One year down…

A year ago today I stepped foot on the campus at Curtin university for the first time. I don't know where the time has gone - how can it really be a year?!

It's been a pleasant surprise - especially given how my last attempt at a PhD went - to realise how much I'm enjoying myself. I'm making slow but steady progress; I even have data which is quite incredible! All this means that I have some idea of what I'm actually doing and to preempt the traditional 'what are you working on?' questions when I next come home I thought I'd take the time to explain just that.

I'm working on two projects which I've nicknamed 'fisheries' and 'fossils' as I'm really unimaginative. In this post I'll talk about the fisheries project as it's the one I've been working on most so far and I've written a lot more than I expected so think this is better off as a two-parter. 

'Fisheries'

The 'fisheries' project is funded by WAMSI (the Western Australia Marine Science Institution) and is one small component of a huge multi-million dollar project called the Kimberley Marine Research Program. In case you don't know (and I didn't when I began) the Kimberley is the north-western part of Western Australia - there's a handy map at the research program link above. 

I'm using otolith geochemistry to investigate ontogenetic niche separation and connectivity of fish populations. That's a lot of complicated words so I'll try and translate…

Otoliths

So, otoliths. I love otoliths. I first came across them when I worked in the Falklands, though I can't remember if I properly wrote about them so I'd better do it now. Otoliths are found in the heads of fish:

Inside a fish head (from Secor et al. 1992)

There are three of them - the sagitta, lapillus and asteriscus - and there is one set of them on each side of the fish. They are found around the semi-circular canals and help the fish with balance and hearing. In most fish the sagitta is the biggest otolith with the lapillus and asteriscus being so small they can easily overlooked, so unless otherwise stated, when I'm describing the work I do on 'otoliths', I really mean the sagittae.

So why should you care about otoliths? Well, the amazing thing about them is that they grow with the fish. They're made of calcium carbonate (which is the same thing that a lot of shells are made from) and each day they grow by a very small fraction. The amount of calcium carbonate deposited will vary over the seasons and this produces light and dark bands, similar to the bands you see in tree trunks, which can then be counted to age the fish. The centre of the otolith is called the 'core' and this is the oldest part of the otolith, forming when the fish is still in its egg. The outside of the otolith is called the 'margin' and this is the newest part of the otolith which formed just before the fish died.

Annotated otolith

Otoliths come in a wide variety of shapes and sizes and the precise shape and size is unique to each species. Fossil otoliths, and otoliths in the stomachs of animals, can be used to identify the fish they came from which is really useful if you're trying to understand diets and food webs.

A selection of otoliths from the Falkland Islands

Otolith Geochemistry
The other exciting (and more pertinent) aspect of otoliths is that it's not just calcium carbonate that they're laying down as they grow. Trace elements and isotopes from the environment also get incorporated into the structure, and it's these elements that I'm interested in. Trace elements come from the earth, mostly from rocks. They get washed into streams and rivers and then flow out to sea. This means that coastal areas will have different trace elements depending on the geology of the nearby land. Further out to see the water gets all mixed up and the trace elements get more homogenised, but close to land they can vary quite a bit. This enables fish from one part of the coast to be distinguished from fish from another part of the coast by their trace elements. And if the fish moves around the coast during its life then this can also be seen through changes in the trace elements in different parts of the otolith.

Isotopes of strontium and oxygen vary according to the salinity of the water, so if fish move from freshwater to saltwater this can be seen in changes in these isotopes. By combining information from trace elements and isotopes we can get a pretty clear picture of how the fish are moving during their lives. And by taking that information from lots of fish we can see how the population is structured. Which is what I am doing…

Ontogenetic Niche Separation

So that's otolith geochemistry. Next up is ontogenetic niche separation. This is just a very scientific (or poncy, depending on how generous you're feeling) way of saying how animals partition up the environment at different points in their life cycle. Ontogeny is how organisms change as they grow. A niche is the set of environmental characteristics that an organism inhabits - the climate, type of habitat and so on. And niche separation just means how competing organisms use their niches to try and reduce competition. Ontogenetic niche separation can occur when the animals all live together - for example butterflies where the caterpillars eat leaves while the adults drink nectar; or when they live apart - for example a lot of insects, where larvae will live underground while the adults live above ground. For a lot of coastal fish this second option is preferable. The adults will spawn in the open ocean and the larvae will float around in the plankton for a bit before settling in coastal mangroves and seagrass beds - habitats often termed 'nurseries' due to the presence of so many juvenile fish. Once the fish have grown bigger and matured they will move out to the more marine habitats where they will live as adults, spawn and start the cycle over again. 

