Introduction
Human-wildlife conflicts is a state of hostility or fight in which either one part has an adverse effect on the other (Redpath et al., 2015). The human-opossum conflict is a current concern in metropolitan areas of the Antioquia province (Colombia) (Delgado, 2007) as reported by Fundación Zarigüeya (FUNDZAR), a Colombian NGO aimed to increase welfare standards of opossums. From 2018 to 2020 FUNDZAR received 3,008 opossum-related calls from citizens asking for advice. From this total, 62.6% were road-kills, 14.9% were dog or cat attacks, 10.1% were hard-objects hitting, and 9.7% orphaned opossums (F. Flórez FUNDZAR director; personal communication, June 15, 2021).
The common opossum (Didelphis marsupialis, Didelphidae) is a neotropical mammal distributed from Mexico to Argentina. Opossums have a highly unrestricted and opportunistic diet, are skillful in arboreal and terrestrial environments, and have high reproductive potential of up to 10 youngsters twice a year (McManus, 1970; Flórez-Oliveros and Vivas- Serna, 2020). Due to these characteristics, opossums are ecologically successful in a wide range of habitats (McManus, 1970; Sunquist et al., 1987; Vaughan and Hawkins, 1999). In addition, deforestation in Antioquia is high and it is associated to pasture establishment, urban expansion, and wildland fires (González- Caro and Vásquez, 2017). These conditions are difficult to manage, and the human-opossum conflict may worsen (Rueda et al., 2013) if no practical solutions are implemented to discourage their encounters.
Several repellents have been proposed to mitigate wildlife-human conflicts, including acoustic, visual, odor, electric, and irritant methods (Mason, 1998; Gerisoli and Pereira, 2020). Electrical fencing, trip alarms, and warning calls have also been used. However, they are expensive and not viable in the long term (O’Connell-Rodwell et al., 2000). Other methods, such as burning animal feces with ground chillies to produce a noxious smoke, are more effective and inexpensive (Osborn and Parker, 2002). Plant oils are also used as olfactory repellents, and wolf urine as anti- predator to repel deer; however, these were not effective perhaps due to rapid habituation (Elmeros et al., 2011).
To our knowledge, no odor repellents have been systematically tested on opossums; thus, herein we tested citronella, ammonia, and creolin, as potential odor repellents. We set camera-traps to estimate the frequency of opossum visits to the study sites in two localities.
Materials and Methods
Ethical considerations
This study was approved by the Committee on Ethics in Animal Research of the Universidad de Antioquia, Colombia (Act 114, December 5th, 2017).
Study sites
We chose a total of six sites in two localities in the Municipality of Envigado, Province of Antioquia (Colombia), to test the effectiveness of three chemicals as odor repellents for opossums. Four sites were located at a natural forest named La Morena Ecologic Reserve (Morena 1, 2, 3, 4), and two sites at the urban park La Heliodora (Heliodora 1, 2). Morena is a 37-hectare peri- urban forest located in the rural area named El Escobero at 2,200 m.a.s.l., while Heliodora is a 24-hectare park at 1,575 m.a.s.l. (Alcaldía de Envigado, 2016). We chose four sites in Morena and only two in Heliodora because the latter is more visited by the local community, thus there was a higher risk of losing the equipment. We installed one camera-trap per site (three Bushnell Trophy Cam Essential 12mp, two Bushnell Trophy Cam Hd Aggresso 14mp, and a Cuddeback 20mp IR plus) which was set to record 20-second videos with 10-second intervals, and high sensitivity to movement. The camera-traps recorded the presence of opossums and other species from April to October 2018. The total number of sampling days slightly varied on the sites due to rough environmental conditions.
Chemicals and odor devices
Three chemicals were used in the study: citronella, ammonia, and creolin. Citronella (CAS registry number 8000-29-1) is a water- insoluble oil extracted from an aromatic plant (density 0.85 g/mL). It has a light-yellow color, citric smell, and is efficient as insect repellent (Sharma et al., 2019). Ammonia or ammonium hydroxide (CAS registry number 7664-41-7) is a solution of NH3 in water (24-28%), colorless and highly irritant. Citronella and ammonia were bought from local chemical retail stores, e.g., Quimicos JM S.A., and Protokimica. Creolin (CAS registry number 12751-04-1) is a natural over-the-counter disinfectant mainly composed of phenol (17%). It is commonly used as antiseptic to clean wounds, bathrooms, and barns at low concentrations, although in higher concentrations can cause severe toxicity (Vearrier et al., 2015). This chemical was purchased from different labs with the same phenol concentration (e.g., Fenolgan® produced by Farmagan Colombia S.A.S. laboratory, or -Específico® produced by Rotam-Vet Colombia).
To contain the chemicals, six devices were prepared using inexpensive materials. They consisted of five-gallon plastic containers (40 cm diameter) with four 2-cm holes below the lid. Then a polyurethane foam strip (100 × 4 × 4 cm) was wrapped inside the container and the liquid chemical was poured to cover the foam so its gases spread out of the container through the holes.
