Article Science & Technology Studies 32(4)

On the Plurality of Environmental Regimes of

Anticipation: Insights from Forest Science and
Management
Antoine Dolez
Université Grenoble Alpes, CNRS, Sciences Po Grenoble ; Irstea Lessem, France / antoine.dolez@umrpacte.fr

Céline Granjou
Université Grenoble Alpes, Irstea Lessem / LISIS (Laboratoire interdisciplinaire sciences innovations
sociétés), France

Séverine Louvel
Université Grenoble Alpes, CNRS, Sciences Po Grenoble, France

Abstract
In recent years, the social sciences have increasingly investigated ways in which futures are anticipated,
fostered, and pre–empted. However, less attention has been given to how various predictive
approaches inform different ways of acting in the present. Our article presents the results of an
investigation into the current practices and agendas of forest scientists and managers in France. We
first suggest how an anticipation of environmental futures is coming to the fore as an emerging field
of expertise and practices in forest sciences, including predicting but also monitoring, preparing
and adapting to projected futures. We then account for the co–existence of three ‘micro–regimes’ of
anticipation combining a certain approach to the forest, a certain vision of the future, and a certain
type of scientific predictive approach, including different anticipatory objectives, different modelling
practices, and different interactions between research and management: i/ Adapting forestry to future
climates; ii/ Predicting Future Tree Biology; iii/ Monitoring forests as indicators of climate change.

Keywords: regime of anticipation, climate change, forest science, foreknowledge

Introduction
In recent years, the social sciences have increas- Granjou et al., 2017). A recent special issue in the
ingly investigated the ways in which futures are Sociological Review is emblematic of the call for a
anticipated, fostered, and pre–empted (Adams new “engagement with and interrogation of the
et al., 2009; Tavory and Eliasoph, 2013; Andersson future in social sciences” (Coleman and Tutton,
and Duhautois, 2016; Coleman and Tutton, 2017; 2017: 441). It suggests “shifting the emphasis from

78
 Dolez et al.

looking into the future to looking at the future— change because they are often perceived as the
that is, to engage with the future as an analyti- “lungs of the world” and are thus a key factor in
cal object” (Coleman and Tutton, 2017: 441). This reducing carbon dioxide emissions. In this article,
article is located at the crossroad between two we document the way in which a growing range
streams of literature: first, the emerging schol- of forest scientists, field observers and managers
arship which builds on cultural geography and are realigning their agendas, practices and goals
anthropology research to question how ideas of around new anticipatory agendas and standards
the future inform actions in the present; second, associated with the circulation of climate change
the specific insights of Science and Technology projections and anticipatory concerns. By doing
Studies into the production and role of scientific so, we shall highlight the plurality of anticipatory
forecasts and models in various academic fields research agendas and predictive technologies that
and disciplines. In addressing how the anticipa- forest scientists have developed and how they are
tion of socio–environmental futures in a changing embedded within various visions of forest and
climate are coming to the fore as a new scientific forest futures as well as within contrasted relation-
and political agenda, this article aims to scrutinize ships between research and forest management.
the coexistence of various ‘regimes’ of predic- We will first present brief historical insights
tive and anticipatory knowledge production and into the evolution of forests and forestry practices
their embedment in several and partly conflict- in France and emphasise how recent concerns
ing politics of environmental anticipation. It fol- about climate change have fostered new antici-
lows Mike Michael’s suggestion to account for the patory agendas and practices in forest science
‘ecology of futures’ at play in the intertwinement and management. We shall describe how climate
of both the ‘Big Futures’ at stake in broad societal change topics and concerns are transforming
and ecological narratives, and the ‘Little Futures’ forest science organisations, collaborations and
pertaining to everyday social life and interactions material infrastructures of knowledge, including
(Michael, 2017). Our key question is: how does cli- practices of data production, and how this process
mate change, as the embodiment of a ‘Big Future,’ of ‘climatisation’ (Aykut et al.., 2017) involves
play out on the practices of forest modellers, and increasing exchanges and collaborations between
specifically on the way they anticipate the evolu- forest science, ecology and climate science and
tion of their research agendas and contributions the models that were previously developed in
to forest management? Our contribution shall isolation from each of those fields. Then we shall
account for the various Big and Little Futures and document the co–existence of three micro–
their relations at stake with the growing focus regimes of anticipation in the case of French
of environmental sciences on climate change forest science and management and eventually
anticipation. account for their tensions and relations. Each of
Our contribution focuses on the case of forest these micro–regimes combines a vision of the
science and forestry and unpacks how the antici- future with an approach to the forest, including
pation of forest evolution in future climates is a certain type of scientific predictive approach
currently gaining ground in a rapidly growing associated with modelling practices. In particular,
field of research, expertise and management. we will discuss Aradau and Van Münster’s (2013)
Forest management is embedded within complex and Amoore’s (2013) idea that the rise of future–
and interwoven issues of ecological sustain- oriented knowledge agendas destabilises pre–
ability and profitability. Its future prospects in existing scientific approaches based on the
a changing climate are highly uncertain and interpretation of past data and requires disrup-
foster new concerns about how to anticipate the tive epistemology and practices. We argue that
changing patterns of tree growth, species distri- the future does not necessarily disrupt previous
bution, plant disease outbreak, forest produc- epistemic practices and organisations; instead,
tivity and economic profitability, as well as the the production of foreknowledge is embedded in
overall adaptation of forestry practices. Forests various, situated visions of the specific future and
are at the forefront of the “battle” against global knowledge.

79
 Science & Technology Studies 32(4)

Empirical and theoretical Theoretical background: ‘climatisation’ and

background the ‘micro–regimes’ of anticipation
The case of French forest science and The STS literature has long addressed the produc-
management tion of predictive knowledge and the particular
status and role of predictions in science–policy
This article is based on a sociological investigation
interface (in systems dynamics: Bloomfield, 1986;
into forest science and management in France,
economy: Collins and Pinch, 1996 and geo-
including about 30 interviews with forest scien-
sciences: Sarewitz et al., 2000; Dahan–Dalmedico,
tists, forest managers and field correspondents
2006; Edwards, 2010). STS scholars have notably
and backed up with the reading of forest science
criticized the “quest for a scientifically legitimated
articles, administrative literature and institutional
view of the future” through the development of
websites. The forest scientists we interviewed are
scientific predictive models (Sarewitz et al., 2000:
members of the main disciplines involved in forest
367). Much attention has been given to why pre-
science (mostly ecology, but also biology, genet-
dictive models do not or cannot produce accu-
ics and computer science) and various research
rate predictions (Collins and Pinch, 1998) and
institutions, including France’s National Center for
how those predictions should be communicated,
Scientific Research (CNRS), the Research Institute
received and used (Sarewitz et al., 2000). Recent
of Science and Technology for Environment and
STS and post–ANT scholarship also insist on the
Agriculture (IRSTEA) and the National Institute
performative role of models, scenarios and simu-
for Agriculture Research (INRA). Semi–structured
lations and their effects on shaping the reality,
interviews were focused on scientists’ professional
for instance, how economics shape and perform
trajectory and career, their vision of the scientific
economy (Callon, 1998; Mackenzie et al., 2007)
field including modelling and predicting the evo-
and more broadly how differing scientific mod-
lution of forests, and their vision of the future and
els and approaches shape various “ontologies”
the type of knowledge they build on it. In the case
(Mol, 2002; Law and Mol, 2002), as they encom-
of forest managers, interviews included a focus
pass several different ways of knowing and inter-
on management practices and potential uses of
vening upon it. Our point, however, is concerned
models and simulations.
less with these ontological politics than with the
What makes France an interesting example of
embedment of predictive models within various
the development of various predictive and antici-
and partly conflicting politics of environmental
patory regimes regarding forests and climate
anticipation.
change? It is mostly the strong interdependence
STS scholars started scrutinizing the internal
between forest science and forest management
plurality of predictive approaches and models in
which can be traced back to the French histor-
the case of glaciology (yet see: Skrydstrup, 2017).
ical tradition of centralised forest science and
However, less attention has been given to how
forest management, linked to the high economic
differing predictive approaches inform various
and social importance of forests to the nation
ways of acting in the present. Recent literature
(Decocq et al., 2016). Today forests cover around
examined the investigation of official counter–
30% of the French mainland with 75% being
terrorism programmes and practices and elabo-
private forests and 25% public forests, a third of
rated on the expansion of a ‘politics of possibility’
which are managed by the National Forests Office
that aims “not to prevent the playing out of a
(ONF) while the rest is owned by local councils.
particular course of events on the basis of past
The ONF, which employs around 10,000 people
data tracked forward into probable futures but
today, was created in 1964 and is in charge of
to pre–empt an unfolding and emergent event in
wood production, forest protection and tourism.
relation to an array of possible projected futures”
The Research and Development department of
(Amoore, 2013: 9). Aradau and van Münster
the ONF promotes exchanges, collaborations and
also depicted an all–encompassing ‘Regime of
knowledge transfers between forest science and
Anticipation’ including the development of a
forestry.
new “conjectural episteme” in which imagina-

