Biodiversity and Conservation

, Volume 23, Issue 4, pp 963–978

Urban sprawl and species movement may decimate natural plant diversity in an Afro-tropical city

Authors

    • Faculty of Forestry and Nature ConservationSokoine University of Agriculture
  • Abubakari Said
    • Faculty of Forestry and Nature ConservationSokoine University of Agriculture
  • Kuruthumu A. Mwamende
    • Range Management Section, Department of Animal Science and ProductionSokoine University of Agriculture
  • Shombe N. Hassan
    • Faculty of Forestry and Nature ConservationSokoine University of Agriculture
  • Seif S. Madoffe
    • Department of Forest Biology
Article

DOI: 10.1007/s10531-014-0646-1

Cite this article as:
Rija, A.A., Said, A., Mwamende, K.A. et al. Biodivers Conserv (2014) 23: 963. doi:10.1007/s10531-014-0646-1
  • 306 Views

Abstract

Although urban sprawl is increasingly becoming a worldwide problem, the effects associated with urbanization on local biodiversity particularly in the developing countries, is still poorly understood. We investigated plant species richness along urban–rural gradients in an Afro-tropical metropolitan Morogoro city in Tanzania, and examined patterns of species movement by humans within and outside the city and its potential impact on habitat homogenization of urban ecosystems. Biodiversity information was assessed in 120 sample plots (25 × 25 m) distributed in core urban, sub-urban and peri-urban (rural) zones, while patterns of plant resource use and species movement were surveyed from 120 households in the study area. We found the highest tree species richness in the urban core (82.6 ± 1.4 species) whereas tree density decreased from peri-urban (465.60 ± 152.03 tree/ha) to urban core (244.00 ± 120.86 tree/ha) and species diversity decreased from urban core (α = 2.82 ± 0.01) to peri-urban area (α = 2.61 ± 0.23). Further, the proportion of exotic species was significantly higher (75.6 %) than that of native species in the study area (p = 0.001). Tree community assemblages showed least differences among the study zones (dissimilarity = 10 %) and appear to have been influenced by local cultural differences, species movement as well as local resource use. Furthermore, species movement exhibited a random and multi-directional pattern perpetuated by human and nonhuman factors. Movements were significantly higher between backyard gardens within the city than from outside. Alien species by far outnumbered native plant species moving across urban landscapes, potentially increasing species homogenization. Conservation awareness is needed to avoid habitat homogenization due to problems associated with urban sprawl and to maintain heterogeneous habitats for the urban wildlife.

Keywords

BiodiversitySpecies introductionSpecies movementSpecies richnessUrbanizationUrban–rural gradientsMorogoro-Tanzania

Introduction

There is now plenty of evidence that urbanization is accelerating loss of biodiversity. McKinney (2008) reviewed various aspects of biodiversity in relation to urbanization across the globe. This review revealed that despite urbanization being a global problem, many studies have mainly focused on the developed world (McKinney 2008) and very little has been done in developing countries (Savard et al. 2000; McKinney 2002). Even in the context of developing countries, disparities do exist due to disproportionately fewer publications in the Afro-tropical regions than in the Oriental and Neotropical regions. Studies conducted in the Neotropical region (Pauchard et al. 2006; Ortega-Álvarez et al. 2011) show patterns of urbanization effects on biodiversity being consistent with observations made in cities of the developed world (Kuhn et al. 2004; McDonnell and Hahs 2008; McKinney 2008). However, such results cannot be generalized for all the developing countries due to local and regional differences in the level of development, residents’ culture, behaviour and socio-economic conditions among the inhabitants of different countries (Hope et al. 2003; McKinney 2008). Existing local conditions caused by these factors have a direct bearing on habitat deforestation, urban planning and management as well as on species introductions in urban landscapes (Dow 2000; Hope et al. 2003; Breuste 2004; Kim and Pauleit 2005) with consequences for the biodiversity of many countries world-wide (McKinney 2008). Many studies have explored the broad effects of urbanization on different animal and plant taxa and report that species richness generally increases at mild disturbance levels (McKinney 2002; Hope et al. 2003; Tait et al. 2005) although for some taxa such as plants results may differ (Burton et al. 2005; Kim and Pauleit 2005; Pautasso 2007; McKinney 2008); suggesting that urbanization effects on biodiversity is influenced by several other factors operating at local spatial scales (von der Lippe and Kowarik 2008; Bongers et al. 2009).

