Public Green Spaces

Urban forests consist of trees, shrubs and associated herbaceous vegetation located within the boundaries of a city. It also includes lines of trees on sidewalks and median strips in streets and avenues, and in general, any type of assemblage of plants ( ^{Benavides-Meza & Fernández-Grandizo, 2012} ) (graph 1). Together with streams and other water bodies, urban forests comprise what is known as the urban ecosystem or, as Bolund and Hunhammar (1999) put it, the green and blue spaces located within the urban area.

Source: Adapted from ^{Kardinal-Jusuf & Nuyuk-Hien (2016)}.

Figure 1. Types of Urban Forests  

In the present study, the PGS was defined as a subtype of urban forest and a type of public space; it consists in public squares, parks, and gardens; most of the time, the municipal government is responsible for its administration and maintenance. Urbanistic terminology includes different terms to refer to these spaces; the Mexican regulatory system for urban recreational equipment prepared by the Secretariat of Social Development provides additional examples ( ^{SEDESOL, 1999} ) (see table 1). Finally, a PGS must include three necessary components: vegetation, urban infrastructure, and public access. Although all three of these dimensions are important, this work focuses on the plant component of PGSs.

Vegetation: Green Cover

The term green area is a metaphorical expression that refers to the color of photosynthetically active vegetation, whether trees, shrubs, or grasses, in a given location. On the contrary, constructions, buildings, and infrastructure are referred to as gray infrastructure. The area of the vegetation layer that covers the soil in relation to the total area of the PGS is termed plant cover (PC).

Although the literature has not achieved consensus concerning the minimal proportion of PC for an area to be considered a PGS, it is expected to be sizable, since plants are responsible for many ecological functions that translate into urban ecosystem services (UES). Vegetation can improve the thermal sensation of outdoor locales as a result of evapotranspiration, and its shade reduces albedo values: trees can deflect between 75% and 90% of solar radiation (Kardinal-Jusuf & Nuyuk -Hien, 2016), increasing the comfort of the PGS. For instance, as early as 1975, the Chapultepec Forest in Mexico City was identified as a “cold island” of better air quality than the rest of the city ( ^{Jáuregui-Ostos, 1975} ). Other UESs provided by the urban forest are carbon sequestration, rain delay, improvement of air quality and sound insulation.

^{Nowak et al. (1996)} measured the urban forest tree cover in 58 cities in the United States, and their results ranged from 55% in Baton Rouge, Louisiana, to 0.4% in Lancaster, California; the authors conclude that the natural environment surrounding a city is the best predictor of its tree cover. For example, it varied from 0.4 to 26% in arid locations.⁸ The authors also found differences within each city as a function of land use; tree cover averaged 11.3% in urban parks located in arid areas and 47.6% in locations where the natural surroundings were forested.

For its part, ^{SEDESOL (1999)} describes a general architectural program for urban recreational areas in Mexico where the presence of PC is implicitly suggested. Although this program does not imply that the surface should be completely covered with vegetation, it refers to green areas, forests, or open areas (Table 1).

The percentage of PC and the social attractiveness of a PGS are not linearly associated. Attempts have been made to determine optimal tree density based on measurements of scenic beauty, security, and safety. For instance, ^{Schroeder & Green (1985)} found that the perception of scenic beauty in a park and the actual density of trees maintain an inverted-u relationship, that is, too few or too many trees result in unattractive landscapes. The optimal level was found to be between 150 and 160 mature trees per hectare, which is equivalent to 1.55 trees per 100 m². In the case of cities such as Phoenix or Hermosillo, an average tree is estimated to have a crown radius of three meters, which means that 1.55 trees per 100 m² represent 44% of PC.⁹

Sustainable PGS

The traditional urban landscape is usually dotted with green spaces including many exotic species, poorly adapted to the local climate, that require considerable maintenance, especially irrigation ( ^{Spinti, Hilaire, & VanLeeuwen, 2004} ). An arid climate, low water availability, and a strong preference for non-native species introduced from rainy climates in tropical or temperate regions have resulted in a shortage of plant life in PGs. In Hermosillo, for instance, parks, plazas, and gardens are often places where a brownish-gray color predominates, and vegetation is scarce or inexistent ( ^{Clemente-Marroquín, 2007} ). The intense solar radiation and absence of vegetation typical desert climates create the phenomenon known as urban heat islands, where the social function of this spaces is impossible due to the unattractive microclimate.

