Earth Sciences thesis of Jesse Schoenmakers
Date and place: 2-7-2018, Amsterdam Supervisor: Crystal McMichael, second supervisor: William Gosling
The Amazon Basin in the Anthropocene, spatial and temporal
distribution of human impacts in eastern Peru
Abstract
Although the Amazonian forests provide abundant ecosystem services, the dynamics and structure of these forests are poorly understood. Past disturbances, which can significantly affect current forest dynamics, are often unaccounted for in Amazonian forest research. This research assessed charcoal abundance, an important indicator of anthropogenic impact in the Amazon rainforest, from 12 soil cores in the Tapiche-Blanco area in Eastern Peru to determine the spatial and temporal distribution of both pre- and post-Columbian human impacts. Two charcoal fragments were dated to have formed 1627-1738 calibrated years before present (BP) and 1816-1934 calibrated years BP. However, this research was unable to compose a high resolution timeline of particles, it did however show pre-Columbian charcoal to be significantly more abundant compared with post-pre-Columbian charcoal. Furthermore, spatial distribution results suggest that the distance to rivers does not significantly influence the distribution of human impact in the Tapiche-Blanco region. This research concludes that the Tapiche-Blanco region was subjected to only small, infrequent and highly localized human impacts which predominantly occurred in interfluvial forests (> 10 km from river). Future research should integrate a more complete dataset of past disturbances with current vegetation data to enhance our understanding of the Amazon basin as a whole.
Keywords: Amazon forests, Tapiche-Blanco area, ancient human impact, disturbance, charcoal abundance, forest dynamics
Table of contents
1. Introduction 1
2. Methods
2.1 Tapiche-Blanco region site description 4
2.2 In field methodology 2.2.1 Soil charcoal 4 2.3 Laboratory methodology 2.3.1 Macroscopic charcoal 4 2.3.2 Charcoal dating 5 2.4 Statistical methodology
2.4.1 Spatial distribution and temporal analysis 5
2.4.2 Proximity to rivers 6
3. Results
3.1 Charcoal age 7
3.2 Spatial patterns 7
3.3 Pre- and post-Columbian analysis 9
3.4 Distance to rivers and likelihood of ancient human impact 9 4. Discussion
4.1 Charcoal age 12
4.2 Spatial and temporal distribution 12
5. Conclusion 16
6. Acknowledgements 17
7. References 18
1. Introduction
The Amazon basin, an area of approximately 5,800,000 km2 (Salati & Vose, 1981), is arguably the most species-rich terrestrial ecosystem in the world (Hoorn et al., 2010; Olson et al., 2001; Olson & Dinerstein, 2002). Furthermore, the Amazon basin provisions abundant ecosystem services such as carbon sequestration, water cycling and nutrient cycling (Zhao & Running, 2010; Olson & Dinerstein, 2002; Boerner et al., 2007). Most of these ecosystem services sustain human life at local, regional and global scales (Metzger et al., 2006).
The Amazon forests are a key component in the global carbon cycle (Phillips et al., 2009). However, the magnitude of the contribution of the Amazon forest to the carbon cycle is debated (Brienen et al., 2015). The models used for predicting Amazonian contribution to the carbon cycle are subjected to assumptions and are not able to take both annual and seasonal carbon fluxes into account (Stephens et al., 2007). Furthermore, because observation plots may be inherently located close to rare mortality events, studies are likely to overestimate the extent of anthropogenic disturbances which result in unreliable study results (Fisher et al., 2008). Most recent estimates indicate that the Amazon basin has a net storage of carbon of 0.56GtC annually, i.e. it is a carbon sink (Rödig et al., 2018). This estimate does not correspond with another recent study of Baccini et al., 2017, which implies that tropical forests are a carbon source, with most carbon emissions occurring in America’s tropical forests (Baccini et al., 2017). However, most studies strongly support the idea that carbon potential of the Amazon is declining, induced mostly by slowed tree productivity, declined carbon density, lowered tree longevity, deforestation and drought (Baccini et al., 2017; Brienen et al., 2015; Malhi et al., 2008).
To assess the role of the Amazon forests in the global carbon cycle, measurements of forest productivity, structure and biomass are often used (Rödig et al., 2018; Fearnside, 2018). Forest structure is defined as the spatial arrangement of the various components of the forest, such as the canopy level height and vegetation distribution (McElhinny, 2002). The forest structure has a substantial impact on forest productivity and forest biomass (Rödig et al., 2018). The accumulated vegetation of the Amazon forests are directly proportional to the carbon stock (Fearnside, 2018), with 1% of tree species being responsible for 50% of carbon storage and productivity (Fauset et al., 2015). Most aboveground biomass studies concern the Brazilian part of the Amazon. Biomass estimates show major differences between studies conducted with Light Detection and Ranging (LiDAR) (Saatchi et al., 2011; Baccini et al., 2012) and studies that used plot data with geographic information system analysis (GIS) (Mitchard et al., 2014). More accurate and corresponding biomass measurements requires improved remote sensing and more extensive ground-based measurements (Fearnside, 2018). The Amazon productivity is also influenced by the forest structure, with faster-growing genera causing higher growth rates, mortality and recruitment in western Amazonia (Phillips et al., 2008). Current forest productivity, biomass and structure are influenced by both climatic and edaphic conditions (Quesada et al., 2009).
Past disturbances also have substantial impacts on current Amazonian carbon sequestration by altering forest biomass and structure (Frolking et al., 2009). Trees that can live up to 1,400 years (Chambers, 1998) and gap dynamics that can alter forest composition for over a century (Hartshorn, 1978) are responsible for the legacy of past disturbances on current forest compositions. The implications of these long-term disturbance histories on our understanding of the Amazonian biodiversity and carbon dynamics are largely unknown (McMichael et al., 2017), as these
disturbances generally exceed the timespan of established inventory plots such as reported by Gloor et al., 2009. Depending on the duration and intensity of these past disturbances, the recovery of a
tropical system can take centuries (Chazdon et al., 2003). Therefore, ancient people, who form a likely source of past large-scale disturbances, should be taken into account when assessing forest dynamics (McMichael et al., 2017; Foster et al., 1999).Ancient people have been active in the Amazonian forest since the onset of the Holocene (Roosevelt et al., 1996). The occupation distribution of these ancient people is believed to be influenced by the presence of rivers which provide a source of protein and transportation (Bush et al., 2015; Denevan, 2012). Furthermore, neighbouring varzea floodplain forests provided nutrient rich soils and large bodied game (Bush et al,. 2015). Nonetheless, ancient people also occupied the interfluvial terra firme areas with evidence of villages of around 1,000 inhabitants in terra firme areas (Clement, 2015), this is debated. Research conducted by Souza et al., 2018 concluded that the presence of rivers in the southern rim of the Amazon basin had only little influence on ancient settlement patterns compared to more important factors such as bioclimatic variables and elevation (Souza et al., 2018). Humans actively transformed the Amazonian ecosystem by domesticating crops (Clement, 2006), the construction of fertile soils (McMichael et al., 2014) and the practice of agroforestry (Denevan & Padoch, 1988) which led to highly productive and orderly landscapes (Erickson, 2010). Historic Amazonian population estimates are highly debated, some estimates range between 8 to 20 million by the end of the pre-Columbian era, the time period prior to European colonization of the Americas (Clement et al., 2015; Denevan, 2012).
