Friday, June 22, 2012

Charlie's First Meal After Returning Home

In the post Intimate Encounters with the Northern Goshawk, I told the story of a nestling goshawk's return to its nest. Here is some video captured from his first meal after returning home. Charlie is the smallest of the four nestlings, but after being left out at first, dives right into the middle to ensure that he gets his share! Awesome! It is also cool that the adult male brings in a second Belding's Ground Squirrel while they are still eating the first. (The video is presented in double time).

Wednesday, June 20, 2012

Intimate Encounters with the Northern Goshawk

Welcome to the third update for this year's field season studying the Northern Goshawk for my Master's thesis in Raptor Biology. My field assistants and I have just returned from nine days in the field with a new collection of adventures! The season continues to be a huge success. It is hard to believe the season is over half way completed with only three weeks left in the field!

Adult female Northern Goshawk. South Hills Idaho, 2012.

We have now discovered and are monitoring 14 occupied goshawk nests! This is four more than I had last year. During my last post I indicated a possible issue with the late season snow storm. During that snow, we were only able to confirm that four of the ten nests were still occupied.

Goshawk nest buried in snow. (Photo: Emmy Tyrrell).

The good news is that all of the nests were still occupied after the storm. So far, we have observed that 13 of the 14 nests have successfully hatched young. The only nest where we have yet to confirm hatching is occupied by a sub-adult female. Sub-adult females are known to breed later and have much lower hatching success than full adults, but we are still hoping for success. We continue to survey for additional nests and I won't be surprised if we find one or two more.

Raptor Rob (me) broadcasting goshawk alarm call to solicit a response. (Photo: Karyn deKramer)

Nest Cameras. During this bout in the field we installed three nest cameras to help me identify and quantify the breeding season diet of the nestlings. This information is the foundation of the first chapter of my thesis on goshawk diet and provides critical support for the second chapter which deals with the influence of prey abundance on nest occupancy and success. The installation of nest cameras require a significant disturbance to the birds, something that we do not take lightly. Every measure is taken to limit our time at the nest and get out as soon as possible. We also wait until the nestlings are 8-10 days old to limit the chance of nest abandonment. Of the three other studies which have deployed a similar procedure, none have resulted in nest abandonment. This year I had a very eager volunteer to climb the trees and install the cameras - Dr. David Anderson. While I am capable and willing to climb myself (completed 9 climbs last year), I was very willing to share the excitement of climbing the trees and to learn from someone with the depth of experience that David brings. David has many years of experience climbing large trees in the tropics as the focus of both his master degree and Ph.D. degree. It was a great opportunity to see him in action. I should note that these were the first climbs that David has performed into goshawk nests.

Upon reviewing my gear David was reluctant to use the Lacrosse helmet which I specifically acquired after last year's in-nest attacks. His impression was that the extra bulk and reduced visibility were not warranted. However, the morning of his first climb he was broadcasting a goshawk alarm call to solicit a response. A response is exactly what he got! The adult goshawk swooped toward his head, circled around and hit him from behind! His hat went flying through the woods as his ears rang from the impact. As I have personally experienced this on numerous occasions, it has a way of building a deep respect for these birds and their striking force! Anyway, David decided the helmet was a pretty good idea.

David's self-portrait in the tree.
One of my encounters with an angry goshawk!

The first camera install was in a tree with a female that has to date not been very bold in defense. While she protested loudly and circled the tree, she never actually made contact with David during the climb. The team performed well, everyone fulfilling their critical roles. David climbed, I provided climbing support, Mike managed the video equipment on the ground, and Emmy took pictures and provided the most important function of all - to warn the climber when the goshawk is coming in for an attack. You don't want to be in a dynamic move when you take a direct hit from a bird!

David climbing first tree. (Photo: Emmy Tyrrell)
Me providing climbing support. Hard to look up in noon sun! (Photo Emmy Tyrrell)
David on rappel. (Photo: Emmy Tyrrell)

David didn't take photos in the first nest, but this is what he found in the second!

