Paper published on Deforestation Frontier Development in the Amazon

We recently published a scientific study on Deforestation Frontier Development in the Brazilian Amazon. The study called “Recent transformations of land-use and land-cover dynamics across different deforestation frontiers in the Brazilian Amazon” maps current deforestation frontiers and shows how agricultural development and land-use change coincide with frontier development.

We show that new frontier development is characterized by expanding crop production dynamics and that all frontiers are characterized by improvements in cattle productivity. As for settlements we observe increasing deforestation rates and a persistence of cattle-ranching which requires a larger amount of areas compared to annual or perennial crops. In all frontier types large deforestation events increased again after a steady decrease up to 2012 (see Rosa et al., 2012) which could mean that large landholders are returning to business-as-usual under the current policy regime. The study is published as Open-Access and can be viewed on the Science Direct Homepage here.

Deforestation in the Brazilian Amazon shifts to less accessible areas

Deforestation decreased in the Brazilian Amazon considerably between 2004 and 2012 and and is since more less stable. Research indicates that those reductions where not uniform across the basin which bears important consequences for policy design. The following graph shows deforestation as a function of travel time. It depicts that the participation of deforestation increased in remote areas over time which might ultimately increase the costs for command and control policies to reduce illegal deforestation. Command and control policies where however  an integral part of deforestation reductions in the past and this trend is alarming given the steep budgetary cuts for environmental agencies under current Brazilian government.

I found an easy reproducible code example to make animated GIFs in R which allowed me to produce these plots very quickly. They show how clear-cut deforestation developed developed over time since the all time high in 2004.

Overall Brazilians successfully reduced deforestation until 2012. Since 2012 the rates increased slightly. We can observe in this animated GIFs how the share in overall deforestation of different Amazon states changed.


States with traditionally high deforestation shares are Rondônia, Mato Grosso and Pará. They account  for the very large part of the total deforestation. Pará inherited the fist place from Mato Grosso with almost 60% share of total deforstation in 2009. However since 2010 also Mato Grossos share increased again. The state of Amazonas is a new emergent player where forest fall especially in the southern part (Boca do Acre). Amazonas state almost levels the share of Rondônia which was was in the 80s and 90s the state where large part of the deforestation took place. The five other states have overall low shares:


However if we compare how the states performed to reduce deforestation by their own standards the picture changes. In the following plot we compare the deforestation rates relative to the base year 2004 which had the highest rates of all times. We see that Pará, Rondônia and especially Mato Grosso where able to reduce overall deforestation even with the slight increase since 2012. The Amazonas state however is almost back to its level of 2004. This could indicate that the Amazonas state is a region to look out when it comes to future deforestation trends in the next couple of years.


The other five states also reduced deforestation to different degrees however we observe that Roraima, Maranhão and Amapá had even higher levels for some years after 2004 compared to their baseline. Acre and Tocantins reduced deforestation quickly but had sharp increases in 2016. These regional shifts show how diverse the region reacted to the environmental policies that where introduced in the last two decades two foster deforestation control.


Deforestation in the Brazilian Amazon after 2004, animated!

Radio interview on recent deforestation increase in the Brazilian Amazon

Latest deforestation data from PRODES suggest that there was an increase in clear cut deforestation of 29% in 2016. An area of almost 8.000 square-kilometers, equivalent to 1.6 Mio. football fields was deforested in 2016 with sharp increases along the agricultural frontier. Is this a sign that the trend for reducing deforestation in Brazil is going to revert again?

In the In this short radio-interview with Christoph König from SWR2 Impuls I discuss some of the eventual drivers of this recent change and possible solutions  to address this problem in the future. You can download the german podcast here.


Annual relative variation of clear cut deforestation in the Brazilian Amazon. Source: PRODES


Brazilian municipality evolution and time-series analysis

Sumário em português:

Analisando dados estatísticos do Brasil em escala municipal encontramos o problema da divisão de municípios ao longo dos anos. A evolução da malha municipal dificulta a análise de dados em series temporais por que os novos municípios não tem dados para os anos prévios a sua criação,- Além disso os velhos municípios que foram divididos não podem ser comparados antes e depois da sua divisão por que perderam território e gente. No seguinte texto apresento uma metodologia (em inglês) para lidar com este problema agregando os municípios em grupos baseados na evolução da malha municipal ao longo dos anos. Os grupos baseados nessa agregação podem ser baixados para os anos 1980, 1991 e 2000 referente à malha municipal de 2010. As agregações permitem analisar todo tipo de dado socioeconómico e demográfico do Instituto Brasileiro de Geografia e Estatística – IBGE e de outras instituições académicas e estaduais.

Long version

In statistical analysis we may encounter the problem that our units of observation change over time. This is true for the case of Brazilian municipality data where the units of observation change due to the fact that a lot of new municipalities were created during the last decades. In 2010 Brazil possessed 5567 municipalities of which 3991 where created between 1940 and 2010. Those new municipalities lack data for the previous years to their existence. If a municipality was e.g. created in 1994, there is no demographic data previous to 1994. Furthermore, since municipalities are not created out of nothing, one or multiple other municipalities might suddenly experience a decrease in population in the subsequent year (here 1995) because their areas and populations were divided. This is also true for a large set of other socio-economic variables that are produced in the studies of the Brazilian Institute for Geography and Statistics (IBGE) and others. If we analyze the development of municipalities over time we have to account for this problem in order to avoid artificial increases or decreases in our variables, that might be largely a legacy of administrative change.

In Brazil, municipality creation spurred especially in the 1980s in course of the decentralization policies and ongoing population increase in the rural and urban areas. Although IBGE provides information on the evolution of the so called “Malha Municipal“ I was not able to identify a method or data-set that specifically solved that problem.

Therefore, I developed a set of algorithms in R that help to group municipalities on a base year e.g. 1991 in order to create an aggregation level that is neutral to municipality change. I applied these algorithms for the years 1980,1991 and 2000, which are the years of the population censuses prior to 2010. Furthermore, I limited the analysis to the states of the Brazilian Amazon, since some manual edits have to be made to the raw data, if municipalities where e.g. misspelled by the staff that created the documentation on the evolution of the „Malha Municipal“. Doing it for the whole country with the R script should be still feasible with a few hours of extra work. If you are interested, I can provide you the script and some explanations on how to use it.

For the theoretical part I have two images that show how the aggregation is done. The fist image shows a typical situation on how municipalities might split up over the years. In this example between 1991 and 2000 the municipality “d“ developed out of „a“ alone. Between 2000 and 2010 “b“ developed out of „a“ and „c“.


Here we already get a notion of the complexity of the task since a new municipality often develops out of multiple others and group affiliation changes over  time. If we group these municipalities together we create aggregates from municipalities, that did not form an administrative unit in the past. This is a little drawback but there is no other way around it since “b“ might contain both, population data from „a“ and „c“.

Image 2 shows how the algorithm for creating the groups actually works.


For my analysis I used this Excel sheet from IBGE that contains information on how municipalities developed over time. IBGE also provides detailed, yearly information on the development of the “Malha municipal” in each state. With this raw-material it is theoretically possible to go back to any year of interest to get the minimum amount of aggregation necessary for your analysis. However, it takes some time to download and prepare the data in order to be comparable to the table linked above.

The results of my aggregation are shown below. You can download the Shapefile that contains the groups based on the year 1980,1991 and 2000. The data is based on the Malha Municipal from 2010 with three columns called “m_1980″,”m_1990” and “m_2000”. Those columns contain either the group affiliation or the geocode of the municipality if it was not affiliated to any group.

  • “m_1980” contains muncipalities that where part of a group in 1970 and where splitted during 1970 and 1980. Altogether 139 municipalities where created and 221 municipalities where part of the splitting process, hence they have a group affiliation.
  • “m_1991” contains muncipalities that where part of a group in 1980 and where splitted during 1980 and 1991. Altogether 263 municipalities where created and 497 municipalities where part of the splitting process, hence they have a group affiliation.
  • “m_2000” contains muncipalities that where part of a group in 1991 and where splitted during 1991 and 2000. Altogether 15 municipalities where created and 32 municipalities where part of the splitting process, hence they have a group affiliation.

Furthermore, you can find seven columns with the geocodes of those municipalities that gave origin to another municipality in each respective year.My proposed methodology gives the minimum amount of aggregation necessary.

In the data-set I utilized there were no new municipalities created between 2000 and 2010. I cannot tell for sure however, if this is due to a lack of actual creation of municipalities in the region or if the dataset from IBGE is outdated. Other sources point to the fact that at least between 2000 and 2010 the number of municipalities increased from 5507 to 5565. It is unclear however how many municipalities where created in the Amazon States. Checking with official shapefiles on the “malha municipal from IBGE” we observe that

  • in 2000: only 792 municipalities existed in the nine Amazon states
  • in 2007: 807  and
  • in 2015: 808

It is most probable therefore, that the data-set is somehow incomplete.