In the Amazon basin, forest fires are a synergetic consequence of drought and human activity (Bush et al., 2008; de Oliviera, 2008). Forest fires can lead to drastic changes in forest structure and composition with cascading effects (Barlow & Peres, 2008). Furthermore, research by Cochrane & Schulze, 1999 found forests that experienced fire events to be extremely heterogeneous with pioneer species dominating the understory (Cochrane & Schulze, 1999). Whereas it is plausible that fires occurred during the dryer millennia of the Early/Mid-Holocene between 8000 and 4000 cal. yr before present (BP) (Mayle & Power, 2008), the occurrence of naturally occuring forest fires are so rare that charcoal is interpreted as an indicator of human activity (Bush et al., 2008). Human activity likely transformed the Amazon basin through fire ignition for infield buring (Denevan, 2012), clearing of forests for hunting grounds and agriculture (Denevan & Padoch, 1988). Evidence of these micro paleoenvironmental changes can be found in the presence of charcoal combined with the presence of Asteraceae, which is an indicator of a cleared canopy paleoenvironment (Behling & Hooghiemstra, 2000). Human activities such as slash-and-burn agriculture can affect the forest vegetation
composition for up to 500 years (Riswan et al., 1985). More intensive or extensive types of
agriculture could however impose an even longer lasting legacy on current forest systems compared to slash-and-burn agriculture (Guariguata & Ostertag, 2001).
In addition to modifying forest biomass and structure through the ignition of fires, both pre- and post-Columbian people have been selecting certain tree species in areas where they lived (Levis et al., 2017). It is argued that the selection of economically valuable plants has caused certain species to become ‘hyper dominant’ and thereby leaving a legacy on current vegetation composition (Levis et al., 2017). In the Amazon basin as a whole, these hyperdominant species are believed to play a disproportionately larger role in forest productivity and carbon sequestration than less-common species (Fauset et al., 2015). The implications of these long-term disturbance histories on our understanding of the Amazonian biodiversity and carbon dynamics are largely unknown (McMichael et al., 2017). Depending on the duration and intensity of these past disturbances, the recovery of a tropical system can take centuries (Chazdon et al., 2003).
The magnitude of past human disturbances in Amazonian forests has only been assessed for three forest plots, the Los Amigos and Cocha Cashu sites which are both located in Peru and the
Manaus sites which are situated in Brazil. To gain a wider perspective on the extent and magnitude of past disturbances in the Amazon Basin, this research is set out to investigate the fire history for three sites in Eastern Peru. These sites are located between the Tapiche river and Blanco river, situated in the lesser studied western Amazonia. This research could provide a new step towards
comprehending the greatest “carbon machine” on Earth (Malhi et al., 2009) and contribute to our understanding of the Amazonian basin as a whole. To further assess the magnitude and distribution of past disturbances the following research questions will be answered:
· What is the time since the last fire in the forest plots?
· What is the spatial and temporal distribution of past fires within the plots?
It is hypothesized that there is no historic human impact in the Tapiche-Blanco is as there were no archeological artifacts found in the region (Pitman et al., 2015). Furthermore, on the basis of previous research by Bush et al., 2015, it is expected that soil charcoal frequencies will decrease as distance-to-river increases. Especially bluffs are believed to offer excellent settlement conditions (Bush et al., 2015). Therefore, this research hypothesizes that cores derived from riverine bluffs close to the river will contain more charcoal compared to cores that are located further away from the river. Lastly, the research hypothesizes that charcoal will be more abundant at pre-Columbian depths compared to charcoal abundance at post-Columbian depths due to a population collapse soon after European arrival (Bush and Silman, 2007).
2 Methods
2.1 Tapiche-Blanco region site description
The Tapiche-Blanco sites are located between the Tapiche river and its tributary, the Blanco river, located in Eastern Peru, close to the Brazilian border (Fig 1). The sampled area included camp 1 (Wiswincho; 5°48'30.59"S 73°51'56.06"W ), camp 2 (Sismica; 6°16'18.84"S
73°54'58.35"W) and camp 3 (Pobreza; 6°16'31.58"S 73°55'2.56"W). The sites are located at an elevation of around 150 meters above sea level, in one of the wettest parts of the Amazon basin (Burd et al., 2008). The mean annual rainfall in the area is approximately 2,300 mm, with the dry period ranging from May to October and the wetter months occurring between November and April (Hijmans et al., 2005). Average temperatures range from 25-27°C, with
high temperatures of 31–33°C and low temperatures of 20-22°C. The area comprises large wetlands, white-sand soils with stunted forests and upland hills and
terraces. This lowland Amazonian wilderness lacks infrastructure, and is subjected to active logging and hydrocarbon concessions.
2.2 In field methodology
2.2.1 Soil charcoal
Three to five soil samples were taken in each plot for the assessment of charcoal abundance.
When biomass is exposed to temperatures of 250 – 500°C, charcoal is formed (Antal et al., 1990). In this study “charcoal” is defined as solid residual macroscopic charcoal particles ≥ 500μm. The camp 1 samples were derived from a blackwater floodplain forest and surrounding peatland savanna and varillal. The other camps were located in terra firme areas. An auger was used to retrieve a core to a depth of 80 centimeters. The 80 centimeter soil core was subdivided into 8 subsamples, representing the 10 centimeter depth intervals (Pitman et al., 2015). An age depth-generalization was applied on all particles at a depth >20 centimeter as being pre-Columbian and <20 cm depth as being
post=Columbian as most published records of soil charcoal are pre-Columbian below this depth (Bush et al., 2008). The samples were sent to the laboratory of the University of Amsterdam for further analysis.
2.3 Laboratory methodology
2.3.1 Macroscopic charcoal
For the quantification of macroscopic charcoal, methods from McMichael et al., 2012a were adopted. These methods were adjusted methods of Whitlock & Larsen, 2002. First of all, a third of each subsample was selected by the use of a cylinder with a 2.42 centimeter diameter. The volume of the selected subsample was calculated in cm3, a total of 96 subsamples were volumetrically measured (N=96). These samples were deflocculated on a cookplate at 150 °C with 3% H2O2 for a
Table 1, description of sites Figure 1, Map of the study area
period of approximately 15 minutes. Subsequently, the finer subsample particles were passed through a 500μm sieve in order to retrieve macroscopic particles.
