29

Global Environmental Research ©2010 AIRIES 14/2010: 29-36 printed in Japan

Simultaneous Monitoring of Livestock Distribution and

Desertification

Kensuke KAWAMURA1* & Tsuyoshi AKIYAMA2

1Graduate School for International Development and Cooperation, Hiroshima University, Japan 1-5-1 Kagamiyama, Higashi-Hiroshima, Hiroshima 739-8529, Japan

2River Basin Research Center, Gifu University, Japan *e-mail: kamuken@hiroshima-u.ac.jp

Abstract In the arid and semi-arid regions of Northeast Asia, grassland degradation has become a major environ-

mental and economic problem, so sustainable utilization of resources is crucial in terms of not only supporting the local and animal production but also thinking about the global environment. Especially in Mongolia and North China, grassland degradation has mainly been induced by artificial factors such as overgrazing by livestock, inadequate management of arable land and uncontrolled cutting of trees for fuel. In particular, overgrazing is one of the main causes of grassland degradation, and conservation and sustainable use of the grasslands can be achieved by proper grazing management to match carrying capacity estimates. How, though, can we achieve the ability to maintain valuable ecosystems and sustainable livestock farming at the same time? Satellite remote sensing and the Geographic Information System (GIS) are two of the most powerful tools to assist in grazing management. There is an increasing number of development analysis tools that can aid in collecting data on seasonal changes in grazing behavior over long-term periods. Quantitative assessment methods for monitoring spatial and temporal distributions of grazing animals at a regional scale are required in studies designed to determine the driving causes of desertification. This paper first reviews previous studies monitoring desertification through the use of remote sensing and tools for assessing livestock distribution, and then discusses current developments and future perspectives of remote sensing of grasslands in Northeast Asia.

Key words: grazing pressure, phenology, qualitative and quantitative assessment, remote sensing

1. Introduction

It is currently widely accepted that desertification is

one of the most serious environmental issues in semi-arid and arid areas. According to Lepers et al. (2005), Asia currently has the greatest concentration of areas showing rapid land cover change, particularly dryland degradation. In Northeast Asia, desertification is extremely serious due to poor climate conditions, such as drought, severe wind erosion and long-term unsustainable human activities, including overgrazing, extensive cutting and over-reclamation (Wang et al., 2006). Especially in Mongolia and North China, grazing pressure may play a critical role in maintaining ecosystem productivity through concentration of limited resources (Wang & Wang, 1999) as well as the meteorological effect (Li et al., 2000).

The environmental impact from overgrazing in Mongolia and North China can be reduced by more effi-cient forage utilization, and grassland diversity enhanced through improved understanding of grazing selection. In

addition, the physical and economic productivity of grazing systems is intrinsically linked to the efficiency of single grazing units. To date, numerous grazing trials have been done by scientists in Mongolia and China in order to determine the effects of grazing on plant production (Li et al., 1999), plant species composition (Han et al., 1999, Tsutsumi et al., 2003), soil organic matter (He et al., 2008) and soil structures (Jia et al., 1999). However, when applying these results in free ranging areas, some difficulties arise when relying on direct observation alone because some ground-based methods are not practical for assessing both plant bio-mass and grazing distribution over extensive geographic areas. Satellite remote sensing and the Geographic Information System (GIS) are two of the most powerful tools for assisting in grassland resource inventories and integration of data and as mechanisms for analysis, modeling and forecasting to support decision-making (Tueller, 1989). However, it is difficult to evaluate live-stock intake from satellite data alone because remotely sensed signatures only can provide information on exist-

30 K. KAWAMURA & T. AKIYAMA

ing surface properties, such as soil, vegetation and water bodies. Quantitative assessment methods for monitoring spatial and temporal distributions of grazing animals at a regional scale are required in studies designed to deter-mine the driving causes of desertification. The Global Positioning System (GPS), as one potential tool, is now widely applied in monitoring the spatial distribution of grazing animals.

In this review paper, we introduce ways to examine grassland desertification using remote sensing technol-ogy with a focus on Mongolia and North China. We in-clude recent developments in sensor technology for remote sensing and tools for monitoring livestock distribution, with future perspectives for integrated use of these technologies to combat desertification issues. The sections include an overview, remote sensing for monitoring grassland degradation, tools for monitoring livestock grazing behavior and their spatial distribution, and future perspectives on efficient grassland manage-ment.