Last but not least let me point you to the fact, that besides the separation of municipalities also municipality boundaries and hence their areas changed during the years, which is not documented in a comprehensive way and might pose a problem depending on your type of analysis. You see a comparison in in the last image of the Acrean municipalities in 1991 (purple with black boundaries) and 2010 (gray with red boundaries).


One might assume that aggregation based on intersects might be also an adequate tool for this problem, however if combining intersections and the grouping based on the evolution of the “Malha Municipal”, the groups might get too many members to be useful in your analysis.

The implementation costs of forest conservation policies in Brazil

Tropical forest conservation is considered a low-cost option for climate change mitigation. But mitigation cost assessments have featured opportunity costs, neglecting policy implementation costs. Here we use official data to identify the Brazilian federal government’s operational and institutional budgets related to forest conservation policies implemented from 2000 to 2014. We distinguish the allocated and executed budgets of these policies, and provide scenario-based estimates of their cost-effectiveness. On average, Brazil spent US$ 1 billion/year on forest conservation policies at the federal level. Brazil’s substantial reduction in annual forest loss after 2004 was accompanied by a higher operational budget execution of disincentive-based policy instruments, and an absolute increase in both allocated and executed institutional budgets. The post-2004 successful mitigation effort represented additional implementation costs to the Brazilian federal government of US$ 308–923/ha of avoided deforestation, or US$ 0.87–2.60/tCO2 of avoided emissions. Factoring in also approximate municipal and state expenditures, these costs increase to US$ 385–1153/ha or US$ 1.09–3.25/tCO2. We conclude that implementations costs are non-trivial in size, including compared to estimates of land users’ opportunity costs. This has important implications for REDD + policy design, in the sense that implementation costs need to be adequately considered.

The full article can be found at this link:

Field trip to the Brazilian Amazon

These photos will show some impressions from my field trip to the Brazilian Amazon region in 2015. All in all I stayed about three month in the region to make interviews with local farmers about their production systems and their attitudes towards protection of the natural rain forest. My interviews covered a large area of the Amazon in the states of Acre and Mato Grosso.

1-mapThe state of Mato Grosso and Acre in red

My studies are part of a research-project at the Centre for Development Research in Bonn that is funded by the Robert-Bosch-Fundation.  The project seeks to understand how environmental policies can contribute to lower deforestation and foster sustainable land-use in tropical rain-forest regions.

During the field research I interviewed persons about their land-use choices including the use of new technologies to increase productivity and the role of deforestation in current land-use practices. I hope to understand how social and economic forces drive these phenomena and how the effectiveness of current policy approaches is influenced by those circumstances.

All in all we were performing more than 100 interviews with farmers in different socio-economic settings. We spoke to very poor producers who rely on a relatively small piece of land as their sole income source and very rich producers who possess thousands of hectares of agricultural land and have very diversified income portfolios the span into other sectors of the economy as well.Interview participantsTraditionally cattle ranching is one of the most important drivers of  deforestation in the Amazon region. In Brazil cattle ranching was the preferred development model to open up and occupy vast areas of pristine rain forest along the main infrastructure projects in the North. Throughout decades large amounts of public money were spent to incentivize and subsidize the expansion of cattle-ranching in the region. 


Growth of export value of Brazilian Cattle. Source: Mongabay

Up to date, a huge part of the cattle production-systems in the Amazon are characterized by very low technological production-levels. Farmers miss important inputs such as fertilizers or herbicides to maintain the productivity of pastures. Also there is also a lack of additional fodder in the dry season and adequate veterinary assistance to increase nutrition levels and productivity of the animals.

4-cowAnimals might get very skinny in the dry season if not fed with additional fodder.

With few inputs, productivity declines rapidly in the pasture and producers rely on deforestation to gather new land suitable for production. Since the environmental impacts of deforestation where not a political concern for many decades and land was relatively abundant, the cheapest form to maintain or increase production was to open up new areas for pasture creation.

5 def 4Freshly  deforested area in Northern Mato Grosso.

Like this more forest areas got burned to the ground. Up to date around 20% of the Amazon rain-forest have been cleared in Brazil and an equal amount of area is thought to be selectively logged. Since the early 1990s Brazilian Government tried to increase the control of deforestation by implementing new environmental regulations and  building up governmental entities to control land-take and usage.

6 ibama police

Heavily armed military forces  have to back up field  enforcement actions of  environmental  agencies like IBAMA.

New policies where successful to reduce deforestation in the last ten years by increasing law enforcement and putting more pressure on the local governments to address the problem. Still it is questionable to which degree those approaches will be able to maintain or further lower deforestation rates in the future as more recent data suggests.

7 prodes

Clear cut deforested areas in square kilometres as observed by the Brazilian Forest Monitoring Program PRODES from 1988 to 2015. Source: INPE.

Inside the agricultural sector, environmental policies have created huge tensions because they are associated with negative impacts on production levels and the income of farmers. A lot of producers criticize environmental policies  for being incongruent with what was supported or even demanded by policies in the past.

One main challenge to reduce the pressure on the forest is therefore to overcome the old production paradigm and develop land-use options that demand less areas be it either by increasing productivity inside cattle-systems or diversifying the production of cattle ranchers as such. By this environmental and socio-economical goals could be achieved in line.

8 sustainable 1

Animal confinement is a form of  intensive cattle-ranching that is able to reduce the amount of necessary land considerably.

8 sustainable 2A farmer showing palm- fruits from his palm-trees that are part of a combined silvo-pastoral land-use system that is able to increase income and carbon storage on existing pasture areas.

One big challenge to change the current production systems is the lack of adequate infrastructure. Even though there is evidence, that expanding infrastructure is a precondition for opening up new areas, this does not automatically mean that the worse the infrastructure the better off the environment.

We continuously observed that bad infrastructure might be worse for the environment than good infrastructure. With less access to reliable infrastructure, farmers prefer to have a product that can be commercialized throughout the whole year (cattle) instead of a crop that is normally commercialized after harvest where the farm might not be accessible due to rains. This is especially true for regions with few storage capacities (e.g. grain silos).

9 infra 4Bad infrastructure reduces accessibility through the rainy season…

9 infra 5…and puts  burdens to daily live e.g. through the  inability to travel. On our journey from Mato Grosso to Acre we got surprised by a sudden rain that took us of the road for several hours.

Furthermore high transportation costs from bad infrastructure might also increase significantly the costs for production inputs thereby increasing the attractiveness of deforestation over land reform (e.g. more expensive fertilizers or agricultural lime for reducing soil acidity).

9 infra 7Bad access to farms might eliminate the chance for more intensive land-use. Animals however can walk over broken bridges and wade through riverbeds themselves. Some farmers make their cattle walk over 200 km to take them to the next market.

Being a more viable option under poor infrastructural conditions cattle ranching is also favourable under difficult biophysical conditions such as steep slopes where mechanized agriculture is not viable. During our field-trip, regions with a very undulated topography where often partially characterized by cattle ranching.

10 hillyA pasture of a relatively small farm in Northern Acre

Besides economic reasons, cattle ranching also brings a lot of other benefits to farmers that are commonly less addressed in the literature. Cattle creation is for example significantly less labour demanding and often easier to apply and manage than agriculture. Even though farmers could significantly increase their income changing production e.g. to coffee plantations, they prefer cattle over crops for the relative easiness in  implementation.

Another benefit is the herd itself functioning like a rural bank account. A growing herd is like a  growing capital stock  that is easy exchangeable for cash throughout the whole year. Cattle can therefore serve as a form of social security since it provides money for emergency spendings or investments, particularly in regions, where financial services are sparse. As long as the herd is visibly growing the farmer feels to be on the safe side.

Cattle also brings financial stability to producers since it is less vulnerable to inflation and other economic shocks than other investment forms. By that it is a very attractive investment for either retirement or additional income of people that do not engage traditionally  in agriculture. To our surprise a lot of elder people like retired school teachers or state workers engage in cattle ranching to invest and increase their pensions.

Our upcoming publications will address those issues since they are key to understand the logic of cattle production, its attractiveness and strong persistence in the region as well as its implications for promoting sustainable land-use options in the future. Below you can find some more photos from the trip.

5 def 3A huge fire from deforestations close to the city of Colniza in Northern Mato Grosso. Heavy smoke during the dry season often covers huge areas and affects people living in cities nearby.

9 infra 10Infrastructure development everywhere. Besides huge construction sites it is difficult to maintain the existing road network that is often poorly paved and utilized by very heavy trucks…

IMG_20151209_145425…carrying for example trunks from legal and illegal wood extraction.

