, , ， ZG*

1.College of Urban and Planning, Yancheng Teacher’s University, Yancheng 224007, China; 2. School of Geoscience and Technology, Southwest Petroleum University, Chengdu 610500, China；3.College of Geography, Nanjing Normal University, Nanjing 214007, China

Supported by National Natural Science Foundation of China (41771199); Basic Research Project of Jiangsu Province, China (BK20171277).

1 Introduction

Suaedasalsa, an annual saline herb, is a kind of plant with very strong salt tolerance[1], mainly grows in the marsh saline environment[2], and is distributed in the Northeast, North China, northwest inland and coastal areas to the north of Yangtze River[3]. It also has stronger adaptation to the saline environment, called indicator plant of saline soil[4]. Thus,S.salsais pioneer species of plant in coastal wetland. As it can absorb and accumulate much NaCl or Na2SO4to reduce the soil salinity, it is awarded the title of an admirable breed of saline soil improvement, which exerts an important influence on maintaining ecological system stability and succession[5-8]. However, with increasingly intensive human management activities, and the success of theS.alterniflora’s introduction, the landscape evolution of Salsa marsh has changed apparently and imposed negative effects on regional biodiversity and eco-environment.

Yancheng coastal wetland, including two national reserves, Yancheng Nature Reserve and Dafeng elk nature reserve, is one of the typical original coastal wetlands in China and even in the world. And it keeps natural ecological system structure and function basically[8]. Under the natural and artificial circumstance, coastal wetlands landscape has apparently changed especially the trend ofS.salsamarsh’s deviation from natural laws.S.salsaas an important part of local eco-system, is indigenous in Yancheng coastal wetland. At present, the research ofS.salsamainly focuses on the physiological characteristics, scale biological ecological adaptation, soil improvement and saline-alkali land ecological restoration from the aspect of ecological system[3-4, 10-13]. However, there are few researches on space-time evolution from the scale of landscape. By choosing Yancheng Nature Reserve core region as typical area and making a comparison between artificial administrative zones and the naturalS.salsamarsh, this paper attempts to reveal the evolution characters of Yancheng coastal area under the artificial effects, and provide a reference for guiding coastal wetland’s ecological conservation.

2 Data sources and methods

2.1StudyareaThe study area is in the core zone of Yancheng National Nature Reserve, which is north to Xinyanggang River, south to Doulonggang River and west to seawalls and is regarded as the typical growth tidal flat of wetlands in Jiangsu with a total area of 1.740×104ha. At present, the reserves are divided into north and south central parts (Fig.1).

Fig.1Locationofstudyarea

Fig.2Landscapechangesinthecoastalwetlandsinartificialareafrom2000to2011

The northern area is about 0.540× 104ha. As the reserves focus on protecting the key wintering habitat for red-crowned cranes, it is necessary to create a good habitat for red-crowned cranes and other rare species, and the artificial wetland and reed marsh restoration experiment in its core area has set up since 1993, becoming the typical artificial management area (Fig.2). Southern area is approximately 1.100×104ha. The area by human disturbance is fragile, the landscape pattern as well as evolution is mainly affected by climate, topography, hydrology, soil, vegetation and other natural factors, becoming a typical wetland area under the control of natural conditions (Fig.3)[14].

Fig.3Landscapechangesinthecoastalwetlandsinnaturalareafrom2000to2011

2.2DatasourcesThis research took three pictures of ETM+images on May 4, 2000, May 21, 2006, and September 24, 2011 as the basic data sources. ETM+images include 7 multi-spectral bands (resolution rate at 30 m) and one full-color band (resolution rate at 15 m) from the same sensor, and two kinds of resolution data can realize high precision fusion. To more accurately extract information, other auxiliary materials were used, including one diagram of 1∶400 000 coast land use map in Jiangsu province and the field investigation of 53 points GPS data. In ENVI 4.7, we used the non-supervision classification and decision tree classification method in combination with the field investigation to improve the interpretation precision, which reached 90% (Fig.2 and Fig.3). We used the ArcGIS9.3 to do overlay analysis of different images. The Fragstats 3.3 was used to calculate landscape index, and grid size was set to be 100 m. For unity in the study area, considering other landscape types of integrity at the same time, and on the basis of image in 2006 as a benchmark, 2000 and 2011 research area were cut in ArcGIS9.3.