Experimental design
All sites were initially baited with fruit (ripe mango or plantain, wrapped in a piece of veil and hanging from a tree) or canned sardines to habituate the resident opossums for approximately three to four months. Immediately afterwards, the chemical devices were installed besides the bait, one device per site for two weeks. A new cycle started by removing the device and baiting again for two weeks, and then adding a different chemical for two weeks again. Each of these cycles were repeated on each site only changing the chemical until all three chemicals were tested on each site. Citronella was tested first, then was ammonia, and lastly creolin (Table 1). The cameras permanently recorded the activity at each site, and every 14 days they were checked to retrieve the videos and identify the wildlife species visiting the sites. Site Morena 1 was chosen as a control site and the device installed there had no chemical inside. The control site was used to test that the container itself did not affect the frequency of opossum visits.
Table 1
Experimental design of the study indicating the treatment (only bait or bait+chemical) used in each site.
Data analysis
There were nights with several videos obtained at different times; however, it was not possible to differentiate whether the opossum was the same or a different individual as many videos were recorded within 60 min at the same site. Thus, to be conservative, we counted only the number of nights with opossum records (and nights without opossum records) regardless of the nightly number of videos obtained. The capture success was calculated by dividing the number of nights that opossums were recorded by the number of camera trap-nights and multiplying the result by 100 (Srbek-Araujo and Chiarello 2013). Fisher´s exact tests of independence were used to test the null hypothesis that the proportion of opossums/night (number of nights with opossum/number of nights without opossums) without the chemical (citronella, ammonia, or creolin) is not different than the proportion of opossums with the chemical. This null hypothesis would indicate that the chemical did not repel the opossums. P-values below 0.05 were considered significant to reject the Ho and accept the alternative hypothesis that the proportions varied (i.e., that the chemical repelled the opossums). Fisher's exact test is more accurate than the Chi-square test when the expected numbers are small (i.e., when sample size is less than 20 in a 2×2 contingency table; McCrum-Gardner 2008). Fisher´s exact tests were done using the function fisher.test in package stats version 4.0.2 in R environment.
Results
Repellent effectiveness
We compared the proportion of opossums recorded (nights with opossums/nights without opossums) using bait alone vs. bait plus one of the three chemicals expected to drive the opossums away. The proportions were statistically different at some sites with ammonia and creolin, but not with citronella (Table 2). Specifically, there were less opossums when ammonia was added at Heliodora 2 and overall (counting all sites together; p<0.0008 and p<0.0007, respectively); and there were less opossums when creolin was added at Morena 4 and overall (p<0.0003 and p<0.0113, respectively). Citronella showed no effect on the presence or absence of opossums at either of the sites tested nor overall (p>0.05 all sites and overall).
The results were significant in spite that other factors also affected the proportion of opossums recorded (i.e., the sampling site and the bait used). There were more opossums at the urban park (n=43, in two sites) than at the peri-urban forest (n=17, in three sites) when bait alone was used (p<0.0001). Also, overall, there were 38% more opossums recorded when sardines were used as bait (21 opossums/187 nights) compared to fruits (14 opossums/331 nights; p<0.007).
Nocturnal activity of opossums
Although the chemical devices were let in place from May to October 2018, the cameras stayed in place longer, until February 2019. In this way we obtained a total of 244 videos recording opossums at five sites (Heliodora 1 and 2, and Morena 1, 2 and 4). The activity peak of opossums was recorded by midnight in both the urban park and the peri-urban forest, starting by 18:00 hours and ending by 8:00 (Figure 1). Although there seems to be a tendency to start the activity early in the afternoon and finish late in the morning at the urban site, we found no statistical difference in the frequency distributions between both sites (p=0.297).
Richness of wildlife species
We recorded a higher richness of species at the peri-urban forest than at the urban park after pooling the results from all sites within each one. This is, we recorded nine different species of mammals of medium and large body size (including Dasypus novemcinctus, Potos flavus, Eira barbara, Mustela frenata, Cerdocyon thous, and Herpailurus yagouaroundi), and nine birds at Morena. At the urban park we recorded three mammals of small and medium body size (Notosciurus granatensis and Cerdocyon thous) and three birds (Table 3).
Table 2
Number of days with opossums/number of days without opossums, discriminated by site and treatment (bait or bait+chemical).
[i]Bold p values reject the null hypothesis and suggests the chemical had a repellent effect, e.g.: *Ammonia decreased the number of days with opossums (from 10 to 0) and increased the number of days without opossums (from 10 to 16). NA = not possible to test the potential repelling effect of the chemical because there were no opossums to repel.
Figure 1
Nocturnal activity of opossums recorded at Heliodora urban park and Morena peri-urban forest.
