1# RECORD 51b11fac87d973139852a88a8c12a9cdbf87b57a2# TITLE: Hedonic Analysis of Forest Pest Invasion: the Case of Emerald Ash Borer3# YEAR: 2019.0 VENUE: Forests DOI: 10.3390/f100908204# AUTHORS: Xiaoshu Li; T. Holmes; Kevin J. Boyle; E. Crocker; C. Nelson5# PAGES: 15 CANDIDATE TABLE PAGES (1-based): [7, 10, 11, 12]6# ABSTRACT: The emerald ash borer (EAB) was first detected in North America in 2002, and since its introduction, this invasive pest has killed millions of ash trees. While EAB kills native North American ash trees in all settings, its impacts have been especially large in urban areas where ash has been a dominant street tree, especially in residential areas. While some management costs, such as insecticide treatment, tree removal, or tree replacement, are relatively straightforward to compute, the impact that EAB has had on residential property values is less clear. To better understand the economic cost of EAB in urban settings, we conducted a hedonic property value analysis to evaluate the impact of ash tree damages due to EAB infestation on housing sales prices. This study was conducted in the city of Milwaukee, Wisconsin, which had high stocking levels of ash trees prior to EAB’s arrival. The objectives of the study are to investigate: (1) how EAB-infested ash trees affect property values; (2) whether the benefits from healthy ash trees to property value change after arrival of EAB; and (3) whether healthy ash trees located within infested neighborhoods provide the same benefits as the healthy ash trees located outside of infested neighborhoods. In general, our results show that the EAB outbreak has had a negative impact on home values for properties located in close proximity to the ash tree component of the urban forest. This result holds true for neighborhoods where EAB does not yet pose an imminent threat, and is amplified for neighborhoods where EAB has been detected. Our results highlight the early stages of a dynamic economic process that impacts urban residential property owners subject to the risk of EAB or other tree pests and diseases. In general, we find that forward-looking behavior of residential property owners is capitalized into property values during the process of forest pest infestation.789===== PAGE 1 (score 13, numeric-paren 0, table-words 0, var-words 13) =====1011Article12Hedonic Analysis of Forest Pest Invasion: the Case of13Emerald Ash Borer14Xiaoshu Li 1,*15, Thomas P. Holmes 216, Kevin J. Boyle 3, Ellen V. Crocker 417and C. Dana Nelson 518119Center for Health Services Research, University of Kentucky, 740 South Limestone, J515 Kentucky Clinic,20Lexington, KY 40536, USA21222Forestry Sciences Lab, Southern Research Station, USDA Forest Service, Research Triangle Park, NC 27709,23USA; tom.holmes@usda.gov24325Agricultural and Applied Economics Department, Virginia Polytechnic Institute and State University, 430A26Bishop Favrao Hall, Blacksburg, VA 24060, USA; kjboyle@vt.edu27428Department of Forestry and Natural Resources, University of Kentucky, 121 Thomas Poe Cooper Building,29Lexington, KY 40546, USA; e.crocker@uky.edu30531Forest Health Research and Education Center, Southern Research Station, USDA Forest Service, 208 Thomas32Poe Cooper Building, Lexington, KY 40546, USA; charles.d.nelson@usda.gov33*34Correspondence: Xiaoshu.li@uky.edu; Tel.: +86-859-323-707835Received: 29 July 2019; Accepted: 17 September 2019; Published: 19 September 2019363738Abstract: The emerald ash borer (EAB) was first detected in North America in 2002, and since its39introduction, this invasive pest has killed millions of ash trees. While EAB kills native North American40ash trees in all settings, its impacts have been especially large in urban areas where ash has been a41dominant street tree, especially in residential areas. While some management costs, such as insecticide42treatment, tree removal, or tree replacement, are relatively straightforward to compute, the impact43that EAB has had on residential property values is less clear. To better understand the economic cost44of EAB in urban settings, we conducted a hedonic property value analysis to evaluate the impact of45ash tree damages due to EAB infestation on housing sales prices. This study was conducted in the46city of Milwaukee, Wisconsin, which had high stocking levels of ash trees prior to EAB’s arrival. The47objectives of the study are to investigate: (1) how EAB-infested ash trees affect property values; (2)48whether the benefits from healthy ash trees to property value change after arrival of EAB; and (3)49whether healthy ash trees