
Mapping unequal accessibility to habitat quality
A spatial analysis approach in the capital city of Colombia
By Liliana Perez and Yenny Cuellar

01 / 08
Bogota's urban wetlands
Found in and around cities or their suburbs (Ramsar 2018), urban wetlands are unique and play a fundamental role in the well-being of the metropolitan area.

Bogota and some of its wetlands problems
These ecosystems offer multiple ecosystem services such as water supply, water purification, diminishing the urban heat island effect, providing habitat to critical plant and animal species, flood regulation, and recreational opportunities (Millenium Ecosystem Assessment 2005).
However, the 2005 Millennium Ecosystem Assessment (MEA) reported that over 80% of wetlands near urban areas have disappeared.

This trend is not far from the historical changes of urban wetlands in , Colombia, which have been affected by anthropic actions such as urban expansion, sewage disposal, illegal construction waste dumping, and grazing activities.
La Vaca wetland before its restoration.
Being Latin America and the Caribbean amongst the most urbanized regions in the world, during the 60s and 70s of the 20th century, Colombia's urban growth was fueled mainly by migration from the countryside to the city due to violence (Universidad Externado de Colombia 2007). Since then, the city has experienced substantial growth and spread westward.

People illegally inhabit the La Vaca wetland.
It is estimated that at the beginning of the 20 th century, the area occupied by lakes and wetlands was more than 50.000 hectares. However, around 700 hectares remain today being 17 ecosystems recognized by the city's Environment Secretariat, of which 11 are under the Ramsar convention.
Purpose
Bearing in mind that these transition ecosystems provide ecosystem services (ES) such as wastewater treatment, water cycle, and temperature regulation, carbon dioxide reduction, recreation, and ecotourism, which are beneficial to society, assessing the accessibility and equal distribution of them is important for an improved urban planning and development.
Therefore, our study uses the definition of inequality as a positive description of the variance in the distribution and access to an ES.

Methodology
In this study, we estimated the habitat quality and analyzed its response to land cover changes using the Integrated Valuation of Ecosystem Services and Tradeoffs (InVEST) model.
Additionally, a spatial analysis was applied to determine if there is a relationship between the city’s population socioeconomic profile and the access to the ES evaluated.
What is habitat quality?
Refers to the ability of the environment to provide suitable conditions for biological survival and development (Ding et al., 2021).
InVEST model
We used the InVEST Habitat Quality (HQ) model to map and estimate the habitat quality in the urban wetlands of the study area.
This spatially explicit tool, created by the Natural Capital project, allows mapping, quantifying, and valuing ESs with 16 models, including the Habitat Quality model.
The model represents the biodiversity of a landscape, estimating the magnitude of habitat types and their state of degradation (Natural Capital Project). The HQ model estimates HQ and Habitat Degradation (HD) maps from the combination of LUCC maps and threats to biodiversity. HQ ranges between 0 and 1, indicating one of the habitats with the highest suitability.
Spatial Correlation and Regression Analysis
We applied a bivariate spatial autocorrelation method to analyze the spatially coupled correlation between the Habitat Quality and the Quality of Life Index. For this, we used the Multivariate LISA tool of GeoDa 1.20.0.22.
Further, to quantitatively analyze the influence of socioeconomic factors on habitat quality, we applied the spatial regression model.
Data
Two input land use land cover maps for 2010 (left), and 2022 (right) were used.
Besides the LUCC maps, the model requires threats data, the maximum distance over which each threat affects the habitats, the relative impact weight of each threat, the type of decay over space of each threat, and the sensitivity of each land use class to the threats. The previous parameters were determined from a literature review and expert knowledge (Jiang et al., 2012).

Threat types
Also, we used some variables of the Multipurpose survey for 2010 and 2022 to calculate the quality of life index.
Variables used to calculate the quality of life index.
Results
The habitat quality (HQ) results show different spatial patterns in the study area. Visually Habitat degradation is higher where anthropogenic concentration is highest. On the contrary, HQ values are higher in the peripheral wetlands of the study area (, , , the northwestern region of , and ).

Habitat quality trend for each wetland in 2010 and 2022.
In 2020, two projects caused deep concern among the population: an elevated concrete bridge in the and a series of hard paths for pedestrians and bicycles in the and wetlands.
Wetlands in the south are highly affected as a consequence of urbanization. An example is the wetland which continues to be affected. The passage of roads such as Avenida Ciudad de Cali divides the wetland in two, transforming the wetland and letting the ecosystem with only 0.2 ha of water. Despite the efforts of the mayor's office to recover the wetland, wastewater discharges continue to be its biggest problem.
Spatial Autocorrelation Analysis
To analyze the spatial autocorrelation analysis, we applied the bivariate Moran’s I index for the habitat quality (HQ) in 2010 (left) and 2022 (right) with the quality of life index (QLI). The maps represent the spatial autocorrelation with p-values less than 0.05.
The high values for both variables are recorded in the northern of Suba and Usaquen, western of Fontibon, and Tunjuelito for 2010. But for 2022, the high-high pattern changes, being in west Fontibon and Barrios Unidos.
Low values for both variables in both years are identified in the Kennedy, for 2010 concentrated in the west of the locality, while in 2022, it was almost all its extension.
Spatial Regression Analysis (SLM)
To explore the spatial dependence of the habitat quality and the socioeconomic variables, we carried out a spatial regression analysis.
Applying the spatial lag model for 2021 results shows that the only significant variables are connectivity and security. Therefore, there is a negative spatial relationship between connectivity and HQ and a positive one with security.
Regression results of the SLM
The results showed that connectivity is the dominant socioeconomic variable in decline in HQ of the wetlands in Bogota with the most significant magnitude effect. Furthermore, as the connectivity increases during the study time, this shows that the HQ of the wetlands would decline because of the expansion of the road network and the acquisition of more cars, which are aspects related to the connectivity index.
Example of road projects in Bogota.
- Juan Amarillo elevated bridge "green corridor".
Elevated concrete bridge in Juan Amarillo wetland.
References
Dirección Nacional de Estadística. (2021). Encuesta Multipropósito-Bogotá.
Millenium Ecosystem Assessment. (2005). Ecosystems and human well-being: wetlands and water Synthesis. World Resources Institute. Washington, DC.
Ramsar. (2018). Urban wetlands: prized land, not wasteland.
Universidad Externado de Colombia. (2007). Ciudad, espacio y población: el proceso de urbanización en Colombia.