Modelling for monitoring biodiversity:

How to predict, follow and assess species shifts in a

dynamic world

S TUDENT : Joaquim Filipe Polónia de Faria

Master’s Degree in Biodiversity, Genetics and Evolution – MBGE

Biodiversity of Land Ecosystems

May 2018
ABSTRACT

Monitoring biodiversity has always been one of the major focus of the scientific
community dedicated to biology. In later times, the difficulty of this task has been
massively increasing due to the effects of climate changes, as species are moving out of
their typical territories as they become unsuitable for their survival; with this, more and
better information — which is in itself another current crisis — becomes necessary to
effectively account for the needs in terms of actions and policy to save the planet
biodiversity. At the same time, new tools are being developed, which may to solve both
the biodiversity and the information crisis at once: models. These statistically driven
methods use existing information — that can be from different sources and of different
quality — to predict for the species presence not only for today, but also for the past and
future, thus helping greatly with the task of monitoring biodiversity. In this work the
topics of the biodiversity information crisis and modelling for monitoring biodiversity are
critically discussed and analysed based on the author’s current knowledge and research
of the available literature.

1
INTRODUCTION

Nowadays, planet earth is facing continuous environmental changes over all its biomes,
and though the main causes are still to be asserted, the effects on biodiversity are clear to
view: either be them at the ecosystem, organism, or even gene level (cf. Bachelet, Neilson,
Lenihan, & Drapek, 2001; Garrett, Dendy, Frank, Rouse, & Travers, 2006; Haeberli &
Hoelzle, 1995; IPCC, 2007; Lejeusne, Chevaldonné, Pergent-Martini, Boudouresque, &
Pérez, 2010; Loehle & LeBlanc, 1996). In order to maintain efficient conservation
management and protection efforts in this shifting system, researchers working on the
subject need more information and new tools that allow them to analyse, quantify and
estimate the impacts of these changes over biodiversity, and, ultimately, convey their
findings to those in charge of decision-making so the they may have the necessary
knowledge to — hopefully — make the best choices.
Regarding this necessary information, the current problems with it may not even be
actually its quantity, but issues more like its access, reliability, usability, or uneven
distribution — whether geographically or across taxa —, all of which makes much of it of
no use for relevant conservation programs and studies (Boakes et al., 2010; Hortal,
Jiménez-Valverde, Gómez, Lobo, & Baselga, 2008), or, when used even despite that, ending
up backing untrue conclusions (Ioannidis, 2005). This is a matter that may be solved —
or at least mitigated — with the embracing of standardised procedures for data collection
and storage, such as those proposed by the Global Biodiversity Information Facility (GBIF)
or the Group on Earth Observations Biodiversity Observation Network (GEO BON; Scholes
et al., 2012).
About the “new tools”, one that can surely be used to improve biodiversity monitoring
efforts is modelling. Put simply, this modelling consists on gathering the current and past
data regarding a certain biological entity of interest — e.g. a population, a species, a
community — and the environment it inhabits, analysing it side by side with the forecasts
for how that entity may behave — e.g. predictions for how climate change affects the
required resources and climate for the entity —, and generating a bunch of reality based
and scientifically supported scenarios of how the subject is going to behave in the
following years; these scenarios can then be used to design better conservation and
management programs and to improve existing ones (Honrado, Pereira, & Guisan, 2016),

2
that consider not only the present, but also plan for the future. Besides improving the
quality of the collected data, the usage of models when designing monitoring networks
for biodiversity also helps with cutting the costs — both economic and time related — by
implementing targeted observation methods (McDonald-Madden et al., 2011; Vicente et
al., 2016), and may be able to help with the above mentioned problem of using current
data, as models can integrate data from several sources and with different variables and
correlate them in order to still deliver coherent results (Honrado et al., 2016).
During the course of this work, the problems with information on biodiversity and
modelling for monitoring biodiversity were explored more in depth through research of
the at the time available literature. Both topics strengths and flaws were analysed, as it
was their relation — to assert for the true value of the former regarding its application on
the latter. The ensuing discussion, obtained conclusions and backing arguments are thus
presented in the following pages, accompanied by practical examples of the uses of
modelling and results of its implementation provided under the form of published articles
to further support the given thesis.