80
 Dolez et al.

tion mostly replaces the use of past data sets to of agendas and practices with climate issues. We
attempt to make the future “knowable” (Aradau argue that ‘climatisation’ importantly involves
and van Münster, 2011). the realignment of knowledge practices toward
We argue that those theorisations do not anticipatory objectives and agendas, including
properly account for how scientists, experts, predicting but also monitoring, preparing and
policymakers and managers associate practices, adapting to projected futures. We shall describe
infrastructures and imaginaries to anticipate the the ‘climatisation’ of forest science organisa-
‘not yet’. We propose to develop the notion of tions, collaborations and material infrastruc-
‘micro–regime’ of anticipation in order to empiri- tures of knowledge, including practices of data
cally account for those various assemblages and production and modelling, and how this involves
their mutual tensions. We further criticize the idea increasing exchanges and collaborations between
there is only one Science (Knorr–Cetina, 1999) forest science, ecology and climate science—as
by unpacking the diversity of foreknowledge well as between the models that were previously
and predictive models in the case of forests and developed in isolation in each of those fields.
climate change. Our notion of ‘micro–regimes’ of Eventually, we shall document the coexistence of
anticipation is inspired by Pestre’s approach to different competing micro–regimes of anticipa-
the micro–historical embedment of the produc- tion that forest scientists and managers deploy
tion of knowledge into socio–economic regula- and how those anticipatory micro–regimes are
tion (Pestre, 2003) and departs from descriptions embedded in various (and partly competing)
of the historical development and succession of processes of research agenda setting and environ-
broad regimes or ‘styles of knowing or reasoning’ mental issue framing.
from past to present (Hacking, 1994; Kwa, 2011).
While Pestre’s notion of ‘regime of knowledge’
Anticipatory pluralism in forests
puts forward the interactions between science,
science and management
politics and society on a macro–social level, the
notion of ‘micro–regimes’ is located at the smaller The ‘climatisation’ of forest science and
level of mundane research agendas and practices management
(Shinn, 1999). ‘Micro–regimes’ of anticipation are
ways of negotiating the co–production of ‘Big Almost all the forest scientists and managers we
Futures’ (here embodied in narratives of climate met spontaneously referred to climate change
change and its impacts on forests) together with in their answers when asked about on–going
‘Little Futures’, which pertain to routine research changes in their research agenda and practices,
practices and interactions, including developing indicating that aligning one’s research and agenda
models, collaborations and projects, etc. (Michael, with climate change had become necessary in
2017). order to attract funding. Beyond the rhetorical
The construction of climate change as a global reference to climate change, forest scientists and
concern and expertise has been extensively practitioners also suggested that they were now
documented (see in particular: Jasanoff and de confronted with very practical questions related
Martello, 2004; Edwards, 2010). However, research to the anticipation of forest growth and productiv-
is only emerging on how climate change issues ity under future climates, such as: How will rising
and concerns are now reframed as local concerns temperatures, soil acidification and water scarcity
in a wide range of sectors and activities whose influence tree growth and forest species compo-
practices, communities, jobs and identities are sition? Which species will be the most resilient to
being transformed and re–aligned toward antici- future droughts, heat waves or storms? When will
patory objectives which relate to various situated be the best moment to harvest timber produc-
activities and agendas. Sociologists Aykut, Foyer tions? Our field work thus confirms bibliometric
and Morena (2017) proposed the notion of ‘clima- analyses which suggest that climate change has
tisation’ to depict this multi–level and highly become a central topic in forest science, along
contextual process of re–alignment of a range with others such as “carbon dioxide” and “adapta-

81
 Science & Technology Studies 32(4)

tion” (Aleixandre–Benavent et al., 2017). However, sation’ of forest science and management has
we found that ‘climatisation’ (Aykut et al.., 2017) triggered the convergence of empirical, forestry–
occurred not only at a discursive level—meaning oriented models with other types of models that
the integration of the topic of climate change into have long developed in isolation from forestry
forest science agendas and discourses—but also concerns and practices, i.e. process–based, statis-
within organisations, collaborations and mate- tical models (Korzukhin et al., 1996; Adams et al.,
rial infrastructures of knowledge. For instance, 2013). The latter models have been developed in
the ONF’s R&D department was reorganised in biology and in ecology since the beginning of the
2005 along five topics including climate change. twentieth century in the wake of the equations
A range of new research and management net- of Lotka–Volterra (Leslie, 1948). They aim to
works were created in order to address the impact understand simple or fundamental biological
of climate change on forests and forestry, such processes such as photosynthesis (Farquhar et al.,
as the GIP ECOFOR (“ECOsystèmes FORestiers”) 1980) or carbon allocation and to translate them
which was founded in 1993 and has been support- into equations. These equations can then fuel a
ing research on forests and climate change since computer program that simulates “virtual experi-
the beginning of the 2000s, including research ments” (Legay, 1997).
on the future distribution of French forests in a These two types of models have long been
changing climate1. Forest scientists and managers developed in isolation from each other, as
also started implementing new climate–related forestry–oriented models were taught in forestry
data collection and new collaborative arrange- schools and process–based models in ecology
ments around climate data sharing and use, for and biology master’s degrees. In the early 2000s,
instance within the network of forest observation growing interest in understanding the evolution
sites called RENECOFOR, initially created in rela- of forests in the context of climate change led to
tion to concerns over the effects of acid rain on the development of new models that blur that
forests. In 2012, RENECOFOR was also integrated distinction by mixing the characteristics and
into a Long–term Environmental Research Moni- objectives of the two former categories of models,
toring and Testing System (SOERE), whose goal is as a forest scientist explained:
to produce data on the “System Earth” dynamic,
illustrating how the collection of forest and cli- Process–based models attempted to summarise
mate related data becomes embedded within ecosystem functioning without any predictive
new collaboration arrangements between a objective […] however, with climate change issues
we discovered that this dichotomy did not work
broad range of disciplines such as ecology, forest
anymore […].We understood that empirical models
science, genetics and population biology.
should also explicitly integrate climatic data into
The ‘climatisation’ of forest science and
their architecture and equations; on the other
management also fostered the development of hand, process–based models should also address
new methods for modelling forest growth and prediction and applications…
making decisions regarding which tree species to
choose and when to fell trees for timber produc- As a result, many of the forest scientists we inter-
tion. Such decisions were first made on the basis of viewed use climate scenarios designed by the IPCC
the production quotas, which calculate the annual and other climate–oriented research institutions
growth of one species in a particular place, and (such as Météo–France which develops climate
served as decision guidelines for when to fell the projections with great precision at the local level
trees and which tree species will grow faster. From in France) and integrate them into pre–existing
the mid–1980s, forest engineers have developed forestry–oriented models in order to simulate the
empirical models in order to predict the timber potential growth of trees in a changing climate.
productivity of a particular forest in a constant On the other hand, while climate projections
environment, including soil and climate. Those are increasingly used and integrated into forest
models are fuelled by long–term data produced models, forest models are also increasingly inte-
by the National Forest Inventory. The ‘climati- grated into climate change modelling which