Charcoal fragments were identified on their visual and physical characteristics by means of a Bausch & Lomb microscope , with a 10x-25x magnification. The charcoal particles are visually
characterized as opaque angular and usually planar fragments (Whitlock & Larsen, 2002). Charcoal is very brittle and will fracture under pressure instead of compressing or impaling (Whitlock et al., 2002). The colour of charcoal is either a dark gray, grayish-black or deep black, in contrast to many other coal colors which have brownish hues (Sander & Gee, 1990). After the identification and quantification of macroscopic charcoal fragments the fragments were photographed using a fujifilm camera mounted on an Olympus stereo microscope with a 7x-40x magnification. The area of the photographed charcoal surface was calculated in mm2/cm3 using Image J software as described by Abràmoff et al., 2004. Lastly, the mm2/cm3 results were converted to mm3/cm3 by applying the simple calculation of ∑Ai3/2 proposed by Weng et al., 2005. The spatial distribution of these fire events were subsequently calculated as the proportion of cores per plot containing charcoal (McMichael et al., 2015). Stratigraphic diagrams were created using C2 software.
2.3.2 Charcoal dating
Charcoal particles are selected on size, the biggest particles were send to DirectAMS radiocarbon dating service in Seattle, USA. The uncalibrated radiocarbon age before present (BP) were obtained by means of accelerator mass spectrometry and send back to the University of Amsterdam.
Subsequently, the raw radiocarbon dating data were calibrated using Calib 7.10 software provided by
www.calib.org. The IntCal13 calibration curve was selected, this calibration curve was developed for latitudes in the Northern Hemisphere. However, the IntCal13 calibration curve was applied in this research because of the air masses interaction between both Northern and Southern Hemispheres (Marsh et al., 2018). Radiocarbon age BP and the standard deviation in age provided by DirectAMS was used to retrieve calibrated ages of the charcoal particles.
2.4 Statistical methodology
2.4.1 Spatial distribution and temporal analysis
The abundance of charcoal area per core was tested between camps and at different depths. The spatial distribution of charcoal abundance within the campsites and between the campsites was visualised with both a table and a map. For this analysis, a charcoal area threshold of >0.25 mm3 was taken into account to distinguish non-significant and significant charcoal presence, hereafter ‘trace’ and ‘present’ amounts of charcoal respectively (McMichael et al., 2012a). The table was drawn within Microsoft Excel and the map was drawn using the packages 'vegan', 'analogue',
'RColorBrewer', 'classInt', 'raster', 'rgdal', 'dismo', 'rJava', 'mapTools', 'SDMTools' within R.
A nested analysis of variance (ANOVA) was performed using R software. The formula for the nested ANOVA was set to be ‘Charcoal abundance is a function of campsite, site nested within campsite and pre- or post-Columbian depths nested within site and campsite (see appendix 8.1 for code).
2.4.2 Proximity to rivers
The proximity to rivers of every distinct site was measured using the measure tool in Google Earth Pro. An ANOVA was performed within R for each site. Rivers were selected by analysis of aerial photos and descriptions of Pitman et al., 2015, it was assumed that observable rivers were large enough to form a protein source and facilitate transport of small vessels such as canoes. The Blanco river formed the closest river to both camps 2 and 3, whereas a tributary of the Blanco river was the closest to camp 1. The formula used for this ANOVA was ‘Charcoal abundance is a function of distance to river’. Furthermore, the likelihood of ancient human impact (AHI) model (McMichael et al., 2017) will be applied on the coordinates of each Tapiche-Blanco campsite. The modeled
probability of ancient human impact of each campsite will be compared and tested on correlation to actual findings to further validate the model.
3. Results 3.1 Charcoal age
Two charcoal particles from camp 2 site 4 were dated, these particles were retrieved at 30-40 and 60-70 centimeters depth. The radiocarbon age for the particles at 30-40 centimeters was determined at 1788 years BP with an 1 sigma error of 27 years. The radiocarbon age for the particles at 60-70 centimeters was determined at 1920 years BP with an 1 sigma error of 28 years. Calibration revealed that particles from 30-40 centimeter depth were most likely formed between 1627-1738 years BP. The charcoal particles derived from the cores at 60-70 centimeters depth were formed between 1816 - 1934 calibrated years before present (table 2).
Table 2, calibrated age of charcoal particles with the area underneath the probability curve (probability)
3.2 Spatial patterns
The cores sampled in camp 1 contained the lowest charcoal amounts, with 40% of the cores lacking any charcoal at any depth (figure 2). The charcoal abundance in camp 1 ranged from 0.0101
mm3/cm3 to 1.4043 mm3/cm3 found at site 3 and site 2 respectively (figure 2). Within camp 1, ‘Present’ charcoal was most abundant at site 2 which revealed 50 ‘present’ particles compared to 1 and 4 ‘present’ particles found in sites 1 and 3 (figure 3).
75% of the cores derived from camp 2 experienced some form of fire event, with site 3 showing no evidence of having a fire history. Site 4 contained the highest charcoal abundance of all 12 cores presented here with a measured volume of 11.6172 mm3/cm3. The lowest amount of charcoal recorded in camp 2 was found in site 2 with a 0.0438 mm3/cm3 charcoal volume at a depth of 60 to 70 centimeters depth. Most ‘present’ charcoal particles were derived from site 4, with site 1 and 2 showing similar abundance in ‘present’ charcoal particles (figure 2).
Within the Tapiche-Blanco area, camp 3 showed the most extensive fire history. All 3 sites within camp 3 contained charcoal, ranging from 0.0175 mm3/cm3 to 6.4 mm3/cm3 in volume. Site 2 revealed the most extensive fire record, containing 74 ‘present’ charcoal particles. A nested ANOVA showed a significant variance in spatial distribution and explained significant charcoal abundance variance between campsites (d.f. = 2, F statistic = 7.791, p = 0.000917) and sites (d.f. = 9, F statistic = 5.734, p = 0.00000715), see appendix 8.1 table 5.
Figure 3, ‘present’ charcoal count in all camps, pre-Columbian ‘present’ charcoal is shown underneath
the dashed red line, post-Columbian ‘present’ charcoal is presented above the dashed red line Figure 2, Stratigraphic diagram of the charcoal abundance in camps ,1 2 and 3. The y-axis describes the depth in centimeters from the surface, grey blocks indicate presence of charcoal abundances which are not apparent on the selected scale. Scales are different per camp, white blocks indicate values exceeding scales.
3.3 Pre- and post-Columbian analysis
Nested ANOVA indicated that there were significant differences in the temporal distribution of charcoal. Charcoal was significantly more abundant in depths > 20 centimeters (d.f. = 12, F statistic = 2.705, p =0.005). In pre-Columbian times, most abundant charcoal records were found in camp 3 (figure 4). Camp 2 contained the highest amounts of post-Columbian charcoal. As shown in the boxplot figures 5 and 6, the variance between pre- and post-Columbian charcoal abundance was higher within camps than between camps.