IMG_20151209_132546Machines to mix animal fodder for intensive cattle ranching are gaining momentum in central Mato Grosso …

IMG_20151209_132947As well as the expansion of industrialized grain production that often leads to the displacement of cattle ranching to more remote areas.

IMG_20151209_150349Besides being one of the most important rural economic activities, cattle ranching is also a social phenomena with the cowboy culture rising (very nicely documented by Jeffrey Hoelle). This photo shows a gathering of the local cattle elite at an auction of breeding animals at the agricultural fair ExpoAcre .

Making a movie with R

Since some models (like mine) do not always give a direct interface with results, R provides us with a powerful tool to play around with big piles of data models spit out. Obviously, you also may want to do this despite existing model interfaces!  The program R is making life often happy due to its endless possibilities. Since I am performing a vegetation model with individual trees modelled, one of my wishes was to see the forest in real life and 3D. So the forest is growing, waning, stabilizing and so forth. Making a movie with R!

All right, I directly admit: the movie is not made with R, but is calling another program from outside R; namely Image Magick in combination with the FFMPEG software. Nevertheless the code is totally in R and it works really nice.
If you have time-dependent graphs, plots, maps or whatsoever, you can use this neat function to make a moving pictures file. Of course you can also use a sleeper function within the R studio display, but it is easier for use to make a seperate movie.

First I will show you my personal code, which is making a movie of a growing tree stand and then explain it command by command:

TOT=floor(tlen/interval)    # decide on number of graphs needed, based on total time and time steps
for(step in 0:TOT){      # start the loop of making graphs here!
  t=1+(interval*step)             # step counter  
  # decide on some specific graph-parameters before and put them in 1 dataframe
  # open the file in which the graph is drawn: the following if-statements are to make sure the graphs are numbered well, 
  # so 100 is not appearing before 1 and so on...
  if(step<10){             # make sure generated graphs are numbers from 1-999
    jpeg(filename=sprintf("%sMovie/%s/Flash_00%s.jpeg",Folder,nameshort,step), width=1000, height=800, units="px")
  } else if(step<100){
    jpeg(filename=sprintf("%sMovie/%s/Flash_0%s.jpeg",Folder,nameshort,step), width=1000, height=800, units="px")
  with(TreeStruct, {
    s3d <- scatterplot3d(gridyx1, gridyx2,HTrees,             # x y (grids) and z (height) axis in scatterplot3d package
                         color=CTrees, pch=FTrees,            # color (species) and form (canopyform) of symbols
                         cex.symbols=STrees,                  # symbol size (canopy size)
                         zlim=c(0:40),# alpha=0.2,                        # maximum value for z-axis (height)
                         type="h", lty.hplot=2,               # lines to the horizontal plane and its line type (2=dashes)
                         grid=TRUE, main=sprintf("A Hectare of Tree Growth \n (time = %s)",t), # include the grid and graph title
                         xlab="", ylab="", zlab="Height (m)") # axes titles
  # Sys.sleep(0.5)      # possibility to let the movie run in the R-studio plot (might be disfunctional when using the whole script)   }    # the loop for making graphs ends here!
# create morphing images with same size (morph 3 = 3 images per graph):
cmd_morph <- paste0("convert ", sprintf("%sMovie/%s/*.jpeg",Folder,nameshort),
                    " -morph 3 ", sprintf("%sMovie/%s/",Folder,nameshort), "%05d.morph.jpg")
# create the movie from the morph images (-r 10 = movie speed 10 fps; -qscale 2 = quality class 2):
cmd_mov <- paste0("ffmpeg -r 10 -i ",sprintf("%sMovie/%s/",Folder,nameshort), 
                  "%05d.morph.jpg -qscale 2 ", sprintf("%sMovie/%s/%s_%.0f%.0f.mp4",Folder,nameshort,nameshort,x,y))
# run the command lines created above in linux terminal

As one can see, I first make the graphs, that form the basis of the movie. I did not provide all details here, since it would make the script too long, due to the specifics of trees. Of course you can skip this all, when you have your own nice graphs. Just be aware you put them in a logical numbering order to read for the movie maker program later!
I also wanted to show here the s3d plotting function of R, which is really cool. It makes it possible to have 3D graphs. There are even options to look at the whole grid from different angles. (do not forget to download and call the package at the start of your R-script)

So, the real movie making starts at the very end of this script with the programs Image Magick and FFMPEG. In R you can create a command using the paste0-function, which is then called on the system with “system(command line)”.
We use two commands: the first I call “cmd_morph”, which creates the morphed pictures, so the movie is flowing; and the second one I call “cmd_mov”, which in fact makes the movie.

> cmd_morph <- convert inputfile/*.jpeg -morph x outputfile/%05d.morph.jpg
x = number of frames (morphed pictures) in between input graphs – the higher this number, the more flowinf the movie will be, but also the more time it needs to generate – standard is 2 to 5.
inputfile/ = file address of input graphs in jpeg or other picture type – take care this file only entails the graphs for the movie (or use inputfile/rootname*.jpeg).
outputfile/ = file address, where morphed output files will be stores.
%05d.morph.jpg = name of output file using 5 numbers (for max. 99999 frames) – change this 5 into 3 (max 999 frames) or any other number, depending on the frame number x.

> cmd_mov <- ffmpeg -r a -i inputfile/%05d.morph.jpg qscale b outputfile/name.mp4
a = the r-factor, meaning the number of frames per second – standard is 10.
b = the qscale-factor, meaning the quality class of the movie – standard is 2.
inputfile/05d.morph.jpg = file addres of the morphed input files, that resulted from the previous command cmd_morph.
Outputfile/name.mp4 = file address for the movie – consider a specific name for the location and time.

Hope this bit of code has entertained you a little bit.

Burning forests across the globe

The tropical forests seem to be burning at a extremely high level this year – worldwide!

In the Brazilian indigenous reserve of Arariboia is burning unstoppable since more than 4 weeks, the guardian reports here. It seems the fire was set on purpose due to the Agricultrual interests.

In Indonesia fires this year have released more CO2 than Germany’s industry. A very impressive report summarizes the research of the World Resources Institute linking the fire to palm oil plantations.

In this context: Who wants to make a beautiful map with Brazil’s public fires and protected areas? The satellite fire foci are downloadable here in shapefile format. Spatial boundaries of protected reserves are available for example here from the Ministry for Environment (MMA).

Please be invited to feel free to comment on these two reports!

From shame to fame? On Brazil’s experiment using public disclosure for conservation

Finally our paper “Naming and Shaming for Conservation: evidence from the Brazilian Amazon” has come out in PLOS ONE.

Brazil has probably been the first country in the world experimenting with the “naming and shaming” approach in the forestry sector and at a large (Amazon) scale. We started looking at the outcomes already back in 2013, with Sophie ’s preliminary results being presented at the International Workshop: Evaluating Forest Conservation Initiatives: New Tools and Policy Needs. Early in 2014, we discussed our findings with colleagues in Brazil who later published a working paper (Assunção and Rocha, 2014) using a similar empirical strategy.

Shortly after we had submitted our manuscript, Arima et al. (2014) came out with a paper using the blacklisting policy as a quasi-experimental setting to evaluate Brazil’s surge in field-based law enforcement activities in the Amazon.

How do our results compare to the other two studies? Assunção et al.’s counterfactual simulation suggests that blacklisting has conserved ca. 12,000 sqkm of forests, whereas Arima et al. provide a range from 2,300 to 12,000 sqkm. Like Arima et al., we pre-process our data set using matching techniques. Probably because we rely on a relatively large and diverse set of matching covariates, our results lie at the conservative (lower) end of the range provided by Arima et al..

However, inspired by Paul Ferraro’s and Merlin Hanauer’s (2014) exciting recent PNAS piece on causal mechanisms behind the welfare effects of protected areas, we wanted to learn more about how the blacklisting policy has worked.
We started with what we thought was most obvious: blacklisted districts should have seen more environmental enforcement action than others. But then, reviewers pushed us further to include public credit flows and the environmental cadaster (CAR) coverage as additional mechanisms. To our surprise, none of these mechanisms seems to have played a major role in explaining the additional conservation effect of the blacklist. For the time being, we conclude that “shaming” must have been an important driver behind the type of collective conservation action at the local level that we had heard a lot about from organizations on the ground.

Of course many caveats apply: For example, we’d need to know more about potential qualitative differences in field-based enforcement action – perhaps enforcement operations were as frequent, but more rigorous in blacklisted than in non-blacklisted districts… more research is needed.

More importantly, what can we learn from the Brazilian experience?

