Thursday, March 20, 2014

Logistics

So here's how Tuesday unfolded...

2:50 pm - Alternator bearing froze in the middle of the Mara. We had already been suspicious of the new sound coming from our engine, and a friend's mechanic had checked it out, diagnosed the problem, and determined we could safely drive the 2 hours to the nearest town. We broke down a few hundred meters later.

2:53 pm - We had already ordered a new bearing from Nairobi as soon as we heard the diagnosis. Determined we couldn't get the part from Nairobi until the next day. Called our mechanic in Narok, 2 hours away, to source a more local alternative.

3:07 pm - Called our field assistant in Mulot, 1 hour away, to look for a back-up option.

3:11 pm - Called our friend whose family owns a shop in Aitong, 30 minutes away, to look for a back-up, back-up option. The closer the towns, the smaller they were, and the less likely we could find the part we needed, but it was worth the try.

3:49 pm - Determined a used alternator from Narok was the best/only option. Sent all the money we could (not enough) via cell phone to our mechanic to purchase the alternator.

4:05 pm - Our mechanic left Narok on a motorbike with the used alternator, heading for the Mara. The roboticists and I went to hang out with our friends we had been visiting. Chris, like a true sea captain, wouldn't leave the vessel, and insisted on sitting outside the gates in the afternoon sun with the Land Rover.

6:49 pm - Our mechanic arrived at our Land Rover in the Mara. She (the Land Rover) and Chris are outside the gate, and it's starting to get dark.

7:30 pm - Chris rolled into our friend's place, new (used) alternator in place and the Land Rover purring like her normal self again. Decided it was too late to drive, and a great opportunity to spend more time catching up with friends, so set up our tents for the night.

4 hours and 40 minutes from breakdown to repair in the middle of the Mara-- not bad! Of course, we have some great friends to thank for their help in getting things running again so quickly, and having a comfortable place to camp for the night. In Kenya, logistics are all about who you know, and how easily you can contact them. I can't imagine having worked here before cell phones!

Tuesday, March 18, 2014

True safari experience

Wouldn't be a true safari experience without a breakdown! Alternator bearing froze approximately 500 meters after a mechanic told us we could make the 2 hour drive to town. Fortunately we're near some friends who can help out, and we're pulling out all our logistics tricks to sort this out... wonder where we'll end up tonight...

Wednesday, March 12, 2014

Can a robotic boat tell us what's at the bottom of a hippo pool?

What happens inside of a hippo pool? Is it deep with a hard bottom, because the hippos are always wallowing out the mud and sediment? Or is it shallow with a soft bottom, because there are lots of hippos defecting inside of it all the time? How do the pools compare to the river reaches up and downstream? Wouldn't you love to know what is happening at the bottom of a hippo pool? So would we! But it's hard to study a place you can't get into, and it's hard to get into a place inhabited by a number of 1,500 kg animals widely considered to be the most dangerous animal in Africa. 

We posed this problem to a group of robotics engineers from Carnegie Mellon University who have started an innovative company called Platypus. They design robotic boats that can autonomously create detailed spatiotemporal maps of water quality parameters. There are lots of uses for these boats, such as sampling in otherwise difficult-to-sample areas. Well, we have some difficult-to-sample areas! 

After lots of discussions, the folks from Platypus agreed to come to the Mara to spend two weeks with us in the field, using some of their robotic boats to map the insides and bottoms of hippo pools in the river. We warned them that some of their boats may not survive the trip, but they were up for the challenge! 

John and Chris from Platypus rigging up their robotic boat
To help the odds of boat survival, on the advice of a Maasai fellow, they decided to disguise the boat as a crocodile. They found this amazing foam croc head, which we attached to the front of the boat, and we covered the body of the boat with a camouflaged tarp. Other than the depth sensor and camera mounted on the back, it looked pretty convincing cruising through the water!
Robotic crocodile cruising the Mara
The first day they were here, we took the boats out for a test run on a stretch of river without any hippos. The boat looked great and worked well, but the real question is, will it survive encounters with hippos and crocodiles?
Deploying the boat for a test run
 I have to admit, I have no idea if this will work, and I'm both nervous and excited about it. If all their boats get eaten right away, this will be a long and awkward two weeks together. But if they work, we're going to discover some of the hidden secrets of the bottom of the Mara River, and gain invaluable knowledge about how hippos influence river ecosystems. Fingers crossed for the latter!

Saturday, March 8, 2014

Closing the streams

After about a month of running our experimental streams, it was time to finish out the experiment. The main thing we wanted to measure was how much biofilm and of what different types grew in the streams under different treatments. Biofilms are groups of microorganisms which grow on a surface. We usually think of biofilms as being comprised of algae, but they can also have a lot of bacteria in them as well.

I have to admit, I was a little worried nothing would grow, which would make it difficult to measure any effects of the treatments. Fortunately, that was an unnecessary fear! After 4 weeks of running, it was incredible how much biofilm our streams had grown! You could clearly see it on both the tiles and gravel we had placed in all the streams.