Urban Africa is currently experiencing rapid development due to the increasingly high influx of people from rural areas in search of jobs and better living conditions (Cohen 2006).This has caused the natural biota of most peri-urban areas to change due to land clearance for agricultural farms and settlements. There is also the introduction of non-native plant species accompanied with urban sprawl, which may lead to homogenization (Mack and Lonsdale 2001; Kowarik 2003). Nevertheless, there is often a net increase in species richness due to the addition of non-native species (McKinney 2002) particularly for species that require relatively small area to maintain stable and viable population such as plants (McKinney 2008).

Like many other cities, Morogoro municipality in Tanzania is experiencing urban sprawl with potential impacts on local biodiversity due to its location at the foot of Uluguru Mountains, part of the Eastern Arc Mountains biodiversity hotspot (Myers et al. 2000). There is an increasing wave of local immigrants from rural areas as well as from other regions across the country as the city is strategically located at the main road connecting other regions of Tanzania and bordering countries. While this situation poses high risks of biodiversity, there is yet no information about the effects of urbanization on biodiversity either for this city, or of any in Tanzania, to guide city planning and sustainable biodiversity conservation. The aims of this study were to, (i) describe the tree structure, composition and species diversity along a gradient of urbanization in Morogoro municipality, (ii) understand patterns of tree community structure between urbanization zones (along the gradient) and (iii) assess local use of tree species and species movement patterns within and outside the city. We predicted that tree density would increase from urban core to peri-urban due to fewer open spaces available in urban cores, and more open spaces in the peri-urban that could be filled by introduced plant species. Also species richness should be lowest at urban core where disturbance levels are high. Further, we expected higher richness of non-native tree species in the urban core than in the peri-urban area due to introduction of non-native species for shade and other uses such as poles and wind break. This study provides new insights into the effects on local biodiversity associated with urbanization in Afro-tropical context and contributes to the growing scientific literature in urban ecology.

Materials and methods

Study area and study design

Morogoro is located at about 200 km west of Dar es Salaam city between latitudes 5°00′ and 7°40′S of the equator and longitudes 37º10′ and 38º33′E of Greenwich Meridian. The municipality has a population of approximately 600,000 people (Mayor- Morogoro municipality pers. com, 2012) in an area of more than 65 km2 at the foot of the Uluguru Mountains. The annual rainfall ranges from 600 to 1,000 mm with bimodal pattern characterized by short rains from November to January and long rains during March to May. The mean monthly temperature varies between 21 and 27 °C. The predominant vegetation cover is miombo woodland (Morrison and Lind 1974). However, in the recent decades the natural vegetation type has been cleared and replaced by a spatially complex mix of plant communities owing to the urbanization process (Rija 2003). Further, based on the development plan available for this municipality, three distinct zones were identified (herewith named urban zones) namely urban core (cU), sub-urban (sU) and peri-urban (pU) corresponding to the human population density—‘high density’, ‘medium density’ and ‘low density area’ categories being used by Morogoro Municipal Council when allocating land properties to the urban residents. In this study, these urban zones were used to study composition and diversity, tree community structure, local use of plant resources as well as species movements from outside the city and within along this gradient of urbanization from the center of the city.

Sampling of trees, local use of plant biodiversity and species introduction and movement

Survey plots (25 × 25 m) spaced at least 150 m apart representing independent sampling units were used to record tree data between January and April 2012. Forty plots were sampled within each study zone for a total of 120 plots. In each plot all trees and shrubs available were documented in terms of number per species and categorized as exotic or native. Presence of vegetables growing along with trees in the sampling plots was also recorded: results will be reported in a separate paper (Rija et al. in prep). To collect data on local use of biological resources in the area and the sources of non-native plant species available in the household backyards, a questionnaire was distributed to one household chosen randomly in close proximity to each surveyed tree plot. A total of 40 questionnaires were used in each zone: urban core, sub-urban and peri-urban, making a total of 120 questionnaires. The questionnaire survey included questions on the type of plant species available in the back yard gardens, the sources of plant species that are considered not native to the area, the uses of the plant species, and their knowledge of spreading of such plant species from one garden to another within and outside Morogoro city. Plant identification was done using a field guide book for East African plant species (Dharani 2002). For plant species whose names were not easily identified in the field, voucher specimens were collected and identified using specimens stored in the herbarium and some were identified with the help of a botanist at the Faculty of Forestry and Nature Conservation, Sokoine University of Agriculture.

