Balancing development and sustainability: Assessing risks and protecting aquatic biodiversity on US college and university campuses

Extensive urbanization impairs biological communities through landscape alteration and physiochemical changes to stream ecosystems. Analogous to urban development in cities, new building and facility expansion on university campuses can lead to dramatic changes in impervious cover and consequently, increased stormflow impacting downstream ecological communities. Here, we analyze the extent and relevance on a nationwide scale to determine campuses with the highest risk for ecological impairment. From the US‐wide analysis of 5761 college/university campuses, ~45% of campuses were within critical aquatic species watersheds, and >5% were identified as buffering critical habitat. The highest risk campus our study identified was selected to compare the impacts of planned, conventional development versus sustainable development on a sensitive aquatic species, the federally threatened Jollyville Plateau salamander (Eurycea tonkawae). Impervious cover, simulated increases in stormwater runoff, and total suspended solids (TSS) from the conventional expansion is estimated to lower salamander density. Conversely, an alternative campus expansion plan allowed for increased development but permitted provisions for reduced runoff and TSS, thereby maintaining current salamander densities. Our findings show that sustainable campus development plans have the potential to mitigate ecological disruption within watersheds, and that campus management and policies are critical for preserving biodiversity in the future.


| INTRODUCTION
Urbanization is an increasing global concern as the human population continues to climb.Globally, about 55% of the population resides within urban areas and is expected to increase by 2.5 billion within the next 30 years (Gao & O'Neill, 2020;Jiang & O'Neill, 2017;United Nations, 2022).In the United States, the concentration of individuals in urban areas is even higher than the global average, currently at 80.7% and projected to increase to 87.4% by the year 2050 (U.S. Census Bureau, 2019).Dense settlements of people lead to major alterations to infrastructures (buildings, impervious surfaces, etc.) and consequently, the environment, to accommodate large population of people living in proximity (Seto et al., 2010;Sun et al., 2020;Zang et al., 2011).Student population concentration in college/university areas experiences a similar phenomenon.Between 2001 and 2019, the number of Americans enrolled in a postsecondary school as a full-time or part-time student increased by over 20% (4.3 million students) (NCES, 2019).Analogous to urban development in cities, increasing enrollment requires new infrastructure to accommodate more students, specifically building and facility expansion on university and college campuses.The 5709 college and university campuses located in the contiguous US are distributed widely across the country (HIFLD, 2019; Figure 1).Like cities, these campuses typically have a centralized community area with high development and impervious cover, comparable to downtown areas of a city (Chapman et al., 2018).
Cities are required to meet certain water quality standards set by environmental policy, including the Safe Drinking Water Act (SDWA) and the Clean Water Act (CWA).In addition, the US Environmental Protection Agency (EPA) has suggested documentation for additional environmental protections, but it is up to the state and city governments to establish enforceable policy.College and university campuses are comparable, with some bearing remarkably little environmental protection policies, whereas others are leading the way in sustainable procedures.City development is driven by commercialization and economic growth, and requires multiple clearances, such as zoning clearance, electrical permits, and plumbing permits, that are awarded by various entities (Branch, 2018;Cullingworth, 2004;Giles-Corti et al., 2016;Nolen, 1916;Southworth, 1989).College and university campuses are subject to the same water quality policies and follow similar procedures for the development of new buildings and facilities (Dalton & Davis, n.d.;Illinois State University, 2011, Texas A&M University, 2017;Texas Tech University, 2014;University of Maine, 2008;University of Tennessee, 2016).Both cities and campuses are required to acquire land, obtain funding, receive building clearance and approval, and procure all service, utility, occupancy, permissions and approvals, etc. (Branch, 2018;Cullingworth, 2004;Dalton & Davis, n.d.;Giles-Corti et al., 2016;Illinois State University, 2011;Nolen, 1916;Southworth, 1989; Texas A&M University, 2017; Texas Tech University, 2014; University of Maine, 2008; University of Tennessee, 2016).
However, there are notable differences in the structural composition of campuses and cities.While campuses may have significant building infrastructure and impervious surfaces, campuses may differ from cities regarding percent of outdoor recreational areas, such as "quad" spaces or outdoor study zones, although these potential differences have yet to be directly studied in scientific literature.
