A GIS-Based Assessment of Land Cover Changes and Tree Species Suitability for Urban Greening Policies in the City of Kitwe, Zambia

2.2.1.

Remote sensing and GIS methods

The study used Landsat images of Kitwe for two distinct years (1990 and 2015).

To undertake preliminary treatment of the images (see Figure 2) and processing, the study used EDRAS IMAGINE 2014 while the mapping was done using ArcGIS 10.4.

The Landsat images obtained had the characteristics outlined in Table 1 and were examined for the appropriate bands required for the project.

2.2.2.

Urban tree species assessment

Once the magnitude and direction of the land cover changes in the district were determined, the study focused on determining which vegetation tree species were commonly found within the town. This was done by determining the number of tree species present within the urban part of Kitwe. The initial community analysis was carried out during the rainy season to capture the peak growth of trees. This was carried out between October 2017 and December 2018.

2.2.3.

Data analysis

The information collected was analysed through the identification of common patterns, equations, comparisons of primary findings, and the statistical package SPSS. Interpretation of results and attempts to rationalise or understand the meaning of these figures and/or numbers were also considered as follows:

2.2.4.

Determining above-ground biomass (AGB)

2.2.5.

Statistical assessment

The study used biochemical data to measure urban tree species’ adaptability and test the hypothesis (H₀).

3.1.1.

Land cover classification

According to the study findings, the area is primarily defined by four major classifications (see Figures 4 and 5). Between 1990 and 2015, the built-up area’s LC rose dramatically while bare soil and vegetation decreased. The change vector images (Figure 5) produced from the two study periods permitted verification that the deforested area in Kitwe was 28,140 ha between 1990 and 2015. The NDVI index (see Figure 6) for this impacted area revealed a 3.45% rise in low vegetation for grasslands and agricultural fields, from 22,918 ha (1990) to 23,709.87 ha (2015). While the indigenous forest cover decreased by 24.93%, from 14,458.8 ha in 1990 to 10,853.01 ha in 2015, the NDVI index also decreased. This coincided closely with the Copperbelt Province’s annual average deforestation rate of 0.84%, according to the Forestry Department (2016).

Water resources decreased by 31%, from 13,619.53 ha in 1990 to 9,393 ha in 2015, which could be linked to changes in rainfall patterns and increased water abstraction [56]. This loss could also be a direct result of expanded, developed areas and the destruction of crucial recharge sites in this district as a result of mining and agriculture [57]. Figure 4 shows a decrease in bare soil, which could be a signal to increase concrete, or Figure 7 shows a continuum ranging from high vegetation conditions to exposed soil conditions in 1990 and 2015.

3.1.2.

Direction and magnitude of change in Kitwe

Figure 8 depicts the magnitude of the changes, which were relatively low within 92.94% of the district (see Table 6), demonstrating that while changes were occurring, the majority of them occurred within existing and much older communities rather than in new ones. According to the study, 6.36% of the district underwent medium-level alterations, particularly in the Northeast in regions such as Kafue and Kamfinsa, as well as parts of the CBD. The growth in built-up area is primarily owing to the city’s expansion, and it is likely to continue towards the northeast as more land parcels are approved for construction.

The results suggest that 60% of Kitwe’s LC is degraded (see Table 7), with the remaining 40% showing no change (see Figure 9) and covering the main CBD of Kitwe as well as mine regions. Persistence may be a trend related to the lengthy existence of urban areas and the urban activities that take place in these regions.

Other than normal urban activities, most mining-related land use increased significantly in the 1970s and early 1980s during the copper industry’s boom years [58] and continue to influence land use in Kitwe. High alterations have been observed (Figure 9) in the Mindolo and Twatasha areas as a result of heavy mining activities, with much of the region being used as garbage disposal sites or tailings dams. When compared to the opportunity cost of using the land for other purposes such as agriculture and recreation, this is a loss of productive land. With the majority of the land transformed into a waste storage area, dangers from the loss of isolated, unique microhabitats within an otherwise undisturbed habitat have been discovered [59] to cause the local extinction of certain plant and animal species.

3.1.3.

Relative abundance of urban trees in Kitwe

As expected, the majority of Kitwe’s built-up regions are composed of evergreen species rather than deciduous plants, and their continuous presence in the town is extremely likely to be the result of intentional introduction. Both introduced and indigenous species predominate. A post-hoc analysis using the Bonferroni correction found that there were no significant differences between the species groups in any of the urban tree species categories. This supports the theory that, because this is part of Kitwe’s urban tree inventory, it could have come from secondary urban forest programmes that happened at varied times following the city’s expansion. The top-ranking species identified (see Figure 10) as the most abundant species within the urban built-up region were Delonix regia (15.9%) and Mangifera indica (12.3%). More intriguingly, despite the fact that it was assumed that largely built-up areas would have introduced species rather than indigenous species, this was not the case. The more indigenous trees were less numerous within the built-up natural environment due to extensive forest loss caused by clearance during urbanisation.