Data analysis

To describe composition and structure of plant communities in the study area we calculated density of trees per plot extrapolating to a mean value per hectare for each zone, and for the whole area, based on pooled data. Data distribution assumptions of normality were tested using Kolmogorov–Smirnov test (p > 0.05), and homoscedacity of variance examined to allow use of analysis of variance for tree density between study zones (Bartlett’s test, p = 0.001). Therefore, a non-parametric analysis of variance test was used to compare tree density between the zones as data were not normal even after transformation. Species richness and diversity were estimated using Chao1 richness estimator and Shannon–Wiener diversity index respectively under program EstimateS (Colwell 2012). To evaluate richness and diversity indices, we set the number of runs to 100 to get smoothed curves at all levels of species accumulation (Gotelli and Colwell 2001). Randomization protocol was set at “randomization with sample replacement” with the respective diversity index setting to obtain species richness and diversity. Shannon diversity index was checked. Species richness between exotic and native species and between study zones was compared using Mann–Whitney and Wilcoxon signed rank tests respectively (Kolmogorov–Smirnov test, p > 0.05).

To gain knowledge of plant community structure in different study zones, cluster analysis was performed based on a Bray–Curtis similarity matrix of grouped variables with the program PRIMER v6 (Clarke and Gorley 2006). This was done after square-root transforming the data to downweight the effect of high abundance species (Clarke and Warwick 2001). Non-metric multidimensional scaling (NMDS) plots were used with cluster overlays to explore the similarity of the sample clusters within and between study zones. Where this was evident a further similarity test was performed to determine if there was statistically significant evidence of genuine clustering in the tree samples (Clarke and Gorley 2006). The similarity profile test (SIMPROF) is normally applied to the samples which are not a priori divided into groups (Clarke and Gorley 2006). To explore whether there were significant differences in tree assemblages between the study zones, analysis of similarities ANOSIM- (Clarke and Gorley 2006; Clarke and Warwick 2001) test was used between pairs of study zones. This test resembles the classical ANOVA and tests for the hypothesis that there is no significant difference in tree community structure between the study zones (Clarke and Gorley 2006) and has been used for other taxa (Hore and Uniyal 2008). Because this test is based on probability sampling, we set the number of permutations to 999 runs based on randomisation protocol of the samples. This test generates a global R-statistic which is a measure of distance between tree assemblages and can range from 0 to 1, indicating relative strength in distinct separation between the groups being compared (Clarke 1993). Furthermore, data on plant resource use and patterns of species introduction into the city and movement within and between study zones were analyzed descriptively using simple statistics.

Results

Tree density and community composition across urban–rural gradients

A total of 46,784 individual trees and shrubs comprising of 82 species in 33 families were recorded in the study area. Forty-five species were found in the urban core, 57 species in the sub-urban and 66 species in the peri-urban zone (Table 1). Of the 82 species recorded, 75.6 % (n = 62) were exotic while the remainder was native species. Total tree density in the study area was 390.93 ± 119.29 trees/ha. There was a significant difference in tree density between the three urban zones; peri-urban, sub-urban and core (Kruskal–Wallis test, χ2 = 16.582, median = 368 trees/ha, df = 2, p = 0.0001). Mean tree density was lowest in urban core (244.00 ± 120.86 tree/ha) and highest in peri-urban (465.60 ± 152.03 tree/ha). Exotic species contributed a much greater proportion than native species (333.47 ± 195.77 vs. 66.40 ± 21.89 tree/ha) to tree density totals in the area. Densities for both exotic and native species were significantly different among the three study zones (Exotic species, Kruskal–Wallis = 18.2, median = 296 tree/ha, df = 2, p = 0.0001; native species, Kruskal–Wallis χ2 = 20.652, median = 32.0 tree/ha, df = 2, p = 0.0001) and showed a relatively similar increasing pattern from urban core to the peri-urban area (Fig. 1).
Table 1

Plant species (woody and non-woody) recorded in the surveyed plots across different study zones in the study area

Species botanical name

Family

Native/exotic

Plant form

Peri-urban

Sub-urban

Urban core

Adansonia digitata

Malvaceae

N

T

x

  