Urbanization fueled alterations to the abiotic and biotic functioning of ecosystems have been widely documented across both terrestrial and aquatic environments (Grimm et al., 2008).Aquatic ecosystems are especially vulnerable to the impacts of urbanization, particularly from increased impervious cover.The expansion of impermeable surfaces in cities reroutes natural water flow paths of stream networks, resulting in decreased landscape water storage capacity and consequently, increased rapid runoff, flooding, erosion, and pollutant deposition (total suspended solids, TSS) (Fletcher et al., 2013;O'Driscoll et al., 2010;Shuster et al., 2005).A consistent decline of species richness in aquatic ecosystems is observed in areas impacted by these urbanization consequences (Paul & Meyer, 2001).This "urban stream syndrome" phenomenon necessitates improved management techniques and innovative drainage design (O'Driscoll et al., 2010).Like cities, campuses with neighboring naturalized assets and sensitive species are at greater risk of experiencing detrimental ecological consequences of urban encroachment, such as new building, facility, and parking lot development (McManamay et al., 2018).However, the extent and relevance of this issue for aquatic species conservation has yet to be studied.
Cities and campuses alike have begun adopting green infrastructure programs to manage stormwater and protect vital water resources and aquatic species (Saygın & Ulusoy, 2011).Sustainable or "green" infrastructure, such as green roofs, rain gardens, and permeable pavement, have displayed significant positive impacts on watershed hydrology (Golden & Hoghooghi, 2018;Pennino et al., 2016).Consequential improvements to watershed hydrology from green infrastructure can improve aquatic ecosystem health and biodiversity (McManamay et al., 2017).Thus, implementation of green infrastructure and sustainable practices has the potential to improve sustainability and ecological functioning in both city and college areas.
When compared to cities, college and university campuses seem to experience similar tendencies for population expansion, environmental policy, and ecosystem impacts.However, where cities have a multitude of various stakeholders and zoning issues to contend with, universities frequently have monetary control over their entire property, allowing them developmental freedom that cities lack.This provides college and university campuses with a unique opportunity to make large-scale, uniform infrastructure and policy changes.For example, University of California, Berkley, which is ranked first in environmental sustainability performance in higher education institutions by QS World University Rankings, has established a Sustainability Steering Committee that ensures sustainable goals are implemented campus-wide (QS Sustainability University Rankings 2023, n.d.; Sepasi et al., 2018).This unilateral goal creation and enforcement of policy provides a rare opportunity for environmentally beneficial change.While the implementation of sustainable policy and infrastructure on campuses is favorable, it is not yet known if the impacts of campus development are significant enough on a large-scale to warrant infrastructure changes.
Here, we analyze the extent and relevance of this topic on a nationwide scale and determine the campuses with the highest risk for ecological impairment.In addition, we evaluate the implications of campus development on ecological aquatic systems through a case study analysis.This research endeavors to determine the impact of campus development and assess whether it justifies a transition towards sustainable infrastructure and practices.

| College and university campus risk assessment
We obtained the Colleges and Universities dataset from Homeland Infrastructure Foundation-Level Database to establish the campus boundaries (HIFLD, 2019).However, this dataset included some inconsistencies in the spatial representation of campus boundaries.For instance, some campuses in the shapefile only included buildings and did not include open land or parking lots.For this research, a college or university campus is defined as the buildings, impervious surfaces, and natural/undeveloped lands owned by an individual college campus.Thus, we manually compared a subset of college and university campuses to a series of incremental-sized buffers (0.5, 1, and 2 km) to ensure accurate representation of campus land-use.We found that the 1-km buffer zone was an optimal solution, as it conservatively and realistically represented campus boundaries and important land-use factors.
Within each buffer area, the percent urban change was calculated as the percent change in urban area from 2001 to 2019 using the National Land Cover Dataset (NLCD) Land Cover Change Index (Dewitz, 2021).Similar means were used to determine the percent of natural lands per campus zone using the Natural Land classes from the 2019 NLCD (Table S1).We also determined campuses proximate to critical habitat for US federally endangered or threatened aquatic biota by intersecting campus buffer zones with the United States Fish and Wildlife Service (USFWS) Critical Habitat polygons.Lastly, we identified campuses located within watersheds containing sensitive aquatic species using spatial distributions of aquatic organisms at the 8-digit hydrologic unit code (HUC8) scale, obtained from NatureServe HUC-8.A smaller, HUC-12 watershed would have been an ideal resolution to conduct our analysis; however, spatial contagious and consistent information on the distribution of sensitive species was not available across the United States at this resolution.Aquatic species in critical endangerment of extinction were classified as G1 and G2 Aquatic Animalia species (Master et al., 2012; Table S2).All spatial analyses were conducted within ESRI ArcMap 10.8.