^{Clemente-Marroquín (2007)} indicates that drought or water scarcity periods have drastic effects on public park maintenance in Hermosillo; a considerable amount of plants dries out and much additional work is required to keep irrigation systems running. Exotic species such as Ficus nitida, Ficus benjamina, and Grevillea robusta, among others, which are widely distributed in the city, are affected by diseases such as Texas root rot ( ^{El Imparcial, 2014} ). ^{Navarro and Moreno (2016)} found that the arboreal landscape of the Hermosillo sidewalks was dominated by two non-native species: F. benjamina (45.34%) and black olive (Bucida buceras) (23.79%). The relationships among lack of vegetation, an insufficient water supply, and scarce use of local plant species has also been highlighted by ^{Ojeda Revah and Álvarez (2000, p. 18)} , who focused their study in Tijuana.

The preference for exotic species over native vegetation is not unique to the present case study. ^{Raxworthy (2010)} highlights how in Australia, an arid country where native vegetation has evolved mechanisms to cope with drought, inhabitants continue to adopt the European paradigm of a green landscape, which requires abundant water and is therefore unsustainable. A similar experience is described by ^{Alam, Khattak, Ppoyil, Kurup, and Ksiksi (2017)} concerning the United Arab Emirates. In all cases, cities are faced with the dilemma between favoring plant cover to maximize climate comfort or minimizing water use ( ^{McPherson & Dougherty, 1989} ; ^{Guhathakurta & Gober, 2007} ).

On the other hand, an ecological landscaping model has the potential to reconcile scarce resources for the maintenance of PGSs, especially irrigation water, with the demand for urban ecosystem services and people’s landscaping preferences ( ^{Hitchmough & Dunnett, 2004} ). The first step in such direction would be to select plants adapted to the local temperature and water availability (rain and irrigation) ( ^{Beck, 2013} ). On the other hand, xeriscape landscaping involves the design of gardens or PGSs that require little or no supplemental irrigation ( ^{Tyman, 2011} ).

Thus, this approach considers non-native, alien, imported, introduced, or exotic species as inadequate for the local environment. Unwanted species are usually brought from mesic, temperate, or tropical environments; these species are adapted to humid conditions, and once established in the dry environment and high temperatures prevalent in cities such as Hermosillo, they require large amounts of water.

Strictly speaking, native plants are those among the species existing before the city engulfed the original environment, i.e., the natural vegetation one can find surrounding the city. Naturalized or introduced species and those that have become invasive are excluded from this category. For instance, the type of natural vegetation surrounding the city of Hermosillo was described by the Consultative Technical Commission for the Determination of Pastureland Coefficients ( ^{COTECOCA, 1986} ) as arborescent scrub, a combination of medium-sized trees, shrubs, as well as some cacti. The local vegetation has also been classified as xerophilous scrub ( ^{Rzedowski, 1990} ), xerophilous mesquite, sarcocaule (thick-stemmed) scrub, and desert microphyllum ( ^{Inegi, 2013a} ).

Water Demand for Urban Forests

Outdoor water consumption in single-family homes with landscaped spaces has been estimated at an average of 54.62% of the home’s total water consumption in arid and semi- arid cities in the United States southwest. A particular example of this consumption pattern is Tempe, Arizona, where 67.3% of the home’s water is consumed outdoor ( ^{Western Resources Advocates, 2003, p. 63} ).

The large amounts of outdoor water consumed in arid and semiarid cities are reflected by the presence of water demand peaks during the summer. ^{Johnson (1980)} described a 124% increase in residential water demand in the city of Tucson during the month of July in comparison with the month of January. The same phenomenon was reported in Phoenix, where water demand in July was found to be more than twice as high as in February ( ^{City of Phoenix, 2011} ). In the case of Hermosillo, the lowest monthly consumption levels occur in December whereas the highest occur in June, when demand is 30% higher.¹⁰

Most of the water consumed outdoors is assumed to be used for watering gardens. ^{Sovocool (2005)} took measurements before and after transforming a traditional garden into a xeriscape garden in Las Vegas and found that the changes resulted in 30% savings in annual water consumption for an area equivalent to a traditional lawn.¹¹

^{Moll (1989)} points out that cities such as Tucson, Arizona, face two major limitations to the development of urban vegetation: temperature and water. Native species have adaptive advantages over these limitations. For example, a mulberry tree (Morus spp.) can demand as much water as four paloverdes (Parkinsonia spp.) ( ^{McPherson & Dougherty, 1989} ). This adaptive capacity can be observed in plants such as ironwood (Olneya tesota) ( ^{Smith, Monson & Anderson, 1997, p. 73} ), littleleaf false tamarind (Lysiloma watsonii) ( ^{Yetman, 2003, p. 203} ), and hopbush (Dodonaea viscosa), ( ^{Felger, Johnson, & Wilson, 2001} ), which remain green in spite of the lack of water during the dry season.