Albizia lebbeck

Fabaceae

E

T

x

  

Anacardium occidentale

Anacardiaceae

E

T

 

x

 

Annona muricata

Annonaceae

E

T

x

x

x

Annona senegalensis

Annonaceae

N

T

x

x

x

Artocarpus artilis

Moraceae

E

T

x

x

 

Artocarpus heterophyllus

Moraceae

E

T

x

x

x

Arundinaria alpinia

Poaceae

N

S

x

x

 

Azadirachta indica

Meliaceae

E

T

x

x

x

Bambusa vulgaris

Poaceae

E

S

x

x

x

Brasaia actinophylla

Araliaceae

E

T

  

x

Caesalpinia pulcherrima

Fabaceae

E

S

x

  

Callistemom citrinus

Myrtaceae

E

T

 

x

x

Camellia sinensis

Theaceae

E

S

x

  

Casuarina cunninghamiana

Casuarinaceae

E

T

 

x

x

Casuarina equisetifolia

Casuarinaceae

E

S

x

  

Cedrela ordorata

Meliaceae

E

T

x

x

x

Ceiba pentandra

Malvaceae

E

T

x

 

x

Chamaecyparis lawsoniana

Cupressaceae

E

T

x

  

Cinnamomum zeylanicum

Lauraceae

E

T

x

  

Citrus limon

Rutaceae

E

T

x

x

x

Citrus sinensis

Rutaceae

E

S

x

x

x

Citrus spp 1

Rutaceae

E

T

x

x

x

Citrus spp 2

Rutaceae

E

T

x

x

 

Cocos nucifera

Arecaceae

E

T

x

x

x

Coffea arabica

Rubiaceae

E

S

x

x

x

Cupressus lusitanica

Cupressaceae

E

T

x

  

Cycas revoluta

Cycadaceae

E

T

 

x

 

Dalbergia melanoxylon

Leguminosae

N

S

x

 

x

Datura suaveolens

Solanaceae

E

S

x

x

x

Delonix regia

Fabaceae

E

T

x

x

x

Diplorhynchus condylocarpon

Apocynaceae

N

T

x

  

Elaeis guineensis

Arecaceae

E

T

x

x

x

Eucalyptus globulus

Myrtaceae

E

T

x

x

x

Euphorbia heterochroma

Euphorbiaceae

N

S

x

  

Euphorbia tirucalli

Euphorbiaceae

N

S

x

x

x

Ficus benjamina

Moraceae

E

T

 

x

 

Ficus capensis

Moraceae

N

T

x

  

Ficus stuhlmanii

Moraceae

N

T

x

  

Flacourtia indica

Salicaceae

N

T

x

x

x

Gossypium hirsutum

Malvaceae

N

S

x

x

 

Grevillea robusta

Proteaceae

E

T

x

x

x

Jacaranda mimosifolia

Bignoniaceae

E

T

x

x

 

Jatropha curcas

Euphorbiaceae

E

S

x

x

x

Khaya anthotheca

Meliaceae

N

T

x

x

x

Kigelia africana

Bignoniaceae

N

T

x

x

 

Lantana camara

Verbenaceae

E

S

x

x

 

Leucaena leucocephala

Fabaceae

E

S

x

x

 

Mangifera indica

Anacardiaceae

N

T

x

x

x

Markamia platycalyx

Bignoniaceae

E

T

  

x

Milicia excelsa

Moraceae

N

T

x

x

 

Morus alba

Musaceae

E

T

x

x

x

Newtonia buchananii

Fabaceae

N

T

 

x

 

Opuntia vulgaris

Cactaceae

E

S

x

x

 

Pandanus utilis

Pandanaceae

E

T

x

  

Persea americana

Lauraceae

E

T

x

x

x

Polyathia longifolia

Annonaceae

E

T

x

x

x

Psidium guajava

Myrtaceae

E

T

x

x

x

Pulmeria rubra

Apocynaceae

E

S

x

x

 

Punica granatum

Punicaceae

E

S

 

x

x

Rauvolfia apendiculata

Apocynaceae

E

T

 

x

 

Rauvolfia caffra

Apocynaceae

E

T

  

x

Ravenala madagascariensis

Strelitziaceae

E

T

 

x

x

Roystonea regia

Arecaceae

E

T

 

x

x

Schinus molle

Anacardiaceae

E

T

x

  