We calculated critical habitat risk and critical aquatic species risk using a binary system, and natural land risk and urban change risk using a quantile system (Table 1).We first assigned risk factor weights (RFWs) to campuses, based on the values of each variable of conservation concern.Campuses with "Very High" RFWs occurred within critical species watersheds, intersected critical habitat, and contained >18% natural lands."High" RFWs included campuses with 2%-18% natural land and/or greater than 6% urban change.Urban change of 3%-6% was considered a "Moderate" RSW.Campus "Low" RFWs were less than 3% urban change, less than 2% natural land, not buffering critical habitat, or outside of critical aquatic species watersheds.We then calculated Conservation Risk Scores for each campus using following equation: where P ch and P cas are binary indications (1, 0) of the presence (1) or absence (0) of critical habitats or critical aquatic species in the watershed, respectively.RFW refers to risk weights for each respective variable.
We calculated the risk score for each campus using the established 1 km buffer.We used these calculated risk scores to determine the Top 30 High Risk Campuses Index.The results from the risk study were filtered by campuses with enrollment greater than 1000.The Carnegie Classification of Institutions of Higher Education defines colleges and universities with less than 1000 students as "Very Small" (Indiana University Center for Postsecondary Research, n.d.).Campuses with under 1000 students have experienced a 34% decrease in average total student enrollment since 2010, while campuses with enrollment greater than 1000 are only seeing a 3% decrease (NCES, 2019).Due to the declining trends found in the Very Small college and university classifications, 12 campuses with this characteristic were determined as lower risk and omitted from the Top 30 High Risk Campuses Index.

| Study area and biodiversity
The highest risk campus identified by our scoring system, Concordia University Texas (CTX), was selected as a case study for sustainable development versus conventional development.CTX was founded in 1926 near downtown Austin and relocated to its current northwest Austin campus in 2007.The 389-acre property is surrounded by the Balcones Canyonlands Preserve System (BCP), a 32,000-acre urban preserve established in the early 1990s to protect critical habitat for several endangered/ threatened species.About 250-acres of Concordia's campus is part of the BCP and was delegated as protected habitat for the endangered golden-cheeked warbler (Dendroica chrysoparia), and the Jollyville Plateau salamander (Eurycea tonkawae) which was listed as a threatened species by the United States Fish and Wildlife Service (USFWS) in 2013.The Jollyville Plateau salamander (JPS) is endemic to the Central Texas area and is listed as a G1 critical species in the NatureServe HUC-8 species dataset (Bendik, 2017).The salamander's range is almost entirely encompassed within the Travis and Williamson County watersheds, where it inhabits the interstitial spaces between cover objects (e.g., rocks, leaf litter, plants) of spring-fed streams and creeks, water-filled caves, and small hillside seeps (Adcock et al., 2022).The salamander spends most of its life underground, retreating into the Edward's aquifer when streamflow declines.The majority of the JPS range is either downstream of or within highly urbanized areas of Central Texas, and many of these areas are expected to see rapidly increasing urbanization within the next 10-20 years.Stormwater runoff from nearby urban areas collects pollutants and sediments that are eventually deposited into salamander habitats.These pollutants contaminate water quality, and the sediments block the interstitial spaces, effectively destroying salamander habitat (Bendik et al., 2014).

| Current and future campus expansion
The campus is currently composed of six academic buildings, two residence halls, a gymnasium, two parking lots plus the connecting roads, one baseball field, and one softball field (Figure 2a).The existing infrastructure totals to approximately 1.84 acres of impermeable surfaces, and approximately 0.46 acres of open fields.
In 2017, CTX completed their 30-year Master Plan for the development of new infrastructure on the CTX main campus (Figure 2b).This plan featured an additional 6 residence halls, 16 parking lots and connecting roads, a nature center, 7 new academic buildings plus an extension of the existing gymnasium, sport/tennis courts, a running track/soccer field, and a PAC facility.This totals to a 250% increase in impervious surfaces, approximately 4.54 acres and approximately 1.86 acres of additional open fields.