Three nestlings. Front left is definitely female! Look at the size of those legs! (Photo: David Anderson)

We were in and out of the territory just under the allotted time goal that we had set. We then moved on to the next nest for the second camera install. The female waited patiently on the nest as we set up for the climb. Once David moved part-way up the tree, she started her attack. Unlike the first female, this one was playing for keeps. David took many direct hits during the climb. On his back, on the helmet, wow! When he was just 3 feet under the nest, the female returned to the nest. She looked prepared to defend the nest via hand to talon combat! I had never seen this before. I warned David that this was unusual and he would have to make the call to abort if necessary. It didn't come to that as she left the nest, although she continued her aerial assault. The second camera was successfully installed!

The Rescue. David returned to earth and started to un-clip from the rope. Just then he spotted a nestling on the ground. Was it one of the three or was it a fourth nestling? Did we cause this or was it there before? It was amazing that we had not stepped on it. This produced an ethical dilemma and we had little time to consider. Should we put it back? If we had caused it, then definitely. If we had not, then returning the bird to the nest could negatively affect the surviving nestlings. Returning the bird could affect the lifetime reproductive success of the remaining birds. Hmm. None of us were going to just leave it to die. In true rockstar fashion, David once again went up the tree to return the bird. This nestling would hereafter be referred to as "Charlie" (not to be confused with "Chuck" from last year).

David depositing "Charlie" back into the nest! (Photo: Emmy Tyrrell)

We were reasonably sure, although not entirely, that we had not caused the nestling's departure from the nest as next to it on the ground was a nearly whole, decapitated Northern Flicker. Either the flicker also fell, or the parents were trying to feed the bird on the ground. Regardless, what is a 12 day old nestling going to do with a whole flicker? Anyway, we proceeded with the rescue plan. Here is the nest camera footage of "Charlie" being returned to the nest (in double time).

I am happy to report that as of four days later all four nestlings were still alive and doing fine. The third and final install went off without a hitch. Thanks again to David for putting in the tough work.

Nestlings in third camera nest. (Photo: David Anderson)

Changing Dynamics. As the season progresses, the dynamics of goshawk detections change. First, with time, the adults become more responsive to our call broadcasts. However, they also spend more and more time away from the nest increasing the chance that we will miss them at a nest or detect them somewhere else. Now that the females are spending more time away from the nest, we also generally experience many more sightings as both the males and the female are out hunting.

One of our observations was that a nesting female had a color band on her leg. This color band is consistent with the type I apply to the nestlings. Before my study, color bands were last applied to nestling goshawks in the South Hills in 2004. It was important to understand dispersal behavior to identify this bird. As a result, we spent five hours in the nest stand with a spotting scope to be able to read the numbers on her band. We were positioned far enough away so as not to disturb the bird, but close enough that a 60x scope could be used to read the band. After five hours I finally got the code. The band number reveals that this female was indeed banded in the South Hills in 2004 as a nestling. Usually females will disperse a greater distance than this (average 50km), but the island type nature of this forest forces them to stay close to home (within 20km) or move much further than the average dispersal distance (80-200km). This bird stayed home.

While watching this female and trying to read her band number, I witnessed some great interactions. First, her nestling started to stir. The adult female then left the nest in a flash and chased off the adult male which was perched nearby, as if to tell him to go get dinner. She called off and on for a few minutes. 26 minutes later the male returned with a Ruffed Grouse chick! Dinner on demand! Before I was done I would see another similar incident.

I took my good camera to get some better goshawk photos. Here are some of my results.

Female Northern Goshawk.
Female Northern Goshawk in nest in dead Aspen.
Karyn's shot of a female on a nest in live Aspen. (Photo: Karyn deKramer).
Duck! Me being bombed by a male goshawk. Shot from the hip!
Each nest is unique. This one appears new this year.
Northern Goshawk.
And another.

The Moose. The South Hills are a great place for moose. Each of us have had a number of encounters with these beasts so far. A nice bull welcomed us back to our cabin.

David posing with bull moose.

I even had one on a prey survey! Do you think a goshawk could take a moose? This guy paralleled my path and then turned toward me. I was at least 500 meters from the nearest tree. Wow, for a while there I felt somewhat vulnerable. I tried to look big and he eventually waved off.

