An assessment of the role of geodiversity in species diversity of the bird of paradise in New Guinea

(1)

An assessment of the role of geodiversity

in species diversity of the bird of paradise

in New Guinea

Demi Coenraads

University of Amsterdam

Supervisor: dr. K.F. Rijsdijk

(2)

2

Abstract

The main island of New Guinea is situated on the convergent plate margin of the continental Australian plate and the oceanic Pacific plate. The area is characterized by products of subduction and fault zones, making it a highly geologically complex area. The bird of paradise is mainly endemic to New Guinea and has the highest species diversity on the island. This research assesses the role of geodiversity on the species diversity of the bird of paradise and if land use change is endangering this species. GIS is used to calculate a GDI on different resolutions for New Guinea with geomorphological, geological and hydrological. In addition, a correlation analysis is performed with the species diversity and the relationship between species diversity and land use is investigated. The results of this research have shown that there is a positive relationship between geodiversity and species diversity of the bird of paradise in New Guinea. This is mainly caused by the strong positive correlation with geological and geomorphological units, as species diversity is weak negatively correlated to hydrology. The high geological and geomorphological variety is caused by the tectonic movements of the past and present, resulting in a variation of landforms and rocks. The hydrological variety in this area is very low, indicating that the species diversity is not affected by hydrological variety. The bird of paradise is the most diverse in the forested highlands, which largely protects itself due to its inaccessibility through ruggedness. Further research should focus on finding an appropriate resolution and method to assess geodiversity. Additionally, including ruggedness and slope into the geodiversity index could help with predicting which areas are vulnerable to land use change and where this comes together with high endemic species diversity.

Keywords: geodiversity index, species diversity, bird of paradise, land use change, conservation, New Guinea

(3)

3

Introduction

The main island of New Guinea is divided between Indonesia in the west and Papua New Guinea in the east. The island is situated on a convergent margin of the Australian and Pacific microplates (Figure 1). The oceanic Pacific microplates South Bismarck (SBS), North Bismarck (NBS) and Caroline (CAR) subduct beneath the continental Australian plate. Major fault zones are found at the Bird’s Head Block (BHB) and above the Mobile Belt (Baldwin, Fitzgerald, & Webb, 2012). At the fault zones terranes have accreted. Terranes are crust material which is uplifted onto another part of crust. Accreted terranes often include ophiolites, which is an uplifted and exposed part of the oceanic lithosphere (Heads, 2002). Tectonics divided the main island of New Guinea in three regions (Figure 1). The middle region is a mountainous region as a result from the subduction. This region is called the New Guinea Highlands (NGH). The mountains in this region peak at 4900 m above sea level. Furthermore, this area is characterized by volcanic activity, high variety and accumulation of sediments, land uplift and formation of methamorphic rocks (Davies, 2012; Gray, 2008). The NGH are also described as the major geological boundary of New Guinea (Heads, 2002). The part situated south from the NGH is a sedimentary basin and the north is characterized by an aggregation of oceanic and metamorphic rocks and accreted terranes (Davies, 2012).

The climate in New Guinea is throughout the year characterized by high precipitation, humidity and temperatures. Dry conditions last from June to October and from December to April, monsoons increase the precipitation in the whole country. Due to the decline in cloud cover the drought is most severe in the highlands. Furthermore, at altitudes starting at 2200 m the frost is increasing (Allen & Bourke, 1997). Climate models predict increases in precipitation in wetter areas, higher annual temperatures and larger seasonal extremes due to climate change (Pratt & Beehler, 2015).

Figure 1. Tectonic map of New Guinea. Abbreviations: ADB, Adelbert block; AOB, April ultramafics; AUS,

Australian plate; BHB, Bird’s Head block; CM, Cyclops Mountains; CWB, Cendrawasih block; CAR, Caroline microplate; EMD, Ertsberg Mining District; FA, Finisterre arc; IOB, Irian ophiolite belt; KBB, Kubor & Bena blocks (including Bena Bena terrane); LFTB, Lengguru fold-and-thrust belt; MA, Mapenduma anticline; MB, Mamberamo Basin block; MO, Marum ophiolite belt; MHS, Manus hotspot; NBS, North Bismarck plate; NGH, New Guinea highlands block; NNG, Northern New Guinea block; OKT, Ok Tedi mining district; PAC, Pacific plate; PIC, Porgera intrusive complex; PSP, Philippine Sea plate; PUB, Papuan Ultramafic Belt ophiolite; SB, Sepik Basin block; SDB, Sunda block; SBS, South Bismarck plate; SIB, Solomon Islands block; WP, Wandamen peninsula; WDK, Woodlark microplate; YQ, Yeleme quarries (Baldwin, Fitzgerald, & Webb, 2012).

