Introduction 

Invasive species are recognized as one of the main threats to European forests, especially at regional level ([29]). One of the most dangerous alien tree all around the world is Ailanthus altissima (Mill.) Swingle (hereafter A. altissima), which is particularly widespread in temperate and Mediterranean regions ([24]). The species was introduced from China to Europe in the second half of the 18^th century, and was initially used as an ornamental plant but later also for the urban forestation and for erosion control of steep slopes ([4]). Nowadays it is considered a major invader in many parts of the United States ([2]), in most of temperate and Mediterranean Europe ([25]), and throughout Italy ([14]) including Sicily ([4]). A. altissima has a wide ecological range, due to its very high growth rate, effective reproductive and dissemination systems, and general lack of natural enemies ([24]) in its secondary range. For these reasons A. altissima can compete effectively against the native flora in disturbed and degraded sites, such as roadsides and other urban areas, including archaeological sites ([13]). A. altissima has also invaded natural habitats, along forest edges and open gaps ([23]), especially in the presence of anthropic disturbance. For instance, occasional fires favors A. altissima spread because of its impressive capacity of resprouting by injured stumps and superficial roots ([26], [17]). In addition to damaging ecosystems, A. altissima causes serious problems for forest management ([26], [15]) because it competes directly with the woody species of silvicultural interest and makes forest utilization more difficult and expensive. Under favorable conditions, A. altissima can form almost pure stands, which obviously reduce biodiversity and change the abiotic characteristics of the ecosystems, including the chemical and physical characteristics of the soil ([34]). Within only a few years after its initial invasion and in the absence of appropriate control, this fast-growing tree makes the recovery and restoration of native vegetation extremely difficult. For these characteristics, A. altissima is considered a fast-growing and contamination- resistant species ([12]).

The dramatic ecological consequences of A. altissima invasion have long been recognized especially in the Mediterranean basin, where it has been recently included in the EPIDEMIE project (“Exotic Plant Invasions: Deleterious Effects on Mediterranean Island Ecosystems”), supported by the European Commission ([31], [35], [1]). One of the main reason why A. altissima is a successful invader in Mediterranean-type regions is its ability to cope with summer drought. In particular, this species is able to reduce water loss through a marked reduction in stomatal conductance and in root hydraulic conductivity ([19], [32]).

Like other Mediterranean island ecosystems, Sicily has experienced substantial A. altissima invasion, and despite this its establishment and spread process have seldom been documented or studied. In particular this tree is able to invade a wide range of pedo-climatic environments and is increasingly common throughout the island in natural and semi-natural habitats, from sea level up to 1300 m a.s.l. ([4]).

Because of the serious damage caused by A. altissima worldwide, many control trials have been carried out in many different regions, including Mediterranean Europe, the United States, and Canada ([10], [27], [16], [15], [7]). These field experiences have clearly demonstrated that the eradication of A. altissima is extremely difficult and expensive, as mechanical treatment alone (cutting the trunk at the root collar or girdling the tree) does not kill the tree, but rather stimulates vigorous resprouting from the remaining trunk and roots ([10], [5], [24]). Even a “double-cut stump treatment” performed each year for 5 consecutive years failed to significantly reduce the ability to resprout ([15]). Only the application of herbicides combined with physical treatment has effectively controlled regrowth and significantly reduced the presence of A. altissima in invaded areas ([10], [27], [16], [15], [7]). For example, the combined application of heavy mechanical treatments and herbicides suppressed A. altissima in Carrascal de la Font Roja Natural Park (Alicante Province, SE Spain), where the first signs of recovery by the natural vegetation were observed 5 years after the treatments were applied ([15]).

One of the herbicides that has given satisfactory control of A. altissima is glyphosate, so that it is considered “a mainstay of woody plant vegetation management” ([6]). In our study we used this chemical, and we attempted to identify a practical way to control A. altissima in woodlands and other remote areas where main electricity is not easily accessed.