Biofilm growth in tiles in the streams 
Biofilm growth on rocks in the streams
 A lot of the biofilms were green and looked like algae, like this rock below, but we took measurements that will help us determine the quantity and composition of the biofilms in each stream.

A rock covered with algae
 We will have to analyze all our samples and run statistical tests on the data before we will know if the experiment "worked" scientifically speaking-- that is, did the treatments have significantly different effects on biofilm development that we could measure? However, after running smoothly for one month and growing lots of great biofilms, I can happily say the experiment was a success! When we started this process, we didn't even know if we could get these things to hold water, much less act like small streams. Truly, we owe all the success of this experiment to Paul Geemi, our tireless field assistant.

Geemi changing water in the streams
Geemi camped next to the streams for one and half months, just to help maintain them and keep them running. He changed out 50% of the water in each stream every day-- that's a lot of pumping water and filling and carrying buckets! He monitored the power that kept the paddlewheels turning and switched things over to battery when there were power outages, even in the middle of the night. He also did a great job doing outreach and education for the many people who came by and wanted to learn about the experiment. And through it all, he kept the most thorough and well-written field notes I have ever seen!

Geemi doing outreach about the streams
Field biology is full of unsung heroes-- the people actually out there in the heat and the rain and the dark, doing the work, carrying the heavy weight, monitoring the results, keeping things working. Often times, these are not the same people who end up analyzing and writing up the data, so they don't always get a lot of credit for the work they do. Geemi is definitely the reason these streams worked as well as they did (or even at all!), and we are super grateful to be able to work with him. Thanks Geemi!

Geemi, the unsung hero of the experimental streams

It all comes back to silica

After Emma and David left, we hosted two colleagues from Antwerp University, Belgium, in the Mara. Jonas Schoelynck and Eric Struyf study silica cycling and are particularly interested in the role animals might play in the silica cycle. 

What is silica? It's commonly found in nature as quartz, but it can also be found in living organisms like grass (it gives grass the sharpness that can cut you). Silica is taken up by diatoms in the ocean, that are at the base of the food web, and it is often limiting to their growth. That it, the more silica there is, the more diatoms there are, and the more overall productivity in that region of the ocean. A lot of the silica in the ocean is brought in by rivers, and it has generally been assumed that all of that silica originated from weathering of rocks. But recent advances in the last decade have shown that animals can actually contribute a great deal of silica by eating grass and defecating it out, and the silica defecated by animals is more reactive than that simply eroded form rocks. If it involves animals eating and defecating grass, we have the perfect animal to study...

Hippos basking in the Mara River

Jonas and Eric came to the Mara to work with us to investigate the role hippos might be playing in the silica cycle. To study this, they collected water and sediment samples from multiple reaches of the river with varying densities of hippos.

Eric collecting a pore water sample from the Mara

Jonas and Eric collecting a sediment core from the riverbank
We did this throughout a 100 km reach of the Mara, which was a great opportunity to see some beautiful places, and to learn a lot about silica.

Jonas and Eric on the banks of the Mara

They also took sediment samples in the middle of the grasslands for comparison.

Jonas and Eric collecting a sediment core in the savanna
These guys are super passionate about their subject and a lot of fun to work with, and after a week in the field together, they pretty much convinced me that everything comes back to silica cycling! Best of all, they took a system and an animal about which we have thought a lot, and introduced a whole new component to think about that we hadn't considered before. It will be really exciting to see what these new analyses can teach us about the Mara.

Friday, March 7, 2014

If optical brighteners are in the Talek, shouldn't the water be "whiter and brighter"?

After spending two great weeks with us, David had to return to the US, but we had one more week in the field with Emma. During this time, we spent some more time investigating the Talek River and its tributaries. A lot of Emma's research is focused on understanding the impact of humans on river ecosystems, and she is working with us to untangle the role of urban developments vs. tourism facilities vs. hippo inputs in the Talek River and its tributaries. This is an important river system to understand because it is the main tributary in the middle reaches of the Mara, and it enters the Mara River in the middle of the reserve. However, water quality in the Talek can be very poor, which may have a big impact on the water quality we see in the downstream portion of the Mara.

Build-up of rotting hippo feces and in a tributary of the Talek
As one tool in our toolbox, we worked on a new method that can be used to detect something called "optical brighteners" in river water. I had never heard of optical brighteners until recently, but they are in many kinds of laundry detergent, and are what make your clothes look "whiter and brighter." It's not that optical brighteners are necessarily bad for the river, but if they are detectable in a river, they indicate the presence of large human inputs of other kinds, like nutrients and sewage. 

