Sunday, March 15, 2015

Hello Instant Wild!

"They are going to stand right here"
My name is Frank.  My wife, Kate and I run the camera trap project in Costa Rica.  I'd really like to thank ZSL , Edge, and Instant Wild for allowing us to be part of such a great idea.  I thought that you may be interested in knowing a bit about camera trapping in the jungle, these particular wireless cameras, both camera trap locations, and finally about some of the animals.  We are currently in South Carolina in the United States and get our images via email seconds before they are posted on Instant Wild.

About the Ocho Verde Camera Trap Project

We acquired the property in 2010. It had been classified as a preserve by the original owner who was interested in maintaining a population of the Endangered Red-Backed Squirrel Monkeys. He did this by planting fruit trees in the cleared areas and allowing the secondary and primary jungle continue to develop.  He probably did know it at the time, but the squirrel monkeys actually do better in recovering secondary rainforest as opposed to primary rainforest.  Unfortunately when the property changed hands, the new owners soon lost interest in the property and the land was hunted for Pacas, Curassows, and Tinamous.  Even some of the large trees were slashed and 'milked' for a medicinal sap. When we arrived, we brought in a full time caretaker family who had a non violent way to curb trespassers and poachers.   

Prior to our first visit as property owners, Kate reached out to an now defunct organization called Yaguara who were camera trapping all over this part of Costa Rica. We could not offer much financial support to their organization, but we were able to transport 8 camera traps for them duty-free from the United States.  We even purchased a camera trap for ourselves.  On Christmas day 2010, we looked at the memory card and had a visit from an Ocelot.  That was exciting, but we wanted to find out what else lived on the property and were animals recovering from being poached.  In 2013, we launched and Indiegogo project that helped us buy 12 cameras that were spread throughout the property.  We learned about so much more. We learned patterns of activity, movements by individuals and that we had a great amount of biodiversity.

Camera Traps in the Jungle
If you have ever been to a tropical rainforest one thing that you will notice is that it is dark.  The thick canopy of trees obscures so much light that photography with a DSLR, point and shoot, or camera phone is a real challenge.  Most contemporary camera traps were designed to capture images of feeding deer and not see animals in the dark wet humid jungle. Those who know about photography realize that in low light you need long shutter speeds. Long shutter speeds give blurry images when there is movement.  That is one of the image conundrums we are faced with in dense jungle camera trap photography.

The ants are beginning to obscure the lens. (jaguarundi)

A few days later this was the last recognizable image. (raccoon)
We have some other problems being in the jungle.  Rain can get on lenses making for strange and blurry images.  Worse still, bad rubber seals allow water to actually enter into the cameras short circuits occur and render them useless.  Plants can grow very fast in the rainforest and often times a plant will grow smack in the middle of our frame.  This obscures animals and sends off false triggers that eventually drain batteries.  We also have problems with ants and spiders.  A species of tiny spider finds the little hood over the lens a perfect place to spin a web.  It is a nice and dry space, but the web obscures our view of all animals.  The ants will eat the rubber gasket where the camera door closes and lets water in.  They can also just invade the camera and eventually close off the viewing portal.  Cameras often quit working for no apparent reason and sometimes batteries inexplicably corrode.

About these Cameras
The Spartan Camera(showing adjustable antenna) and my iPhone 5 for scale.

Like all of the cameras that I have seen on Instant Wild, these are made by HCO Scoutguard.  They are called Spartan cameras. They were new in the Fall of 2014.  I don't have the ability to name the cameras on the images.  The cameras use infrared flash for night images.  There is not a white flash wireless camera on the market yet, but there will be in another month or so(fingers crossed).

The cameras use 2g (second generation) cellular technology.  There is currently 3g and 4g available.  We use Spanish telecom operator Movistar as our provider as it is the only one that is within range of the cameras as of now.  The Spartan can be set to send images via text or via email.  They are currently set to send email images.  

