DEGREE OF INFESTATION AND PREFERENCES OF HEMIPARASITES IN URBAN ARBORIZATION

ABSTRACT Mistletoe is an important nutritional resource for the frugivorous ornithological fauna in several ecosystems. However, in great intensity, they can reduce the vigor of their host trees, requiring adequate management in urban afforestation, without depleting this source of food for birds. The hypothesis that there are hemiparasites with host specificity was the reason for this study. Thus, a floristic census of the trees was carried out at UFMT, campus Cuiabá; the infestation by genus of hemiparasite present was recorded, in quantity and degree of infestation, in addition to the quantification of seeds of the hemiparasite germinated without the development of the vegetative part, whose cotyledons necrotized and died. A total of 4265 tree individuals were evaluated, distributed in 46 families, in which 16,52% were identified of the genera Phoradendron, Psittacanthus, and Struthanthus, of which about 90,0% were Psittacanthus. The highest infestation occurred in 24 families, highlighting Fabaceae (28,3%) and Combretaceae (25,6%), and the species Terminalia catappa, Anadenanthera peregrina, and Vatairea macrocarpa. The presence of germinated seeds of the hemiparasite with dead cotyledons varies between tree species and between individuals of the same species. The highlight was in Rutaceae, Arecaceae, and Rubiaceae, and in the species Acrocomia aculeata, Caryocar villosum, and Genipa americana L. that present individuals with natural control potential, even with the presence of frugivorous ornithological fauna and the dispersion of the hemiparasite. In general, the intensity and degree of infestation are low, and there is no infestation in the Caryocaceae, Rutaceae, and Rubiaceae families, highlighting the species Caryocar villosum, Murraya paniculata and Alibertia edulis.


INTRODUCTION
Urban aff orestation provides ecological, aesthetic, physical and health benefi ts to the population. However, there are numerous abiotic and biotic factors that interfere with the development of trees, and among the biotic are hemiparasites, also known as mistletoe.
Its infestation can aff ect development, as it alters the architecture of the trees and can compromise the aff orestation program (Rotta, 2001). The application of techniques for its control, such as pruning, if performed incorrectly, causes injuries to the plant and its exposure to external agents and disfi gures its structure (Martins et al., 2010).
The development of the hemiparasite occurs with the adherence of the seeds to the bark of the tree with the aid of the mucilaginous substance called viscin and, in places irregular, the haustoria of the hemiparasite form an intumescence in the host (Rotta et al., 2006). This structure enters the cortex and with the anastomosis there is a connection between the phloem and xylem cells of the hemiparasite and the tree, occupying part or almost all of the canopy (Tainter, 2002). Its structure diff ers from it and is used by the fauna as a shelter, perch, nesting, hibernation or refuge from predators (Griebel et al., 2017).
The action of hemiparasites results in harmful eff ects on host plants and relevance as a resource for frugivorous animals, such as birds (Arruda et al., 2012) and mammal, Dromiciops gliroides, in Argentina (García et al., 2009). Birds are an initial fi lter for the frequency and distribution of seeds for the available hosts, while the host is the fi nal fi lter for the establishment of the hemiparasite (Arruda et al., 2012). Thus, the successful development of the hemiparasite only occurs when biologically, chemically and physically compatible with the host; these associated variables determine whether a tree is infected or not (Fadini, 2011). Consequently, the specifi city of the hemiparasite to a host.
However, when the hosts have characteristics that confer resistance to the hemiparasite, due to a cellular response by the plant, such as a hypersensitivity reaction (Király et al., 2007) or biochemical resistance to toxic substances (Pascholati and Dalio, 2011b), prevent host infestation.
Thus, the objective was to analyze the infestation of hemiparasites in the trees of the UFMT campus Cuiabá, identify the host species, quantify the degree of hemiparasite infestation and the host species that react to the infestation process.

MATERIAL AND METHODS
The study was carried out at the Federal University of Mato Grosso -UFMT, campus Cuiabá. The region has a predominance of the Cerrado biome. The climate is tropical with a dry season (Aw), according to the Köppen-Geier classifi cation (Souza et al., 2013).