82
 Dolez et al.

tends to become more and more integrative and of the type of knowledge that matters for forest
biological (while initial climate models relied on and forest management, including certain visions
atmosphere physics and chemistry only). Science of the extent to which the future disrupts past and
historian Amy Dahan–Dalmedico accounted for present scientific practices and technologies.
how climate change models shifted in the 1990s Following the analytical distinction introduced
from a focus on the atmosphere to a broader by Mike Michael (2017), each ‘micro–regime’
focus on ‘Earth systems’ integrating oceans and of anticipation shapes and performs both “Big”
terrestrial surfaces, i.e. vegetation and forests and “Little” futures including ecological futures
(Dahan–Dalmedico, 2010). As a result, not only (climate change), economic futures (forestry
do forests scientists use climatic models in forest evolution), and academic and scientific futures
research, but forests models also fuel broader (research agenda setting, maintenance or creation
climate models that simulate the interactions of collaborations, publication writing, etc.). These
between the atmosphere and the biosphere. For are ways of negotiating their coexistence and
instance, the Laboratory of Climate and Envi- potential tensions at various levels.
ronmental Sciences in Paris–Saclay, which hosts
climate scientists actively involved in the IPCC, First ‘micro–regime’: Adapting forestry to future
has developed the ORCHIDEE model, which climates
simulates the role of tree development and life In the first ‘micro–regime’ of anticipation,
cycles in carbon flows in the biosphere, including researchers and managers seek to predict the
the ORCHIDEE–FM submodel which integrates composition and geographical distribution of
the effects of various forestry strategies on carbon forest and forest socio–ecologies in future cli-
cycle. mates. Their scientific practices are those tradi-
The ‘climatisation’ of forest sciences thus tionally used in forest science and engineering.
involves increasing exchanges and collabora- They assess timber stocks using forest inventories
tions between forest science, ecology and climate and maps and they construct statistical models
science and the models that were previously that build on correlations between a wide range
developed in isolation in each of those fields. Yet of ecological, geophysical and socio–economic
far from an all–encompassing alignment towards parameters in order to estimate the productivity
a unique anticipatory “episteme” (Aradau and van of future forests, to write guidelines and design
Münster, 2013), our fieldwork also points to the forests policies. That foreknowledge is meant to
co–existence of a plurality of research agendas help produce guidelines for present and future
dedicated to anticipating the future of forests forestry practices and to contribute to forestry
under a changing climate. These research agendas economic planning and adaptation. The vision
differ in two ways: they develop forest–driven of the future focuses on securing future forestry
vs. climate–driven science on one hand. On the activities. Forests are considered to be anthropo-
other hand, they handle forestry–oriented versus genic, managed socio–ecosystems.
ecological and biological process–based research. The anticipatory logic of this ‘micro–regime’ is to
improve forestry strategies by taking into account
Three ‘micro–regimes’ of anticipation climate change. What matters is that the model
This section describes three ‘micro–regimes’ of can be applied in order to guide forestry choices. A
anticipation which became apparent during our key objective is to identify the forest practitioners’
fieldwork. Each of these combines three dimen- leverage actions to secure forestry in a changing
sions: a certain vision of the forest, a certain idea climate. This first ‘micro–regime’ thus gathers
of the future, and a certain type of scientific pre- forest scientists and engineers from various
dictive approach, including different anticipatory research institutes (IRSTEA, INRA), the R&D depart-
objectives, different types of models and model- ment of the National Forest office, and members
ling practices, and different interactions between of forest research networks such as GIP ECOFOR.
research and management (see Table 1). The three Close relations and collaborations between forests
‘micro–regimes’ encompass actors’ various views scientists, managers and decision–makers are an

83
 Science & Technology Studies 32(4)

Table 1: Three ‘micro–regimes’ of anticipation.

Adapting forestry Predicting Future Monitoring forests as indi-

to future climates Tree Biology cators of climate change
1. Vision of forests Forests are con- Forests are consid- Forests are considered to
sidered to be ered to be a functional be an observatory of the
socio–ecosystems. system governed by evolution of climate change.
ecological processes.
2. Visions of The future is viewed The future is disruptive. The future becomes pal-
the future on past and pres- Researchers and manag- pable and knowable,
ent trends. ers assume that studying and thus governable.
fundamental ecological
processes in a chang-
ing climate is essentially
different from study-
ing them under stable
climatic conditions.
3. Type of scientific predictive practices
3–i/ Anticipatory Adapting forestry Understanding the Producing indicators of
logic and objectives strategies in a ecological processes at climate change, assessing
changing climate stake in the evolution of and mapping its evolution
forests (such as carbon
flows and water scarcity)
3–ii/ Model- Statistical “meta– Simple ecological and Simple model illustrat-
ling practices models” that aggre- process–based models ing the causal relation
gate a broad range of between the chosen indica-
ecological, social and tor and climate change
economic variables
3–iii/ Interactions Strong collabora- Researchers contribute Contributions to the French
between research tions between forest as experts to biodiversity Ministry of Environment;
and management science and forest and nature conserva- Co–construction of indi-
management: data tion international and cators between forest
sharing, collaboration national organizations researchers and managers.
in research projects,
and co–production
of forestry guidelines

essential characteristic of this micro–regime; they mortality, forests composition, light interception)
result both from collaborations and from indi- and socioeconomic variables (i.e. forestry strat-
vidual mobility between organizations dedicated egies, expected timber stock). The researchers’
to forest management (such as the National Forest objective is to run multiple simulations with
Office) and organisations dedicated to scientific different models to foresee the consequences of
research and to producing technical support and their potential forestry’s strategies. As a result,
advice for foresters (such as INRA and IRSTEA). models tend to accumulate and integrate an
The models at stake in this first ‘micro–regime’ ever increasing number and variety of variables,
are constructed and calibrated using a large as suggested for instance by the case of the
amount of data collected during forest inven- SAMSARA model, a tree–growth model whose
tories and by research teams. They integrate many versions were developed over time in order
both ecological variables (i.e. tree growth and to achieve the integration of an ever wider range