So what happens when an animal triggers the camera?  The camera records two images at 8mp to the internal SD card(see top image for a card saved sample).  The second image is then compressed into a small file and sent to me and the Instant Wild image address.  The Spartan can send larger images, but that will use slightly more battery power and twice as much data as the current images.  We opted for the small images as we were unsure of data costs and battery drain, but this can be adjusted in the future.

The cameras have a neat feature called SMS.  This gives one the ability to control different aspects of the camera by calling a number for the phone and then entering in commands via text.  One can tell the camera to take a picture and send it right a way or enter other email addresses, etc. We opted to turn off SMS control as we were told it would consume much more battery power.  We won't be back in Costa Rica for a few more months and want the batteries to last a long time. 

We have had these cameras running since the end of January.  They were a challenge to set up as they had to be programmed with a PC and we are Mac people.  Movistar had never heard of these remote cameras so we had to teach them about needing a sim card, but no phone plan as we were only going to be using data for email transmissions.  The cell signal can be spotty in Costa Rica. If the camera is ready to send an image and cannot get a signal it aborts the attempt to send after a few tries.  Fortunately, the images are held on the SD card for us to view later.

The M in the information strip tells us that the camera was MOTION activated as opposed to another type of trigger command.  The P5 tells us how much battery power is remaining in the cameras. P5 indicates full power.  It should drop to P4---->P1 as the batteries run down.  The current batteries are AA Lithium batteries(12 of them) that we put into the cameras on February 11. The info strip also tells us the current phase of the moon as well as time and date and temperature.

One of the cameras was purchased with funds from a donor and the other was generously offered to us by HCO Scoutguard. A note about the Spartan camera shown above.  The model pictured has a blackout screen over the infrared LED's.  The two cameras that we are using does not have the blackout screen.

About the Locations
Images from this camera arrive from OCHOVERDE

Images from this camera arrive from ochoverde.

We chose the locations for the cameras due to the sheer numbers of animals that passed through the jungle at these locations.  The only animals that we have not seen in these locations are the Neotropical River Otter, some of the birds, and the wrong people.  We learned that these areas are very good because people trails and animal trails intersect. For identification purposes we'll call each camera OV and ov.

The OV camera is at a spot that we call "Tigrillo Woods" since it was where we have seen the most ocelots(tigrillos).  In the upper right corner is part of a huge whorled vine known as 'mi favorite'.  'Mi favorite' is the favorite vine of our caretaker.  It can also be seen numerous time in the videos and images in our previous blogs.  As you may recall from the Instant Wild images, the animals usually cross right to left or vice-versa.  That is precisely how the person trail is oriented.  There several 'animal' trails here as well as you can see in this crude schematic:
There are 5 possible ways for animals to enter frame.
 I considered placing the camera on the tree to the far right, but it would have to look up the trail and is looking west.  Cameras in that spot in the past get rain drops on lenses and light flares from the sun.  The Spartan cameras have a bit of a telephoto lens, which I'm not a fan, that tends to compress the field of view.  Across the trail from the camera is a fairly steep embankment.  In actuality, the animals often walk over and climb onto that vine instead of going up the embankment.

The other location where camera ov is placed is called 'Tres Hermanas', three sisters.  The 3 large trees are the three sisters.  We have had cameras here for over 2 years and again it is a confluence of people and animal trails.  Last year the top of the tree above the buttress on the left fell and gave us this nice log.  This log actually gave us a new animal trail.  To the left side are the branches of the tree where many animals feed on lizards and insects, etc.  The camera is not there because of the intense rays of sunlight during most of the day which makes it difficult to see the animals.  Here is how the trails lay out in another of my crude schematics:
The camera location is on the tree marked C.