DATA FROM MONITORING BIODIVERSITY

Following a somewhat recent trend to protect and maintain biodiversity and stop the
current snowball effect of species loss (Tittensor et al., 2014) — which came to be due to
the notion that it is indispensable for our own survival (Bernstein, 2012) —, a series of
policy instruments have been created in the last years (Geijzendorffer et al., 2016).
In order to report for the actual effects of those policies or the need for new ones, it is
required that shifts in biodiversity are effectively monitored and that their causes are
correctly identified, so to assist those in charge of making decisions with relevant
information for them to knowingly choose for the best. The problem is that most of the
data we have from biodiversity monitoring suffers from some big limitations: (1) a biased
taxonomic and spatial coverage; (2) unusable data sets due to different data gathering
methodologies; and (3) insufficient integration at different scales (Geijzendorffer et al.,
2016; Green et al., 2005; Pereira, 2006). This altogether makes up for a very small
percentage of data that, in the end, can effectively inform on biodiversity shifts and be
applied in the decision making and designing of biodiversity protection policies.

3
BIASED DATA

It is widely known across the scientific community that our current knowledge on
biodiversity suffers from a great bias, both in terms of studied taxa and geographically —
not considering the chronological bias (Boakes et al., 2010; Hortal et al., 2008), since our
access to study the past is obviously quite limited and there is not much we can do to
balance the lacking knowledge of past times when compared to the present other than
studying the fossil record and ancient DNA (e.g. Pacioni et al., 2015), which though useful
are both quite incomplete. The taxonomic bias prevalent in biodiversity research results
in a lack of relevant information on biodiversity as a whole, with the information coverage
throughout taxa being incongruent with their frequency and variation in nature, as more
“popular” or economically important organisms receive many times as more attention
than “unpopular” ones (Bonnet, Shine, & Lourdais, 2002; Clark, 2002). To worsen the
situation, this less studied groups mostly correspond to more threatened taxa (Stein,
2002), making information on them a more immediate need; and, when they eventually
get studied, authors are somewhat forced to broaden their approach on the study,
focusing less on the organism and more on related general topics, in order to be able to
publish their findings (Bonnet et al., 2002). On the geographical side, accessibility and
development of the to be studied countries and locations play a major role on tipping the
scales, with places like the deep sea being overwhelmingly under-studied due to their
very difficult and costly access (Webb, vanden Berghe, & O’Dor, 2010), developing
countries being unable to sustain monitoring programs without external donor funds or
too dangerous to safely conduct scientific research on their territories (Danielsen et al.,
2003), and the tropics suffering from both difficult access and being in less wealthy
countries plus having external exploration interests forced onto them and prioritized over
the care for those environments (Sheil, 2001) — to picture some examples.
This is probably the hardest of the three addressed issues to completely solve, as it calls
not only for a compromise from the scientific community to invest their time and
attention in the “not-so-interesting” organisms, but also for aid and understanding from
outside sources — both governmental and from private lobbies — to overcome the
problems of funding, external pressures or political instabilities that haunt most of the
less studied areas of the world.

4
SHARING AND INTEGRATING DATA

When it comes to how data is collected, the second mentioned issue arises: the unusable
or incompatible data. Regarding this, the following must be stated beforehand: in science,
experiments and observations must be conducted according to a set of rules if they are to
be considered valid and their results usable: they have to be representative and replicable
(Bollen, Cacioppo, Kaplan, Krosnick, & Olds, 2015) — amongst others, more specific ones,
though those two are golden rules of science —, and any data collected outside this scope
or through non-described methods is unreliable and thus useless for research purposes
(Ioannidis, 2005).
But even when these conditions are met, some studies are performed with such specific
purposes in mind — not few times corresponding to non-scientific principles, like
political/sponsor interests or limited time —, that the data they obtain can’t be used for
pretty much anything else outside that precise scope (Geijzendorffer et al., 2016; Honrado
et al., 2016). This gives rise to the third mentioned issue, as this data cannot then be used
together with data from other studies, thus spatially restricting the knowledge of changes
in biodiversity to the locations where each study is executed and blocking us from
assessing its global trends (Henry et al., 2008).
In order to solve this problem, authors and research groups have proposed methods
such as (1) the creation of standardized frameworks for the planning and execution of
biodiversity related studies (Green et al., 2005; Pereira, 2006) — suggesting a series of
steps, more or less detailed, that every study should oblige to, independently to their
actual goal, in order to make the data collected usable in further studies and for more
general biodiversity monitoring purposes —, to be used both on a regional and global
levels; (2) new, less work intensive and remote ways of monitoring global biodiversity,
allowing the monitoring of large areas instead of conducting local field studies through
the use of satellite captured images — those images would be used to monitor key
indicators from which diversity can, directly or indirectly, be evaluated, allowing for
general idea of biodiversity at a national and international level that can be used as an
early warning system to possible major shifts in the distribution of organisms (Duro,
Coops, Wulder, & Han, 2007) —; and (3) the introduction of a set of defined variables that
should be included in every study, that would remain useful in the analysis of biodiversity
not only on that specific study but also for further future evaluations of the area, on this