Based on the architecture of the Master Plan, we developed an Alternate Campus Expansion Plan (ACEP) for the CTX campus as a species-conscious design to mitigate stormwater runoff and nutrient loading (Figure 2c).This plan was designed to accommodate all the new infrastructure included in the Master Plan, but with means to mitigate stormwater runoff through use of pervious surfaces and green infrastructure.Green roofs have the potential to reduce peak runoff rates by 60%-80%, however, they do tend to increase phosphorus within the first few years of implementation (Palla & Gnecco, 2018).In contrast, rain gardens have little stormwater retention capacity, but have been shown to effectively remove suspended solids and nutrients from stormwater runoff (Jeon et al., 2021).Pervious pavements have been shown to greatly reduce both peak runoff and TSS transport.Pervious surfaces can trap between 90% and 100% of TSS within stormwater and reduce peak runoff rates by 43%-55% (Brattebo & Booth, 2003;Dreelin et al., 2006;Gilbert & Clausen, 2006;Pezzaniti et al., 2009;Rowe et al., 2009;Yong et al., 2008).Thus, these three infrastructure types were chosen for the ACEP to model changes to impervious surfaces, peak runoff, and TSS.The ACEP features 18 green roofs, 9 rain gardens, and permeable pavement for all new parking lots and roads.Compared to the Master Plan, the ACEP only adds 0.12 acres of additional impermeable surfaces and 1.86 acres of additional open fields.

| Ecological implications of planned and alternative development
We chose to model the ecological impacts of sustainable development versus conventional development on the JPS as a representative critical species.Ecological impacts were modeled using previously documented relationships between watershed variables (percent impervious cover and TSS) and JPS density for several watersheds neighboring CTX (Figure 3) (Bendik et al., 2017).For each JPSinhabited watershed in Bendik et al. (2014), we estimated impervious cover, runoff, and TSS.We then related these variables to JPS density values obtained in the same study.This provided a predictive relationship for JPS density responses to changes in these variables, which provided a path to evaluate ecological impacts of the future CTX campus expansion scenarios.
We used the NLCD 2019 Impervious Descriptor dataset to characterize percent impervious cover in each watershed, according to different infrastructures and facilities present (Dewitz, 2021).This dataset includes 12 distinct impervious cover and non-impervious cover classifications.Our sites contained six of these classifications (Table S3).We also chose to incorporate the 17 distinct classifications included in the NLCD 2019 Land Cover dataset to determine the amount of each land cover type lost for the multiple CTX development scenarios (Dewitz, 2021).We determined potential changes in impervious cover for the current campus conditions, the proposed CTX 30-year Master Plan, and the student-developed ACEP.Using the NLCD 2019 Impervious Descriptor dataset and a delineated CTX watershed shapefile, percent impervious cover was calculated for the existing conditions in the CTX watershed.To model impervious cover for the CTX Master Plan and the ACEP, we digitized all proposed infrastructure into polygon shapefiles in ArcMap.Each proposed infrastructure type for the CTX Master Plan was classified into one of the NLCD Impervious Descriptor types (Table S4), then amalgamated into the NLCD Impervious Descriptor raster.Using the new raster, impervious cover percentage was calculated for the CTX Master Plan.The same techniques were applied to calculate percent impervious cover for the ACEP (Table S5).
Based on changes in landcover, particularly impervious surface, we then used the Model My Watershed (MMW) application to simulate storm runoff and TSS (mg/L) for current conditions in each watershed from  The MMW app uses an enhanced Generalized Watershed Loading Function (GWLF-E) model to simulate average annual loads from 30-years of daily fluxes (Haith et al., 1992).This product was selected due to its' compatibility with data from Bendik et al. (2014).The GWLF-E model employs land cover, soils, elevation, and stream network geographic layers and USEPA National Climate weather data as a base input file.Nutrient, sediment, and pathogen loads are then calculated using the base input file and the delineated area defined by the web app user.MMW also allows the user to define the area of different features, such as parking lots and stormwater mitigation resources.Therefore, we were able to explicitly map alternative scenarios to account for runoff generation or mitigation from different campus expansion strategies.