The island of New Guinea supports a great variety of endemic bird fauna, meaning that these bird species are found nowhere else and speciation only takes place on the island (Pratt & Beehler, 2015). The bird of paradise has the highest species diversity of the bird fauna in New Guinea (Figure 2). The birds of paradise have 42 species. 38 species, from which 36 endemic species, are found in New Guinea. In addition to this, four species, from which two endemic are found in Australia and the Moluccas have two endemic species. Together with the bowerbird family, the birds of paradise form

(5)

5 the most distinctive group of fauna (Heads, 2001). The high endemism of the bird of paradise can be explained by the distinction of oceanic and land-bridge islands (Pratt & Beehler, 2015). Oceanic islands are separated from New Guinea by deep water. Land-bridge islands are separated from New Guinea by relatively shallow water, less than 100 m deep, and therefore they were connected to New Guinea when the sea level was low. The bird of paradise is a sedentary bird with a small range, so it does not fly easily across barriers (Heads, Birds of paradise, vicariance biogeography and terrane tectonics in New Guinea, 2002). Therefore, this species is mainly found on the land-bridge islands. The most endemic species are found on larger islands, due to the higher variation in habitats (Pratt & Beehler, 2015). Species of the bird of paradise appear at altitudes differing from 0 to 4000 m. The highest number of species is found at an altitude between 1000-2000m. The landscape at this altitude is lower montane forest with high trees and with little moss cover (Heads, 2001).

Pratt & Beehler (2015) describe how the landscape of New Guinea is changing due to human actions. Habitats are decreasing due to deforestation. Former forested areas are now being used for agriculture purposes or natural resource development. The population density in the New Guinea Highlands between altitudes of 1400 m and 2850 m is high. The sloping land is suitable for subsistence agriculture, because soil saturated conditions can be avoided during increased rainfall periods (McVicar & Bierwirth, 2001). Population growth causes a growing demand for food, which can increase deforestation for agricultural intensification (Shearman, et al., 2008).

This research assesses the geodiversity of New Guinea and how this affects the species diversity of the bird of paradise. Geodiversity captures the spatial diversity in topography, geology, geomorphology and hydrology (Hjort, Heikinnen, & Luoto, 2012). The biotic components of Earth are captured by biodiversity. This is measured by measuring different components, such as all species richness, vegetation and organisms, in an area (Swingland, 2001). Many people see biodiversity and geodiversity as two separate concepts. However, conservation management needs an approach that recognizes the relationship between biodiversity and geodiversity (Matthews, 2014). For instance, Schrodt, et al. (2019) illustrate the importance of understanding geodiversity for biodiversity with an example about resource extraction. Geological and mineral diversity is important for local ecosystems. By removing natural resources, the geological and mineral diversity is decreased and deforestation for mining removes habitats as well (Pratt & Beehler, 2015).

Matthews (2014) describes the need for a geo-ecological approach that acknowledges geodiversity as an important component in the process of deciding which areas need to be protected. Geodiversity affects biodiversity in many ways, for example climate influences species distributions and a geomorphological change, such as a landslide, can remove or even create new habitats. Biodiversity and geodiversity need to be combined together in one overarching term: natural diversity. When conservation plans are made, both need to be observed and measured. Assessment of all protected areas is necessary to determine what geodiversity is already protected and which hazards of future Figure 2. Number of species of Paradisaeidae in 110 km grid

(6)

6 environmental change are present. The integration of geodiversity in systematic conservation planning will add an extra dimension in selecting areas that need protection (Matthews, 2014).

Faith et al. (2000) emphasize the importance of including species data for conservation planning. They state that the current knowledge about biodiversity in New Guinea is based on the descriptions of the environment and forest types and that the focus needs to be shifted to the locations of species. Many other countries and regions with available species data use this as a biodiversity surrogate for conservation planning. Species data in New Guinea was very limited and is therefore used significantly less for conservation planning than in other countries. However, species data is available for the bird of paradise and Sodhi et al. (2005) points out that research on the effects of land use changes on bird richness in tropical montane and submontane regions in Southeast Asia is rare.

