???????

ORIGINAL PAPER

The inf l uences of biotic and abiotic factors on the occurrence and severity of poplar canker disease in Qingfeng County,China and the management implications

Zhigang Ma1,2?Jingle Zhu1,2?Zhiqiang Sun1,2?Jun Liang3?Zhaoxin Zhang4?Limin Zhang4?Lijuan Sun4?Wenjuan Li4

?Northeast Forestry University and Springer-Verlag Berlin Heidelberg 2015

Landscape pathology is a research approach that can provide validation of the effectiveness of regional controls of forest disease at a landscape scale.In this paper, we analyzed the effects of stand features,management approaches,and geographical locations on poplar canker disease incidence(DI)and disease severity index(DSI)of individual trees at a 10 km×10 km mesoscale landscape in Qingfeng County,China.DI varied signif i cantly with stand age,tree densities,and the degree of canopy closure. DI in stands younger than 4 years old was signif i cantly lower than that in the stands over 6 years old and reached the highest value at a stand age of 8–10 years.Overall,DI was positively correlated with stand age,stand density,and the degree of canopy closure.DI was signif i cantly lower in agro-forest stand patches than in other three patch types, i.e.isolated patch,pure stand patch,and mixed stand patch. Poplar plantations distributed around and near to villages exhibited signif i cantly higher DI mainly due to human activities and herbivores.Fragmentation or connectivity in this mesoscale landscape seemed not impact disease occurrence.DSI was not signif i cantly correlated with stand density,but varied signif i cantly with tree varieties and trees ages.DSI was highest in stands of 10–12 year trees for all poplar varieties we studied here.Plantation density and plantation age were thus critical factors in determining DI and DSI.A logistic predictive model of disease occurrence was developed for the study area,considering varieties, age,height,density,canopy cover,stand types,patch types, management status,and stand geographical locations.Our study here shows that adjustment of stand density by thinning at different plantation ages is an effective approach controlling the occurrence canker disease in short-rotation poplar plantations at the landscape scale.

Landscape pathology?Poplar plantation?Canker disease?Disease incidence?Disease severity incidence?Stand features?Adaptation

Introduction

Poplariswidely distributed in China and itsculture goesback several millennia(Li and Zhou 2000;Jin 2003;Freer-Smith et al.2007).There are over 100 species of poplar around the world and over 53 species in f i ve taxonomic sections across China(not including intersectional hybrids)(Li and Zhou 2000).Nowadays,Populusclonesare used asone ofthe most economically and environmentally important sources ofrenewable energy and raw material for afforestation and short-rotation forestry(SRF)plantations in China(Li et al. 2005;Liang et al.2006).

The total area of poplar plantations reached 10 million ha by 2008,with 7 million ha of poplar SRF plantation,to provide biomass and saw timber.In addition to providing raw materials,these plantations also provide ecological services such as wind protection,water purif i cation,pollution control,carbon sequestration,waste disposal,and site reclamation(Coyle et al.2005;Liang et al.2006;Fang et al.2007;Christersson 2008).

China accounts for 73%of the world’s total poplar plantation area,while it accounts for only 53%of the global plantations for annual wood production(Ball et al. 2005;Liang et al.2006).Widespread adoption of poplar SRF production in China,however,is hindered by the occurrence of serious diseases since poplars are host to numerous pathogens that can lead to diseases in all parts of the tree(Dickmann et al.2001;Nenad et al.2008;Hui 2009).Xu et al.(2009)outlined the Chinese literatures on poplar diseases in the past 40 years and listed 246 pathogen organisms from 98(including species,clones,and cultivars)Populus,included 205 fungi(30 Ascomycota,99 Basidiomycota and 68 Mitosporic fungi),12 bacteria,etc. (Xu et al.2009).

Over the past few decades,many poplar pathogen outbreaks have occurred over regional rather than local scales. A leaf anthracnose caused by a species ofMarssoninahas damaged poplar clones all over the world,and limits the planting of many susceptible clones(Steenackers et al. 1996).In addition to growth reductions of 40–60%in some cases,repeated early defoliations deplete food reserves,subjecting trees to winter damage and infection by secondary fungi,such asDothichiza,Dothiorella,andCytospora,which cause cankers and dieback.

Dieback caused byCytosporafungi leads to 50% mortality rate in some plantations in central China,whereas cankers byDothichizaandDothiorellafungi lead to 50 to 100%disease incidence all around China(Liang et al. 2005,Zhang et al.2008).Recently,a new canker disease caused byFusarium solanihas led to 50–70%rapid mortality in SRF plantations compromised byPopulus×euramericanacv.‘74/76’,P.euramerican‘zhonglin 46’andP.deltoidsBartr.CI.Lux I-69/55 in Henan and Shandong Province,China(He et al.2009).

Although we now have a good understanding of the biology and impact of pathogens causing many of the important diseases of hybrid poplars at f i ne-scale(i.e. within nurseries and plantations),the situation of control of poplar SRF diseases over large areas show no sign of improving.In fact,the spread of exotic pests brought about by trade and established pests continue to impact host populations.In this regard,many authors agree that understanding(and,ultimately,managing)of diseases on a regional scale requires a broader scope of investigation (Holdenrieder et al.2004,Plantegenest et al.2007;Sun et al.2010).As a result,there has been a widespread adoption of epidemiological approaches for managing diseases.

