Thursday, March 20, 2014


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...

Friday, March 14, 2014

The Hippo Test

After our test run with the robotic boat the other day, we were feeling pretty confident the boats would be fine in the river... it was just the hippos we were nervous about. Hippos can be pretty territorial about their pools, even attacking one another and sometimes pushing subordinant males out of the pool completely, with plenty of deep scars to show for it. While hippo skin is tough enough to sustain the sharp tusks of their fellow pool-mates with usually minimal damage to the animal, we didn't think our boat could handle a bite. Our goal was to avoid a hippo interaction altogether, which we hoped the crocodile disguise would help with. Hippos and crocodiles share pools, albeit sometimes grudgingly, but all we needed was for a hippo to think twice before attacking the boat.

We went down to the water's edge and gently set our craft afloat. I could feel my stomach in my throat, watching this boat with all this important scientific equipment being sent out into such dangerous waters. 

Of course, it wasn't just the boat that was in a precarious position. Going down to the water's edge at a hippo pool can be dangerous, due to both crocs and hippos, so we hired an armed ranger from the Mara Conservancy each day to accompany us. Still, it made me hyper-vigilant every time one of our guests approached the water, and I was constantly urging them to minimize their time at the water's edge.

The hippos were definitely curious about the boat, and they kept popping their heads above water and blowing forcefully out of their nostrils, as if issuing a warning to the "crocodile" not to get too close. Fortunately, the hippos also congregated closer and closer, allowing us to get a pretty thorough survey of the majority of the pool. We managed to get within a few meters of the hippos, but we decided not to push our luck any further.

It wasn't just the hippos that were curious-- we also had a vervet monkey climb out on a tree limb over the river in order to get a closer look at this strange creature!

Everything went remarkably well, and we surveyed hippo pool after hippo pool, increasingly confident in the boat's ability to deter the hippos, and our boat driver's ability to avoid the hippos. Of course, just when you're getting comfortable is when nature will choose to remind you who's really in charge. We were surveying the last hippo pool of the day, which was down inside a deeply incised river channel. The hippos at this pool seemed especially ornery, and we had to launch the boat about 100 meters downstream of the hippo pool in order to approach the river safely. It was difficult to walk down to the river and launch the boat, because the banks were so steep and the edge of the riverbank dropped off quickly. All of this contributed to a general sense of unease as we watched our pilot slowly navigate the boat upstream. 

As the boat approached the pod of hippos, suddenly one huge male broke away from the pod and came lunging towards the boat. The pilot cranked up the speed on the boat, but the hippo kept gaining ground. We all watched breathlessly to see if the boat would be able to escape, and suddenly, something seemed to spook the hippo, and he made one last giant splashing lunge towards the boat and then turned around. The boat zoomed away, and the hippo grudgingly returned to his pod. 

The video of this event is pretty exciting to watch, and it got a lot of coverage online. You can see a list of links to stories about our work and a video of the chase here

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


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


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...


Wednesday, February 5, 2014

Building the streams

Setting up our experimental stream array turned out to be no small feat! In some bizarre moment of optimism, I had scheduled one day to get the streams up and running. Impressively, it only took two and half days, which I still consider a win. There were a number of steps along the way, and I don't think we could have done it without everyone's contributions and help. 

There were a number of opportunities for innovation and ingenuity, like when we realized we had the wrong couplers for our shafts, and David and Emma improvised a solution out of electrical tape and tubing. Or when we realized that all of the electronics we had ordered to control the motors to turn the paddlewheels came without any wiring, and Chris ordered wiring from Nairobi and spent the day figuring out how to wire US devices to an unreliable Kenyan power supply.

Chris wiring our motors
There was also lots of just hard work, like having to measure out the exact number of liters of water to add to every stream and the hand carry it with buckets.

Emma measuring out water
Chris and Emma filling buckets
David filling streams
Or realizing the ground had to be perfectly level to have a 6 m long shaft turn 6 paddlewheels at equal depths, and having to adjust the ground under each stream by hand, despite the risk of cuts from the sharp flashing edges.
WRUA members helping with the setup
But by the end of two days, we had all the streams set up and filled with water, the motors running, and the paddlewheels circulating the water around the streams. But just when we were ready to declare victory, an unexpected problem arose...

Tuesday, February 4, 2014

Emma in the Mara!

After David had been in the Mara for one week, one of my other professors, Dr. Emma Rosi-Marshall, flew in for a two week visit as well. Emma is a scientist at the Cary Institute of Ecosystem Studies, and her research focuses on large river ecology and the influences of human inputs on river ecosystems. Emma's knowledge and experience working on rivers has been invaluable in helping us learn how to study the Mara, and her research on human inputs is critical to helping us sort out the effects of hippos versus people on the river.

Emma also has an experimental stream array she designed at the Cary Institute, and we based our experimental streams in the Mara on her designs, so it was great to have her here to help out with the start of the experiment.

Michael, Amanda, Emma, and David picking up the stream parts
Emma's first stop in Kenya was at a welding shop in Narok, where a very skilled welder named Michael Otieno had crafted paddlewheels, shafts, motor mounts and ball bearing stands to impressive precision out of simple materials based on some sketches in my notebook. Not the most glamorous way to be welcomed to the Mara, but as an avid river ecologist, Emma was excited to see the Kenyan version of her streams coming to life. By the end of the her first day in Kenya, all the pieces were in place to start actually building our streams.

Monday, February 3, 2014

David in the Mara!

The last few weeks have been some of the busiest, but most fun and successful, of our entire field season. My advisor, Dr. David Post, came in from Yale in mid-January to spend two weeks in the field with us. David visited us in the Mara in 2011, and his insights on the system were invaluable in helping shape our research questions, so we were really excited to have back in the Mara to see what we had accomplished since then. 

David was excited to see some more of the Mara than he had seen during his last visit, and I think we delivered. Within his first week in Kenya, we had driven 1,600 km (mostly on dusty, bumpy dirt roads) and he had only spent two nights in the same place once.

David Post and Amanda at the Rift Valley overlook
Our primary goal during David's first week in Kenya was getting the experimental streams up and running. After determining that running the streams off of solar power would be cost-prohibitive, we decided to run the experiment at the Water Resource User's Association (WRUA) in Mulot. This location has several advantages. It has a main source of power, a water pump and tank system already in place, good security, and a location for Geemi to camp while he helps out with this experiment. Most importantly, it allows us to work with the WRUA and inform them more about our research in the Mara. The WRUA has been an active and important force for improved water and land management in the Mara for a number of years now, and they just recently won the Thiess International Riverprize!

David talking with some of the WRUA members about the streams
By the end of David's first week, we had laid the groundwork for not one, but two exciting experiments to undertake in the Mara! More details to come...

Saturday, January 18, 2014

Last minute

Yes, that's our gearbox on the road. Trying to fix a major leak before David Post arrives in a few hours.