2.3AnalysismethodsLandscape index points to highly concentrated landscape pattern information, reflecting the structure and spatial configuration and some other characteristics of simple quantitative index, which can quantitatively describe and monitor landscape structure characteristics changing with time[15]. This study selected landscape type percentage (PLAND), the degree of polymerization index (AI), average plaques area (PA_MN) to describe the basic characteristics of landscape; and used annual change area (A_RAT) to compare the rate of change with different drivers in different period of landscape. The study method of the centroid change in land use was introduced to the space changes of landscape. The centroid changes ofS.salsamarsh were used to reveal its spatial evolution rules[16-17]. Formula is as follows:

whereXandYare the weighted centroid coordinates according to the area;XiandYiare the centroid coordinates ofS.salsamarsh’s patch;Ciis the No.iarea ofS.salsamarsh’s patch;nis the number ofS.salsamarsh’s patches.

3 Results and discussions

3.1Spatial-temporalevolutionoflandscapestructureinsalsamarshesDue to the different drivers, in the southern and northern parts of the study area, there was obvious differentiation in spatio-temporal changes. The area ofsalsamarsh in these two areas decreased overall. Comparing the two area we found that thesalsamarsh area of northern artificial zone was reduced 87.16%, almost a decrease of nearly 22% compared with southern natural area. At the same time, the average patch area ofsalsamarsh was reduced 94.50% in artificial area, nearly 11% more than that in the southern area.AIin the artificial zone fell from 95.780 to 65.455 and fell from 95.716 to 81.337 in natural area during 2000-2011. In addition, from the rate of landscape changing, we saw that the rate ofsalsamarsh evolution changes was accelerated in artificial area, but under the control of natural conditions, the rate changed little. From 2000 to 2006, the area was reduced to 296.998 ha/a. the evolutional rate ofsalsamarsh was 214.822 ha/a in natural area during 2006-2011, but in artificial area, the evolutional rate ofsalsamarsh obviously increased rapidly first and then slowly from 2000 to 2006, the area was reduced to 118.167 ha, only 51.500 ha/a during 2006-2011.

Table1Landscapefeaturesofsalsamarshinnaturalareaandartificialarea

LandscapeindexesArtificialareaMay2000May2006Sept．2011NaturalareaMay2000May2006Sept．2011PLAND22．3038．7652．86429．74818．46310．301PA＿MN(ha)389．33365．57121．429298．071112．60948．167AI96．78087．54265．45595．71690．62681．337A＿RAT(ha/a)118．16751．500-296．998214．822

Through the further analysis of landscape zone changes insalsamarsh by superimposing images of different periods, we can draw that artificial cofferdam accelerated the change of landscape zone. In 2000-2006, the average width ofsalsamarsh decreased from 2 570.800 m to 1 010.272 m in artificial area, decreasing by 60.70%. Compared with the natural areas, it decreased more than 23%;salsamarsh in artificial area showed the characteristics of contraction from two directions to center, but in natural area showed the contraction features to the sea directions. During 2006-2011,salsamarsh presented the properties from the sea and land two directions to center both in the south and north. But the average width ofsalsamarsh decreased by 67.320% in artificial area, decreased more than 23% compared with natural area.

The landscape centroid ofsalsamarsh in different times was calculated by using the formula one. We can see that the spatial evolution direction changed insalsamarsh under artificial cofferdam. During 2000-2006, the centroid ofsalsamarsh migrated by 1 042.710 m to north-east under the control of natural conditions. Under the influence of human, the centroid ofsalsamarsh migrated to south-east by 666.350 m. From 2006 to 2011, the centroid ofsalsamarsh migrated 88.329 m to the east and 320.029 m to the north in natural area, the main migration to the north. However, under the influence of human, it migrated 502.474 m to the east, 67.276 m to the north, to the east primarily.

3.2TheeffectofartificialcofferdamsandS.alternifloraexpansiononS.salsamarshLandscape centroid changes inS.salsamarsh by human activities and natural processes from 2000 to 2011 were shown in Fig.4-5.

