Maps to predict population sizes
 

Hi,
I am a computer cartographer, but also a dedicated hunter, and have been working on creating detailed maps that can be used to forecast which areas will hold large wildlife populations. My first map - forecasting Gambels quail in Arizona, worked out pretty well and is based on sound scientific studies - but I am looking to make more maps for other states as well. The size of quail populations in Arizona are directly correlated to the precipitation levels during the winter months preceding the hunting season. I would like to make maps for other states (and species) as well. The reason I am posting this request is if anyone knows of correlations between rainfall and wildlife populations. If you know of a known relationship - please let me know! If you let me know of som sort of relatiosnip I will make the map and email it to you for free. The rain data I have available is detailed down to one square mile - so if strong relationships/correlations exists one can predict population sizes locally.

Please send an email to [email protected]
or visit the webpage at http://www.rainmaps.com

In advance, thanks a lot!

-Andreas
Tucson, AZ

RE: Maps to predict population sizes
 

Hi, I sincerly wish that predicting wildlife populations were that easy! Unfortunately, one variable such as precip is probably not a strong indicator of population numbers. If you want to model populations, a multi-variate model is needed and can be pretty easily implemented in ArcMap. Unfortunately, such models are usually limited by the data available. While it may be mildly interesting, and possibly coincidentally predict population trends, precip alone is not sufficient to accurately model populations as multiple factors influence survival and reproduction of all wildilfe species.

RE: Maps to predict population sizes
 

Wow, I couldn't agree with Brutal more. There are simply to many variables to accurately predict wildlife trends. Factor in the "Butterfly Effect" principles and it becomes nearly impossible. That is not to say there cannot be some short term success. The RMEF has used GIS discipline to track core wintering area for elk, then over lapped poplutaion desity and growth maps to identify critical habitat areas, however, actually predicting wildlife populations via precipitation etc is difficult. An obscure disease can throw poplutaions into a downfall, or a change in predation etc. Additionally, population trends tied into precipitation would have to be adjusted yearly would they not?

RE: Maps to predict population sizes
 

Well, you would have to arbitrarily pick a "cut-off" time for the precip data (like Dec 31 or something) and then determine the precips effects on the species (which will vary depending on the ecology of the species being predicted). Then, you would have to arbitrarily pick a lag time in which the species will then realize the supposed effects of this precip. For example, "a high precip year may not impact seral shrub growth until the following spring" or something of this nature. Now, if your talking about direct impacts on survival such as quail, you could assume an immediate impact on the population such as a very wet spring causing increased chick mortality or something of this nature.

Again, though it's basically armchair population dynamics with no real trendpredictor value.

RE: Maps to predict population sizes
 

Thank you for your comments! To a certain degree I can understand that you are skeptical, but let me explain a few things in greater detail.

"I sincerly wish that predicting wildlife populations were that easy! Unfortunately, one variable such as precip is probably not a strong indicator of population numbers."

Some species are easier to predict than others - quail for instance is simpler to predict compared to dove, elk, deer, etc. Reason being that quail do not move very large distances, therefore, local meteorological factors have a direct impact on local populations.
The same relationships do not exist everywhere, and definately does not apply to every species (or sub-species). There are different correlations for different types of quail, different correlations in different states, and of course, in different regions of some states. In AZ, at least three scientific studies have shown an extremely strong relationship with precipitation for certain months and the population sizes. One study explains 73% of the variation in quail populations with precipitation data alone. No scientific method will ever predict a 100% accuracy, nor has that ever been my intention. Since the strong, yet simple, relationship exists, it makes sense to hunt areas where those predictors are in the hunter's favor.

"While it may be mildly interesting, and possibly coincidentally predict population trends, precip alone is not sufficient to accurately model populations as multiple factors influence survival and reproduction of all wildilfe species. "

Yes, many factors play a role when forecasting. But some factors are more important than other factors. For instance, if one factor can explain 90% of the variation, it makes little sense to involve additional variables if each added variable only explains an additional 1%. For the purposes of predicting quail, it is very fortunate that precipitation is such a strong predictor.

