which model? mark-capture

[MatthiasJurczyk]

Hi evrybody,
my name is Matthias and I am an student of biology in germany.
I've done a mark and capture study n small geckos on an island.
But after reading a lot about MARK, I'm still not sure, which modell i should use.

The island is complete inhabited by those animals. I chose a small study area.
In the first night, i walked an transect and captured every gecko i found, marked it with a number
and a special color, and released it.
The following night i did the same - i captured evry gecko, noted the number, if it had one,
or gave it a ne color and a new number and released it.
The same with the third night.

Now i have an encounter history file with codes like this:
100
110
111
101
010
011
001

My question now is:

Is there a modell in MARK, which allows me to use these codes? Most of the examples have only one "marking event" at the beginning.

I would be very glad if any of you could give me a little help with this program.
Would it be easier with the programm CAPTURE or POPAN?

What i almost forgot to tell: what i try is to estimate the size of the population in the islands.

Thank you very much!
With kind regards,
Matthias

[bacollier]

Matthias,
Sounds like you have a simple closed capture model. You might want to read a bit farther, especially Chp 14 of the MARKBOOK as there is an example that roughly mirrors yours around page 14-11.

Bret

[sbonner]

Mattias,

I think there are a couple of questions you need to consider before choosing the model:
1) How you think the geckos move about the island?
2) What aspect of the population are you trying to estimate?

How the geckos move about the island is important for a couple of reasons. You say you chose a small study area and captured individuals by walking a transect line through that area. Assuming that the transect line covered the small area sufficiently so that all geckos in the area had the same chance of being captured then the population will be closed if the individuals are pretty much stationary so that they don't move in and out of the study area between days. However, if the individuals can move in and out of the study area then you'll have to consider this an open population and fit open population models (like CJS or JS).

Whether the population is open or closed will also determine what you can estimate. If the population is closed then you can easily estimate population size and density for your study area (again, assuming that all individuals in the area have the same probability of capture along your transect line). If the population is open, then estimates of population size/density don't make sense because it isn't clear what area your study actually covers. Also, estimates of "survival" will be estimates of apparent survival and will incorporate both survival and emigration rates.

Hope that helps a bit.

Simon

[MatthiasJurczyk]

Hi Bret,
thank you for your answer. If I understand it right, a closed capture implies, that there is no imigration or emigration. But the geckos in my study do both and in a high frequency.
The study area is a flat, rocky area. There is a high chance of finding nearly all geckos that are present.

Hi Simon,
thank you, too, for your answer.
You're right, there are several questions, that should be answered before.
1) The geckos DO move in and out of the area, so there is no 'real closed population'.
2) I'm trying to get an estimation of the population size. I know that a closed population design would be much more acurate. But there was no chance.

So you say, that there is chance to get a 'good' population estimation with my data? Mh.

Thank you both very much.

Greets,
Matthias

[sbonner]

Hey Matthias,

This is a problem that my PhD supervisor, Carl Schwarz, drilled into my head so I'm on a constant lookout.

Imagine that each individual has a home range in which it moves at random. The probability that an individual is in your study site then depends on the size of the ranges, and so it's not possible to know what fraction of the entire islands population your really sampling.

The exception would be if you believed that the home ranges for the geckos covered the entire island (i.e., the geckos move about the entire island at random). In this case, you could consider the population to be closed and that you are estimating the number for the entire island (and I would have spoken too quickly).

Hope that helps.

Simon

[MatthiasJurczyk]

Hi,
that gives me a little hope.
These geckos don't have anything like a concrete "home". The island has a size of 34ha, and is more or less equally structured.
They move around and search for food. I found out, that they can walk up to 30 or more meters in one night.
But truly 'open' ..mh .. i don't know.
Best regards,
Matthias

[sbonner]

Hey Matthias,

Useful to know the distance that you expect that each individual might move between capture occasions. Given that the island is 34 ha and individuals can only move 30m per night, it's not reasonable to believe that you are sampling from the entire population of the island. This means that you can't justify using a closed population because the population on the island is closed.

So, the question really becomes how big is your study area is relative to this 30m? If your area is much larger, then it might be good enough to treat the population as closed and ignore the edge effects. Note that the estimate of the population size in this case will only reflect the population within the study area and not the whole island.

If your study area is on the same magnitude then you could be observing completely different subsets of the islands population each night. This is going to be hard (impossible?) to get at with only three samples, and it also begs the question of what population you are estimating the size of. Parameters like the rate of immigration into and emigration out of the study area survival might be possible assuming 100% survival, but are again going to be very hard to estimate with only 3 days of study.

Do you have any feeling of what your capture probability is (i.e., the probability of catching a gecko on one night given that it is inside the study area)?

Can I ask if you are still on the study site and able to add to your study? If so, it might be worth thinking of some alternative approaches to get at the information you need.

Cheers,

Simon

[MatthiasJurczyk]

Hi Simon,
my study area was just 20x35m.
The capture propability was nearly 100% - its a rocky, but flat area, without any vegetation.

It is very hard to get on this island. There is no chance to get there in the next few months.
But what i'm going to do is using my other smaller transects, but without MARK oder CAPTURE.
Thank you all for your help.
Matthias