We used MMW to simulate runoff and mean annual concentration TSS for each of the three development Figure 3).A regression analysis was conducted to determine the best-fit relationship for the data.When analyzed, the data followed a nonlinear, negative exponential pattern, where the decline begins rapidly and slows as it approaches zero.Thus, using site location, survey area, surveys conducted, and individuals captured from Bendik (2017), we developed exponential decay equations to predict salamander density for the CTX current conditions, CTX MP, and CTX ACEP based on changes in impervious cover, runoff, and TSS.The equation: y = 23.949eÀ3.048x was produced to represent the relationship between salamander density and percent impervious cover, while the salamander density relationships with max surface runoff (cm) and TSS (mg/L) were y = 12664e À3.54x and y = 25.454eÀ0.05x , respectively.
F I G U R E 4 Distribution of college/university campuses within critical aquatic species watersheds (CASW; blue), buffering critical habitat (CH; pink), both within CASW and buffering CH (purple), and not within CASW or buffering CH (gray).Point size indicates percent of natural land for each campus.
T A B L E 2 Count of college and university campuses within critical aquatic species watersheds (CASW), buffering critical habitat (CH), both within critical species watersheds and buffering critical habitat, and not within critical species watersheds or buffering critical habitat for each campus type.3 | RESULTS

| Campus risk analysis
The results of our nationwide campus analysis identified 2558 campuses buffering critical habitat (CH) or within critical aquatic species watersheds (CASW; Our risk score analysis identified 92 campuses with a score of 200 or higher.We further identified 30 campuses with the highest development risk, all of which were within CASW and buffing CH (Table 3).Concordia University Texas was ranked as the highest-risk campus, with 13.98% urban change and 69.88% natural lands within its 1-km buffer zone.Other high-risk campuses contained between 19% and 80% natural land and 1%-30% urban change.Second ranked University of California, Santa Cruz, contained the highest percent natural lands (74.18%), while University of Texas at San Antonio experienced the highest percent urban change (29.38%).Seven out of 48 states were represented in the Top 30 high-risk campuses, 19 of which were concentrated in the state of California.North Carolina had the second largest number of high-risk campuses with three campuses in the Top 30.
Texas, Alabama, and Missouri each had two campuses within the Top 30, and Georgia and New Mexico were each represented by only one campus.

| Impervious cover, surface runoff, and TSS
Watersheds from the Bendik et al. (2014) study and the CTX watershed garnered varying results in terms of impervious cover and TSS (Figure 5a,b).Tributary 6 had the largest percent impervious cover (70.8%), the highest maximum surface runoff (2.27 cm), and the highest TSS (43.09 mg/L), while the Franklin catchment area had the smallest percent impervious cover (0.3%), lowest maximum surface runoff (1.78 cm), and lowest TSS (0.94 mg/L).
The CTX current conditions have 28.4% impervious cover, maximum surface runoff of 2.05 cm, and TSS of 21.2 mg/L.The CTX Master Plan model predicted an impervious cover percent of 34.1, maximum surface runoff of 2.16%, and TSS of 24.98 mg/L.The CTX model predicted percent impervious cover (28.4%) and TSS (21.3 mg/L) almost identical to the current conditions for CTX, and a decrease of the maximum surface runoff (1.92 cm).

| Salamander density analysis
Our analysis determined a negative exponential relationship between salamander density and impervious cover (Figure 6a), maximum surface runoff (Figure 6b), and TSS (Figure 6c).The produced exponential regression analyses were highly correlated for all three factors (Table 4).However, the standard error was significant due to the limited number of sites with access to data (Figure 7).The highly correlated exponential regression equation produced for salamander density and impervious cover predicted a salamander density of 9.9 (standard deviation (SD) = 5.8) for the CTX current conditions, 8.5 (SD = 5.5) for the CTX MP, and 9.9 (SD = 5.76) for the ACEP.The salamander density and maximum surface runoff regression equation predicted a larger alteration to salamander density with 8.5 (SD = 15.3) for CTX current conditions, 5.6 (SD = 15.1) for the CTX MP, and 14.0 (SD = 15.6) for the CTX ACEP.The relationship between runoff and salamander density displayed the highest r 2 (0.9465) and resulted in the only predicted increase in salamander density from the CTX current conditions to the CTX ACEP.The regression equation for salamander density versus TSS predicted a salamander density of 8.6 (SD = 7.3) for the CTX current conditions, 7.2 (SD = 7.1) for the CTX MP, and 8.5 (SD = 7.3) for the CTX ACEP.