Moose on prey survey.

Karyn even got her moose. While riding down to the southern portion of my study area, I was just telling her that this was an area that I had seen moose tracks. Just that instant she calls out "Moose!" There she was with the youngest calf I have ever seen.

Moose and calf south of Monument Peak.
Quite a show!

Of course, all living things must eventually meet their demise. On another prey survey, Karyn and I happened upon this complete carcass.

Dead bull.

Then there are the other animals that we happen across: deer, elk, jack-rabbits, cottontails, coyotes, marmots, etc. When you walk randomly through the woods, who knows what you might find.

Mule deer calf I almost stepped on.
Horned Lizard.
Hundreds of types of butterflies!

Thank you for reading. My next report will be out in approximately two weeks.

Sunday, June 10, 2012

GIS Terrace Extraction

I post here quite a bit about my thesis research into the Northern Goshawk breeding ecology. However, I am also studying to complete my graduate certificate in Geographical Information Analysis. Shhhh, don't tell anyone! A wildlife biologist friend of mine once said that GIS knowledge is absolutely critical to be a good biologist, but not to tell anyone you know how to do it or that is all that you will ever do! Well, I will break the rules a bit here.

This last semester our GIS class was working to reproduce a process described in a published paper to analyze a river corridor and extract where the terraces are located within the canyon. While my focus of GIS is on ecological applications, this particular course was very heavy in geology.

Demoulin, A., B. Bovy, G. Rixhon, and Y. Cornet. 2007. An automated method to extract fluvial terraces from digital elevation models: The Vesdre valley, a case study in eastern Belgium. Geomorphology 91:51-64.

As is often the case, some of the key details to complete the process were left out of the paper. The class struggled with a few steps and ultimately we simplified the process down to what is described in Terrace Extraction I. However, I continued with the original assignment and used a bit of automation to complete the process. I co-authored this post for my professor's web site: GIS 4 Geomorphology. I am reposting it here , with a few small modifications for presentation and consistency with this blog.

Robert Miller Raptor Biologist, Boise State University
Skye Cooley G4G Editor

INTRODUCTION
GIS projects with a large number of repetitive steps are ripe for automation. Automation requires increased upfront investment in learning and time, but results are often more reliable and highly reproducible.

This case study is an example of automated terrace extraction. The study area is a 70km reach of the Okanogan River valley extending south from the US-Canada border to Omak, WA. The procedure was inspired by a paper by Demoulin et al. (2007). They developed a method of teasing out subtle fluvial terrace remnants intermittently preserved along a narrow gorge in Belgium. We developed a new method for glacial outwash terraces in Washington state.

One challenge was the fact that ArcGIS required approximately 20 steps for each reach (500m long). With a 70km river to analyze, the result is several thousand steps, each with a high risk of operator error. While the following procedure will require some modification to accommodate other geographic areas, the general concept and the code that implements it should be readily adaptable.

We used ArcGIS 10 and the free, open-source program R (www.r-project.org). R is a powerful, flexible program with tools for both vector and raster data. R’s power in this application is its elegant handling of Digital Elevation Model (DEM) data (XYZ data).

PART I: ARCGIS
1.) Download 30m Digital Elevation Model from the USGS Seamless Server (ned.usgs.gov) or another source.

2.) Load DEM into ArcGIS.

3.) Do not project the DEM. The R script assumes no projection.

4.) Use the Draw tools to create a polygon encompassing the center of the river and far enough up the canyon wall to be above any possible terraces. This should include the whole length of the river to analyze. This can be repeated to analyze the other side of the river.

5.) Convert this graphic to a shapefile (right-click on Layers > Convert graphics to features)

6.) Clip the DEM with this newly created shapefile (Data Management Tools > Raster > Raster Processing > Clip). Fill the checkbox for “Use Input Features for Clipping Geometry”, which will clip the raster to the polygon and not the rectangular polygon extent.

7.) Export this newly created clipped DEM to a GEOTIFF file (right-click DEM > Data > Export Data).

8.) Record the rectangular extent of this raster for use in customizing the R script listed below (Lat/Lon in DD units). You are now done in ArcGIS!