The fl oristic census used the methodologies of Silva Filho et al. (2002) and Finger (2011) with collections from 2014 to 2017. The number of occurrences of hemiparasites in each individual tree was counted. The infestation was the frequency of the hemiparasite in relation to the population and the degree of infestation was the proportional area occupied by the hemiparasites in relation to the total area of the host canopy, expressed as a percentage. The degree of infestation was considered low with up to 10% of crown occupation, medium above 10 to 30% and high, greater than 30%. The criterion was associated with the maximum recommended level for removing the volume of live branches from the canopy in the application of pruning or trimming of plants (Samek, 1974).
The crown area was visually obtained by its diameter and length, associated with the shape of the crown, according to the criteria of Ramalho (1976). On the trunk and branches of each individual tree, using a monocle, the presence of hemiparasite cotyledons expanded to germinate was observed and counted, but which were dry, necrotic, dead, and lacking in vegetative growth. Data were analyzed with descriptive statistics, with the structuring of tables and fi gures. With the use of software R version 4.0.5 (R Development Core Team, 2021) the box diagrams "boxplot" for the average number of expanded and dead cotyledons per species.
The highest frequency of infestation, 14.8%, was Psittacanthus sp., both for species native to Brazil, native to the Cerrado and exotic, with a distribution of 9.46%, 71.62% and 18.92%, respectively (Table 1).
For the gender Struthanthus sp. an infestation of 2.0% was recorded, whose distribution among natives of Brazil, natives of the Cerrado and exotics was respectively 20.0%, 60.0% and 20.0% (Table 2).
The lowest frequency of infestation, 1.4%, was Phoradrendon, and for native, Cerrado native and exotic species, the distribution was respectively 7.41%, 62.96% and 19.63%. Hosts with greater abundance had less infestation by Phoradrendon, however, for the species Senna siamea, Simarouba versicolor and Terminalia catappa infestation was greater than 10% (Table 3).
Of the analyzed trees, 582 individuals had very low incidence and degree of infestation or the absence of hemiparasites. With highlight on the families Lechytidaceae, Rubiaceae, Arecaceae, Dilleniaceae, Caryocaceae and Ruthaceae. The most frequent species were Cariniana rubra, 29.5%, Genipa americana, * Origin (Or.): Native to Brazil (N), Exotic (E). Natural occurrence in the Cerrado phytogeographic domain (*); TN = total number of tree individuals per species; TNW = total number without hemiparasite; TNH = total number with incidence of hemiparasite; H (%) = percentage of infestation by tree species; **Specialized tree species: the only ones found with the hemiparasite. * Origem (Or.): Nativa do Brasil (N), Exótica (E). Ocorrência natural no domínio fi togeográfi co do Cerrado (*); TN = número total de indivíduos arbóreos por espécie; TNW = número total sem hemiparasito; TNH = número total com incidência de hemiparasito; H (%) = percentual de infestação por espécie arbórea; **Espécie arbórea especialista: as únicas encontradas com o hemiparasito. The highlight is that in 14.2% of individuals the seeds of the hemiparasite Psittacanthus germinated, as the cotyledons expanded, but they were necrotic and dead. Although the mean number of expanded and dead cotyledons is 22.93%, there is only one individual of Bertholletia excelsa with 100%, while Caryocar villosum and Caryocar brasiliense were the only ones present in the fourth quartile. Of the ten species with expanded and dead cotyledons, seven were below the median (Table 4).   The species Genipa americana, Cariniana rubra and Acrocomia, without the incidence of hemiparasites, they presented a greater abundance of specimens. However, the average number of dead cotyledons per host was 5.3%. In specimens of the Rutaceae family, there was no record of the hemiparasite or its dead cotyledons, likewise the species Attalea phalerata, Copernicia alba, Copernicia prunifera, Moringa citrifolia and Moringa oleífera.
Thus, hemiparasites are plants that occupy a habitat with a distinct spatial structure (Overton,  1994). The author mentions that this habitat is composed of hosts that are colonized trees; non-host trees that are not colonized; and an inhospitable matrix that is the area between the trees. Thus, the monitoring of hemiparasites is facilitated, which allows the measurement of colonization, recolonization and extinction. Therefore, the classic metapopulational dynamics is characterized between colonization and extinction rates of hemiparasites (Arruda et al., 2012), as observed in (Figure 1).
