Hi John - very interesting question. All the literature I have seen on this relates to maturation reaction norms where there is a clear selection advantage to maturing younger/smaller. Your observations sound very intriguing although I really cannot suggest any other studies I know of on this - is there any chance that thermal reaction norms and maturation could be linked in some way?

Thanks for your thoughts Clive. Unfortunately, we don't have any maturation data for these fish to formally test whether variation in thermal reaction norms is related to altered maturation schedules. However, the time frames for the observed changes (5-10 years) suggests that it is highly unlikely that fish are now maturing earlier. This is why I was thinking along the lines of some sort of indirect density effect on thermal response. It has been shown, for example, that temperature plays a role in the manifestation of predator induced phenotypes (e.g. in tadpoles); I guess our data suggests that predation (in this case fishing) can in turn affect the expression of thermal phenotypes. Anyway, I'll keep digging and hopefully find a few more hints.

I think the most straightforward link is as follows: fishery are generally size-selective, causing a demographic inbalance between age classes in a population. Thus, reaction norms may changes simply due to this fact (a new demographic structures, affecting both the mean and variance). Thermal response varies between size-age classes, due to shifting surface-volumen relationships (larvae vs. adults) and capacity of a given species to regulate body temperature

Not sure exactly how John has calculated his reaction norms but it should be possible to factor out the age effect?

I'm pretty sure it is not an age effect as I have rigorously checked the analyses using same age fish from the two periods (e.g. reaction norms calculated from 12 year olds) and the magnitude of the variance difference remains the same. The modelling framework I used to estimate reaction norms (mixed model using within subject centring, Morrongiello and Thresher 2014) also explicitly accounts for ontogenetic growth etc. There is, however, the potential that the obvious answer as per Alvar's suggestion is what is really going on- always very hard to be sure with correlative data!

very little is known about thermal reaction norms for individuals, and how they covary for metabolism, growth and behavior.  we are currently exploring this in yabbies, at least in part.

good luck john, and congrats on the job/postdoc or whatever you have going on in melbourne.

Hi Pete

Thanks for your response on this. After some further digging, I can support your point about there being little known about individual-level thermal reaction norms in this context. I've been able to make what I think is a sensible case for the variation being mediated by density, but as this is only a field-based correlative study (on the scale of yearly growth), I am not able to experimentally test it.

I really liked your recent paper in proc roy soc on yabbies- nice work!

John