Nevertheless, in cities such as Phoenix, xeriscaping fails to reduce temperature and increase people’s thermal comfort as in the case of mesic gardens with grass and exotic plants ( ^{Chow & Brazel, 2012} ). As already mentioned, the adoption of native vegetation, available maintenance resources, and people’s landscaping preferences must be taken into account to achieve an optimal balance ( ^{Hitchmough & Dunnett, 2004} ).

Quantitative Strategy

The unit of analysis was defined as the PGS, represented by individual parks and gardens administrated by the municipality. The Parks and Gardens Department of the municipality of Hermosillo provided a spreadsheet listing the city’s parks (updated to 2015), which were categorized as squares, rural locations, and boulevards. Only the category of squares was used in the present study; these venues totaled 669. The list included the name of the park, its location, and its area in square meters. The polygons of 590 squares were manually georeferenced and outlined based on this information and data from Google Earth.¹²

Determination of Plant Cover

A sample of 112 of the identified 590 squares was selected for analysis. Each of these polygons was photo-interpreted to differentiate masses of vegetation, gray infrastructure, and bare ground. Photo-interpretation consists in interpreting the features and characteristics of high-resolution satellite images based on shape, color, texture, pattern, association, and other a priori information that allows for the categorization of a geographical object (polygon, point, or line) within a pre-established class ( ^{Loelkes, Howard, Schwertz, Lampert, & Miller, 1983} ). The analyst assigns a category and creates the object manually, simultaneously establishing its extension. PC was estimated as the quotient between the area photo-interpreted as vegetation and the total area of the PGS.

Floristic Composition

The number of trees in each PGS was counted with the help of experts, and each specimen was investigated using clues such as its common name and characteristic traits (size, color of flowers, fruits, etc.). Additionally, their genus, species, and families were determined to identify local species.

Qualitative Strategy

This part of the analysis consisted in a documentary review of laws and regulations related to the afforestation policy that governs PGS management.

a) International Recommendations

The city has adopted a norm of 9 m² of green area per inhabitant. Reportedly, this volume is an international standard for PGSs recommended by the World Health Organization (WHO); however, we were unable to locate the original source where this recommendation was made, and how it was determined remains unknown. Noticing the same inconsistency, other authors have suggested that the parameter was never put forward by the WHO ( ^{Henríquez-Ruíz, 2014} ; ^{Huizar-Contreras & Ojeda Revah, 2014} ). However, multiple publications and different diagnoses ( ^{Alanís-Flores, 2005} ; ^{Jiménez-Pérez, Cuellar, & Treviño, 2013} ) focused on PGSs in cities around the world use the 9 m² value as a reference ( ^{OECD, 2017, p. 236} ; ^{Mohapatra, 2016, p.12} ; ^{Ros-Orta, 2013, p.31} ; ^{Sorensen, Barzetti, & Williams, 1998, p.4} ).

In any case, the standard is difficult to implement due to different reasons. In the first place, it is not clear which spaces should be defined as green areas; in this regard, ^{Rapoport and López-Moreno (1987, p.14)} indicate that private urban gardens are often not included in green area calculations even though they represent an important amount of a city’s vegetation. A second issue is that the variables of even distribution and convenient access are seldom considered, despite that green areas should be no further than a 15-minute walk from potential users’ homes ( ^{Sorensen, Barzettiy, & Williams, 1998} ). Finally, the standard says nothing about the quality of these spaces, for example, optimal PC or infrastructure and maintenance parameters or guidelines concerning social functionality.

Mexico has made international commitments related to green areas and urban ecosystems. For example, the 2011-2020 Strategic Plan for Biodiversity and Aichi Biodiversity Targets. Specifically, Target 14 concerns the safeguarding of ecosystems that provide essential services, considering that urbanization is, essentially, replacing a natural ecosystem with an urban ecosystem. The 2030 Agenda, adopted on September 25, 2015, established Sustainable Development Goal (SDG) 11 with the purpose of guaranteeing the safety, resilience, and sustainability of human cities and settlements. SDG 11.7 establishes the need for universal access to green areas and safe, inclusive, and accessible public spaces by 2030, particularly for women and children, the elderly, and the disabled. Explicit linkages between these international commitments and national, state, or local planning instruments are unavailable.

b) Federal Regulatory Framework

The General Law on Human Settlements, Territorial Zoning and Urban Development, a federal regulation, empowers states and municipalities to ensure an adequate balance between green areas and infrastructure construction based on national regulations on the subject (articles 75 and 76).