An increasingly important interdisciplinary f i eld–landscape pathology—has emerged from the incorporation of landscape ecological concepts and methods into the science of forest pathology.Hence,by utilizing landscape concepts and mathematical tools,spatial modeling methods,and landscape metric analyses have proved useful for investigating the spread of plant pathogens and the incidence and severity of tree diseases(Holdenrieder et al.2004,Ostfeld et al.2005,Plantegenest et al.2007;Sun et al.2010).

Recent work highlights the importance of landscape pathology for understanding the mechanisms of tree disease epidemics over large spatial scales,including trembling aspen(Populus tremuloides),which suffered rapid mortality in concentrated patches covering 56,091 ha on the Mancos-Dolores Ranger District,San Juan National Forest,Colorado in 2006(Worrall et al.2008).These studies suggested that agents that typically killed mature trees in aspen stands were not signif i cant factors for tree mortality.Instead,a group of interchangeable,usually secondary agents were most commonly associated with mortality,includingCytosporacanker(usually caused byValsa sordida),aspen bark beetles(Trypophloeus populiandProcryphalus mucronatus),poplar borer(Saperda calcarata),and bronze poplar borer(Agrilus liragus).The recent drought accompanied by high temperatures reduced trees vigor and indirectly accelerated trees death(Worrall et al.2008).

On the other hand,landscape pathology can help discern impact of abiotic factors on host susceptibility.La Manna et al.(2008)analyzed impacts of landscape climatic, topographic,and edaphic attributes on the incidence of theAustrocedrus chilensisdisease syndrome in the‘16 de Octubre’Valley(Chubut,Argentinean Patagonia),using remote sensing,geographic information systems,and statistical methods.They suggested that the occurrence and incidence of theA.chilensisdisease syndrome was strongly related with poor soil drainage(La Manna et al.2008).

Another analyses of longitudinal data from unmanaged old forests in the western United States showed that background(non-catastrophic)mortality rates have increased rapidly in recent decades,with doubling periods ranging from 17 to 29 years among regions.Regional warming and consequent increases in water def i cits are likely contributors to the increases in tree mortality rates (Van Mantgem et al.2009).

Zhang et al.(2005)compared disease incidence and severities in 14-year old poplars within 5817 ha in HongqiForest Farm in Heilongjiang province,China.They indicated that disease incidence and severities differed significantly among 25 poplar species and varieties;however, they did not report the impact of local biotic and abiotic factors on occurrence of disease(Zhang et al.2005).

To date there are rare studies designed to understand mechanisms of poplar plantation disease epidemics over large spatial scales at human-generated landscapes in China,through methodology and principles of landscape pathology.Consequently,there is little guidance available to forest managers on the long-term disease control for managed poplar SRF communities at a landscape scale.As China possesses the largest poplar SRF plantations around the world(Ball et al.2005),the need for such insight is critical.

From an applied point of view,landscape pathology studies can help identifying both biotic(plantation features)and abiotic characteristics(i.e.site conditions and management approaches)impacting poplar SRF plantation disease risk and further help designing SRF plantation protection and management strategies with the objective to reduce the disease risk.The objectives of the present study were to identify the relationship between the canker disease incidence of the poplar SRF plantation and stand characteristics,management status,and site features at the landscape scale,and to create a predictive model that relates the disease occurrence to these features.We hypothesized that spatial distribution of the poplar canker disease at a landscape scale is controlled by stand age and plantation densities,management approaches,and environmental features and predicted that the disease incidence would be greater on sites with higher stand density,in unmanaged stands, and nearer to villages where there were more anthropogenic interferences.

Materials and methods

Study area

Qingfeng County—located on alluvial plain of the Yellow River delta at the junction of Henan and Shandong Province at 114°27′–115°23′E and 35°45′–36°5′N—comprises a total land base of approximately 872 km2.The climate of this region is characteristic of humid monsoon temperate climatic zones,with a frost-free period of 210–215 days and a mean annual temperature of 13.4°C. Annual precipitation ranges between 700 and 900 mm. Elevation is between 45 and 55 m.Soil is deep and mainly sandy loam.The water table ranges from 2.1 to 3.7 m.

This region has undertaken extensive afforestation by converting farmland into commercial,intensively managed SRF plantation and agro-forestry systems since the 1990s. These commercial SRF plantations and agro-forestry systems have been rapidly adopted by farmers for economic gain.

Most plantations are cut on an 8–12-year rotation for pulp or saw timber;small holdings are cut unsystematically leading to considerable variation in stand age.At present, the region is a human-generated mosaic of plantations of varying ages and varieties/clones.Populus euramerican‘zhonglin 46’(P-46)andPopulus×euramericanacv.‘74/ 76’(P-107)make up ca.35 and 25%of forest density, respectively.Other varieties includeP.tomentosatriploid clone(P-triploid),P.langfangnsis,P.tomentosa,P.deltoidsBartr.CI.Lux I-69/55(I-69),P.×euramericanacv.’Guariento’(P-108)in the region,representing 40%of total forest compositions.These varieties are planted either as pure SRF plantation or mixed with one another or withPaulowniaspp.to form agro-forestry and shelterbelts.