3.4 Distance to rivers and likelihood of ancient human impact
Lastly, an ANOVA was performed to test whether ‘Charcoal abundance is a function of distance to river’, the result was significant with a p-value of 0.000017 (d.f. = 11, F statistic = 7.91), see table 6 appendix 8.2. The latter indicates that the distance to rivers which ranged from <2 km for camp 1 to >20 km for camp 2 (table 5 appendix 8.2) had a negative effect on charcoal abundance. The AHI model predicted no human impact at the campsites in pre-Columbian times. According to the AHI model, the probability for pre-Columbian presence at the campsites was 29.4%, 14.3% and 24.2% for camp 1, 2 and 3 respectively. The probabilities of human presence in post-Columbian times increased for camp 1 but declined for camp 2 and 3 (table 3).
Figure 4, spatial distribution of both pre- and post-Columbian charcoal
Table 3, AHI model results with corresponding campsite, Pr is the chance of encountering a form of human impact in an area. >0.5 indicates human presence, <0.5 indicates human absence.
Figure 5, boxplot of charcoal abundance and campsites, red or green rectangles represent 50% of the data. Red rectangles represent pre-Columbian formed charcoal, green rectangles represent charcoal formed in the post-Columbian era.
Figure 6, boxplot of charcoal abundance and sites, S stands for site and C for corresponding campsite, red or green rectangles represent 50% of the data. Red rectangles represent pre-Columbian formed charcoal, green rectangles represent charcoal formed in the post-Columbian era.
4. Discussion
The data presented in this paper suggest that all sampled areas in the Tapiche-Blanco area were subjected to past fire events of some form. The data also suggest that interfluvial areas experienced significant more fire events than areas close to rivers. The most charcoal abundant cores of camp 3 (overall ‘present’ charcoal cores/all cores = 1) were situated 10 kilometers from the river. Camp 2, which had an overall ‘present’ charcoal cores/all cores ratio of 0.75 was situated 22 kilometers from the closest river. These findings contradict findings of Bush et al., 2015 who states that soil charcoal frequencies decreased as distance-to-river increased. The study reported an overall ‘present’ charcoal cores/all cores ratio of approximately 0.22 at a distance of 10 kilometers from river and a value of at least under 0.08. The results represented here reject the hypothesis that the campsite closest to the river contained the highest charcoal frequencies. The temporal scale on which these fires occurred, and an accurate estimation on which the last fire per plot occurred were not retrieved. The resolution of dated particles was too low to draw conclusions for most charcoal presented in this research.
4.1 Charcoal age
Table 4 represents carbon dates of other particles at similar depth in the western Amazon Basin. When comparing the dated charcoal particles and
corresponding depths at the Tapiche-Blanco area with other areas in the western Amazon, more similarities were found between the Algodón plots (Visser, 2018) than with the Los Amigos
Table 4, calibrated ages of particles derived at similar depths. The Iquitos and Los Amigos data is retrieved from McMichael et al.,
plots (McMichael et al., 2012a). These similarities were however, due to a lack of data points, not statistically tested.
The calibrated ages by McMichael et al., 2012a were calibrated using the Fairbanks et al., 2005 calibration curve instead of the IntCal13 calibration curve used in this research and the Algodon age calibration. The usage of the IntCal13 calibration curve in the Tapiche-Blanco area is subjected to discussion, as this area is influenced by both Northern Hemisphere air masses and Southern
Hemisphere air masses. A mixed calibration curve as suggested by Marsh et al., 2018, which accounts for influence of both Hemispheres, could therefore enhance the accuracy of the Tapiche-Blanco 14 C calibration. A calibrated timeline of the Tapiche-Blanco sites was not established due to a lack of time and funds, the 14C dating results were insufficient to assess when the last fires occurred in the investigated plots.
4.2 Spatial and temporal distribution
As fire events are so rare in the research area, macro charcoal particles are interpreted to be an important indicator anthropogenic environmental impact (Bush et al., 2008). Although there were no archaeological artifacts found within the Tapiche-Blanco area, macro charcoal was abundant in the Tapiche-Blanco site cores. However, it is more likely that without stone tools it would have been very difficult for people to substantially alter or clear forests without using fire Bush et al., 2015 as cited by Kelly et al., 2018, thus magnifying the ‘fire footprint’ of humans in the area.
The spatial distribution patterns of past anthropogenic disturbances are debated.
Publications by Bush et al., 2015 and Denevan, 2012 state that the presence of rivers is an important factor for settlement of ancient people. The research presented here comprises data of few sites, which reduces the power of the performed statistical tests concerning spatial and temporal distribution of charcoal particles in the investigated cores. However, when retrieving the ratio of cores containing ‘present’ charcoal per overall amount of cores, this research contradicts model results as presented by Bush et al., 2015. According to Bush et al., 2015, the percentage of
encountered cores with ‘present’ charcoal is 22% at 10 km distance from rivers. The data presented here showed 100% of the cores contained ‘present’ charcoal at 10 km distance from rivers. The results obtained at campsite 2 which was situated 22 km from the Blanco river (appendix 8.2, table 5), showed 75% of the cores containing ‘present charcoal’, the latter far exceeded the expected values of Bush et al., 2015.
Cores derived from campsite 1 located in a floodplain and the surrounding bluff close to the Blanco river, revealed only a small amount of macroscopic charcoal. The lack of macroscopic charcoal within the floodplain corresponds to earlier studies by Denevan, 1996 who states that floodplains are high-risk habitats because of regular annual flooding and periodic extreme flooding events that overflow even the highest terrains (Denevan, 1996). Furthermore, it is also likely that any charcoal particle evidence was washed away during flooding events. However, the results presented in this paper do not support the hypothesis that neighboring bluffs, elevated areas close to rivers, are more suitable for human settlement compared to the floodplains themselves (Denevan, 1996) and
therefore more likely to contain evidence of ‘present’ charcoal. These bluffs are believed offered excellent settlement conditions (Bush et al., 2015), as they are situated above the high water level but still close to the river which provides protein and transportation (Bush et al., 2015; Denevan, 2012). The cores situated on the riverine bluff (Camp 1, site 4 and 5) were however the least abundant in charcoal when compared to the other campsites. These results do not correspond to earlier findings by McMichael et al., 2012. Out of several riverine and interfluvial sample sites,
Distance from sample sites to the closest river positively influenced the charcoal abundance. This significant positive influence contradicts any previous research. Because of the low statistical power however, no strong conclusions were drawn from these results. Furthermore, this study did not consider the likeliness of particles to have been washed away by the river. Moreover, the study did not test the effect of habitat types on charcoal abundance. Habitat types with fertile soils where resources were abundant and could easily be captured were also important factors for
pre-Columbian settlement patterns (McMichael 2012b). However, this research showed the campsite with the lowest mammal abundance (Pitman et al., 2015) to be most abundant in charcoal (campsite 2). Lower mammal abundance could be a legacy of human impact (C.N.H. McMichael, personal communication, July 3rd, 2018), the latter is however debated and unknown. It has to be stressed that no statistical tests concerning soil fertility and spatial distribution of charcoal
abundance were conducted. Future research should be performed to statistically test habitat types and charcoal abundance. Without considering the latter, data retrieved from floodplains could significantly disturb our analysis. This study does however, combining floodplain data in the analysis, call into question the prevalent idea of strong correlation between major riverways and human settlement.
