It is always hard to find habitability percentage for the land birds ( Black winged stilt, Oriental Pratincole, etc) as their babies soon after hatching leave the nests. How to find this in such an area where we cannot put cameras due to human interference? And specially if their nests are on small islands, and in large numbers, and even walking on that island can cause considerable damage to birds and can attract predators.
According to various papers and nest monitoring guidelines we should not visit nests too often and should wait for minimum 3 days to visit that place again. So in presence of all these problems how to get an accurate hatchability percentage for those birds?
as your data is incomplete (no daily visits) I suggest you to use capture-recapture.
take a look to that paper: http://dx.doi.org/10.1675/063.036.0211
Building on what Fernando suggested, you will probably want to use Program MARK to analyze nest success. It shouldn't be a problem to have incomplete data for this - I think that's very common for nest success/ hatchability analyses, as these same limitations often apply. An older methodology that you might also look into is Mayfield nest success.
I agree that most studies ignore this and typically assume offspring successfully leave the nest even if the nest is last checked a few days before fledging. This can really influence success rates, but is never really discussed. It is not necessarily true that you have to wait 3 days between nest checks. Whether predation increases or decreases with an increased frequency of nest checks depends on the species and what predators you have. You could potentially check some nests more frequently when the offspring are closer to hatching and then you could take this into account in your model. For example, if you have some nests where you checked every 3 days and others every day or 2 days, you could take into account the frequency of nest checks in your model. Another option would be to check from a distance if possible. Could you use a spotting scope and look for feeding trips by parents to the nest? As for analyses, logistic exposure also does not require you to have known fates.
I agree with Corey. Checking every three days is a simple guideline and not a rule. As for the recommendation to use program MARK, please be careful when reporting the outputs. MARK reports symmetric standard errors in its output, despite the fact that the errors are not symmetric (owing to the fact that most mortality in these young life-history stages is exceptionally high...therefore, the values are at the edge of the 0-1 distribution). Worse yet is that practitioners of MARK often report a single (i.e. symmetric) standard error as the measure of uncertainty. This single value is actually an average of the upper and lower asymmetrical error values. Even worse is that many practitioners shrug their shoulders when confronted with this reality. Be careful with the default MARK output. It pretends to help you when, in fact, it is attempting to stab you in the back. In sum, I would go with logistic-exposure, implemented in R as a GLMM (if you think nature is linear) or as a Generalized Additive Model (if you like to let nature tell you what pattern it holds).
Dear Rehan
I suppose that you interest with hatchability of these birds in natural habitat and wild life , because we can domesticate most wild birds and study its biology and behavior , and the study can be performed in natural reserve .
In case of wild life the excellent method for the study can performed with camera monitor .
Good Luck
- Badges
- Science topic
- Similar topics
- Biological Science
- Zoology
- Ornithology