All three analyses predicted CTX to see a decrease in salamander density from its current conditions to the CTX MP.Predicted decreases in salamander density were À14.5% using impervious cover, À34.1% using maximum surface runoff, and À15.4% using TSS.Alternatively, the salamander density reduction from the CTX current conditions to the CTX ACEP using impervious cover and TSS were 0% and 0.13%, respectively.In addition, the predicted salamander density using maximum surface runoff showed an 64.2% increase from the CTX current conditions to the ACEP.

| DISCUSSION
Our study identified a subset of college and university campuses as high-risk cases, but campuses at risk of jeopardizing sensitive aquatic species and critical habitat are widespread across the United States (Figure 4).Unfortunately, environmental and sustainability policy on campuses is extremely understudied, so little is known about the success rates and overall trends of campus development.Our CTX case study analysis indicates that sustainable urbanization on campuses can help preserve protected aquatic species, particularly in the case of highrisk campuses.While our case study was focused on only one species and one campus, studies have shown that TSS and impervious cover have been widely shown to induce declines in aquatic species biodiversity (Paul & Meyer, 2001).Sustainable development and green infrastructures-specifically, the forms included in our case study-have the potential to mitigate urbanization effects while still providing essential infrastructure needed for campus growth.Our analysis presents a F I G U R E 7 Predicted salamander densities for the CTX current conditions, CTX MP, and CTX ACEP using the produced exponential regression equations for percent impervious cover, maximum surface runoff, and TSS.ACEP, Alternate Campus Expansion Plan; CTX, Concordia University Texas; MP, Master Plan; TSS, total suspended solids.
baseline for college and university campuses at high risk of jeopardizing sensitive aquatic species and critical habitat.
As previously mentioned, campuses buffering CH and within CASW were primarily located in the south/southwest, specifically California.However, the largest concentration of schools overall is in the east/northeast.In part, this is because the Eastern US is more urbanized, generally, and therefore has less CH and fewer natural lands in the campus area.Campus zones that were previously developed or occur in highly urbanized areas have a lower development risk simply because development has already reached maximum capacity, or close to it, and further development will not require full land transformation and the subsequent environmental damage.However, sustainable retrofitting and development, especially green stormwater infrastructure, may be beneficial for fully urbanized campuses (Towsif Khan et al., 2020).Rain gardens, bioswales, and bioretention ponds are easily implemented into urbanized campuses, but can greatly improve stormwater retention and runoff conditions for that watershed (Ishimatsu et al., 2017;Jefferson et al., 2017).
Most of the larger, well-known college and university campuses in the United States are found in the Colleges, Universities, and Professional Schools category.These schools tend to have the largest enrollment and campus land area and are therefore more likely to see increasing development.Alternatively, Junior Colleges and Technical/Trade Schools tend to have lower enrollment and smaller campus land areas.While these smaller campuses may experience similar urbanization in their area, the land is not autonomously owned, which may hinder the ability to facilitate beneficial infrastructure changes.Thus, campuses in the Colleges, Universities, and Professional Schools category have more potential for large-scale watershed improvements from green infrastructure implementation.
While our analysis provides a baseline, it does not incorporate all the potential factors required for formally assessing risk and benefits of development on college and university campuses.Campus size was not included in our analysis due to the inconsistencies in infrastructure representation on campuses from the HIFLD dataset.However, it is an important factor that should be considered by college and university campuses when assessing individual risk.Another limitation of this study was the lack of incorporation of hydrological routing of landscape alteration risks to downstream environments.Future research could incorporate more hydrologically meaningful analyses, such as the distance from infrastructure to receiving water bodies.Campuses with developing infrastructure proximate to water bodies, or with elevation and flow paths that transport water to those water bodies more efficiently, may have additional risk that was not included in our analysis.In addition, our case study location, Concordia University Texas, was unique in that it had protected lands as part of its property.Aquatic species in protected areas tend to be more sensitive to ecosystem changes and react quicker and more adversely to development impacts (Piczak & Chow-Fraser, 2019).However, campuses near unprotected natural lands are much more likely to experience large-scale urban development, and therefore adverse impacts to the aquatic species within the watershed.Campuses buffering CH versus those within CASWs experience a similar phenomenon.CH areas are frequently federally protected due to the sensitive species living on the lands, and therefore have some reduced risk for development directly on the campus land (Tickner et al., 2020).Species in CASWs will be sensitive to urbanization too, but these areas also have less protection against direct development and therefore may experience a higher risk.