Why does any of this matter? It matters for conservation. If you're more economically-minded you might like to think of it as mattering for sustainable exploitation of natural resources. When you think about conservation, chances are you think about protecting species like pandas and tigers. A lot of conservation used to think like this too. But it's become increasingly apparent that protecting a species without protecting its habitat is useless (and impossible - at least in the wild). To allow populations to exist in the long-term they need enough habitat to sustain themselves. With all the pressures on the natural world this is really difficult, but it's even harder when the animals undergo ontogenetic movement and use different habitats at different stages of their life cycle. While not quite the same, think of the birds that winter in the UK every year. Geese and swans and other birds come to places like Slimbridge and the Norfolk Broads to over-winter every year. We try and protect these habitats because we know that the birds need them, but what we're not doing is making sure that they're summer habitats are also being protected yet there's little point protecting their wintering grounds if they have nowhere to go in summer.

It's a similar thing for coastal fish. A lot of focus these days is placed on protecting spawning grounds by preventing them from being fished too much, or at all. This is great, and very important, but it's pointless if we're catching all the juveniles before they even have a chance to become spawning adults. We need to understand how the fish are using the entirety of their environment, at every stage of their lives in order to protect them from being over-fished. The economic argument should be obvious - catching all the fish and leaving nothing for next year is not a very sensible business plan. Essentially it boils down to that old adage, 'information is power.' The more we know and understand about how the fish interact with their environment, the better equipped we are to manage that environment to ensure the long-term health of the fish populations. And, the great thing about this 'ecosystem management' strategy is that even if you only care about one or two species, you end up protecting many more at no additional cost.

Connectivity
So far I've been talking about movements of individual fish, but fish don't generally live alone,  instead they live in schools (or more broadly, populations, though I will stick with schools for the benefit of the analogy). And, like human schools, these fish often grow up together, and do the fish equivalent of moving from infant school to junior school to senior school (or whatever they're called these days) along with all their classmates (this would be their ontogenetic niche separation, with the different age groups using different environments to minimise competition and conflict). What the question of connectivity raises is 'how many fish/pupils transfer to schools in different areas'? Is it that everyone stays with their schoolmates for their entire life with no-one new coming in, or are kids (or fish) transferring between schools every week? You may ask why we care. And the reason comes down to diversity. In the case of fish I mean genetic diversity. If the fish stay in the same school all their lives then their potential partners are limited, and over generations this will mean that the genetic diversity will be very different to that of another school as there's no interbreeding. If, however, fish are transferring between schools a lot then there's much more genetic diversity but it also means that it's hard to tell one school apart from another.

But why should we care? I hear you repeat. It comes back to conservation/management. In the same way that niche separation means we need to protect all the habitats used during the life of the fish if we want to keep populations going, connectivity means we also need as much genetic diversity as possible. If each school is genetically distinct then we need to protect as many of them as possible, which may mean protecting an awfully big area. If, however, there's a lot of transfer between schools - a lot of connectivity - then the loss of one of these schools isn't quite such a big deal and we can maybe get away with protecting a smaller area. By understanding how the schools (or, more broadly, populations) are connected, we can manage them more efficiently.

Working out connectivity is commonly done using genetics, but there can be problems with this, particularly in terms of time lag which means that recent changes in connectivity may not be easy to detect. It is, however, fairly cheap and easy to do. When using otoliths connectivity is assessed by looking at the cores to determine the number of spawning populations. However, the power of otoliths is that, unlike genetics, it is possible to track the fish as they grow, which means that it is possible to see if they form different groupings at different ages - do all the fish from one spawning ground spend their entire lives together, or do they go to 'infant school' together then go to different 'junior schools' only to meet back up for 'senior school'? - for example. Genetics can't tell you that but otoliths may be able to.

So there are pros and cons to the two methods but it's not clear if one is better than the other. And that's where one of the most interesting aspects of this project lies. The genetic connectivity of the fish has already been analysed so I will be able to compare my results - once they're done - with these and see how they stack up.