84
 Dolez et al.

of processes, including wind damage (Ancelin et I’m not interested in ecology as a science. I take an
al., 2004), colonisation processes (Cordonnier et interest in forest as a socio–ecosystem; it’s my point
al., 2006), intraspecific competitions (Vieilledent of view, but I can’t avoid ecology because it is one
et al., 2010) and biodiversity’s stock, ecosystems of the analytical and theoretical sciences I use to
study forests. (Modeller working at the National
services (Courbaud et al. 2017; Lafond et al. 2017)2
Forest Office and INRA)
and climatic parameters (while the model was
initially designed under a constant climate) (Lagar-
As a result, in the first micro–regime, the future
rigues, 2016). The implementation of the CAPSIS
is mostly deduced from past and current trends
modelling platform in 1998 (Dufour–Kowalski et
detected by using ever bigger sets of data. As a
al., 2012) also illustrates the trend towards the
modeller reported: “Yes, I’m interested in the
increase in the number of variables as the platform
future, but most of the time, it is the past that I
integrates about 70 different forest models in only
study.” However, both modellers and managers
one simulation software and makes it possible to
are aware that the future may destabilize and dis-
run all of them together in order to predict how
rupt past and present trends. They address the
any given variable will react in a changing climate.
disruptive character of the future by accumulat-
An engineer in charge of CAPSIS told us:
ing models and variables—thus, rising comput-
In CAPSIS, we have tree–growth models. These
ing power—and by developing new statistical
models grow trees; they create virtual forests. Then, methods, such as the Bayesian approach, which
you can add other models to them, for example, “aims to artificially break with the linear structure
timber quality models or risk models, to address of time”, a modeller said. These tools are standard-
how the forest may resist wind storms (…) or ized for example in R-packages (Jabot et al., 2013).
economic models. All those second–level models Referring to Michael’s distinction between
can be added to CAPSIS tree–growth models in the ‘Big’ and ‘Little’ futures (Michael, 2017), one could
same simulator. say that the Big Future envisioned by forest
researchers and managers in this micro–regime is
Accordingly, the models used in the first ‘micro– about securing forestry under a changing climate,
regimes’ must be user–friendly—the calculation while the Little Futures that are at play in the
speed is a key element in this: “If it takes too much everyday life of forest research teams include:
time to set up the data and run the simulation, it is • Developing ever more sophisticated and
off–putting. It is a beautiful theoretical tool but in integrative statistical models that take into
practice it is useless” (Forests modeller). When we account an increasing range of parameters in
asked modellers and managers about potential order to produce forestry guidelines;
future improvements of the models, all of them • Developing new technologies of data–collec-
referred to an increase in the calculation’s speed. tion, including remote–sensing technologies,
The development of remote–sensing using LIDAR in order to improve the quantity and quality
(Laser Imaging Detection and Ranging) is also of the data available;
expected to help collect ever larger amounts of • Reforming forest management and the insti-
data to fuel the models with the idea that scien- tutions in charge of it in order to secure future
tific progress meant both more and more past and forestry under a changing climate.
present data for more and more anticipative mod-
els. Accordingly, the first micro–regime does not Second ‘micro–regime’: anticipating future
focus on knowing the ecological processes but tree biology
relies on statistical methods applied to big data- In the second ‘micro–regime’ of anticipation,
sets to find correlations between climatic, biologi- researchers aim to understand and predict
cal and socio–economic information. Modellers the evolution of forest ecological functioning
use ecological processes such as tools to parame- by modelling how future climates—including
terise the models; however, producing knowledge extreme events such as severe droughts (Estiarte
on those processes is not part of their objectives: et al., 2016; Lempereur et al., 2017)—will affect
the way trees use an array of resources (water,

85
 Science & Technology Studies 32(4)

carbon, nutrients, light). This ‘micro–regime’ is no ger environmental conditions that will be essen-
longer related to utilitarian objectives (i.e. design- tially different from the environmental past and
ing forestry practical guidelines and economic present, as one of the forest ecologists we met
planning). Instead, it should be understood in explained:
the more environmentalist perspective of under-
standing the ecological processes at stake in the As soon as we have something calibrated to the
evolution of forests under future climates. Some present… I mean, it is tempting to apply the model
of them contribute as experts to biodiversity and to the future and to see what will happen. But
the question is: Is the knowledge of the system
nature conservation organizations, such as the
in the current climate sufficient to be applied to
Intergovernmental Platform on Biodiversity and
future scenarios, including extreme conditions?
Ecosystem Services or the French Foundation for
… I want to know whether my little model, which
Biodiversity Research. This micro–regime aggre- is calibrated to current conditions, using 10 years
gates what scientists call “simple” process–based of data collection, could be applied to extreme
biological modelling, in natura experiments and events, such as a six–month water shortage. That
ecological theories. is to say, things that you have rarely or never
Ecologists and modellers involved in this observed so far…
‘micro–regime’ tackle ecological questions such
as: What is the link between diversity and stability? In a similar way, another forest ecologist explained
(Morin et al., 2011) How resilient is the ecosystem that “modelling is not interesting when everything
when faced with scarcity? Which function of the is all right,” meaning he did not expect simulation
ecosystems will be the first to react to a changing models to make a linear business–as–usual predic-
climate (Gustafson et al., 2015)? Compared with tion based on past data, but instead to be able to
the first ‘micro–regime’, ecology—especially integrate future extreme events and to deal with
functional ecology—plays a central role and the disruptive nature of climate future. With that
is no longer merely considered a tool. For the goal, ecologists and modellers seek to capture the
researchers involved in this ‘micro–regime’, under- non–linear responses of ecosystems by develop-
standing the dynamic of forest ecosystems is an ing both modelling and in natura experimentation
opportunity to understand fundamental ecolog- (Perez–Ramos et al., 2010). For example, the same
ical processes at stake, such as photosynthesis ecologist developed an experimentation consist-
or the allocation of carbon between plants (Gea– ing of excluding rain for a few months (using artifi-
Izquierdo et al., 2015), atmosphere and soil and cial covers to protect plants from the rain) in order
between different parts of the plants, as explained to enrich his model:
by a forest scientist whom we interviewed:
It has allowed me to add a few modules that were
I’m not interested in the holm oak, but rather in… not in the first version of the model. The model
how forests respond to scarcity (…) In that case, we calibrated to the current climate works up to
worked on the holm oak. Yet what matters for us a precise threshold, and once this threshold is
is really the functioning, the functional aspects in crossed, you have to add a [new] module... it is
terms of ecosystems, carbon flow, growth, primary something that my models initially did not take
productivity … into account, and now I am developing it in order
to simulate the non–linear relationship between
The anticipatory logic of this second ‘micro– water scarcity and fire risk.
regime’ addresses climate change as a disruptive
event that forces forest modellers to renew their Field experimentations are meant to help set
practices and develop collaborations with experi- and observe the possible future climatic condi-
menters and functional ecologists. Researchers tions and their impacts on trees and forests. The
assume that studying fundamental ecological development of experimentations on the effects
processes in a changing climate is essentially dif- of climate change on ecosystem functioning is
ferent from studying them under stable climatic emblematic of what some interviewed forest
conditions. Climate change is thought to trig- ecologists called an “experimental turn” that dates