located within infested neighborhoods provide the same benefits as the50healthy ash trees located outside of infested neighborhoods. In general, our results show that the51EAB outbreak has had a negative impact on home values for properties located in close proximity to52the ash tree component of the urban forest. This result holds true for neighborhoods where EAB does53not yet pose an imminent threat, and is amplified for neighborhoods where EAB has been detected.54Our results highlight the early stages of a dynamic economic process that impacts urban residential55property owners subject to the risk of EAB or other tree pests and diseases. In general, we find that56forward-looking behavior of residential property owners is capitalized into property values during57the process of forest pest infestation.58Keywords: emerald ash borer; property value; forward-looking behavior; urban forests; forest health591. Introduction60The emerald ash borer (EAB) was first detected in North America in 2002, and is thought to61have been introduced from Asia by packaging materials [1]. Since its introduction, this invasive pest62has spread out from Michigan to over 30 states and killed tens of millions of ash trees. EAB is an63invasive wood-boring beetle that can kill ash trees within 2–4 years from infestation, as larvae tunnel64into the vascular system of the tree and cut offtranspiration of water and nutrients. Ash trees are65Forests 2019, 10, 820; doi:10.3390/f1009082066www.mdpi.com/journal/forests676869===== PAGE 7 (score 33, numeric-paren 0, table-words 5, var-words 20) =====7071Forests 2019, 10, 820727 of 1573Table 1. Descriptive Statistics of Property Characteristic Variables (n = 4502). (EAB = emerald ash74borer).75Variable76Means77Standard Deviations78Description79Ln(Sale price)8011.642810.00982Log transformation of sale price ($2005)83Living area841.462850.010861000 ft287Lot size887.647890.106901000 ft291Baths921.900930.01294Number95Bedrooms963.200970.01298Number99Age10065.0001010.372102Home age at sale; Years103Stories1041.2001050.006106Number107Air conditioning10867%1090.7%1101 if air-conditioned; 0 otherwise111Fireplace11232%1130.9%1141 if has fireplace; 0 otherwise115Pool1162%1170.2%1181 if a pool present; 0 otherwise119Basement area1201.0291210.0051221000 ft2123Garage area1240.4101250.0021261000 ft2127Distance to highway1282.3001290.030130Kilometers131Distance to railway1324.7001330.100134Kilometers135Distance to Airport13616.8001370.200138Kilometers139Water1401%1410.1%142Percentage of land covered by water in 1 km buffer143of parcel144Trees/Forest %14527%1460.1%147Percentage of land covered by trees and forest in 1148km buffer of parcel149Grass/herbaceous plants %15027%1510.1%152Percentage of land covered by grass & herbaceous153plants in 1 km buffer of parcel154Ash_parcel15531%1560.7%1571 if ash trees on the parcel; 0 otherwise158Ash_buffer0.115930%1600.2%161Percentage of properties located within 0.1 km162buffer of parcel that have ash trees163Ash_buffer0.516428%1650.2%166Percentage of properties located within 0.5 km167buffer of parcel that have ash trees168Ash_buffer1.016927%1700.1%171Percentage of properties located within 1 km buffer172of parcel that have ash trees173EAB_parcel1740.0021750.001176Number of EAB infested ash trees within the parcel177EAB_buffer0.11780.0471790.011180Number of EAB infested ash trees within 0.1 km181buffer of parcel182EAB_buffer0.51831.2171840.142185Number of EAB infested ash trees within 0.5 km186buffer of parcel187EAB_buffer1.01883.8031890.298190Number of EAB infested ash trees within 1 km191buffer of parcel192post1930.6201940.0071951 if sale occurred after 2011; 0 otherwise196outbreak1970.0331980.0031991 if EAB infested ash tree within 1 km of parcel200Within the study area, roughly equal proportions of the land area within 1 km of sales properties201were covered by trees/forest and grass/herbaceous plants (27% each; Table 1). Nearly one third (31%)202of sales parcels contained ash trees, and similar proportions of parcels within buffers up to 1 km203surrounding sales parcels also contained ash trees (27–30% of parcels). These statistics suggest that,204in terms of the ash component of the urban landscape, sales parcels were generally representative of205other residential parcels within the study area.206The summary statistics which describe the spread of EAB are listed in Table 2. These measurements207were based on the field surveys consistently conducted over four years. It shows that 16 ash trees were208infested in 2012, with 34 infested trees in 2013, 240 infested trees