How do I investigate the onotogenetic niche separation and connectivity of fish using otolith geochemistry?

The first thing to do is pick a fish species. I picked two: Lutjanus carponotatus and Pomacentrus milleri. Both these fish are common in the Kimberley which is why they've been chosen - it's a remote area and it's hard to get samples so the easier we can make things the better. Unfortunately all my samples were collected for me so I haven't had the chance to visit but I hope to get up there one day even if it's just for a holiday. 

The next thing to do is prepare the otoliths so their geochemistry can be analysed. This is what I've spent most of my time doing. It's a fairly involved process to go from otolith to data. I need to get a section of the otolith through the core - they're three dimensional structures so I need to cut it in half to see the centre, like cutting a peach to get to the stone. They're very small so it's not possible to do this by just holding the otolith and cutting, they're embedded in resin first to give me something to hold on to. 

The process begins by removing the otoliths from the head of the fish and then storing them in either paper envelopes or little plastic tubes depending on their size. That's the point at which I get them. The first thing I do is give them a clean. They're far too fiddly to clean individually so I use a machine called a sonicator. This is a machine that holds water and then vibrates at a very high rate. (Jewellers use them to clean jewellery). Once they've been cleaned and dried the next step is to embed them in the resin. This is done through the high-tech method of using ice-cube trays. A small base of resin is made and then when this has hardened the otolith is placed on top and more resin poured over. This is then left to harden. 

Otoliths embedded in resin

 Once the resin is hard it's time to cut the otolith - or 'section' them to use the proper term. For this I get to use a slow-speed circular saw.

Sectioning set-up with saw and microscope

You end up with three pieces from your one 'resin cube'. The end piece, the section through the core (hopefully), and the rest of the otolith. How do I know I'm cutting through the core? There's three stages to ensuring I hit it. The first is to mark the sulcus - this is a grove that runs most of the length of the back of the otolith - with a pencil.

Marking the sulcus (and yes, those are my fingers, which gives you an idea of the size of these things)

The next steps comes when the otolith is embedded. You can see the core when you hold the otolith up to the light (or look down a light microscope) and you can then mark the resin directly above the core with a pen. The final step is to take a sheet of paper with a big cross marked on it, line up the sulcus along one of the lines and then mark the other part of the cross on the resin. Why do I do this? It's because I need to cut through the core cleanly, not at an angle. The sulcus is running along one axis of the core but if I cut down it then I get a really long and awkward section. A Much easier cross-section to work with is going perpendicular to the sulcus and that's what the cross allows me to do. But where I drew my cross is pretty arbitrary so by marking a dot for the core I have my target and all I need to do is take aim. 

Once I have the section then it's time to check it's gone through the core which I can do under the microscope, and then I polish it up. When you're aging fish it's often best to have the sections really thin as the three-dimensional shape gets in the way - the lines can sort of 'smush' together if you cut too thick and make it really hard to differentiate them. With the work I'm doing I can get away with the sections being a bit thicker and in fact given the polishing I need to do and the fact the laser (oh yes, I use lasers!) will drill holes in them, I like having them a bit thicker. I need them polished because the smooth surface is better for the laser and it also makes it easier to see things. In the same way that the rings of a a roughly sawn tree trunk are fairly difficult to count in comparison to a beautifully polished one, so too for otoliths. I polish them using lapping film (which seems to be a fancy term for 'really fine sand paper') and a bit of water. Once they're polished I then cut off the excess resin and then mount them. For some of my otoliths I mounted them in the same way that geological samples are mounted which involves embedding them in more resin in a circular mould. For others I just stuck them onto microscope slides. This latter method was definitely easier, but the former has its uses.

Geological-style mount vs microscope slides

Geochemical Analysis - Now with Lasers!
At last the otoliths are ready to go for analysis. I'm using three different methods of analysis to measure three different aspects of otolith geochemistry. The first is trace element analysis. This uses laser ablation inductively coupled plasma mass spectrometry (or LA-ICP-MS for slightly shorter). The name is quite descriptive, saying that a laser is used to vaporise (ablate) a bit of the surface of the otolith and then puts it through a mass spectrometer to determine which elements are present. 

