Introduction

Society requires a sufficient area of urban forests and green areas that allow it to carry out and develop social, educational, cultural, and civic activities (^{Ruiz-Montiel et al., 2014}); in particular, urban trees offer many benefits, as they improve the quality of the environment (air and water), produce a sense of well-being, promote lower temperatures and reduce ultraviolet radiation (^{Nowak et al., 2006}). However, they face stress conditions caused by the action of man, particularly in areas where pedestrian and vehicle traffic is frequent. For this reason, it is important to assess the state of the trees and know if they are in optimal conditions and if the risk they mean to people is minimized (^{Vogt et al., 2015}), as well as damage to infrastructure (^{Koeser et al., 2016}).

A risk tree is described based on its instability due to a defect in its structure, that is, any part of the specimen: trunk, branches or crown can collapse and fall on a person or infrastructure causing them injury or damage (^{National Tree Safety Group, 2011}). Therefore, the evaluation of urban trees aims to identify and qualify such structural condition, in order to determine its potential risk and probability of causing damage (^{Matheny and Clark, 2009}).

The challenge is to provide the right tree management to ensure a low level of risk (^{Tomao et al., 2015}). A long-lived specimen without maintenance is more likely to become harmful (^{Albers et al., 2003}); thus, it is important to identify the specimens that represent a risk as a result of their deterioration due to water or thermal stress, atmospheric pollution, urbanization, attack by pests or diseases (^{Restrepo et al., 2015}), or also as a result of a bad management (^{Hauer and Johnson, 2003}). Tree defects are evaluated through a visual inspection to identify and understand the indicators of potential danger (Calaza and Iglesias, 2016).

Seven categories of potential damage features are identified: decayed wood, cracks, root problems, weak branch joints, cankers, architecture, branches and crown or dead individuals (^{Pokorny and Albers, 2003}).

In the Coyoacán mayoralty, there is the Viveros de Coyoacán National Park, which is one of the parks with the greatest flow of people, who use the space to carry out sports activities (athletics, mainly), family or single recreation, school education and academics, among others; it is estimated that around 2 500 to 3 000 people attend daily (^{Semarnat, 2018}). There is no formally published information on the condition of the trees and management; thus, and from the importance of the Park, the objective of this study was to identify and evaluate risk trees in the Viveros de Coyoacán National Park, in Mexico City.

Materials and Methods

Sampling design

A random sampling design of 14 circular sites representing the center of the sampling units was used, which were 500 m² (12.62 m radius). The sampling intensity was 2 % accepted for urban areas (^{Schreuder et al., 2004}) (Figure 1). These units were located in the work areas using a Global Positioning System (GPS) 12XL model Garmin geopositioner, with a <10 m error; the dimensions of each site were delimited with a longimeter and rope. All the trees within each sampling site were evaluated.

Figure 1 Location of the sampling and polygonal sites of the Viveros de Coyoacán Park, Mexico City. 

Results and Discussion

A total of 299 trees of 21 different species were counted, of which Cupressus lusitanica Mill., Fraxinus uhdei (Wenz.) Ligelsh and Liquidambar styraciflua L. represented 53 % of the total (Table 2). The most abundant stage of development corresponded to juvenile individuals (84.6 %), while 15.4 % were mature. The latter mostly corresponded to Alnus acuminata Kunth, Casuarina cunninghamiana Miq., Cupressus lusitanica, Eucalyptus globulus Labill., Fraxinus uhdei, Pinus ayacahuite Ehrenb. ex Schltdl., P. greggii Engelm. ex Parl., P. oocarpa Schiede ex Schltdl. and Ulmus parvifolia Jacq. This indicates that there is a diversity of species, but only a few dominate, and they are generally exotic, including: Ficus sp., Casuarina sp., Ligustrum sp., Cupressus sp. and Eucalyptus sp., genera that predominate in the Viveros de Coyoacán park, and also prevail in other green areas of Mexico City, with a latent risk due to pests and diseases (^{Román-Guillén et al., 2019}, ^{Saavedra-Romero et al., 2019b}).