Methods

Our study site was concentrated in a 10 km×10 km area around Shaoyang Town,in Qingfeng County.Most plantations in this county were planted in widths ranging from 200 to 3000 m along two provincial roads across this area. Forest distribution was surveyed by point sampling at the intersections,in a 1 km grid yielding 10 cells/km2.Sample cells(15 m×15 m each)were randomly located according to a plantation distribution map provided by local forest bureau.In addition,a 100 m resolution distribution map of the poplar SRF plantation satellite images was used.The location and perimeter of each sample cell was determined using a global positioning system on the ground.

We classif i ed patch types by the stand types in the study area(Perkins and Matlack 2002).Stand types were classif i ed to pure stand,mixed stand,agro-forestry stand,and isolated patch stand(a minimum size of 700 m2,see La Manna et al.2008).Other stands were mainly planted as shelterbelt,such as along road sides,on ditch orriver bank, or in surrounding villages.

From July to September 2009,we recorded,at ten cells in each grid,the plantation types(pure,mixed,agroforestry,or shelterbelt)and age;varieties or clones;individual tree status(healthy/diseased and dead);cause of disease and death(Dothichiza,Dothiorella,Cytospora,and newly foundFusarium solani;longhorn beetle,defoliators);height and diameter at breast height(DBH);and the spacing of individual trees.We measured the canopy cover of each cell using the CI-110 Plant Canopy Digital Imager (CID Inc.,Vancouver,Washington,USA).

Disease status for poplar trees in each cell was described by disease incidence(DI%)and disease severity index (DSI%),respectively.DI was calculated by following formula:

here diseased trees referred to those with canker symptoms.

DSI was used to compare severity of disease occurrence. The number and area of current-year cankers on a trunk ranging 0.8–1.8 m high from ground were visually investigated and recorded.For each individual tree,DSI was calculated by following formula: year rotation,data from stands 12 years old were excluded because of their low proportions(＜4%).

A one-way ANOVA using Fisher’sftest(p＜0.05)was conducted to detect any signif i cant differences among plantation age,density and canopy cover in DI,using Duncan’s multiple comparison(Tukey’sttest,α=0.05)to determine signif i cant differences of DI among these plantation characteristics by using harmonic mean of the group

For each sample cell falling in the forest,management status was classif i ed as unmanaged(with no pruning,no thinning,no fertilizing and no weeding)and managed(with pruning,or thinning,or fertilizing,and weeding),respectively.Due to the vast investigation area,we did not classify management status in detail.For example,managed plantation might have been tended by one or more approaches described above.Any plantation,which had received at least one tending approach,was classif i ed to managed class for determining the effect of management on canker disease occurrence and severity.

For each cell,the shortest distance to raoad,village, river,ditch,and farmland was recorded.If any cell simultaneously had the same distance to more than one of the above variables,the distance to road or village was recorded in priority for determining anthropogenic inf l uences.

Data analyses

In order to analyze these data with statistical tests,1000 cells were drawn at random from the total area.A sub sampling of 239 cells was obtained because these cells were separated from each other＞150 m in distance (Ramsey and Schafer 1997l;La Manna et al.2008),and they were located within patches of differentPopulusvarieties,plantation types and varying ages.These 239 cells were considered for statistical analysis.

In this study area,stands with age ranging 2–4 years old make up 15%of whole plantation compositions accounting for 23%of forest area.Forty-eight percent of stands were 5–8 years old and comprised 47%of plantation compositions.The other 30%forest area was comprised by stands more than 7 years old,in which there were less than 4%of stands more than 13 years old.Due to an 8–12-sizes,among which canopy coverwascategorized to group of＜30,30–70 and＞70%,and stand density was classif i ed to group of＜750 trees/ha,750–1200 trees/ha and＞1200 trees/ ha.Correlations between DI and plantation age,density,and canopy cover were performed by Pearson correlation analysis and tested by two-tailed signif i cance tests.

Attest was conducted to assess management intensity impacting DI between extensively and intensively managed stands.In order to compare relations between DI and stand geographical location,a one-way ANOVA using Fisher’sftest(p＜0.05)was conducted to detect any signif i cant differences among stands with the four locations (i.e.shortest to road,or to village,or to river/ditch,or to farmland),using Duncan’s multiple comparison(Tukey’sttest,α=0.05)to determine any signif i cant differences by using harmonic mean of the group sizes.

A one-way ANOVA using Fisher’sftest(p＜0.05)was conducted to detect any signif i cant differences among plantation density in DSI.Correlation in DSI and plantation density was performed by Pearson correlation analysis and tested by two-tailed signif i cance tests.A two-way ANOVA using Fisher’sftest(p＜0.05)was conducted to detect any signif i cant differences of DSI among poplar varieties and ages,including P-46,P-107,P-triploid,P.langfangnsis,P. tomentosa,I-69,and P-108,using Duncan’s multiple comparison(Tukey’sttest,α=0.05)to determine any signif i cant differences of DSI among these varieties by using harmonic mean of the group sizes.

A logistic regression model was used to analyze poplar canker disease occurrence.Poplar canker disease in this area was mainly caused by secondary pathogens,in which individual tree vigor was a key determinant for disease occurring.Therefore,disease occurrence in each individual tree could be treated as a dependent event.The binarydependent variable was disease occurrence(i.e.,diseased tree)and disease absence(i.e.,healthy tree).The independent variables were:Populus varieties,age,tree height, plantation density,canopy cover,stand types(also patch types),management status,and stand geographical locations.Three categorical variables were also considered: varieties(seven clones),stand types(four categories).Since there was only one soil type in the study area,we assumed that site conditions(including soil and weather)had an equal effect on all poplar varieties/clones in this area,we therefore did not consider soil type(plus weather condition) as an independent variable in the data analysis.This allowed us to discern the adaptation ability of poplar clones in the study area according to their disease severity indexes.

