Floral representation of 10 native (N) and 10 non-native ornamental plants selected for this study. Photos courtesy of Joanna J. Silva and Walker Bensch.
Overall effects of provenance (native and non-native) on the (A) plant growth index and (B) total floral abundance in year 1 and year 2 of the study. Different letters indicate significant differences at P < 0.05.
Click on author name to view affiliation information
Pollinator gardening has gained momentum with an increased consumer interest in selecting native rather than non-native plant species to reduce water dependence and maximize the biodiversity value in both public greenspaces and residential gardens. Native plant species can enhance biological control and benefit ecosystems and wildlife. Often, they are also better-adapted to local environmental conditions, including temperature and rainfall, thus increasing their survival and reducing associated maintenance costs, primarily by requiring less water. Commercially available pollinator-friendly plant mixes often include both native and non-native species. A 2-year study was conducted to determine the main effects of plant provenance (native or non-native) and irrigation (full or partial irrigation) on landscape performance and flowering of 20 plants, including 10 congeneric pairs of native and non-native taxa that were planted in two locations (north and northcentral Florida). Native and non-native taxa were paired by genus to analyze the effect of the plant native status on vegetative and floral traits while controlling for variations in leaf and floral morphologies, growth habits, and blooming periods, which was a key and novel component of our study design. Represented native species included Spanish needles (Bidens alba), false rosemary (Conradina grandiflora), tickseed coreopsis (Coreopsis leavenworthii), blanket-flower (Gaillardia pulchella), swamp rosemallow (Hibiscus grandiflorus), inkberry (Ilex glabra), spotted beebalm (Monarda punctata), azure blue sage (Salvia azurea), Florida scrub skullcap (Scutellaria arenicola), and Walter's viburnum (Viburnum obovatum). Non-native taxa paired with native congeners included Beedance® painted red bidens (Bidens ferulifolia) or Goldilocks Rocks® bidens (Bidens ferulifolia ‘BID 16101’), barbeque rosemary (Salvia rosmarinus ‘Barbeque’), Jethro Tull coreopsis (Coreopsis × ‘Jethro Tull’), Arizona sun blanket-flower (Gaillardia ×grandiflora ‘Arizona Sun’), Ruffled Satin® rose of Sharon (Hibiscus syriacus ‘SHIMCR1’), dwarf Burford holly (Ilex cornuta ‘Dwarf Burford’), pardon my pink beebalm (Monarda didyma ‘Pardon My Pink’), big blue salvia (Salvia longispicata × S. farinacea ‘PAS1246577’), Malaysian skullcap (Scutellaria javanica), and Sandankwa viburnum (Viburnum suspensum). Overall, the results revealed that native plants outperformed non-native plants and exhibited greater survival, more vegetative growth, and greater floral abundance regardless of the irrigation treatment. Although there was no overall effect of irrigation on plant size or flower abundance, there were some species-specific responses, especially during the establishment year, and plants under full irrigation had greater survival in the establishment year. Thus, in general, the effects of plant provenance were stronger and more consistent across years than irrigation. Additional studies are underway to determine the floral rewards of these species and their attraction to diverse pollinators.
During the past three decades, ecologically friendly landscaping has strived to implement plants that not only require less water and other inputs but also create esthetic value and biodiversity in gardens (Brzuszek et al. 2010; Potts et al. 2002; Zadegan et al. 2008). Among these efforts, pollinator-friendly gardening has become a popular strategy of addressing the discernable global issue of pollinator decline and typically involves planting a diversity of flowers in cultivated environments (Garbuzov and Ratnieks 2014; Majewska and Altizer 2020; McFrederick and LeBuhn 2006; Salisbury et al. 2015). Research has shown that urban habitats with pollinator gardens can support abundant and diverse bee communities that exceed those found in rural and even semi-natural habitats (Baldock et al. 2015; Fetridge et al. 2008; Theodorou et al. 2020). Furthermore, surveys have found that many people, including diverse groups of scientists, gardeners, and the general public, are willing to participate in pollinator conservation efforts (Kalaman et al. 2020; Wiggins et al. 2018). However, guidelines to optimize the ecosystem services of pollinator gardens while still meeting other goals associated with feasibility, sustainability, cost, and aesthetic appeal, remain limited.