2. Overview

China is one of the most severely desertified countries

in the world. Almost 60% of its population is living in areas at risk, covering approximately 3.3 million km2 in thirteen provinces and autonomous regions, mainly in North China (CNC-UNCCD, 1996). There are eight large deserts and four large sandy lands (Zha & Gao, 1997) (Fig. 1). Most lie in regions above 35ºN and have an annual rainfall of <450 mm (Wang et al., 2008), or <400 mm according to Zha and Gao (1997). Historically, in China, ground surveys have been conducted to collect

information on desertification, and this method has been gradually replaced by aerial photographic interpretation at small scales. Since the emergence of space-borne remote sensing, satellite images have been utilized to delineate the extent of desertified areas (Zha & Gao, 1997).

Mongolia is a landlocked country with a land area of over 150 million ha, of which over 83% is rangelands (Angerer et al., 2008). Its climate is continental with extremely cold, dry winters and warm summers. Live-stock production is its major agricultural industry, accounting for approximately 20% of the gross domestic product. Mongolia’s livestock production depends much on the productivity of natural grasslands. Wuyunna and Okamoto (2006) reported on changes in Mongolia’s agriculture, stock raising, ecosystems and ecological dynamics.

How then can we get an overview of the present and future states of grasslands ? Grassland conditions will be decided by the balance of grass production (GP) and herbage intake by grazing animals (HI) (Fig. 2). When HI is greater than GP, grasslands face degradation. On the other hand, if GP is greater than HI, grasslands will be conserved and the land will recover. Akiyama and Kawamura (2007) proposed a system for measuring GP and HI to predict grassland conditions. GP is determined by climate, soil, plant species and so on, whereas HI depends on grazing intensity, including the stocking rate, animal species, palatability and so on. This system is intended for real-time monitoring of GP and HI using satellite remote sensing data, the GIS, GPS and mathe-matical models.

Fig. 1 Landsat mosaic image for China and Mongolia with locations of desert and sandy lands in China (Source: modified from

Zha & Gao, 1997), and boundary line (yellow line) at 400 mm annual precipitation.

Simultaneous Monitoring of Livestock Distribution and Desertification 31

3. Remote Sensing for Monitoring Grassland Degradation

3.1 Changes in land cover type Desertification dynamics can be accurately monitored

and assessed via a combination of satellite observations and in situ information (Collado et al., 2002). Satellite remote sensing data can provide alternative data sources for large areas, allowing classification of the land cover types and estimations of above-ground biomass and primary production. In general, changes in vegetation are the most direct indicator of land degradation (Yang et al., 2005). Satellite-based remote sensing data have been applied to the monitoring of desertification in Northeast Asia, the mapping of land use and land cover (LULC) and the ascertainment of desertification indicators.

Numerous studies of large-scale land cover mapping have used data from the National Oceanic and Atmos-pheric Administration’s (NOAA) Advanced Very High Resolution Radiometer (AVHRR). AVHRR-derived nor-malized-difference vegetation index (NDVI) data may be the most widely used data source and has providing some success at assessing the condition of vegetation in arid and semi-arid regions of North China (Akiyama & Kawamura, 2003, Kawamura et al., 2004, Lee et al., 2002) and Mongolia (Kogan et al., 2004). Vegetation indices (VIs) are spectral transformations of two or more bands designed to enhance the contribution of vegetation properties. Perhaps, NDVI is the most widely used VIs because of their simplicity, which can be calculated by red and near-infrared bands. The AVHRR sensors, however, were originally designed for meteorological applications and only cover two spectral bands (red and near-infrared) that can be used to generate spectral indices of vegetation greenness. A new generation of optical sensors designed for terrestrial applications was launched in the 1990s. These include the VEGETATION (VGT) sensor onboard the SPOT-4 satellite and the Moderate Resolution Imaging Spectroradiometer (MODIS) onboard the Terra and Aqua satellites.

In the Temperate East Asia (TEA) region, Boles et al. (2004) applied the multi-temporal Land Surface Water Index (LSWI) and Enhanced Vegetation Index (EVI) to

map land cover types and investigate the seasonal dynamics of various land cover types, distinguishing five classes of barren or sparsely vegetated land. Later, in TEA region, Huang and Siegert (2006) investigated the characteristics and usefulness of low resolution SPOT VGT satellite imagery for assessment and monitoring of desertification in North China. They found that more sparsely vegetated land can be detected better by optimized VGT-NDVI data than by Global Land Cover 2000 or MODIS Vegetation Continuous Field data. The areas at risk of desertification were modeled, and the results suggested 1.60 million km2 were at risk of deser-tification.