Clipped DEM for river Right.

PART II: R
The simplifying factor in our case is that the river runs north-south. A river running east-west would work equally as well. If the river were to run diagonally, a more advanced algorithm for the moving clip window would be required. With N-S or E-W, only one dimension of the moving clip window is required and a rectangular clip region can be used. You can always rotate the DEM in ArcGIS prior to bringing the data into R.

………………………………………………………………………………………………..
Overview of R Workflow: The high level process is outlined in the lettered steps below. I’ve added many instructional comments in the code itself (comments follow # symbols):

a.) Read in the clipped DEM from above as a GEOTIFF file.

b.) Create a pdf file to capture all of the graph output (two graphs per page, one per moving clip window). This example produces 56 graphs!

c.) Determine the size of the moving clip window and how many clips exist in the valley (56 in this case). Clip windows should be in the range of 500-1000m of river, but can extend well beyond the edge of the map in the other direction (full extent width in my case). Greater than 1000m produces too much noise and crisp terrace definitions cannot be seen. The example below uses approximately 1000m, but a few of the sections could benefit from a smaller region.

d.) For each clip region, clip the raster.

e.) Find the minimum elevation.

f.) Subtract this elevation from all points to produce a relative elevation raster (normalizes all of the graphs).

g.) Remove data points for the lowest 5 meters of relative elevation (you don’t want to count the modern floodplain as a terrace).

h.) Plot the integer slope versus the integer relative elevation.

i.) Analyze each graph for local minima to identify terraces per the Demoulin et al. paper.
………………………………………………………………………………………………..

Data processing now moves into R and is fully automated within the script below. Here is the R code which implements this procedure for our study area.

# Load the various spatial libraries that I will use.
library(raster) # key routines for reading in and clipping raster files (DEMs)
library(SDMTools)  # Includes routines for Slope and Aspect
 
# set the working directory on my local machine. 
# All data files are relative to this location
setwd("c:\\users\\wolf21m\\Desktop\\BSU\\GIA\\Terraces")
 
# Read in raster datafile from above mentioned directory
ras <- raster("newRRFinal1.tif")
 
# Either run the windows() command to open a plot window, or save to a pdf file.
# windows(record=T) # opens plot window. record=T enables scrolling thru previous graphs
# instead let's save to pdf
pdf('RiverRight.pdf', width=7.5, height=10, paper="letter", pointsize=11)
 
# arrange plots for two plots per page (two rows, one column)
par(mfrow=c(2,1))
 
# For loop repeats 56 times for this study area. Each time "i" is incremented
for (i in 1:56){
  # Setup moving clip window. Easting is constant for this example, 
  # but northing moves north to south by 0.01 degrees for each iteration (~1km).
  # remember the "i" is incremented each time thru.
  ext <- extent(-119.76, -119.375, 48.98 - i*0.01, 48.985 - i*0.01)
 
  # Clip the raster to the just established extent and keep in tmpras
  # the original ras object remains unaltered for the next pass thru.
  tmpras <- crop(ras, ext)
 
  # extract the minimum elevation from this portion of the DEM
  rmin <- minValue(tmpras)
 
  # Subtract the minimum elevation from all points in raster.
  # this time tmpras is changed.
  tmpras <- calc(tmpras, fun=function(x){x-rmin})
 
  # convert relative elevation values less than 5m to NA or NoData
  # This is to omit the flood plane from being identified as a terrace
  tmpras <- calc(tmpras, fun=function(x){x[x<5] <- NA; return(x)})
 
  # Now plot the slope versus the relative elevation
  # X-axis - round(tmpras,0) # integer value from the DEM
  # Y-axis - round(slope(tmpras, latlon=T)/27.7, 0) # calculate the slope of DEM
  # but since it is not projected, we have to divide by the cell size in meters
  # then round to an integer. Rounding isn't required for the process, but it looks better
  # xlab and ylab define the labels
  # xlim and ylim define the sclae of the axis - clips everything outside
  # cex sets the point size, pch chooses the point symbol.
  # main defines the title for each chart, paste combines strings of text. In this
  # case the chart number and the extent covered.
  plot(round(tmpras,0), round(slope(tmpras, latlon=T)/27.7, 0), xlab="Elevation(m)", ylab="Slope", 
       ylim=c(0,30), xlim=c(0,200), cex=0.5, pch=20,
       main=paste("Terrace Map #", i, " X=",ext@xmin,":",ext@xmax,", "," Y=",ext@ymin,":",ext@ymax, sep=""))
  # complete for this clip window. increment at for loop and begin again
}
 
