Hi-
I am pretty new to mark/recapture so bear with me. I am doing a study on wading bird prey availability and will have stocked enclosures with a set number of fish (open population since they are prey). I was wondering if there is a way to figure out how many fish I have by fin clipping a certain percentage of the fish but not individually marking them. Any thoughts? Basically I need a set number of fish so I will have to restock each day after figuring out my numbers using minnow traps to recapture them. Thanks.
beginner
2 posts
• Page 1 of 1
Estimating number of fish removed from a batch mark.
by cschwarz@stat.sfu.ca » Tue Apr 18, 2006 1:13 pm
This is a bit tricky to do and there are a number of problems.
Simply clipping a fraction of the fish won't work. For example, suppoe you could clip ALL of the initial fish. Then 100% of the fish are clipped. If the wading birds eat some fish, the clipping percentage will stay at 100% regardless of the number of fish left.
In general, if deaths are the only source of non-closure, than simple batch marking is equivalent to a petersen experiment, and you can only estimate the number of fish at time 0 which is not of interest.
If you are only doing a batch mark (e.g. fin clip) that doesn't identify the capture history of each fish, then I think you are stuck.
Can you identify newly added fish with a different fin clip? If so, you might be able to get the individual capture histories and do a classic jolly-seber analysis.
If you can clip new additions differently than fish currently in the POND, then you migh be able to do something like the following:
(i) Add the known number of newly marked fish BEFORE you take the second sample.
(ii) Take the second sample. Find the fraction of newly marked fish in the your second sample. This can then be expanded to estimate the total population size at the second sample, e.g. if you add 10,000 new fish, and your second sample has 10% newly marked, then your population size must be 100,000.
This is nothing more than a simple Petersen experiment with the "marking" event being done by adding newly marked fish.
You may also wish to do something like a robust design where you do a petersen at each time point you want to add fish. But you will need a separate mark for each event.
How about precision. I'm afraid that unless you have marked a substantial fraction of the population, you estimates of population size will be so poor as to be useless. A general rule of thumb is that if you RECAPTURE m marked fish, you cv of the population estimate is 1/sqrt(m). So if you get back 100 marked fish, you cv of the population estimate will be about 1/sqrt(100)=.1=10% and your 95% ci will be +/- 20% which may not be precise enough for your needs.
Alternatives to a mark-recapture study include a catch-per-unit effort (your minnow trapsx number of hour soak time).
Could you provide more details on your experimental design and conditions?
Simply clipping a fraction of the fish won't work. For example, suppoe you could clip ALL of the initial fish. Then 100% of the fish are clipped. If the wading birds eat some fish, the clipping percentage will stay at 100% regardless of the number of fish left.
In general, if deaths are the only source of non-closure, than simple batch marking is equivalent to a petersen experiment, and you can only estimate the number of fish at time 0 which is not of interest.
If you are only doing a batch mark (e.g. fin clip) that doesn't identify the capture history of each fish, then I think you are stuck.
Can you identify newly added fish with a different fin clip? If so, you might be able to get the individual capture histories and do a classic jolly-seber analysis.
If you can clip new additions differently than fish currently in the POND, then you migh be able to do something like the following:
(i) Add the known number of newly marked fish BEFORE you take the second sample.
(ii) Take the second sample. Find the fraction of newly marked fish in the your second sample. This can then be expanded to estimate the total population size at the second sample, e.g. if you add 10,000 new fish, and your second sample has 10% newly marked, then your population size must be 100,000.
This is nothing more than a simple Petersen experiment with the "marking" event being done by adding newly marked fish.
You may also wish to do something like a robust design where you do a petersen at each time point you want to add fish. But you will need a separate mark for each event.
How about precision. I'm afraid that unless you have marked a substantial fraction of the population, you estimates of population size will be so poor as to be useless. A general rule of thumb is that if you RECAPTURE m marked fish, you cv of the population estimate is 1/sqrt(m). So if you get back 100 marked fish, you cv of the population estimate will be about 1/sqrt(100)=.1=10% and your 95% ci will be +/- 20% which may not be precise enough for your needs.
Alternatives to a mark-recapture study include a catch-per-unit effort (your minnow trapsx number of hour soak time).
Could you provide more details on your experimental design and conditions?
- cschwarz@stat.sfu.ca
- Posts: 43
- Joined: Mon Jun 09, 2003 1:59 pm
- Location: Simon Fraser University
2 posts
• Page 1 of 1
Who is online
Users browsing this forum: No registered users and 3 guests