While our study only focuses on aquatic species, our approach can also be extended to nonaquatic species in an analogous fashion.Here, we utilized the NatureServe dataset and isolated sensitive aquatic species, however, this methodology could also be applied to sensitive bird species, sensitive mammal species, or other terrestrial species of concern.Our techniques can be used to calculate risk scores for campuses experiencing forest removal or habitat fragmentation in areas that house endangered or threatened terrestrial species.These analyses, as well as those for aquatic organisms, could also be amalgamated into one large analysis to consider overall species conservation risk for college and university campuses.Future research may focus on the development of a tool that analyses college and university campuses on an individual basis to provide a comprehensive risk assessment based on factors examined in this analysis.Our analysis used the CTX case study to investigate the applicability of these techniques in evaluating risks to aquatic species on college and university campuses.
Our case study investigated the differences in impervious cover, runoff, and TSS for CTX current conditions, 30-year Master Plan, and ACEP and found that the ACEP consistently had improved ecological conditions for all three factors.The ACEP utilized only three types of green infrastructure-rain gardens, green roofs, and pervious surfaces-yet showed significant benefits to stormwater management on the CTX campus.Other green infrastructure practices, such as rainwater harvesting, bioswales, and urban tree canopies have shown significant potential in improving aquatic conditions and may improve ecological conditions in campus watersheds even further (Pataki et al., 2021;Xiao et al., 2017;Xu et al., 2023).It is important to note that the impervious cover, TSS, and runoff for the CTX scenarios were reported on a watershed scale (Figure 5).CTX is a small campus, with only a few buildings and a relatively small percentage of impervious surfaces when compared to other major colleges and universities.Including green infrastructure in development plans for quickly expanding campuses, or retrofitting large campus buildings, facilities, and impervious surfaces with green infrastructure, could have an even larger impact on TSS and runoff within a watershed.Our study specifically highlights the importance of sustainable expansion planning for small to medium-sized campuses that have the potential for considerable future growth, and thus, have the greatest potential for proactive management to prevent biodiversity loss.More studies are needed to investigate green infrastructure impacts on a larger, watershed scale.
The CTX case study's exploration of salamander density responses to varying factors adds a layer of complexity to the broader discussion.As seen in our study, salamander density was found to decline strongly with increased impervious cover, runoff, and TSS for all the assessed watersheds.The predicted salamander density calculated from the CTX scenarios demonstrated the same trend, suggesting a steep decline should CTX choose to execute their Master Plan as planned.Alternatively, protection and sustainability of the salamander population is possible through use of the ACEP.In fact, the ACEP alterations to maximum surface runoff showed an increase in salamander density, suggesting that green infrastructure may be able to not only serve as compensatory mitigation for new development, but also improve existing aquatic conditions in urbanized campuses.Standard development is also likely to contribute to declining trends for other species' diversity and density.Research expansion is necessary to determine if sustainable campus development could mitigate watershed-scale urbanization impacts for other sensitive aquatic species.Due to their habitat preferences, salamanders are impacted by flow conditions and sediment loading.However, other sensitive aquatic species may be more heavily impacted or not at all impacted by specific factors.Fish responses to increased flood magnitude and frequency are variable, however, increased TSS may result in decreased visual acuity and therefore decreased feeding, reproduction, and overall density (Kjelland et al., 2015;McManamay et al., 2013).Thus, sustainable infrastructure on campuses may be able to mitigate some urbanization impacts, but success may vary depending on the species and habitat.Additionally, our study investigates campus impact on aquatic systems under the current climate conditions.Changes to climate as represented by the four main Representative Concentrations Pathways (RCP) include alterations to precipitation duration, timing, and quantity, as well as temperature shifts (Shukla, 2019).Since impact to aquatic systems is highly dependent on precipitation conditions, any development effects-sustainable or otherwise-would vary based on the climatic environment.This presents an interesting concept for future research on campus development impact.
