Integral to all of our daily, monthly, and yearly activities is the locations where we perform our tasks, be they recreational, professional, or personal maintenance. We don’t necessarily have to define ourselves based solely on our locations, and even though Dave Matthews would probably disagree, where we are can certainly provide plenty of information about our lives.
Birds are no different. They go to certain locations to sleep. They go to certain locations to eat breakfast. And even though birds don’t have an economy or traditional ‘jobs’, they still have work to do. Thus, if we can figure out where an individual bird is, and better yet, why that individual is there, we can start to piece together the rich tapestry that is the life of a bird. And with more information about where birds go and how they get there, conserving habitat and populations becomes that much easier and more effective.
INSERT PHOTO
A Prairie warbler surveys his breeding territory. Pennsylvania, summer. Photo credit: Steve Brenner
First big point to establish: this is all about tracking individuals and then using that spatial information to answer a variety of questions about birds. Tracking animal movements at the population or species level is possible, albeit with slightly different methodological frameworks, but we can save that for another post. There are many ways to track individual birds, and the methodology is usually defined by the questions you want to answer and the technology available. This is quite a robust topic with a deep history, but alas, we must contain the ever-growing urges of scientific curiosity bubbling inside and focus on the overall purpose of tracking individuals. What kind of information can we gain, and what can we say or do with that information?
Let’s look at one of the basic and fundamental measurements in spatial ecology - an individual’s home range. The simplest definition of a home range is the space where an animal lives. Think of the daily routine example from above. Where we sleep, eat, and work exists within a certain space. Usually this space is contained within a town or city, and within that space would be your house, your office, your favorite places for recreation. Likewise, a bird’s home range is the space that contains the locations where it forages, nests, preens its feather, and sleeps.
To generate a home range, the first things we need are the locations in space and time of an individual (think GPS points on a map). Next, we need to choose a period of time we are interested in. For migratory birds, this could be a variety of periods throughout the year that each encompass different ecological behaviors and have different implications. For example, the breeding season, roughly May-August for North American birds, would be the time to construct a classic home range that contains a nest location, feeding areas, and locations for protecting young from predators.
INSERT PHOTO
Nestling Dark-eyed juncos, hoping their parents picked a safe nest location. Arizona, summer. Photo credit: Steve Brenner
To properly construct a home range, we need to make sure we have enough locations that we are gathering (or sampling from) that are representative of the bird over different times of the day and over the entire period of interest. Once we have our representative locations, we can plot them on a map and build the home range. But enough with the words, Steve, give us an example!
I’ve been studying towhees for the past two years in an effort to assess the effectiveness of statewide early successional/young forest management strategies for songbirds. Towhees are perfect representative of young forest or shrubland birds. Think of all the thorny, scrubby, bushy places you avoid on a daily basis…this type of habitat is perfect for towhees, and it’s in short supply in southern New England. Gathering spatial and nesting data on towhees and other shrub birds in Rhode Island will help us understand how (and if) these animals are using state-managed forests.
INSERT PHOTO
Male Eastern towhee, looking sharp and ready to provide spatial data with his new transmitter. Rhode Island, summer. Photo credit: Steve Brenner
Let’s look at the locations of an adult male Eastern towhee between June and August 2016 in Rhode Island. This individual was tracked after he successfully fledged 2 young, and was subsequently caring for his fledglings. Here are some of his GPS locations mapped out.
INSERT PHOTO
Already this is pretty cool to visualize. Just from seeing his points in context with aerial imagery is neat on its own. Also, the imagery provides a general context for the type of forest towhees are hanging around. But let’s create his home range and see what else we can find out. The simplest way to do this is by a method called ‘Minimum Convex Polygon”, or MCP. Essentially, this entails drawing the smallest box possible around all of our sampled points.
INSERT PHOTO
Cool! Already we can look at this map and say some things about this bird’s life. For instance: the size of this polygon is just about 1 hectare, which is roughly the size of a football field. Thus, this particular bird seemed to consistently spend a lot of time within a hectare-sized area while his young slowly grew up over the summer. But this whole straight-lined polygon thing seems a bit…unnatural. What are the odds that this towhee didn’t stray outside the blue lines on the map, or put another way, the likelihood the summer home range of this bird doesn’t include space beyond these lines? Fortunately, scientists have devised other ways to estimate home ranges beyond MCPs. A common method to account for the likelihood of an animal occurring outside this arbitrary polygon is by using kernel density estimation, or KDE. These methods can be a little complex and depend on many factors including sample size (how many points did you gather per bird?), autocorrelation (the influence of one location on the next), and bandwidth estimators (for statistics!).
I know what you’re thinking: we have reached the section of the article filled with multi-syllabic words that sound like math and are intentionally complex, and the only people who understand this are folks that like tofu and listen to jazz. Fear not. The extremely short explanation of KDE is that by using the distances between the sample locations themselves, one can more accurately estimate the probability of space used by an animal, and thus build a better home range. So let’s rebuild this towhee’s home range using KDE.
INSERT PHOTO
Well isn’t this just a pretty looking bit of spatial data! The size of this polygon is 2.8 hectares, which is much larger than the square box from earlier. But think about why this makes more sense. Giving the layout of this points, the odds that this bird wouldn’t use areas outside the GPS points I sampled it at are slim. These points are daily samples of one point in space - not direct minute-by-minute tracks of the bird, or even it’s path from one point to the next. Thus, this type of home range estimation takes this fact into account and by the magic of statistics you get this purple polygon. With a measurement of home range, we can compare this bird’s movements to other towhees that are raising young, or even to other towhees but during different stages of the life cycle (for example: does the home range size between the post-fledging period and the nesting period? I don’t know, but that’s a great question for future research!) Think about the habitat/environmental questions we can answer with this home range. What if I wanted to know how much towhees utilized previously managed forest clearcuts? Well, first I can add this GIS layer that outlines areas that were previously managed by the state.
INSERT PHOTO
Sweet! Then I could overlap with the home range, and calculate a quick percentage (~65%). Seems like this critter was happy to use a regenerating clearcut to raise his young, which makes a whole lot of sense. Early successional forests are full of densely packed shrubs and young trees. This provides excellent cover for vulnerable, recently fledged baby birds.
INSERT PHOTO
A fledgling towhee, wondering if it's about to be fed or eaten. Rhode Island, summer. Photo credit: Steve Brenner
We could also overlay the different type of forests in and around this bird’s home range in order to get a better sense of the immediate landscapes around towhees.
INSERT PHOTO
As in most research, if we can gather similar data and create home ranges for a larger sample of towhees, we can more confidently use this data to answer some really interesting questions about our ecosystem. Think of the possibilities!
-Does proximity to developed areas or edges influence survival?
-Do females use more or less space, or do birds with our without young use more or less space?
-Does distance to other managed forests impact spatial movements?
The possibilities are endless! These are the types of questions I’ve been working on with these little songbirds, and with a rapid increase in tracking technologies, all sorts of spatial questions are starting to be answered and will be addressed in the near future. But a home range is always a good place to start!
(All maps were created with ArGis.)
About the author:
Steve Brenner studied the impacts of habitat management on avian spatial ecology in the Scott McWilliams lab at the University of Rhode Island