The present study was conducted in the suburban forest of Thessaloniki city (Seich Sou), which constitutes one of the most significant suburban forests in Greece. In 1997, more than the half of the forest area was destroyed by a wildfire, after which soil erosion and flood control works (check-dams) were constructed in the burned areas. The aim of the study is to estimate the annual soil erosion rate for the last 30 years (pre- and post-fire periods) applying the WaTEM/SEDEM model, in order to investigate the impact of this wildfire on soil erosion, the effectiveness of the flood- and erosion-control works and the level of forest regeneration. It is the first time that WaTEM/SEDEM was calibrated and validated in Greece, taking into account soil erosion records from the 18 check-dams that were constructed in the study area in 2001. The mean annual erosion rate was 0.0419 t/ha/year, 0.998 t/ha/year and 0.08 t/ha/year for the pre-fire period, the first 3 years and 20 years after the fire, respectively. The results showed a very low erosion rate for the pre-fire period, an expected significant increase 3 years after the wildfire and a gradual decrease in the subsequent years until 2021. However, it seems that the post-fire regeneration of the forest has not been fully achieved, since the annual soil erosion rate at the long-term post-fire period is double compared with the pre-fire period. Concerning the check-dams’ effectiveness, it was observed that after 20 years of operation, they were non-silted, and most of them retained a small amount of sediments. This fact could be attributed to multiple factors such as the very thin soil depth, fire severity and catchment geomorphology, though the main reason seems to be the time elapsed between fire occurrence and the check-dams’ construction. The results of this study advance/strengthen the knowledge concerning the pre/post-fire soil erosion processes in already degraded ecosystems, while the calibrated model could serve as a useful tool able to be applied in other Mediterranean catchments of similar characteristics.

The present study was conducted in the suburban forest “Seich Sou”, which is located at the north-east of Thessaloniki city (north Greece). “Seich Sou” is one of the most significant suburban forests of North Greece. However, many disastrous events have taken place in the suburban forest during the last 25 years, caused by either human interference or other biotic/abiotic factors, such as insect outbreaks. In the present study, erosion measurements and field data were collected using a combination of silt fences (USLE plots) and depositions retained behind the constructed check-dams, aiming to monitor the impact of the significant biotic and abiotic disturbances (forest fires, insect outbreaks, logging) in Seich Sou forest. Specifically, the aim of this study is to present a cost-effective and time saving methodology towards the achievement of accurate and reliable soil erosion measurements and field data acquisition. Additionally, the installation details, data recording and collection, field work, supplementary materials and the advantages of silt fences and check-dams are presented in detail, as well as the limitations of the methods and the difficulties during the installation, maintenance and data collection period. The proposed methodology could be effectively applied in many environments and in the context of varied purposes, to quantify the erosion and runoff processes with high accuracy, as well as to increase the accuracy of soil erosion modeling performance, through implementation of calibration and/or validation processes, which is a major issue for the scientific community.

<p>Natural-related disasters like insect outbreaks might alter how the surface runoff is intercepted in affected forests, leading to more intense floods and soil erosion. The quantification of soil erosion and its relationship to the phenomenon of insect outbreaks are currently not well-explained or insightfully understood in the literature.</p> <p>In addition, wildfires are the most destructive phenomena that significantly affect the soil erosion rate. Most studies primarily focus on the "disturbance window" (ranging usually 3 to 10 years). There are few studies investigating the long-term post-fire consequences on soil erosion rates, particularly those examining the time period beyond 20–30 years after the wildfire.</p> <p>The present research is conducted in Thessaloniki's suburban forest (Seich Sou), North Greece. Over 50% of the woodland area was burned by a wildfire in 1997. The bark beetle <em>Tomicus piniperda</em> caused a severe insect infestation (in May 2019) . More than 300 ha of <em>Pinus brutia</em> forest were lost as a result of the infestation. To minimize and prevent the spread of the infestation, diseased trees have been cut down and removed from the forest.</p> <p>The aim of the research is to investigate the long-term effects of wildfire (1997) on erosion rate and, for the first time in the literature, to estimate the effect of insect outbreak on soil erosion.</p> <p>Soil erosion was monitored by installing silt fences on 25 field plots (each plot area: 5 x 22 m) in five representative forest areas: (1) areas naturally reforested after the wildfire of 1997, (2) areas characterized by reforestation failure, (3) areas infected by insects, (4) areas with insect outbreak and logging of dead trees, and (5) undisturbed areas as a control.</p> <p>Soil erosion was lowest in (5) undisturbed plots (control) (0.023 t/ha/year), whereas it was highest in (2) areas of reforestation failure (0.51 t/ha/year). When the four affected sites are compared, the plots with (3) "insect outbreak (no logging)" present the lowest soil erosion rate (0.027 t/ha/year). Furthermore, the difference between (3) "insect outbreak (no logging)" and (5) "control" plots was negligible, revealing that the influence of entomological infestation on soil erosion was observable and measurable, although quite low. The plots with (4) "insect outbreak and logging of dead trees" showed double amount of erosion compared with the plots of (3) "insect outbreak (no logging)" and (5) "control" plots. This fact revealed that logging, as a method to prevent infestation spread, might have positive results, but it could also have negative impacts on soil erosion rates.</p> <p>Concerning the long-term consequences of the 1997 wildfire on soil erosion, the results from plots (1) with "natural reforestation" indicated that erosion rates are triple (0.062 t/ha/year) 25 years after the wildfire compared to the (5) "control" plots (0.023 t/ha/year). It appears that the "window of disturbance" in the reforested region has not been closed, and the forest environment has not been substantially recovered. Restoration of soil erosion rates to normal levels under Mediterranean conditions, where soils are generally thin and stony, can take more than two decades, depending on site quality, geomorphology and meteorological conditions.</p>