There was no evidence of Asteraceae, which is an indicator of a cleared canopy paleoenvironment (Behling & Hooghiemstra, 2000) and is an excellent indicator of human disturbance (Piperno et al., 2015). The latter, combined with the lack of archeological artifacts, suggests that the sampled areas in the Tapiche-Blanco area were neither cultural parklands nor virgin forests both before and after European arrival. Furthermore, the variance in charcoal between sites in a camp suggest that people only impacted small patches of land. The latter results also correspond with earlier research by McMichael et al., 2012a which suggests that interfluvial forests were
subjected to only small, infrequent and highly localized human impacts (McMichael et al., 2012). Analysis of the vertical distribution of macroscopic charcoal in the Tapiche-Blanco sites (figure 2 and 3) revealed a significant difference between pre- and post-Columbian charcoal
abundance. Pre-Columbian charcoal were significantly more abundant compared to post-Columbian charcoal, the latter accepts the hypothesis that pre-Columbian charcoal records are more abundant compared to post-Columbian charcoal records. This significant temporal difference in charcoal abundance could be attributed to higher pre-Columbian population estimates compared to
population estimates after the time of European arrival. It is believed that the Amazonian population ranged between 8 to 20 million by the end of the pre-Columbian era, this population may have collapsed by 95% soon after European arrival as a result of European diseases (Bush and Silman, 2007; Denevan 1976).
Comparable studies conducted by Visser 2018 and McMichael 2012a presented dissimilar results. Of the 103 interfluvial cores investigated by McMichael et al., 2012a, 11 % contained pre-Columbian ‘present’ charcoal, 46% was found at depths associated with post-pre-Columbian times (McMichael et al., 2012a). The same study showed similar results for 74 plots closer to the river, 5% of the plots contained pre-Columbian ‘present’ charcoal compared to 36% of the cores that
contained post-Columbian charcoal. Visser 2018, reported no significant temporal difference
between the investigated cores, Visser 2018 did however report more consistency between pre- and post-Columbian ‘present’ charcoal compared to McMichael et al., 2012a (figure 8). It has to be stressed however that the accuracy of pre- and post-Columbian differentiation varies per site.
Differences in calibrated age BP and depth between sites are likely caused by different accumulation rate of layered deposits (Blaauw et al., 2007) which are site specific and subjected to environmental dynamics and temporal context of a sample site (Bennett, 1994; Polo-Díaz et al., 2016). Furthermore, the stratigraphic integrity of soils is often much more affected when compared to for example lake cores (McMichael et al., 2012b).
Both
spatial and
temporal distribution patterns of charcoal abundance in the Tapiche-Blanco area strongly contradict human impact probabilities as modeled by the AHI model. The latter suggests that the parameters of the AHI demand further calibration and alteration. This research proposes to run the model with a reduced importance for the parameter ‘distance to rivers’ to see whether results would correlate more with the observed values in the Tapiche-Blanco region. Furthermore, the temporal differences between pre- and post-Columbian human impact probabilities (table 3), suggest that the model fails to integrate accurate time dependent parameters.
The methodology used within this research was directly adopted from the ‘easily replicable and affordable [methods]’ used by McMichael et al., 2012a. This study supports the idea that the adopted methods are easily replicable and affordable. The latter methods could be easily integrated into research with similar objectives, thereby expanding our knowledge on the Amazon Basin as a whole. It has to be stressed however that these easy replicable and affordable methods do rely on some assumptions that increase the measurement error. Firstly, the accuracy of the calculations adopted from Weng et al., 2005 to convert charcoal area to volume is highly influenced by a human factor. Photographed area of charcoal can significantly be altered by photographing the charcoal particle at a different angle. Lastly, the methods that were not adopted by McMichael et al., 2012a to determine site distance to rivers. These methods consisted of using the measurement scale in Google Earth Pro. In order to limit measurement error, proximity tools in ArcGis should have been utilized instead.
When taking the above into account, further research should integrate the presented data with data from previous research to enhance statistical power. Furthermore, future research should elaborate more on the current state of the Amazon Basin and compare it with the extent of human disturbance presented in this paper. Current vegetation data on each plot should be compared to assess what the extent of past human disturbance is on the present vegetation plots. The latter would be an essential step to integrate the results from this study in the state of the current Amazon forests as a whole. Considering the magnitude of past charcoal reported in this study, it is
hypothesized that impact on the current vegetation is minor.
Figure 8, binary map of present charcoal reported by Visser 2018, white rectangle indicates no present charcoal found, black rectangle indicates presence of 'present' charcoal. Top part of the rectangle represents pre-Columbian times, top part represent post-Columbian times
Figure 7, binary map of present charcoal, white rectangle indicates no present charcoal found, black rectangle indicates presence of 'present' charcoal. Top part of the rectangle represents pre-Columbian times, top part represent post-Columbian times
5. Conclusion
The charcoal record from the Tapiche-Blanco sites provide an important insight in the extent of past human impacts in the region. The temporal and spatial distribution analysis of charcoal particles in the Tapiche-Blanco area provide two conclusions. Firstly, there was no evidence found of intensive landscape modifications by past humans such as the establishment of a cultural parkland in the Tapiche-Blanco area. However, human presence the Tapiche-Blanco region resulted in only small, infrequent and highly localized human impacts on the Amazon forests. These impacts were proven to be more substantial in pre-Columbian times compared to post-Columbian times. The temporal scale on which these impacts occurred, and an accurate estimation on which the last fire per plot occurred were not retrieved. The resolution of dated particles was too low to draw conclusions for most charcoal presented in this research.
Secondly, the spatial distribution of human impact in the Tapiche-Blanco area does not correlate with the modeled predictions of human impact by the AHI model. Evidence of human were predominantly found in interfluvial forests with a distance of 10 kilometers or more from the closest
river. The study calls into question the prevalent idea of strong correlation between major riverways and human settlement. This research does stress the importance of other factors which are
important for human settlement such as habitat types, these factors were not taken into account in this study. The results presented in this research do however contribute to our understanding of the history of the Tapiche-Blanco area and should be taken into account when assessing current forest dynamics in the area.
6. Acknowledgements
I thank the people of the Tapiche-Blanco area for allowing the extraction of soil cores. I am indebted to dr. C.N.H. McMichael and dr. W.D. Gosling at the University of Amsterdam for their assistance during the research period. Furthermore, I would like to thank Gemma Koelmans and Eva Visser for the successful cooperation and good times in the lab.
7. References
Abramoff, M.D., Magalhaes, P.J. & Ram, S.J. (2004) Image Processing with ImageJ. Biophotonics International, 11, 36-42.
Antal Jr, M. J., Mok, W. S., Varhegyi, G., & Szekely, T. (1990). Review of methods for improving the yield of charcoal from biomass. Energy & Fuels, 4(3), 221-225.