  Materials and methods 

The study sites

The trials were carried out at two sites near Palermo (NW Sicily) at altitudes between 50 and 100 m a.s.l. and over surfaces of about 500 m². The first site was an abandoned orchard in a suburban environment (38° 06’ 28.91” N, 13° 19’ 24.91” E), and the second site was a semi-natural community afforested with Pinus halepensis Mill (38° 05’ 40.82” N, 13° 15’ 41.32” E). Both sites have been heavily invaded by A. altissima and share a typical Mediterranean climate, with about 600 mm of rainfall per year, concentrated during the coldest months. The average annual temperature is 19 °C ([18]). Despite similar climatic conditions, the two sites are ecologically different because of differences in management. At the first site, the abandonment of cultivation has enabled the rapid invasion by A. altissima, and although some orchard trees are still evident, A. altissima is the dominant species. At the second site, the presence of a simplified forest ecosystem, due to frequent human disturbance, along with the absence of control interventions, have favored A. altissima spread. This non-native tree is able to establish quickly in forest clearings and along forest borders, from which it gradually spreads into adjacent forested areas.

Treatments

Experiments were carried out in early August, late in the vegetative growth of A. altissima, when an injected herbicide is readily translocated to the roots ([21]). In each site a total of 40 trees, randomly chosen, were treated with undiluted glyphosate (Myrtos^® 360 SL, NUFARM Italia srl, Milan - 360 g lt^-1) that was injected into holes drilled into the stems. One ml of the herbicide was used when DBH (diameter at breast height) of trees was lower than 8 cm (20 trees per site), while 2 mL were used when DBH was larger than 8 cm (20 trees per site). The average DBH of the stand was 14 cm. Trunk holes were located 1 m above the soil surface, about 5 cm deep into the stem or trunk. One hole was drilled into each tree. We also compared drill bits (Krino^®, Italy) with a diameter ranging from 8 to 10 mm. Drilling a few holes, we preli-minarily verified that 10-mm diameter drill bits are best suited to put the fixed rate of glyphosate.

The herbicide was injected with a plastic syringe. The injected holes were pointed downward (at 45°), and after injection each hole was sealed with a wound-sealing compound in order to reduce the chance of herbicide spreading from the application site to the surrounding vegetation. The canopy of the treated trees was visually assessed about 1 month (September 2011, short-term evaluation) and 8 months (April 2012, long-term evaluation) after treatments.

The control techniques applied in the field

The availability of main electricity in remote areas is a recurring problem, and we therefore compared a drill with rechargeable battery (cordless drill, Black & Decker^®), and a car battery (12 V, 70 Ah, Unienergy^®, Italy) plus one car battery charger (Life^®, Italy), and one inverter (GBC, Italy) to transform a 12 V CC to 220V AC, as to connect the car battery to the drill - see Fig. 1). In the first site 30 stems were drilled with power system one and 10 with the power system two, while in the second site we applied only the power system two. We took into account not only the working speed, but also technical and operational feasibility of these two possible sources of power.

Fig. 1 - Recommended equipment for drilling for stem-injection technique. Before the battery is taken to the field, it is charged with the blue inverter, which is connected to the electricity grid via a plug (A) and connected to the battery via the positive and negative poles (B and C, which are not connected to the battery in the figure). In the field, the charged battery feeds a second inverter via cables (F and G - via the positive and negative poles D and E), and the inverter supplies power via another cable to the drill.

  Enlarge/Shrink   Download   Full Width  Open in Viewer

  Results 

In both studied areas more than 90% of the glyphosate-treated A. altissima trees appeared to be dead one month after treatments and these results were confirmed 8 months later (Tab. 1). Non-treated Ailanthus trees and non-target woody species within the sampling areas remained healthy.

Comparing the two drilling systems adopted in the present study, we found that cordless drill usually works for no more than one hour, allowing the treatment of about 30 stems. By contrast, the second method, consisting of a fully charged car battery connected to an inverter and to the drill, allowed to drill the remaining 50 trees (10 in the first site and 40 in the second site) before running out of power. As we verified that a fully charged car battery lasts about 6 hours, up to 180 stems could be treated with a single battery recharge (Tab. 2).