Despite the city having lost most of its indigenous green spaces, which have been converted into residential or commercial sectors over time, more indigenous species are regularly scattered about the city in small spaces. Albizia antunesiana, Albizia versicolor, Baphia bequaertti, Branchiostegal boehmii, Brachystegia utilis, Isoberlinia angolensis, Julbernardia paniculata, Pericopsis angolensis, Uapaca kirkiana, and Uapaca nitida are among the indigenous species identified in Kitwe City. These findings are consistent with local authority reports [38], which reveal that many of the current green spaces still support the identified species as well as many additional indigenous species.

In comparison to other towns like Mufulira, Kalulushi, Ndola, and Lusaka, these species are located along comparable sidewalks, roads, and avenues surrounding the city’s main metropolitan areas, forming homogenous habitats or green spaces. This analysis indicated that the prevalent selection procedures were derived from Commonwealth town planning systems, which were modelled after British practise and legislation. The study refers to research by Home [60], who confirms that the roots of these introduced species can be traced back to pre-colonial Northern Rhodesian times (now known as Zambia), when these areas were still being planned for as townships. Home [60] goes on to explain Lusaka’s reputation as a garden city, referring to “Dutton’s initiative” that constructed a chain of nurseries and planted various trees in Lusaka and sections of the Copperbelt, including the ones in this article.

3.1.4.

Adaptive capacity of Kitwe’s urban trees

After determining the most abundant species, the study analysed the plants’ ability to cope with environmental conditions by measuring plant sensitivity to their surroundings and carbon sequestration. The tree’s sensitivity was characterised in connection with its exposure to its environment and air pollution, which was very likely to occur in built-up areas. The focus of adaptive capacity was the ability of species groupings to respond to specific sorts of hazards—in this example, drought, plant health, and biochemical air pollution. These responses assist the urban trees in regenerating and/or adapting in order to reduce susceptibility and increase reaction time. According to Flórez Bossio et al. [61], urban adaptive capacity traits can be viewed as a collection of variables that allow urban trees to adapt. After all, the resilience of urban trees and their socioeconomic value influence their abundance as well as the health of any urban forest.

The laboratory test findings revealed biochemical differences (p-value 0.05) in the means of pH, RWC, AA, and total chlorophyll content (TChl) among the three urban tree species groups (Table 8). There are significant comparison pairings in the post-hoc comparisons using the SPSS Bonferroni correction. Trees are more flexible due to their drought resilience and water efficiency. The study looked at how different tree species react to air pollution. This research is founded on three fundamental assumptions. As high pH increases stomatal sensitivity, urban trees are more resistant to air pollution. Urban tree species with high RWC maintain physiological balance and improve drought resilience. Ascorbic acid is required for cell wall growth, defence, and cell division in trees, in addition to photosynthesis for carbon dioxide fixation and antioxidant defence in urban plants. As TChl decreases with pollutant levels, urban trees with high TChl can survive air pollution and are healthier.

The overall pH ranged between 4.67 and 6.80, with Bauhinia variegate and Toona ciliate having the highest pH. As a result, urban trees in Kitwe’s built-up area have better tree health and are more resistant to diseases caused by stomatal sensitivity as a result of air pollution. Compared to Pinus oocarpa, Jacaranda mimosifolia, and Eucalyptus grandis, which had the lowest pH and were below the supposed ideal pH (between 6.0 and 7.0), at pH levels above 6.4, urban trees are vulnerable to insect attacks, while at pH levels below 6.4, trees are vulnerable to specific diseases.

Table 8 shows that the RWC was regularly distributed, with a mean of 65.39% (SD = 1.17). RWC values for urban tree species in Kitwe were in the higher range, i.e., between 50% and 90%. This was greatest in Schefflera actinophylla and Bauhinia variegate and lowest in Jacaranda mimosifolia. Since urban trees can improve water resilience in cities, the majority of Kitwe urban tree species identified have sufficient RWC to maintain tree physiological balance and improve other key sensitivity aspects such as drought tolerance. Drought tolerance indices [62] can be used to learn more about how environmental conditions influence this.

There was a statistically significant difference in the means of TChl between the various urban tree species groups. When compared to other species, Toona ciliate and Jacaranda mimosifolia have the highest TChl, while Delonix regia and Bauhinia variegate have the lowest. The presence of a high concentration of chlorophyll in these urban trees boosts their tolerance to air pollution and serves as an indirect indication of urban tree health. It should be emphasised, however, that the TChl declines as pollution levels rise. Significant comparison pairings were discovered using post-hoc comparisons with the SPSS Bonferroni correction.