  1. The blacklist was published only after Brazil had started a major crackdown on illegal deforestation in the Amazon. Transferability of the Brazilian experience to other tropical forest margins thus probably hinges on law enforcement or other economic consequences being a real threat. Naming and shaming can then render this threat more imminent even if it is not actually accompanied by targeted action.
  2. If you look at our S1 Fig. in our paper, you’ll wonder how the official blacklisting criteria were used, if at all, to compose the list. The authorities were probably wise to keep the process relatively in-transparent. Deforestation monitoring in Brazil is very good, but not good enough to avoid lengthy debates with blacklisted districts about how clouds and classification errors could have messed up the latest blacklist. You also want to avoid that districts strategically deforest up to some “safe limit”.
  3. Whatever criteria are used for putting districts on and off the blacklist, deforestation should probably be weighed against other criteria, such as district size or economic performance.

As a fun example, consider organizing the blacklist like a football league:

The upper league would consist of districts that perform well, i.e. low levels of deforestation per unit of agricultural GDP (or total district area). The lower league (blacklist) would consist of the districts that consistently perform poorly in terms forest loss over agricultural GDP. There are currently 771 players, divided into two leagues, the upper league would have 721 members and the lower league would have 50 members. Each period (e.g., 2-3 years given high annual remote sensing error margins) the 5 blacklisted districts that perform best among their league shift to the upper league. Simultaneously the worst-performing 5 upper-league districts now move to the blacklist. The advantage over the current blacklisting system is threefold. 1) The two criteria at which districts are ranked within their leagues can be made public without making selection predictable. 2) The two combined criteria reflect the fact that larger / economically more powerful districts naturally tend to deforest more than smaller districts. 3) Such a system could create competition for first league membership among Amazon districts and perhaps lead to overall conservation effects beyond the current system. Don’t most of us prefer fame over shame?

Germany to invest 550 million € in rainforest protection and renewable energies in Brazil in the upcoming two years

The German government has been supporting Brazilian efforts to control deforestation in the Amazon for several decades within the so called PPG7 program. Currently German Federal Development Cooperation cooperates with Brazil in financial and technical areas to foster rainforest protection. German chancellor Angela Merkel, who is currently visiting Brazil, complimented Brazilian Policy Makers for their efforts during the last years and their ambitious goal to stop deforestation until 2020 . At the same time she committed around 550 million € for investment in rainforest protection and renewable energies in the upcoming two years as German newspaper “Die Zeit” reports on their website.

Pre-Columbian human history in the Amazonia remains disputed

This month two contesting papers have been published on the question how pre-Columbian society in the Amazon looked like. One is from Piperno et al. in The Holocene and the other one is from Clement et al. in Proceedings of the Royal Society. Both are putting forward an alternative to the now-abandoned concept of a ‘virgin’ Amazon. Previously, is was believe the Amazon rainforest was a pristine untouched wilderness before European settlers came over. The discovery of so-called ADE (Amazonian Dark Earth), prehistoric geometric ditches and the preposition of certain plant concentrations being a relic of prehistoric management led researchers to the conclusion that the region has actually been relatively densely populated. In 2013, Stéphan Rostain wrote an interesting book on the topic called Islands in the rainforest.


Ditches of old settlements found in Acre state, Brazil (from Rostain, 2012).

Today, researchers are divided.

Clement et al. explain all archaeological sites as a connected system. Densely populated regions were found along the rivers, which also spread along smaller tributaries and locally in between rivers. They describe the region as a social-ecological mosaic landscape with production systems. The lack of more archaeological evidence in the dense forests is due to its large area and its remoteness, rather than the preposition nobody had lived there. Instead, they state the current findings actually proof these regions might as well be full of undiscovered man-made features.

On the other hand, Piperno et al. is trying to urge for caution in proposing wild hypotheses and stresses that the present archaeological findings are still open for interpretation and rather local, not at all wide-spread. They acknowledge a much higher population number before the arrival of Europeans and also propose large communities along the main rivers, but not much more. The press seemed to pick up more on the paper of Clement et al.

Archeological sites

Locations of pre-Columbian settlements (from Clement et al., 2015)

p.s. That same week, the press picked up on related news from the Amazon in genetics. The tribes living in the Amazon for millennia seem to be closely related to Australasian people (Australian and Papuan indigenous people). The discovery of Skoglund et al. is, just as the research of Piperno et al. and Clement et al., breaking with the well-established theories. Until now the roots of native Americans have been believed to come from Eurasians crossing the Bering street during the end of the Ice Age. These wanderers then moved south. The new research shows it might have been a little bit more complex back in the days. Who were / are the Amazonian tribes?

Reasoning Conservation: Prudence, Justice and the Good Life

This week I had the pleasure of attending a lecture by Uta Eser on the PopeWHY of nature conservation. It seems a stupid question in our field of research and is actually seldom questioned. Moreover, even the pope is fighting climate change nowadays! The story on climate change and extinction of species is well-known. The facts are known. However, there is often an implicit task in this story: an urging plea to do something about it. For action nevertheless, one needs to rationalize and justify why the facts are worth doing something about.

Uta Eser discussed that the rationality of protecting nature, biodiversity and traditional landscapes can be divided in three pillars: Prudence (Intelligence), Justice and the Good Life (Happiness).

Prudence has been mostly used and is referring to the existential need for nature; it serves us in the basic needs to survive. In communication, one often uses the picture of the guy sawing the branch on which it is sitting: you must be out of your mind to saw the very branch which is supporting you! The analogy holds for the sustenance and insurance of the human inhabitants of planet earth. One could think of the scientific approach on ecosystem services and its (monetary) value. However, we get into a pitfall: the collective WE does not always have the same values and wishes as the individual WE. The WE is becoming a generalization and its rhetoric is concealing important issues.Branch

Justice is the second pillar and one of these issues. It might help to redeem the pitfall of prudence. Justice is looking at who is sawing the branch and who will fall out of the tree when the branch is cut. It goes without saying that industrialized western countries are very active in handling the saw, mostly indirect. Our needs (USA is needing 4 planets, while India is needing 0.4 planet!) are being supplied by developing countries, where resources and cheap labour are available. Their branch is being sawed through (think of oil palm plantation, open pit mining and oil exploitation in the Amazon to name a few). In a broader spectrum one could also call upon human being within the chain of living species (biodiversity) or the future generation, who will not have much left over from planet earth. What is justifying us to exploit its resources in such a way? This justification is also used by the pope. However, justification gives us an obligation and people start to rethink the thesis and actually question whether we really need all this nature to survive as a species.

At this point the third pillar comes in: Happiness for All. This appeals to our subjective reasoning of feeling good and comfortable. As the Club of Rome stated in 1972, the crux is not whether human species can survive, but whether its existence will be worthwhile. By protecting nature we provide a better life, improving our relation to ourselves, human kind and earth, involving values, ethics and happiness.

More reading?
Eser, Nuereuther, Seyfang and Muller (2013): Prudence, Justice, and the Good Life. A typology of ethical reasoning in selected European biodiversity strategies

Deforestation and environmental governance in Paraguay

Paraguay is one of the countries with the most accelerated deforestation rates worldwide. Especially in the dry northern part called Chaco vast amounts of areas where cleared in recent years. The impressive deforestation patterns as observed by satellite imagery show that probably large agribusiness engages in a planned clearance of the Chaco. You can have a look at these patterns in the Hansen data-set here:

Screenshot from 2015-04-29 21:43:46

With weak institutions and most of the land hold in private hands, it is doubtful that Paraguay will be able to curb this development in the near future. Furthermore corruption puts a burden on preserving natural vegetation even inside indigenous reserves. This recent article in Spanish informs about the arrest of Ruben Quesnel former employee at the Paraguayan institute of indigenous people – Instituto Paraguayo del Indígena (INDI). Quesnel is responsible for the sellout of 25.000 hectares of indigenous land and was sentenced to six years in prison:

As environmental governance improves in Brazil, there are rumours that large actors from Brazilian Agribusiness change their strategy to occupy land in Paraguay that is cheaper and less controlled by environmental agencies. Market integration within the Mercosur furthermore facilitates foreign investment. The expansion of soy-bean production in Paraguay contributed to the increase in economic growth within recent years.

The spatial reorganization of land-clearance is also discussed under the terminology of leakage effect and is a very important topic when discussing land-use change on a global scale. Future research on this topic is required to make more reliable statements about inter-regional and cross-border effects of environmental policies to control deforestation.

Fragmentation patterns and ecosystem services

Landscape fragmentation and human-transformed forests have been subject of discussion in the comparison between land sharing and sparing, as suggested by Phalan et al. 2011. They elaborate on the ecological benefits of intensified agriculture in contrast to organic agriculture, due to its protection potential of a larger size of core forests. This view is supported by research of Gibson et al. 2011, who found that the ecological value of primary forest is far superior to any other human-transformed land cover, such as secondary forests.