Therefore, the aim of this research is to provide a framework on how geodiversity can affect species diversity and how geodiversity should be implemented future conservation plans. Accordingly, the main research question is: how can assessing geodiversity help with maintaining the species diversity of the bird of paradise in New Guinea?This research question will be answered through three sub questions.

1. What does geodiversity look like in New Guinea in a resolution of 110 km and how does this differ from smaller resolutions?

2. Do geodiversity and the different aspects of geodiversity have a positive effect on the species diversity of the bird of paradise in New Guinea?

3. How does land use change affect species diversity?

Firstly, the methods and data will be described. Secondly, the results will be shown. Thereafter, the results will be explained in the discussion and finally, the conclusions will be disclosed.

Methods and data

The workflow (Figure 3) is divided in two parts. One describes how the geodiversity is calculated and used to find a relationship with species diversity. This part of the workflow is partially based on the method

described by Seijmonsbergen,

Guldenaar & Rijsdijk (2017). The other part describes how the relationship between land use and species diversity is assessed. The software ArcGIS Pro 2.4 (GIS) is used to produce the results, conduct the statistical analyses and to digitalize the species diversity map (Figure 2). Geology, geomorphology and hydrology maps are collected (Table 1) to calculate the geodiversity index (GDI). Furthermore, a coastline feature is used to clip all maps for New Guinea. In addition, a land cover map (Table 1) is used to assess the influence of land use on species

diversity. The following steps were taken in GIS to collect the results.

Figure 3. Workflow of calculating geodiversity index, correlation

with species diversity, and assessment of relationship between land use and species diversity.

(7)

7 1. The World Coastline raster dataset (Table 1) is used to extract a polygon feature of the coastline of New Guinea. This new feature is used to clip all the maps so that they only show data for New Guinea.

2. Hjort, Heikinnen & Luoto (2012) describe a method to calculate GDI. The GDI will be summing the variety in geomorphology, hydrology and geology (Table 1). Hjort, Heikinnen & Luoto (2012) describe geology, geomorphology and hydrology as the main components of geodiversity. The land systems dataset (Table 1) shows geomorphological and hydrological features. Therefore, the hydrological components are extracted from the land systems map. This results in two separate maps, one showing the geomorphological features and another showing the hydrological features.

3. The geological, geomorphological and hydrological maps are converted to raster datasets using the Feature to Raster tool. To calculate the variety in elements, a fishnet grid is created. The calculations of variety will be done for each grid cell.

4. The recommended boundaries of the grid resolution are calculated using the formula as described by Hengl (2006). The recommended resolution is calculated by multiplying the scale number of the coarsest dataset by 0.0005 (Hengl, 2006). The coarsest dataset used is the geology map with a resolution of 1:1,000,000 (Table 1). The recommended resolution for the calculation of the geodiversity index is 500 m. The coarsest legible resolution is calculated by multiplying the scale number with 0.0025. The coarsest legible resolution is 2.5 km. To assess the geodiversity in New Guinea, the GDI will be calculated in both resolutions.

5. In addition to two different resolution, the GDI will be calculated for a grid cell size of 110 km as well to calculate the spatial correlation with the species diversity map (Figure 2) which is visualized in grid cells of 110 km by 110 km. The Create Fishnet tool is used to create a fishnet grid with a grid cell size of 110 km by 110 km.

6. The Zonal Statistics tool is used to calculate the variety of statistics of all three maps within the cells with the same size of the fishnet grid that was created. Subsequently, the zonal statistics layers are reclassified using the Reclassify tool. The Natural Breaks method (Jenks, 1967) is used to classify five classes of variety within geomorphology, geology and hydrology: 1 Very low; 2 Low; 3 Medium; 4 High; 5 Very High. The new reclassified raster datasets are then combined into one raster dataset using the Cell Statistics tool. The input of this tool are the geology, geomorphology and hydrology indices and creates the geodiversity ranges as an output using the sum function. After summing the three raster datasets, the output is reclassified into the five classes as described earlier. The output is the geodiversity index, a map that will show the geodiversity of New Guinea.

7. Then, the species diversity map is digitalized. The same fishnet grid with the grid cell size of 110 km is used. A new field ‘Number_Species’ is added to the attribute table of this feature layer. The data of number of species in each grid cell is added in this field. Then, feature with the species diversity data is converted into a raster dataset using the Feature to Raster tool. This new raster layer is reclassified using the Reclassify tool. The Natural Breaks method (Jenks, 1967) is used to classify five classes of diversity in species: 1 Very low; 2 Low; 3 Medium; 4 High; 5 Very High. The species diversity raster map shows the spatial distribution of the species diversity of the bird of paradise in New Guinea.

