Degree of urbanization and predation pressure on artificial lepidopteran caterpillars in Cali, Colombia

Growing urban expansion results in the alteration of ecological processes (i.e. predation) within trophic networks. Predation on herbivores is known to vary with the size of the area covered in vegetation, successional stage, altitude, and the structure of the predator community, but there are gaps in information regarding how this occurs in urban and suburban environments. The purpose of this study was to determine whether the predation pressure on artificial models of lepidopteran larvae varied with degree of urbanization, type of substrate, and group of predators (birds or arthropods) in Cali, Colombia. Five hundred and eighteen artificial larvae were placed in two areas of the city (urban vs. suburban) and in two types of substrate (leaf vs. stem) for 30 continuous days and with two replications over time. Total predation was measured as the number of models with evidence of attack by predators. The overall incidence of predation was 24.13%, and was significantly higher in the urban area (63.20%) when compared to the suburban area (36.80%). The leaf substrate was attacked significantly more than the stem (60% vs. 40%). The proportion of attacks carried out by birds was significantly higher (74.40%) than that carried out by arthropods (24.80%). Together, these results suggest that the incidence of predation varies with the disturbance caused by urbanization and by the type of substrate in which prey organisms are found. In addition, the study confirms that birds are the main controllers of herbivorous insects in urban environments.


42
The drastic transformation of pristine natural landscapes into fragments within urban 43 landscapes [1][2] causes changes in community composition, structure, abundance, and 44 trophic relations [3]. The size of these fragments of vegetation and the degree of 45 disturbance in the interior and surrounding areas affect the permanence of species that 46 require large continuous areas of habitats and particular food resources [4][5][6]. When species 47 disappear from a habitat, predator-prey relationships change and some species are 48 temporarily favored. For example, herbivore communities undergo rapid growth in habitats 49 where populations of its natural enemies are low or have disappeared due to human induced 50 changes [7,9]. 51 Predation pressure varies both within and between habitats due to differences in 52 community and predator density, vegetation structure and complexity, and intensity of 53 anthropogenic disturbance [2, 10-11]. Compared to the interior, the edges and patches of 54 forest fragments are considered as areas of greatest predatory risk for many species [5, 8, disturbance. The degree of prey exposure also influenced detection by predators. For 60 example, predatory rate on artificial larvae located exposed on leaves was greater compared 61 to those that were hidden [13].

62
Usually predatory events happen quickly and are often hard to measure because 63 predators may hide while consuming prey, or they may be nocturnal leading to lower 64 detection. Similarly, the degree of predation is complex to measure because, frequently no 65 traces or only fragments of the consumed prey are found. One technique to measure 66 predator pressure is using dyed prey organisms that allows tracking of prey. However, one 67 of the disadvantages of this method are the low number of individuals effectively tracked, 68 time-intensive observation, and difficulty in identifying the natural enemies of prey [14].

69
To resolve these difficulties, we can use artificial plastic prey that conserve marks of 70 attacks received by natural enemies on their surface. This method has been used to make The city has a mean temperature of 24°C, average annual rainfall of between 1000 and 88 2000 mm, and a bimodal climate with dry periods during January-February and July-  butterfly larvae, were made from odorless, non-toxic modelling clay (Fig 1).  Initially we carried out a pilot test to evaluate whether predators at these sites 104 responded to the artificial models and whether they were attracted to a particular shape 6 105 (elongated or spherical). For this test, 20 P. dulce separated by a distance of at least 30 m 106 were used in each area (urban, suburban). Ten trees received sphere-shaped models (five on 107 stems and five on leaves), and 10 received elongated larvae (five on stalk and five on leaf).

108
Five models were placed on each tree. This test lasted for 30 days.

109
For the main predation test (PT), ten trees were selected and five artificial larvae were

161
Predation tests 162 In the pilot test, 23.48% of the models had predator marks and significant differences 163 were detected between the proportion of attacks and the type of model and predator. 164 Elongated models (n = 36, 62.07%) were attacked significantly more than spherical (n = substrate showed a greater number of attacks than the stem (60% vs 40%). Season (rainy-172 dry) had no significant effect on incidence of predation nor was there an interaction effect 173 among the variables used (Table 1). Birds were the most important predators (74.40%) 174 followed by arthropods (24.80%) ( Table 2). Only one mammal attack was registered 175 (0.80%). There was a positive correlation between the abundance of birds (0.95) and the 176 number of attacks on the models; the contrary was found for arthropods (0.47) (Fig 3).    The greatest incidence of predation on lepidopteran larvae models in the urban area 213 when compared to the suburban area could be related to the increased abundance of birds 214 associated with urbanization. In urban settings, the abundance of some bird species 215 increases due to the absence or reduction of the predators that control them; in these 216 environments, the survival of predators such as snakes, or birds of prey diminishes.  higher level of exposure of artificial larvae increased their incidence of depredation; those 277 exposed on leaves were attacked more significantly than those hidden in rolled up leaves.

278
In some environments, stem predation was found to be 1.9 times greater than on leaves perceived as harmless by predators [41,59] to models in the shape of their prey.

338
Our results suggest that predation pressure on a prey organism can vary significantly according to Colombia. Bird families putative predators of the plasticine models (X-axis) and the number of 352 species for each of these families (on the Y-axis).