Sclerocarya birrea

Anacardiaceae

E

T

x

x

x

Senna grandis

Leguminosae

E

T

x

x

x

Senna siamea

Fabaceae

E

T

x

x

x

Solanum incanum

Solanaceae

E

S

x

  

Spathodea campanulata

Bignoniaceae

E

T

x

 

x

Sterculia appendiculata

Sterculiaceae

N

T

x

  

Syzygium aromaticum

Myrtaceae

E

T

x

  

Syzygium cuminii

Myrtaceae

E

T

x

x

x

Tamarindus indica

Fabaceae

N

T

x

x

 

Tecoma stans

Bignoniaceae

E

S

x

 

x

Terminalia catappa

Combretaceae

E

T

 

x

x

Terminalia mantaly

Combretaceae

E

T

x

x

x

Terminalia superba

Combretaceae

E

T

 

x

x

Tephrosia vogelii

Fabaceae

N

S

x

  

Thevetia peruviana

Apocynaceae

E

S

x

x

x

Thrinax floridana

Arecaceae

E

T

 

x

 

Trichilia emetica

Meliaceae

N

T

x

x

 

Native (N)/Exotic (E) shows species originality, plant form indicates whether it is a tree (T) or shrub (S) and a study zone with ‘x’ shows presence and without ‘x’ indicates absence of plant species in that particular zone. Non-native species contributed highest proportion to the vegetation community in this urban ecosystem (see text for statistics)

https://static-content.springer.com/image/art%3A10.1007%2Fs10531-014-0646-1/MediaObjects/10531_2014_646_Fig1_HTML.gif
Fig. 1

Tree density (mean ± SD) for exotic (white bars) and native (grey bar) species decreases along rural–urban gradient in Morogoro municipality. Density varied significantly among the study zones

Mean species richness for the peri-urban zone was lowest (50.2 ± 15.3 species), sub-urban had 75.2 ± 3.5 species (Fig. 2a) and the urban core had the highest richness (82.6 ± 1.4 species). Tree species richness was significantly different among the three urban zones (Kruskal–Wallis χ2 = 80.0, median = 76.3 species, df = 2, p = 0.0001). Overall, total species richness was on the average 69.3 ± 16.6 species for the combined data. On examining tree species origin across the study area, the exotic trees had a higher species-richness than the native (mean species richness—Exotic, 55.8 ± 2.3 species vs. native 17.6 ± 2.1 species). For both groups, species richness increased towards the urban core and was significantly different in peri-urban (Wilcoxon signed rank test Z = 7.49, p = 0.001), sub-urban (Wilcoxon signed rank test Z = 7.73, p = 0.001) and at the urban core segment (Wilcoxon signed rank test Z = 7.85, p = 0.001; Fig. 2a, b).
https://static-content.springer.com/image/art%3A10.1007%2Fs10531-014-0646-1/MediaObjects/10531_2014_646_Fig2_HTML.gif
Fig. 2

a Tree species richness (mean ± SD) increases from peri-urban to the core urban in the study area. Species richness was significantly different between study zones. b Native and exotic species richness (mean ± SD) in peri-urban (white bars); sub-urban (dotted bars) and urban core (grey bars). Native and exotic species richness were significantly different across the three urban zones

At the level of study zones, peri-urban had an alpha diversity of 2.61 ± 0.23, sub-urban had 2.79 ± 0.01 and the urban core had a diversity of 2.82 ± 0.01. These were significantly different between the zones (Kruskal–Wallis χ2 = 87.36, df = 2, p = 0.0001). Across the study area total Shannon diversity was 3.0 ± 0.69 species. Mean Shannon species diversity index for the exotic species was 2.74 ± 0.12 species, significantly higher than for native species (1.90 ± 0.13) in the area (Mann–Whitney U = 159.0, p = 0.0001). Taken separately and for each zone, Shannon diversity indices for exotic species were higher than for the native species (Fig. 3) and were significantly different across the three study zones; peri-urban (Mann–Whitney U = 22.0, Z = −7.49, p = 0.001), sub-urban (Mann–Whitney U = 2.11, Z = −7.73, p = 0.001) and urban core zone (Mann–Whitney U = 2.01, Z = −7.85, p = 0.001).
https://static-content.springer.com/image/art%3A10.1007%2Fs10531-014-0646-1/MediaObjects/10531_2014_646_Fig3_HTML.gif
Fig. 3