5
 subject, it is worth mentioning and further developing the concept of the Essential
Biodiversity Variables (EBVs), as proposed by GEO BON (Pereira et al., 2013).

ON THE ESSENTIAL BIODIVERSITY VARIABLES

EVBs are defined, as stated by GEO BON, as “the derived measurements required to study,
report, and manage biodiversity change, focusing on status and trend in elements of
biodiversity”, and they “should play the role of brokers between monitoring initiatives and
decision makers”. Ideally each EBV should correspond to (1) a biological and (2) state
variable, which is (3) related to critical scales and dimensions of biodiversity, (4) sensitive
to change, (5) as equally as possible applicable to any ecosystem, and (6) technically and
economically viable and sustainable through time, thus being useful for both conveying
information to policy makers about biodiversity and still gathering relevant data for other
scientific studies on the field (https://geobon.org/ebvs/what-are-ebvs/).

EBV CLASS EBV CANDIDATE (VARIABLE)

Co-ancestry
Allelic diversity
Genetic composition
Population genetic differentiation
Breed and variety diversity
Species distribution
Species populations Population abundance
Population structure by age/size class
Phenology
Body mass
Natal dispersion distance
Species traits
Migratory behaviour
Demographic traits
Physiological traits
Taxonomic diversity
Community composition
Species interactions
Net primary productivity
Secondary productivity
Ecosystem function
Nutrient retention
Disturbance regime
Habitat structure
Ecosystem structure Ecosystem extent and fragmentation
Ecosystem composition by functional type

Table 1 . EBV classes and candidates, according to GEO BON (https://geobon.org/ebvs/what-are-ebvs/), as of

May 2018.

6
 Taking in mind these aspects, an extensive list of EBVs was created comprising putative
key variables to evaluate changes in biodiversity; still it is in an ever-changing state due
to constant works to assess the actual value of each variable as an EBV and reduce the
list’s length (https://geobon.org/ebvs/working-groups/). Currently, the list is at 6 EBV
classes, with a total of 22 variables among them (Table 1).
However, when stress-testing this concept against its usefulness for reporting for
several policy instruments, its performance varies greatly, and it shows several
constraints that are still in need of tweaking before the concept can be applied in full as
intended (Geijzendorffer et al., 2016).

MODELLING FOR MONITORING BIODIVERSITY

When talking about modelling for biodiversity, one mostly refers to species distribution
models (SDMs), which employ information on the environment to infer on a species’
habitable space (Kearney & Porter, 2009): put simply, these models are tools for mapping
a species’ ecological needs and plot its distribution throughout time.

MODELLING STRATEGIES

Before going into the models usefulness, one should first talk about the models
themselves. Currently, for the building of these models, there are two main schools of
thought: (1) the correlative and (2) the mechanistic. These differ mainly in the data from
the study entity that they require.
The first consists in mapping the presences and absences of a target entity and
assessing the ecosystem variables throughout a sampled region, from which the entity’s
tolerance ranges to the ecosystem variables inserted in the model are delimited, and then
projecting that tolerance across a bigger area to statistically produce a probability value
of occurrence of the entity throughout the whole — both sampled and not sampled —
region (Kearney & Porter, 2009). Within the inserted parameters — assuming there are
enough and good parameters —, goes a representation of the abiotic processes and
interactions that affect the entity’s survival, which represent a big plus for the strategy;
however, the observed causalities may not be direct, maybe being in reality interlinked
with biotic interactions or non-inserted variables collinear with the selected ones, thus