The creation of PGSs is concomitant with the urbanization process. The planning, design, and extension of these areas are firstly in charge of the developer or real estate developer, who must observe the relevant federal, state, and municipal regulatory frameworks. When urban areas grow at the expense of lands bearing natural vegetation and are part of the nation's forests, developers require authorization to change the land use, as described in the General Law on Sustainable Forest Development (article 93). A favorable decision concerning environmental impact, in accordance with the General Law of Ecological Equilibrium and Environmental Protection, must be obtained prior to this process. These authorizations are granted by a federal agency: the Semarnat.

It should be pointed out that Semarnat usually recommends or mandates the use of plant species present in the local geographical context for the afforestation of the green areas, boulevards, and median strips included in urban design, and it Promotes the rescue of species included in NOM 059-SEMARNAT-2010, which recommends making them available to the municipality. In addition, it suggests that the permanence of original vegetation should be privileged as much as possible.

Finally, ^{SEDESOL (1999)} elaborated a normative system for recreational areas that specifies the required architectural components, dimensions, service radius, and population to be benefited by the space. However, according to ^{Huizar-Contreras (2012, p.23)} , these recommendations are only guidelines, which means that their application is decided individually by the state or municipality.

c) State Regulatory Framework

In quantitative terms, state legislation determines the extent and, more recently, the possible PC of PGSs. For at least 30 years, from the Urban Development Law (February 18, 1985) to the Territorial Ordinance and Urban Development Law (September 28, 2006), which repealed it, until decree number 149 (July 16, 2015), which added and reformed the latter law, the policy was for the developer to donate 2% to 3% of salable area for parks and gardens.¹⁵

Decree number 149 amended the Law on Land Use Planning and Urban Development in favor of green areas via the following stipulations:

For example, considering housing developments with a density of 50 houses per hectare, 125-133 m² per lot, and an average density of 3.63 inhabitants per house, if 3% of the salable area was allocated to PGSs, the result was around 1.1 m² per inhabitant. Given these minimum requirements, it was unlikely to meet the minimum WHO indicator unless the municipality built neighborhood parks to supplement the surface area donated by the developers, independently of its extension. On the other hand, the amendment to article 103, by decree number 149, fails to specify how the donation of green areas is proportional to population density. The 6.2 m² per dwelling factor can be translated into a fixed 7% for 90 m²-dwellings, which would result in a PGS density below the WHO recommendation.

As explained above, the extension of PGSs is parallel to urbanization via donated areas or the construction of parks and gardens by the municipality. However, the recommended tree density for these spaces was never regulated. Recently, the state government approved the Law for the Protection, Conservation, and Development of Trees in Urban Areas in the State of Sonora, better known as “Ley Arbol.” Considering the specificity of this law in the matter of urban forest measurement, we decided to describe it and analyze it in detail in the following section.

d) Municipal Regulatory Framework

Article 115, section III, subsection g of the Mexican Federal Constitution, states that municipalities are in charge of parks and gardens. The state regulations described above are to be observed by the municipalities. The role of the municipality is subdivided into PGS planning, construction, maintenance, and species cataloging.

PGS Maintenance and Inventory

PGSs are managed by CIDUE via the General Directorate of Urban Development and the Parks and Gardens Direction (PGD), which has a Tree Control Department (TCD). The purpose of these administrative areas is to provide conservation, maintenance, and afforestation services for public green areas, which are the responsibility of the municipality; they are also in charge of maintaining median strips in boulevards, planting trees, and irrigating. The TCD acquires and grows the trees planted in PGSs. Therefore, this government organism is partly responsible for deciding the species to plant.

In May 2016, the PGD was asked to inventory its green areas. The resulting list divided the city into four sectors and included the name of the green area, its surface, and its location. They were also classified into plazas, boulevards, and rural areas; since the latter were outside the polygon of the city, they were discarded from the analysis. Boulevards were not considered as PGS since most of them lack social infrastructure. The information provided by the PGD (summarized in Table 3), reveals that the area of PGS corresponding to plazas was 273 hectares, which represents 3.08 m² per inhabitant.

