130925-studie-wildlife-comeback-in-europe
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
Construct<strong>in</strong>g historical distribution maps – pitfalls, biases and advances <strong>in</strong> technology<br />
As with population time series, knowledge of both historical<br />
and current species distributions can help underp<strong>in</strong> understand<strong>in</strong>g<br />
of <strong>wildlife</strong> <strong>comeback</strong> and decl<strong>in</strong>es and help to provide<br />
tangible solutions to conservation issues [1] . While locality<br />
records for species are generally widely available, for example<br />
through museum data, literature data, atlas publications and<br />
onl<strong>in</strong>e databases [2] , reconstruct<strong>in</strong>g species distributions over<br />
time often relies on a variety of sources, each of which may<br />
harbour dist<strong>in</strong>ct biases and shortcom<strong>in</strong>gs, which <strong>in</strong> turn may<br />
have a direct bear<strong>in</strong>g on the accuracy and resolution of the<br />
result<strong>in</strong>g distribution map.<br />
Compared to other regions of the world, European <strong>wildlife</strong><br />
has received a large amount of research attention over time.<br />
As a result, there is a large pool of knowledge available on<br />
current species occurrences and distributions, and obta<strong>in</strong><strong>in</strong>g<br />
current data is made even more straightforward through the<br />
establishment of records centres and databases which conta<strong>in</strong><br />
up-to-date <strong>in</strong>formation. In the case of the IUCN Red List of<br />
Threatened Species [3] , current distribution maps are verified<br />
by experts and regularly updated; moreover, the data are freely<br />
available.<br />
Construction of historical distributions is much less straightforward.<br />
Most often, distributions are amalgamated from<br />
different sources, and this can lead to biases with<strong>in</strong> the result<strong>in</strong>g<br />
distribution data. It is therefore imperative to understand the<br />
shortcom<strong>in</strong>gs when construct<strong>in</strong>g historical distributions, many<br />
of which have been discussed <strong>in</strong> the literature (see [2] for a good<br />
overview). Here, we summarise the three most likely pitfalls<br />
when reconstruct<strong>in</strong>g historical distribution maps:<br />
1. Data from different sources are likely to vary <strong>in</strong> terms of<br />
spatial resolution and may be biased towards certa<strong>in</strong> parts of the<br />
species range, while other areas with<strong>in</strong> the species range may<br />
only be broadly covered or even overlooked.<br />
2. The age of technology has advanced our ability to map species<br />
distributions: while <strong>in</strong> the past, distribution estimates were<br />
generally based on species occurrence records and broad <strong>in</strong>ferences<br />
about suitable habitat, we now have the use of advanced<br />
habitat suitability models which are fed by detailed data layers<br />
on climatic and habitat factors. For example, some mammalian<br />
range maps on the IUCN Red List have been produced that way.<br />
This creates a dichotomy <strong>in</strong> spatial resolution between current<br />
and past range maps.<br />
3. Focus on previously under<strong>studie</strong>d taxa (for example as a<br />
result of <strong>in</strong>creased conservation focus) may have led to recent<br />
discoveries of new populations and locations. Such new records<br />
suggest range expansion, while <strong>in</strong> fact the species may have<br />
persisted <strong>in</strong> that location undiscovered for a long period of time.<br />
Certa<strong>in</strong> precautions can be taken to avoid these pitfalls and<br />
biases <strong>in</strong> the result<strong>in</strong>g data. For example, smooth<strong>in</strong>g of overly<br />
detailed distribution maps may help to f<strong>in</strong>d some middle ground<br />
between different spatial resolutions. Includ<strong>in</strong>g areas for which<br />
species presence is uncerta<strong>in</strong> <strong>in</strong> our construction of historical<br />
distributions can help to reduce bias towards over<strong>studie</strong>d areas.<br />
However, these sources of bias rema<strong>in</strong> a major issue when<br />
consider<strong>in</strong>g range changes from historical basel<strong>in</strong>es.<br />
References<br />
1. Willis, K.J., Arauo, M.B., Bennett, K.D., et al. 2007.<br />
How can a knowledge of the past help to conserve<br />
the future? Biodiversity conservation and the<br />
relevance of long-term ecological <strong>studie</strong>s. Philosophical<br />
Transactions of the Royal Society of<br />
London B, 362: 175–186.<br />
2. Boakes, E.H., McGowan, P.K.J., Fuller, R.A., et al.<br />
2010. Distorted views of biodiversity: spatial and<br />
temporal bias <strong>in</strong> species occurrence data. PLoS<br />
Biology, 8: e1000385.<br />
3. IUCN. 2011a. The IUCN Red List of Threatened<br />
Species. 2011.1 Edition. [Available from: http://<br />
www.iucnredlist.org/.<br />
the Waterbird Population Estimates (WPE) <strong>in</strong>formation<br />
portal [18] , an onl<strong>in</strong>e database provid<strong>in</strong>g<br />
<strong>in</strong>formation on the current status of waterbird<br />
species, <strong>in</strong>clud<strong>in</strong>g long-term population trend<br />
analyses carried out us<strong>in</strong>g TRIM software [19] .<br />
Pan-European trends of breed<strong>in</strong>g population<br />
size for two species (White stork and Common<br />
crane) are available <strong>in</strong> the form of Pan-European<br />
Common Bird Monitor<strong>in</strong>g Scheme (PECBMS)<br />
Index trends [20] . PECBMS is a jo<strong>in</strong>t <strong>in</strong>itiative of the<br />
European Bird Census Council (EBCC) and BirdLife<br />
International, which aims to collate data on the<br />
breed<strong>in</strong>g population trends of common well-monitored<br />
species <strong>in</strong> Europe. PECBMS comb<strong>in</strong>es the<br />
results of national bird monitor<strong>in</strong>g schemes to<br />
produce yearly population <strong>in</strong>dices of bird species<br />
across Europe, us<strong>in</strong>g TRIM software [19, 21] . The<br />
method takes <strong>in</strong>to account differences <strong>in</strong> survey<br />
methodologies between countries, as well as differences<br />
<strong>in</strong> population size, and imputes any miss<strong>in</strong>g<br />
values for survey localities and years [21] . It was<br />
possible to adapt this method to calculate pan-European<br />
trends <strong>in</strong> abundance for five raptor species<br />
[White-tailed eagle (Haliaeetus albicilla), Eastern<br />
Imperial eagle (Aquila heliaca), Lesser kestrel (Falco<br />
naumanni), Red kite (Milvus milvus) and Peregr<strong>in</strong>e<br />
falcon (Falco peregr<strong>in</strong>us)], us<strong>in</strong>g the time-series<br />
of estimated population size and treat<strong>in</strong>g each<br />
16