7
ending up with the SDM’s not actually representing the mapping of the entity’s
fundamental niche (Fordham et al., 2018).
For a mechanistic approach, the entity’s functional characteristics are what is
important to evaluate. Physiological variables of the entity are asserted and then plotted
on a map according to the region ecosystem variables — thus suppressing the need for
sampling presences and absences on a training region to calibrate the rest of it —
quantifying the relation between the entity’s characteristics and the environment, and
producing a map of the entity’s fitness across the target region (Kearney & Porter, 2009).
This has the great upside of not requiring an extensive previous knowledge of the species
distribution, but adds to the process the difficulty of understanding the fundamental
relations between the entity’s characteristics and the many environmental factors.
In terms of results, both models seem to behave similarly, producing generally
overlapping results, with the correlative approach usually overestimating the study
entity’s range (Kearney, Wintle, & Porter, 2010), and thus the choice between them mostly
comes to what data is available.

SYSTEMATIC CONSERVATION PLANNING

Independently of whichever strategy one chooses when modelling, this tool provides
substantial advantages when planning for conservation — be it to just monitor or actually
manage it through the designing of protected areas and such. With model assisted
systematic planning, in comparison to classic random sampling based methods, it is
possible to accomplish more precise and longer time-spanning assessments of the
biodiversity of the areas worth of such efforts, therefore improving their cost-efficiency
relation (Silvia B. Carvalho, Gonçalves, Guisan, & Honrado, 2016; Margules & Pressey,
2000).
When using models for this, Carvalho et al (2016) suggests a flexible seven stages
workflow for designing multispecies monitoring schemes using correlative modelling
techniques, though the subjacent planning can be used for single species and with
mechanistic modelling too. The workflow goes as following: (1) clearly determine the
objective, for as in any kind of practical science, one should know what one wants to
accomplish; (2) choose taxa to monitor — this step is exceptionally important for
assessing biodiversity in general, when the conservation work is to be addressed to a

8
single species or even a whole small genus, this choice is already implied in the previous
step —, since monitoring everything is an idyllic but impossible task, the choice of what
to monitor must be taken into serious account, as the selected taxa have to be
representative of the target region biodiversity (cf. Kati et al., 2004); (3) compile
occurrence data for the selected taxa — if going for a mechanistic approach instead it is
needed to experiment or search for existing data on the taxa ecophysiology —; (4) predict
the taxa distribution using the data from the previous step to fill the gaps between the
sampling results and its true range — or, again if using a mechanistic approach, to find
out from scratch the taxa distribution, since it is not considered before —, and allowing
for an optimization of the costs and efforts by predicting the spots where the quantity of
taxa and/or frequency of each is higher; (5) optionally collect data on the study region
environment variables and sectioning to improve the diversity of the to-be-selected
monitoring sites, as knowledge of the existence of “sections” within the region — be them
areas with different land cover, policy protection, etc. — can allow for a better assessment
of changes in the taxa presence throughout time; (6) set minimum quantity of monitoring
sites to effectively obtain the information stated as the objective in the first step; and
finally (7) design the monitoring network by choosing the optimal sites, maybe even
designing more than one network with different degrees of complexity and according to
different budgets and, preferably, making them compatible so that if it is mandatory to
start with a more modest plan, it remains possible to later upgrade it to a more complete
one (cf. Amorim, Carvalho, Honrado, & Rebelo, 2014).

MODELLING FOR THE FUTURE

One of the greatest strengths of using models when tackling biodiversity issues is, as it
has already been stated, their predictive abilities (Mouquet et al., 2015), therefore they
can be used to help conservation biologists in forecasting many species shifts and
movements and thus better advice for policies that can tackle many of the oncoming
problems on the rise with the ever present threat of climate change (Sílvia B Carvalho,
Brito, Crespo, Watts, & Possingham, 2011; Urban et al., 2016), including the issue of
invasive species that benefit from those changes and threaten the native biodiversity
(Vicente et al., 2016).

9
 Models can do this by integrating within their computations the several predictions for
climate change on certain time intervals and then, following a framework as the
aforementioned (Amorim et al., 2014), crossing the environmental needs and resilience
of a taxa with the changes on the environment itself for each of the climate change
scenarios, produce snapshots of that taxa occurrence as far as across the globe for each of
those intervals, which provide a dynamic prediction of the available suitable locations for
the taxa for as long as the prediction goes — with the uncertainty obviously increasing
with the distance from the present.