# If opened a pdf file, then must close it before viewing. Don't call dev.off()
# if using a plot window as it will close the window.
dev.off()

Created by Pretty R at inside-R.org

Okay, now we have a 56 page PDF file with these odd looking graphs of Slope versus Relative Elevation. If you refer back to Demoulin et al. (Fig. 6), you will notice that terraces are defined by points where the slope is zero surrounded by points where the slope is very different from zero. Here is a good example (see Terrace Map #7). Terraces in the figure below are located at approximately 42m, 98m, and 165m above the river.

Here is an example that did not work out so well (see Terrace Map #19). Some terraces can be seen at approximately 80m, but what is going on at 110m or 160m? This may be cleaned up by choosing a narrower section of river (500m versus 1km). This can also be an artifact of side canyons and tributaries.

There are other methods for identification of terraces (see ‘Terrace Extraction I’ post).

Refs:

Demoulin, A., B. Bovy, G. Rixhon, and Y. Cornet. 2007. An automated method to extract fluvial terraces from digital elevation models: The Vesdre valley, a case study in eastern Belgium. Geomorphology 91:51-64.

Friday, June 08, 2012

As slow as possible, as fast as necessary!

Welcome to the second update for this year's field season studying the Northern Goshawk for my Master's thesis in Raptor Biology. My field assistants and I have just returned from nine days in the field with a new collection of adventures! The season continues to be a huge success, but there are some indications of a possible dark cloud on the horizon.

This bout of effort included the launch of new procedures for the season. We have started our prey surveys. Each team member conducts two prey surveys per day in two separate goshawk territories. The prey surveys consist of walking 750 meter transects and noting the distance to each potential prey item. Each survey requires a little over an hour to complete when you factor in transit time back from the survey. In total, we completed 41 surveys during this bout falling a bit short of our stretch goal of 50. The snow that I will describe later had a little bit of influence on that! Regardless, we do have time in the schedule to make it up.

The second new method launched May 31st, is the use of call broadcast surveys. In territories where we have not yet discovered an occupied nest, we play a recording of a goshawk alarm call to solicit a response by resident birds. The team and I have complete 121 call points! This has led to the discovery of three more occupied nests for a total of ten which matches last year's count. We have had a response in an eleventh territory, but have not yet found the nest. At this point last year I had only identified 5 nests.

Nestlings! While we have yet to get visual confirmation of hatched nestlings, we believe that nestlings have hatched in at least four of the ten occupied nests. Clear signs include the female sitting a little higher in the nest than when incubating and her being more fidgety than while incubating. If true we should get clear signs early next week.

The team. The team is really coming together. I think we have all learned a great deal about each other and respect the unique contributions that each person makes. Mike, the map guy, studies the territory maps and plans out each route meticulously. He seems to have memorized every road sign and turn in the hills. He is so focused on navigation that while sleep walking one night at 2am, he went to get his GPS; apparently fearing he would get lost in dreamland! Mike has thoroughly studied birding by ear using Larkwire, and has quickly surpassed my own abilities. Emmy could not be more unlike Mike and I. Her outgoing nature puts the two of us introverts to shame and definitely keeps us on our toes. Nothing seems to get her down. She is a great spotter having just stumbled across an occupied goshawk nest while walking in the woods. On organized searches, she is often the one to find the nest. We give her a hard time as most of our equipment issues have had some connections to her! Regardless, she jumps right in on any task and gets the job done. It really is a great team!