  Discussion 

Effective methods for controlling A. altissima are urgently needed all around the world and should be immediately implemented, especially into woodlands and other forested sites heavily invaded by this fast-growing tree. Previous studies concerning the control of A. altissima by stem-injection of herbicides ([10], [27], [16], [15], [7]) do not provide detailed information on which application methods are practical and actually feasible in the field. In this work, we applied an effective and simple way to clear Ailanthus-infested areas also in faraway, hardly-accessible, rugged locations, as in many forest sites in southern Italy.

We have chosen to use the glyphosate - a non-selective systemic herbicide - because of its effectiveness in other control trials concerning A. altissima (e.g., [15]), and we chose stem-injection technique because it can be applied to many trees in fairly short time, minimizing the quantity of required herbicide ([16], [27]). In other experiments, glyphosate has also been used to fill the cut in trunks caused by “machete” ([9]), while in Palermo city (Sicily) local gardeners are used to control Ailanthus-infested sites by injecting glyphosate in holes created by drill.

Herbicide injection has to be carried out very carefully, as glyphosate may cause skin or eyes irritation after direct exposure in humans, and also some animal species, in particular amphibians and fishes, are susceptible to poisoning ([8], [22], [3], [28]). This is particularly true for some glyphosate-based herbicide, which are specifically toxic in aquatic environments. However, in terrestrial ecosystems glyphosate has low mobility, as it is rapidly adsorbed by soil colloids and eventually inactivated by microbial degradation ([30], [11]). For this reason, glyphosate is one of the most commonly used herbicides in natural areas ([33] - see also references therein). To prevent any possible environmental risk, stem injection joined with the subsequent cover of the drilled holes is the safest technique to be used in natural habitats or in areas to be preserved. Our experiments confirmed that glyphosate injection could be a safe method of control even in presence of native significant species.

According to DiTomaso & Kyser ([16]), we found that the application rate of 1 mL of glyphosate per tree is effective when DBH is < 8 cm, but 2 mL is needed when DBH ≥ 8 cm. Our results are consistent with other researches which provided a high mortality of plants treated by stem-injection or with the EZ-Ject® application system ([27], [16], [7]), also in the case where sprouts following a cut-stump treatment ([20], [27], [16], [15]).

Although previous studies demonstrated that herbicide is best applied via holes drilled into the stem (e.g., [16]), the electrical source to power the drill was not previously taken into consideration. Use of the electrical drill is only feasible when the electrical supply is close to the interventions areas, though the presence of long electrical cables in the field is unsafe, mainly when the terrain is rough or steep. Generators are quite efficient but are usually heavy to be carried and thus are suitable only for treating small and easily-accessible areas. A cordless drill would be the best solution, but the batteries usually cannot supply adequate power; its use is therefore discouraged unless the unit can be readily recharged or treatment has to be applied to few trees. Based on our field trials, we recommend the use as power supply of a fully charged car battery (usually delivering a peak current of 450 A). Car batteries are relatively light, can therefore be readily moved to different locations in the field, and can be used with short cables. Although they are heavier than a cordless drill, car batteries are able to supply power for a much longer time (see Tab. 2) and so they should be suitable for Ailanthus control in natural settings, over wide areas and on hillsides. As for the drill regards bits, 10-cm-long and 10 mm-diameter helical bits enable to drill holes quickly and with little effort.

  Conclusion 

In conclusion, invasion by A. altissima is a serious problem all around the world, and calls for suitable solutions to hinder its unobstructed spread. Although injection of glyphosate (and of other herbicides) into holes drilled into stems kills the trees, a power supply is needed for the drill, and obtaining electrical power in remote areas can be difficult. We think that our technique could be readily applied in the field, not only in suburban areas but also in natural areas such as forests and other woodlands.

  Acknowledgments  

The research was performed under the Agreement with the Department of Agriculture and Forestry of the Sicilian Region: “Analisi dei sistemi seminaturali e degli agro ecosistemi nei sistemi insulari mediterranei: Isola di Lampedusa, pantani di Vendicari e sistemi agricoli regionali ad Alto Valore Naturale (AVN)”. We thank Giuseppe La Mantia, John J. Borg, Salvatore Pasta and Davide Molone for their assistance.

  References