CONCLUSIONS

Starting by the most straightforward conclusion, thus regarding the modelling topic of
this work, it seems most obvious that models are indeed and invaluable tool for the
improvement of many of the monitoring schemes nowadays employed, and that one can
only expect their popularity and usage to increase with time. Their integrative power also
appears to be a welcome tool to help with the problems comprising the disperse and
incompatible — for previous methods — existent information, though it shouldn’t be
considered as an excuse to not improve the ways of data gathering and storage.
Moving back to the first discussed topic, it seems also clear that the way much of the
global scientific community has been dealing with their data in the past years — being not
at all global — needs to be completely revised and changed, as the globalization that we
see witness in everything else should also be applied to the way we do science, allowing
for an around the globe sharing of good information that can be used in favour of
biodiversity and every one of us.

10
REFERENCES

Amorim, F., Carvalho, S. B., Honrado, J., & Rebelo, H. (2014). Designing Optimized Multi-Species Monitoring Networks to Detect Range
Shifts Driven by Climate Change: A Case Study with Bats in the North of Portugal. PLoS ONE, 9(1), e87291.
http://doi.org/10.1371/journal.pone.0087291
Bachelet, D., Neilson, R. P., Lenihan, J. M., & Drapek, R. J. (2001). Climate Change Effects on Vegetation Distribution and Carbon Budget
in the United States. Ecosystems, 4(3), 164–185. http://doi.org/10.1007/s 10021-001-0002-7
Bernstein, A. (2012). Biodiversity and human health. In A. A. Aguirre, R. S. Ostfeld, & P. Daszak (Eds.), New Directions in Conservation
Medicine: Applied Cases of Ecological Health (pp. 45–55). New York: Oxford University Press.
Boakes, E. H., McGowan, P. J. K., Fuller, R. A., Chang-qing, D., Clark, N. E., O’Connor, K., & Mace, G. M. (2010). Distorted Views of
Biodiversity: Spatial and Temporal Bias in Species Occurrence Data. PLoS Biology, 8(6), e1000385.
http://doi.org/10.1371/journal.pbio.1000385
Bollen, K. A., Cacioppo, J. T., Kaplan, R. M., Krosnick, J. A., & Olds, J. L. (2015). Social, behavioral, and economic sciences perspectives on
robust and reliable science. Report of the Subcommittee on Replicability in Science Advisory Committee to the National Science
Foundation Directorate for Social, Behavioral, and Economic Sciences.
Bonnet, X., Shine, R., & Lourdais, O. (2002). Taxonomic chauvinism. Trends in Ecology & Evolution, 17(1), 1–3.
http://doi.org/10.1016/S0169-5347(01)02381-3
Carvalho, S. B., Brito, J. C., Crespo, E. G., Watts, M. E., & Possingham, H. P. (2011). Conservation planning under climate change:
Toward accounting for uncertainty in predicted species distributions to increase confidence in conservation investments in
space and time. Biological Conservation, 144(7), 2020–2030. http://doi.org/10.1016/j.biocon.2011.04.024