If the truth were told, I would have placed the camera on the trees(marked with red circles) on one or the other side of tree #1.  I did have it placed on the small tree between #1 and #2 and it gave us great images:
Current camera location is off frame on tree past the one on far right(female Curassow!)
The other camera location is one that we had a camera at for a long time as well.  It mostly recorded videos that you can see in other blogs.  Here is a decent screen capture of the view from that location--it is on the tree to the left of tree #1:
Camera ov is currently placed on the viny tree behind the anteater/tamandua.
Again, this spot gave us a nice wide viewing area.  So, why did I decide to place the camera where I did.  There are a couple of reasons.  The main one being that tree #2 is about to lose another huge section of trunk.  It looks to me as if it will fall directly on top of each of our other camera location options, when it does fall.  The second reason is that there is good evidence of animal activity on the trunk of the tree in our current frame.  We think as this tree decays, it could be a popular feeding spot.  ov very well could get smashed when sister #2 falls so we'll have to keep that in mind.

ochoverde UPDATE MARCH 17, 2015
It looks as if a large chunk of tree has fallen to the left of sister #1. 

Empty Images

I really dislike empty images.  There are several reasons for them.  The most common are that the animals are moving fast.  The Spartan are reported to have a 1.2 second trigger speed which is considered slow in the camera trap world, but given the other advancements, we'll give it a pass. There are not a lot of reasons for animals to be moving fast.  One is that it is chasing something or being chased.  Another is that the camera is triggered by a flying bird or bat.  It could be a Tayra which is always on the move.  Trees and leaves moving in the wind can trigger the camera.  Hot spots can trigger them as well.  A hot spot is an intense area of sunlight that heats up more than the rest of the area around it. A hot spot example is in one of the empty ov images above as well.  The animals can be just off frame and trigger the cameras. Sometimes it is a mystery trigger.  The woods at Ocho Verde are indeed mysterious with many strange unexplained happenings.  Check out our "Mystery Smoke" video: 
I have no idea what triggered this.  If it is a person smoking early in the morning in the jungle, the cameras all around missed him coming and going.   He also would have had to kneel or sit as the camera is just a foot or so off the ground. He then would have let the smoke drift into frame.  Spooky.
The best cure for being blanked completely are multiple or 'burst' images. The cameras are programmed to take two pictures about 1 second apart.  This is done to insure that we get at least one image.  Unfortunately, the Spartan is programmed to send the second image instead of the first image.  I would change that if I could. The good news is that our blank images will have a companion image to give another view of the animal or to see what triggered our camera.  We also have non-wireless cameras very close by.  I will be able to sync up the times and we will ID the missing animals and those we could not recognize in the Instant Wild images. We will just have to wait a while to find out for sure.
Identifying Animals
The internet is full of strange and wonderful pictures of people captured at exactly the wrong times.  The same is true of camera trap images.  Animals in weird positions can be tricky to recognize.  People routinely mis-identify tayras and jaguarundis.  They are close in size, color, habits, and movement.  It is often the slender tail that clearly separates the two animals in a cam trap image.  When ID'ing animals, the guides are great resources.   Pacas and Agoutis look a lot alike at certain times.  However, you rarely see Pacas during daylight or Agoutis in hard darkness. Sure there are exceptions, but it is not often. Knowing an animals habits is part of the identification process.

Often times when looking through images, I'll have a few "What the hell is that?" statements.  They are usually figured out in images before or after when we get another piece of the puzzle to add to the equation.  The wireless Spartans give us one chance at the moment. We have to go on animal behavior and physical characteristics.  Even then we can't always be sure.  Take this image from a few weeks ago:

 is it this Tayra on the same log...

 or this capuchin monkey on the same log?

Hard to say, right?  We are leaning toward capuchin monkey.
What about this image...

This could be a Tayra, Jaguarundi, or labrador.  However, we have never seen a lab on the property nor are they very common in our part of Costa Rica...let alone a well fed lab.  The Tayras on the property are pretty healthy. The thin tail suggests that it could be a Jaguarundi.  We won't know for sure until we see the second image or even other images from the companion cameras to the left and right of our Spartan.

Someone asked about ocelot recognition.  We are doing that along with some paca(side patterns) and tamandua(tail patterns).  To date we have followed Crooked Tale/Roi, a female ocelot and her cub Pizza slice.  We saw Pizza Slice on the OV camera last week.  Ocelot recognition is addressed in this blog from last year.