Diff erences in results may be associated with the greater or lesser presence of fruit trees, the intensity of fl ow of ornithological fauna and the application or not of eradicating pruning in the analyzed areas. However, in Macapá-AP, Soares et al. (2021), detected an average of 12.42% with hemiparasitism, mostly in the Anacardiaceae family. Similar result to the present study, where hosts from the same family, among others, also suff ered greater infestation.
The low infestation of hemiparasites, from an ecological point of view, contributes to the supply of food for the local fauna without harming the host plant. Meanwhile, the lowest percentage of trees with medium and high degrees of infestation would require greater urban monitoring for the control of hemiparasites. Pazini et al. (2020) state that aff orestation of the campus of the UFMT in Cuiabá is rich, with great diversity of species with ideal proportions of native Brazilian species. Corroborating with Silva and Fadini (2017), who claim that population diversity with native species can reduce the infestation of hemiparasites and costs with urban aff orestation.
Pouteria gardneri presented only Psithacanthus, already Terminalia catappa and Anadenanthera peregrina were infested by the three hemiparasites. Plant parasites can be considered keystone species, and their impacts are dictated by the host range of the parasite and by its preference and selection of particular host species (Press and Phoenix, 2005). The authors reinforce that they can change the physical environment around them and, therefore, can be considered as ecosystem engineers, which can modify the supply of resources and behavior of other organisms, within communities of parasitic plants.
Phytopathogenic microorganisms access the host, by direct penetration, natural openings or injuries, and if the reaction is not effi cient, tissue colonization will occur, necessary for the pathogens to remove the nutrients stored in the form of carbohydrates in the host tissues, for their growth and reproduction. For this purpose, the pathogen uses substances such as enzymes, toxins and hormones (Pascholati et al., 1998;Pascholati and Dalio, 2011a), in addition to genes, whose mechanisms are important to prevent the activation of the plant's defense system (Medeiros et al., 2003).
Usually, hosts with greater hemiparasite intensity in an ecological niche are competitive dominants. In this case, parasitism will facilitate the maintenance of competitively subordinate species (Press and Phoenix, 2005). In addition, the leaves of hemiparasites are rich in nutrients, and their deposition along the litter provides an increase in soil fertility (Mellado et al., 2016).
In urban aff orestation, management for plant fertilization is more complex, therefore, the lowgrade incidence of these hemiparasites adds the campus. Given that, hemiparasites can also alter the spatial and temporal distribution of nutrient cycling while increasing the amount of nutrients available to host plants (Arruda et al., 2012). These authors point out that in nutrient-poor ecosystems, such as tropical grasslands and savannas, their presence was considered of great importance. According to Ndagurwa et al. (2018), the soils under hemiparasite hosts showed higher levels of silt, organic matter, sodium, potassium, magnesium and nitrate.
Pruning branches with hemiparasite infestation is recommended when the incidence in the crown is greater than 50%, or replacing the individual with autochthonous species with greater resistance (Coelho et al., 2020). However, individuals with an infestation greater than 30% may have a developmental defi cit and reduced vigor. When there is no removal of hemiparasites, such as Psittacanthus spp. and Struthanthus spp., there may be crown covering, aff ecting the photosynthesis of the host tree (Caldeira, 2013).
Although some hemiparasite species are harmful agents to plants, most do not aff ect economically valuable crops and forest products, but play key roles in forest ecosystems (Mathiasen et al., 2008). The spatial distribution of hosts and the high incidence of light in isolated host trees can lead to high prevalence in urban areas (Maruyama et al., 2012). The authors argue that the eradication of hemiparasites in urban areas should be replaced by ecological management, their importance for the conservation of bird species should be considered and studies should be carried out to determine which bird species are favored by hemiparasites in these regions.
The specifi cities of the hemiparasites regarding their host are highlighted in the results found. Some hemiparasites have a greater number of hosts in diff erent tree families, while others manage to parasitize only one host species (Mathiasen et al., 2008). In addition, the authors point out that some hemiparasites have host specifi city, and some of them are prone to parasitize other hemiparasites.