Pinus oocarpa had the highest ascorbic acid concentration, while Plumeria rubra had the lowest (Table 8). These urban tree species with a high ascorbic acid content are more suitable because ascorbic acid is required for cell wall synthesis, defence, and cell division in trees, is required for photosynthesis for carbon fixation, and acts as an antioxidant, increasing urban tree resistance to their natural environment.

Vulnerability, as it relates to the socioeconomic elements of the tree species, covered numerous aspects such as the aesthetic value of trees as well as the minimal demand for care and maintenance, which is the major criterion for use, particularly in urban management [63]. Take, for example, Delonix regia, the most prominent species in the research area, which is known for its dense clusters of orange-red flowers that bloom in early summer and have a vibrant hue that makes it appealing. Furthermore, the species was discovered to be native to Madagascar and Zambia and has been introduced in many other countries, becoming widespread in most tropical and subtropical areas of the world [64]. Another popular species is Jacaranda mimosifolia, which is known for its clusters of beautiful bell-shaped, blue-violet flowers that provide both beauty and shade.

The observations in Kitwe reveal that natural indigenous plants are subject to challenges from invasive species that are known to enter forest gaps, plantations, roadsides, and riparian zones (banks of watercourses) (ibid.). These have a large impact on Kitwe’s urban forest inventory because they are abundant in the built-up areas. CAB International classifies Jacaranda mimosifolia as an invasive species because this urban tree, which is native to South America, is highly competitive and does not allow other urban trees to grow beneath it [63]. T. ciliate and S. campanulate are two other instances. While both are invasive species, it was discovered that T. ciliate is a more aggressive invasive species than S. campanulate. Furthermore, S. campanulate is a more popular ornamental tree than T. ciliate, which has no attractive flowery attributes but is widespread due to its indigenous status [63, 65]. Other species, such as Mangifera indica, are abundant because of their fruit-bearing ability and ability to provide shade for residents [66]. As a result, these evergreen trees are more sheltered and grow well in locations with a short dry season.

Kitwe city land use dynamics posed an additional threat to the city’s urban forest’s ability to survive change. According to Phiri et al. [55], while examining trends in spatial distribution of such urban green areas improves understanding of forest management, this is not always the case. Kitwe is distinguished by complex terrain changes induced by natural and man-made forces interacting with one another. Threats such as a lack of soil naturing programmes to enhance soil quality, on the other hand, continue to shorten the life span of trees in what were formerly green regions with degraded soils [67]. This became an opportunity and the primary focus of urban forest management.

3.1.5.

AGB of urban trees

Kitwe’s urban forest had the highest WD (0.662 g cm⁻³) among Eucalyptus grandis species and the lowest (0.220 g cm⁻³) among Spathodea campanulata species, as shown in Table 10. The variation in WD results could be best described by Deng et al. [70], who show that such results can be greatly influenced by site quality, relative heights, tree age, and social class of the forest inventory. These signals must be used with caution, as there may be minor differences in findings between places. The results also show that the largest contributors to carbon sequestration are Delonix regia and Toona ciliate (see Table 11), both of which are the most abundant species in town when compared to the other nine species. Senna siamea (Cassia siamea) came in third, and Jacaranda mimosifolia came in fourth, respectively. The majority of the urban trees discovered contribute to CO₂ absorption at the city level. However, it was discovered that species with high tolerance, such as Schefflera actinophylla, contribute less to carbon capture.

The fact that variance is modest (see Figure 12) may be attributed to the species’ ability to sequester carbon as well as the species’ abundance in urban areas. The ability of urban forests to sequester carbon is determined by various aspects, including (1) the age of the tree, (2) its size, (3) environmental considerations, and (4) policy and methods of generating deliberate enabling elements that enable such greening programmes. These considerations stem from the observation that there are considerable changes in carbon sequestration results between tree species (see Table 10) due to the difference between young and older urban trees. Anwar [68] verified this, arguing that older trees have lived longer and have had more time to store carbon than younger urban forests. According to the findings of this study, the two commercial tree species are young urban trees with extremely low biomass (Table 10). This has a substantial impact on each tree’s biochemical characteristics. For example, TChl varies over the tree’s life cycle. Table 8 displays how TChl changes depending on species, leaf age, pollution level, and other biotic and abiotic variables [71]. Certain pollutants increase TChl, while others decrease it [72].

3.1.6.

