Vinsamlegast notið þetta auðkenni þegar þið vitnið til verksins eða tengið í það: http://hdl.handle.net/1946/21127
The overall objective of this study was to develop statistical methods to detect trends with applications to two ecological monitoring programs, a) monitoring of contaminants in the marine environment around Iceland and b) monitoring of the population of the rock ptarmigan in Iceland. Polynomial models were used to account for trends with no consistent direction, mixed models were used to analyze data from multiple sites simultaneously and to describe correlations between observations. A changepoint (CP) model was investigated and a new method proposed which takes autocorrelation into account when detecting a CP in short time-series. A population reconstruction model was developed for the ptarmigan population in NE-Iceland which allows for the possibility of including a CP. The statistical analyses revealed that the concentration of the persistent organic pollutants have been decreasing both in mussel and cod over the recent years. However, there were signs of local pollution that could be traced back to a whaling station, aquaculture and waste incinerator. There was no consistent trend for the trace elements. A population reconstruction model was developed for the population of the rock ptarmigan in Iceland. It estimates the abundance, natural survival and hunting mortality for two age groups. This model allows for the possibility of modeling the natural survival as a function of density and the hunting mortality as functions of either density or hunting effort with a CP. A CP was included in the function for hunting mortality in 2003 when the hunting regulations were changed. The model indicates that changes in the hunting regulation did indeed have an effect in reducing the hunting mortality and also changing the harvest strategies of hunters. Still, management goal of reducing the total annual mortality to 37% has not been achieved and a further change in regulation may be needed.
|Statistical analysis of trends in data from ecological monitoring.pdf