College and university campuses are distinct in that they have the autonomy to make unilateral land-use decisions, leading to dramatic changes across a large contiguous area in a relatively condensed period, which ultimately may threaten biodiversity.Likewise, as primary owners and managers of the land, buildings, and facilities, colleges and universities also have a unique opportunity to change policy and infrastructure plans to more sustainable alternatives, with far less regulatory obstacles than municipalities.The results of our case study emphasize that even small universities can have impacts on aquatic species in their larger watershed, indicating the potential ecological significance of this autonomous decision-making.However, benefits from sustainable infrastructure extend past ecological preservation and into human welfare.Sustainable infrastructure provides jobs, recreational and study areas, and improves overall health conditions (EPA, 2018).On college campuses, green infrastructure implementation provides unique opportunities for student and faculty engagement, such as research and experimentation to advance knowledge and practice.While large-scale sustainable infrastructure changes require college and university administration approval, even small changes implemented by students and staff have the potential to promote sustainability on a larger scale.This ability for an individual, for example, students and faculty, to be involved in sustainable infrastructure changes is yet another rare opportunity available for college and university campuses.In addition, campus involvement in sustainable development plans would provide an opportunity for small scale "green" urbanization research.This research has the potential to be applied to larger scale urbanization planning, such as colleges and universities as mesocosms for sustainable city growth.

| CONCLUSION
Infrastructure development undertaken on college and university campuses across the United States poses a significant risk of aquatic ecosystem degradation.Impacts of campus expansion are likely analogous to urbanization-increases in impervious cover, peak runoff, and physiochemical responses in receiving waterbodies.Negative impacts from these factors can affect entire watersheds, not just their immediate area.These impacts can be mitigated through research and implementation of sustainable development plans, especially for campuses at high risk.College and university campuses are unique in their land ownership of multiple buildings, facilities, and impervious surfaces in condensed area, allowing them increased freedoms to make uniform changes for aquatic species conservation.Results from future sustainable campus development may provide critical sustainable expansion information to be applied to large scale urbanization.
T A B L E 1 Risk Factor Weights (RFWs) for use in calculating Conservation Risk Scores for college and university campuses.Buffers critical habitat and within critical aquatic species are binary indicators (B).Natural land and urban change divisions were determined by quantiles (Q) represented in their respective columns.

F
I G U R E 2 (a) Concordia University Texas, CTX, original conditions for the CTX preserve (light green textured), CTX boundary (black dashed line), CTX watershed (gray dashed line), Tornado Trail (red hatched), open fields (light green), buildings (yellow), and pavement.(b) CTX 2017 Master Plan additional infrastructure, including proposed buildings (orange), open fields (dark green), and impervious pavement (dark gray).(c) CTX Alternate Campus Expansion Plan infrastructure for green roofs (green hatch), rain gardens (blue), and permeable pavement (gray hatch).
Bendik et al. (2014), and for the three CTX scenarios in the CTX watershed(Stroud Water Research Center, 2021).
scenarios for CTX.Landcover changes were input into the scenario template provided by MMW.Using the developed proposed infrastructure shapefiles and the NLCD 2019 Land Cover dataset, we tabulated existing NLCD land cover for each of the new infrastructure polygons.The tabulated polygon NLCD data was then entered into a Land Cover Change Analysis tool we created in Microsoft Excel to determine the hectares (ha) of each NLCD land cover type lost and gained.Results from this tool were entered into MMW along with the CTX watershed shapefile to model TSS for each of the proposed CTX development scenarios.Finally, we sought to determine the influence of the simulated physiochemical variables on salamander density, as well as the impact of potential development scenarios on the CTX watershed.To do so, we developed a relationship between percent impervious cover, runoff, and TSS on JPS density using the six salamanderinhabited catchments defined by Bendik et al. (2014, F I G U R E 3 Six catchment areas defined by Bendik et al. (2014) and the CTX delineated watershed.Travis County boundary (purple line) and Concordia University Texas location (red star).CTX, Concordia University Texas.

F
I G U R E 5 CTX scenario breakdown of (a) percent impervious cover, (b) maximum surface runoff in centimeters, and (c) TSS in milligrams per liter.CTX, Concordia University Texas; TSS, total suspended solids.

Table 2
Top 30 High Risk Campuses Index.Development risk for college and university campuses in the contiguous US determined by scoring system and ranked by top 30.
).These campuses were distributed across the entire US, with the highest concentrations occurring in the Southwest and Southeast regions of the United States (Figure4).About 44% of US college and university campuses are located within CASW, while only about 4% of T A B L E 3 Note: