Native Plants, Habitat Restoration, and Other Science Snippets from Athens, Georgia

Sunday: 25 November 2012

Forty Four Box Turtles, 2012, Part 3  -  @ 09:13:06
We'll get to the exciting stuff below, but first:

This was another of the new box turtles things I saw this year, mating. Many years ago I was doing a little weeding, late in the summer, and saw two box turtles chasing each other across the fairy ring. It was delightful, but I didn't have a camera.

But in this case I came upon these two right at the edge of the newly added study area. I decided that's where they rightly belong, so they didn't get included in this year's population estimate (next year they will). If you don't know what I'm talking about, ask - this is part of the mark and recapture thing, so I can't use data from the newly added territory this year.



I'll still add this warning: the following exciting stuff is mostly for my purposes, but you're welcome to follow along! You wouldn't believe how often I go back to the blog to retrieve numbers and thoughts from previous years. This year was an important year in box turtle observations, therefore more numbers and more thoughts. This adds to the previous three posts here, here, and here.

I'm interested in following the size of the population, to determine if our guys are in decline or otherwise. I actually think that my conservation efforts are encouraging a stable population, but it's better to confirm this. Turtles don't stay in one place necessarily, and there are some predators around.

For analyzing numbers, I'm using mark and recapture experiments. You can find more about the simple Lincoln-Peterson model that I'm using here.

I plotted two indicators to determine if numbers of new turtles discovered were plateauing by the end of the season. Once of these, the red one, is a simple count of turtles previously unseen this season (regardless of whether I'd found them in earlier years). The last three months begins to show some leveling off. Even though I found plenty of turtles Aug-Oct, they had already been seen in 2012. I was finding turtles, but they were largely old friends.



The green line (right axis) is the ratio of new turtles to total 2012 turtles. It too shows flatlining, which is not surprising since it's really the same measure as the red plot.


Notice that the red line starts to level off at 25 new turtles. That's not to say that there are only 25 turtles, it's just to say that new ones are becoming less likely to encounter. To get a measure of the full population, you do mark and recapture.

To do a mark and recapture experiment you need two visits separated in time. In the first visit you capture and mark (in my case, photograph) the turtles in the population.

This year I used the 12 unique turtles I found in 2011 as my first visit, and then all the turtles I found this year as my second visit. That's just one way I could have put the data together.

And so I've plotted that calculated population size at each turtle discovery in the plot below. The red line is one estimate based on the simple equation given here, along with the rationalization. The green line is based on a slightly modified form that is supposed to be less biased. The good thing is that it comes with an error estimate. The population size is then 44 +/- 8 turtles in the old study area.



There's one thing puzzling about the plots. We rather quickly achieved an estimate of 40 or so turtles, in early May. But then there was a rapid rise in the estimate of population size, to 55 turtles by the end of June. Then the estimate fell back to around 45 around mid July, which was maintained through the end of the season.

What's that all about? I speculated about this when I noticed it happening, and that the new turtles seemed to be mostly female (4 new females to 1 male, during this period). It may represent an influx of females into nesting areas during nesting season. They arrive, do their thing, and then leave.

Now, these mark and recapture experiments are usually done much more quickly. Net a bunch of fish and mark their fins. Return them to the environment and let them merge back in. Check back in an hour and do the same thing. With turtles we come back a year later. That's just the way it is, with turtles. I've earlier discussed my concerns about whether I'm doing mark and recapture appropriately, but have since found a number of published population estimates done pretty much this way.

So next year I'll have some additional ways I can use the data to calculate a population. I'll have at least two or three possible first visits to choose from - 2011, 2012, and portions therein. I'll also be doing my second visit in the new study area, and will have an estimate for that extra acreage.


Wednesday: 21 November 2012

Box Turtles 2012, Part 2  -  @ 11:05:12

This was one of several new things I saw with box turtles this year. The fellow below is quite fierce, isn't he? I decided that this was a dominance fight between two males. I think we have a clear winner.



Warning: the following is mostly for my purposes, but you're welcome to follow along! You wouldn't believe how often I go back to the blog to retrieve numbers and thoughts from previous years. This year was an important year in box turtle observations, therefore more numbers and more thoughts. This adds to the previous two posts here, and here.

I'm interested in following the size of the population, to determine if it's in decline or otherwise. I'm also interested in getting an idea of what proportion of the turtles are stay at home, and what part are wanderers.

So with that in mind, here are the overall numbers for box turtles from 2005-2012:

Within the enlarged study area, I've seen 49 unique turtles. This is a total of 24 males, 22 females, and 3 uncertain. This year I found 22 new turtles. (Remember that I extended the study area. The 10 turtles I'd found there were all new, not seen before, and not rediscovered again, this year.)