During this bout, we also hosted a few volunteers. Volunteers are great as they help with the large workload we have to do and usually even fix us dinner! How cool is that! Karyn showed up for a few days and helped me on prey surveys and on nest searches. Michelle J. came for a few days to march through the woods on her own in search of new territories. 11 call points, a new goshawk sighting, and some great venison pasta for dinner... awesome! Lastly, Greg and Dusty came down to try to find some old nesting structures that Greg had visited years ago. The weather didn't cooperate, but they still fixed bar-b-que elk steak and dutch oven potatoes. In the end, they did get out on questionable roads to find the nest structures we were hoping for. And we were all fat and happy!

Luck was on my side. Sometimes it just takes a second set of eyes. Mike was on a roll with the call broadcasts. One day he received a "vocal response" to a call broadcast. This almost always means that there is a nest nearby. He began searching for the nest, but was not able to find it. The bird rarely responds a second time during one day. After giving up for the day, I returned the next day to take over the search. I played a call with no response, then just happened to walk right under the nest! Sometimes, this is how it goes. Meanwhile, Mike had another response in another territory. Once again, he did not find the nest. It is much, much harder than you think! Michelle and I took over and out of sheer luck, I found the nest within five minutes. I truly walked into the area, sat my pack down, and looked up to see the nest. Unbelievable. I was on fire! But, all good things must come to an end... It was Emmy's turn to get a response and I confidently jumped at the opportunity to once again show them how it's done. Unfortunately, it didn't work out that way. Karyn and I searched for hours without a nest... Don't worry, it's there and we will find it.

Equipment woes. My female field assistants have a history of naming their ATVs. Last year, the blue ATV from the forest service was affectionately named "Old Yeller" by Lauren. This year Emmy renamed it Earth, Wind, and Fire. Late last week it became Earth, Wind, and no-Fire... (Karyn came up with this one). We, and Emmy specifically, were lucky that it chose to loose its fire while at the base. A 20 mile walk from the far reaches of the South Hills is not something that I wish on me or my team. Anyway, plans are in the works for a temporary replacement and repair. We hope the fire will soon return. Fieldwork can definitely take its toll. I think the average equipment is designed for one or two trips into the woods per year. We put it to the test day in and day out for months. The result - wear, tear, and downright failure. So far this season we have destroyed a GPS unit, ripped holes in otherwise good clothing, and my new boots, just months old, are already coming apart. Off trail hiking in rugged terrain, through brush, over rocks, etc is probably outside the standard design parameters of today's manufacturers. The human body generally gets stronger under strain, but the products that we manufacture just can't hold up. If there was a time to copy plans from nature, now would be it!

View from our cabin on June 5 mid-day. It kept snowing... (Photo: Mike Zinn)

The Adventure. Sometimes our greatest strengths can also be our greatest weaknesses. One of my strengths is that I like to complete activities. If I have a list then crossing things off that list provides me great satisfaction. This pushes me to complete assignments well ahead of schedule, keep my responsibilities on track, and can be credited for most of the success I have had in my various careers. Of course, there are situations where it might just be better to throw away the list...

This last week the South Hills were hit with a big snow storm. At 6 am it started raining. I told the team that it would all be ok. The rain would stop and we would get our tasks for the day completed. I was right, the rain did stop. It started snowing... Ok, we canceled the first prey transect, but the second one would surely be ok. It started snowing harder... Plan C: we will hike to check some nests. This we did accomplish.

Hard to spend time looking up in these conditions with snowflakes cutting our corneas.

The rest of the day was a bust. Late in the afternoon, I determined that prey surveys must also be cancelled for the following day. I really didn't like this. Only one task remained for our week - to check the status of all ten occupied nests. A plan was drawn up - Emmy would take the close nests in the truck, Mike the west-side nests on the remaining ATV. My task included a 60 mile tour on my motorcycle over the most remote roads in the South Hills. My commitment to check off this last remaining task of the week would lead to a collection of questionable decisions, which luckily turned out all right as I am here typing this post.