Many commercially available pollinator-friendly plant mixes often include both native and non-native plants (Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services 2016; Seitz et al. 2020), as do those from retail nurseries and other commercial plant vendors (Zinnen and Matthews 2022). Native plants are particularly known for their resiliency in gardens because they are locally adapted to the climate, soil conditions, and natural pests of the region from which they originated (Matrazzo and Bissett 2020). In recent years, efforts have been directed toward optimizing propagation and production practices of native species that have ornamental value and ecological value (Wilson and Deng 2023). Similarly, many state programs such as Florida-Friendly Landscaping™ (FFL) have been launched to promote the use of both native and non-native plants that are low-maintenance and reduce water, fertilizer, and pesticide consumption (Florida-Friendly Landscaping Program 2024).
Along with native plants, many cultivated non-native species have attractive traits for home and community landscapes that could also be beneficial for pollinators, such as different forms, larger flowers, and prolonged flowering periods (Kalaman et al. 2022b). In general, studies of plant performance and pollinator preference associated with native and non-native plant species have shown mixed results. For example, Salisbury et al. (2015) reported increased pollinator preferences for native rather than non-native plants; however, Martins et al. (2017) observed no such effect. Similarly, Kalaman et al. (2022a) found that pollinator visitation rates vary across plant species and locations, but not necessarily by native status. Tartaglia and Aronson (2024) performed a systematic literature review of studies related to the use of non-native plants rather than native plants in urban horticulture and found evidence that native plants outperformed non-natives in the provisioning of biodiversity and human health.
Part of the variation across studies may have been attributable to the focal pollinators studied; for example, non-native plants may be primarily attractive to generalist and non-native bee species, whereas native, specialist bee pollinators would likely require, and be attracted to, native plant resources (Fetridge et al. 2008; Frankie et al. 2019; Matteson et al. 2008). The informed selection of ecologically friendly plants also plays a major role in creating attractive landscapes that not only provide adequate floral resources for wildlife (like pollinators) but also meet water conservation goals, including reducing irrigation, among other conservation aims (Matrazzo and Bissett 2020; Ogden and Ogden 2011).
Periodic droughts throughout much of the US have led to increased restrictions of water use in domestic and public green areas (Ruiz et al. 2020). In Florida, despite a high average rainfall, irrigation is often used because of the nonuniform distribution of rainfall throughout the year, leading to seasonal droughts, as well as the limited water-holding capacities of Florida’s sandy soils. Thus, irrigation throughout the year is common in Florida garden and landscape settings, and reducing water use is one of the top priorities for the garden and landscape industry (Boyer et al. 2014). Because of the sensitivity of many plants to water stress, landscapers and consumers are selecting plants that can survive with minimal watering during the periodic droughts in Florida, including ornamental native plants (Ferrarezi et al. 2020; Knox 1990).
Water availability can affect pollinator attraction and interact with plant provenance, thus affecting the resource value of the plant. For example, previous studies showed that environmental conditions like soil moisture, temperature, and water availability were the primary factors that affect the vegetative and floral traits and attractiveness of pollinator-friendly plants rather than the native status of plants (Burkle and Runyon 2016; Carroll et al. 2001; Caruso 2006; Descamps et al. 2021; Kuppler et al. 2021; Plos et al. 2023; Scheiber et al. 2008; Shober et al. 2010). Additionally, changes in environmental conditions, like irrigation, can influence vegetative and reproductive traits in different and complex ways. For example, Wu et al. (2023) evaluated how soil water and nutrient availability interactively influenced pollinator-mediated selection of Tibetan primrose (Primula tibetica) and found that such abiotic factors increased plant height and reproductive capacity (fruit production and seed production) of this species. These results highlight the need to examine how native and non-native plant species respond differently to water availability, thus affecting both vegetative and floral traits.
The overall goal of this study was to ascertain vegetative trait responses of a broad range of pollinator plant species that differed in native status and irrigation treatment and were transplanted in two different locations. The specific objectives were as follows: to determine the effects of the native and non-native status on plant growth, plant height, plant width, and flowering; to determine the effects of full and reduced irrigation on these same plant traits across native and non-native plants; and to determine the interaction between irrigation and native status on vegetative responses and floral density. To address these objectives, we selected 10 native and 10 non-native ornamental plants paired by genus and grew them under two different irrigation treatments in two locations in north and northcentral Florida. W also present the species-specific results of 20 taxa and the effects of irrigation and location on plant performance. Additional studies are underway by the research team to determine specific floral rewards of the plant species included herein and their attraction to diverse pollinators.
