Trend detection in source-sink systems: when should sink habitats be monitored?
(2005) In Ecological Applications 15(1). p.326-334- Abstract
- We determine the power of population monitoring in source or sink habitat to detect declining reproductive success in source habitat using a stochastic population model. The relative power to detect a trend in the source by monitoring either the source or the sink Varies with life history parameters, environmental stochasticity, and observation uncertainty. The power to detect a decline monitoring either source or sink habitat is maximized when the reproductive surplus in the source is low. The power to detect a decline by monitoring the sink increases with increasing reproductive deficit in the sink. If environmental stochasticity in the source increases, the power in the sink goes down due to a lower signal-to-noise ratio. However, the... (More)
- We determine the power of population monitoring in source or sink habitat to detect declining reproductive success in source habitat using a stochastic population model. The relative power to detect a trend in the source by monitoring either the source or the sink Varies with life history parameters, environmental stochasticity, and observation uncertainty. The power to detect a decline monitoring either source or sink habitat is maximized when the reproductive surplus in the source is low. The power to detect a decline by monitoring the sink increases with increasing reproductive deficit in the sink. If environmental stochasticity in the source increases, the power in the sink goes down due to a lower signal-to-noise ratio. However, the power in the sink increases if environmental stochasticity is increased further, because increasing stochasticity reduces the geometric mean growth rate in the source. Intriguingly, it is often most efficient to monitor the sink even though the actual reproductive decline occurs in the source. If reproductive success is declining in both habitats, censusing the sink will always have higher power. However, the probability of Type 1 error is always higher in the sink. Our results clearly have implications for optimal population monitoring in source-sink landscapes. (Less)
Please use this url to cite or link to this publication:
https://lup.lub.lu.se/record/147417
- author
- Jonzén, Niclas LU ; Rhodes, J R and Possingham, H P
- organization
- publishing date
- 2005
- type
- Contribution to journal
- publication status
- published
- subject
- in
- Ecological Applications
- volume
- 15
- issue
- 1
- pages
- 326 - 334
- publisher
- Ecological Society of America
- external identifiers
-
- wos:000227120700026
- scopus:13944269243
- ISSN
- 1051-0761
- DOI
- 10.1890/03-5168
- language
- English
- LU publication?
- yes
- id
- 6723ea0a-d22c-49fb-841d-6e1a4f6b1a09 (old id 147417)
- date added to LUP
- 2016-04-01 17:13:59
- date last changed
- 2022-02-13 03:42:11
@article{6723ea0a-d22c-49fb-841d-6e1a4f6b1a09, abstract = {{We determine the power of population monitoring in source or sink habitat to detect declining reproductive success in source habitat using a stochastic population model. The relative power to detect a trend in the source by monitoring either the source or the sink Varies with life history parameters, environmental stochasticity, and observation uncertainty. The power to detect a decline monitoring either source or sink habitat is maximized when the reproductive surplus in the source is low. The power to detect a decline by monitoring the sink increases with increasing reproductive deficit in the sink. If environmental stochasticity in the source increases, the power in the sink goes down due to a lower signal-to-noise ratio. However, the power in the sink increases if environmental stochasticity is increased further, because increasing stochasticity reduces the geometric mean growth rate in the source. Intriguingly, it is often most efficient to monitor the sink even though the actual reproductive decline occurs in the source. If reproductive success is declining in both habitats, censusing the sink will always have higher power. However, the probability of Type 1 error is always higher in the sink. Our results clearly have implications for optimal population monitoring in source-sink landscapes.}}, author = {{Jonzén, Niclas and Rhodes, J R and Possingham, H P}}, issn = {{1051-0761}}, language = {{eng}}, number = {{1}}, pages = {{326--334}}, publisher = {{Ecological Society of America}}, series = {{Ecological Applications}}, title = {{Trend detection in source-sink systems: when should sink habitats be monitored?}}, url = {{https://lup.lub.lu.se/search/files/4915185/625127.pdf}}, doi = {{10.1890/03-5168}}, volume = {{15}}, year = {{2005}}, }