Thanks for reading this far.  I'll be waiting for the next click of the camera...just like you!

If you have wandered onto my blog by accident and want to know what's being discussed, be sure to check out Instant Wild and their live camera trap images: LINK TO INSTANT WILD!   

Thursday, March 5, 2015

Hello and Happy 2015.

Roi, aka Crooked Tail, with a big belly.

We have big news to report about our project a little later in our update.

We have just returned from Costa Rica where we downloaded nearly 9000 images and videos that had been recorded since late July. This many images is a bit out of whack for our time-frame, but we had 2 cameras that recorded continuously until the batteries died.  One was due to a short due from high humidity, the other was due to a few blades of grass that popped up in front of the sensor and triggered it every time a breeze blew.  We also had one camera that developed a corrupt memory card and sadly I could not download any of the photos.  If you know of an easy way to retrieve them, send me an email and we can take a look at the images.

Is a cub on the way, for Roi?

On the animal front, we are seeing pretty much what we thought that we would.  Crooked Tail(Roi) is still around as is her cub Pizza Slice.  Crooked Tail has a bit of a saggy belly in the later pictures indicating that she may have given birth in the last few months.  This would go along with the fact that we have seen far few male ocelots recently. Pregnant or parental female ocelots do not attract mates until the young are no longer with them. We are hoping to see her with a cub later this month.

This blurry because a) it is from a video and b) Tayras are always on the move.
Our macho male Tayra is still around, but this time he is with a love in the air?  I had a chance encounter with the female in early January on a trail.  She climbed a tree groaned at me to continue to back away, which I did.  

The Pacas seem to be abundant.  The agoutis are common as they ever were at one point one occupied 350 of nearly 400 videos from a single camera.  There are plenty of ant eaters.  The Jaguarundi seem to be around as well although only singles were seen this past Fall.  The Capuchin monkeys came down to several cameras and the coatis were seen at nearly every location.  The hog-nosed striped skunk was seen more often and this may indicate him being more of a resident than passer by.  The Curassows have been strutting around like they have owned the forest.  One new animal that we had not seen was the 4 Eyed Opossum.  It seems to have a little more bounce to it’s step that the larger Common Opossum.  

Here is video quick recap:

We had 9 cameras that failed and had to be returned to the states. One new Cuddeback camera failed 3 weeks into the new year.  Ants caused a few of the failures and I believe that the rest was due to the humidity of extra rainfall in November.  One camera was full of water from a bad seal, however, the images were just fine. 
The cameras were all brought back to the states for warranty repair(I hope).

We have been invited to join the Osa Conservation Camera Trap Network.  Project director Juan Carlos Cruz came over twice to help get us coordinated with the data input and looked at the trails.  We are part of a 14 station network that helps monitor the wildlife in the Osa Region of Costa Rica.  Read more about the project here:
Osa Conservation Camera Trap Network

We are currently on the eastern edge of the network's boundary as it makes it way down towards Panama.

We were also invited to share our data(nearly 30,000 images) with the MAPCOBIO Project which records biodiversity from camera trap projects from all over Costa Rica. 

In August, we received two Spartan wireless cameras, one from a private donor and one from the camera manufacturer HCO Scoutguard.  With a month fiddling with the cameras, we finally had the cameras up and working.  We are able to receive nearly instantaneous email images from our cameras. As we sat on the plane in Miami, this shot of an iguana dead center of the frame, appeared on my phone. 

 Using this technology we have become one of the very first to receive live images from Central America in the United States.  This technology has been incorporated into an app by the  Zoological Society of London.  “Instant Wild” uses wireless camera trap images from select parts of the world to increase awareness of biodiversity.  Now, we can all see what is going on in our forest along with everybody else.  Here is the link to Instant Wild

I'm going to do a special blog on the 2 wireless Instant Wild cameras in the next week or two.  I'll discuss some of the remarkable technology within these cameras.

In the meantime, something weird is happening with Ocho Verde Facebook page, so until I get that figured out, you can

Tayra and Banana