In relation to the previous results, four species constitute 64.1 % of all the individuals inventoried. This figure is very close to that obtained by ^{Velasco-Bautista et al. (2013)} in the Bosque de San Juan de Aragón urban park, in Mexico City, where seven species represent 74 % of the total trees; and similarly, ^{Saavedra-Romero et al. (2019a)} recorded five taxa with the highest frequency (76.2 %) of plantation in the same park. Likewise, ^{Castillo-Islas et al. (2008)}, on the campus of the Universidad Autónoma Chapingo, cite that 74 % of the total population belongs to four species. ^{Pérez et al. (2018)} on the campus of the Instituto Tecnológico Superior Venustiano Carranza document 143 individuals, and only 12 taxa.

In the Tezozómoc Cultural and Recreational Park (PCyRT), 3 758 trees were counted, belonging to 30 species grouped into 16 families; the most abundant and with the greatest cover were Eucalyptus camaldulensis Dehnh., Populus tremuloides Michx., Pinus radiata var. binata (Engelm.) Lemmon, Fraxinus uhdei, Schinus molle L. and Cupressus lusitanica (^{Reséndiz et al., 2015}). The general trend is that a large number of trees of five or six genera are planted in the urban parks of Mexico City, including: Casuarina sp., Eucalyptus sp., Fraxinus sp., Cupressus sp., Ligustrum sp. and Grevillea sp. In this regard, ^{Santamour (1990)} recommends that no species should predominate above 10 %, no more than 20 % of a genus and 30 % of a family.

In the study area, it was estimated that 30 % of the trees have DN <20 cm and a height of less than 15 m, 51 % reach heights greater than 18 m; while 26 % have diameters >30 cm. The average height of the clean shaft was 5.80 m, which corresponds to the height at which the electrical wiring is located, and below this the signs are located within the park. The crown cover showed that they are unbalanced or incomplete, mainly due to the high tree density throughout the park and the lack of maintenance (Table 3).

Based on the number of trees, the number of species represented and the stages of development, most of which are juveniles, it can be stated that according to ^{Baró et al. (2014)}, is a green area that provides quality ecosystem services, such as the removal of atmospheric pollutants, carbon sequestration and storage capacity, water infiltration and noise reduction, among others.

Regarding the location of the trees, they predominated inside the gardens, which have a high density and competition for nutriments; there are trees established throughout the area of free transit for visitors; therefore, the soils have a high degree of compaction. In this regard, ^{Martins-Nieri et al. (2018)} indicate that in compacted soils the main factor considered is the loss of aeration due to the decrease in pore spaces and organic matter, therefore, the roots do not have a good development, consequently, growth is weak. ^{Hernández et al. (2011)} point out that soil compaction is the main cause of death in urban trees, since their roots need permeable or grass-covered soils to promote greater infiltration and the development of secondary roots.

The presence of some insects in Eucalyptus camaldulensis and fungi in Cupressus lusitanica, Eucalyptus globulus, Pinus ayacahuite and Ulmus parvifolia was observed. It should be noted that few studies have been carried out on pests and diseases in urban trees in Mexico City; thus, ^{Reséndiz et al. (2015)} identified in the PCyRT several species with foliage damage caused by different fungi, in particular, Fusarium sp., which exhibits chlorotic staining. In a similar way, in the Third Section of the Chapultepec Forest, it was determined that the condition of the trees, in general, is poor due to the presence of phytophagous insects that damage its vegetative structure (^{Cervantes et al., 2019}).

In the trees of Viveros de Coyoacán Park, in particular, low affectation was observed to the urban infrastructure inside, since only 12 % of the individuals interfere with the electrical wiring, 37 % with signs and 9 % obstruct the walk paths. ^{Velasco et al. (2013)} recorded in the San Juan de Aragón Forest that 7.9 % of the trees had some risk associated with electricity cables, fences, sidewalks and constructions such as fences or planters. These conditions are different from what is recorded for other urban areas, such as gardens, parks or in medians where the main damage to infrastructure is directly related to the incorrect selection of species, due to ignorance of their biology. development and poor management (^{Román-Guillen et al., 2019}).