86
 Dolez et al.

back to the mid–2000s. Yet some forest ecologists tree biology, developing in natura experimental
whom we interviewed emphasized the difficulty infrastructures, improving the linkages between
or even impossibility of conducting experiments process–based models and field experimenta-
on trees in completely controlled ‘live labs’ such tions, and contributing to biodiversity conserva-
as the controlled experimental enclosures called tion and management organizations.
‘ecotrons’ that have been constructed in France in
the 2000s (and elsewhere in the world3) because Third ‘micro–regime’: Monitoring forests as indi-
of scale issues (ecotrons are not designed to host cators of climate change
more than two–meter–high plants) (Granjou and In the third ‘micro–regime’ of anticipation,
Walker, 2016). As a consequence, they have devel- researchers and managers consider the evolution
oped field experiments that consist of condition- of forests as a case–study for observing and pre-
ing gas concentrations or simulating climates dicting the evolution of climate and its broader
(for instance, droughts) in the field by using, for impacts on ecosystems and society. They monitor
instance, flux towers which enrich the air in CO2 forest fauna and flora, collect field data and com-
and measure gas concentration and temperature bine the data with population biology models
at the bottom and at the top of the studied trees and laboratory experiments in order to develop
(Misson et al., 2010). indicators of the growing intensity and impacts
In this ‘micro–regime’, relationships between of climate change. While the models used in this
researchers and managers are less close than ‘micro–regime’ are meant to be simple models,
in the first micro–regime. This is partly because like in the second ‘micro–regime’, the ultimate
researchers would rather belong to laboratories objective is not to understand basic biology but
and centres favouring fundamental research over to represent the state and evolution of climate in
applied research (such as the French National order to inform policy–makers, in particular from
Center for Scientific Research CNRS)—even the French Ministry of Environment, in order to
though the institutions of belonging do not design environmental and climate policies. This
systematically determine the development of ‘micro–regime’ mostly builds on scientific results
applied versus fundamental research by their produced by the first two ‘micro–regimes’ (as, for
members. This is also because when researchers instance, the budburst indicator which builds on
produce knowledge on ecological processes, it the modelling of the various steps of plant growth
makes it difficult for them to connect to forest and life cycle, Chuine, 2000). Eventually the third
management issues and concerns, as a forest micro–regime differs from the first two as it does
scientists working at the CNRS (Center for Evolu- not tackle “the forest” per se but focuses on cli-
tionary and Functional Ecology) explained: mate change and only makes use of the forest as
a proxy to make climate change visible: the very
Relationships with managers remain difficult object of anticipation differs between the first two
because we face a cultural issue. I mean, we ‘micro–regimes’ and the third one.
consider forests as an ecosystem. It is an ecological In the third ‘micro–regime’ of anticipation,
point of view, and they consider forests as a
researchers from the National Institute for Agri-
production means, something cultivated. In
culture Research (INRA) and field correspondents
foresters’ minds, the forest is something we
from the ONF monitor and record forest data
cultivate, and it is not a natural ecosystem. They
are concerned about productivity, plantations
in order to document and assess the ongoing
or species selection and not really about how evolution of climate. Here, forest data are useful
trees use what they have and how they manage to the extent that they can be directly linked to
available resources. (…) [W]e are definitely not on climate change, such as, for instance, shifting
it. budburst dates which are thought to be caused
by warmer springs (Chuine and Cour, 1999). Forest
In the second ‘micro–regime’, the Big Future is data are conceived as climate change indicators
about anticipating future tree biology under a when the relation between the indicator and
changing climate. This is enacted through ‘Little climate change is considered to be simple and
Futures’ that include improving simple models of almost causal. The ONERC (National Observatory

87
 Science & Technology Studies 32(4)

on the Effects of Global Warming), which is part or E. Coli in order to study fundamental biological
of the French Ministry of Environment, defines mechanisms. While the idea in both cases is to
an indicator as “information tied to a phenom- focus on a ‘simplified case’ in order to understand
enon [that shows] its evolution through time in an a more complicated general issue, in our third
objective way.”An indicator is like a thermometer: ‘micro–regime’, however, forest scientists’ ultimate
“As the body’s temperature gives a hint about a goal is not to study basic ecological mechanisms
patient’s health, climate change indicators tell us (like in the second micro–regime) but to deduce
about the Earth’s state” (National Observatory on trends of on–going climate change from certain
the Effects of Global Warming, 2010). As in the first characteristics of forest evolution. Accordingly, in
‘micro–regime’, indicators are conveyed to both a similar manner to the statistical models of the
forest managers and politicians to inform them first ‘micro–regime’, researchers and managers
of the evolution of climate change and allow for expect to be able to collect ever more field data in
designing appropriate forest policies. Therefore, order to improve the accuracy of climate change
the form and aesthetic of the indicator are central indicators.
because this is the way in which climate change Results are presented under the form of a map
is made visible and palpable. This ‘micro–regime’ that shows the past, present and future progres-
gathers scientists from a broad range of disci- sion of the “colonisation front” of the caterpillar.
plines such as genetics, population biology or Therefore, the anticipatory logic at stake here
entomology. Researchers produce indicators and puts past, present and future in linear succes-
information for the French Ministry of Environ- sion in order to make climate change visible and
ment that are meant to be collective reference palpable. Compared to the two other ‘micro–
landmarks on the intensity and impacts of climate regimes’, the focus of the third micro–regime
change for policy–making. is on raising the alarm and alerting managers
One of the best–known indicators in France and decision–makers on the progress of climate
is the pine processionary caterpillar (Rossi et change with the idea that the future is already
al., 2015), which is a forest parasite that causes here and we have to act now in order to adapt to
tree death and health problems inhumans and it. Accordingly, the vision of the future is both in
animals. Pine processionary caterpillars have continuity with past and present at the level of
very stinging hairs that can cause skin problems scientific and modelling practices that use past
to both humans and pets. Researchers initially data of caterpillar populations and distribution,
started studying this insect because it is one of and disruptive at the political level because it
the most dangerous European parasites for forests invites both managers and politicians to act now
as they eat pine needles and cause the tree to die. in order to adapt to the future.
The progressive change of focus in the research In this ‘micro–regime’, Big Futures are about
devoted to the caterpillar, from pest to climate assessing the progress of on–going climate
indicator, gives another example of the ‘climati- change and alerting decision–makers, involving
sation’ of forest research agendas and practices Little Futures that include finding ever more
(Roques, 2015). Since the creation of the ONERC accurate and simple–to–use indicators of climate
in 2001, the caterpillar has indeed become a change and their impacts, collecting ever more
central indicator for assessing the evolution of data to fuel indicator levels and maps, and
climate once the relation of its growing numbers fostering policy changes and adaptation strate-
and shifting geographic distribution with climatic gies.
parameters was established in the literature. The
processionary caterpillar was thus progressively Relations and interactions between the
built as a “reference model” meaning “a model of three micro–regimes
response to climate change” (to quote a forest Our inquiry accounted for the co–existence of
researcher whom we interviewed). This latter three ‘micro–regimes’ of anticipation that seek
notion of ‘model’ has some similarities with the to anticipate the evolution of forests under a
way biologists use ‘model organisms’ such as mice changing climate. We found that researchers in