Various high-resolution satellite data such as Landsat TM and Terra ASTER, were also used to investigate desertification processes (Collado et al., 2002). Imagawa et al. (1997) attempted to detect desertified areas using multi-temporal images, and developed a monitoring method in the Horqin desert area. Using three indices obtained using a Landsat TM-band combination, they determined the yearly changes in desertified areas. Based on a GIS approach, Zhang et al. (2003) investigated desertification of transitional zones of the Desert Loess in the Yulin region using two Landsat TM images between 1987 and 1999. Their results showed that desertification is still severe and that the desertified land accounts for 67.7% of the total land area.

3.2 Yearly changes in net primary production

At the regional to global scale, there have been numerous attempts to predict terrestrial net primary pro-duction (NPP) or above-ground NPP (ANPP) from space using NOAA AVHRR (Paruelo et al., 1997; Wylie et al., 1991; Tucker & Sellers, 1986), Spot VGT (Sabbe & Veroustraete, 2000), and Terra MODIS (Running et al., 2004). For instance, Fig. 3 shows the improved MODIS NPP product (MOD17A3) data of 2000 and 2006 for Mongolia and China (Zhao et al., 2005). The original MODIS NPP/GPP algorithm is described in Running et al. (2004) and Heinsch et al. (2003). The difference in images in Fig. 3 shows that NPP had decreased in Mongolia and North China between 2000 and 2006. According to Angerer et al. (2008), recent studies in Mongolia have indicated that NPP will be reduced by 5%~30% in the forest steppe and steppe zones by 2080. High mountain areas and desert steppe zones, however, are expected to have higher NPP (25%~75%) due to ris-ing temperatures in the high mountain zone and increas-ing rainfall in the desert steppe zones.

3.3 Monitoring aboveground biomass and forage

quality In Inner Mongolia, Lei and Yokoyama (1999) esti-

mated mean above-ground biomass using AVHRR- NDVI on a regional scale in 1996. A similar study in Kawamura et al. (2003) using AVHRR-NDVI found that the mean aboveground biomass of the entire Xilingol steppe in the summer of 2001 was 1,189 kg/ha, which was 40% lower than the amount that Xiao et al. (1997)

Fig. 2 Concept of grassland condition balance between

grass production and herbage intake (Akiyama & Kawamura, 2007).

32 K. KAWAMURA & T. AKIYAMA

obtained by analyzing Landsat TM data in 1987. It demonstrated the decrease in steppe biomass over those fourteen years. In the same region, Kawamura et al. (2005b) compared the capability of spectral VIs derived from AVHRR and MODIS in estimating biomass,

standing crude protein (CP) and CP concentrations in leaves. A regression analysis revealed that MODIS-VI achieved a good coefficient of determination (R2 = 0.77 – 0.83) with regard to estimations of total and live biomass. Furthermore, MODIS-EVI, which uses a blue

Fig. 3 NPP maps of China and Mongolia in 2000 (a), 2006 (b) and the difference between 2000 and 2006 (c),

using the improved MODIS primary production product (MOD17) (Zhao et al., 2005).

Fig. 4 Seasonal changes in the MODIS-EVI estimated live biomass ( ), standing CP (■) and CP

concentration (○) in (a) NG: non-grazed grassland, (b) LG: lightly grazed grassland, (c) MG: intermediately grazed grassland, and (d) HG: heavily grazed grassland in the Xilingol steppe (Kawamura et al., 2005a) with pictures showing respective NG, LG, MG and HG sites.