8. To assess if there is a positive relationship between species diversity and geodiversity, Pearson’s R correlation coefficients are calculated. The Band Collection Statistics tool is used for this calculation. This produces an ASCII file that provides a correlation matrix among other statistical information. The correlation coefficient can vary from -1, indicating a strong negative correlation, to +1, indicating a strong positive correlation.

9. For the land use analysis, a global land cover is used (Table 1). This data consists out of two separate maps for the continents of Asia and Oceania. The Mosaic to Raster tool is used to combine these two raster files into one raster file. From this map, the land cover data for New Guinea is selected using the earlier created coastline of New Guinea feature and the Clip Raster tool. This data is reclassified, combining all forest classes into one class for all forest types.

(8)

8 Subsequently, the Zonal Histogram tool is used to calculate the number of cells of every land cover type within the predefined zones of the species diversity map.

Table 1. Metadata of the used datasets.

Map Description Data

type Coordinate system Scale/cell size Publication date Source

Geology Geological units Polygon WGS 1984 Lambert Conformal Conic 1:1 000 000 1999 Steinshou er et al. (1999) Landsystems of Indonesia and Papua New Guinea Geomorphological and hydrological units

Polygon GCS_WGS_1984 1:500 000 2010 Saxon & Sheppard (2010) World Coastline Coastlines Line GCS_WGS_1984 1:10 000 000 2009 Natural Earth Global Land

Cover Land use Raster GCS_WGS_1984 100 m 2019 Buchhorn et al. (2019)

Results

Geodiversity in New Guinea

The GDI of New Guinea is calculated in three different grid cell sizes. The GDI is calculated in grid cells of 500 m (Figure 4a), 2500 m (Figure 4b) and 110 km (Figure 4c). See Appendix 1-3 for an enlargement of the figures.

Figure 4. The geodiversity index of New Guinea calculated from hydrology, geomorphology and geology

indices visualized in three different resolutions and in five classes varying from very low to very high. a) The GDI calculated in grid cells of 500m. b) The GDI calculated in grid cells of 2500 m. c) The GDI calculated in grid cells of 110 km.

(9)

9

Species diversity and geodiversity

The number of species of the bird of paradise per grid cell of 110 km by 110 km in New Guinea reclassified into five classes (Figure 5). The map shows five classes varying from yellow which means very low species diversity (three or less species) to red squares meaning very high species diversity (more than 14 species).

Figure 5. The Species diversity of the bird of paradise in New Guinea in grid cells of 110 km by 110 km. The

species diversity is visualized in five classes. Adapted from Heads (2001).

The correlation between species diversity (Figure 5), geology (Appendix 5), geomorphology (Appendix 6) and hydrology (Appendix 7) and these components combined as the geodiversity (Figure 4c) is calculated (Table 2). The correlation between species diversity and geodiversity is 0.44934. The highest correlation between species diversity and a component of geodiversity is 0.61228 and this is the relation between species diversity and geomorphology. All calculated correlation coefficients are positive except for the correlation coefficient for the relation between species diversity and hydrology. This correlation coefficient is negative, -0.15820.

Table 2. Correlation between species diversity, the components of geodiversity and geodiversity.

Layer Species diversity Geodiversity

Species diversity 1.00000 0.44934

Hydrology -0.15820 0.16201

Geomorphology 0.61228 0.76002

Geology 0.45792 0.78991

(10)

10

Land use and species diversity

The number of cells per land use type for each level of species diversity is calculated (Table 3). High and very high species diversity is mainly found in forest and barely found in urban or agricultural areas. Table 3. Number of cells of land use type within the predefined zones of 110 km of species diversity of the bird of paradise.