88
 Dolez et al.

the three micro–regimes belonged to different started to become more frequent. As a result, he
scientific institutions and published in differ- stopped making predictions and started devel-
ent scientific journals, suggesting that the three oping research on more basic processes. Another
‘micro–regimes’ operate in relative isolation from modeller started his scientific career by working
each other. The first micro–regime spans across on trees’ large–scale distribution in various forests
traditional agriculture and forest science and and using correlative models and inventory data
management institutions such as INRA, IRSTEA sets—a research activity that falls under the first
or ONF, whereas the second one unfolds in more micro–regime of anticipation. However, during
basic research institutions such as CNRS. The third his career, his interest shifted to theoretical
‘micro–regime’ involves both traditional forest ecological processes such as the functioning and
research and management institutions (INRA, evolution of the diverse tree species in a given
IRSTEA, and ONF) and policy–making institutions forest and he began studying the general link
(the French Ministry of the Environment). Bib- between an ecosystem’s diversity and its stability,
liometric analysis also shows that researchers in then meeting the approaches favoured under the
the first ‘micro–regime’ publish in academic for- second ‘micro–regime’.
est science and management such as, for exam- A number of researchers also criticized the
ple, Annals of Forest Science or Forest ecology and first ‘micro–regime’ as ‘fashionable’ but not robust
management, while researchers in the second enough in scientific terms. They meant that too
‘micro–regime’ publish in ecology and global many variables and data put together as models
change journals such as Global Change Biology are run in order to obtain long–term previsions
or Ecological Letters. In the third ‘micro–regime’, with little attention to the precise biological
they publish both in genetics or population biol- mechanisms and diversity at stake. Importantly,
ogy journals such as Journal of Applied Entomology while researchers in the first ‘micro–regime’
or Insect Biochemistry and Molecular Biology; they construct the future as the follow–up to both the
also publish papers in more applied forest man- past and the present, researchers in the second
agement journals and reports for the French Min- ‘micro–regime’ consider the future to be disrup-
istry of the Environment. tive. Therefore, they doubt models and simulation
However, certain scientists moved from the based on past data series to be able to properly
first to the second ‘micro–regime’ as, after being predict the future. They think that there are things
initially trained in forestry schools—such as the and rules that just cannot be known in advance,
National School of Water and Forests in Nancy— also implying a different vision of which sort of
they found positions in ecology laboratories in knowledge is worth developing, as the head of
research institutions such as CNRS. This is the case an important research centre in ecology in France
of a forest ecologist trained in the French National explained:
School of Forestry, who now develops funda-
mental ecological research into tree competi- I’m not able to model the rules of carbon allocation
tion, which falls under the second ‘micro–regime’. in a tree: how much carbon is allocated to the roots,
Other researchers also moved from the first to to the trunk, or some other thing? I’m not able
the second ‘micro–regime’ as they became more to give an equation and say: This is how it works.
Hence, I’m unable to make a prediction. Anyone
aware of the uncertainties and limits related to
who predicts the forest productivity or the timber
running big correlative models. This is, for instance,
stock in the 2100s if rainfall drops by 30% relies on
the case of a forest modeller who works on the
current carbon allocation rules. We know that these
evolution of fire risks in a changing climate. While rules will change, but we don’t know how to model
his research was initially focused on developing it. It doesn’t prevent us from making predictions,
correlative and statistical models, he progressively but what is their validity? I don’t know.
became convinced that he could not extrapo-
late future conditions by relying solely on past He also suggested that the recent development
and current data, because the system would not of the first ‘micro–regime’ was strongly driven
have the same behaviour if the extreme events and supported by managers and decision–mak-

89
 Science & Technology Studies 32(4)

ers’ high expectations in the capacity of forest the representation of forest ecological func-
science to produce long–term predictions on the tioning. This project led to comparing and inte-
future of forestry. In particular, he criticised the grating the correlative models developed in the
increasing development of integrative, statistical first micro–regime (such as BIOMOD: Thuiller,
models producing maps of species distribution at 2003) and the process–based models developed
very long term, for instance, 2100. To him, these in the second micro–regime (such as Phenofit:
correlative statistical models produce “beautiful Chuine, 2000). A third research project, Climator,
maps” but fail to address fundamental scientific was eventually developed in order to apply the
issues such as how ecosystem functioning would results of this integrative modelling approach to
be impacted by a changing climate. He argued the production of standardised guidelines for
that modelling practices should not be taken as agriculture and forestry. Its results were translated
a scientific result per se: “So, we have a model, into a Green Paper aiming to help forest and agri-
and that’s it. We fuel it with anything and it out- culture managers and policy–makers anticipate
puts something. A model always gives you some the adaption of the timber and paper production.
result.” Instead, models should support the search Overall, we found that researchers’ commit-
for scientific answers to problems regarding eco- ments to ‘micro–regimes’ of anticipation are
logical mechanisms. partly related to, and dependent on, the types
Researchers working under the second ‘micro– of relations they have with forest managers.
regime’ thought that they were “closer to the Their relations with forest managers should not
biological reality” by developing models that take be considered end–products of their research
into account a very limited set of variables in a (as researchers disseminate scientific results to
very precise manner: managers in the form of guidelines, advice or
technical support) but also as determining the
I reduce everything to one parameter: quality Q. type of anticipatory research agenda and practices
But in fact, the precision with which I calibrate that researchers are developing (Granjou and
my “black box” takes into account the chemical Mauz, 2012). For instance, having close relations
diversity of the species... while the big categories
with research managers will provide resources
[used in statistical models] do not take it inaccount.
(i.e. funding, project partnerships) for developing
So, they told me that my research is a “black box,”
research agendas and activities aiming to predict
but it is their research that is a “black box”! (A forest
ecologist)
the future conditions of forestry and to help
forest managers (i.e. commitments to the first
The friction between the first and second ‘micro– micro–regime). On the other hand, researchers
regimes’ is thus linked to their visions of which committed to the second ‘micro–regime’ (antici-
sort of scientific advancement matters, i.e. to dif- pating future tree biology) and working in funda-
ferent visions of what future knowledge agendas mental research institutions usually have less
are worth developing and to different ideas of close relations with forest managers, who are not
what “good” forest science is. that interested in understanding the very basic
We also found a range of cases of combina- ecological mechanisms of trees in a changing
tion and collaboration between researchers climate. Eventually, researchers committed to the
from the first and second ‘micro–regimes’ of third ‘micro–regime’ (monitoring forests as indi-
anticipation, such as in the case of the CarboFor cators of climate change) have developed close
research project (2002 – 2004) (Loustau, 2004). relations with policy–makers (especially from the
The CarboFor project was the first French scientific Ministry of the Environment via the ONERC), and
project to develop integrative statistical models those relations help provide resources that fuel
based on IPCC scenarios in order to foresee the the agenda of monitoring forests as indicators
impacts of climate change on the distribution of of climate change. As a result, researchers’ antici-
trees over the long term. As these models progres- patory agendas and practices tend to co–evolve
sively appeared as being not precise enough, a with the nature and proximity of their relations
second research project, Qdiv, aimed to improve with managers: their various relations to forest