Nevertheless, the conditions in the Viveros de Coyoacán Park are similar to those of other forests or parks in Mexico City, such as the Chapultepec Forest (^{Benavides and Fernández Velasco, 2012}) and the San Juan de Aragón Forest (^{Velasco et al., 2013}; ^{Saavedra-Romero et al., 2019a} and ^2019b), or in alignment trees (^{Román-Guillen et al., 2019}), all located in limiting conditions due to water supply, induced opening of the space below the canopy for security reasons and for recreational activities.

The characteristics that can be defined as defects associated with the trees (Figure 2) showed that just over 70 % of the trunks of the sampled population have cracks and dead wood caused by cankers, weak unions of branches, injuries due to mechanical damage or by vandalism. These factors caused the formation of cavities, which in most of them were the entrance point for pathogenic agents that caused the deterioration or reduction of the stability and the physical-mechanical properties of the standing tree, since the phytopathogens were present in branches, trunks and roots. In half of the trees, weak unions were recorded in the branches, which include the branches produced by epicormic shoots in response to injuries or environmental stress, and 60 % developed ulcers or cankers in the branches. Fractures or cracks often originated in these areas, with a high risk of causing the specimen to fall.

Figure 2 Defects presented by risk trees in the Viveros de Coyoacán Park, Mexico City. 

Regarding the degree of risk, it was determined that 131 trees of the total represent extreme risk (32 %) and high risk (12 %) (Table 2, Figure 2). Related defects are rot, cracks, weak branch joints, cankers and dead wood. ^{Pokorny and Albers (2003)} indicate that these characteristics are the main causes of a high risk for the population in any urban area. Likewise, root rot caused by fungi constitutes an additional factor for the potential fall of the tree individual (^{Calaza and Iglesias, 2016}).

Regarding the risk categories, the species with the highest percentages of red and orange codes (extreme risk and high risk, respectively) were Cupressus lusitánica (25 %), in addition to the Pinus genus (22 %), Ulmus parvifolia (13.7 %), Fraxinus uhdei (10.7 %) and Liquidambar styraciflua (9.9 %); those with the lowest percentage were Alnus acuminata, Casuarina cunninghamiana, Eucalyptus camaldulensis, E. globulus, Ligustrum lucidum W. T. Aiton and Morus celtidifolia Kunth.

In a similar study by ^{Castillo-Islas et al. (2008)}, estimated that 37 and 26 % of the trees were classified as red and orange, respectively, and the species that prevailed were Fraxinus uhdei, Cupressus lusitanica and Casuarina cunninghamiana in three high attendance areas of the Universidad Autónoma Chapingo; the foregoing implies a potential or severe risk, with rotten wood and cavities, severe root damage, weak unions between branches, tips and dead branches, high density and old specimens. A similar situation occurs in areas near the Venustiano Carranza Higher Technological Institute, where 52 % of the trees are of high risk to the population because they are in poor condition, are prone to falling branches or entire individuals on users and real estate (^{Pérez et al., 2018}).

In the present work, the highest proportion of the inventoried individuals were classified with a yellow code (47 %), that is, moderate risk (Figure 3). Finally, only 9 % corresponded to the category without any risk, and the species that made up this group were Phoenix canariensis Chabaud and Yucca elephantipes Baker.

Figure 3 Percentage of trees at risk in the Viveros de Coyoacán Park, Mexico City. 

Based upon the results, the management actions are basic to improve the conditions of the Park and provide security to the users; formation and sanitation pruning, surgeries in the hollows and cavities of the trees are recommended, in addition, carry out work to loosen the soil inside the planters. On the other hand, in future programs of planting or replacement of specimens, it is suggested to diversify the species; therefore, it is important to undertake a training program for the operating personnel, with the advice of engineers or technicians who participate in the work of the Park, with proper work and safety equipment.

Conclusions

The criteria and indicators used for the characterization of the risk trees in the Viveros de Coyoacán Park allow visually identifying and evaluating the defects and risks of the individual trees that represent a danger to the user population of the Park. The most frequent defects are rot, cracks, brittle branch joints, cankers and dead wood. The species with the highest percentages of extreme risk and high risk are Cupressus lusitánica, Pinus sp., Ulmus parvifolia, Fraxinus uhdei and Liquidambar styraciflua. There is little impact on the Park's infrastructure. This information is relevant for administrators responsible for implementing preventive measures to mitigate or avoid potential damage to Park visitors.