each of the above groups, so we are now
looking at a minimum of 160 taxa in the analysis of only the Pentatomidae. Complicating matters is the
fact that some of the groups contain only a single (or few) species, and these are represented only by a
few specimens in collections.
2.3.2 Interpretation of Characters
It is obvious that several important characters that have been used in past analyses have evolved multiple
times. Additionally, various characters, including those that have arisen multiple times, have been sec-
ondarily lost. A good example of this is the armature of the base of the abdomen, which is a character
that often has been used in establishing classifications. In the early 1980s, this character was used to
segregate the majority of New World genera into three groups: (1) those lacking a spine or tubercle,
(2) those having a spine or tubercle with the apex free, and (3) those with a spine or tubercle that was
met in apposition with an elevated metasternum (Rolston et al. 1980; Rolston and McDonald 1981, 1984).
For practical purposes, this classification worked well and allowed for further keys to be developed, thus
aiding mainly in the identification of the New World genera. It did not, however, reflect phylogeny, espe-
cially once we began to integrate this with Old World classifications. It became obvious that the abdomi-
nal armature had evolved a number of times and when certain groups were separated off using other
characters, those that remained began to make more sense. Similarly, it appears that the development of
the structures associated with the metathoracic scent gland involves similar problems. The length and
shape of the ostiolar rugae often has been used in helping develop classifications. Initial data seem to
indicate that a short, auriculate ostiolar ruga may be considered primitive, and a longer, groove-shaped,
apically acuminate or disc-shaped ruga may be more advanced (Schaefer 1972, Linnavuori 1982). But
it also appears that this does not always support phylogenetic relationships (see Kment and Vilímová
2010a). That is, the rugae appear to have lengthened multiple times and then, in some cases, may have
shortened. This makes it extremely difficult to interpret this type of character. We also believe we need
to be careful in interpreting the structure of the thoracic sternal characters, and characters associated
with the internal male genitalia. Again, certain aspects of these characters also appear to have evolved
multiple times or have been lost secondarily.
2.3.3 Lack of Specimens
The lack of specimens has been mentioned in connection with conducting molecular research, but it
is also true for morphological work. The greatest asset for conducting taxonomic research, especially
phylogenetic work, is to have access to specimens for study. Although much taxonomic work on the
Pentatomoidea has been completed, there is still much to do. There is still much diversity to be explored.
To illustrate this, we will provide a couple examples. The Neotropical discocephaline genus Phoeacia
(Figures 2.13B, 2.17L) currently has three described species. Several new species were discovered, so
loans were requested to complete a review of the genus. There are now over 30 distinct species known.
Similarly, the New Guinean rhynchocorine genus Pegala Stål currently contains six species, but at least
one or two should be removed to other genera. Again, two undescribed species were discovered, so a
revision was planned. Now, after accumulating museum specimens, there are over 25 new species and
several new related genera in need of description. Although these two examples may not be typical, they
certainly are not exceptional. Many new genera and species await discovery, hence the need for further
collecting.
The lack of fresh specimens is an even more severe problem for DNA work. Molecular research
requires access to relatively fresh material, preferably material collected in the last five to ten years,