[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Subject Index][Author Index]

Re: Elephants and hyenas


From: Raptorial Talon <raptorialtalon@gmail.com>
>To: Jura <pristichampsus@yahoo.com> 
>Cc: "dinosaur@usc.edu" <dinosaur@usc.edu> 
>Sent: Sunday, 7 April 2013 9:35 PM
>Subject: Re: Elephants and hyenas
>Question from a novice: couldn't convergence to fill comparable niches lead to 
>convergences in behavior? If, for example, large ceratopsians and rhinos have 
>evolved broadly similar biomechanical solutions to quadrupedal locomotion, 
>such that their gaits are more comparable to each other's than either's is to 
>a modern crocodile, couldn't something similar be true of their behaviors? 
>(Obviously physiology is a bigger factor in determining k- versus 
>r-strategists and so on, but even so.)
>I mean, sure, the value of such
 inferences would be low because they're so broad, but isn't it possible
 that convergence could make a dinosaurian group more similar to a 
mammalian group than to a croc or lepidosaur, whether we're speaking 
behaviorally or biomechanically? I guess I'm suggesting that the 
constraints of specific ecological roles can overwrite the signal from 
phylogenetic bracketing to some extent, making the latter potentially 
dubious for highly derived taxa . . . I don't entirely trust volant 
hyper-endotherms and aquatic ambush predators to elucidate the 
strategies of relatives which had radically different and 
non-overlapping niches.
>Certainly. Extensive convergence is something that we see in many animals 
>groups including extant birds and mammals. Both mammals and birds share a 
>suite of behavioural and (grossly) physiological traits to the exclusion of 
>other amniotes. These gross similarities were enough for Owen (1866) to erect 
>the haematothermia. So extant birds and mammals show that extensive 
>convergence can happen in distantly related groups subjected to the same 
>environmental pressures.
>So mammals and birds show many features in common with one another. 
>Unfortunately, few of these features leave marks on the bones. Physiology a
prehistoric animals. This brings us to the whole bugaboo regarding inferring 
dinosaur behaviour based on mammals. When we use crocs and birds (and even 
lizards) we are limiting what we can say about dinosaurs based on what we know 
of their extant relatives. However, when we do infer things like nest making / 
guarding, or unidirectional flowing respiratory systems in dinosaurs based on 
what extant archosaurs are doing, we at least have some level of confidence in 
our inferences. Since birds and crocs both show this particular trait we can 
hypothesize that said trait was present in their last common ancestor. Should 
that trait leave a mark on the bone then we can even test our hypothesis by 
looking for those osteological correlates in the fossils. 
>This then leads us to the problem. If we infer that mammals and dinosaurs had 
>some type of similar physiology / behaviour then we are making what Witmer 
>(1995) called a level 3 prime inference. If mammals and dinosaurs shared a 
>particular trait in common, and this trait was to the exclusion of extant 
>archosaurs, then the trait must have evolved two separate times (and was later 
>lost on the lineage that gave rise to birds). For behavioural and 
>physiological traits that means inferring mammalian qualities on dinosaurs 
>without any way of backing that inference up. It is basically just 
>speculation. Yes it may have happened, but what we see in mammals may also be 
>one of dozens of solutions to the problems posed by that particular 
>environment. We have no way of knowing and no way of testing these level 3' 
>inferences out. 
>So yeah, using extant archosaurs, or even extant diapsids as our hard limits 
>for dinosaurs is bound to result in the underestimation of certain dinosaur 
>qualities. However by forcing this conservative view on our interpretations, 
>we (theoretically) keep ourselves from having our speculations run beyond the 
>reach of our data.
>All this isn't to say that a strict extant phylogenetic bracket approach is 
>the only way. Mammals can provide insights into how
ngs. Traits in dinosaurs may even be hypothesized to be analogous to traits 
seen in mammals, especially when these traits seem to have no real 
analogue/homologue among extant diapsids. Though given what little we know of 
extant diapsids, I'm often left wondering how many of those traits actually 
>Owen, R. 1866. On the Anatomy of Vertebrates, Volume 2. Longmans Green and 
>Co., London.
>On Sun, Apr 7, 2013 at 7:36 PM, Jura <pristichampsus@yahoo.com> wrote:
>>----- Original Message -----
>>> From: Dann Pigdon <dannj@alphalink.com.au>
>>> To: dinosaur@usc.edu
>>> Cc:
>>> Sent: Sunday, 7 April 2013 6:42 PM
>>> Subject: RE: Elephants and hyenas
>>> On Mon, Apr 8th, 2013 at 4:39 AM, john-schneiderman@cox.net wrote:
>>>>  I think that the ways of birds and crocodylia are our best models for
>>>>  parental care and/or defense of the young within the dinosauria.
>>>>  Dinosaurs were not mammals and mammal behaviour shouldn't be used as a
>>>>  guide to what we consider dinosaur behaviour. Tempting as it is.
>>> Then again, birds and crocs are highly specialised archosaurs. We can't
>>> assume that non-avian
>>> dinosaurs behaved any more like modern archosaurs than they did modern 
>>> mammals.
>>Despite their specializations, the similarities between these two extant 
>>groups make for our best inferences of what dinosaurs did. We could extend 
>>our bracket out a bit further to lepidosaurs, to look for traits that were 
>>shared with birds and crocs, and incorporate those too. We might even use 
>>seeming lepidosaur only traits if we can justify it in dinosaurs.
>>The point is that dinosaurs and mammals are extremely far removed from one 
>>another. Even when we take into account the specializations seen in our 
>>extant archosaurs, inferences based on their behaviour are still better for 
>>reconstructing dinosaurs than comparisons with basal amniotes like mammals.