in 2014, and 528 infested trees in 2015209(Table 2). The sizes of the infested ash trees ranged from 3 to 40 inches, with a mean of 17.0 inches.210However, very few EAB-infested trees occurred on sales properties. In 2015, there were only four211property sales with infested ash trees identified on the parcel. If we assume the sale price could be212impacted by the EAB infested ash trees located within a 1 km buffer, there would be 222 property sales213impacted in 2015.214215216===== PAGE 8 (score 10, numeric-paren 0, table-words 2, var-words 6) =====217218Forests 2019, 10, 8202198 of 15220Table 2. The Number of emerald ash borer (EAB) Infested Ash Trees and Impacted Properties with Sales.221Year222Total EAB223Parcel224Buffer 0.1 km225Buffer 0.5 km226Buffer 1 km2272012228162290230123113232322332013234342351236723729238662392014240240241024212243812441642452015246528247424834249137250222251Total2525253542542602554842564. Model Specification2574.1. Traditional Hedonic Model258A fixed-effect hedonic model was estimated to address the potential for spatially correlated259omitted variables in the model specification. The base model used to estimate the effects of ash trees260ln Pit = Ziα + Liβ + Ashiγ + τt + ωi + σit261(1)262where Pit is the sale price for property i at time t, Zi is a vector of property-specific characteristics and263distance measurements, measurements (distance to road, railway, and airport) are included in the264model in both linear and quadratic functional forms, τt is the time fixed-effect for the year of sale, ωi265is the spatial fixed effect specified based on the zip code of the sales property, and σit is the equation266error. This model is estimated for the years 2008–2011 to test the hypothesis that ash trees enhanced267residential property values prior to the EAB outbreak.268After the outbreak of EAB, the damages from EAB infestation would impact the property sales. In269Model 2, we constructed a fixed-effect hedonic model following the previous literature, which linked270the forest pest infestation together with property sale values.271ln Pit = Ziα + Liβ + EABitδ + τt + ωj + σit272(2)273The property sale data for estimating Model 2 run from 2012 to 2015. A statistically significant274negative coefficient estimate associated with EABit would suggest that the presence of infested trees in275the vicinity of residential properties decreases property value. The model is separately estimated for276each of the four different spatial specifications—parcel level, and the 0.1 km, 0.5 km, and 1.0 km buffers277(Table 1).2784.2. Quasi-Experimental Design279A limitation of the Model 2 specification is that it does not provide information on whether the280capitalized value of the entire stock of ash trees is affected by the EAB outbreak. After the arrival of281EAB, both infested and healthy ash trees had the potential to affect property values. This limitation282is addressed in Model 3, which estimates whether a structural change occurred in the capitalized283value of ash trees after the arrival of EAB. We construct a difference-in-difference model following the284quasi-experiment design, which could test the specific treatment effect by comparing the changes in sale285prices before and after the outbreak of EAB between the properties with ash trees in the neighborhood286and the other properties in the study area.287The property sale data for estimation is from 2008 to 2015. The hypothesis being tested here is288that the presence of infested ash trees within a 1 km buffer of the sales parcel provides a signal that289other ash trees near the property are at risk of becoming infested in the near future.290ln Pit = Ziα + +Liβ + Ashiγ + Ashi ∗posttρ + Ashi ∗outbreakitθ + τt + ωj + σit291(3)292293294===== PAGE 10 (score 204, numeric-paren 162, table-words 2, var-words 20) =====295296Forests 2019, 10, 82029710 of 15298Table 3. Coefficient Estimates of Hedonic Model 1, Home Sales 2008–2011 (N = 1710).299Variable300Parcel301Buffer 0.1 km302Buffer 0.5 km303Buffer 1 km304Living area3050.212 ***3060.214 ***3070.215 ***3080.214 ***309(0.029)310(0.029)311(0.029)312(0.029)313Lot size3140.006 ***3150.004 **3160.004 **3170.004 **318(0.001)319(0.002)320(0.002)321(0.002)322Baths3230.069 ***3240.070 ***3250.068 ***3260.068 ***327(0.014)328(0.014)329(0.014)330(0.014)331Bedrooms3320.0103330.0083340.0103350.008336(0.012)337(0.012)338(0.012)339(0.012)340Age341−0.002 ***342−0.002 ***343−0.002 ***344−0.002 ***345(0.0004)346(0.0004)347(0.0004)348(0.0004)349Stories3500.082 ***3510.084 ***3520.081 ***3530.083 ***354(0.029)355(0.029)356(0.029)357(0.029)358Air conditioning3590.105 ***3600.106 ***3610.105 ***3620.105 ***363(0.015)364(0.015)365(0.015)366(0.015)367Fireplace3680.092 ***3690.089 ***3700.091 ***3710.092 ***372(0.016)373(0.016)374(0.016)375(0.016)376Pool3770.0143780.0073790.0133800.011381(0.052)382(0.052)383(0.052)384(0.052)385Basement area3860.125 ***3870.131 ***3880.124 ***3890.127 ***390(0.037)391(0.037)392(0.036)393(0.036)394Garage parking area3950.329 ***3960.323 ***3970.325 ***3980.324 ***399(0.049)400(0.049)401(0.049)402(0.049)403Distance to highway4040.041 ***4050.043 ***4060.044 ***4070.043 ***408(0.014)409(0.014)410(0.014)411(0.014)412Square Distance to highway413−0.007 ***414−0.007 ***415−0.007 ***416−0.007 ***417(0.002)418(0.002)419(0.002)420(0.002)421Distance to Railway422−0.036 ***423−0.034 ***424−0.032 ***425−0.030 ***426(0.011)427(0.011)428(0.011)429(0.011)430Square Distance to Railway4310.003 ***4320.003 ***4330.003 ***4340.003 ***435(0.001)436(0.001)437(0.001)438(0.001)439Distance to Airport440−0.030 ***441−0.030 ***442−0.028 ***443−0.028 ***444(0.006)445(0.006)446(0.006)447(0.006)448Square Distance to Airport4490.001 ***4500.001 ***4510.001 ***4520.001 ***453(0.0002)454(0.0002)455(0.0002)456(0.0002)457Water4580.1514590.1324600.059461−0.009462(0.322)463(0.321)464(0.323)465(0.326)466Trees/Forest4671.664 ***4681.514 ***4691.437 ***4701.429 ***471(0.206)472(0.211)473(0.225)474(0.220)475Grass/herbaceous plants4760.517 ***4770.461 **4780.440 **4790.454 **480(0.194)481(0.194)482(0.196)483(0.195)484Ash4850.0054860.165 ***4870.757 **4880.380 ***489(0.015)490(0.056)491(0.095)492(0.129)493constant49410.67849510.700 ***49610.665 ***49710.539 ***498(0.154)499(0.154)500(0.154)501(0.161)502Adjust R25030.75585040.75705050.75675060.7570507Note: *** denotes significant at the 1% level, ** denotes significant at the 5% level, * denotes significant at the 10%508level. Standard errors in parentheses.509510511===== PAGE 11 (score 24, numeric-paren 5, table-words 7, var-words 4) =====512513Forests 2019, 10, 82051411 of 155155.2. After the Outbreak of Emerald Ash Borer (EAB)516The presence of EAB-infested trees in nearby locations, combined with media attention given517to the growing outbreak, likely raised awareness among city residents that the entire stock of ash518trees in Milwaukee was at risk of infestation and death. This type of forward-looking behavior might519be detected in sales data, where EAB-infested ash trees are located in relatively close proximity to520sales properties. As the impact of nearby EAB-infested ash trees on home sales prices may be rather521subtle, a substantial number of observations that possibly include forward-looking behavior would be522needed to detect a measurable effect with statistical significance. Within our data, this possibility is523suggested, as the number of property sales within 0.1–1 km of EAB-infested trees ranges from 54 to524484 observations, respectively, representing roughly 2–17% of the 2792 sales after the EAB outbreak.525Model 2 includes EAB detections in the model specification, and results indicate that residents in526the general vicinity of known EAB detections were negatively impacted in terms of property values527by the outbreak (Table 4). While results are only statistically significant at the 1 km buffer level, and528the impacts per tree on housing price are rather small, this result may reflect an inadequate number529of impacted property sales to detect significant impacts for the other buffer levels. However, Model5302 suggests that nearby EAB detections provide a signal to homeowners that ash resources in their531neighborhood are at risk of infestation.532Table 4. Coefficient Estimates of Hedonic Model 2, emerald ash borer (EAB) Detections, 2012–2015533(n = 2792).534Variable535Parcel536Buffer 0.1 km537Buffer 0.5 km538Buffer 1 km539EAB5400.00065410.0055542−0.0005543−0.0006 ***544(0.054)545(0.0060)546(0.0005)547(0.0002)548Adjusted R25490.77375500.77385510.77385520.7742553Note: *** denotes significant at the 1% level, ** denotes significant at the 5% level, * denotes significant at the 10%554level. Standard errors in parentheses.555Model 3 provided a more nuanced test of the hypothesis that residential property owners were556forward-looking and perceived a risk that healthy ash trees on, or close to, their property may become557infested in the future. The results of Model 3 estimation generally indicated that the capitalized value558of the entire stock of ash trees located near sales properties decreased as the EAB infestation spread559(Table 5). Two measures illustrate