Emma and Chris collecting water samples in the Ntiaktiak River
We collected water samples along a transect down the Talek River and its tributaries and into the Mara. Then we compared the fluorescence of the water samples at a certain wavelength to a solution made completely of Omo (the most popular detergent in Kenya). Then, because optical brighteners degrade in UV light, we exposed the water samples to UV light and tested them again. The samples with declining fluorescence after UV exposure were positive for optical brighteners.

Sure enough, we did find optical brighteners present at certain sites in the river, but not necessarily where we expected them. This is an exciting new tool to have, but as usual with new methods, it only raised more questions for us.

Emma testing for optical brighteners
After we finished that round of surveys, we started some hippo pool surveys. Because river flows are so lo right now, it's a great time to do them, and we could get Emma's insight on our methods.

Collecting hippo feces from hippo pool surveys
With all this exciting stuff going on, we were actually in the field working until one hour before Emma's flight departed out of the Mara. I think that was a record! We had such a great visit with both Emma and David, and feel very fortunate they were both able to spend so much time in the field with us this year. Looking forward to many years ahead!

Amanda and Emma at the Mara Serena airstrip

How to flush a hippo pool

I know our last post was several weeks ago, so you all probably think we have been off celebrating ever since getting our experimental streams up and running... Well, we did do some celebrating, but we've also been super busy with a lot of other cool science work. Having the streams work well just opened up a lot of time for us to do other things. 

One of the things we have been particularly excited about it trying to re-create the dissolved oxygen (DO) crashes we have seen in the Mara. We have developed the hypothesis that it is the flushing of a lot of water through hippo pools that causes these events, but how can you test that experimentally? If we could flush a lot of water through a hippo pool without adding or changing anything else, and if that movement of water alone caused a DO crash, that would be pretty convincing, right? But how would one pull that off?

We toyed around with hiring a huge water truck to dump a bunch of water into a hippo pool, but there were some logistical challenges to consider. In running this idea by our hyena researcher colleagues, David Green suggested, "Why don't you just build a small dam and then remove it quickly?" 

Brilliant! Six tons of sand later and we were in business!

This is what 6 tons of sand looks like
First we found a narrow reach of small tributary we thought we could reasonably block off. Then we just started piling up sandbags.

David showing off his dam-building skills
 After several hours of work, we had constructed a pretty good dam! It was pretty funny to watch Emma and David working on this dam, because they have both done research on the harmful effects dams can have on river ecosystems, but knowing it was temporary made a big difference!

Our completed dam
 We left the dam overnight so it could back up as much water as possible. By the next morning, it had built up almost a 1 foot head, which reach back at least 100 m. That's a lot of water!

Our dam the following morning
After taking lots of measurements and getting our meters spread out downstream, we pulled all the sandbags out of the middle of the dam quickly, letting the water rush through. I was amazed by how big a wall of water we had managed to build up!

Breaching the dam
Chris will be presenting on this experiment and what we learned from it at the Joint Aquatic Sciences Meeting this May, so stay tuned to learn more about our results. Did we re-create a DO crash, or just build a lot of muscles moving six tons of sand around? I'm going to leave you with a little bit of suspense... but I will say that, 24 hours after we started, the river was completely back to normal, but we had learned so much about how these events work! Thanks to David Green for the great idea, and to David, Emma, Chris and the sand guys from Talek for all the hard work!

24 hours later

Friday, February 7, 2014

Success!

After several months of planning, several weeks of construction, several long days of set-up, and lots of hard work and ingenuity by all involved, we finally got the streams running smoothly! It was so exciting to finally see the whole array spinning along, 12 little streams ready to grow biofilms under different treatments and hopefully tell us something about how inputs from large wildlife influence the way the whole river functions.


After getting the experiment started and running smoothly, we finally made it down to our main camp in the Mara, where we celebrated our success with a sundowner!

Celebratory sundowner!

Thursday, February 6, 2014

Leaks!

Again, I can only blame irrational optimism for not thinking we might have some trouble with leaks from streams made out of PVC canvas. Chris and I had done a test on one of the streams, by filling it with water and leaving it outside for a couple of days. When it still held water, we were excited and moved on. But filling 12 streams sitting on bare ground with a precise amount of water was whole different level of testing. 

Just as we were wrapping up our set-up and getting ready to officially start the experiment, we noticed water slowly leaching out of about half of the streams and onto the bare ground around them. We had rulers in each stream, so we could see how much the level was falling in each one. There was some banging of heads against the wall (mine), and some reassuring this was an easy fix (Chris), and some general debate about the importance of this factor (all) before deciding we needed to try to patch all the leaks... in our streams already filled with gravel and water. 

David and Emma watching a stream leak
First we had to pump enough water out of each leaking stream to be able to dry the leaking area, and keep track of that volume of water so we could replace it once the leak was fixed.

David pumping water out of a stream to patch the leak
 Then we patched the inside of the streams with liquid electrical tape and the outside with waterproof silicon and waited for it to dry.

Emma waiting for the silicon patches to dry
 Then we re-filled all the streams and waited to see if the leaks had stopped...

Waiting...