Baccini, A. G. S. J., Goetz, S. J., Walker, W. S., Laporte, N. T., Sun, M., Sulla-Menashe, D., ... & Samanta, S. (2012). Estimated carbon dioxide emissions from tropical deforestation improved by carbon-density maps. Nature Climate Change, 2(3), 182.
Baccini, A., Walker, W., Carvalho, L., Farina, M., Sulla-Menashe, D., & Houghton, R. A. (2017). Tropical forests are a net carbon source based on aboveground measurements of gain and loss. Science, 358(6360), 230-234.
Barlow, J., & Peres, C. A. (2008). Fire-mediated dieback and compositional cascade in an Amazonian forest. Philosophical Transactions of the Royal Society of London B: Biological Sciences, 363(1498), 1787-1794.
Barlow, J., Peres, C. A., Lagan, B. O., & Haugaasen, T. (2003). Large tree mortality and the decline of forest biomass following Amazonian wildfires. Ecology letters, 6(1), 6-8.
Behling, H., & Hooghiemstra, H. (2000). Holocene Amazon rainforest-savanna dynamics and climatic implications: high-resolution pollen record from Laguna Loma Linda in eastern Colombia. Journal of Quaternary Science, 15(7), 687-695.
Bennett, K. D. (1994). Confidence intervals for age estimates and deposition times in late-Quaternary sediment sequences. The Holocene, 4(4), 337-348.
Blaauw, M., Christen, J. A., Mauquoy, D., van der Plicht, J., & Bennett, K. D. (2007). Testing the timing of radiocarbon-dated events between proxy archives. The Holocene, 17(2), 283-288.
Boerner, J., Mendoza, A., & Vosti, S. A. (2007). Ecosystem services, agriculture, and rural poverty in the Eastern Brazilian Amazon: Interrelationships and policy prescriptions. Ecological Economics, 64(2), 356-373.
Brienen, R. J., Phillips, O. L., Feldpausch, T. R., Gloor, E., Baker, T. R., Lloyd, J., ... & Martinez, R. V. (2015). Long-term decline of the Amazon carbon sink. Nature, 519(7543), 344.
Bush, M. B., McMichael, C. H., Piperno, D. R., Silman, M. R., Barlow, J., Peres, C. A., ... & Palace, M. W. (2015). Anthropogenic influence on Amazonian forests in pre history: An ecological perspective. ‐ Journal of biogeography, 42(12), 2277-2288.
Bush, M. B., & Silman, M. R. (2007). Amazonian exploitation revisited: ecological asymmetry and the policy pendulum. Frontiers in Ecology and the Environment, 5(9), 457-465.
Bush, M. B., Silman, M. R., McMichael, C., & Saatchi, S. (2008). Fire, climate change and biodiversity in Amazonia: a Late-Holocene perspective.
Philosophical Transactions of the Royal Society B: Biological Sciences, 363(1498), 1795-1802.
Chambers, J. Q., Higuchi, N., & Schimel, J. P. (1998). Ancient trees in Amazonia. Nature, 391(6663), 135.
Chazdon, R. L. (2003). Tropical forest recovery: legacies of human impact and natural disturbances. Perspectives in Plant Ecology, evolution and systematics, 6(1-2), 51-71.
Clement, C. R. (2006). Fruit trees and the transition to food production in Amazonia. Time and complexity in historical ecology: Studies in the neotropical lowlands, 165-185.
Clement, C. R., Denevan, W. M., Heckenberger, M. J., Junqueira, A. B., Neves, E. G., Teixeira, W. G., & Woods, W. I. (2015). The domestication of Amazonia before European conquest. Proc. R. Soc. B, 282(1812), 20150813.
Cochrane, M. A., & Schulze, M. D. (1999). Fire as a Recurrent Event in Tropical Forests of the Eastern Amazon: Effects on Forest Structure, Biomass, and Species Composition 1. Biotropica, 31(1), 2-16.
Denevan, W. M. (1996). A bluff model of riverine settlement in prehistoric Amazonia. Annals of the Association of American Geographers, 86(4), 654-681
Denevan, W. M. (1976). The aboriginal population of Amazonia. The native population of the Americas in, 1492, 205-234. Denevan, W. M. (2012). Rewriting the late pre-European history of Amazonia. Journal of Latin American Geography, 11(1), 9-24. Denevan, W. M., & Padoch, C. (1988). Swidden-fallow agroforestry in the Peruvian Amazon. New York Botanical Garden.
Erickson, C. L. (2010). The transformation of environment into landscape: the historical ecology of monumental earthwork construction in the Bolivian Amazon. Diversity, 2(4), 618-652.
Fairbanks, R. G., Mortlock, R. A., Chiu, T. C., Cao, L., Kaplan, A., Guilderson, T. P., ... & Nadeau, M. J. (2005). Radiocarbon calibration curve spanning 0 to 50,000 years BP based on paired 230Th/234U/238U and 14C dates on pristine corals. Quaternary Science Reviews, 24(16-17), 1781-1796. Fauset, S., Johnson, M. O., Gloor, M., Baker, T. R., Monteagudo, A., Brienen, R. J., ... & Pitman, N. C. (2015). Hyperdominance in Amazonian forest carbon cycling. Nature communications, 6, 6857.
Fearnside, P. M. (2018). Brazil’s Amazonian forest carbon: the key to Southern Amazonia’s significance for global climate. Regional Environmental Change, 18(1), 47-61.
Fisher, J. I., Hurtt, G. C., Thomas, R. Q., & Chambers, J. Q. (2008). Clustered disturbances lead to bias in large scale estimates based on forest sample ‐ plots. Ecology Letters, 11(6), 554-563.
Foster, D. R., Fluet, M., & Boose, E. R. (1999). Human or natural disturbance: landscape scale dynamics of the tropical forests of Puerto Rico. ‐ Ecological applications, 9(2), 555-572.
Frolking, S., Palace, M. W., Clark, D. B., Chambers, J. Q., Shugart, H. H., & Hurtt, G. C. (2009). Forest disturbance and recovery: A general review in the context of spaceborne remote sensing of impacts on aboveground biomass and canopy structure. Journal of Geophysical Research: Biogeosciences, 114(G2).
Glaser, B., & Birk, J. J. (2012). State of the scientific knowledge on properties and genesis of Anthropogenic Dark Earths in Central Amazonia (terra preta de Índio). Geochimica et Cosmochimica acta, 82, 39-51.
Guariguata, M. R., & Ostertag, R. (2001). Neotropical secondary forest succession: changes in structural and functional characteristics. Forest ecology and management, 148(1-3), 185-206.
Hijmans, R. J., Cameron, S. E., Parra, J. L., Jones, P. G., & Jarvis, A. (2005). Very high resolution interpolated climate surfaces for global land areas. International journal of climatology, 25(15), 1965-1978.
Holmes, K. W., Chadwick, O. A., Kyriakidis, P. C., de Filho, S., Eliomar, P., Soares, J. V., & Roberts, D. A. (2006). Large area spatially explicit estimates of ‐ tropical soil carbon stocks and response to land cover change. ‐ Global Biogeochemical Cycles, 20(3).