However, assisting the restoration of degraded landscapes might proof to be effective, as mentioned by Jakovac et al. 2015. No data has thus far been analysed thoroughly. Another type of human-transformed forests are the selectively logged forests, and these seem to provide a rather high potential for many ecosystem services, as recently picked up by Bicknell et al. 2015. He suggests to close off logging roads as being the most effective and quick way to sustainably manage tropical forests and acquire the highest conservation value.

Contradictory to these statements, Mitchell et al. 2015 recently reframed ecosystem services in fragmented landscapes by including its accessibility to people, pointing to the possible positive effect of fragmentation. The authors urge for a better understanding of the importance of the accessibility of ecosystem services to underpin the correct decision-making activities. One of the tasks within our team will be to use the latest models in testing the above mentioned hypotheses.

The Amazon Carbon Sink remains a mystery – feedbacks and landscape transformation

Recently, Brienen et al. 2015 published results from different long-term projects on Amazon forest dynamics, showing a generalized decline of the biomass carbon sink. While more than a decade ago, Betts et al. 2004 modelled the so-called die-back of the Amazonia, recent models have been showing an increase of the biomass carbon sink well into the 21st century due to CO2 fertilisation (for example Huntingford 2013). Brienen et al 2015 now argue a decline in the capacity of the carbon sink, caused by an accelerated tree life cycle, increasing the biomass mortality. The carbon uptake is simply getting oversaturated due to the high atmospheric CO2 concentrations. Will the new-generation models show a return of the die-back theory?  As yet, no effect of CO2 fertilisation or accelerated tree life cycles have been found in tree ring studies, such as done by Van der Sleen 2015.

Simultaneously, Haddad et al. 2015 published another study from long-term projects on world-wide forest dynamics. They show fragmentation has long-lasting impacts on biodiversity, nutrient cycling and ecosystem services. They furthermore state that currently more than 70% of the remaining forests are within a single kilometre from its edge and thus severely fragmented! The tropical forest around the Amazon is one of the two regions with a relatively contiguous area. Nevertheless, the fragmentation of the Amazon still results in a considerable carbon loss of approximately 10% of the total, according to Pütz et al. 2014.

The decline of the biomass carbon sink and the effects of encroaching human-transformed landscapes, including fragmentation, show dire consequences of declining ecosystem services potential and conservation value. The two processes might have a coupled effect, decreasing the total carbon storage of the tropical forests. Can this have consequences for the Payment for Ecosystem Services?

Defining PES: a long expected revisit

The most influential article written on PES has got a long awaited follow up by its author. Sven Wunder’s 2005 “Payments for environmental services: some nuts and bolts” has proposed a definition of PES that has since then been embraced and contested throughout the PES academic community. In the last 10 years, a number of works have tried to seek their own definition of PES, several times taking Wunder’s definition as a starting point for the discussion.

In the new article, before proposing a new definition, Wunder provides a very interesting discussion on the usefulness of definitions and on what one should want from a definition of PES. Especially interesting are his points about how a definition should be a tool for explaining phenomena, instead of a description of the issue in hand, and why a PES definition should single out PES in relation to other types of policies.

The new definition replaces “buyers” for “service users” dissipating an “an inadequate market association”, which has brought some conceptual confusion and practical distrust of PES in many circles. It also replaces “ES provision (or land use proxy)” for “Agreed rules of natural resource management” and scrapes the idea of a “well defined ES” due to technical and cost challenges of ‘well-defining’ an ES. Finally, it adds the idea of “offsite services”, which “deliberately links PES to the […] subset of environmental externalities, which the resource-use rules aim to address”.

The very worth reading can be found at

Severe droughts and deforestation

Science has already made the link between deforestation of the amazon basin and decreasing precipitation in the southern parts of Latin America. See here or here where Antonio Nobre presents a summary of his literature review on that topic. Now this connection is harshly being impeded into everyone’s life in São Paulo, where water scarcity has lead authorities to curb water to 2 days a week, see here (germ.) But deforestation, droughts and climate change aren’t the only reasons. Mismanagement of watersheds, corruption and the absence of a market based price signals for households are most probably more significant contributors to this summer’s water scarcity in southern Brazil.

Environment, Climate, Humans, Change….

January was cold and grey in Germany. Despite it gave me a bit of fire to dare and challenge some thinking on global changes. Hopefully inspirational… Trying to maybe put the world as we know it into perspective.

Our ancestors in the Ice Age
A few weekends ago, I visited the LVR (Landschaftsverband Rheinland) museum in Bonn, which is currently showing an exhibition on humans living during the final countdown of the last Glacial Maximum, some 15,000 years ago. It gave an amazing picture on the totally different world in Western Europe, compared to what we are used to in the 21st century. Our current forested and agricultural land was a vast savannah, with gigantic animals roaming it.  The human population was less than 10,000 souls and thus had enough food. We wandered around over the floor of the North Sea, where enormous paleo-rivers bursted bewildered over the landscape. However, climate change forced the sea level to rise dramatically, increasing temperatures and rainfall (the 6 degrees lower temperatures fixed much of the available moisture in ice caps). The climate change process made the Europe suitable for forests and thriving plant communities. People gradually became the architects of their surrounding landscape.

Human architects…
A few days later I went to a lecture on the effects of climate change, organised by the Frankfurter Geographische Gesellschaft. My supervisor, Prof. Hickler spoke on the effects related to biodiversity and ecosystems. The IPCC report on the devastating amounts of carbon and other pollutants pushed into the atmosphere was one part. Another part was on the physical effect of human being on the landscape. It reminded me of a recent study, stating that Europe nowadays has more forest than a century ago. In Europe, massive deforestation occurred a few centuries ago, when instead of oil, wood was used as fuel. The landscape was being exhausted and clearcut, something which is happening currently in some tropical countries. The whole ecosystem changed dramatically and large mammals disappeared.

Tropical forests in the climate change century
The two stories came together while reading a somewhat outdated textbook, titled ‘Tropical Forest in Transition’, edited by Johann Goldammer in 1992. This book makes clear that the stable status of tropical rainforest is a fairytale. We want to believe the large Amazon forest has always existed the way we know it nowadays. This is simply not the truth. Climatic changes were not restricted to the Northern latitudes. Incursions of drier savannah landscapes in the transitional forest of the Amazon happened regular and rainforests got a patchy occurrence. Refugia of tropical species surely existed during the ice ages and from here the rainforest developed to its large extent. These refugia could have existed in secluded river catchment, compared to the subtropical forests we find today. Rainforests always existed, but altered. They had to survive paleo-Indians setting fire and huge paleo-El Nino events roaring around the region. The rainforest resided.

The visit to the LVR-Museum, the lecture of the FGG and the insights from Goldammer let me to the following:

  • In the past, climate change has drastically altered the landscape in Northern latitudes, changing a savannah with mega-fauna into a forest without mega-fauna.
  • In the past, humans have drastically altered the landscape in Northern latitudes, deforesting an enormous area, resulting in a destruction of ecosystems.
  • Although tropical forests seem to be relatively stable, it is also subject to frequent environmental changes, showing stages of a more open and patchy landscapes.
  • Currently, we push such large amounts of carbon into the atmosphere that the temperature rises more than the ice age fluctuation.
  • Currently, we are deforesting large extents of forest, similar (and more) to the deforestation in the Northern latitude.
  • Therefore, we can expect alterations in the tropical regions. We probably see ecosystems, which are common today, pushed into refugia, while others take over. This might force us to change our living standards and think once again carefully about system earth we live on.


The geological temperature fluctuations according to the wikipedia site for ‘Geologic Temperature Trend’ (Glen Fergus)

New study, co-authored by NWF, shows Amazon Soy Moratorium saves more rainforest

Brazil’s Soy Moratorium

From: Gibbs, Rausch, Munger et al. 2015

Brazil’s Soy Moratorium (SoyM) was the first voluntary zero-deforestation agreement implemented in the tropics and set the stage for supply-chain governance of other commodities, such as beef and palm oil. In response to pressure from retailers and nongovernmental organizations (NGOs), major soybean traders signed the SoyM, agreeing not to purchase soy grown on lands deforested after July 2006 in the Brazilian Amazon. The soy industry recently extended the SoyM to May 2016, by which time they assert that Brazil’s environmental governance, such as the increased enforcement and national implementation of the Rural Environmental Registry of private properties (Portuguese acronym CAR) mandated by the Forest Code (FC), will be robust enough to justify ending the Agreement. We argue that a longer-term commitment is needed to help maintain deforestation-free soy supply chains, as full compliance and enforcement of these regulations is likely years away. Ending the SoyM prematurely would risk a return to deforestation for soy expansion at a time when companies are committing to zero-deforestation supply chains.