Native and exotic (mean ± SD) species diversity in peri-urban (white bars), sub-urban (dotted bars) and urban core (grey bars) in the study area. Native and exotic species diversity were significantly different across the three study area

Tree community structure and pattern of species movements within and between urban zones

Analysis of species composition and community structure revealed three broad groups of plant communities across the study area (Fig. 4). There was no significant internal structuring within groups of plant communities (SIMPROF-test, Pi = 0.641; p < 0.3 %). Across the three urban zones, tree samples showed only 10 % dissimilarity with most species overlapping between tree samples across the zones (Fig. 5). Pair wise comparison using ANOSIM test indicated there was least difference in tree assemblages between peri-urban and urban core zones (R = 0.34, p = 0.001) and between sub-urban and urban core tree communities (R = 0.14, p = 0.001). Tree community structure occurring between peri-urban and sub-urban zones was also barely separable (R = 0.092, p = 0.013). Further, about 45 % (n = 54) of the sampled households (N = 120) had planted native and exotic species for various purposes ranging from ecological to socio-economic importance. Most of the respondents (68.7 %) said that potential benefits from the plant species influenced their choices of species to introduce into their backyard gardens. Some of the benefits included timber, building poles, prevention of windstorms, fruit, protection fence, shade and local medicinal properties (Table 2). The pattern of species movement (Table 3) within a particular study zone was random and multidirectional (Fig. 6). A small proportion (9.3 %) of household respondents indicated to have introduced some tree species from out Morogoro city. About 23.7 % of respondents in peri-urban zone introduced species either from sub-urban or urban core areas, while 29.1 % in sub-urban introduced species both from peri-urban and outside the city. Most respondents (71.9 %) in urban core areas said the tree species available were mostly introduced from outside the city. Moreover, most of the plant species growing in the urban were either introduced by humans or by other mechanisms such as dispersal by birds. The garden–garden (G–G) movements within any particular zone were either facilitated by humans or non-human factors such as wind, water runoffs and birds.
https://static-content.springer.com/image/art%3A10.1007%2Fs10531-014-0646-1/MediaObjects/10531_2014_646_Fig4_HTML.gif
Fig. 4

Dendrogram of tree samples from across the study area showing three broad clusters of plant communities, A–C based on Bray Curtis similarity of the tree samples. There was not significant internal structuring of samples within individual community cluster

https://static-content.springer.com/image/art%3A10.1007%2Fs10531-014-0646-1/MediaObjects/10531_2014_646_Fig5_HTML.gif
Fig. 5

MDS ordinatination plots of tree samples from three urban zones in Morogoro city generated based on tree abundances in peri-urban (cross sign) sub-urban (plus sign) and urban core (open square) in the study area. There is clearly great overlapping of the samples of different zones indicating high similarity in tree samples between urban zones

Table 2

A list of ecosystem services and goods as incentives driving introduction and movement of plant species in the study area, showing relative proportions of respondents (%, n = 240) from different zones

Ecological services/goods

Peri-urban

Sub-urban

Core-urban

Increase soil fertility

20

7.5

7.5

Provide shelter for other organisms

22.5

17.5

15

Prevent soil erosion

17.5

10

5

Moisture retention

27.5

12.5

7.5

Source of food for urban wildlife

32.5

10

12.5

Provide shade to humans

25

15

17.5

Regulation of wind speed

35

42.5

55

Species refugia

12.5

5

7.5

Provides clean air

10

5

5

Carbon sequestration and storage

10

17.5

12.5

Seed dispersal

20

2.5

27.5

Provide local medicine extracts

7.5

12.5

7.5

Rainfall formation

37.7

32.5

40

Provides fruits for both humans and animals

5

20

17.5

Table 3

Plant species (woody and non-woody) considered to be non-native that was introduced in the study area by human and non-human factors

Species botanical name

Peri-urban

Sub-urban

Urban core

Albizia lebbeck

 

x

x

Anacardium occidentale

x

 

x

Annona muricata

   

Artocarpus artilis

  

x

Artocarpus heterophyllus

   

Azadirachta indica

   

Bambusa vulgaris

   

Brasaia actinophylla

x

x

 

Caesalpinia pulcherrima

 

x

x

Callistemom citrinus

x

  