Hoorn, C., Wesselingh, F. P., Ter Steege, H., Bermudez, M. A., Mora, A., Sevink, J., ... & Jaramillo, C. (2010). Amazonia through time: Andean uplift, climate change, landscape evolution, and biodiversity. science, 330(6006), 927-931.
Kelly, T. J., Lawson, I. T., Roucoux, K. H., Baker, T. R., Honorio Coronado, E. N., Jones, T. D., & Rivas Panduro, S. (2018). Continuous human presence ‐ without extensive reductions in forest cover over the past 2500 years in an aseasonal Amazonian rainforest. Journal of Quaternary Science, 33(4), 369-379.
Levis, C., Costa, F. R., Bongers, F., Peña-Claros, M., Clement, C. R., Junqueira, A. B., ... & Castilho, C. V. (2017). Persistent effects of pre-Columbian plant domestication on Amazonian forest composition. Science, 355(6328), 925-931.
Maezumi, S. Y., Whitney, B. S., Mayle, F. E., de Souza, J. G., & Iriarte, J. (2017). Reassessing climate and pre-Columbian drivers of paleofire activity in the Bolivian Amazon. Quaternary International.
Malhi, Y., Roberts, J. T., Betts, R. A., Killeen, T. J., Li, W., & Nobre, C. A. (2008). Climate change, deforestation, and the fate of the Amazon. science, 319(5860), 169-172.
Malhi, Y., Saatchi, S., Girardin, C., & Aragão, L. E. (2009). The production, storage, and flow of carbon in Amazonian forests. Amazonia and Global Change, 355-372.
Marsh, E. J., Bruno, M. C., Fritz, S. C., Baker, P., Capriles, J. M., & Hastorf, C. A. (2018). IntCal, SHCal, or a Mixed Curve? Choosing a 14 C Calibration Curve for Archaeological and Paleoenvironmental Records from Tropical South America. Radiocarbon, 1-16.
Mayle, F. E., & Power, M. J. (2008). Impact of a drier Early–Mid-Holocene climate upon Amazonian forests. Philosophical Transactions of the Royal Society B: Biological Sciences, 363(1498), 1829-1838.
McElhinny, C. (2002). Forest and woodland structure as an index of biodiversity: a review. A literature review commissioned by NSW National Parks and Wildlife Service.
McMichael, C. H., Correa-Metrio, A., & Bush, M. B. (2012a). Pre-Columbian fire regimes in lowland tropical rainforests of southeastern Peru. Palaeogeography, Palaeoclimatology, Palaeoecology, 342, 73-83.
McMichael, C. N., Matthews-Bird, F., Farfan-Rios, W., & Feeley, K. J. (2017). Ancient human disturbances may be skewing our understanding of Amazonian forests. Proceedings of the National Academy of Sciences, 114(3), 522-527.
McMichael, C. H., Palace, M. W., Bush, M. B., Braswell, B., Hagen, S., Neves, E. G., ... & Czarnecki, C. (2014). Predicting pre-Columbian anthropogenic soils in Amazonia. Proceedings of the Royal Society of London B: Biological Sciences, 281(1777), 20132475.
McMichael, C. H., Piperno, D. R., Neves, E. G., Bush, M. B., Almeida, F. O., Mongeló, G., & Eyjolfsdottir, M. B. (2015). Phytolith assemblages along a gradient of ancient human disturbance in western Amazonia. Frontiers in Ecology and Evolution, 3, 141.
McMichael, C. H., Piperno, D. R., Bush, M. B., Silman, M. R., Zimmerman, A. R., Raczka, M. F., & Lobato, L. C. (2012b). Sparse pre-Columbian human habitation in western Amazonia. Science, 336(6087), 1429-1431.
Metzger, M. J., Rounsevell, M. D. A., Acosta-Michlik, L., Leemans, R., & Schröter, D. (2006). The vulnerability of ecosystem services to land use change. Agriculture, ecosystems & environment, 114(1), 69-85.
de Oliveira, R. R. (2008). Environmental history, traditional populations, and paleo-territories in the Brazilian Atlantic Coastal Forest. Global Environment, 1(1), 176-191.
Mitchard, E. T., Feldpausch, T. R., Brienen, R. J., Lopez Gonzalez, G., Monteagudo, A., Baker, T. R., ... & Steege, H. (2014). Markedly divergent ‐ estimates of Amazon forest carbon density from ground plots and satellites. Global Ecology and Biogeography, 23(8), 935-946.
Olson, D. M., & Dinerstein, E. (2002). The Global 200: Priority ecoregions for global conservation. Annals of the Missouri Botanical garden, 199-224. Olson, D. M., Dinerstein, E., Wikramanayake, E. D., Burgess, N. D., Powell, G. V., Underwood, E. C., ... & Loucks, C. J. (2001). Terrestrial rEcoregions of the World: A New Map of Life on Earth: A new global map of terrestrial ecoregions provides an innovative tool for conserving biodiversity. BioScience, 51(11), 933-938.
Phillips, O. L., Aragão, L. E., Lewis, S. L., Fisher, J. B., Lloyd, J., López-González, G., ... & Van Der Heijden, G. (2009). Drought sensitivity of the Amazon rainforest. Science, 323(5919), 1344-1347.
Phillips, O. L., Lewis, S. L., Baker, T. R., Chao, K. J., & Higuchi, N. (2008). The changing Amazon forest. Philosophical Transactions of the Royal Society B: Biological Sciences, 363(1498), 1819-1827.
Piperno, D. R., McMichael, C., & Bush, M. B. (2015). Amazonia and the Anthropocene: What was the spatial extent and intensity of human landscape modification in the Amazon Basin at the end of prehistory?. The Holocene, 25(10), 1588-1597.
Polo-Díaz, A., Benito-Calvo, A., Martínez-Moreno, J., & Torcal, R. M. (2016). Formation processes and stratigraphic integrity of the Middle-to-Upper Palaeolithic sequence at Cova Gran de Santa Linya (Southeastern Prepyrenees of Lleida, Iberian Peninsula). Quaternary International, 417, 16-38. Quesada, C. A., Lloyd, J., Schwarz, M., Baker, T. R., Phillips, O. L., Patiño, S., ... & Lloyd, G. (2009). Regional and large-scale patterns in Amazon forest structure and function are mediated by variations in soil physical and chemical properties. Biogeosciences Discussion, 6, 3993-4057.
Riswan, S., Kenworthy, J. B., & Kartawinata, K. (1985). The estimation of temporal processes in tropical rain forest: a study of primary mixed dipterocarp forest in Indonesia. Journal of Tropical Ecology, 1(2), 171-182.
Rödig, E., Cuntz, M., Rammig, A., Fischer, R., Taubert, F., & Huth, A. (2018). The importance of forest structure for carbon flux estimates in the Amazon rainforest. Environmental Research Letters.
