... sleep deprivation and a wet head.
Night zooplankton sampling one of twenty seven. Actually an extension of day two because of some really bad weather coming in.
After dark we take a 5 litre integrated sample from each tank. We do this at night to make sure of an equal distribution of zooplankton within the mesocosms - during the day they will hide in the macrophytes from their fish predators. The water is then filtered through a 20µm mesh and then the zooplankton caught are washed into a bottle before being preserved with ethanol. Identification and counts (the most fun part) happens at a later date.
How wrong I could be. At the height of summer sampling start times were as late as 11.30 p.m, and end times as late as 2.35 a.m because of course now that it is summer its light for longer and more of the tanks (all of them it seems) have high algal biomass so trying to filter five litres of water per tank through a 20μm gauze can take some time. I have never been so happy for the days to get shorter again.
Sometimes when algal biomass is really high its just not possible to get the sample through a 20μm gauze so we filter through a 50μm instead. This still captures most of the zooplankton but the mesh size is too big for some of the smallest rotifers. During the winter all of the tanks would go through a 20μm gauze without any problems, below shows you the situation at the end of August - mainly 50μm filters. It demonstrates nicely how different the biomass is between the winter and summer months.
This last ever night zooplankton sampling finished at 10:25pm, all went well, a bit windy but no rain, until I was packing up the buckets and out of nowhere one big torrential downpour hit me. From bone dry to soaked in a matter of seconds. Boy am I glad that its all over!
All in the name of science. Plankton will make me do the strangest things.
... a very long day and almost the death of my filtering pumps.
I then drive the samples back down to the lab, pop them in the fridge and get ready to filter for the next four or five hours. First I preserve 100ml of unfiltered water from each tank with Lugols, this is for algae identification and bio-volume counts. I send off 50ml unfiltered water from each tank to chemistry for total nutrient analysis (the nutrients found in the water and also within organisms).
The rest of the water sample gets filtered like you can see below. I need one filter for chlorophyll a analysis (we do laboratory analysis as well as using the sonde ) and one filter for cyanobacteria cellular toxin analysis. Toxins can be within cells or released into the water so I also need to take a certain volume of the filtered water for in-water toxin analysis. Some of the tanks are so high in biomass that obtaining the right volume of water for toxin analysis takes a long time and a lot of filters. Below is pump one of two, they run continuously. I started filtering at 10am and finished at 3.15pm - the poor pumps felt a bit hot by the end. The remainder of the filtered sample is split into another sample for nitrate analysis which the chemists process and the other part for srp (soluble reactive phosphorus) analysis which I will do later in the afternoon. Nutrients can be held within organisms or are available in the water column, total nutrients is a measure of both whilst srp calculates what is available in the water. Cells will release these nutrients when they die and so a high srp in a closed system like the mesocosms can be indicative of a break down of biomass.
Once the filtering is done I get on with srp analysis. This involves chemistry, reactions, sulphuric acid which smells pretty bad, wearing safety goggles which make me look very attractive and a spectrophotometer, which I can't say properly.
And that marks the end of Wednesday, well usually it would. Because of the weather day three, Thursday is being moved to day two so I have to now bike home to get the head torches (which I forgot) to start on part three, night zooplankton sampling.
... lots of data and some dead fish.
This is actually the last sampling event (one of twenty seven) for the mesocosm experiment. So in case I miss it here is a break down of what happens when we go and sample.
Day 1: in situ data collection (sondes, algaetorch and nephlometer), periphyton and maintenance
Day 2: collection of water samples, filtration in the lab for toxins and chlorophyll a, srp analysis and sample fixation for bio-volume counts
Day 3: Night time zooplankton sampling
Day 1 - Tuesday, in situ data collection
As soon as I get to the mesocosms the Exo2-sonde goes into the first tank, we take averages from at least six readings which takes four minutes per tank, four times thirty two equals lots of minutes so I get that started straight away. The EXO2 is a muti-parameter sonde that measures temperature, conductivity, salinity, pH, dissolved oxygen, chlorophyll and blue-green alage (cyanos). Measurements are logged internally which means I can get on with taking some other readings. The algae torch also measures chlorophyll a and cyanobacteria biomass using fluorescence of algae cells and is used to get an in-situ measurements. I, as a plankton geek think it pretty cool! The nephelometer measures turbidity (suspended particulates) by recording the light reflected from the particles (from a beam of light emitted from the instrument).
As I go around I also check all the tanks for any maintenance issues and clean all the sensors (temperature, oxygen and PAR - photosynthetically active radiation) located in the tanks. I also make any observations of 'unusual' activity in the tanks, like these dead fish found below. I'm not entirely sure why there were so many dead right now, the only thing immediately different about this tank is the amount of dissolved oxygen (%) - 27% compared to >65% in the rest of the tanks, which could be a factor.
