This study used comprehensive surveillance data to profile RIFA invasions in time and space on an isolated island. By using this surveillance data, which were collected regularly together with information on land-use in different years, distinctions of RIFA severity can be compared, and RIFA SIRH were therefore identified. Our statistical model decomposed the spatial invasion risk into four geographic and anthropogenic factors: land-use characteristics, distances from RIFA sampling location to the nearest road, and spatial factors. For land use from 2014 to 2017, agricultural land, transportation usage, and land-use change had significantly higher odds of RIFA SIRH than natural land cover. Regarding the distance from the nearest road, RIFA invasions were most likely (> 60%) to occur within 350 m from the nearest road on the transportation usage land. Meanwhile, it was likely (> 60%) to have RIFA invasions within 150 m from the nearest road in areas where land-use change had occurred between 2014 and 2016. Finally, the highest risks of RIFA SIRH were identified around the pier area and the area of the earliest RIFA invasions on Kinmen. Our study provided an example showing how RIFA gradually expanded to the entire isolated island.
Highest risks for agricultural land, transportation usage, and land-use change
Agricultural land
The vulnerability of agricultural lands to RIFA invasions has been reported in many studies. For example, a review by Apperson and Adams showed that RIFA often infested soybean fields in the United States28. Way and Khoo reviewed the RIFA infestation of crop plants, including sugar cane and cotton29, and indicated that crop invasion by RIFAs was a common occurrence. The study conducted by Stuhler et al. demonstrated that in unthinned patches, RIFA mounds were likely to occur in agricultural lands compared to woodlands in South Carolina30. Thus, the results of our study align with the literature in finding that agricultural land tends to be highly assailable by RIFAs.
The large majority of agricultural lands on Kinmen Island include sorghum farms, peanut farms, and other food crop farms31. These farms need to be plowed or cultivated at least twice per year. Therefore, soil disturbances by humans could be the reason for the defenselessness against RIFA invasions. The potential mechanism is that soil disturbances destroy habitats for all living organisms, including RIFA. However, RIFAs reestablished their colonies faster than others30,32. Thus, RIFAs became one of the dominant species in highly disturbed areas. Higher soil disturbances associated with higher RIFA abundances were evidenced by the study by Stuhler et al.30 in which the authors compared the thinned areas to unthinned areas, identifying more RIFA mounds in thinned plots. King and Tschinkel also conducted a field experiment on different levels of soil disturbances. They demonstrated that higher numbers of RIFAs persisted at higher levels of disturbance (i.e., plowing) than at lower levels (i.e., mowing)32.
Land for transportation usage
The land-use type for transportation purposes, including roads and ports (i.e., seaports and airports), was also identified as a risk factor for RIFA SIRH in this study (Table 2). Among the 1814 sampling tubes in the transportation area, there were 1768 sampling tubes for roads and 46 for ports. As most of the sampling tubes were set along roads in the present study, it could be deduced that roadsides or road cuts were at risk of being infested by RIFA. This result was in compliance with previous studies in the U.S., showing that areas beside roads such as roadsides and road margins provided suitable habitats for RIFA development11,33,34,35,36,37.
Roadsides or road cuts had significant risks of RIFA SIRH in Kinmen, which could be due to frequent disturbances from vehicles. In Kinmen, most roads have only one lane or two narrow lanes. When two vehicles traveling in opposite directions pass each other, they will sometimes take turns or pull over onto the side, resulting in frequent soil disturbance. Roadsides or areas near roads are generally considered highly disturbed10,11,34,38, and narrow and disturbed areas suitable for RIFA establishment were demonstrated by Stiles and Jones12.
In addition to disturbances along roads, some vehicles may also transport RIFAs in potted plants and soil. Newly-mated queens may potentially attach to the surface of vehicles and fall during transportation, further facilitating invasions near roadsides. This traffic-related dispersal process has been documented in many plant species39,40,41.
Road maintenance could also be a reason for the high risks near roadsides. Road maintenance involves moving soil from one place and adding soil to construction sites. If the transported soil is contaminated by RIFAs, the maintenance areas will likely be occupied by RIFA. A case report by King et al. revealed how RIFA spread to roadsides by road maintenance32.
Ports, in addition to roads, are another land type for transportation usages. Our finding was in line with previous studies showing that airports or seaports were common areas of RIFA invasion in Taiwan and neighboring countries. For example, Taoyuan International Airport was considered one of the earliest RIFA infestation locations in Taiwan42,43. RIFAs were also detected in container yards in Taiwan’s Kaohsiung commercial port in 201844. In other Asia–Pacific countries, such as China, South Korea, Japan, and Australia, RIFAs have also been reported at ports in the last decade44,45.
Ports in this study consist of one seaport and one airport (Fig. 1). Based on the predicted risk of RIFA SIRH (Fig. 8a), one of the highest risk areas was around Shuitou Pier in Jincheng township (Fig. 1). The Pier area had high risks could be because it is one of the cargo container entrances on Kinmen Island. Shipping cargo containers have been suggested to facilitate the movement of RIFAs from abroad or between domestic ports42,43,44. Container yards can become infested when RIFA-contaminated cargo containers are unloaded44,46. In addition to possible contributions from cargos, the pier area had high risks of invasions, which could be due to environmental conditions. This can be supported by the risk of spatial factors, showing that the Pier area had high risks (Fig. 8c). One of the possible environmental factors could be that floating rubbish tends to accumulate in the Pier area47. Studies have shown that nonnative species, including ants, can travel with marine litter to new locations32,48,49,50,51.