24 of these 49 have never been seen again, while 25 have been rediscovered at least once.

However, 15 of those one-timers were just seen for the first time this year and have hardly had a chance for rediscovery (although 6 of this year's 22 new ones were actually rediscovered at least once this year).

I went over the dates of discoveries and rediscoveries, and processed the data into this table of individuals versus the days since previous encounter. I broke them down into three major groups as indicated on the right of the table.

The first group, in red, are 4 turtles that I've found repeatedly over the last 6-7 years. One explanation for this is that they are strongly territorial and their territory fits in well with my usual routes.



The second group, in green, is of 5 turtles seen once in previous years, but not rediscovered until this year for the first time. 3 of those 5 were seen more than once in 2012 (Corey was rediscovered five times this year!). I might have suggested that these turtles don't overlap my route so much, except for those three that were seen several times this year.

The third group, in blue, the largest with 9 members, were discovered for the first time in 2012 and then rediscovered at least once again this year. Reuben, the first of the season on March 20, was rediscovered five times!

It might seem that these groups indicate how closely the turtles' territories overlap my study area. However, I think much of what I'm seeing here is artifactual, a result of my casual turtle watching up until late summer 2011, when I really began watching carefully over planned routes at much more frequent intervals.

So I'd expect most of the members of the second, and perhaps even the third group would have been rediscovered more frequently and would merge with the members of the first group.

So, too, might some of those 24 turtles that have been seen once and then never again. If we subtract out the 15 turtles found for the first time this year, that leaves 9 turtles, or 18% of the population, that really and truly may just be wandering through.

Just for fun, I plotted a histogram of the rediscovery numbers in the third column in the above table. The first bar is of the number of turtles never seen again. From that point on the bars increment every 100 days - the second is the number of turtles I've rediscovered 1-100 days later, and so forth.



27 of the 30 rediscoveries made within 100 days come from this year's encounters, 2012. I think that's a pretty solid indication that this year's numbers are heavily skewing any conclusions I might have drawn from the many fewer observations made prior to 2011.

So that's why I think there's not much significance to the apparent clustering of rediscoveries made in the first 500 days, no rediscoveries made between 500 and 900 days, and then a clustering of rediscoveries made within 900-1800 days. I think this is just a consequence of observational quality.



Monday: 19 November 2012

Witch Hazel  -  @ 08:48:56
I've loved it since I first saw it, many years ago, at Amicalola Falls in late autumn. So about 15 years ago I planted a dozen or so small witch hazels here and there. They've grown v e r y slowly, with most of them succumbing to probably too-dry too-shady conditions. Several have survived, and yesterday I caught the largest in flower:


Hamamelis virginiana is a shrub or small tree of eastern North America, as the USDA Plants map shows. Doubtless there are now planted trees all over the US, but the east is its original range.

There is a closely related species, H. vernalis, found in OK, TX, AK, and MO, so overlapping with its sister's range. If you know your Latin, you'll know that vernalis means spring. Ozark witch hazel flowers maybe a little later than H. virginiana, up into early spring.

You probably know of the medicinal uses attributed to witchhazel.


Witch hazel is the last woody plant to flower. Was this the first year this individual has flowered? Maybe, although I certainly could have missed it in earlier years. It might look spectacular here, but that was only because I waited til just the right sunlight, late in the afternoon, against the right backgrounds. Otherwise it just sort of blends in with the other autumn foliage.

Here's a very nice photo essay on witch hazel. It also shows photos of the fruits, which take about a year to develop after flowering. The seeds are ejected explosively from the fruit.



The spidery flowers have a dozen or so very narrow, long petals that look crimped along the length. There are variants with red centers. Ours are fully yellow.

Glenn said they had a "not unpleasant" odor. I've read that it is somewhat objectionable - the kind of odor that attracts pollinators like flies. And indeed a number of flies were visiting while I took photographs.



Above is the "spider's body," the business part of the flower. There are four almost green, chunky little stamens, which will dehisce as the pollen matures. And a single pistil in the center of the four stamens - not very visible here. The woody structure behind may be the remains of one of this year's mature capsules, but I think it's more likely a gall, something which witch hazel is prone to. See this entry, 14 Sep 2006, on witch hazel cone gall maker.

Here's another nice treatment of the plant, especially the flowers. I especially like the description of the stamenodes, stamens modified to produce nectar.

There are a several other shrub or tree species that are in the same family, Hamamelidaceae, as witch hazel. Witch alder (Fothergilla spp), and sweet gum (Liquidambar styraciflua) are two of the better known plants.

That made for an exciting afternoon! I'll have to see if today I can get a photographic census of the flies visiting this plant.