this result. First, by summing the coefficient estimates obtained for560ash trees over the entire study period (γ) and during the post-outbreak period (ρ), a diminishment in561the capitalized value of ash trees is observed that is statistically significant within each of the buffer562zones (Sum 1 = γ + ρ). For example, prior to the EAB outbreak, a one percentage point increase in the563number of properties containing ash trees located within 0.5 km of sales properties increased property564values by about 0.76% (Table 3). However, after the outbreak began, the marginal percentage increase565in home value from properties containing ash located in that buffer was only about 0.20%. Both of566these parameter estimates are significant at the 0.01 level. Second, adding the parameter estimate for567the variable indicating that an EAB infected tree had been located within 1 km of the sales parcel to the568previous measure (Sum 2 = γ + ρ + θ) can be seen to further reduce the capitalized value of ash trees.569The existence of a healthy ash tree would no longer statistically beneficial to the community.570571572===== PAGE 12 (score 86, numeric-paren 65, table-words 5, var-words 2) =====573574Forests 2019, 10, 82057512 of 15576Table 5. Coefficient Estimates of Hedonic Model 3 for Ash and emerald ash borer (EAB) variables,5772008–2015 (n = 4502).578Variable579Parcel580Buffer 0.1 km581Buffer 0.5 km582Buffer 1 km583Ash5840.0075850.183 ***5860.298 ***5870.393 ***588(0.016)589(0.049)590(0.075)591(0.101)592Ash*Post593−0.002594−0.066595−0.102596−0.204 **597(0.020)598(0.056)599(0.077)600(0.098)601Ash*Outbreak602−0.021603−0.066604−0.093 *605−0.095 *606(0.028)607(0.048)608(0.051)609(0.054)610Adjusted R26110.76476120.76576130.76586140.7657615Sum 1616−0.0056170.117 ***6180.196 ***6190.188 **620γ + ρ621(0.040)622(0.042)623(0.064)624(0.090)625Sum 2626−0.0266270.0506280.1036290.093630γ + ρ + θ631(0.043)632(0.059)633(0.078)634(0.100)635Note: *** denotes significant at the 1% level, ** denotes significant at the 5% level, * denotes significant at the 10%636level. Standard errors in parentheses.637In Model 4, which includes the severity of EAB outbreak within a 1 km buffer in the model638specification, it is noted that the coefficient estimates for Ash*Outbreak are still negative but no longer639significant (Table 6). This is due to a high degree of correlation between this interaction variable640(indicating the occurrence of EAB) and the severity of an EAB outbreak (the number of infested trees).641Further, the negative effects from EAB-infested ash trees within a 1 km buffer are shown to be significant642at the parcel level and all buffer levels. These results suggest that one additional EAB-infested ash tree643within a 1 km buffer of a property decreases sales price by 0.1%. Based upon the average home price,644this result suggests that the decrease in sales price due to one additional EAB-infested ash tree within a6451km buffer is about $114. Looking at estimates of Sum 1 and Sum 2, it can be seen that these coefficient646estimates are similar to those estimates from Model 3 and indicate that the capitalized value of nearby647stocks of ash trees diminished after the onset of the EAB outbreak.648Table 6. Coefficient Estimates of Hedonic Model 4, Ash and emerald ash borer (EAB), 2008–2015649(n = 4502).650Variable651Parcel652Buffer 0.1 km653Buffer 0.5 km654Buffer 1 km655Ash6560.0076570.182 ***6580.301 ***6590.393 ***660(0.016)661(0.049)662(0.075)663(0.101)664Ash*Post665−0.005666−0.070667−0.104668−0.210 **669−(0.020)670−(0.056)671−(0.077)672−(0.098)673Ash*Outbreak6740.001675−0.011676−0.033677−0.044678(0.029)679(0.054)680(0.058)681(0.060)682EAB_Buffer3683−0.001 ***684−0.001 **685−0.001 **686−0.001 *687(0.0002)688(0.0003)689(0.0003)690(0.0003)691Adjusted R26920.76506930.76596940.76606950.7658696Sum 1697−0.0116980.112 ***6990.198 ***7000.183 **701γ + ρ702(0.040)703(0.042)704(0.064)705(0.090)706Sum 2707−0.0107080.1007090.164 **7100.139711γ + ρ + θ712(0.044)713(0.063)714(0.083)715(0.103)716Note: *** denotes significant at the 1% level, ** denotes significant at the 5% level, * denotes significant at the 10%717level. Standard errors in parentheses.718719720===== PAGE 13 (score 8, numeric-paren 0, table-words 0, var-words 8) =====721722Forests 2019, 10, 82072313 of 157246. Discussion and Conclusions725In our study, we conducted a hedonic analysis to investigate the impacts of the outbreak of EAB726on household property values in Milwaukee, Wisconsin. Our results show that, in general, tree cover727was positively capitalized into property values, and