Evaluating and modeling the effects of thinning on stream thermal regimes in a forested headwater catchment

概要

Stream thermal regimes represent as stream temperature, which is an important indicator for understanding both hydrological and biological processes in stream ecosystems. In forested headwater catchments, timber harvesting by removal of forest stands along riparian areas increased stream temperature due to decreases in shading by vegetation and resultant increases in solar radiation inputs. Strip-thinning is a unique forest practice for effective timber harvesting by removal of trees within linear strips. Strip-thinning reduced canopy cover by creating the coherent patches of shade and sunlight areas, which possibly influenced to stream temperature. Hence, the effects of strip-thinning on stream temperature are not fully investigated by comparing clear-cutting and random thinning. Moreover, stream thermal regimes controlled by atmospheric conditions, channel morphology and hydrological processes such as stream discharge and groundwater inflow. Among these factors, the influence of hydrological processes to the rate of increases in stream temperature can vary within a catchment depending on locations of monitoring. Thus, stream discharge and groundwater inflows need to be incorporated into for evaluating the effects of strip-thinning on stream thermal regimes. Furthermore, recovery of riparian vegetation related to the growth of remained trees and understory vegetation possibly alters thermal regimes in a stream channel.

This study was conducted in two headwater catchments, located in FM Karasawayama. 17-ha catchment was subjected to thinning treatment, while 8.9-ha catchment was maintained as control catchment. Paired catchment analysis involving both control and treated catchment was applied to the responses of stream temperature to 50% of strip-thinning. The continuous discharge and stream temperature data were obtained with ± 0.3°C resolution at 5 minutes interval. In the first year after thinning, the treatment effects during April to September elevated by an average of 1.3 °C (range 0.6 to 3.9 °C). The magnitude of stream temperature responses after strip-thinning was similar to the previous studies with random thinning. Daily variability of treatment effects was positively related to solar radiation and negatively associated with stream discharge. In particular, the rate of increases in stream temperature was up to 3.9 °C under streamflow decreasing due to a reduced thermal capacity. In contrast, the moderate discharge (1 to 5 mm day-1) had higher streamflow volume and higher velocity, which mitigate the increases in stream temperature associated with heating from solar radiation.

Differences of stream temperatures and flow volumes within a catchment were examined by monitoring of nested gauging stations. Nested observations were conducted in 10.2-ha (middle catchment), 3.7-ha (middle catchment), and 5.1-ha (upper catchment) sub-catchment within 17-ha catchment affected by thinning. Monitoring nested gauging stations and paired-catchment analysis revealed the rate of increases on stream temperature differed depending on the locations within a catchment. Increases in stream temperature after thinning was the greatest in the catchment outlet (ranges from 0.6 to 3.9 °C), while the treatment effects of middle and upper catchments tended to be low ranging from 0.5 to 3.0 °C. Flow volume during 5 days without rainfall of middle catchment was 2 to 6 times higher than that of upper catchment possibly associated contribution of flow by internal hydrological processes. Summer thermal sensitivity, which is an indicator for groundwater inflows, based on a linear regression coefficient between daily maximum stream temperature against daily maximum air temperature showed that catchment outlets (0.5 °C/°C) was much thermal sensitive than upper and middle catchments (0.3 °C/°C). This results suggested that groundwater inflows occurred within channel reach from upper to middle gauging stations likely mitigated increases in maximum stream temperature during summer.

For evaluating thermal regimes in the headwater channel, heat budget model was applied based on monitoring of net radiation, latent heat, sensible heat, bed heat conduction, stream discharge and, channel dimensions as parameters of model. Microclimate station including net radiation, air temperature, relative humidity, wind speed and streambed temperatures of channel reach was measured every 5 minutes interval. Detail channel morphology and wetted width were measured using structure from motion (SfM) by a camera with zig-zag cable system located 2-3 m above a 40m of headwater channel. The estimated energy fluxes suggested that net radiation was the major heat energy during daytime, while latent and sensible heat fluxes and bed heat conduction were the secondary terms. Estimating of channel wetted varied by the complexities of channel morphology (e.g., presence of boulders). The accurate measurement of wetted width by SfM improved performance of heat budget model by 18% because this channel reach had small depth compared to the particle’s sizes.

The changes in riparian vegetation cover before and after strip-thinning affecting to stream thermal regimes were examined with three scenarios: (1) pre-thinning, (2) 1st year of post-thinning, and (3) 8th year of post-thinning by using a heat budget model. The net radiation in the first year of post-thinning was 1.6 times higher than that of pre-thinning canopy cover. The mean of riparian canopy openness in 8th year of post-thinning was 11.4±0.9% and similar to pre-thinning period (10.9±1.3%). In the 8th year of post-thinning, the leaf development of remaining overstory and the growth of understory vegetation, which could increase canopy closure and resulted in decreases daytime net radiation. By comparing the simulated stream temperature between 1st year of post-thinning and 8th year of post-thinning, the growth of overstory from remained riparian trees and understory vegetation could reduce stream temperatures ranged from 0.4 to 1.8°C. These results suggested that the recovery rate of stream temperature in this study was faster than previous studies, which showed that the recovery of stream temperature occurred after 10 to 16 years of forest harvesting.

The findings of this study showed that hydrological processes such as groundwater inflows and understory vegetation were important to evaluate and mitigate the impacts of forest harvesting on stream temperature. The application of these findings can be used in a guideline of forest management practices to support decision making of forest managers for protecting the stream ecosystems in forested areas. The countries, where had high frequency of forest harvesting such as Viet Nam and Japan, can consider assessing hydrological processes of watershed and remaining understory vegetation along streams to minimize the impacts of forest harvesting on stream temperature.

この論文で使われている画像