Carvalho, S. B., Gonçalves, J., Guisan, A., & Honrado, J. P. (2016). Systematic site selection for multispecies monitoring networks.
Journal of Applied Ecology, 53(5), 1305–1316. http://doi.org/10.1111/1365-2664. 12505
Clark, J. A. (2002). Taxonomic Bias in Conservation Research. Science, 297(5579), 191b–192.
http://doi.org/10.1126/science.297.5579.191b
Danielsen, F., Mendoza, M. M., Alviola, P., Balete, D. S., Enghoff, M., Poulsen, M. K., & Jensen, A. E. (2003). Biodiversity monitoring in
developing countries: what are we trying to achieve? Oryx, 37(4), 407–409. http://doi.org/10.1017/S0030605303000735
Duro, D. C., Coops, N. C., Wulder, M. A., & Han, T. (2007). Development of a large area biodiversity monitoring system driven by
remote sensing. Progress in Physical Geography, 31(3), 235–260. http://doi.org/10.1177/0309133307079054
Fordham, D. A., Bertelsmeier, C., Brook, B. W., Early, R., Neto, D., Brown, S. C., … Araújo, M. B. (2018). How complex should models be?
Comparing correlative and mechanistic range dynamics models. Global Change Biology, 24(3), 1357–1370.
http://doi.org/10.1111/gcb.13935
Garrett, K. A., Dendy, S. P., Frank, E. E., Rouse, M. N., & Travers, S. E. (2006). Climate Change Effects on Plant Disease: G enomes to
Ecosystems. Annual Review of Phytopathology, 44(1), 489–509. http://doi.org/10.1146/annurev.phyto.44.070505.143420
Geijzendorffer, I. R., Regan, E. C., Pereira, H. M., Brotons, L., Brummitt, N., Gavish, Y., … Walters, M. (2016). Bridging the gap between
biodiversity data and policy reporting needs: An Essential Biodiversity Variables perspective. Journal of Applied Ecology,
53(5), 1341–1350. http://doi.org/10. 1111/1365-2664.12417
GEO BON. What are EBVs? (2018, April 11). Retrieved May 17, 2018, from https://geobon.org/ebvs/what-are-ebvs/
GEO BON. Working groups (2018, April 11). Retrieved May 18, 2018, https://geobon.org/ebvs/working-groups/
Green, R. E., Balmford, A., Crane, P. R., Mace, G. M., Reynolds, J. D., & Turner, R. K. (2005). A Framework for Improved Monitoring of
Biodiversity: Responses to the World Summit on Sustainable Development. Conservation Biology, 19(1), 56–65.
http://doi.org/10.1111/j.1523-1739.2005.00289.x
Haeberli, W., & Hoelzle, M. (1995). Application of inventory data for estimating characteristics of and regional climate-change effects
on mountain glaciers: a pilot study with the European Alps. Annals of Glaciology, 21, 206–212.
http://doi.org/10.3189/S0260305500015834
Henry, P.-Y., Lengyel, S., Nowicki, P., Julliard, R., Clobert, J., Čelik, T., … Henle, K. (2008). Integrating ongoing biodiversity monitoring:
potential benefits and methods. Biodiversity and Conservation, 17(14), 3357–3382. http://doi.org/10.1007/s10531-008-
9417-1
Honrado, J. P., Pereira, H. M., & Guisan, A. (2016). Fostering integration between biodiversity monitoring and modelling. Journal of
Applied Ecology, 53(5), 1299–1304. http://doi.org/10.1111/1365-2664.12777
Hortal, J., Jiménez-Valverde, A., Gómez, J. F., Lobo, J. M., & Baselga, A. (2008). Historical bias in biodiversity inventories affects the