Link to the article:

Link to the blog:


On the Geology of the Amazon River

Being a geologist from principle, I could not resist myself to have a look into literature on the paleo-Amazon. This information will help in the understanding of the Amazon vegetation structure nowadays. Perhaps even to understand future trends, although these future trends that we tend to predict have a very different time scales.

Initially the Amazon did not flow from west to east
The first indications of the Amazon River can be found back some 40 million years ago, during the Paleogene. The recently formed Andes and the existing high-topography shields of Guyana and Central Brazil formed a huge interior sea, in which the headwaters of the Amazon formed. This was the so-called Pebas basin (lake) and coincided with the current dense rainforest of the western Amazon region. In this basin, a network of smaller rivers was formed, which we might better categorize as being tributaries of the paleo-Orinoco. In those times, the Amazon did not flow to the east, but instead followed a route north draining into the present-day Gulf of Mexico. For a long time, the Pebas basin was a huge drainage pot for its surrounding mountains (Andes and Guyana Shield), and functioned like a kind of extension of the Gulf of Mexico. From the Gulf, sea level fluctuation, resulting climate change, induced the occasional flooding (incursions) and draining of the basin, forming the lacustrine (lake) sediments nowadays found in the western Amazon. The real Amazon was only a minor river draining on the west side of the Guyana shield.

A breakthrough!
During the Miocene things changed. We are jumping to approximately 20 million years ago. World-wide temperatures were decreasing from a very hot climate to a more temperate climate (similar to present-day). Meanwhile, the uplifting of the Andes continued rapidly and mountain ranges formed in the northern part of Latin America (Venezuela and Colombia). The Pebas basin was slowly cut off from its drainage point in the Gulf and the Amazon experienced more resistance to flow in that direction. Searching another route, the Amazon River cut itself loose from the Orinoco River and moved eastward. The Pebas basin developed into a large inland wetland, vegetated mainly with palm swamps and lowland riverine forest. Only at the end of the Miocene, some 10 million years ago, the river managed to breach through the so-called Purus Arch and connected to the much smaller eastern Amazon River. This break-through must have forced a violent burst and ecological disasters. In the fan delta of the current Amazon River,we find a sharp change in its sedimentary succession from this time onwards with significantly higher sedimentation rates. During the first couple of million years, the Amazon River is still not well developed and the Pebas wetland coexisted along with the river draining it.

Fluctuations during ice ages
In the Pliocene (around 5 million years ago), ice ages started to appear. The era of ice ages meant a rapid fluctuation in temperature, rainfall and sea level. The waxing and waning of polar ice caps did not just happen once, but many times. During this phase the Amazon River developed to its present form, incising deeper and deeper into the gorge, which it had formed by connected east and west breaking through the Purus Arch. Swamps and lakes still persisted, and flooding events due to reformations of the river course were common. Only since the last ice age the current Amazon River channel along Manaus formed: 5,000 to 2,500 years ago major flooding events occurred establishing this flow channel. This was while the Romans conquered the old world.

 Thus, although the Amazon river nowadays seems to be a very stable river, it did not always exist in the way we know it and it shows to be a living and moving system.

Some papers on the Miocene river reversal: Hoorn, C.; Guerrero, J.; Sarmiento, G.A.; Lorente, M.A. (1995); Andean tectonics as a cause for changing drainage patterns in Miocene northern South America; Geology 23 (3); p. 237-240  &  Figueiredo, J.; Hoorn, C.; Ven, P. van der; Soares, E. (2009); Late Miocene onset of the Amazon River and the Amazon deep-sea fan: Evidence from the Foz do Amazonas Basin; Geology 37; p. 619-622
A paper on the Holocene Amazon: Fatima Rossetti, D. de; Toledo, P.M. De; Goes, A.M. (2005); New Geological framework for Western Amazonia (Brazil) and implications for biogeography and evolution; Quaternary Research 63; p. 78-89

A tree strategy to survive droughts: HD

After some time of literature and even more literature study on climate change, droughts, upcoming climatic disasters and the total dy-back of the Amazon rainforest, I was most amazed by the technology which trees develop in surviving these horrible futuristic scenarios.
Let me explain. The Amazon rainforest is a system, which keeps itself alive in different ways. One of them is the water recycling by ways of high evaporation rates, bringing moisture back in the atmosphere and pumping water-loaded air deeper into the forest.
The other is a mechanism called hydraulic distribution, which I want to address here. Despite the above mentioned water recycling, large parts of the Amazon rainforest experience an annual dry period of around three months. Nevertheless, the rainforest does not lose its leaves or reduces its biomass production (such as the trees in German winter!). Instead, the tree itself has the ability to govern its own water resources.

What happens exactly? The trees have two sets of root, namely the taproots and lateral roots. The taproot goes deep into the soil to reach deeper water resources in aquifers, while lateral roots spread directly from the tree to get water from the topsoil. Normally both the taproot and lateral roots will suck water from the soil into the trunk, which transports it to the canopy (middle picture b).
However, when rains fall down, this system changes overnight. Instead of tapping water from deep aquifers, the tree will only tap water from the topsoil and actually transports the water via the taproot down into the aquifer (right picture c). The tree is saving the rich harvest of fresh rain in the deeper soil layers! Now the tree will have some savings for the dry period.
However, trees seem to be very social and practical. During the dry period, it applies another process, which also happens overnight. While the taproot keeps on sucking soil water from deeper aquifers, the lateral roots use some of the water and push it into the topsoil, wetting it for its own use during the day and meanwhile providing water for fellow plants around (left picture a).

Hydraulic redistribution

The system is actually typical for trees in the Savanna and other dry regions, but seems to work well in the humid tropics as well. However, we do not really understand when the tree decides to change its fluid transport from storing to distributing it. Therefore, in physically based vegetation models it is often not modeled, but more or less lumped within some measure of resilience to drought. Does it remain too mystical or complicated to include the amazing ability of trees to redistribute soil water?

Want to read more? Oliveira and associates published a very interesting paper in Ecophysiology (March 2005), titled “Hydraulic redistribution of three Amazonian trees”. A bit later, Lee and associates published another paper in PNAS (October 2005), titled “Root functioning modifies seasonal climate”, that also gave an in-depth description, including the picture used in this post.

Import multiple files to R

I have been recently asked a few times how you can import a bunch of data (let’s say for example .csv files) to your R-Environment without copying and pasting a lot of code. I’m not aware of a built-in-package in R that does that for you (although I can imagine that somewhere there might be one) but I will show a little example on how you can do this manually. The advantage is that you can easily modify the code to import other file-types and if you are a beginner with R you might get some feeling for automating, loops and lists in R.

The main idea behind the following code is, that you put all your files into one directory and read them into R with a loop. Therefore you will have to work with lists which serve as a “container” to receive the incoming data. Note that your .csv files need to have the same characteristics in order to automate the process. If you have for example  csv. files with different separators (one with commas and another with semicolons) the import will not work as expected. If you have never imported data into R before, try the read functions on single files before you go to automating.

The code is quite self-explanatory:

## import_multiple_csv_files_to_R
# Purpose: Import multiple csv files to the Global Environment in R

# set working directory

# list all csv files from the current directory
list.files(pattern=".csv$") # use the pattern argument to define a common pattern  for import files with regex. Here: .csv

# create a list from these files

# create an empty list that will serve as a container to receive the incoming files<-list()

# create a loop to read in your data
for (i in 1:length(list.filenames))

# add the names of your data to the list

# now you can index one of your tables like this$deforestation2010.csv

# or this[1]

# you can make a function out of this
tmp.list.1<-list.files(mypath, pattern=mypattern)
for (i in 1:length(tmp.list.1)){tmp.list.2[[i]]<-read.csv(tmp.list.1[i],...)}

# use it like this
# note: with ... we enable the function to refine the import with parameters from read.csv.
# here we define the separator of entries in the csv files to be comma.

# save it to the folder with your custom functions

# load it like this whenever you need it in another script with

# end

Vídeo sobre áreas protegidas

Prezado Leitor,

Estamos contentes em iniciar nosso novo Blog sobre políticas ambientais nos trópicos com um interessante vídeo mostrando a evolução espacial das áreas protegidas no Brasil de 1920 a 2012. Nós criamos esse vídeo para visualizar os esforços realizados pelo governo brasileiro, ONGs e institutos de pesquisa de todo o mundo para reduzir o desmatamento em um a das fronteiras agrícolas mais dinâmicas do mundo. Estudos mostram que o estabelecimento de áreas protegidas foi chave para controlar o desmatamento na Amazônia brasileira, onde as taxas anuais de corte raso caíram desde 2005. Além disso, a demarcação de terras indígenas ajudou a preservar o habitat de comunidades tradicionais e evitar conflitos sobre o uso da terra.