I left the cabin at 7am on a cool morning over snow covered roads. There was six inches of snow piled up on the cabin but only a dusting on the main highway. As I climbed in elevation toward Monument Peak, the snow started getting a little deeper. By a little deeper, I mean a lot deeper. Then there were snow drifts near the ridge. Wow, those were tough to ride through. The motorcycle would spin as I used my legs to help push through. It's worth noting that if I was still riding my Yamaha TW200 from last year, I would have been finished about 2 miles from the cabin, either stuck or headed back home. However, a Suzuki DR650 has so much raw power that it can blow through a lot of snow. This power is a blessing in many situations as the quote "when in doubt, power out" is truly something to live by. In this situation it just enabled me to get deeper into trouble. I should back up and address the title of this article. To be authorized to drive a Boise State vehicle, we were required to watch a 4x4 training video. The main rule of the video is to proceed "as slow as possible but as fast as necessary." This has become the theme of our field work as I instruct the team to approach a goshawk nest to "keep as far away as possible, but as close as necessary" to get the job done, or to causes as "little disturbance as possible, but as much as necessary". It applies well in many situations. In 4x4 work and motorcycle work, it is truly a great thing to remember. I was traveling as slow as possible to avoid a significant slide out event, yet had to maintain a critical minimum speed to bust through the snow mounds confronting me.

As I punched through snow drift after snow drift, only 5 miles into the day's task, I somehow convinced myself that once I got past Monument Peak, it would get better.... It did not... Another 15 miles along the ridge equated to another 15 miles of snow drifts to work through. In between the drifts were puddles of water frozen over. On more than one occasion I plunged my foot deep into the water to prevent a more catastrophic and intimate encounter with our planet. Nothing better than a boot full of water in 32 degree weather with 3 hours of riding ahead. But all would be ok, I was planning to take a different route back home. I kept repeating a quote from a recent movie I watched - The Best Exotic Marigold Hotel - "Everything will be all right in the end. So if it is not all right, then it is not yet the end". But what end are we talking about?!?

I arrived at the first nest I was supposed to check after about an hour and fifteen minutes of riding. I usually get here in less than half of that. Regardless, the female was brooding her young as she should. Things were looking up. I would simply drop down the east side and avoid most of the snow. I could see the green in the valley, the sun was shining, was it the end?

No. I descended the rocky slope to the valley floor to find a whole new challenge - mud and lots of it. The power of the motorcycle was evenly matched by the soul sucking quagmire I had entered. You couldn't even stand on this stuff. Inch by inch I and those fossilized dinosaur molecules in the tank pushed the bike forward. We eventually broke through and headed up the other side. I was exhausted, but surely my progress would improve. I was two hours into a three-hour-tour (hmm, that sounds familiar for some reason...).

For a while, the roads did improve as did my progress. The combination of snow on mud required the most of my senses. I was constantly battling the bike to prevent a slide out. Splash, another foot in the puddle! Bigger splash, wow, that puddle could be called a pond! My hands were frozen, my feet were frozen, and my neck and shoulders ached as I pulled into the last territory to check. There she was, a goshawk perched high on her nest as she should be. I had completed my task, but there was one thing left to do. I was still 20 miles from the cabin. It was not yet the end, and things were not yet right. One hour and 15 minutes later I pulled into home base as my team was quickly assembling a rescue plan. Each day we set a time when we expect to be back and at what time a rescue attempt should be mounted - the curfew. I had informed the team that I would be back in 2.5 to 3 hours, but not to worry until 4 hours. I pulled in at exactly 4 hours. Ouch. For better or for worse, I put my check next to the task - Complete - Nests checked. Karyn's question - "Did you learn anything from this experience?" ... Hmm... I would like to think that I have... Hmm... Probably not...

After 4 days off, next week promises new adventures as we install three nest cameras to monitor the diet of the nestlings. This involves climbing 30-50 feet up a tree while simultaneously being attacked by the adult birds. Believe me, they take their nest defense VERY seriously! We will also continue our prey surveys, call broadcasts, and monitoring the status of nests.

What is the possible dark cloud I mentioned in the introduction? Of the ten nests we checked on the final day of the week, we could only see four adults on the nests. The other nests were buried in snow. We hope under that snow were the dutiful parents keeping their eggs or nestlings warm. This is the critical stage in their development and the point at which nests commonly fail. I look forward to finding out on Monday, but only if the news is good...