Artificial upwelling – in our experiment, the partial exchange of nutrient-depleted water in the mesocosms by nutrient-enriched Atlantic deep water, mimicking natural upwelling – can lead to enhanced surface productivity. We have had the chance to see this with our own little eyes during the last few weeks, by watching the originally blue water in some of the mesocosms turn green [LINK], or by looking at the differing amounts of organic matter that end up in the sediment traps at the bottom of the mesocosms (sediment subsamples depicted in the header of this page). But do these observed phytoplankton blooms also lead to enhanced CO2 sequestration? In other words, how much of the detritus that we have caught in the sediment traps (export flux) would have made it (would have been transferred) to the deep ocean, carrying with it the CO2 that was taken up by the phytoplankton during photosynthesis in the sun-lit surface layer? As a step towards estimating this so-called export efficiency, we have been measuring how fast the detritus collected in the sediment traps sinks. I was doing this (and probably have been doing this for too long) by simply dropping bits of detritus into filtered seawater, watching them, and stopping the time they need to sink a certain distance. It is surprisingly entertaining. Especially when you get to look at the winning detritus aggregates under a stereo-microscope also called “lupa”.
Photo of a dried detritus aggregate taken through a lupa. The scale in the background is in millimeters
To get statistically solid results for all mesocosms and all sampling days, stopping aggregates by hand is not very efficient, as you can imagine. And I would miss a lot of the small stuff that is hard to see. Luckily, we have two FlowCams here that can be turned into detritus speed cameras (see image below). They basically work like me with the stopwatch, just faster, and probably more accurately. And they can keep track of many (also smaller) particles at the same time. First, the detritus from the mesocosm sediment traps is brought to the lab, diluted in filtered seawater, and the mix is then pulled up into a pipette using a rubber bulb. Subsequently, the detritus that starts to sink within the then stationary filtered seawater in the pipette is photographed 20 times per second while passing a flashlight (see violet light in animated gif below). And from the y-positions of the detritus aggregates in the photos and the times at which the photos were taken, the sinking velocities are computed. So much more efficient!
FlowCam converted into a detritus speed camera
Detritus sinking in the cuvette / the heart of the speed camera. The camera can take up to 20 images per second, and for each photo, the flashlight is only on for a few microseconds. The fast flashes are difficult to capture in a coarse-in-time animated gif like this.