Opossums were observed at all sites in Heliodora urban park and at Morena peri-urban forest (except at Morena 3), and they were always among the top three species more common at both sites. However, and as expected, opossums were more common at the urban park than at the peri-urban natural forest. This is, we recorded opossums in 42 nights using 324 camera trap- nights at Heliodora (capture success = 12.96%), compared to 30 nights using 768 camera trap- nights at Morena (capture success = 3.9%). The real number of opossums was probably higher given that we counted only the number of nights with at least one video record; not opossum records to avoid counting the same individual several times.
Discussion
When data from all sites was pooled together (regardless of the site), ammonia and creolin, but not citronella, had a significant effect to repel opossums even when bait was still available. However, in some sites (i.e., Morena 2, 3 and 4 Table 2) no opossums, or too few, visited the site tested, making it impossible to test the chemicals at those specific sites with no opossums. Real situations of human-opossum conflict are the scenarios where these chemicals should be tested next without having to bait the area to attract opossums. Ammonia and human urine have been used successfully to deter and drive away black bears when humans encountered them in Montana (Hunt 1977), and red pepper (capsaicin) sprays in Alaska (Smith et al., 1998) as well. Furthermore, it should be considered that the effectiveness of any chemical tested may be affected by several factors, including weather conditions such as rain, relative humidity, wind temperature, variability of the opossum population density throughout the year (Mason, 1998; this study), and the bait used (this study).
Chemical repellents act in different ways, they may produce sensory irritation, semiochemical mimicry (e.g., pheromones or allomones), or digestive malaise (Brown et al., 1970; Borden, 1989; Mason, 1998).
Ammonia gasses can be very harmful to humans; they may cause larynx blocking and lung distension and congestion (ATSDR, 2004). Creolin is similar to ammonia in that exposure to phenol, its main constituent, may be rapidly absorbed through the skin, respiratory and digestive systems, conducing to a systemic toxicity (Vearrier et al., 2015). Thus, similarly to the effect of capsaicin -the active component of chili peppers- on some mammals (Norman et al., 1992; Smith et al., 1998) the repellent efficacy of ammonia and creolin in opossums was likely sensorial; its efficacy may lie in that they produce irritation and short-term pain.
Table 3
Richness of wildlife species recorded during the study at Morena peri-urban natural forest and Heliodora urban park.
Its avoidance is immediate, no learning is required, and adaptation to learn its avoidance is minimal (Mason, 1998; Osborn and Parker, 2002). Further toxicity effects on animals are unknown.
On the other hand, although citronella is mainly used as mosquito repellent (Muller et al., 2009), finding that it does not repel opossums supports the thesis that irritants are effective within some taxa (insects), but rarely among others (mammals or birds) (Mason, 1998). In addition, citronella is used as a fragrance ingredient in cosmetics, and it has antibacterial, antifungal and antiparasitic properties (Sharma et al., 2019; Kamal et al., 2020) that seem desirable rather than unpleasant.
An important question is whether opossums could get habituated to chemical repellents. Previous studies have shown that mammals can get habituated to pungent chemicals, such as wolverines (Gulo gulo) to lambs carrying a mixture of olfactory aversive oils in a dispenser attached to the neck and ear-tags. This result however was observed when wolverines did not have untreated lambs as an alternative prey (Landa et al., 1998; Landa and Tømmerås 2015; Smith et al., 2000). Whether opossums can get used to ammonia or creolin would need to be properly tested. Meanwhile, if chemical repellents are to be used, it would be advisable to use them only when really needed and not as a preventive measure for long periods of time.
Finally, the wildlife diversity recorded in our study was known already in Envigado and the Área Metropolitana (Alcaldía de Envigado, 2018). The urban park held lower species richness, but opossums were relatively more common in comparison to the natural forest. This was expected given that opossums are highly unselective in their food habits and may take advantage from any resources (McManus 1970), including trash cans in urban parks (personal communication from park rangers). They also have been reported in diverse environments associated to humans, such as crops and roads (Orjuela and Jiménez 2004). This finding is rather important because opossums may have been underestimated in their ecological role in urban environments as evidenced by the current human-opossum conflict going on in Antioquia province (FUNDZAR personal communication). It is common species -not species richness- the ones expected to shape the environment by having more interactions with other species and the habitat itself (Gaston, 2010; Winfree et al., 2015). Populations of common species may decline because they are the first to suffer from any pressure on biodiversity (Gaston, 2010), and even more serious is the case of species in direct conflict with humans. Thus, future studies to monitor opossum abundance are desirable.
In conclusion, ammonia and creolin have the potential to repel opossums. Citronella, on the other hand, had no effect on reducing the number of opossum visits. These chemicals were tested in natural habitats, thus next they should be tested on real human-opossums conflict scenarios to probe their repellent potential. However, caution is warranted given its irritant, flammable, and corrosive properties. Nonetheless, we suggest that any management action to improve the welfare of mistreated opossums should be accompanied by education so we, humans, learn to value and cohabit peacefully with synanthropic wildlife.