Fig.4LandscapecentroidchangesinSuaedasalsamarshbyhumanactivitiesfrom2000to2011

Fig.5LandscapecentroidchangesinSuaedasalsamarshbynaturalprocessesfrom2000to2011

The models in the artificial area are recovering reed marshes with artificial efforts and breeding mainly by building dams in north of the core region of Yancheng national natural reserve. The succession process fromsalsamarsh to reed marsh can be accelerated by man-made methods to further change the hydrological process and improve the environment ofsalsawith artificial cofferdam preventing the coming of tide and artificial breeding reeds. According to the results of soil and surface water salinity monitoring in Yancheng nature reserve in April 2011, we can see (Table 2) that: soil moisture in the artificial area was significantly higher than in the nature area; soil salinity in the artificial area was lower than in the nature area exceptS.alternifloramarsh. Through monitoring the salinity of the surface water in reed marshes both in the tide ditch in natural wetland and cofferdam area in artificial area, we can find that the salinity of the former was as high as 1.090%, while the latter was only 0.230%. So we can say that artificial cofferdam makes the hydrological conditions more conducive to the development of fresh water reed marshes in artificial area. At the same time, soil moisture increase and soil salinity decline insalsamarsh, changed the habitats of salsa which rendered it more conducive to the direction of the development of reed marshes.

Table2Thecontrastofsoilmoistureandsalinityofcoastalwetlandsbetweennaturalareaandartificialarea

ArtificialareaSpartinamarshSalsamarshReedmarshNaturalareaSpartinamarshSalsamarshReedmarshMoisture47．36342．07638．83446．33240．70336．820Salinity2．0060．8170．3081．3280．9530．347

From 2000 to 2006, in the artificial area, through the artificial cofferdam, 539 hasalsamarshes changed into reed marshes and aquaculture ponds, accounting for 46.14% ofsalsamarsh area, and compared with natural conditions, the transfer rate was nearly 4% higher. From 2006 to 2011, 178 hasalsamarshes in the artificial area were transferred into reed marshes, accounting for 38.78% ofsalsamarsh area in 2006, compared with natural conditions, the transfer rate was nearly 20% higher (Table 3).

Table3Landscapetransitionmatrixbynaturalprocessesandhumanactivities

LandscapetransitionHumanactivities2000—20062006—2011Area∥haTransitionrate∥%Area∥haTransitionrate∥%Naturalprocesses2000—20062006—2011Area∥haTransitionrate∥%Area∥haTransitionrate∥%Ponds15713．4400370．8900Reed marsh38232．717838．78173241．5151419．85Spartinamarsh17815．2413830．07882．1153820．77Suaedamarsh45138．6114231．15231655．5153859．38

At the same time, due to the impact of human during 2000-2006, nearly 50% ofsalsamarshes were transferred into reed marshes, aquaculture ponds and dams. However, during 2006-2011, there was no new cofferdam area,salsamarsh in the cofferdam area disappeared, all transferred into reed marshes and aquaculture ponds, so the change rate ofsalsamarsh in this period slowed down. Therefore, the evolution rate ofsalsamarsh under the artificial cofferdam presented the obvious characteristics of being fast at first and then getting slower afterwards. In addition, the centroid change ofsalsamarsh and reed marsh has consistency in artificial area. In 2000-2006, reed marshes in south-east direction were offset 429.720 m; in 2006-2011, reed marshes in northwest direction were slightly offset 199.246 m (Fig.6).

S.alterniflorawas introduced in the 1980s in Jiangsu coastal wetlands, and formed a large community in the 1990s, then its area expanded quickly, becoming the dominant wetland salt vegetation of marine marsh.S.alterniflorawas distributed in the upper intertidal zone, and the top edge to the mean high tide, the lower edge to the mean sea-level, tide assault frequency was between 20% and 80%[18]. BecauseS.alterniflorahas very strong function of promoting deposition, which causes the elevation in-

Fig.6Landscapecentroidchangesinreedmarshbyhumanactivitiesfrom2000to2011

crease inS.alternifloramarsh, and makes the tide cross it and reachsalsamarsh,S.alternifloraexpands to the lower edge ofsalsamarsh along tide ditch. At the same time, becauseS.alterniflorahas the characteristics of wide ecological range, it can settle down on the lower edge ofsalsamarsh, which results in the appearance of niche overlapping and interspecific competition.S.alterniflorahas two modes of reproduction, namely sexual reproduction and asexual reproduction. Sexual reproduction has certain advantage in adapting to different environment, the offspring produced by asexual reproduction have the same genetic composition as the parent. However, the reproduction ofsalsadepends on seed dispersal. The plant is short and small, they have little chance to survive when the tide assault frequency is more than 20%. So in the competition ofsalsaandS.alterniflora, the growth space is easier to be used and occupied byS.alterniflorain the bottom edge ofsalsamarsh[19-20], which hasformed the evolution pattern ofsalsamarsh toS.alternifloramarsh. In addition, artificial cofferdams effectively play the corridor function, not only stoppingS.alternifloraexpansion to land direction, but also changingS.alterniflora’s expansion direction, and turning to expand in south to occupy the living space ofsalsa. Through the calculation ofS.alternifloralandscape centroid in 2006 and 2011, we found that it was offset in south-east direction slightly by 223.169 m, accelerated the transformation fromsalsamarsh toS.alternifloramarsh (Fig.7).

Fig.7LandscapecentroidchangesinSpartinaalternifloramarshbyhumanactivitiesfrom2000to2011

Through the comparison of the two areas, during 2000-2006, 178 hasalsamarsh in the artificial area was transferred intoS.alternifloramarsh, accounting for 15.24% of the area ofsalsamarsh in 2000, and compared with the natural area, the transferring rate was higher than 13%. From 2006 to 2011, 138 hasalsamarsh was transferred intoS.alternifloramarsh, accounting for 30.07% of the area ofsalsamarsh in 2006, and under natural conditions, the transferring rate was nearly higher than 10% (Table 3). At the same time, due to the expansion ofS.alternifloramarsh, it made the trend ofsalsamarsh fragmentation obvious, from 2000 to 2011, the average patch area ofsalsamarsh decreased by 94.50% in the artificial area, more than 11% than in the natural area; landscape gathering index in the artificial area dropped from 95.780 to 65.455, decreasing by 31.66%, more than 16% than in the natural area.

4 Conclusions

Through the spatial-temporal analysis, this paper revealed the influence of artificial cofferdam andS.alternifloraexpansion on the evolution ofS.salsamarsh in Yancheng coastal wetland. We drew the following conclusions.

Artificial cofferdam affected the landscape evolution insalsamarsh. Artificial cofferdam made artificial area transform into freshwater ecosystem, and meanwhile the freshwater reed marsh did benefit a lot from it. During the year 2000 to 2006, in the artificial area, by artificial cofferdam, 539 hasalsamarsh was transferred into reed marshes and aquaculture ponds, of which the transformation rate was nearly 4% higher than in the natural area. While from 2006 to 2011, the transformation rate was 20% higher than in the natural area, 178 hasalsamarsh was transferred into reed marsh. In the artificial management area, the consistency of landscape centroid change between reed marsh andsalsamarsh showed that artificial cofferdam was an important factor forsalsamarsh evolution.

Owing to the powerful ability of spreading and competition ofS.alternifloraspecies, the coastal wetland has formed the pattern of "Salsa-Spartinamarsh". In the artificial area and the natural area, during 2000-2006, 15.24% and 2.11% ofsalsamarsh was transferred into theS.alternifloramarsh; from 2006 to 2011, 30.07% and 20.77% ofsalsamarsh was replaced by theS.alternifloramarsh. At the same time, the trend of thesalsamarsh was quite apparent as a result ofS.alternifloramarsh expanding.

Salsamarsh is mainly developed in salt marsh environment. It is the original landscape type in the coastal wetlands and plays an important role in maintaining landscape diversity. However, human activities and invasion ofS.alterniflorahave decreased the area ofSalsamarsh and even made it disappear, having a negative impact on regional biodiversity. Therefore, starting from protecting native species ofS.salsaand maintaining landscape diversity, we identified temporal spatial evolution characteristics and response mechanism ofsalsamarsh in the coastal wetlands under different drivers, which was of positive significance to managing and using the coastal wetlands scientifically.

[1] DING HR, HONG LZ, YANG ZQ,etal. Progress of study on halophyteSuaedasalsa[J].Acta Agriculturae Jiangxi, 2008, 20(8): 35-37.

[2] MA DZ, LIU HL. The herbal of Ningxia [M].Yinchuan: Ningxia People’s Press,1986: 256-258.

[3] MENG QF, YANG JS, YAO RJ,etal. Improvement of saline soil by plantingSuaedasalsaand adding nitrogenous and phosphate fertilizers in the northern Jiangsu Province [J].Pratacultural Science, 2012, 29(1): 1-8.

[4] ZHANG XJ, FANSJ, LI FZ. Development and utilization ofSuaedasalsain China[J].Chinese Wild Plant Resources, 2003, 22(2): 1-3.

[5] MAO PL, CHENG WL, LIU YH,etal. Biomass allocation inSuaedasalsapopulation in different habitats of coastal zone[J].Ecology and Environment, 2011, 20(8-9): 1214-1220.

[6] ZHANG LB, XU HL, ZHAO GX. Salt tolerance ofSuaedasalsaand its soil ameliorating effect on coastal saline soil [J].Soils, 2007, 39(2): 310-313.

[7] LIN XZ, SHEN JH, LIU KZ,etal. Study on remediation effects ofSuaedasalsaL.planting on coastal saline soil [J]. Advances in Marine Science, 2005, 23(1):65-69.

[8] ZHAO KF, FAN H, JIANG XY,etal. Improvement and utilization of saline soil by planting halophytes [J].Chinese Journal of Applied and Environmental Biology, 2002, 8(1):31-35.

[9] LIU QS, LI YF, ZHU XD. Characteristics of coastal wetland ecosystems and their healthy design: a case study from Yancheng Natural Reserve, Jiangsu Province, China[J].Acta Oceanologica Sinica, 2003, 25(3): 143-148.

[10] GUAN B, YU JB, LU ZH,etal. The ecological effects ofSuaedasalsaon repairing heavily degraded coastal salinealkaline wetlands in the Yellow River Delta[J].Acta Ecologica Sinica, 2011, 31(17): 4835-4840.

[11] LI HS, LI CH, SHEN YX,etal. Fealtures of the ecological distribution and growth rivalrousness ofSuaedasalsa[J].Jiangsu Agricultural Sciences,2009, 37(2): 296-298.

[12] CUI BS, HE Q, ZHAO XS. Researches on the ecological thresholds ofSuaedasalsato the environmental gradients of water table depth and soil salinity[J].Acta Ecologica Sinica, 2008, 28(4):1408-1418.

[13] MOU XJ, SUN ZG,WANG LL,etal. Characteristics of nitrogen accumulation and allocation ofSuaedasalsain different growth conditions of intertidal zone in Yellow River Estuary[J].Wetland Science, 2010, 8(1): 57-66.

[14]ZHANG HB，LIU HY, HAO JF,etal. spatio-temporal characteristics of landscape change in the coastal wetlands of Yancheng caused by natural processes and human activities[J].Acta Ecologica Sinica，2012，32(1)：101-110.

[15] WU JG. Landscape ecology(2nd ed)[M]. Beijing: Higher Education Press, 2007: 106-106.

[16] WANG XL, BAO YH. Study on the methods of land use dynamic change research[J]. Progress in Geography, 2009, 18(1): 81-87.

[17] GONG ZN, ZHANG YR, GONG HL,etal. Evolution of wetland landscape pattern and its driving factors in Beijing[J].Acta Geographica Sinica, 2011, 66(1): 77-88.

[18] ZHANG RS, SHEN YM, LU LY,etal. Formation ofSpartinaalterniflorasalt marsh on Jiangsu coast, China[J].Oceanologia et Limnologia Sinica, 2005, 36(4): 358-366.

[19] YUAN HW, LI SZ, ZHENG HZ,etal. Evaluation of the influences of foreignSpartinaalternifloraon ecosystem of Chinese coastal wetland and its countermeasures[J].Marine Science Bulletin, 2009, 28(6): 122-128.

[20] CHEN ZY,WANG GX, LIU JN,etal. Relative competitive ability ofSpartinaalterniflorapatches to native species in tidal zone ecotone of north Jiangsu [J].Ecology and Environmental Sciences, 2011, 20(10)：1436-1442.