Simultaneous Monitoring of Livestock Distribution and Desertification 33

wavelength, was a good predictor of standing CP (R2 = 0.74) compared to AVHRR (R2 = 0.53). These results suggest that MODIS-VI can reliably detect the phenology and pasture quantity and quality of steppe grassland area. The NDVI exhibits scaling problems, asymptotic (saturated) signals under high biomass con-ditions and is very sensitive to canopy background variations; NDVI degradation is particularly strong with higher canopy background brightness (Huete, 1988). The EVI on the other hand, was developed to optimize the vegetation signal with improved sensitivity in high biomass regions and improved vegetation monitoring through a decoupling of the canopy background signal and a reduction in atmosphere influences. Kawamura et al. (2005a, 2005b) also conducted a study to determine the potential suitability of Terra MODIS for monitoring short-term phenological changes in pasture conditions in a semiarid region. Applying the regression model, the EVI accounted for 80% of the variation in live biomass, 77% of that in the total biomass, 11% in the CP concen-tration and 74% in the standing CP. Applying these results, seasonal changes in live biomass and the standing CP could be described at the four selected sites with different degrees of grazing intensity (Fig. 4).

4. Remote Sensing in Conjunction with GIS

and GPS

4.1 Tools for monitoring grazing behavior and its spatial distribution

Monitoring changes in spatial distribution and forag-ing behavior of grazing animals provides knowledge that can help farmers understand the relationship between

their internal state (i.e., nutritional requirements, health, etc.) and their environment (i.e., sward state, climate, etc.) (Penning & Rutter 2004). This knowledge would help also land managers provide optimum resource management and forecasting to support decision making.

To date, many investigations in behavior analysis have been conducted at various scales ranging from plant-part to landscape (Bailey et al., 1996; Senft, 1989; Hirata & Ogura, 2003). At a landscape level, GPS tech-nology has been increasingly applied in animal science to monitor spatial distribution and track routes (Barbari et al., 2006; Ganskopp, 2001; Ganskopp et al., 2000), and is often combined with equipment for monitoring animal activity (Champion et al., 1997). For example, Penning (1983) developed a jaw movement recording system for sheep. Combined with a microcomputer the recording system could analyze the movements to deter-mine periods of eating and ruminating and the total numbers of jaw movements, using “Graze” software (Rutter, 2000).

More recently, a simple neckband bite-counter (BC) was developed by Umemura et al. (2009) for cattle, that records jaw movements only when the cattle are foraging. Using BC, Kawamura (2005) monitored the daily pattern of grazing times for sheep in the Xilingol steppe (Fig. 5), where the herd in study grazed freely in the area between sunrise and sunset, and then were kept in fenced enclosures near the farmhouses overnight. The results elucidated the diurnal patterns of grazing activities in Mongolian sheep throughout the grazing season. Higher rates of grazing were observed in the morning and evening, while lower rates tended to occur as tem-peratures increased around 15:00 in the afternoon. The

Fig. 5 Changes in daily patterns of grazing time with air temperature

between June and September in 2004 (Kawamura, 2005).

34 K. KAWAMURA & T. AKIYAMA

results confirmed previous findings from animal behav-ior observations that daily patterns of grazing behavior and total times spent grazing were influenced by hot summer environments (Kurosaki et al., 1982; Matsui, 1996). 4.2 Quantifying grazing intensities using integrated

RS, GIS and GPS As a powerful information technology system which

stores, analyzes and displays both spatial and non-spatial data, the GIS is expanding rapidly with the growth of computer technology. Kawamura et al. (2005c) developed a methodology, using Terra MODIS-derived NDVI data combined with GPS data and GIS technology for quantifying the effects of grazing intensities (GI) on plant productivity in the Xilingol steppe, Inner Mongolia. To quantify the grazing pressure, a GI map of three herds of sheep was created using a grid cell method, capturing the tracking data by GPS. The relationship between GI and estimated plant biomass showed a poor negative correlation. When the plant biomass data were separated into two different vegetation types, marshy meadow and typical steppe, a stronger negative correlation was obtained between GI and biomass (R2 = 0.887) for the typical steppe data. The higher values of GI were showed around the farmer’s house and near the Xilin riverside area. In addition, the grazing area was used to extended area in June and to narrow area in July (Kawamura, 2005). It may be explained because of that; in early spring and autumn season, sheep walk long distance with looking for grass, while in late spring to summer season, there are enough grass to eat.

Using remote sensing integrated with GPS and GIS, the effects of grazing on forage quantity and quality can be determined in a free ranging area. However, consid-ered the other way around, animal performance and their spatial distribution are themselves affected by forage quality and quantity (Ganskopp & Bohnert, 2009). Until now several factors have been identified in observations as having an effect on animal behavior, including air temperature (Kurosaki et al., 1982; Matsui, 1996), degree of slope (Ganskopp & Vavra, 1987). These factors are interrelated and influence both grass production and grazing distribution. In future studies, it will be important to determine the interactions between grass production and animal performance in order to promote sustainable grassland management.