90
 Dolez et al.

managers are an important part of the shaping, tions). They also tend to co–evolve with the nature
stabilization and evolution of their commitments and proximity of their relations with forest man-
to ‘micro–regimes’ of anticipation. agers, as, for instance, having close relations with
research managers will favour research agendas
and activities aiming to predict the future condi-
Conclusion
tions of forestry and help forest managers. On the
Overall, our results show that ‘climatisation’ (Aykut other hand, researchers committed to the second
et al.., 2017) occurred in forest science and man- regime have fewer close relationships with forest
agement not only at a discursive level—meaning managers, while researchers committed to the
the integration of the topic of climate change into third ‘micro–regime’ have rather developed rela-
forest science agendas and discourses—but also tionships with national policy–makers in the field
within organisations, collaborations and material of forest, agriculture and environment.
infrastructures of knowledge, especially practices Let us briefly return to the recent literature on
of data collecting and modelling (i.e. the network the emergence of a ‘regime of anticipation’ that
of forest observation sites called RENECOFOR, challenges previous models of predicting the
and the platform of models CAPSIS). The ‘climati- future through the calculations of risk probability
sation’ of forest sciences also involves increasing based on past data series (Amoore, 2013; see also
exchanges and collaborations between forest sci- Aradau and van Münster, 2013). Instead of the rise
ence, ecology and climate science and the mod- of a global, all–encompassing regime of anticipa-
els that were previously developed in isolation in tion accompanied by a new “conjectural episteme,”
each of those fields. Yet, our fieldwork also points our fieldwork in the case of French forest science
to the co–existence of a plurality of research and management suggests that the transfor-
agendas dedicated to anticipating the future of mations of forest science and forestry practices
forests under a changing climate. These research aiming to anticipate climate change entail the
agendas differ in two ways: they develop forest– co–existence of various, partly conflicting antici-
driven vs. climate–driven science on one hand, patory ‘micro–regimes’ at work, whose goals and
forestry–oriented versus ecological and biological approaches to science, forest and the future are
process–based research on the other hand. Our different. Our fieldwork suggests that Amoore’s
results do not only suggest how groups of envi- and Aradau and Van Münster’s thesis does not
ronmental scientists, experts and decision mak- do justice to current changes in the knowledge
ers hold various and potentially conflicting views production practices which aim to capture envi-
regarding which research fields, predictive tech- ronmental changes and futures. Instead of one
nologies and anticipatory governance are worth unique way of constructing the future as an object
developing to produce sound science insights of knowledge and action, we found several antici-
into environmental futures, they also highlight the patory assemblages that seek to foresee the future
embedment of the predictive models developed evolution of French forests in a changing climate,
by forest scientists and experts within various and various visions of the extent to which the future
partly conflicting politics of environmental antici- is disruptive and a variety of practices and strat-
pation, including concerns for adapting forestry egies for producing future–oriented knowledge.
practices to future climates (‘micro–regime’ 1), While in the first ‘micro–regime’, predictions rely
for anticipating the ecological resilience of trees on assembling past and present data into ever
and forests (‘micro–regime’ 2) and for mapping more sophisticated and integrative predictive
and preparing the advent of climate change and projections, in the second regime, anticipating
its impacts (‘micro–regime’ 3). Forest researchers’ ecological changes requires scientists to modify
commitments to the ‘micro–regimes’ and their the core of their models in order to predict how
various visions of what knowledge matters not basic ecological mechanisms will evolve. In the
only correspond to their professional trajectory third ‘micro–regime’, researchers aim to produce
and affiliation (i.e. belonging to fundamental or ever more accurate indicators of the on–going
more management–oriented research institu-

91
 Science & Technology Studies 32(4)

and future progress of climate change drawing on certain futures being completely unaddressed
data collection. and unscrutinised (for instance, should the
Lastly, the embedment of a predictive research anticipation of forestry practices adaptation
agenda setting within an environmental issue become hegemonic over the anticipation of tree
framework and political choices suggests that species extinctions and ‘natural’ forest ecosystem
anticipatory pluralism is important in keeping destabilization). Documenting the variety of
a broad range of futures open to scientific and predictive scientific practices is instrumental in
public scrutiny. In line with the new political understanding the various and partly conflicting
sociology of science (Frickel et al., 2010), we ways in which environmental futures are known,
argue that this plurality is essential in avoiding predicted and acted upon.

92
 Dolez et al.

References
Adams HD, Williams AP, Xu C, Rauscher SA, Jiang X and McDowell NG (2013) Empirical and process–based
approaches to climate–induced forest mortality models. Frontiers in Plant Science 4: 438.
Adams V, Murphy M and Clarke AE (2009) Anticipation: Technoscience, life, affect, temporality. Subjectivity
28(1): 246–265.
Aleixandre–Benavent R, Aleixandre–Tudó J, Castelló–Cogollos L and Aleixandre J (2017) Trends in scien-
tific research on climate change in agriculture and forestry subject areas (2005–2014). Journal of Cleaner
Production 147: 406–418.
Amoore L (2013) The politics of possibility: risk and security beyond probability. Durham, NC: Duke University
Press.
Ancelin P, Courbaud B and Fourcaud D (2004) Development of an individual tree–based mechanical model
to predict wind damage within forest stands. Forest ecology and management 203: 101–121.
Andersson J and Duhautois S (2016) Futures of Mankind: The Emergence of the Global Future. In: Van
Münster R and Casper S (eds) In the Politics of Globality: Assembling the Planet. Oxford: Routledge, pp.106–
125.
Aradau C and van Münster R (2011) Politics of Catastrophe. Genealogies of the Unknown. London: Routledge.
Aykut SC, Foyer J and Morena E (eds) (2017) Globalising the Climate. COP21 and the ‘climatisation’ of global
debates. London: Routledge.
Bloomfield BP (1986) Modelling the world. The social constructions of systems.London: Blackwell.
Callon M (eds) (1998) The Laws of the Markets. Oxford, UK: Oxford University Press.
Chuine I (2000) A unified model for tree phenology. Journal of Theoretical Biology 707: 337–347.
Chuine I and Cour P (1999) Climatic determinants of budburst seasonality in four temperate–zone tree
species. New Phytologist 143: 339–349.
Coleman R and Tutton R (2017) Introduction to Special Issue of Sociological Review on ‘Futures in Question:
Theories, Methods, and Practices.’ The Sociological review 65(3): 440–447.
Collins H and Pinch T (2002[1998]) The Golem at large: What everyone should know about technology.
Cambridge: Cambridge University Press.
Cordonnier T, Courbaud B and Franc A (2006) The effect of colonization and competition processes on the
relation between disturbance and diversity in plant communities. Journal of Theoretical Biology 243(1):
1–2.
Courbaud B, Pupin C, Letort A, Cabanettes A and Larrieu L (2017) Modelling the probability of microhabitat
formation on trees using cross–sectional data. Methods in Ecology and Evolution 8(10): 1347–1349.
Dahan–Dalmedico A (2006) Les Modèles du futur. Changement climatique et scenarios économiques: enjeux
scientifiques et politiques. Paris: La Découverte.
Dahan–Dalmedico A (2010) Putting the Earth System in a numerical box? The evolution from climate
modeling toward global change Studies in History and Philosophy of Science Part B: Studies in History and
Philosophy of Modern Physics 41(3): 282–292.
Decocq G, Kalaora B and Vlassopoulou C (2016) La forêt salvatrice. Reboisement, Société et Catastrophe au
prisme de l’Histoire. Ceyzérieu (Ain): Champ Vallon.
Dufour–Kowalski S, Courbaud B, Dreyfus P, Meredieu C and de Coligny F (2011) Capsis: an open software
framework and community for forest growth modeling. Annals of Forest Science 69(2): 221–233.
Edwards PN (2010) A Vast machine: Computer Models, Climate Data and the Politics of Global Warming.
Cambridge and London: MIT Press.