Roosevelt, A. C., Da Costa, M. L., Machado, C. L., Michab, M., Mercier, N., Valladas, H., ... & Sliva, J. (1996). Paleoindian cave dwellers in the Amazon: the peopling of the Americas. Science, 272(5260), 373-384.
Saatchi, S., Marlier, M., Chazdon, R. L., Clark, D. B., & Russell, A. E. (2011). Impact of spatial variability of tropical forest structure on radar estimation of aboveground biomass. Remote Sensing of Environment, 115(11), 2836-2849.
Salati, E., & Vose, P. B. (1984). Amazon basin: a system in equilibrium. Science, 225(4658), 129-138.
Sander, P. M., & Gee, C. T. (1990). Fossil charcoal: techniques and applications. Review of Palaeobotany and Palynology, 63(3-4), 269-279. Souza, J. G., Schaan, D. P., Robinson, M., Barbosa, A. D., Aragão, L. E., Marimon Jr, B. H., ... & Iriarte, J. (2018). Pre-Columbian earth-builders settled along the entire southern rim of the Amazon. Nature communications, 9(1), 1125.
Stephens, B. B., Gurney, K. R., Tans, P. P., Sweeney, C., Peters, W., Bruhwiler, L., ... & Aoki, S. (2007). Weak northern and strong tropical land carbon uptake from vertical profiles of atmospheric CO2. Science, 316(5832), 1732-1735.
Weng, C. (2005). An improved method for quantifying sedimentary charcoal via a volume proxy. The Holocene, 15(2), 298-301
Whitlock, Cathy, and Chris Larsen. "Charcoal as a fire proxy." Tracking environmental change using lake sediments. Springer, Dordrecht, 2002. 75-97. Zhao, M., & Running, S. W. (2010). Drought-induced reduction in global terrestrial net primary production from 2000 through 2009. science, 329(5994), 940-943.
8. Appendix
8.1 Code used for spatial analysis 1
install.packages("nlme") install.packages("ggplot2") install.packages("maptools") library(nlme) library(ggplot2) library(maptools) getwd() setwd("C:/Users/Dell/Documents/Bachelorproject/TapicheBlanco/Methods") s <- read.csv("NestedANOVA2.csv", header = TRUE, sep = ";")
ggp <- ggplot(data = s, mapping = aes(x=as.factor(Camp), y=Charcoal_abun, fill=as.factor(Post_Pre))) + geom_boxplot()
print(ggp)
geom_boxplot() print(ggp) aov1 <- aov(s2$Charcoal_abun ~ as.factor(s2$Camp) + as.factor(s2$Site):as.factor(s2$Camp) + as.factor(s2$Post_Pre):as.factor(s2$Site):as.factor(s2$Camp)) summary(aov1)
Charcoal abundance is a function of' df Sum Sq Mean Sq F value Pr(>F0
camp 2 1.542 0.7708 7.791
0.00091 7
camp and site 9 5.105 0.5673 5.734 7.15E-06
camp, site and post- or pre-Columbian time 12 3.211 0.2676 2.705 0.005
residuals 66 6.539 0.0989
Table 5, results of nested ANOVA with charcoal abundance as function of camp, camp and site or camp, site and post- or pre-Columbian era
8.2 Code used to test distance to rivers and charcoal abundance
install.packages("ggplot2") install.packages("maptools") library(nlme) library(ggplot2) library(maptools) getwd() setwd("C:/Users/Dell/Documents/Bachelorproject/TapicheBlanco/Methods") s <- read.csv("NestedANOVA.csv", header = TRUE, sep = ";")
ggp <- ggplot(data = s, mapping = aes(x=as.factor(DistanceRiver), y=Charcoal_abun, fill=as.factor(Post_Pre))) + geom_boxplot()
print(ggp)
# you will see that there is one really high value at camp3 - this is basically an outlier # to really see differences between these camps
s2 <- s[s$Charcoal_abun < 3, ]
ggp <- ggplot(data = s2, mapping = aes(x=as.factor(Site), y=Charcoal_abun, fill=as.factor(Post_Pre))) + geom_boxplot()
aov1 <- aov(s2$Charcoal_abun ~ as.factor(s2$DistanceRiver)) summary(aov1) Site DistanceRiver C2S3 22240.26 C2S3 22240.26 C2S3 22240.26 C2S3 22240.26 C2S3 22240.26 C2S3 22240.26 C2S3 22240.26 C2S3 22240.26 C2S2 22232.54 C2S2 22232.54 C2S2 22232.54 C2S2 22232.54 C2S2 22232.54 C2S2 22232.54 C2S2 22232.54 C2S2 22232.54 C2S1 22211.56 C2S1 22211.56 C2S1 22211.56 C2S1 22211.56 C2S1 22211.56 C2S1 22211.56 C2S1 22211.56 C2S1 22211.56 C2S4 22249.01 C2S4 22249.01 C2S4 22249.01 C2S4 22249.01 C2S4 22249.01 C2S4 22249.01 C2S4 22249.01 C2S4 22249.01 C1S1 1815.28 C1S1 1815.28 C1S1 1815.28 C1S1 1815.28 C1S1 1815.28 C1S1 1815.28
C1S1 1815.28 C1S2 1790.13 C1S2 1790.13 C1S2 1790.13 C1S2 1790.13 C1S2 1790.13 C1S2 1790.13 C1S2 1790.13 C1S2 1790.13 C1S3 1898.41 C1S3 1898.41 C1S3 1898.41 C1S3 1898.41 C1S3 1898.41 C1S3 1898.41 C1S3 1898.41 C1S3 1898.41 C1S4 3805.44 C1S4 3805.44 C1S4 3805.44 C1S4 3805.44 C1S4 3805.44 C1S4 3805.44 C1S4 3805.44 C1S4 3805.44 C1S5 3641.89 C1S5 3641.89 C1S5 3641.89 C1S5 3641.89 C1S5 3641.89 C1S5 3641.89 C1S5 3641.89 C1S5 3641.89 C3S2 10182.76 C3S2 10182.76 C3S2 10182.76 C3S2 10182.76 C3S2 10182.76 C3S2 10182.76 C3S2 10182.76 C3S2 10182.76 C3S3 10571.2 C3S3 10571.2 C3S3 10571.2 C3S3 10571.2 C3S3 10571.2
C3S3 10571.2 C3S3 10571.2 C3S3 10571.2 C3S4 10915.55 C3S4 10915.55 C3S4 10915.55 C3S4 10915.55 C3S4 10915.55 C3S4 10915.55 C3S4 10915.55 C3S4 10915.55
Table 5, distance of closest river to plots. C = camp and S = site.
Charcoal abundance is a function of' df Sum Sq Mean Sq F value Pr(>F0
distance to river 11 1.542 0.7708 7.791 0.000917
residuals 9 5.105 0.5673 5.734 7.15E-06
Table 6, results of nested ANOVA with charcoal abundance as function of camp, camp and site or camp, site and post- or pre-Columbian era