LA-ICP-MS machine

The machine takes up a large part of the room it's in. The picture above is only one part of it - there's a whole other part that looks a little like the stupidly unrealistic particle accelerator Tony Stark built in Iron Man 2 (this is the mass spectrometer). The machine is controlled by computer and it's so easy even I could do it!

All I had to do was set my laser spots for analysis, the rest was done by the amazing technician. You tell the laser where to point and then it points and does it's thing. It's amazing! The results comes out as .csv files which need processing using an expensive piece software to extract data in a format that can be analysed. Learning how to use this software was the most complicated step but I think I've managed it. 

Next up is strontium isotope analysis. This is being done at the University of Melbourne as there's nowhere in Western Australia that's set up to do it yet (but hopefully Curtin will be soon). It's essentially the same method as for the trace elements, just instead of detecting trace elements it's detecting strontium isotopes. In case you're interested, the method being used is called multicollector inductively coupled plasma mass spectrometry (or MC-ICP-MS). It still uses laser ablation but it uses a multicollector (and no, I don't know what that is) to measure the strontium isotopes.

The final method is the most complicated, but also, arguably, the coolest. It is measuring oxygen isotopes which both relates to salinity. The strontium analysis is also measuring salinity so why am I measuring it twice? Well, because neither of these methods measure it directly. They're what's called 'proxies'. And proxies aren't always entirely accurate. So by using two different methods we can see if they both give the same result. If they do then we can be pretty sure that they're accurate, but if they don't then we know that at least one is off. How do I know both aren't off but just giving the same result? It's a matter of odds. Isotopes of two different elements measured by two completely different techniques; the chances that they both give the same wrong answer is too minute to worry about. 

So what is this final method? It's called secondary ion mass spectrometry or SIMS (yay - an actual acronym!). This is another very cool piece of kit that takes up another room, this time at the University of Western Australia. 

SIMS at CMCA@UWA

The reason it's more complicated is that whereas LA-ICP-MS and MC-ICP-MS don't really care about how polished the surface is (within reason), SIMS really does. It needs to be flat. Really flat. The description I was repeatedly given was 'mirror-like finish'. How do I achieve such a finish? By polishing. For hours. Lapping film is not sufficient for this level of polish - I had to use a diamond suspension. First with a 9 micron suspension, then a 6 micron suspension, then a 3 micron suspension and finally a 1 micron suspension. You put a bit of suspension on a polishing plate with some lubricant and then make a figure-of-eight pattern with the mount. 

Just keep polishing, just keep polishing…

You check it under the microscope periodically to see if you've made any difference and then you go back and repeat for about 2 hours until you have - hopefully - finally achieved the level polish required.

Mirror-like enough?

The next step is taking SEMs of the otoliths to provide a map of the position of the otoliths in the mount. Then they're coated in a thin layer of gold and then, finally, they're ready for analysis. This involves more computers…

Ready for launch!

Surprisingly, it's a very similar set-up to the LA-ICP-MS - set your laser spots and away you go. The only downside to this method (other than the insane amount of polishing required) is that it's very slow in comparison to the other methods. Whereas LA-ICP-MS can analyse 50+ otoliths in half a day, SIMS will take that long to measure 3. This isn't because the machine is really slow but because the data is much higher resolution. For all its benefits this slowness is a major limitation in the number of samples that can be analysed (though one I won't moan too much about because the thought of polishing 50+ otoliths to that level is not one to relish). The reason the SIMS is so cool is that while LA-ICP-MS and MC-ICP-MS are standard techniques for otoliths, SIMS isn't. In fact it's barely been used so it's quite exciting to be doing something really new.

So that's it. I have data so far from the trace element analysis on about 60 L. carponotatus and 40 P. milleri and data from another 60+ L. carponotatus is on its way. I have SIMS data and the strontium isotope data should be coming in a couple of months so things are going well. I've started doing my data analysis and so far I have no idea what any of it is showing but give me time and I'm sure I'll have something interesting to show!

So that's it for the fisheries project so far. I didn't realise how long this post so if you've made it this far well done. I hope it's been interesting. If you've got any questions then please let me know in the comments.

Reference:
Secor, D.H., Dean, J.M., Laban, E.H., 1992. Otolith removal and preparation for microstructural examination, in: Stevenson, D.K., Campana, S.E. (Eds.), Otolith Microstructure and Analysis. Ottowa, Canada, pp. 19–57.