TY - JOUR
T1 - Detection of vegetation trends in highly variable environments after grazing exclusion in Mongolia
AU - Kakinuma, Kaoru
AU - Terui, Akira
AU - Sasaki, Takehiro
AU - Koyama, Asuka
AU - Jamsran, Undarmaa
AU - Okuro, Toshiya
AU - Takeuchi, Kazuhiko
N1 - Publisher Copyright:
© 2017 International Association for Vegetation Science
Copyright:
Copyright 2017 Elsevier B.V., All rights reserved.
PY - 2017/9
Y1 - 2017/9
N2 - Aims: Environmental variability (e.g. in precipitation) has a large effect on vegetation dynamics, and this often makes it difficult to assess the recovery of vegetation after a disturbance. In this study, we assessed vegetation recovery trends in response to grazing exclusion while considering for the annual environmental variability. Location: Two regions with highly variable precipitation: a steppe near Mandalgobi, in Mongolia's Central Gobi province, and a desert steppe near Bulgan, in Mongolia's South Gobi province. Methods: Changes in vegetation were observed along grazing gradients at the above two sites, and vegetation thresholds were identified. We established reference plots in pre-threshold areas along the grazing gradients. We removed the impact of livestock grazing at various locations along the grazing gradients by establishing exclosures, and investigated vegetation from 2005 to 2013. We developed a smoothed hierarchical model within a Bayesian framework, and examined the effect of grazing exclusion on vegetation, focusing especially on the extent of grass cover recovery. In addition, we compared soil nutrient conditions in the reference plots and inside and outside each exclosure along the grazing gradients. Results: Temporal trends in the cover of perennial grass in each plot inside and outside of the exclosures largely coincided, irrespective of grazing intensity, and exclosure had no effect (Bulgan) or a negative effect (Mandalgobi) on vegetation recovery. Soil nutrient content was not significantly affected by exclosure at a given distance from the grazing source, but decreased significantly with decreasing grazing intensity. Thus, recovery of the land from a post-threshold state may not be apparent even after 9 yr of grazing exclusion in environments with highly variable precipitation. Conclusion: The effect of exclosure duration on perennial grass cover was limited, even after controlling for environmental variability. Once a vegetation threshold has been crossed, merely removing livestock from the landscape may not be sufficient for that area to recover.
AB - Aims: Environmental variability (e.g. in precipitation) has a large effect on vegetation dynamics, and this often makes it difficult to assess the recovery of vegetation after a disturbance. In this study, we assessed vegetation recovery trends in response to grazing exclusion while considering for the annual environmental variability. Location: Two regions with highly variable precipitation: a steppe near Mandalgobi, in Mongolia's Central Gobi province, and a desert steppe near Bulgan, in Mongolia's South Gobi province. Methods: Changes in vegetation were observed along grazing gradients at the above two sites, and vegetation thresholds were identified. We established reference plots in pre-threshold areas along the grazing gradients. We removed the impact of livestock grazing at various locations along the grazing gradients by establishing exclosures, and investigated vegetation from 2005 to 2013. We developed a smoothed hierarchical model within a Bayesian framework, and examined the effect of grazing exclusion on vegetation, focusing especially on the extent of grass cover recovery. In addition, we compared soil nutrient conditions in the reference plots and inside and outside each exclosure along the grazing gradients. Results: Temporal trends in the cover of perennial grass in each plot inside and outside of the exclosures largely coincided, irrespective of grazing intensity, and exclosure had no effect (Bulgan) or a negative effect (Mandalgobi) on vegetation recovery. Soil nutrient content was not significantly affected by exclosure at a given distance from the grazing source, but decreased significantly with decreasing grazing intensity. Thus, recovery of the land from a post-threshold state may not be apparent even after 9 yr of grazing exclusion in environments with highly variable precipitation. Conclusion: The effect of exclosure duration on perennial grass cover was limited, even after controlling for environmental variability. Once a vegetation threshold has been crossed, merely removing livestock from the landscape may not be sufficient for that area to recover.
KW - Bayesian statistics
KW - Drought
KW - Environmental variability
KW - Precipitation
KW - Recovery
UR - http://www.scopus.com/inward/record.url?scp=85022331753&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85022331753&partnerID=8YFLogxK
U2 - 10.1111/jvs.12551
DO - 10.1111/jvs.12551
M3 - Article
AN - SCOPUS:85022331753
SN - 1100-9233
VL - 28
SP - 965
EP - 974
JO - Journal of Vegetation Science
JF - Journal of Vegetation Science
IS - 5
ER -