Species Cataloging

As already mentioned, several articles (96, 98, and 100) of the Law on Land Use Planning and Urban Development stress the use of natural, native, or adapted non-introduced vegetation. Article 46 of the Ley Árbol stipulates the creation of a catalog of species.

Article 52 of the Construction Regulations for the Municipality of Hermosillo establishes that existing trees should be preferably preserved and that local species, adapted to the city’s climate, should be favored. The article presents a list of 16 recommended exotic tree species such as bitter orange, black olive, grevillea, nimtree (Azadirachta indica), kapok tree (Ceiba pentandra) and Yucatecan; 11 regional (native) species such as paloverde (Parkinsonia spp.), palo brea (Parkinsonia praecox), ironwood, mesquite, and tepeguaje; six palm species (one native); eight cactus species, and eight species of yucca and agave. Clearly, CIDUE has allowed for exotic species¹⁷ among the vegetation that can be planted in the city to accommodate the interests of real estate development projects.

Floristic Composition and Abundance of Species

To determine the extent to which the use of natural and non-introduced vegetation was observed in PGS, we analyzed a sample of 112 PGS distributed homogeneously throughout the city. A total of 6,065 trees belonging to 66 species were registered. Ten species represented around 85% of the afforestation pattern (Table 4). The average percentage of regional vegetation per PGS was 38; 21 different species were identified, 10 of which belonged to a single plant family—the Fabaceae (legumes)—and 10 species, despite being native, are adapted to riparian or more humid environments than those in the city, for example: uvalama (Vitex mollis), blackbead (Pithecellobium dulce), amapa (Unrelenting Tabebuia), west Indian elm (Guazuma ulmifolia) and desert willow (Chilopsis linearis). The most important native plants, either because of their abundance or their distribution, were mesquite, fan palm, tepeguaje, and blackbead.

An aspect worth mentioning is that, although the spatial arrangement of the species in each PGS was not directly evaluated, it could be observed that native trees were mixed with exotic species. For example, mesquites were growing next to black olive trees, although the recommendations of xeric landscaping are to group trees according to their hydrological needs.

Index of Diversity

The afforestation pattern was dominated by a few species: 64% of the trees were black olives, mesquites, or bitter orange trees. Shannon’s H diversity index (Shannon & Weaver, 1949) was used to estimate the diversity of the landscape; this index determines the degree of diversity in a sample from a study area as compared to its potential diversity based on the overall pattern of species in the PGS universe (n = 112). This index ranges from a value of 0 in the case of a single-species afforestation pattern to 4.19 (natural logarithm of 66 for the case study) for a hypothetical case of maximum equitability where the 66 species were present in one PGS. On average, the studied parks had an index of 1.04, representing 25% of their potential diversity. It is important to mention that 17 of the 66 species were unique to single PGSs.

Frequency of Species

This indicator measures the presence or absence of a species in a given PGS. Therefore, it is a measure of the geographical distribution of the species in the city. In this regard, the best-distributed species were mesquite, black olive, and fan palm (Table 5).

The information obtained from the sample confirmed the second hypothesis, which states that the floristic composition of Hermosillo’s PGSs is dominated by exotic species. The diversity index is low, which suggests the predominance of a small number of species. These results are consistent with those found by Navarro and Moreno (2016), although they studied other components of the urban forest.

Plant Cover and Tree Density

The average area of the PGS sample was 5,850 m² (range 145-94,708 m²); half of the sample had an area equal to or less than 2,154 m². Mean PC was 27% (range 0-68%, graphic 2 ); it is important to mention that three of the PGS were completely devoid of plant cover. No statistically significant correlations were found among the variables of PC percentage, total PGS area, and total tree count. Tree density was 1.15 trees per 100 m² (range 0 to 4.4).

Source: Prepared by the author based on field data (2017).

Graphic 2. Distribution of Plant Cover Percentage in the Sample  

For example, the park with the highest absolute count was Francisco I. Madero, with 1,439 trees (1.52 individual trees per 100 m²). This PGS is the oldest in the city; at the time of the present study, it was the only space receiving irrigation from treated water through a purple line, and its PC was 48%. On the other hand, the irrigation system of 87.50% of the studied PGSs used drinking water from the public grid and relied on drip irrigation, locally known as cajete, although it was not possible to verify whether the systems worked properly; moreover, the presence of this irrigation system was not significantly correlated with tree density or PC.