11
 observed environmental niche of the species. Oikos, 117(6), 847–858. http://doi.org/10.1111/j.0030-1299.2008. 16434.x
Ioannidis, J. P. A. (2005). Why Most Published Research Findings Are False. PLoS Medicine, 2(8), e124.
http://doi.org/10.1371/journal.pmed.0020124
IPCC, I. P. O. C. C. (2007). Climate Change 2007 - The Physical Science Basis. In Solomon S, Qin D, Manning M, Chen Z, Marquis M, Averyt
KB, et al., editors. Contribution of working group 1 to the fourth assessment report of the intergovernmental panel on climate
change (p. 1009). http://doi.org/volume
Kati, V., Devillers, P., Dufrêne, M., Legakis, A., Vokou, D., & Lebrun, P. (2004). Testing the Value of Six Taxonomic Grou ps as
Biodiversity Indicators at a Local Scale. Conservation Biology, 18(3), 667–675. http://doi.org/10.1111/j.1523-
1739.2004.00465.x
Kearney, M. R., & Porter, W. (2009). Mechanistic niche modelling: combining physiological and spatial data to predict spec ies’ ranges.
Ecology Letters, 12(4), 334–350. http://doi.org/10.1111/j.1461-0248.2008. 01277.x
Kearney, M. R., Wintle, B. A., & Porter, W. P. (2010). Correlative and mechanistic models of species distribution provide congruent
forecasts under climate change. Conservation Letters, 3(3), 203–213. http://doi.org/10.1111/j.1755-263X.2010.00097.x
Lejeusne, C., Chevaldonné, P., Pergent-Marti ni, C., Boudouresque, C. F., & Pérez, T. (2010). Climate change effects on a miniature
ocean: the highly diverse, highly impacted Mediterranean Sea. Trends in Ecology & Evolution, 25(4), 250–260.
http://doi.org/10.1016/j.tree.2009.10.009
Loehle, C., & LeBlanc, D. (1996). Model-based assessment of climate change effects on forests: a critical review. Ecological Modelling,
90(1), 1–31. http://doi.org/https://doi.org/10.1016/0304-3800(96)83709-4
Margules, C. R., & Pressey, R. L. (2000). Systematic conservation planning. Nature, 405(6783), 243–253.
http://doi.org/10.1038/35012251
McDonald-Madden, E., Baxter, P. W. J., Fuller, R. A., Martin, T. G., Game, E. T., Montambault, J., & Possingham, H. P. (2011). Should we
implement monitoring or research for conservation? Trends in Ecology & Evolution, 26(3), 108–109.
http://doi.org/10.1016/j.tree.2010.12.005
Mouquet, N., Lagadeuc, Y., Devictor, V., Doyen, L., Duputié, A., Eveillard, D., … Loreau, M. (2015). REVIEW: Predictive ecolo gy in a
changing world. Journal of Applied Ecology, 52(5), 1293–1310. http://doi.org/10.1111/1365-2664. 12482
Pacioni, C., Hunt, H., Allentoft, M. E., Vaughan, T. G., Wayne, A. F., Baynes, A., … Bunce, M. (2015). Genetic diversity loss in a
biodiversity hotspot: ancient DNA quantifies genetic decline and former connectivity in a critically endangered marsupial.
Molecular Ecology, 24(23), 5813–5828. http://doi.org/10.1111/mec.13430
Pereira, H. . (2006). Towards the global monitoring of biodiversity change. Trends in Ecology & Evolution, 21(3), 123–129.
http://doi.org/https://doi.org/10. 1016/j.tree.2005.10.015
Pereira, H. M., Ferrier, S., Walters, M., Geller, G. N., Jongman, R. H. G., Scholes, R. J., … Wegmann, M. (2013). Essential Biodiversity
Variables. Science, 339(6117), 277–278. http://doi.org/10. 1126/science.1229931
Scholes, R. J., Walters, M., Turak, E., Saarenmaa, H., Heip, C. H., Tuama, É. Ó., … Geller, G. (2012). Building a global observing system for
biodiversity. Current Opinion in Environmental Sustainability, 4(1), 139–146. http://doi.org/10.1016/j.cosust.2011. 12.005
Sheil, D. (2001). Conservation and Biodiversity Monitoring in the Tropics: Realities, Priorities, and Distractions. Conservation Biology,
15(4), 1179–1182. http://doi.org/10. 1046/j.1523-1739.2001.0150041179.x
Stein, B. A. (2002). Taxonomic Bias and Vulnerable Species. Science, 297(5588), 1807c–1807.
http://doi.org/10.1126/science.297.5588.1807c
Tittensor, D. P., Walpole, M., Hill, S. L. L., Boyce, D. G., Britten, G. L., Burgess, N. D., … Ye, Y. (2014). A mid-term analysis of progress
toward international biodiversity targets. Science, 346(6206), 241–244. http://doi.org/10.1126/science. 1257484
Urban, M. C., Bocedi, G., Hendry, A. P., Mihoub, J.-B., Peer, G., Singer, A., … Travis, J. M. J. (2016). Improving the forecast for biodiversity
under climate change. Science, 353(6304), aad8466-aad8466. http://doi.org/10.1126/science.aad8466
Vicente, J. R., Alagador, D., Guerra, C., Alonso, J. M., Kueffer, C., Vaz, A. S., … Honrado, J. P. (2016). Cost-effective monitoring of
biological invasions under global change: a model-based framework. Journal of Applied Ecology, 53(5), 1317–1329.
http://doi.org/10.1111/1365-2664.12631
Webb, T. J., vanden Berghe, E., & O’Dor, R. (2010). Biodiversity’s big wet secret: The global distribution of marine biological records
reveals chronic under-exploration of the deep pelagic ocean. PLoS ONE, 5(8). http://doi.org/10. 1371/journal.pone. 0010223

12