Apesar de este vídeo mostrar avanços impressionantes com relação ao tamanho total das áreas protegidas, grandes desafios ainda existem, para que se atinja a sustentabilidade de longo prazo dos esforços realizados. Estes desafios incluem a gestão das áreas estabelecidas e a recuperação de áreas degradadas, além da promoção de alternativas sustentáveis de geração de renda que equilibrem necessidades ecológicas e o desenvolvimento socioeconômico na região.

Nosso Blog informa sobre desenvolvimentos no tema de proteção de florestas tropicais em todo o mundo, mas com uma ênfase especial na região Amazônica. Nós somos uma equipe de pesquisadores baseada no Centro para Pesquisas de Desenvolvimento com experiência de campo no Brasil, Peru, Bolívia e Equador. Caso você se interesse pelo nosso trabalho em políticas ambientais, você pode visitar a página do nosso projeto  ou acompanhar os próximos posts no Blogazonia.

Grupo de pesquisa PA

Você pode fazer o download do video aqui.

Automated Mapping in R

After creating our base-layers within a PostGIS database as described here we used R to map our layers and create a plot showing the area count for each year from 1920 to 2012. In total we created 1840 plots (92 years x 20 buffers) that where then combined to produce this video.

The source code for the mappings is complex and wont be discussed in detail here but a brief description of the single steps as well as the code is provided for anybody interested in R-Programming:

A. Session Setup
Here we set our working directory, load all required libraries and suppress scientific notations (later-on important for the plots)

B. Load the data
We load all relevant data.

  • ucs_cum is a table that contains the area counts for all years and all protected area classes. We will need it for the area count on the left side of the video.
  • brasil and world.countries are two shapefiles that we need for our plots. World.countries contains the administration boarders for  all countries worldwide and brasil the boarders for Brazil. We fortify these shapefiles to make them understandable and plottable for ggplot2.
  • map.picture3 is the picture on the upper left of the video. We rasterize it to make it plottable.
  • Next we load all shapefiles that we need for the maps. All in all we have 3 different protected areas with each 20 shapefiles. To avoid too much code we have to automatize this follwing a simple three step approach: First we create a list with all filenames in a loop. Second we create an empty list where we will store our imports. Third we import the shapefiles with a loop thereby using the list of file-names. shapefiles.list.import contains all layers for all multiple use areas. shapefile.list.upi contains all layers for strictly protected areas and shapefile.list.ind contains all layers for indigenous territories. Afterwards we rearrange the order of the list and plot the minimum years because we will need this information later.

C: Construct parameters for the lineplot

Here we calculate the extension of all protected areas (total_protected_area_sqkm) and the extension of indigenous areas + strictly protected areas (total_uus_ind_sqkm). We need this information later  for the areacount plot.

D: Define viewports for the plots

We define different Viewports for the plots. Inside these Viewports we will later paste the different parts of our video (the map, the area count and the picture)

E: Define color schemes for the plots

We define custom colors for the different protected areas.

F: Define the different themes for the plots

We define how our plots made with ggplot2 should be lay-outed. Actually this piece of code was only created after we created the plots which come later. We externalize this part of the code because the plot function that follows afterwards is very long. If you do testplots and want to adjust small parameters in the layout, it might be an advantage to have the layout outside of the plot function.

G: Define additional parameters for the plots

Here we define the plot-title.

H: Construct the plotfunction

Now it gets tricky. In this function we do the following things.

  1. First we create additional layers that contain the information about protected areas ( These layers will display all protected areas that have been created until the respective year. We have to do it with an if-else statement since we don’t know if there is any data for the given year.
  2. Second we create additional layers that contain information only about the respective year. In this layer we plot our buffers that indicate the newly created areas (tmp.data1-3). Again we have to do it with an if-else statement since we don’t know if there is any data for the given year.
  3. We construct the mapplot by adding the baselayers (countries worldwide, Brazil) and additional layers from 1 and 2. We furthermore add limits, title, projection and theme configurations to the plot.
  4. We construct the lineplot (areacount) with geom_ribbon from ggplot2.
  5. We define the area count as a title of the lineplot and modify the plot.
  6. We load the picture
  7. We plot our plots. Therefore we first open a new png file. Inside this file we create a Grid where we push the Viewports. In each Viewport we place our plots.
  8. We turn the device off and finish our plotting function.

I: Plot the data

We plot the data iteratively over all years and all buffersizes. We therefore define the output path and create two vectors for the iteration. One containing the years to be plotted and another one containing the alpha levels for the different buffer sizes (if you watch the video carefully you might notice that the alpha level changes with buffersize).

We then plot everything which takes about 4 hours to complete.

J: Close the session

We close the session and save all created objects from the Global environment.


# protected_areas_plotting_1
# Author: Johannes Schielein
# Outline:
  # A. Session Setup
  # B. Load the data
  # C: Construct parameters for the lineplot
  # D: Define viewports for the plots
  # E: Define color schemes for the plots
  # F: Define the different themes for the plots
  # G: Define additional parameters for the plots
  # H: Construct the plotfunctions
  # I: Plot the data
  # J: Close the session

# ----------- A. Session Setup -----------
# set wd

# load workspace (if already saved once)

# activate required libraries
#font_import() # only if not already imported

# surpress cientific notation

# ----------- B.Load the data -----------
### Load the base data
# load cumulative uc data

# shp brasil
brasil<-readOGR("World Countries 2/","subset_brazil")

# shp worldcountries
world.countries<-readOGR("World Countries 2/","sa")

### load the PNG foto for the plots
map.picture3<-readPNG("pics/fotos3.png") #load fotos
map.picture3.rastered <-rasterGrob(map.picture3)

### Load all uus shapefiles
## create an interation vector with the names endings
# the endings of all shapfile names
shapefile.iteration.vector<-vector(length=20) #create the vector
shapefile.iteration.vector[1]<-0 # set the first place zero
for (i in 2:20)shapefile.iteration.vector[i]<-shapefile.iteration.vector[i-1]+5 #set the rest places according to the names endings
shapefile.iteration.vector #show the vector
## creat a vector with the complete names
shapefile.list<-vector(length=20) # create the vector
# create a list with the shapefile names
for (i in shapefile.iteration.vector)shapefile.list[i/5]<-paste("_uus_buffer",i,sep="") # paste beginning and end of names
shapefile.list[20]<-("_uus_buffer0") # _uus_buffer0 was lost so set it automatically
## import all shapefiles to the list
for (i in 1:20)shapefiles.list.import[[i]]<-readOGR("buffer_shps/",shapefile.list[i])

### Load all upi shapefiles
## create an interation vector with the names endings
## creat a vector with the complete names
shapefile.list.upi<-vector(length=20) # create the vector

# create a list with the shapefile names
for (i in shapefile.iteration.vector)shapefile.list.upi[i/5]<-paste("_upi_buffer",i,sep="") # paste beginning and end of names
shapefile.list.upi[20]<-("_upi_buffer0") # _uus_buffer0 was lost so set it automatically
## import all shapefiles to the list
for (j in 1:20)shapefiles.list.import.upi[[j]]<-readOGR("buffer_shps/",shapefile.list.upi[j])

### Load all indigenous shapefiles
## create an interation vector with the names endings
## creat a vector with the complete names
shapefile.list.ind<-vector(length=20) # create the vector

# create a list with the shapefile names
for (i in shapefile.iteration.vector)shapefile.list.ind[i/5]<-paste("_ind_buffer",i,sep="") # paste beginning and end of names
shapefile.list.ind[20]<-("_ind_buffer0") # _uus_buffer0 was lost so set it automatically

## import all shapefiles to the list
for (k in 1:20)shapefiles.list.import.ind[[k]]<-readOGR("buffer_shps/",shapefile.list.ind[k])

### rearrange the shapefiles lists to have a better naming

# remove old shapefilelists

# find out the first year of the different type of protected areas
min(shapefiles.list.import.2[[20]]$DATE_CREAT,na.rm=T) # uus = 1927
min(shapefiles.list.import.ind.2[[20]]$DATE_CREAT,na.rm=T) # ind = 1926
min(shapefiles.list.import.upi.2[[20]]$DATE_CREAT,na.rm=T) # upi = 1896