Saturday: 10 November 2012

Box Turtles 2012  -  @ 06:28:51
For those who have arrived late to this particular scene, the box turtle project may need a little introduction. It's become something of a passion of mine, this idea of seizing on a particular species of plant or animal, and observing it. It's something everyone could do, and should do, and it doesn't have to be box turtles.

For me, box turtles are just plain neat, they're here and available where I live, and they're very much deserving of conservation. They are excellent bioindicators of ecological health over the long term. They live many decades, approaching and occasionally passing the 100-year mark, unlike most animals. They are easily identifiable as individuals, and so can be tracked over those many decades. And they don't go far from what they consider as home, or at least most don't (some do travel farther than others).

On Nov 11, 2005, I posted my first photo of a box turtle, and someone suggested in comments then and later that this might be an interesting project. Mountains of posts have followed, then, and so it's not my fault. I've had quite a shallow learning curve, which culminated this year in a peak I plan to maintain.

Everything is a predator to a tiny juvenile box turtle, and that condition persists for several years. Very few make it to maturity, but by the time they're mature adults (5-10 years) only cars and lawnmowers, maybe large pets and habitat disruption - ok ok - pretty much all human activities - really threaten them on the large scale. You won't find any but the most confused lost box turtle in a suburban area, for instance, and it won't be there for long. It would be an amazing thing to find one in a city, and it could only be an escaped "pet".

Here's something I've read or heard, over and over: "When I was a kid, I would see box turtles all the time. Now I never see them." Bless your heart - there's a lot of stuff in there to unpack. Let's do so.

Here is a map of our area, to put my sampling area into perspective:



The blue indicates creeks, mostly small, and the red are roads, with the broader red being the more traveled, at higher speeds, and of course the more dangerous for box turtles. Roads are killers for box turtles, as they are for many animals. The green is an approximate location of my study area, that I see as being 20-25 acres in size.

The map doesn't really give a good idea of usage. Suffice it to say that much that seems free of dangerous vehicular traffic is actually given over to agricultural uses. That brings in tractors, tree and other harvesters, and other killers of box turtles. The area between Moss Creek and Old Edwards Road is largely free of such things, until you get to Lake Oglethorpe and south. West of Moss Creek is either highly agricultural University of Georgia land, hog parlors, or as we go south, a few residential areas. Black Snake Road is a fairly well traveled dirt road, not only by its residents but also by kids who drive ATVs back and forth ceaselessly. The real killers will be US 78 to the north and Wolfskin Road to the south.

If you had to say what areas *were* safe for box turtles, it would be the areas around the creeks. Not the lake, unfortunately, because it is ringed by 80 or so houses, and lawn mowers are active there during box turtle season. The occasional pond is more likely an indication of pasture activity (read: cattle), and therefore not safe for turtles.

I think you could safely say that for this particular species, they are boxed into areas that are usually smaller than their territories. You know what happens then, right? They get smatched.

So here's what happened this year.

A couple of weeks ago I said:
In 2012, I made 95 trips over a study area of 20-25 acres, averaging a little over a mile per trip. I covered a total of 113 miles since March 1, over 204 hours. I had 66 box turtle encounters, most of which were rediscoveries of old friends. And I found 23 new box turtles (10 of which were on a newly opened study area on the east side of the property).


Here is an indication of how that has been going over the years. As you can see, this year was a huge increase of observations over previous years:



2012 was the first year that I really made a strong effort to cover box turtle movements over the course of a season. Other years I've walked as much over the property, but have not actively studied my route for box turtles to the left and right. Active observing has had a five-fold enhancement over simply wandering, in previous years. Sure I've found some in prior years, but most probably watched me walk past them, unobserved by myself. I probably missed some this year, but damned few, I'd guess.

I began, right at the beginning in early March, to use an app (MyTracks, for Android) to keep track of my mileage, time, and route. I had 3-4 major routes that I walked, covering as much of that 20-25 acres overall as possible. An average walk probably allowed me to scan 2 acres - 10-20 feet to my left and right along the path. Not much of a CV workout, but close to 2 miles per trip.

Simply totaling numbers wasn't a good indication of box turtle activity, as I might walk for two hours one day and see one turtle, and walk the same length of time another day and see four. I decided on dividing the number of box turtles observed per hour walked, and compile that number at the end of each month. Here's the resulting bar graph:



Box turtle activity, in numbers per hour walked, peaked in July. Admittedly, my August and September efforts were less numerous, and so less accurate. I was eradicating Microstegium during those two months, it was much hotter and more humid (and therefore less pleasant to contemplate a walk), and that meant my powers of observation were tuned to invasive grass and not to turtles. I also didn't cover as much territory. But I don't think there's much qualitative doubt that box turtles just weren't as active during those months.