5. Future Perspectives on Efficient Grassland

Management A great deal of management knowledge has been

accumulated on grazing in artificial grasslands. It has become possible to estimate the short- and long-term production of each pasture grass under given climate conditions. At the same time, grazing pressure from ani-mals can be calculated from the number of animals, their grazing period, the pasture size and plant production. However, grassland management is more difficult for

native grasslands under free-ranging conditions where the area of grazing grassland cannot be defined. It may be strongly affected by unrestricted animal behavior, biodiversity of plant species and unstable climate conditions. Therefore, a strong need has arisen for accu-rate monitoring of free ranging grasslands from the viewpoint of both sustainable animal production and conservation of fragile grassland ecosystems.

To meet such demands, we need to be able to deter-mine the distribution of plant biomass, its nutritive value and the proportion of unpalatable grass as real-time grassland information. In addition, information about behavior, diet and herbage intake by animals should be comprehended so as to prevent land degradation due to excessive grazing pressure. Up against such challenges, we have been able to get accurate information through precise ecological experiments and field studies in homogeneous artificial grasslands (we call them “eco-logical process studies”), but this cannot be applied to vast native grasslands under free ranging conditions.

Remote sensing is a technology which has been suit-able for approximate surveys of vast areas extending over several hundred square kilometers, but is not good at evaluating functions or providing accurate analyses. In recent years, though, the domains of remote sensing and process ecology have moved closer and have started partially overlapping each other (Fig. 2). As a result, scientists can investigate the same object by independent methods from different viewpoints, thanks to improve-ments in onboard satellite sensors. Entering the 21st century, the spatial, spectral and temporal resolution of satellite sensors has greatly improved (Akiyama et al., 2004), making functional analysis of grassland eco-systems possible for us. At the same time, ground-based instruments for ecosystem study have also improved. For example, the GPS has become an irreplaceable tool for field surveys. Bite counters (BC) are another effective tool for investigating the intake of grazing animals. Many small but useful devices have been developed to record multiple facets of environmental factors, expanding the dimensions of ecological process studies in grassland science.

In the “precision grazing” proposed by Kawamura et al. (2010), a daily grassland biomass map of 250 m × 250 m resolution may be drawn using Terra MODIS images. The botanical composition and nutrient condi-tion of grasslands can be distinguished using hyperspec-tral images, which will improve information on grassland management. As for grazing animals, GPS technology can record the daily behavior of grazing animals, and BC data can show the times and numbers of bites in the grassland. The GIS is a powerful tool for building grass-land management models. System models for grassland management will improve the short- and long-term production of grasslands and livestock. At the same time, sustainable grassland utilization can be initiated.

Simultaneous Monitoring of Livestock Distribution and Desertification 35

References Akiyama, T. and K. Kawamura (2003) Vegetation changes and

land degradation of steppe in Inner Mongolia. Journal of the Japanese Agricultural Systems Society, 19:12-22.

Akiyama, T. and K. Kawamura (2007) Grassland degradation in China: Methods of monitoring, management and restoration. Grassland Science, 53:1-17.

Akiyama, T., K. Kawamura and S. Jia (2004) Remote sensing ecology. Journal of the Japanese Agricultural Systems Society, 20:153-159.

Angerer, J., G. Han, I. Fujisaki and K. Havstad (2008) Climate change and ecosystems of Asia with emphasis on Inner Mongolia and Mongolia. Rangelands, 30:46-51.

Bailey, W.D., E.J. Gross, A.E. Laca, et al. (1996) Mechanisms that result in large herbivore grazing distribution patterns. Journal of Range Management, 49:386-400.

Barbari, M., L. Conti, B.K. Koostra, G. Masi, F.S. Guerri and S.R. Workman (2006) The use of global positioning and geographical information systems in the management of extensive cattle grazing. Biosystems Engineering, 95:271-280.

Boles, H.S., X. Xiao, J. Liu, et al. (2004) Land cover characterization of Temperate East Asia using multi-temporal VEGETATION sensor data. Remote Sensing of Environment, 90:477-489.

Champion, A.R., M.S. Rutter and D.P. Penning (1997) An automatic system to monitor lying, standing and walking behaviour of grazing animals. Applied Animal Behaviour Science, 54:291-305.