### ----------C: construct parameters for the lineplot----------

# calculate sqkm for ucs cumulatice
for (i in 2:5)


# calculate the total protected area

# calculate uus+ind

### --------- D: define viewports for the plots ----------
# define first plot area
vp1 <- viewport(x = 1.05, y = 0.133,
                height = 0.72, width = 0.72,
                just = c("right", "bottom"),
                name = "map"
# define second plot area
vp2 <- viewport(x = 0.05, y = 0.505,
                height = 0.3, width = 0.361,
                just = c("left", "bottom"),
# define third plot area
vp3 <- viewport(x = 0.05, y = 0.12,
                height = 0.36, width = 0.361,
                just = c("left", "bottom"),

# define fourth plot area
vp4 <- viewport(x = 0.5, y = 0.95,
                height = 0.1, width = 0.8,
                just = c("center", "center"),

### ---------- E: define color schemes for the plots ----------

### --------- F: define the different themes for the plots ----------
## for additional information on how to manage them themes see the ggplot2 help

# plot 1 = maps
theme.plot.1<-list(theme(panel.background = element_rect(fill = "lightskyblue2", color="black"),
                          panel.grid.minor = element_blank(),
                          panel.grid.major = element_blank(),
                          axis.text.x = element_text(size=25,colour="white"),
                          axis.text.y = element_text(size=25,colour="white"),
                          axis.title.x= element_text(size=25,colour="white"),
                          axis.title.y= element_text(size=25,colour="white"),
                          plot.title = element_text(size=75,vjust=3,family="DejaVu Sans Condensed",colour="white"),
                          plot.background = element_rect(fill="black",colour="black")
# plot 2 = lineplot
theme.plot.2<-list(theme(panel.grid.minor = element_blank(),
                         panel.grid.major = element_blank(),
                         plot.title = element_text(size=60,vjust=1,family="DejaVu Sans Condensed",colour="white"),
                         panel.background=element_rect(fill = "black"),
                         legend.justification = c(0, 1),
                         legend.text=element_text(size=25,family="DejaVu Sans Condensed",colour="white"),
                         legend.background = element_rect(fill="black",colour="black"),
                         axis.ticks.length = unit(0.3, "cm"),
                         axis.line.y = element_blank(),
                         plot.background = element_rect(fill="black",colour="black")

### --------- G: define additional parameters for the plots ----------
# the title
plot.title<-textGrob("Protected Areas in Brazil",
                     gp = gpar(fontfamily="DejaVu Sans Condensed",
                               fontsize = 90,

#### ---------- H: Construct the plotfunction ----------

  ### 1. fortify all shapefiles for the plot

  ## 1.1 fortify all existing protected areas until the year (unbufferd)
  # layer 1 unbuffered
  else{<-"no area for first unbuffered layer"}
  # layer 2 unbuffered
  else{<-"no area for second unbuffered layer"}
  # layer 3 unbuffered
  else{<-"no area for third unbuffered layer"}

  ## 1.2 fortify protected areas for the respective year if there was a creation/data is available (with buffer)
  # layer 1 buffered
                                       ([[my.buffer.size]]@data$DATE_CREAT>=my.year &
                                                  ([[my.buffer.size]]@data$DATE_CREAT>=my.year &
                                                     {tmp.data1<-"no area for first layer"}
  # layer 2 buffered
                                       ([[my.buffer.size]]@data$DATE_CREAT>=my.year &
                                                  ([[my.buffer.size]]@data$DATE_CREAT>=my.year &
                                                     {tmp.data2<-"no area for second layer"}

  # layer 3 buffered
                                       ([[my.buffer.size]]@data$DATE_CREAT>=my.year &
                                                  ([[my.buffer.size]]@data$DATE_CREAT>=my.year &
                                                     {tmp.data3<-"no area for third layer"}

  ### 2. construct the mapplot
  ## 2.1 add the base layers

  ## 2.2 add all existing protected areas until the year unbufferd
  if (
  # else{print(} # activate this option if you want to trace the layers within each map
  if (
  # else{print(}
  if (

  ## 2.3 add protected areas with buffer if data is available for the respective year
  # layer ind
  if ({tmp.plot<-
                                                          map=tmp.data1, data=tmp.data1,
  # layer uus
  if ({tmp.plot<-
                                                          map=tmp.data2, data=tmp.data2,
  # layer upi
  if ({tmp.plot<-
                                                          map=tmp.data3, data=tmp.data3,

  ## 2.4 add limits, title, projection and theme configurations to the plot

  ### 3. construct the lineplot
  ## 3.1 plot
    geom_ribbon(aes(x=year,# is going to be area for upi
    geom_ribbon(aes(x=year,# is going to be area for uus
    geom_ribbon(aes(x=year,#is going to be area for ind

    ## 3.2. area count as title
            digits=0), # together with round cuts the digits off
          big.mark=","), # together with format creates the thousend seperators)
    ## 3.3 modify the plot
    # scales and colours
                      labels=c("  Strictly Protected Areas","  Multiple Use Reserves","  Indigenous Areas"))+
                        guide="none")+ # to surpress the legend of the outlines
    scale_x_continuous(breaks=seq(1920, 2010, by = 10),
                       limits=c(1920,2013), # set value limits of axis
                       expand=c(0, 0))+ # set visible limits of axis
    # add the themes

  ### 4. Construct the plot with extra information

  ###  Combine the plots and make png outputs
  ## new png
  png(filename=paste(my.path,"map_",my.year,"_buffer",my.buffer.size,".png",sep=""), width=2560, height=1600)
  ## mapplot

  # mapplot
  grid.newpage()# clear plot area
  grid.rect(gp = gpar(fill = "black")) #show the plotting region (viewport extent)
  pushViewport(vp1) # enter vp1
  print(tmp.plot, newpage = FALSE) # plot1
  upViewport(1) # leave vp1

  # lineplot
  print(tmp.plot2, newpage = FALSE)

  # mapinfo
  print(grid.draw(my.picture), newpage=FALSE)

  # title
  print(grid.draw(plot.title), newpage=FALSE)

  ## Turn the device off to finish the print

  ### Finishing the function
  print(paste("Map completed for",my.year/10000,"with buffersize",my.buffer.size,"and opacity set to",my.alpha))

#### --------- I: Plot the data ----------

# define the path for the plot outputs

# create a string to iterate over the input data ucs year by year
for (i in 2:length(1896:2013))iteration.string.plot[i]<-iteration.string.plot[i-1]+10000

# create a vector that contains the transparency/alpha values for each buffer<-seq(from=0.5,to=1,length.out=20)

# plot all years and buffers (over 2gb size!)
for (j in 2:length(iteration.string.plot)){
  for (i in 19)

# ---------- J: Close session----------

PostGIS easy buffers

To create the video Evolution of protected areas in Brazil we created buffers around all protected areas with 20 different buffer sizes. For each of the 3 protected areas types (indigenous terretories, multiple use reserves, stricly protected areas) we used a different layer containing the polygons for each protected area. A display of the 20 buffers in each year within one second creates the impression of a continuous shrinking. The PostGIS code to create a single buffer for one layers is as follows:

create table uso_sustentavel_95 as
 st_buffer(geom,0.95) as geom
from uso_sustentavel_95

We saved the multiple use reserve shapefile on a PostgresSQL server with a PostGIS plugin, named uso_sustentavel. Every territory has an identifier called gid. The creation_date in our dataset was constructed from each decree that inaugurated the reserve. With each decree the publication day is provided. The geom column refers to the information about the geographic properties of each polygon. The st_buffer() creates automatically areas with a buffer of 0.95 degrees around all polygons. To create the 20 buffer zones automatically I used a loop created in STATA12.

Evolution of protected areas in Brazil

Dear reader,

We are happy to open our new Blog about environmental policies in the tropics with an interesting Video showing the spatial evolution of protected areas in Brazil from 1920 to 2012. We created this video to visualize the efforts undertaken by the Brazilian government, NGOs and research institutes from all over the world to curb deforestation in one of the most dynamical agricultural frontiers of the world. Studies show that the establishment of protected areas was key to control deforestation in the Brazilian Amazon, where yearly rates of forest clear-cut dropped since 2005. Furthermore the demarcation of indigenous territories helped to preserve the habitat of traditional communities and avoid conflicts over land use.

Although this video shows impressive advances regarding the total size of protected areas, big challenges still remain to achieve long term sustainability of the undertaken efforts. These challenges comprise the management of established areas and the recovery of degraded land as well as the endorsement of sustainable income alternatives that balance ecological needs and socio-economic development in the region.

Our blog informs you about developments in tropical forest protection from all over the world but with an emphasis on the Amazon region. We are a team of researchers hosted at the Center for Development Research with field experience in Brazil, Peru and Bolivia. If you are interested in our work on environmental policies you can visit our project website or enjoy the upcoming posts on Blogazonia.

Your EP Research Team

You can download the video here.