Biology Reference
In-Depth Information
focus of several studies, including one that examined two factors that might shape dragon-
fly wings: long-distance migration and high-maneuverability mate guarding (
Johansson
et al., 2009
), the impact of temperature and insertional mutations of 16 genes involved in
the formation of the Drosophila wing on wing shape, its variance and developmental stabil-
ity (
Debat et al., 2009
) and the modularity of insects wings (
Klingenberg and Zaklan, 2000;
Klingenberg et al., 2001
). An especially challenging problem for landmark-based studies
was confronted by
Benitez-Vieyra and colleagues (2009)
who analyzed sexual mimicry in
orchids. The orchid's flowers attract sexually excited male wasps that pollinate them;
although chemical signals play a major role in attracting the wasps, the question
addressed by this study is whether plants that are shaped more like female wasps are
favored. The authors thus had to find landmarks on the distal part of the orchid labellum
corresponding to those on the head and most of the thorax of the wasp!
Many more examples of measurement schemes can be found by searching on “geomet-
ric morphometrics” and a keyword relevant to your study, such as “geographic variation”,
“ontogeny”, “quantitative genetics” or “phenotypic plasticity” or the name of a study sys-
tem. Reading several of them before you begin your study can help solve one of the major
problems confronting a researcher, that you need to know how to measure your organisms
before you can analyze your data but you may not determine what you need to measure
until you finish your data collection. Fortunately, this is not as insurmountable a problem
as it may appear to be even if the relevant literature is sparse. That is because the process
of digitizing helps you discover where that variation is. In general, by the time that you
have finished digitizing (and redigitizing) your specimens, you will not be surprised by
the results. That is especially true if you use one of the nice features of the most widely-
used digitizing program, tpsDig (discussed in more detail in the workbook). That feature
is the “template mode” which allows you to copy the landmarks from one specimen to the
next. You need only to reposition them. That ability to copy all the landmarks, in approxi-
mately the correct positions, is very useful when your landmarks are not in a sensible
anatomical order; one of the common measurement errors is to digitize landmarks out
of order. But the template mode is not only a convenience, it is also a way to discern what
varies. Once you have positioned and scaled the copied landmarks to suit the photograph
onto which they are copied, you will find that some landmarks need very little adjustment
but others regularly do. As you scroll through the images, copying and adjusting the
landmarks, you will soon recognize where the variation lies. You can thus design your
measurement scheme by a pilot study, which beings with a preliminary set of landmarks,
perhaps based on a published study, which you modify by adding or repositioning land-
marks and adding curves, until you can measure the variation that you see. As should be
evident from at least two of the examples above, the skull landmarks of the two rodents,
cotton rat (S. fulviventer) and laboratory mouse (M. m. domesticus), as well as the mandibular
landmarks of rodents, measurement schemes evolve.
A pilot study will be especially valuable if you plan to obtain three-dimensonal data
directly from specimens using devices such as a Reflex Microscope or Microscribe (see the
workbook for summaries of methods for capturing three-dimensonal coordinate data).
Such devices offer less flexibility than other technologies that produce reconstructions of
the three-dimensonal images whose coordinates can be digitized. When all that you have
are the coordinates that you chose to measure before beginning the study, you obviously