that prior to the arrival of EAB, ash trees added728a premium to the general value of tree cover. Although this value premium is not found for ash729trees located on individual sales parcels, it is consistently found for ash trees located at various buffer730distances in surrounding neighborhoods, suggesting that healthy ash trees (as well as tree cover of731general types of urban forest) provide a public good that is capitalized into home values. It is suggested732that the ash tree premium may be due to their widespread use as street trees within the study area.733Notably, the value premium attributed to ash trees throughout the study area (northern 10 zip codes734in Milwaukee) is dramatically diminished after the arrival of EAB infestation. This result holds true for735the entire study area, whether or not EAB-infested trees were found near sales properties, and suggests736that homeowners have forward-looking expectations regarding tree health in their neighborhoods.737These expectations are capitalized into diminishing values for ash trees. In general, static models that738only evaluate the impacts from existing damages to trees from insects and diseases will likely produce739downwardly-biased damage estimates in cases where biological invasions are expanding.740Although the capitalized value of ash trees diminished during the study period, our estimates741suggest they still contributed positive value to homeowners. This result may reflect the fact that742preventive measures, such as the use of insecticides, can protect ash tree health. Or, this result743may also reflect the fact that our study was conducted during the early stages of the EAB outbreak.744Whether or not ash trees retain a positive capitalized value during the later stages of an invasion is745unknown. However, we note that previous research has indicated that once an insect invasion reaches746a threshold value, all host trees, regardless of health condition, are capitalized into negative values by747homeowners [21]. The fact that our Milwaukee case study represents just one of many urban forested748areas impacted by EAB suggests that further analysis of this problem is warranted to help urban forest749managers understand the trade-offs between the cost and benefits of protective management activities.750While treatment ash trees are implemented on a two-year cycle in Milwaukee, there have not751yet been enough trees replanted for the lost ones. The plan is to replace removed ash trees with752diverse species of trees that are more resilient to biotic stressors and climate change. In our study, the753outbreak of EAB in Milwaukee is still in the early stage. Therefore, we did not consider the replacement754programs in the analysis. It will generally take a long time for the newly-planted trees to grow back.755The replacement programs should be able to increase property value in the long run. This would also756warrant further hedonic study to investigate the benefits from replacement management practices.757Our study result is highly flexible and useful for other non-native forest insects and pathogens. In758the evaluation studies of these long-lasting effects, more studies examining the dynamic response of759property values from invasive species are needed. Not only the tree health conditions, but also the760treatment and replacement procedures could impact property sale values, as they could potentially761change the expectation of property values based on our analysis results. Tracking the change of property762value related to invasive species during all the stages, including infestation, removal, treatment, and763restoration, would be useful to guide long-term management plans.764The EAB infestation is posing a significant threat of reduction in ecosystem services provided765by ash tree canopies. The value estimates from the hedonic model represent the portion of value766of ecosystem services provided by ash trees to the family residential properties, including largely767aesthetic views, recreational opportunities, and cultural values of trees. Comparing the estimates768from ash trees before and after the EAB infestation, we would be able to check the degradation of769these ecosystem services. However, the total economic value of benefits from numerous ecosystem770services provided by ash trees are not captured in the hedonic study. Further studies are needed to771investigate the reduction of value for other types of ecosystem services from EAB infestation, such as772carbon sequestration, reductions in stormwater runoff, or wildlife habitat.773