Land cover types Very low

species diversity Low species diversity Medium species diversity High species diversity Very high species diversity No data 0 2 1 2 0 Shrubs 40 22 13 10 46 Herbaceous vegetation 270 114 170 116 296 Cultivated and managed vegetation/agriculture (cropland) 61 23 45 16 23 Urban / built up 2 6 1 2 6 Bare / sparse vegetation 1 0 2 3 12

Snow and Ice 0 0 0 0 0

Permanent water bodies

77 71 107 45 33

Herbaceous wetland 137 33 79 8 8

All forest types 7052 7099 6622 8490 7262

Open sea 94 87 62 66 20

Discussion

Geodiversity in New Guinea

The GDI with a resolution of 110 km (Figure 4c) shows very high to medium geodiversity is mainly found in the northern, southeastern and northwestern parts of the island, whereas very low geodiversity is mainly found in the southern part and also in the east and west near the coast. The high geodiversity follows the convergent plate margin (Heads, 2002). The oceanic Pacific plate is subducting beneath the continental Australian plate (Figure 1), causing the highly mountainous landscape of the island at the plate margin (Baldwin, Fitzgerald, & Webb, 2012). The New Guinea Highlands and the northern parts of the island are characterized by accumulation of metamorphic rocks as a result of subduction. Furthermore, north of the New Guinea Highlands, a fault zone exists due to the tectonic activity. This fault zone, the Mobile Belt (Figure 1), is characterized by four major ophiolite belts. (Heads, 2002). Another major fault zone is found at the Bird’s Head Block (Baldwin, Fitzgerald, & Webb, 2012). The high geodiversity is mainly found around the fault zones and at the plate margin, in the New Guinea Highlands. The high geodiversity is a result of high tectonic activity, which causes different landforms, such as the high mountains and also a high variety in rocks, such as ophiolites and metamorphosed rocks. The GDI is low in the south, which is a lowland sedimentary basin, thus has not a lot of variety in geological and geomorphological elements.

Yet, the GDI for New Guinea looks different on different resolutions (Figure 4). The GDI is low for a larger area when it is calculated for the recommend resolution of 500 m (Figure 4a). Taking the coarsest legible resolution of 2.5 km (Figure 4b), more areas of high geodiversity become visible. In addition to this, the GDI is calculated for a very coarse resolution of 110 km (Figure 4c), showing even more areas of high geodiversity. This indicates that a coarser resolution, thus with larger zones, will

(11)

11 show a higher GDI. The resolution of the GDI is an important parameter. Conclusions can differ on different spatial scales. A GDI that will be used for policy should be sensitive to changes in policy (Yue et al., 2007). Choosing the right resolution is therefore highly necessary when the GDI will be used for policy or conservation plans. More research on all scales needs to be conducted to assess geodiversity (Pellitero, Manosso, & Serrano, 2015).

In addition to this, further research is necessary to determine how the GDI should be calculated. Geodiversity is a broad concept and is defined differently in many researches. Seijmonsbergen, Guldenaar, & Rijsdijk (2017) describe that there are many quantative methods to assess geodiversity and that there is not a general approach to assess geodiversity yet.

Geodiversity and species diversity

The GDI of 110 km (Figure 4c) is used to calculate the correlation with species diversity (Figure 5). A correlation of 0.44934 is found between species diversity and geodiversity. This indicates a positive, moderate relation between species diversity and geodiversity. Earlier it was found that geodiversity is high in fault zone areas that are characterized by ophiolites. Heads (2001) points out that ophiolite obduction in New Guinea is a major cause of endemism by functioning as a biogeographical break. Ophiolite terranes have large amounts of endemic species of the birds of paradise. These large outcrops of rocks have prevented the dispersal of species.

Species diversity has a positive correlation of 0.45792 with the geological component of geodiversity. Thomas (2012) explains how landscapes with high diversity in forms and materials with different ages often provide high diversity in habitat niches which increases biodiversity. As explained earlier, the birds of paradise are the most diverse and in the New Guinea Highlands and on ophiolite terranes. These terranes exist out of ultrabasic or ultramafic rocks with high concentrations of magnesium and iron. Furthermore, they contain a large amount of different rocks and have been deformed in the past, making it a highly geologically diverse area (Baldwin, Fitzgerald, & Webb, 2012). The New Guinea Highlands are situated at the plate margin. Continental sediments are deformed. Multiple different metamorphic rocks from different periods in time, for example from the Jurassic and Cretaceous, are exposed in this area. In addition to the metamorphic rocks, ophiolites, volcanic and plutonic rocks also appear in this area, contributing to the high geological variety of this area (Baldwin, Fitzgerald, & Webb, 2012).