93
 Science & Technology Studies 32(4)

Estiarte M, Vicca S, Peñuelas J et al. (2016) Few multiyear precipitation–reduction experiments find a shift in
the productivity–precipitation relationship. Global Change Biology 22: 2570–2581.
Farquhar GD, von Caemmerer S and Berry JA (1980) A biochemical model of photosynthetic CO2 assimila-
tion in leaves of C3 species. Planta 149: 78–90.
Frickel S, Gibbon S, Howard J, Kempner J, Ottinger G and Hess D (2010) Undone science: charting social
movement and civil society challenges to research agenda setting. Science, Technology & Human values
35(4): 444–473.
Gea–Izquierdo G, Guibal F, Joffre R, Ourcival JM, Simioni G and Guiot J (2015) Modelling the climatic drivers
determining photosynthesis and carbon allocation in evergreen Mediterranean forests using multiproxy
long time series. Biogeosciences 12: 3695–3712.
Granjou C and Mauz I (2012) Expert activities as part of research work. The example of biodiversity studies.
Science and Technological Studies 25(2): 5–22.
Granjou C and Walker J (2016) Promises that matter. Reconfiguring ecology in the ecotrons. Science and
Technology Studies 29(3): 49–67.
Granjou C, Walker J and Salazar J (2017) The politics of anticipation: On knowing and governing environ-
mental futures. Futures 92: 5–11.
Gustafson E, De Bruijn A, Pangle R et al. (2015) Integrating ecophysiology and forest landscape models to
improve projections of drought effects under climate change. Global Change Biology 21(2): 843–856.
Hacking I, (1994) Styles of Scientific Thinking or Reasoning: A New Analytical Tool for Historians and Philoso-
phers of the Sciences. In: Gavroglu K, Christianidis J and Nicolaidis E (eds) Trends in the Historiography of
Science. Dordrecht: Springer Netherlands, pp. 31–48.
Jasanoff S and Martello M (2004) Earthly politics. Local and Global in Environmental Governance. Cambridge
and London: MIT Press.
Jabot F, Faure T and Dumoulin N (2013) EasyABC: performing efficient approximate Bayesian computation
sampling schemes using R. Methods in Ecology and Evolution 4(7): 684–687.
Knorr–Cetina K (1999) Epistemic Cultures. How the sciences make knowledge Cambridge and London: Harvard
University Press.
Korzukhin MD, Ter–Mikaelian MT and Wagner RG, (1996) Process versus empirical models: which approach
for forest ecosystem management? Canadian Journal of Forest Research 26(5): 879–887.
Kwa C (2011) Styles of knowing. A new history of science from ancient times to the present. Pittsburgh: Univer-
sity of Pittsburgh Press.
Lagarrigues G (2016) Effects of environmental gradients on demographic processes and the durability of the
uneven–aged forest management: application of Approximate Bayesian computation to improve simula-
tion models. PhD Thesis, Université Grenoble Alpes, France.
Lafond V, Cordonnier T, Zhun M and Courbaud B (2017) Trade–offs and synergies between ecosystem
services in uneven–aged mountain forests: evidences using Pareto fronts. European Journal of Forest
Research 136(5/6): 997–1012.
Law J and Mol A (eds) (2002) Complexities. Social Studies of Knowledge Practices. Durham and London : Duke
University Press.
Legay JM (1997) L’expérience et le modèle. Un discours sur la méthode. Paris : INRA Editions.
Lempereur M, Limousin JM, Guibal F et al. (2017) Recent climate hiatus revealed dual control by tempera-
ture and drought on the stem of Mediterranean Quercus ilex. Global Change Biology 23(1): 42–45.

94
 Dolez et al.

Leslie PH (1948) Some further notes on the use of matrices in population mathematics. Biometrika 35(3/4):
213–245
Loustau D (2004) Final report of the project CARBOFOR. Carbon sequestration in the big forest ecosystems in
France. Quantification, spatialization, vulnerability and impacts of different climatic and forestry scenarios.
INRA/CNRS, France, June 2004.
Mackenzie D, Muniesa F and Siu L (2007) Do Economist Make Markets? On the Performativity of Economics.
Princeton: Princeton University Press.
Michael M (2017) Enacting Big Futures, Little Futures: Toward an ecology of futures. The Sociological Review
65(3): 509–524.
Misson L, Rocheteau A, Rambal S, Ourcival JM, Limousin JM and Rodriguez R (2010) Functional changes in
the control of carbon flux after 3 years of increased drought in a Mediterranean evergreen forest? Global
Change Biology 16(9): 2461–2475.
Mol A (2002) The Body Multiple: ontology in medical practice. Durham and London: Duke University Press.
Morin X, Fahse L, Scherer–Lorenzen M and Bugmann H (2011) Tree species richness promotes productivity
in temperate forests through strong complementarity between species. Ecology Letters 14(12): 1211–1219.
National Observatory on the Effects of Global Warming (2010) Climate Change’s indicators’ list. Report to the
French Ministry of the Environment.
Perez–Ramos IM, Ourcival JM, Limousin JM and Rambal S (2010) Mast seeding under increasing drought:
results from a long–term dataset and from a rainfall exclusion experiment. Ecology 91: 3057 – 3068.
Pestre D (2003) Regimes of knowledge production in society: Towards a more political and social reading
Minerva 41(3): 245–261.
Renecofor’s assessment committee (2006) Renecofor network’s evaluation. Suggestions for the future. 14th
November 2006
Roques A (2015) Processionary Moths and Climate Change: An Update. Springer Netherlands.
Rossi JP, Garcia J, Roques A and Rousselet J (2015) Trees outside forests in agricultural landscapes: spatial
distribution and impact on habitat connectivity for forest organisms. Landscape Ecology 31(2): 243–254.
Sarewitz D, Pielke RA and Byerly R (2000) Prediction: science, decision making and the future of nature. Wash-
ington DC: Island Press.
Shinn T (1999) Change or mutation? Reflections on the foundations of contemporary science. Social Science
information 38(1): 149–176.
Skrydstrup M (2017) Envisioning the future by predicting the past: Proxies, praxis and prognosis in paleocli-
matology. Futures 92: 70–79.
Tavory I and Eliasoph N (2013) Coordinating Futures: Toward a Theory of Anticipation. American Journal of
Sociology 118(4): 908–942.
Thuiller W (2003) BIOMOD – optimizing predictions of species distributions and projecting potential future
shifts under global change. Global Change Biology 9(10): 1353–1362.
Vieilledent G, Courbaud B, Kunstler G, Dhôte JF and Clark J (2010) Individual variability in tree allometry
determines light resource allocation in forest ecosystems: a hierarchical Bayesian approach. Oecologia
163(3): 759–773.

95
 Science & Technology Studies 32(4)

Notes
1 Another example is the creation of the A-FORCE network (“Adaptation des FORêts au Changement Clima-
tiquE” i.e. Adaptation of Forests to Climate Change) which comprises 15 research institutes, engineering
schools and the ONF, with the aim of promoting and supporting climate change knowledge transfer from
research laboratories to forest managers. In particular, A-FORCE seeks to avoid misunderstandings of
modelling and simulation results and their uncertainties among forest managers.
2 The founder of SAMSARA, who works at IRSTEA, is in frequent contact with the National Forest Office in
order to both collect data and to transmit SAMSARA results to forest managers with the goal of improving
forestry strategies. SAMSARA is also used for training forest practitioners.
3 Today, large controlled chambers for the measurement of gas exchanges between plants and the envi-
ronment exist in most major universities and agronomic institutes, for example New Zealand’s Biotron,
the Bioklima project in Norway, the ecotron projects in Germany and in Belgium. Many more ecotron-like
facilities are in progress.

96