"There are simply to many variables to accurately predict wildlife trends. Factor in the "Butterfly Effect" principles and it becomes nearly impossible."

I think most population trends can be fairly accurately predicted with a pretty small set of variables. That said, collecting the data to effectively use established models might be difficult (or not very cost effective). As stated earlier, exact prediction is rarely a goal in science - knowing if there are 999, 1000, or 1001 animals in any given area is not the goal, the goal is usually to be able to say whether there are 500, 1000, or 2000 animals in a given area.
Predicting wildlife population from year to year (or short-term) is usually more robust than that it would be affected by small unpredictable phenomena (the butterfly effect). Long-term studies (prediciting populations 20 years into the future) is different though - very few models have the inherent strength to accurately predict events far into the future.
I am basically using current, or recent, data to predict something in the near future, something that is far simpler.


"RMEF has used GIS discipline to track core wintering area for elk, then over lapped poplutaion desity and growth maps to identify critical habitat areas, however, actually predicting wildlife populations via precipitation etc is difficult. An obscure disease can throw poplutaions into a downfall, or a change in predation etc. Additionally, population trends tied into precipitation would have to be adjusted yearly would they not?"

Yes, disease can wipe out large populations quickly. My forecasts can not take those facts into account. However, drought is probably overall more important than disease, and my maps show area that have received little rain. Areas in distress (drought) are probably also more likely to be strucken by disease.
The maps are made on a yearly basis and will be updated yearly. They use the most recent relevant precipitation.

"Well, you would have to arbitrarily pick a "cut-off" time for the precip data (like Dec 31 or something) and then determine the precips effects on the species "

Yes, I am using a cut-off date. The research I base the maps on are often set up this way.

"Then, you would have to arbitrarily pick a lag time in which the species will then realize the supposed effects of this precip. For example, "a high precip year may not impact seral shrub growth until the following spring" or something of this nature."

I do not arbitrarily pick lag times - this is specified in the scientific literature. This research is very often done by local Game & Fish Departments.
I am currently working on a very interesting map that uses a "drought index", which is calculated using recent historic precipitation data. This index takes into account the frequency of the rain, and the quantity, which in turn relates to vegetation growth. This index has been shown to explain 90% of the variation in the quail populations for certain regions of Texas. Stay tuned :-)

-Andreas
http://www.rainmaps.com

RE: Maps to predict population sizes
 

It doesn't matter.You don't really see many paperswherea single variable is usedto predict anything with any kind of power. That is such a basic concept. If single variate models were accurate and/or precise at all, we would model everything with simple algebraic equations.

If what your doing makes you feel good more power to you.

However, I don't approve of you masquerading as someone who can predict population density with one variable because it doesn't happen and your sending the wrong message. I'm a professionl biologist and I know that for most species no single variable is weighty enough to use by itself to predict anything with any kind of certainty.

RE: Maps to predict population sizes
 

Now I do agree that quail populations are heavily effected by precip, however I'm a little skeptical that you think this one variable can predict population density correctly 90% of the time. Can you post your data?
Thanks.

RE: Maps to predict population sizes
 

The reason we use multivariate methods isn't necessarily to predict population density but to explain population trends which we can then use to predict with some small amount of statistical power. The problem with a single variable model is that is for instance you have a discrete disease event that causes your population to crash and your only looking at your precip data to explain the trend your seeing...then your model isn't even a model it's just arbitrary application of data.

RE: Maps to predict population sizes
 

I understand what your doing, but I am still skeptical of the usefulness of such information. Supposing there is a sybiotic relationship between quail/precip, why would you suppose any other animal would share this relationship? Secondly, most wildlife agencies have even more accurate population information then you do, so why not just consult them? I'm not sure I understand what your angle is here.

RE: Maps to predict population sizes
 

I agree. He's using data he's gleaned from GAP analysis and other sources, which if anything, he has access to less data than state and federal agencies who use multivariate models. It's extremely difficult to model population dynamics using only precip data so of course that is why they gather data on other variables.