Observation bias and its causes in botanical surveys on high-alpine summits
METADATA ONLY
Loading...
Author / Producer
Date
2015-01
Publication Type
Journal Article
ETH Bibliography
yes
Citations
Altmetric
METADATA ONLY
Data
Rights / License
Abstract
Question
To determine long‐term vegetation changes in revisitation studies, it is crucial to know how much of measured species turnover over time can be attributed to pseudo‐turnover (i.e. turnover caused by imperfect data acquisition), and which factors contribute to observation bias and pseudo‐turnover. Independent simultaneous surveys provide a powerful tool to quantify pseudo‐turnover and to indentify factors causing it, which may vary strongly between lowland and mountain areas.
Location
Alpine mountain summits (2616 m to 3418 m a.s.l.) in the southeastern Swiss Alps.
Methods
Plant inventories of 48 summits were collected by two independent observers simultaneously. Pseudo‐turnover between observers was compared to species turnover over one century based on historical species lists of the same summits. Variables linked to observer characteristics and external (observer‐independent) factors were tested for their influence on pseudo‐turnover and number of species missed by one of the observers, and plant characteristics were tested for their effect on species detection probability.
Results
Mean pseudo‐turnover between observers (13.6%) was almost three times smaller than species turnover over one century (41.4%). Pseudo‐turnover and the number of species missed increased with difference in botanizing time between observers and with a longer ascent to the summit, especially in combination with a high species richness on the summit. Species had a higher probability to be missed if occurring on many summits but with a low abundance, if small in stature and if belonging to certain taxonomic plant groups (e.g. Asteraceae).
Conclusions
Our critical evaluation of turnover over time vs pseudo‐turnover confirms that floristic changes on alpine summits over time represent an ecological pattern. In mountainous terrain, factors related to observer characteristics play a major role, as we found the best correspondence between simultaneous records when the difference in botanizing time was small and the ascent was short. Our results help to improve data quality in mountainous terrain by pointing out possible causes for observation bias. Long‐term vegetation studies in alpine ecosystems should make a strong effort to identify and minimize such causes in advance, for instance by reducing between‐observer differences in botanical skills, fitness and time management through appropriate training.
Permanent link
Publication status
published
External links
Editor
Book title
Journal / series
Volume
26 (1)
Pages / Article No.
191 - 200
Publisher
Wiley-Blackwell
Event
Edition / version
Methods
Software
Geographic location
Date collected
Date created
Subject
Between-observer variability; Long-term monitoring; Pseudo-turnover; Revisitation studies; Sampling error; Species detectability; Swiss Alps; Vegetation records
Organisational unit
Notes
Received 17 December 2013, Accepted 30 May 2014, Published online 7 July 2014.