Camellia sinensis

 

x

x

Casuarina cunninghamiana

x

  

Casuarina equisetifolia

 

x

x

Cedrela ordorata

   

Ceiba pentandra

 

x

 

Chamaecyparis lawsoniana

 

x

x

Cinnamomum zeylanicum

 

x

x

Citrus limon

   

Citrus sinensis

   

Citrus spp.1

   

Citrus spp.2

  

x

Cocos nucifera

   

Coffea arabica

   

Cupressus lusitanica

 

x

x

Cycas revoluta

x

 

x

Datura suaveolens

   

Delonix regia

   

Elaeis guineensis

   

Eucalyptus globulus

   

Ficus benjamina

x

 

x

Grevillea robusta

   

Jacaranda mimosifolia

  

x

Jatropha curcas

   

Lantana camara

  

x

Leucaena leucocephala

  

x

Markamia platycalyx

x

x

 

Morus alba

   

Opuntia vulgaris

  

x

Pandanus utilis

 

x

x

Persea americana

   

Polyathia longifolia

   

Psidium guajava

   

Pulmeria rubra

  

x

Punica granatum

x

  

Rauvolfia apendiculata

x

 

x

Rauvolfia caffra

x

x

 

Ravenala madagascariensis

x

  

Roystonea regia

x

  

Schinus molle

 

x

x

Sclerocarya birrea

   

Senna grandis

   

Senna siamea

   

Solanum incanum

 

x

x

Spathodea campanulata

 

x

 

Syzygium aromaticum

 

x

x

Syzygium cuminii

   

Tecoma stans

 

x

 

Terminalia catappa

x

  

Terminalia mantaly

   

Terminalia superba

x

  

Thevetia peruviana

   

Thrinax floridana

x

 

x

A mark ‘x’ shows species absence in a particular zone. A species without ‘x’ indicates its presence, and where two or three zones indicate shared emptiness (i.e. without x) indicate that a species has been introduced into that area from either outside the city or from adjacent urban zone (urban, sub-urban, peri-urban)

https://static-content.springer.com/image/art%3A10.1007%2Fs10531-014-0646-1/MediaObjects/10531_2014_646_Fig6_HTML.gif
Fig. 6

Movement of species from outside the city (bold arrow from outside surrounding matrix) and inside the city of Morogoro, Tanzania. The movement within any particular urban zone (Garden–Garden movement) and across (bold arrows between zones; urban core, sub-urban and peri-urban) have influenced vegetation structure of the study area, and are induced by human and non-human factors

Discussion

In this study, density of tree species increased from urban core to peri-urban areas as expected. This was perhaps due to the limited spaces available at urban core where many trees could be grown than were available in the peri-urban or sub-urban. It appears also that most of the peri-urban areas were planted with various tree species used in part for local household utilization and for serving urban centers with derived goods and services, thus, likely influencing the high tree densities (Dow 2000; Hope et al. 2003). Exotic species exist at higher densities than native species across the study area indicating residents’ higher preferences for such species. In the urban core, the high density of exotics was associated with accidental or deliberate introductions of plant species by humans for the various purposes including ornamentation, cultural and economic needs as has been reported elsewhere (Kowarik 2003; von der Lippe and Kowarik 2008).

Species richness was highest in the urban core which represents the highest level of habitat disturbance, thus contrasting with the general disturbance hypothesis which predicts high richness in mild disturbances i.e. sub-urban areas (McKinney 2008; Bongers et al. 2009). According to the disturbance categories described by McKinney (2002), the highest disturbance level represents urban core with areas having over 50 % hard or paved surfaces, which is a characteristic of cities of most developed countries. However, such a disturbance level hardly exists in the urban cores of most medium-size Tanzanian cities, including our study area. This is probably due to relatively low level of development. Comparatively, the disturbance level in urban cores in our study area is far less pronounced than it is for the developed countries which are characterized by tightly packed buildings, pavements and sealed surfaces, and green spaces (McKinney 2002). These results confirm the previous observations which have recognized context and scale dependent effects associated with urban sprawl on species richness (McKinney 2008), consistent with several other studies conducted across the Oriental and Pacific (Kim and Pauleit 2005; Tait et al. 2005), Neotropical (Pauchard et al. 2006) and Palearctic (Kuhn et al. 2004; McKinney 2008) regions. High species richness was attributable to local use of tree species by the residents particularly for cultural, socio-economic and ecological reasons (Dwyer et al. 1991) and introductions by city residents or traders (von der Lippe and Kowarik 2008).