The logistic regression presents the following formula:

where P is the probability of the poplar canker disease occurrence;is the Y-intercept;andare the coeffi cients assigned to each of the independent variables during regression.The V letters represent the various independent variables.Probability values can be calculated based on the equation below,where e is the natural exponent:

All data was analyzed using the SPSS 17.0 Software Package.For percentage DI and DSI,data were transformed bywhere i was DI or DSI in i th cell.

Results

Stand features related to DI

At landscape scale,DI differed signif i cantly among stands of various ages(F=4.711,p＜0.001),with different densities(F=3.890,p=0.021)and in stands with various canopy cover(F=11.122,p＜0.001).Generally,DI in stands less than 5 years old was signif i cantly lower than that in stands more than 6 years old according to Duncan‘s multiple comparison analysis.DI peaked when stands were eight to 10 years old(Fig.1),and no signif i cant difference was found in DI among stands with age ranging from 7 to 12 years old.

Fig.1 Comparison of disease incidence(DI)among stands with age range 1–12 years old in Qingfeng County.The column shows mean DI and columns with the same letter are not signif i cantly different by Duncan’s multiple comparison using Tukey’s t test(a=0.05)

DI in stands of lower densities(＜750 trees/ha)was signif i cantly lower than that in stands of higher densities (＞1200 trees/ha)according to Duncan‘s multiple comparison analysis(Fig.2),while there was no signif i cant difference in DI between stands of lower density and intermediate density(750–1200 trees/ha).No signif i cant difference of DI was found between slightly dense crown (canopy cover＜30%,n=14)and intermediate crown (canopy cover ranging 30–70%,n=75),while their DI was signif i cantly lower compared to suppressed crown with canopy cover＞70%(n=150)(Fig.3).

According to Pearson correlation analysis,DI was signif i cantly positive correlated with stand age,stand densities and canopy cover(Table 1).Meanwhile,stand density was negatively related to DBH(R2=-0.232,p＜0.001)and stem height(R2=-0.227,p＜0.001).Canopy cover was signif i cantly positive correlated to plantation age (R2=0.389,p＜0.001),DBH(R2=0.361,p＜0.001), stem height(R2=0.415,p＜0.001),but not correlated to plantation density(R2=0.086,p=0.190).

Stand features related to DSI

Fig.2 Differences of disease incidence(DI)among stands in different densities in Qingfeng County.The column shows mean DI and columns with the same letter are not signif i cantly different by Duncan’s multiple comparison using Tukey’s t test(a=0.05)

Although signif i cant differences in DSI was found among stands of various densities(F=8.766,p＜0.001),correlations between DSI and plantation density were not signif i cant(data not shown).A two-way ANOVA analysis indicated that DSI differed signif i cantly among poplarvarieties(F=2.693,p=0.013),and meanwhile DSI differed signif i cantly among different ages for each poplar variety(F=1.917,p=0.004).Among poplar clones, mean DSI of P-107 was lowest with value＜4%,which was signif i cantly lower than that of all other poplar varieties.The highest DSI was found inP.tomentosatriploid clone(P-triploid)with mean 10.48%of DSI,which was signif i cantly higher than that of other poplar clones.In general,DSI was not signif i cantly different among ZL-46,P.langfangnsis,P.tomentosa,I-69,and P-108,and DSI of these varieties was signif i cantly lower than that of P-triploid(Fig.4).Change of DSI was analyzed among stands ranging in age from 2 to 12 years old,due to 8–12 years rotations.

Fig.3 Comparison of disease incidence(DI)among stands with different canopy cover in Qingfeng County.The column shows mean DI and columns with the same letter are not signif i cantly different by Duncan’s multiple comparison using Tukey’s t test(a=0.05)

According to Duncan’s multiple comparison analysis, DSI of two-year-old poplar was the lowest and signif i cantly lower than poplars with older age(Fig.5).DSI of poplar ranging from 2 to 9 years old did not differ signif i cantly, while an obvious increasing trend in DSI was found beginning at 6 years of growth which was signif i cantly lower than DSI in 10–12 year old classes(Fig.5).

Impact of management,stand location to DI and DSI

In the study area,poplar plantations could be distinguished between managed and unmanaged stands.Managed plantations received routine pruning,fertilizing,and wedding, while those unmanaged stands were not carefully managed by approaches mentioned above.According tottest results, overall DI was signif i cantly different between managed and unmanaged plantations(p=0.001),in which DI in managed stands with value of mean 43%was signif i cantly lower than that of unmanaged with value of mean 54%. DSI was also signif i cantly lower in managed stands than in unmanaged stands(p＜0.001).