Generally, non-host species have prominent spines, thorns or the presence of lenticels on stems, leaves and/or fruits. Especially in some species of the Caryocaceae, Rutaceae, Rubiaceae, Lecythidaceae and Arecaceae (Table 4). It is believed that such structures make it diffi cult for birds to land and reduce the dispersion of the hemiparasite. In addition to the angle of the branches, roughness and thickness of the host bark, which are important characteristics to determine the frequency of distribution of hemiparasites in the host, do not interfere with its diversity (Fadini 2011;Wagner et al., 2015). In addition, by choosing tree species with these characteristics, it can be a way to avoid hemiparasites in urban aff orestation.
In species that did not show developed hemiparasites and dead cotyledons (Table 4), such as Genipa americana, Cariniana rubra and Acrocomia aculeata, it was initially interpreted as a reaction to the hemiparasite. According to Király et al. (2007), for pathogenic microorganisms, the hypersensitivity reaction is shown to be an extreme cellular response on the part of the plant, which can cause the high degree of resistance to the pathogen, however, no information was found that can associate this fact with the absence of hemiparasites.
Thus, the presence of dead cotyledons is an indication of the frequency of avifauna in these hosts, and the non-infestation of hemiparasites. The bird lands and the deposition of the hemiparasite seed occurs, the seed swells with water absorption, the cotyledons expand, but the seed necroses and dies. In a study with germination of hemiparasite seeds, it was not possible to observe the eff ect of tree species, whether host or non-host (Maruyama et al., 2012). However, the authors observed that the radicle of the germinated seeds did not manage to penetrate the bark and the seedlings invariably died in the non-host species. According to Pascholati and Dalio (2011b), some trees have the ability to delay or prevent the entry and development of a foreign body in their tissues.

Plants
can defend themselves against phytopathogens either passively or actively, and their resistance mechanisms can be subdivided into two categories: pre-formed and post-formed. In the fi rst case, the plant resists through existing mechanisms, such as the cuticle, which works as a structural barrier, and toxic substances. If, even so, the phytopathogenic agent overcomes the surface structures of the host, it continues to defend itself with new structural barriers, with the formation of halos and lignifi cation around the penetration site (Pascholati and Dalio, 2011b).
These authors also report the biochemical resistance of the host, with phytoanticipins being examples of pre-formed substances, and can be converted into extremely toxic substances, while phytoalexins are post-formed substances in response to the presence of the pathogen. In addition to morphological mechanisms related to the noninfestation of the hemiparasite, the possibility of reaction mechanisms should be studied in relation to the origin of the process, with anatomical studies, as well as the chemical constituents of these substances, similarly to what is understood for the diseases caused by phytopathogens. Gummosis is often produced by plants as a protection mechanism when their cortex is attacked (Andrade et al., 2013), observed in Anacardium occidentale (Figure 1).
The fact that individuals of the same species present diff erent behaviors in relation to hemiparasite infestation may be related to the characteristics of the specimen and not the species. Similar to disease resistance, whose expression is not morphological but physiological, and behaves as a variable characteristic in the population and specifi c in some individuals.
The subject lacks studies and, once the individual characteristic is proven, the vegetative multiplication of individuals that are not infested by the hemiparasite is possible. These will be able to compose the vegetation with other individuals of seminal origin and that, as hosts of the hemiparasite, will exercise the function of off ering this source of food to the ornithological fauna.

CONCLUSIONS
The presence of hemiparasites varies between tree species and between individuals of the same species. In some individuals there is the presence of fauna, the deposition of dispersion units of the hemiparasite without vegetative development. These individuals have potential for the natural control of the infestation and for researches that clarify the morphological or physiological process involved. In the aff orestation of UFMT, campus Cuiabá, the intensity and general degree of infestation by hemiparasites are low and variable for the genders Struthanthus, Phoradendron and Psittacanthus and contributes to the availability of food for the frugivorous ornithological fauna in the urban area.

AUTHOR CONTRIBUTIONS
Theonizi Angélica Silva Albuês designed the research Project, wrote the paper obtained and discussed the statistical, experimental, and estimated data. Daiane de Moura Borges Maria and João Paulo Sardo Madi wrote the paper and supported the data analysis. Kamila Daiany Terres da Silva supported the data analysis and the obtaining of experimental results. Sidney Fernando Caldeira obtained the experimental results and supported the writing of the paper and statistical procedures.