Suitability trees for urban greening

Disaggregating all of the findings from APTI, API, and AGB reveals (see Table 11 and Figures 13–15) that there are disparities between the species groups, with the data being closely spread, showing a high level of similarity, and the rest of the results being scattered between groups. These could be the result of exposure to various environmental factors and the influence of certain functional groups.

In all comparisons, the overall associations were statistically significant (Figures 13–15). Three species, B. variegata, T. ciliata, and P. oocarpa, were identified as the best-performing urban trees for greening programmes that prioritise tolerance and performance (Figure 13). The overall design of such greening programmes would consider the quality and purpose of development programmes that include the vulnerability of urban forests, including climate variability and adaptive characteristics. This necessitates a shift in focus on how to design robust and improved city-focused frameworks that deal with air pollution and new emerging threats [58, 73]. This should provide guidance to urban planning experts on taking carbon stocks seriously [74, 75]. There is no doubt that the best-ranking tree species should be resistant to air pollution while also contributing to carbon capture within Kitwe.

When AGB and APTI (Figure 14) are compared, it is clear that while most species had low AGB, the species varied in terms of their tolerance to air pollution, implying a negative association. A statistically significant linear association (r = −0.195, p < 0.05) exists between an increase in AGB and a drop in APTI. DBH was found to be a somewhat better predictor of APTI than AGB. Despite its weakness, the negative relationship can be attributed to the area’s environmental conditions, and it also suggests that growth parameters are reduced in highly polluted areas compared to those in low air pollution areas (see Figure 14). The findings of AGB and API (Figure 15) are comparable to the combination of APTI and AGB. The amount of biomass accumulated by high-performing urban trees is not the only factor influencing their performance.

After integrating each assessment evaluation, each species was evaluated from best (1) to worst (10) (see Table 11). The most suited trees were B. variegate and T. ciliate, which were placed first and second, respectively. The highest-ranking tree species are resistant to air pollution while also contributing to carbon capture within Kitwe. However, it was discovered that species with high tolerance, such as S. actinophylla, contribute less to carbon capture.

The findings for APTI, API, and carbon sequestration are consistent with those of Jim [67] and Bowler et al. [76], who advocated a more empirical selection criterion for Zambian urban planners that considers factors other than aesthetics. Urban trees with high APTI and API ratings are proposed as species for the area’s urban greening programmes. Because they are based on parameters with significant biological and socioeconomic characteristics, APTI and API can be used.

To address the vulnerability dynamics of urban forests, urban planners should establish socioeconomic scenarios and follow and trace development trends and routes while designing adaptation solutions. Understanding urban trees’ sensitivity to air pollution, socioeconomic performance, and capacity for carbon sequestration will help in achieving these goals the best. Having high APTI and API values, as well as high carbon sequestration, should be promoted for new development areas employing properly developed urban greening programmes and giving specific adaptive capacity ranges [61]. These tree species could be incorporated into the design of an urban greening programme to help with long-term air pollution planning.

Kitwe’s main metropolitan areas have made considerable contributions to reducing air pollution and boosting carbon absorption, thereby increasing the city’s carbon stocks (see Table 10). All of the urban trees evaluated in Kitwe perform admirably in their surroundings, and there are substantial variances in tree species APTI, adaptive capacity metrics, and cumulative biomass. Gmelina arborea, Toona ciliata, and the two commercial trees are the best performing trees for high-pollution areas because they can tolerate the significant pollution from emissions. The best five species identified using rank analysis were Bauhinia variegate, Toona ciliate, Gmelina arborea, Eucalyptus grandis, and Delonix regia. Tables 9 and 10 indicate that each index revealed that greening programmes could combine species during the design or planning stage to achieve specific programme objectives. As a result, the decision should always be based on the purpose, location, and area development requirements.

Depending on the objective goals of the greening programme, one can compare one tree species to another to find the most effective combination that meets developmental needs that prioritise the climate variability of urban tree species as well as air pollution tolerance. The combination in Figure 15 allows urban planners and managers to assess trees suited for greening programmes aimed at improving air quality, providing shade, and improving aesthetics in low-pollution green spaces, houses, and sidewalks.

The application of APTI, API, and carbon sequestration in green infrastructure design gives planners a notion of which tree species might ameliorate air pollution and provide effective ecosystem services for urban greening programmes. The lack of air pollution data to establish a pollution gradient to compare our data between highly contaminated and less polluted areas is one of the study’s limitations. Furthermore, there was insufficient data on urban tree care services near the trees sampled to provide a comprehensive methodological approach to forest management. According to Nayak et al. [77], growth characteristics are lowered in highly polluted areas compared to species found in low-polluted areas. This agrees with Escobedo et al. [78], who state that urban greening strategies suggest that planners develop networks of areas to better harness their socio-ecological components in a way that can conserve ecosystems while providing essential benefits to a given society.