Fig.4 Comparison of disease severity incidence(DSI)among poplar varieties in Qingfeng County.The column shows mean DSI and columns with the same letter are not signif i cantly different by Duncan’s multiple comparison using Tukey’s t test(a=0.05)

Fig.5 Comparison of disease severity incidence(DSI)among stands with age range 2–12 years old in Qingfeng County.The column shows mean DSI and columns with the same letter are not signif i cantly different by Duncan’s multiple comparison using Tukey’s t test(a=0.05)

Table 1 Signif i cance test of relations between DI and Populus stand indexes in Qingfeng County,Henan Province,China

Fig.6 Comparison of disease incidence(DI)among poplar stand types,in which stand types also represents patch types in Qingfeng County.The column shows mean DI and columns with the same letter are not signif i cantly different by Duncan’s multiple comparison using Tukey’s t test(a=0.05)

Fig.7 Comparison of disease incidence(DI)in stands nearest to road,river,village and farmland,respectively,in Qingfeng County. The column shows mean DI and columns with the same letter are not signif i cantly different by Duncan’s multiple comparison using Tukey’s t test(a=0.05)

DI differed signif i cantly among patches(F=5.448,p=0.0003).According to Duncan’s multiple comparison analysis,DI in the agro-forest stand was signif i cantly lower than that in the other four patch types(Fig.6),while DI did not differed signif i cantly among pure,mixed,or isolated patches.

DI in plantations at different locations differed signif i cantly(F=3.30,p=0.021).Among these locations, DI in stands nearest to villages was signif i cantly higher than that at other locations.DI showed no difference among stands near to the road,river,or farmland (Fig.7).Interestingly,no signif i cant difference of DSI was found among stands at these locations(F=1.168,p=0.326).

Table 2 Parameters of the logistic regression model chosen for assessing poplar disease occurrence in Qingfeng County,Henan Province,China

Logistic regression

Logistic regression was carried out considering two groups: diseased individuals and non-diseased individuals,and the logistic regression model were set up successfully (Wald=1.248,p＜0.001).The overall correctness of prediction was 68.2%,in which the correctness for predicting diseased tree was 79.8%.

The selected model parameters are shown in Table 2. According topvalues,biotic factors such as varieties and clones,stand age,stand density and tree height showed the highest relative importance in the model(p＜0.001).The result suggested that the probability of poplar canker disease occurrence increases as stand age,stand density and tree’s height increase,and in varieties/clones corresponding to P-46(p=0.001),P.tomentosa(p=0.01)and P-triploid(p＜0.001)(Table 2).

As to abiotic factors,management status presented a negative coeff i cient suggesting that the probability of disease occurrence increases in lower management levels. Stand locations had a positive affection on disease incidence(Table 2).The result also suggested that poplar disease incidence increases both in connected and in isolated patches,and decreases when inter-cropped in farmland(negative coeff i cient)(Table 2).

Discussion

This work revealed a strong relationship at the landscape scale in the incidence of poplar canker disease and plantation characteristics as well as abiotic factors,such as management approaches and stand locations.The following factors are positively related to the incidence of the disease:elderly stand age,higher stand density,canopy cover,geographical locations associated to nearer to villages,and unmanaged.

We found signif i cant positive correlations between poplar canker incidences and stand density,stand age,and canopy cover(Table 1).Previous studies have suggested that higher stand densities increase stand susceptibility because tree vigor is compromised,and stressed trees are more readily attacked by pathogens/insects(Waring et al. 1982;Larsson 1989).Poplar disease in this area was mainly caused by secondary pathogens such as cankers byDothichizaandDothiorellafungi(Liang et al.2005;Zhang et al.2008),andFusarium solani(He et al.2009),in which individual tree vigor was the key determinant for disease occurring.

As a result,stand density and stand age were considered to be determinant factors for occurrence of canker disease in our study area because these two factors were most likely the main biotic features to affect tree vigor.The relationship between poplar disease and stand density found at a detailed scale,was thus conf i rmed at a landscape scale regardless of poplar varieties/clones(Schiffer Jr 1976,Dickmann et al.2001,Newcombe et al.2001, Dickmann 2006,Nenad et al.2008).

In the study area,disease severity incidence(DSI) exhibited signif i cant difference among poplar varieties (F=2.693,p=0.013＜0.05)and among different age classes for each poplar variety(F=1.917,p= 0.004＜0.01).Lowest DSI was found in poplar clones P-107,I-69,P-108 and P-46(Fig.4).The reason for lower mean DSI found in I-69 and P-46 was probably because of large-scale thinning of 6–12 years old trees infected byFusarium solaniand fewer trees of more than 6 years old involved in the analysis.However,the f i nding of low DSI in P-107 and P-108 was still consistent to some reports in China(i.e.Yang and Fu 1990;Li and Wang 1998;Wu et al. 2000;Zhang et al.2005;Liu et al.2008).This indicated that P-107 and P-108 adapted well to local site conditions compared to other poplar clones in the study area.

Regardless to clones and varieties,DSI peaked when poplars reached 10–12 years old.High mortality of I-69 was found beginning from 6 years old caused byFusarium solaniduring our f i eld observations(Fig.6),while P-triploid was severely attacked by long-horn beetles which led to main trunks that were broken and twisted.The growth of these two clones was thus hindered by diseases leading to low value for commercial use.This indicated that these two poplar clones did not adapt well to local site conditions and therefore were not suitable for large diameter logs. Therefore,landscape pathology can help to discriminate adaptation abilities of different varieties/clones in the same geographical region,particularly under circumstances of the same climate and soil conditions.