To account for this, periphyton growth is measured on a monthly basis in each tank by calculating the dried weight from pre-measured in-situ strips. The picture below shows the difference in the bio-volume and type of periphyton growing in each tank which demonstrated nicely the need to account for this effect.
And that marks the end of the Tuesday session up at the mesocosms. All I need to do now is enter the data and prepare for the sampling marathon tomorrow.
Over the past few weeks Euglena sanguinea has taken a bit of a liking to our mesocosms ...
Aspect number one: What is that? "Hey (generic alien/scary film character name), come and look at this...."
Aspect number two: "Don't touch that s***, it moves!"
It makes you want to double up on gloves!
These birds are the unfortunate casualties of this horror story- don't worry the main character mysteriously survived and is now casually sipping a Martini in the lab.
The surface scum is very densely packed and moves almost as a mass ; its quite hard to describe but its a bit like the skin that forms on custard (but less appetising) or probably more in line, the mouldy skin that forms on a cup of tea when you have left it on your desk for too long (I assure you that I have never done this). I should stress that not all of the tanks are so 'bad' (actually only two, as far as I have observed) and in the one featured above a fight for supremacy has been going on between Euglena sanguinea and Anabaena spirulina.
On a serious note - are there any trends to these blooms? On the surface of it all not really - both of the tanks with the dense scums are nutrient enriched tanks, but one is heated and the other is unheated. Many of the tanks have some reddish tinge to them (confirmed to be E.sanguinea after a peak under the microscope) with varying degrees of scum development; they look more like the 'after' photo below. So, it could be a case of different responses between tanks because of different exposure to mixing from the wind. The clear ability for E.sanguinea to form dense surface blooms means that it can compete with other buoyant species like Anabaena and Microcosytis; seeing as my analysis will be investigating the response of cyanobacteria to different stressors this is definitely one to keep my eye on.
What a difference a windy weekend makes ...
Rather like a Farrow and Ball paint chart, however 'Microcystis Green' or 'Anabanea Green' is far more exciting than 'Ball Green' and 'Ground Green'. If this doesn't turn out well I could change careers to be a paint colour consultant, they have 'Arsenic' as an option so I think they could be interested in some more toxic varieties.
The pots contain samples from each of the tanks (sampled this morning) and will be fixed with Lugols so that I can look at the community composition and also the bio-volume of the most dominant taxa in each tank at a later date.
The experiment is a two factorial block design; factorial meaning that each treatment is cross-matched with all other possible treatment combinations giving eight stressor combinations (heated, unheated, heated + nutrients, unheated + nutrients, heated + flushing, unheated + flushing, heated + nutrients + flushing and unheated + nutrients + flushing), block meaning that each of these eight treatments are randomly assigned to one tank in each block (or row) of eight tanks. We have four blocks (see below) which makes four replicates of each stressor combination and 32 tanks in total.
What you see in the first picture are samples from each tank ordered from number 1 - 32; because the treatments are randomly assigned to tanks within each block this means that treatments are not placed together. Shuffling them around to place them in their treatments looks like this:
The colour gives quite a good indication of the relative amount of algae in each tank, green ones have more algae than clear ones. Here are some chla concentrations to convince you of this (for those unacquainted, chlorophyll a is a pigment used in photosynthesis and is used widely as a proxy for biomass, it is measured in micro grams per litre (μg L-1) in this case using an instrument with fluorescence):
Heated + Nutrients (average): 295 μg L-1
Unheated (average): 60 μg L-1
Here comes the time to make some sweeping generalisations from one sampling date based on some colour charts (and some data too): nutrients support higher algal growth (the 16 tanks to the left) than tanks without nutrient enrichment. It is already generally accepted that nutrients are a strong driver for increased algal growth, especially cyanobacteria; this makes sense, as nutrients are essential for growth. It is also thought that warming also increases algal growth again especially cyanobacteria, the reasons for this are far more complex but it can be seen here that heated tanks seem to have more growth than unheated tanks (without nutrients). But do these single treatments and multiple treatments favour cyanobacteria over other algae? Time at the microscope and some data crunching will soon tell.
The real crux of the matter (of the experiment) is to understand how cyanobacteria respond to combinations of these stressors (treatments) because real lakes can be subjected to nutrient enrichment, warming or changes in flushing rates all at the same time. In the past research has focused on the response of cyanobacteria to individual stressors, and so this experiment along with other the other factorial mesocosm experiments, time series analysis and European scale analysis of the MARS project will contribute to our limited knowledge of how aquatic systems will respond to multiple stressors.
I have managed to turn my rather unhealthy obsession with plankton in to my day job. Things don't get much better than this! This blog documents my PhD research and the plankton delights I encounter along the way.