CNC-UNCCD (China National Committee for the Implementation of the UN Convention to Combat Desertification) (1996) China National Action Programmes to Combat Desertification.

http://www.unccd.int/actionprogrammes/asia/national/2000/ china-eng.pdf [cited 23 March 2010].

Collado, A.D., E. Chuvieco and A. Camarasa (2002) Satellite remote sensing analysis to monitor desertification processes in the crop-rangeland boundary of Argentina. Journal of Arid Environments, 52:121-133.

Ganskopp, D. (2001) Manipulating cattle distribution with salt and water in large arid-land pastures: a GPS/GIS assessment. Applied Animal Behaviour Science, 73:251-262.

Ganskopp D, R. Cruz and D.E. Johnson (2000) Least-effort pathways ?: a GIS analysis of livestock trails in rugged terrain. Applied Animal Behaviour Science, 68:179-190.

Ganskopp, D. and M. Vavra (1987) Slope use by cattle, feral horses, deer, and bighorn sheep. Northwest Science, 61:74-81.

Ganskopp, D.C. and D.W. Bohnert (2009) Landscape nutritional patterns and cattle distribution in rangeland pastures. Applied Animal Behaviour Science, 116:110-119.

Han Y., S. Wang and Z. Chen (1999) Responses of the major plant populations of Inner Mongolia typical steppe to grazing rate based on niche considerations. Acta Agrestia Sinica, 7:204-210. (in Chinese with English summary)

He, N.P., Q. Yu, L. Wu, Y.S. Wang and X.G. Han (2008) Carbon and nitrogen store and storage potential as affected by land-use in a Leymus chinensis grassland of northern China. Soil Biology & Biochemistry, 40:2952-2959.

Heinsch, F., M. Reeves, P. Votava, et al. (2003) User’s Guide, GPP and NPP (MOD 17A2/A3) Products, NASA MODIS Land Algorithm. Version, 2:666-684.

Hirata, M. and S. Ogura (2003) Spatiotemporal variation in herbage mass in a grazed pasture: An approach at a feeding station scale. Japanese Journal of Grassland Science, 49:280-285.

Huang, S. and F. Siegert (2006) Land cover classification optimized to detect areas at risk of desertification in North China based on SPOT VEGETATION imagery. Journal of Arid Environments, 67:308-327.

Huete, A.R. (1988) A soil-adjusted vegetation index (SAVI).

Remote Sensing of Environment, 25:295-309. Imagawa, T., M. Fukuhara and T. Watanabe (1997) A monitoring

method of land cover/land use change in Naiman, Inner Mongolia autonomous region, China using Landsat data. JARQ, 31:163-169.

Jia, S., C. Wang, Z. Sun, S. Li, Y. Chen and C. Jin (1999) Study on grassland dark sandy chestnut compaction by grazing intensity and grazing season. Acta Agrestia Sinica, 7:217-222. (in Chinese with English summary)

Kawamura, K. (2005) Application of satellite remote sensing and GIS/GPS for sustainable use of Inner Mongolia grassland. In: The United Graduate School of Agricultural Science. Gifu University, Gifu.

Kawamura, K., T. Akiyama, O. Watanabe, et al. (2003) Estimation of aboveground biomass in Xilingol steppe, Inner Mongolia using NOAA/NDVI. Grassland Science, 49:1-9.

Kawamura, K., T. Akiyama, H.-O.Yokota, M. Tsutsumi, O. Watanabe and S. Wang (2004) Estimation model for NOAA/NDVI changes of meadow steppe in Inner Mongolia using meteorological data. Grassland Science, 49:547-554.

Kawamura, K., T. Akiyama, H.-O. Yokota, et al. (2005a) Monitoring of forage conditions with MODIS imagery in the Xilingol steppe, Inner Mongolia. International Journal of Remote Sensing, 26:1423-1436.

Kawamura, K., T. Akiyama, H.-O.Yokota, et al. (2005b) Comparing MODIS vegetation indices with AVHRR NDVI for monitoring the forage quantity and quality in Inner Mongolia grassland, China. Grassland Science, 51:33-40.

Kawamura, K., T. Akiyama, H. Yokota, et al. (2005c) Quantifying grazing intensities using geographic information systems and satellite remote sensing in the Xilingol steppe region, Inner Mongolia, China. Agriculture Ecosystems & Environment, 107:83-93.