The highest correlation of species diversity is found in geomorphology. This is a correlation of 0.61228, indicating a strong positive relation between species diversity and geomorphology. This corresponds to the concept of allopatric speciation. Allopatric speciation is based on isolation of population by physical barriers, such as high mountains (Coyne, 1992). Allopatric speciation is seen as the major reason for the high species diversity of the bird of paradise (Irestedt et al., 2009). The isolation, due to the mountainous landscape of New Guinea, caused ecological differences between habitats of the bird of paradise. Different mutations are developing within different populations. Due to the isolations, the populations cannot breed with each other and therefore they can evolve into a new species over time (Turelli, Barton, & Coyne, 2001). Furthermore, previous research also explains that variety in geomorphology is positively related to habitat heterogeneity (Pratt & Beehler, 2015). These high levels of habitat heterogeneity is important for bird richness in New Guinea as they are positively correlated (Marsden & Symes, 2008).

The results show a correlation of -0.15820 between hydrology and species diversity. The hydrological index (Appendix 7) shows that high hydrological variety is mainly found in the southern parts of the island, this is in contrast to where species diversity is found. This indicates that species diversity is barely related to hydrological variety. This could be because the bird of paradise is a montane species that lives in rainforest (Heads, 2001). High precipitation provides abundant water sources in its habitat, however, the bird of paradise has also survived drier periods, as the drought periods are most severe in the mountains (Allen & Bourke, 1997). Therefore, it could be assumed that the bird of paradise is not very dependent on hydrological variety.

(12)

12

Land use and species diversity

The results (Table 3) show that very low to very high species diversity is the most present in forest areas. This is in line with previous research that shows that the bird of paradise is mainly endemic to forests (Heads, 2002). Furthermore, agriculture is situated more in areas of very low species diversity than in areas of high and very high species diversity. This is in line with previous research that states deforestation for agriculture removes habitats. Sodhi et al. (2005) explains that tropical birds living in primary forest are less resilient to habitat changes. If the bird species survives depends on the quality and quantity of the remaining primary forest and their ability to adapt to degraded areas. Primary forest is the most suitable habitat for forest birds (Sodhi et al., 2005). Norder et al. (2020) describe how humans have changed native vegetation into agricultural plots or removed forest for other human activities. This land use change affects biodiversity by reducing and fragmenting habitats. This has led to extinction of species and degradation of ecosystems. Islands are greatly affected by land cover change. More than 60% of the extinct species were endemic to islands (Norder, et al., 2020).

Shearman et al. (2008) did extensive research into the forest cover change over a period of 30 years, from 1972 until 2002. Most species of the bird of paradise occur in the highlands at between 1000 and 2000 m (Heads, 2001). The largest driver of forest change in this area is subsistence agriculture. The high altitudes and ruggedness have prevented forest logging by the timber industry in this area. Between 1972 and 2002, 14% of all forest in the highlands was cleared. The area that is marked as accessible forest is 11% of the total forest in highlands. 14% of this total accessible forest in the highlands has been cleared. Agriculture is intensifying due to population growth. The expection is that more forested areas will be burned down for subsistence agriculture in the highlands where population density is increasing. However, it is expected that the ruggedness is the main protector of the habitat of the bird of paradise. 89% of the total forest in the highlands is inaccessible and thus less suitable for agriculture and logging (Shearman, et al., 2008). Therefore, rugged areas are less likely to be affected by land use change and deforestation (Norder, et al., 2020). This implies that land use change is not yet a large threat to the bird of paradise. Further research should focus on including ruggedness and slope as variables in the geodiversity index for predicting which areas are vulnerable to land use change and where this comes together with high endemic species diversity (Norder, et al., 2020).

Conclusion

The main research question was: how can assessing geodiversity help with maintaining the species diversity of the bird of paradise in New Guinea? Geodiversity is high in the northern, southeastern and northwestern parts of the island, whereas very low geodiversity is mainly found in the southern part and also in the east and west near the coast. High geodiversity has a positive relationship with the species diversity of the bird of paradise in New Guinea. This is mainly caused by the strong positive correlation with geological and geomorphological units. The high geological and geomorphological variety is caused by the tectonic movements of the past and present, resulting in a variation of landforms and rocks. New Guinea’s high geodiverse area is characterized by fault zones, terrane accretement, ophiolites and is highly mountainous. A weak negative correlation is found between species diversity and hydrology.The hydrological variety in this area is very low, indicating that the bird of paradise is not dependent of hydrological variety which could be caused by abundant water supply by high precipitation in the rainforest.