Stanturf et al.(2001)have indicated that successful poplar plantation culture depends on proven clones,highquality sites and timely and appropriate cultural treatment. Our results strongly support this poplar plantation culture triangle theory(Stanturf et al.2001).

Tending measures involved in our study area included systematic pruning during dormancy in winter,watering, fertilizing and weeding.Overall DI in managed stands was signif i cantly lower than that in unmanaged stands (p=0.001).Likewise,DSI was also signif i cantly lower in managed stands than in unmanaged stands(p＜0.001). The results suggested that management approaches were critical for enhancing tree vigor and further decreasing disease risk both at the stand level and landscape scale.

According to our results,DI and DSI were strongly affected by patch types and stand geographical locations at a landscape scale.DI differed signif i cantly among patches comprised of pure stands,mixed stands,agro-forest stands and isolated patches.The reasons for signif i cantly lower DI in agro-forest patches than that in the other four patch types (Fig.7)include:(1)poplar intercropped in farmland may have better growing conditions,such as receiving more water and fertilizer than growing at other locations;(2) intercropped poplar may receive more pruning and pest controls;and(3)poplar in agro-forest system usually have a larger spacing of individual trees resulted in lower stand density.These conditions make poplar intercropped in farm land have a better growing environment and thus keep strong vigor to resistance to pathogens.It is also consistent that DI was lower in stand with lower densities in this area described above.

Isolated patches and patch areas at landscape scale in the natural forest system have positive effect for preventing certain insect herbivores from spreading and outbreak (Floater and Zalucki 2000;Jactel et al.2002;Perkins and Matlack 2002;Ylioja et al.2005).Establishing small polyclonal cultures of 10–20 ha islands of SRF in a matrix of other land uses is a safer alternative that will minimize the spread and impact of pests(Dickmann et al.2001).

Although we found that average DI in isolated patch plantations in this area was a little lower compared to that in pure,mixed,and shelterbelt stands(while still signif i cantlyhigherthan thatin agro-forest),DIdid notdiffersignif i cantly among these patches(Fig.7).Since poplar disease in this area is mainly vigor-induced(Luo and Zhang 2000;Zhang et al.2004),fragmentation or connectivity in this mesoscale landscape seems not to have much effect on canker-disease occurrence in poplar plantations.

There are 34 villages with nearly 50,000 civilians in this 10 km×10 km study area.Human activities were most likely the main reason leading to signif i cantly higher DI in poplar plantations nearer to these villages rather than that at other locations.Repeated wounds on poplar bark by livestock herbivores and inappropriate pruning by humans not only decrease tree vigor but also open routes for infection by pathogens(Black 1992;Bruhn et al.2002). Nevertheless,DSI of a given poplar variety/clone did not differ signif i cantly regarding to its locations.

Management implications

The above information should be considered as a guide to management practices aimed at reducing poplar disease occurrence and severity.Diseased poplar is typically found in stands that have reached saw-timber size(5 year old and above)with higher stand density(i.e.＞1200 trees/ha),and diseased poplar with higher DSI is found in clonesPopulus euramerican‘zhonglin 46’(P-46),P.tomentosatriploid clone(P-triploid),P.tomentosa,andP.deltoidsBartr.-CI.Lux I-69/55(I-69)(Fig.6).

As demonstrated by Song et al.(2003),from a 10-year permanent plot trail in Gansu Province,China,management by thinning in stands experiencing decline in Mongolian scotch pine(Pinus sylvestrisvar.mongolica)caused bySphaeropsis sapinea(Fr.)Dyke et Sutton,greatly mitigate future decline.Provide subject can increase the diameter and volume growth of the residual trees.A thinning approach to adjust stand density for elderly,higherdensity poplar stands experiencing high-disease incidence and severe cankers would be to mark trees for removal, beginning with merchantable trees of 5 years and above with the greatest probability of disease.This general approach is applicable to even-aged stands where improvement harvests or sanitation harvests are prescribed or for selecting residual and reserve trees for largerdiameter log culture.Selecting new clones to replace the un-adapted clones such asP.tomentosatriploid andP. deltoidsBartr.CI.Lux I-69/55,is recommended.

Poplar plantations established for other purposes—than sawlog,pulpwood,and chip—might be an option for avoiding disease risk.For example,carefully tended,highdensity short rotation coppice cultures(SRC)aimed to produce high-f i ber yields has now been adopted and implemented in many countries.SRC not only has a positive impact on biodiversity(small mammals,birds, insects,etc.);nutrient capture;and soil protection from wind and water erosion,but also can enhance terrestrial carbon stocks that have been recognized by the United Nations Framework Convention on Climate Change in the Kyoto protocol(Deckmyn et al.2004a,b,Labrecque and Teodorescu 2005;Afas et al.2008).

In this study area,both poplar DI and DSI exhibit low levels before they are 5 years old.Therefore,establishment of SRC for biomass yields with 2–4 year rotation is most likely effective to avoid high DI and DSI after 5 years of growth.This is being tested in this area using aP.tomentosatriploid clone with itssuperiorfast-growing characteristicsin the f i rst 5 years’of growth,despite that it cannot be cultured for larger diameter logs in an 8–10 year rotation.