Kawamura, K., N. Watanabe, S. Sakanoue and S. Odagawa (2010) Precision farming system on extensive pasture, 8. Estimating spatial distribution of herbage biomass and concentrations of N and NDF from airborne based CASI hyperspectral imaging. Japanese Journal of Grassland Science, 56 (Ext.):26.

Kogan, F., R. Stark, A. Gitelson, L. Jargalsaikhan, C. Dugrajav and S, Tsooj (2004) Derivation of pasture biomass in Mongolia from AVHRR-based vegetation health indices. International Journal of Remote Sensing, 25:2889-2896.

Kurosaki, Z., T. Sonoda and S. Ono (1982) Effects of high environmental temperature on grazing behaviour and respiratory rate in cattle. Japanese Journal of Grassland Science, 28:324-329.

Lee, R., F. Yu, K. Price, J. Ellis and P. Shi (2002) Evaluating vegetation phenological patterns in Inner Mongolia using NDVI time-series analysis. International Journal of Remote Sensing, 23:2505-2512.

Lei, L. and R. Yokoyama (1999) Estimation of primary production by multi-temporal AVHRR data in the semi-arid grassland of Inner Mongolia. Journal of the Remote Sensing Society of Japan, 19:17-29.

Lepers, E., F.E. Lambin, C.A. Janetos, et al. (2005) A synthesis of information on rapid land-cover change for the period 1981-2000. BioScience, 55:115-124.

Li, Y., Z. Chen, S. Wang and D. Huang (1999) Grazing experiment for sustainable management of grassland ecosystem of Inner Mongolia steppe: experimental design and the effects of stocking rates on grassland production and animal liveweight. Acta Agrestia Sinica, 7:173-182. (in Chinese with English summary)

Li, S.G., Y. Harazono, T. Oikawa, H.L. Zhao, Z.Y. He and X.L. Chang (2000) Grassland desertification by grazing and the resulting micrometeorological changes in Inner Mongolia. Agricultural and Forest Meteorology, 102:125-137.

Matsui, K. (1996) Jaw movements during grazing and rumination of a Holstein cow grazed on pasture in hot summer – a

36 K. KAWAMURA & T. AKIYAMA

comparison between late July and middle September. Grassland Science, 41:280-286.

Paruelo, M.J., E.H. Epstein, K.W. Lauenroth and C.I,. Burke (1997) ANPP estimates from NDVI for the central grassland region of the united states. Ecology, 78:953-958.

Penning, P.D. (1983) A technique to record automatically some aspects of grazing and ruminating behaviour in sheep. Grass and Forage Science, 38:89-96.

Penning, P.D. and S.M. Rutter (2004) Ingestive Behaviour. In : P.D. Penning, ed., Herbage Intake Handbook, The British Grassland Society, Reading, 151-175.

Running, S., R. Nemani, F. Heinsch, M. Zhao, M. Reeves and H. Hashimoto (2004) A continuous satellite-derived measure of global terrestrial primary production. Bioscience, 54:547-560.

Rutter, M.S. (2000) Graze: Aprogram to analyze recordings of the jaw movements of ruminants. Bahavior Research Method, Instruments & Computers, 32:86-92.

Sabbe, H. and F. Veroustraete (2000) Demonstration of a Standard Net Primary Productivity Product for the SPOT 4-VEGETA-TION Instrument. VITO, Belgium.

Senft, L.R. (1989) Hierarchical foraging models: Effects of stocking and landscape composition on simulated resource use by cattle. Ecological Modelling, 46:283-303.

Tsutsumi, M., M. Shiyomi, Y. Wang, M. Tsuiki and X, Yu (2003) Species diversity of grassland vegetation under three different grazing intensities in the Heilongjiang steppe of China. Grassland Science, 48:510-516.

Tucker, C.J. and J.P. Sellers (1986) Satellite remote sensing of primary production. International Journal of Remote Sensing, 7:1395-1416.

Tueller, T.P. (1989) Remote sensign technology for rangeland management applications. Journal of Range Management, 42:442-453.

Umemura, K., T. Wanaka and T. Ueno (2009) Technical note: estimation of feed intake while grazing using a wireless system requiring no halter. Journal of Dairy Science, 92:996-1000.

Wang, X., F. Chen and Z. Dong (2006) The relative role of climatic and human factors in desertification in semiarid China. Global Environmental Change, 16:48-57.