The habitat of the bird of paradise is expected to decrease by agricultural intensification, however, this will not affect the habitat greatly. The bird of paradise is the most diverse in the highlands, which largely protects itself due to its inaccessibility through ruggedness. However, including ruggedness and slope into the geodiversity index could help with predicting which areas are vulnerable to land use change and where this comes together with high endemic species diversity. In addition, more research into the most appropriate mapping scale of the GDI and the approach on how to quantify geodiverity is necessary. The GDI differs for different resolutions. A coarser resolution results in a

(13)

13 higher GDI. Additionally, there are different methods to assess geodiversity which could produce different outcomes as well. Therefore, to use the GDI in policy plans, further research into the appropriate resolution and method is necessary to assess New Guinea’s geodiversity.

The results of this research show that geodiversity has affected the habitat and the species diversity of the bird of paradise. These results can form a basis for further research into the importance of geodiversity for not only maintaing the species diversity of the bird of paradise in New Guinea, but nature’s conservation all over the world. As geodiversity is not implemented in conservation methods yet, these findings show that it is necessary to acknowledge the relationship between biodiversity and geodiversity and that both need to be assessed for future conservation plans.

Acknowledgements

dr. K.F. Rijsdijk is thanked for supervision and support during this research. dr. W.M. de Boer and dr. A.C. Seijmonsbergen are thanked for answering all questions related to GIS. dr. S. Sheppard is thanked for providing missing metadata on the landforms map and dr. M. Heads for providing an inaccessible article.

References

Allen, B. J., & Bourke, R. M. (1997). Report of an assessment of the impacts of frost and drought in Papua New Guinea. Port Moresby: Australian Agency for International Development.

Baldwin, S. L., Fitzgerald, P. G., & Webb, L. E. (2012). Tectonics of the New Guinea region. Annual Review of Earth and Planetary Sciences, 40.

Buchhorn, M., Smets, B., Bertels, L., Lesiv, M., Tsendbazar, N. E., Herold, M., & Fritz, S. (2019). Copernicus Global Land Service: Land Cover 110m: epoch 2015: Globe. Dataset of the global

component of the Copernicus Land Monitoring Service. Retrieved from

https://doi.org/10.5281/zenodo.3243509

Coyne, J. (1992). Genetics and speciation. Nature, 355, 511-515.

Davies, H. (2012). The geology of New Guinea - the cordilleran margin of the Australian continent. Episodes - News magazine of the International Union of Geological Sciences, 35, 87.

Faith, D. P., Margules, C. R., Walker, P. A., Stein, J., & Natera, G. (2000). Practical application of biodiversity surrogates and percentage targets for conservation in Papua New Guinea. Pacific Conservation Biology, 6, 289-303.

Gray, M. (2008). Geodiversity: developing the paradigm. Proceedings of the Geologists' Association, 119, 287-298.

Heads, M. (2001). Birds of paradise (Paradisaeidae) and bowerbirds (Ptilonorhynchidae): regional levels of biodiversity and terrace tectonics in New Guinea. Journal of Zoology, 255, 331-339. Heads, M. (2002). Birds of paradise, vicariance biogeography and terrane tectonics in New Guinea.

Journal of Biogeography, 29, 261-283.

Hengl, T. (2006). Finding the right pixel size. Computers & geosciences, 32, 1283-1298.

Hjort, J., Heikinnen, R. K., & Luoto, M. (2012). Inclusion of explicit measures of geodiversity improve biodiversity models in a boreal landscape. Biodiversity and Conservation, 3487-3506.

Irestedt, M., Jønsson, K. A., Fjeldså, J., Christidis, L., & Ericson, P. G. (2009). An unexpectedly long history of sexual selection in birds-of-paradise. BMC evolutionary biology, 9, 235.

Jenks, G. (1967). The data model concept in statistical mapping. International yearbook of cartography, 7, 186-190.

Marsden, S. J., & Symes, C. T. (2008). Bird richness and composition along an agricultural gradient in New Guinea: The influence of land use, habitat heterogeneity and proximity to intact forest. Austral ecology, 33, 784-793.

(14)

14 Matthews, T. J. (2014). Integrating geoconservation and biodiversity conservation: theoretical foundations and conservation recommendations in a European Union context. Geoheritage, 6, 57-70.

McVicar, T. R., & Bierwirth, P. N. (2001). Rapidly assessing the 1997 drought in Papua New Guinea using composite AVHRR imagery. International Journal of Remote Sensing, 22, 2109-2128.

