24.10.2014 Views

Dissertation - HQ

Dissertation - HQ

Dissertation - HQ

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

General questions on behavioural traits 29<br />

of information available on any particular behaviour is limited to relatively<br />

few species, and to only a portion of the larval stage (usually<br />

older larvae). When deciding if behavioural information from species A<br />

can justifiably be used in a model for species B, two things must be<br />

considered at the outset: the evolutionary closeness of the two species<br />

and the similarity of the environment in which the species live.<br />

The vast diversity of teleost fish species — approximately 27000<br />

species in 448 families divided among 40 Orders 66 — means that some<br />

species are very distantly related, with evolutionary histories that have<br />

been separate for tens of millions of years. Particularly among Orders,<br />

there is no reason to assume that behaviours will be similar. Within<br />

mammals, for example, no one would assume that the behaviour of<br />

a Tiger (Order Carnivora) would be similar to that of Dugong (Order<br />

Sirenia), just as no one should assume that the behaviour of a Plaice<br />

larva (Order Pleuronectiformes) would be similar to that of a Herring<br />

larva (Order Clupeiformes). As a general rule, in the absence of other<br />

information, the closer two species are related, the more justifiable<br />

it should be to assume they have equivalent behaviour. The use of<br />

well-corroborated phylogenies that encompass the species under consideration<br />

is essential in assessing the closeness of relationships, but<br />

for many fish taxa such phylogenies do not exist. Even this general<br />

rule should be applied cautiously, because there are many examples of<br />

larvae of confamilials with different behaviours. For example, larvae<br />

of some pomacentrid species are found in midwater, whereas those<br />

of other species prefer the top few centimetres of the water column 67 .<br />

At this point in our knowledge of the behaviour of fish larvae, it is<br />

difficult to make any defendable statement about how closely related<br />

two species must be before it is justified to assume the behaviour of<br />

their larvae is similar. An analysis of behaviour of fish larvae in the<br />

context of phylogeny to help establish if relatedness provides a sound<br />

basis for inferring behaviour would be most useful.<br />

It is unlikely that, even within a family, the larvae of a species that is<br />

pelagic as an adult will behave similarly to the larvae of a species that<br />

lives on a coral reef, or in an estuary. Therefore, if it is not possible to<br />

obtain behavioural data on the species of interest, the species supplying<br />

the behavioural data should at least live in the same habitat as the<br />

species of interest, both in the adult and larval stages. Echoing the<br />

comment above, an analysis of behaviour of fish larvae to determine<br />

to what extent habitat similarity provides a sound basis for inferring<br />

behaviour would be very valuable.<br />

Overall, the use of behavioural data from a distantly related species<br />

that lives in a different habitat should be avoided at all cost. Finally,<br />

there are indications that swimming speed can be predicted from the<br />

morphology of the larvae 68 . Therefore, the use of swimming data from<br />

species with similar larval morphology might be appropriate.<br />

Cross-Order<br />

generalisations<br />

do not make sense<br />

Adult lifestyle influences<br />

larval behaviour

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!