Wang, X., F. Chen, E. Hasi and J. Li (2008) Desertification in China: An assessment. Earth-Science Reviews, 88:188-206.

Wang, Y. and S. Wang (1999) Influence of different stocking rates on aboveground present biomass and hebage quality in Inner Mongolia steppe. Acta Prataculturae, 8:15-20.

Wuyunna and K. Okamoto (2006) State of agriculture and stock raising, ecosystems, and eco-dynamics study in Mongolia. Journal of Japan Agricultural System Society, 22:59-69.

Wylie, KB., A.J.J. Harrington, S.D. Prince and I. Denda (1991) Satellite and ground-based pasture production assessment in Niger: 1986-1988. International Journal of Remote Sensing, 12:1281-1300.

Xiao, X., S.D. Ojima, A.C. Ennis, S.D. Schimel and C. Zuozhong (1997) Estimation of aboveground biomass of the Xilin river basin, Inner mongolia, using Landsat TM imagery. In: Inner Mongolia Grassland Research Station, ed., Research on Grassland Ecosystem, 5, Science Press, Beijing, 130-138.

Yang, X., K. Zhang, B. Jia and L. Ci (2005) Desertification as-sessment in China: An overview. Journal of Arid Environments, 63:517-531.

Zha, Y. and J. Gao (1997) Characteristics of desertification and its rehabilitation in China. Journal of Arid Environments,

37:419-432. Zhang, L., Y. Leping and B. Xia (2003) The study of land

degradation in transitional zones between the MU US desert and the Loess Plateau using RS and GIS – a case study of the Yulin region. Environmental Geology, 44:530-534.

Zhao, M., F.A. Heinsch, R.R. Nemani and S.W. Running (2005) Improvements of the MODIS terrestrial gross and net primary production global data set. Remote Sensing of Environment, 95:164-176.

Kensuke KAWAMURA

Kensuke Kawamura was born in Yamaguchi on 11April 1977, graduated from Gifu University andreceived his Ph.D. in March 2005 under ProfessorTsuyoshi Akiyama’s supervision, majoring inGrassland Ecology and Remote Sensing. Hisdoctoral thesis was entitled “Application of satel-

lite remote sensing and GIS/GPS for sustainable use of Inner Mongolia grassland.” Since then, his major interest has been remote sensing of grassland ecosystems. From April 2004 to March 2006, he was a JSPS post-doctoral research fellow (DC2) at the River Basin Research Center, Gifu University. From April 2006 to June 2008, he received another JSPS research fellowship (PD) at the National Institute for Agro- Environmental Sciences, Japan. As a JSPS research fellow, he had opportunities twice to visit the AgResearch Grassland Research Centre in New Zealand as a visiting post-doctoral researcher in July 2005–March 2006 and October 2007–September 2008. During his stay at the AgResearch, he learned about New Zealand’s advanced grazing management system and studied pasture mass and quality assessment using hyperspectral remote sensing tech-niques. Since July 2008, he has been an Associate Professor at the Graduate School for International Development and Cooperation, Hiroshima University. His current interest is developing environment-friendly precision farming systems applying hyperspectral remote sensing acquiredthrough multi-stage platforms (field, airborne and satellite) for precise assessment of pasture mass and quality, and GPS for monitoring grazing animals’ behavior.

Tsuyoshi AKIYAMA

Tsuyoshi Akiyama got his doctorate at theGraduate School of Agriculture, KyushuUniversity in 1974. He worked in the EcologyDivision of the National Institute of GrasslandScience, Ministry of Agriculture, Forestry andFisheries of Japan from 1973 to 1983. There he

engaged in the study of grassland ecosystems. During that time, he went abroad to study at the Laboratory for Applications of Remote Sensing, Purdue University, USA, during 1978 and 1979. After returning, he was promoted to Chief of the Laboratory of Remote Sensing, National Institute of Agro-Environmental Sciences, to apply remote sensing technology to crops and the agro-environment. In 1996, he transferred to the River Basin Research Center at Gifu University. His work there was analysis of the spatial-temporal dynamics of regional environments, such as propagation of abandoned bamboo forests in the western part of Japan and grassland degradation in Inner Mongolia, China. He is now Professor Emeritus at Gifu University since his retirement in March, 2009.

(Received 27 April 2010, Accepted 13 May 2010)