The Kinmen Shangyi Airport is the other cargo entrance in Kinmen (Fig. 1). Intuitionally, because of cargo containers, the airport area was expected to have risks similar to those in the pier area; however, the risks of RIFA invasions in the airport area were considerably lower (Fig. 8a). The differences in risks could be due to their cargo carrying capacities. In 2018, the airport had 6778 tons of cargo, but the pier had one million tons of cargo52,53. Differences in the types of cargo between the two locations may also play a role in invasion risks. From 2001 to 2018, the majority of goods arriving at the Pier included building stones and block stones from China53. These products have higher risks of being contaminated by RIFAs than goods such as ferrous articles and eggs arriving from the airport of Taiwan53,54.
Land-use change
The land-use change category was identified as a risk factor for RIFA SIRH in the current study. Among land-use change areas, 61.6% were natural land cover in 2014 but were converted to agricultural land, transportation areas, and artificial structures in 2017, which we designated development-related areas (Fig. 6).
As previously mentioned, the reasons why the land-use change category had a high risk of RIFA invasion could be due to anthropogenic disturbances. Taking development-related areas as an example, when natural land cover such as forests are changed to other land usages, the first step may be to remove vegetation by clearcutting or plowing. These activities involve soil or habitat disturbances and could aid in the establishment of RIFA populations55. Then, if lands are changed to build houses or schools (i.e., artificial structures), soil disturbances could also occur during construction activities56. For lands that are changed to transportation usages, moving and adding RIFA-contaminated soil could occur during road construction.
Effects of roads on RIFA SIRH
Distances to the nearest roads were important for understanding invasion where undergoing land-use change, as well in places used as transportation lands (Fig. 7). These land-use categories share a common feature: roads. Meanwhile, agriculture lands had the greatest level of RIFA SIRH, but did not show interaction with distance to roads (Table 2). This could be because agriculture lands were far from roads as compared to land-use change and transportation lands. The median distances to roads from these three land-use categories supported this speculation. Therefore, from this study, it can be deduced that the roads could play a role to transport RIFAs to areas closer to road (i.e., land-use change and transportation). However, the effects of roads on RIFA SIRH did not appear when the areas away from roads (i.e., agricultural lands).
Lowest risk in natural land cover
In the present study, natural land cover were identified as the lowest risk category of RIFA SIRH among the five land-use categories (Fig. 8d). This finding was in line with the study conducted by Brown et al., showing that a high percentage of canopy cover was associated with a low mean number of RIFAs in Texas between 2008 and 201057. In addition, Tschinkel and King investigated longleaf pine forests in Florida in 2012 and found that RIFA had difficulty establishing long-term colonies in the forest35. However, in another longleaf pine forest in Georgia, the ant survey conducted by Stuble et al. revealed that RIFAs were the predominant species in the ant community from 2006 to 200758. Wetlands also had high numbers of RIFAs. In northern Florida, Tschinkel observed that RIFA mounds clustered near pond margins11.
Natural land cover in Kinmen had the lowest risk of RIFA invasions, which could be because most areas (> 75%, data not shown) are forests. The forests are preserved and protected by the Forestry Bureau of Taiwan. Because of protection, forests can avoid most anthropogenic disturbances, such as soil excavation, which are known as one of the factors facilitating RIFA relocation32,59,60. Additionally, the forest environment is cool, humid, and shaded, which are unfavorable environmental conditions for RIFAs1,12,30,34,61,62.
Implications of study findings for RIFA management in Kinmen
Public communications
To date, the Kinmen County Animal and Plant Disease Control Center (KAPCDC) has launched a program aimed at raising public awareness of RIFAs on the island through newspapers, social media, and posters. In addition, for RIFA control, the KAPCDC has listed certified pesticides such as pyriproxyfen and lambda-cyhalothrin for the use of controlling RIFAs on agricultural lands. Nevertheless, our study documented that a greater risk of RIFA invasions still occurred on agricultural lands and lands used for transportation, suggesting communications should target owners of agricultural lands as well as the general public in future campaigns. Many individuals of the general public may not be able to identify ant species, so communications should therefore emphasize the importance of reporting any ant mounds, especially along roads. As different sociodemographic groups react to source information differently, communications have to be tailored to ages and educational levels7. For example, for students in primary school, the study by Madeira et al. showed that by teaching activities including insect specimens and short-film presentations, students increased their awareness of the importance of pest control63. For owners of agricultural lands and workers at ports, educational activities on basic RIFA knowledge and pesticide treatments with suitable communication methods may be needed. Those methods included regular face-to-face discussions on RIFA elimination strategies in the meetings of farmers’ associations or a system sharing updated materials likely to be contaminated with RIFAs64,65.
RIFA control personnel
To prioritize resources, according to the findings from this study, we suggest that government staff focus on the controls within 350 m from the nearest road on transportation usage land and within 150 m from the nearest road on the areas where land-use change occurred between 2014 and 2016. The authorities could consider integrated pest management approaches, which include chemical and biological controls, to preserve the local ecosystem66.
For agricultural lands, RIFA management mainly relies on awareness and reports from owners, as control personnel cannot perform inspections and intervention on private agricultural lands without the owners’ permissions, Although control personnel cannot directly perform interventions on private land, plant quarantine officers in seaports, which were a high-risk area in this study, can prevent RIFA importation by checking cargos to ensure that RIFAs are not stowaways on materials such as plants, rocks, and soil.
Source: Ecology - nature.com