I was not able to reproduce the observations of unusual numbers of new box turtles in the waning months of the season that I saw last year. You might recall that I hypothesized last year that I was seeing box turtles returning to a hibernaculum from places outside of my observing area. That would have been an interesting proposition, but this year I really can't support it, I don't think.

Here is the final daily observation detail for 2012. The green bars indicate a trip in which the indicated numbers of box turtles were found. The red dots indicate days on which a trip was made, but no box turtles were found. You can pretty easily see the August and September wastelands. That's probably more due to my inattention. More to check on, next year.



There are a few additional details that I'll get to in a later post.

Saturday: 3 November 2012

The Month of October  -  @ 09:17:07
It's The Month of October, Number 81 in a series. We were just a degree or so below normal, but with extremely dry conditions after Oct 2.

Here are the usual temperature anomalies products, at the National Weather Service Climate Prediction Center.



For the third month in a row, the West continues with higher than normal temperatures. This anomalies map looks very similar to September's, with just a little more blue and a little less red. The exception would be the New England states, which were considerably warmer than normal for October.

We find the National Weather Service Climate Prediction Center's precipitation plots here.

Drought conditions continued to grow worse, perhaps for the third month in a row, for the northwest quadrant of the US. As was true for the temperatures map above, the precipitation map here is very similar to September's. Florida got an exceptional amount of rain in October, presumably due to Sandy. Elsewhere conditions were mostly close to normal, with surpluses of rain in a swath across the East.


For the Athens, GA area:

Here is a plot of our temperature swings in October, along with precipitation amounts as experienced in Wolfskin. Temperatures declined in steps:



The average monthly temperature for October was 62.0 degF, about a degree below normal. We had 5 days more than 1 standard deviation above normal highs (4.7 normal), and 3 days more than 1 standard deviation below normal lows (5.1 normal). Nothing exceptional there, and we broke no records.


The monthly histogram shows the breakdown of high and low temperature range counts from October 1948 on. There was nothing very significant in the daily high temperature ranges. The number of days with 70-79 degree highs just flirted with being more than normal, while the number in the 60-69 range was just at significance for fewer than normal. We did have significantly more daily lows in the 41-50 degF range than normal, so October was just a bit cooler.



The figure below shows the Athens precipitation data which are official for our area. As usual the green line shows our actual rainfall, the red shows the average accumulation expected. The black dots are rainfall over the last 20 years, the vast river of peach shows the standard deviation.



Our total out here was 1.58", and in Athens (shown here) it was 1.93". 3.55" is normal for October, so we were considerably below normal. There was virtually no rain after October 2.

And so you might ask, what is the cumulative rain log like now? Well, I just happen to have that, below. You'll recall that I project what our rainfall should be if we started taking account of it in January 2005.

The blue line is the most interesting, as it shows difference between the accumulated average and the accumulated actual. As you can see, we've just about returned to the low point we experienced in September 2009. We never really got to zero since then, generally remaining at about a ten inch deficit, and then began dropping steadily beginning in spring of 2011.



Prognosticator stuff:

What is the neat prognosticator telling us? For northeast Georgia, We might expect normal or slightly greater chance of warmer temperatures for November, and then normal temperatures through January. Precipitation is predicted to be basically normal through January, but that's what they said last time about October.

ENSO stuff:

The folks at CPC have a version of PDF or HTML that is much different from their previous presentations, but at least it's there and the link isn't broken.

As of October 29, ENSO neutral conditions continue. What was seen previously as a good change of El Niño developing and continuing into northern hemisphere winter has been modified. It now sounds like it's likely to be ENSO neutral, or at most merely a weak El Niño during the winter months.

As of October 25, the US Drought Monitor continues to have us in extreme to exceptional drought, depending on where you are in Georgia. Much of the US is now under at least abnormally dry or moderate drought classification, with a lot of the country in severe or extreme drought.

NOAA's Monthly State of the Climate product for September is available. The summary for 2011 regionally, nationally, and globally is also available.


I'm only placing five posts on the front page.
Go to the archives on the right sidebar for past posts, or use the search routine at the top of the page.

Copyright and Disclaimer: Unless indicated otherwise, the images and writings on this blog are the property of Wayne Hughes and Glenn Galau and should not be used without permission or attribution. Image thieves and term paper lifters take note.
We are not responsible for how others use the information or images presented here.
Reblogging is not allowed unless you ask for permission. We're sorry to require this but there are rebloggers who refuse to compromise. Thank you.

1.589[powered by b2.]

4 sp@mbots e-mail me