Graduation Year


Document Type




Degree Name

Doctor of Philosophy (Ph.D.)

Degree Granting Department

Civil and Environmental Engineering

Major Professor

Maya A. Trotz, Ph.D.

Co-Major Professor

Thomas L. Crisman, Ph.D.

Committee Member

Sarina J. Ergas, Ph.D.

Committee Member

James R. Mihelcic, Ph.D.

Committee Member

Delcie R. Durham, Ph.D.

Committee Member

Allan Feldman, Ph.D.


Community Engagement, Grand Engineering Challenges, K-12 STEM, Plant Selection Index, Rain Gardens, Stormwater Management


The Green Space Based Learning (GSBL) approach builds on a long-term partnership between a Research I university, surrounding community, and local school district, transforming underutilized community green space into an interactive educational tool to addresses national infrastructure and educational challenges. The GSBL approach is an educational platform for engaging K-12 and the local community in engineering design and construction of sustainable Green Infrastructure (GI) projects. GSBL was piloted as a part of a federally funded Research Experience for Teachers (RET) program in which teachers participated in two intensive 6-week summer research experiences and two consecutive academic year components. The summer experience focuses on the development of Science Technology Engineering and Mathematic (STEM) lessons and activities that meet Common Core and Next Generation Science Standards and the dissemination of the RET research experience. Approximately 400 K-12 students and teachers participated in both formal and informal educational activities that led to GSBL approach outputs throughout the academic year. These outputs included 4 Campus GI Challenge's for identifying areas of implementation and student driven GI design, the publication of 7 curricular products, the design and installation of 70 personal rain gardens and 8 bioretention cells (a type of GI), one of which was designed as a field scale research site within the Hillsborough County Public Schools (HCPS) district.

The eight bioretention cells, seven of which are on three public school campuses and one located at a local community leader's house were designed and implemented as a result of university research, K-12 outreach, and community engagement. These sites were selected based on one or more hotspot factors (e.g. localized areas of flooding, access to site, presence of learning space, willingness to pay, property ownership, visibility of location) and designed to restore the hydrology and water quality to pre-development conditions. The bioretention cells were designed to capture a storm-event ranging from 1.27 cm to 2.54 cm and cost between $550 and $1,650 to construct depending on the design scope, scale, and installation methods. The installed bioretention systems route stormwater runoff to a ponding area sized approximately 2-5% of the total catchment area, are designed to capture between 31% and 67% of annual runoff (March 2010 - March 2015), and attenuate between 97,500 and 226,100 mg N annually.

The educational sites were used to provide insight into hydraulic performance, maintenance requirements, and nutrient management impacts associated with bioretention design. Three of the bioretention cells (BR 1, BR 2, and BR 3) were used as a field research site for collecting bioretention plant performance data on 12 Florida native plant species, Coreopsis leavenworthii, Flaveria linearis, Salvia coccinea, Solidago fistulosa, Canna flaccida, Tradescantia ohiensis, Tripsacum dactyloides, Hymenocallis latifolia, Iris virginica, Sisyrinchium angustifolium, Spartina patens, and Equisetum hyemale.

Mean baseline accumulated nitrogen concentration for tested species was 18.24 ± 5.76 mg N/g biomass. This compared to a harvested mean concentration rate of 12.28 ± 2.23 mg N/g biomass, a reduction of uptake capacity of nearly 33% after two growing seasons. This study found a similarity in mean total nitrogen concentration between baseline and harvested plant species for Flaveria linearis, Sisyrinchium angustifolium, Solidago fistulosa, Canna flaccida, Salvia coccinea, Spartina patens, and Coreopsis leavenworthii and a significant difference in means for Equisetum hyemale, Iris virginica, Salvia coccinea, and Tradescantia ohiensis. These harvested data were used to calculate mean total nitrogen concentration per square meter with Sisyrinchium angustifolium, Equisetum hyemale, Spartina patens, Solidago fistulosa, Salvia coccinea, Coreopsis leavenworthii, Iris virginica ranging from 286 mg N/m2 to 4,539 mg N/m2, and Canna flaccida, Flaveria linearis, Tradescantia ohiensis ranging from 12,428 mg N/m2 to 15,409 mg N/m2. Seven of the twelve species (Flaveria linearis, Equisetum hyemale, Iris virginica, Tripsacum dactyloides, Coreopsis leavenworthii, Salvia coccinea, Tradescantia ohiensis) displayed highly desirable results, ranking (>0.20x̅) when evaluated across 10 quantitative attributes and assessed for their applicability for the subtropical Tampa Bay area.

This research developed a plant selection utility index (PSI) that allows for individual plant scoring based on qualitative and quantitative plant selection criteria. The qualitative PSI was used to evaluate 26 native and regionally friendly plant species commonly found within the subtropical Tampa Bay climate to provide an example and act as a template for selecting plant species. The qualitative PSI scores categorized the identified plant species as highly desirable (n=4, PSI ≥ 80), Flaveria linearis, Tripsacum dactyloides, Salvia coccinea, and Chamaecrista fasciculata; moderately desirable (n=15, 80 > PSI ≥65), Solidago fistulosa, Hymenocallis latifolia, Canna flaccida, Tradescantia ohiensis, Arachis glabrata, Mimosa strigillosa, Callicarpa Americana, Penta lanceolata, Monarda punctate, Muhlenbergia capillaris, Helianthus debilis, Glandularia tampensis, Silphium asteriscus, Stachytarpheta jamaicensis, and Coreopsis lanceolata; and least desirable (n=7, PSI < 65) Spartina patens, Equisetum hyemale, Sisyrinchium angustifolium, Iris virginica, Coreopsis leavenworthii, Myrcianthus fragrans, Zamia puila. The quantitative PSI was used to evaluate attributes of 11 of the 26 species within a 32.5 m2 field-scale bioretention system (BR 1, BR 2, and BR 3) ter two-growing seasons. The tested species scored as highly desirable (n=2, PSI ≥ 70) for Salvia coccinea, Tradescantia ohiensis; moderately desirable (n=5, 70 > PSI ≥ 50) for Equisetum hyemale, Sisyrinchium angustifolium, Solidago fistulosa, Iris virginica, Coreopsis leavenworthii, and least desirable (n=4, PSI < 50) for Spartina patens, Flaveria linearis, Canna flaccida, Hymenocallis latifolia. Both qualitative and quantitative scores were combined on a 0-200 scale to provide a list of recommended species based, ranking from high to low: Salvia coccinea (PSI=160), Tradescantia ohiensis (PSI = 148), Sisyrinchium angustifolium (PSI =127), Flaveria linearis (PSI = 125), Solidago fistulosa (PSI = 124), Iris virginica (PSI =121), Coreopsis leavenworthii (PSI = 117), Equisetum hyemale (PSI = 114), Canna flaccida (PSI = 104), Spartina patens (PSI = 103), Hymenocallis latifolia (PSI =90).