Management approaches will reduce poplar plantation disease risks.Pruning and weeding are widely accepted by local civilians because of the low cost and benef i t from the pruned branches(for f i rewood)and herbs(for fodder and for manure),while the challenge is to extend appropriate pruning techniques that can be implemented by local farmers.SRC is recommended to be planted around local villages,particularly SRC for f i rewood and for fodder.As a result,anthropogenic interference to mature commercial trees could be minimized.

AcknowledgementWe thank Junfang Zhao of Puyang Forestry Institute and staff of Qingfeng Forest Bureau for providing local forest resource distribution map and helping with the f i eld investigations,we also thank Lingjun Cui of Paulownia R&D Center of China for managing and analyzing data.Many thanks for the constructive comments from the reviewer on our manuscript.

Afas NA,Marron N,Van Dongen S,Laureysens I,Ceulemans R (2008)Dynamics of biomass production in a poplar coppice culture over three rotations(11 years).For Ecol Manag 255(5–6):1883–1891

Ball J,Carle J,Del Lungo A(2005)Contribution of poplars and willows to sustainable forestry and rural development.UNASYLVA-FAO 56(2):3–14

Black HC(1992)Silvicultural approaches to animal damage management in Pacif i c Northwest forests.Gen.Tech.Rep. PNW-GTR-287,Portland.http://www.treesearch.fs.fed.us/pubs/ 25665

Bruhn JN,Wetteroff JJ,Mihail JD,Jensen RG.,Pickens JB.2002. Harvest-associated disturbance in upland Ozark forests of theMissouri OzarkForest Ecosystem Project,pp 130–146

Christersson L(2008)Poplar plantations for paper and energy in the south of Sweden.Biomass Bioenergy 32(11):997–1000

Coyle DR,Nebeker TE,Hart ER,Mattson WJ(2005)Biology and management of insect pests in North American intensively managed hardwood forest systems.Annu Rev Entomol 50:1–29

Deckmyn G,Laureysens I,Garcia J,Muys B,Ceulemans R(2004a) Poplar growth and yield in short rotation coppice:model simulations using the process model SECRETS.Biomass Bioenergy 26(3):221–227

Deckmyn G,Muys B,Quijano JG,Ceulemans R(2004b)Carbon sequestration following afforestation of agricultural soils:comparing oak/beech forest to short-rotation poplar coppice combining a process and a carbon accounting model.Glob Change Biol 10(9):1482–1491

Dickmann DI(2006)Silviculture and biology of short-rotation woody crops in temperate regions:then and now.Biomass Bioenergy 30(8–9):696–705

Dickmann D,Isebrands JG,Eckenwalder J,Richardson J(2001) Poplar culture in north America.NRC Research Press,Ottawa

Fang S,Xue J,Tang L(2007)Biomass production and carbon sequestration potential in poplar plantations with different management patterns.J Environ Manag 85(3):672–679

Floater GJ,Zalucki MP(2000)Habitat structure and egg distributions in the processionary caterpillarOchrogaster lunifer:lessons for conservation and pest management.J Appl Ecol 37(1):87–99

Freer-Smith PH,Broadmeadow MSJ,Lynch JM(2007)Forestry and climate change.CABI,Cromwell Press,Trowbridge,pp 4–5

He W,Ren FJ,Guo LM(2009)Pathogen identif i cation ofPopulus×euramericanacanker disease.Scientia Silvae Sinicae 45(6):104–108(in Chinese)

Holdenrieder O,Pautasso M,Weisberg PJ,Lonsdale D(2004)Tree diseases and landscape processes:the challenge of landscape pathology.Trends Ecol Evol 19(8):446–452

Hui C(2009)Occurrence and control of poplar canker disease. Modern Agric Sci 11:17–22

Jactel H,Goulard M,Menassieu P,Goujon G(2002)Habitat diversity in forest plantations reduces infestations of the pine stem borerDioryctria sylvestrella.J Appl Ecol 39(4):618–628

Jin ZD(2003)Present managing situations and intensive management of man-made forest in China.Chin J Eco-Agric 11(1):15–18(in Chinese)

La Manna L,Matteucci SD,Kitzberger T(2008)Abiotic factors related to the incidence of theAustrocedrus chilensisdisease syndrome at a landscape scale.For Ecol Manag 256(5):1087–1095

Labrecque M,Teodorescu TI(2005)Field performance and biomass production of 12 willow and poplar clones in short-rotation coppice in southern Quebec(Canada).Biomass Bioenergy 29(1):1–9

Larsson S(1989)Stressful times for the plant stress:insect performance hypothesis.Oikos 56(2):277–283

Li QH,Wang ZG(1998)Preliminary report on resistance to cankers of 25 poplar clones.J Gansu Agric Univ 33:434–437(in Chinese)

Li C,Zhou X(2000)Status and future trends in plantation silviculture in China.AMBIO 29(6):354–355

Li S,Zhang Z,Luo J,He C,Pu Y,An X(2005)Progress and strategies in cross breeding of poplars in China.For Stud China 7(3):54–60

Liang J,Jiang JQ,Liu HX,Zhao JP,Wang Y(2005)Study on the pathology of poplar-canker pathogen interaction in China.For Res 18(2):214–221(in Chinese)

Liang W,Hu H,Liu F,Zhang D(2006)Research advance of biomass and carbon storage of poplar in China.J For Res 17(1):75–79

Liu LL,Cao ZM,Fan JF,Wang YG,Yu ZD,Fei ZX(2008)A Study of Poplar Resistance toMelampsora larici-populina.J Northeast For Univ 23(1):132–134(in Chinese)

Luo YQ,Zhang XY(2000)Forest protection towards 21th century. Rev China Agric Sci Technol 2(1):67–72(in Chinese)

Nenad K,Milan M,Stania B,Biljana Mirjana N,Radmila K,Jelena M,Martin B,Sini A,Zoran G(2008)Diseases in poplar plantations.Glasnik 97:7–31

Newcombe G,Ostry M,Hubbes M,Perinet P,Mottet MJ(2001) Poplar diseases.Poplar culture in North America.NRC Research Press,Ottawa,pp 249–276

Ostfeld RS,Glass GE,Keesing F(2005)Spatial epidemiology:an emerging(or re-emerging)discipline.Trends Ecol Evol 20(6):328–336

Perkins TE,Matlack GR(2002)Human-generated pattern in commercial forests of southern Mississippi and consequences for the spread of pests and pathogens.For Ecol Manag 157(1–3):143–154

Plantegenest M,Le May C,Fabre F(2007)Landscape epidemiology of plant diseases.J R Soc Interface 4(16):963–973

Ramsey FL,Schafer DW(1997)The statistical sleuth:a course in methods of data analysis.Duxbury Press,Belmont

Schiffer Jr AL(1976)Poplar plantation density inf l uences foliage disease.In:Intensive plantation culture:5 years research.USDA For.Serv.Gen.Tech.Rep.NC-21.Perkin-Elmer Corporation, Washington,DC,pp 8l–84

Song XD,Liu GR,Chen JY,Wu JY,Xu GJ,Li SH,Wu XG(2003) The causes and management of the decline ofPinus sylvestrisvar.mongolica.J Beihua Univ(Natural Science Edition) 4(2):166–169(in Chinese)

Stanturf JA,van Oosten C,Netzer DA,Coleman MD,Portwood CJ (2001)Ecology and silviculture of poplar plantations.In: Dickmann DI,Isebrands JG,Eckenwalder JE,Richardson J (eds)Poplar culture in North America.NRC Research Press, Ottawa,pp 153–206

Steenackers J,Steenackers M,Steenackers V,Stevens M(1996) Poplar diseases,consequences on growth and wood quality. Biomass Bioenergy 10(5–6):267–274

Sun Z,Zhang X,Xiao W,Liang J,Zhang Z(2010)Landscape pathology—a new perspective in forest protection f i eld.Scientia Silvae Sinicae 46(3):139–145(in Chinese)

Van Mantgem PJ,Stephenson NL,Byrne JC,Daniels LD,Franklin JF,Fule PZ,Harmon ME,Larson AJ,Smith JM,Taylor AH (2009)Widespread increase of tree mortality rates in the western United States.Science 323(5913):521

Waring RH,Schroeder PE,Oren R(1982)Application of the pipe model theory to predict canopy leaf area.J For Res 12:556–560

Worrall JJ,Egeland L,Eager T(2008)Rapid mortality ofPopulus tremuloidesin southwestern Colorado,USA.For Ecol Manag 255(3–4):686–696

Wu YZ,Ji YP,Liu Y,Hou YQ,He TG(2000)Resistance to cankers of main poplar clones in Shandong Province.For Pest Dis 19(4):10–13(in Chinese)

Xu MQ,Zhou XD,Piao CG(2009)Populus cultivation-clones in different cultivated area and its diseases in China.For Res 22(5):705–714(in Chinese)

Yang JX,Fu QQ(1990)Investigations on poplar clones resistant to cankers.J Northeast For Univ 5(1):1–10(in Chinese)

Ylioja T,Slone DH,Ayres MP(2005)Mismatch between herbivore behavior and demographics contributes to scale-dependence of host susceptibility in two pine species.For Sci 51(6):522–531

Zhang XY,Luo YQ,Ye JR,Sun JH,Liang J(2004)Forest biological disasters in China in the new forestry times.For Pest Dis 23(6):8–12(in Chinese)

Zhang ZH,Dong XW,Yan DL(2005)On infectious degree of 25 species of poplar byValsa sordida.Prot For Sci Technol 4:28–29(in Chinese)

Zhang XY,Zhao JP,Liang J,Lu Q(2008)Pathogenicity differentiation of tree stem canker pathogenic fungi.For Pest Dis 27(1):1–4(in Chinese)

21 December 2013/Accepted:12 March 2014/Published online:25 July 2015

Project funding:This research was supported by the twelfth f i ve-year science and technology support program of China

(2012BAD19B0801).

The online version is available at http://www.springerlink.com

Corresponding editor:Chai Ruihai

Zhigang Ma and Jingle Zhu contributed equally to this work and should be considered as co-f i rst authors.

?Zhiqiang Sun

zq_sun@paulownia.ac.cn

1Paulownia R&D Center of China,Chinese Academy of Forestry,Zhengzhou 450003,China

2Non-timber Forestry R&D Center,Chinese Academy of Forestry,Zhengzhou 450003,China

3The Key Laboratory of Forest Protection of China State Forestry Administration,Research Institute of Forest Ecology,Environment and Protection,Chinese Academy of Forestry,Beijing 100091,China

4Puyang Forest Research Institute,Puyang 457000,China