Natural Earth. (2009). World Coastline (1:10 million). Retrieved from

https://www.naturalearthdata.com/downloads/10m-physical-vectors/10m-coastline/

Norder, S. J., de Lima, R. F., de Nascimento, L., Lim, J. Y., Fernández-Palacios, J. M., Romeiras, M. M., Elias, R. B., Catarino, L., Ceríaco, L.M.P., Castilla-Beltrán, A., Gabriel, R., Menezes de Sequeira, M., Rijsdijk, K.F., Nogué, S., Kissling, W.D., van Loon, E. E., Hall, M., Matos. M., Borges, P. A. (2020). Global change in microcosms: Environmental and societal predictors of land cover change on the Atlantic Ocean Islands. Anthropocene, 100242.

Pellitero, R., Manosso, F. C., & Serrano, E. (2015). Mid‐and large‐scale geodiversity calculation in fuentes carrionas (nw spain) and serra do cadeado (paraná, brazil): methodology and application for land management. Geografiska Annaler: Series A, Physical Geography, 97, 219-235.

Pratt, T. K., & Beehler, B. M. (2015). Birds of New Guinea (Second Edition). NJ: USA: Princeton University Press.

Saxon, E., & Sheppard, S. (2010). Land Systems of Indonesia and New Guinea. Retrieved from Data Basin:

https://databasin.org/datasets/eb74fe29b6fb49d0a6831498b0121c99?id=eb74fe29b6fb49d0a6 831498b0121c99

Schrodt, F., Bailey, J. J., Kissling, W. D., Rijsdijk, K. F., Seijmonsbergen, A. C., Van Ree, D., Hjort, J., Lawley, R.S., Williams, C.N., Anderson, M.G., Beier, P., Van Beukering, P., Boyd, D.S., Brilha, J., Carcavilla, L., Dahlin, K.M., Gill, J.C., Gordon, J.E., Gray, M., Grundy, M., Lawler, J.J., Monge-Ganuzas, M., Royse, K.R., Stewart, I., Record, S., Turner, W., Zarnetske, Field, R. (2019). Opinion: To advance sustainable stewardship, we must document not only biodiversity but geodiversity. Proceedings of the National Academy of Sciences, 116, 16155-16158.

Seijmonsbergen, A. C., Guldenaar, J., & Rijsdijk, K. F. (2017). Exploring Hawaiian long-term insular geodiversity dynamics. Landform Analysis, 35. Landform Analysis, 35.

Shearman, P., Bryan, J., Ash, J., Hunnam, P., Mackey, B., & Lokes, B. (2008). The state of the forests of Papua New Guinea: mapping the extent and condition of forest cover and measuring the drivers of forest change in the period 1972-2002. Port Moresby: University of Papua New Guinea.

Sodhi, N. S., Koh, L. P., Prawiradilaga, D. M., Tinulele, I., Putra, D. D., & Tan, T. T. (2005). Land use and conservation value for forest birds in Central Sulawesi (Indonesia). Biological Conservation, 122, 547-558.

Steinshouer, D. W., Qiang, J., McCabe, P. J., & Ryder, R. T. (1999). Maps showing geology, oil and gas fields and geologic provinces of the Asia Pacific Region. The Survey: US Dept. of the Interior.

Swingland, I. (2001). Biodiversity, definition of. Encyclopedia of biodiversity, 1, 377-391.

Thomas, M. (2012). A geomorphological approach to geodiversity-its applications to geoconservation and geotourism. Quaestiones geographicae, 31, 81-89.

Turelli, M., Barton, N. H., & Coyne, J. A. (2001). Theory and speciation. Trends in ecology & evolution, 16, 330-343.

Yue, T. X., Ma, S. N., Wu, S. X., & Zhan, J. Y. (2007). Comparative analyses of the scaling diversity index and its applicability. International Journal of Remote Sensing, 28, 1611-1623.

(15)

15

Appendices

Appendix 1

Figure A1. The geodiversity index of New Guinea calculated from hydrology, geomorphology and geology

(16)

16

Appendix 2

indices calculated in grid cells of 2500 m and visualized in five classes varying from very low to very high.

Appendix 3

(17)

17

Appendix 4

Figure A4. Land cover map after reclassifying all forest types into one class for forest (Buchhorn, et al., 2019).

Appendix 5

Figure A5. Geological variety of New Guinea calculated in grid cells of 110 km by 110 km.

Appendix 6

(18)

18 Figure A6. Geomorphological variety in New Guinea calculated in grid cells of 110 km by 110 km.

An assessment of the role of geodiversity in species diversity of the bird of paradise in New Guinea