Community structure of the tree assemblages in the study area showed little differences among the different urban zones due to many tree species occurring in multiple urban zones. Also, human influences on tree communities were observed across the study area as people assembled many plant species both native and exotic for various purposes ranging from aesthetic, socio-cultural to economic gains. Species assembly by people in urban areas may have influenced vegetation structure of the urban landscape indicating the bigger role played by the residents in shaping the urban vegetation communities (McKinney 2002; Ortega-Álvarez et al. 2011). Highest similarity among the clusters of tree communities observed on the chaining dendrogram further suggests potential danger of increasing vegetation homogenization due to the introduction of common plant species in the study area. Similarity in the vegetation structure was probably caused by the observed pattern of species movement from outside the city and between and within particular urban zone. The garden–garden movement of species by humans further increased the dilution effect on the urban vegetation structure. On the other hand, domestication of the tree species that seem to be increasingly threatened could provide a way to serve these species against local extinctions (Msuya and Kideghesho 2009).

Results from the household surveys revealed that local uses of plants influenced the residents’ choices of which tree species to put in the backyards and surrounding homesteads. Like in many other cities (Breuste 2004), the majority of the residents in our study area planted exotic trees for reasons such as shade, wind break, construction poles, fruits and ornamentation perhaps because exotic species are believed to reach maturity at relatively lower age compared to native tree species. Such benefits motivate people to greening their homesteads and the city as a whole (Rija 2010). These observations were consistent with other studies done elsewhere in developed countries where people in cities prefer gardens associated with exotic species and extensive lawns over more natural vegetation thereby causing local biodiversity to dwindle (Henderson et al. 1998; Breuste 2004). Further, lower tree species richness in the peri-urban was presumably attributed to two reasons; low per capita income (McGranahan and Satterthwaite 2003), and the high level of habitat disturbances caused by agricultural activities conducted at the expense of native tree species (Pauchard et al. 2006). This was evidenced by the relatively lower income of local residents in the peri-urban segment than in the Morogoro urban core. Our household survey show that urban core residents earn twice as much as those from the peri-urban gardens (Rija et al. in prep), perhaps because the urban core gardens are continually cultivated throughout the year on a commercial basis to cater for the high demands of urban foodstuff. In cases where gardening and establishment of urban woodlots bring some relief to the income-deprived households through selling of the harvest (fruits, vegetables, fire wood, etc.), domestication of trees could often be taken as a coping strategy thereby increasing tree species richness overall (McKinney 2002). Kideghesho and Msuya (2010) reported that domestication of indigenous plants in northern Tanzania was driven by the derived benefits such as access to nearby herbal medicines and other resources consistent with our observations during the household survey.

Implications on local biodiversity and functioning in urban ecosystem

This study has shown that the pattern of tree species richness in Morogoro municipality is influenced by human activities (economic, socio-cultural and ecological) associated with the spread of urbanization. The urban vegetation communities are largely shaped by the people’s needs, and whose preferences appear to be biased towards exotic plant species over the native ones. This is evident due to highest contribution of the exotic tree species on the overall tree assemblage in the study area. The pattern of species movement within and between the city zones is high and further confirms ongoing spreading of these species by the residents. For whatever reasons these introductions might come about, there is increasing potential for these species becoming invasive thereby impacting the native species negatively (Sax et al. 2005; McKinney 2008). The arrival of exotic species may create unforeseeable results that may affect the urban ecosystem either positively or negatively especially when such species turn out to be invasive. This will have serious consequences to the urban biodiversity and on the health functioning of urban ecosystem as whole. Conservation awareness to the urban dwellers is urgently needed on proper habitat management and conservation to reduce rate of exotic species introduction into the city in order to maintain habitat heterogeneity for the various wildlife and local residents living in the city.

Acknowledgments

We thank the residents of Morogoro municipality for allowing us to carry out research around their homesteads and for taking part during the questionnaire surveys. We thank Andy Bowkett for providing comments on the earlier draft of this paper and Ms Agnes Sirima for the logistical help. Two anonymous reviewers provided comments on the manuscript.

Copyright information

© Springer Science+Business Media Dordrecht 2014