Twenty ornamental plant species were selected for use in this study based on the following criteria: commercially available in nurseries and appropriate for ornamental use in landscapes; predicted to be of value to pollinators (particularly bees) and had diverse bloom times and flower colors; and capable of surviving a 2-year transplanting trial in Florida. The selection of congeneric pairs comprised labeled pollinator-friendly plants native to Florida (native plants) and plants non-native to Florida, including cultivars derived from plants not native to Florida (non-native plants) (Table 1). Native and non-native plants were paired by genus to analyze the effect of plant provenance (i.e., native status) on plant traits and floral abundance while controlling for large variations in leaf and floral morphologies, growth habits, and blooming periods, which was a key and novel component of our study design (Fig. 1). All native plants were sourced from three local nurseries, and all non-native congeners were sourced from three large, national producers for all non-native congeners; all plants within a species were sourced from the same place. All plants were sourced in 1-gallon pots, except for Monarda dydima, which was finished in 1-gallon pots by the research team before planting.
Citation: HortTechnology 34, 5; 10.21273/HORTTECH05470-24
Native species included Spanish needles [Bidens alba (L.) DC], false rosemary (Conradina grandiflora Small), tickseed coreopsis (Coreopsis laevenworthii L.), blanket flower (Gaillardia pulchella Foug.), swamp rosemallow (Hibiscus grandiflorus Michaux), inkberry (Ilex glabra L.), spotted beebalm (Monarda punctata L.), azure blue sage (Salvia azurea Michx. Ex Lam.), Florida scrub skullcap (Scutellaria arenicola L.), and Walter’s viburnum (Viburnum obovatum Walter). Non-native species paired with native congeners included Beedance® painted red bidens (year 1 of the study) (Bidens ferulifolia) or Goldilocks Rocks® bidens (year 2 of the study) (Bidens ferulifolia ‘BID 16101’), barbeque rosemary (Salvia rosmarinus ‘Barbeque’), Jethro Tull coreopsis (Coreopsis × ‘Jethro Tull’), Arizona sun blanket flower (Gaillardia ×grandiflora ‘Arizona Sun’), Ruffled Satin® rose of Sharon (Hibiscus syriacus ‘SHIMCR1’), dwarf Burford holly (Ilex cornuta ‘Dwarf Burford’ Lindl.), pardon my pink beebalm (Monarda didyma ‘Pardon My Pink’), big blue salvia (Salvia longispicata × S. farinacea ‘PAS1246577’), Malaysian skullcap (Scutellaria javanica Jungh), and Sandankwa viburnum (Viburnum suspensum). The non-native Bidens ferulifolia cultivar was switched in year 2 because of a lack of survival through the winter in year 1, followed by cultivar unavailability in year 2 for replanting. Salvia rosmarinus (barbeque rosemary) and Conradina grandiflora (false rosemary) are not congeneric, but they have similar morphological traits and are within the same plant family.
Field plots were prepared similarly in two locations. The first site was located at the University of Florida North Florida Research and Education Center (NFREC) in north Florida (Quincy, FL, USA; USDA cold hardiness zone 8b). The second site was located at the University of Florida Plant Science Research and Education Unit (PSREU) in northcentral Florida (Citra, FL, USA; USDA cold hardiness zone 9a) (US Department of Agriculture, Agricultural Research Service 2023). Two months before early spring installation, eight slightly raised beds at each location were disced and treated with a commercial herbicide before being covered with a commercial-grade black polyethylene woven geotextile fabric (Mutual Industries, Inc., Philadelphia, PA, USA). Each of the eight beds contained 20 plots; each plot had a length of 3 m and width of 0.9 m, with 0.9 m of spacing between each row and 1.20 m of spacing between each plot according to Kalaman et al. (2022a). One plant species was randomly assigned to each plot within each bed for a total of 20 plant species per bed (10 native and 10 non-native species). Two (for woody) or three (for herbaceous) plants of each species were assigned to each plot following the recommended spacing for woody and herbaceous plants. Each of the eight rows was assigned to either full irrigation or partial irrigation, resulting in a sampling size of four plots per plant species per irrigation treatment per site.
Plants were installed at PSREU on 26 Feb 2021, and at NFREC on 1 Mar 2021. Following installation, plants were initially drip-irrigated for 2 h once per day. Once established (after 4 weeks), four of the eight beds were drip-irrigated for 2 h per day (full irrigation), whereas the other four beds were irrigated at 10% volumetric soil moisture using a SMRT-Y- Soil Moisture Sensor Kit (Rainbird Inc., Tucson, AZ, USA) (reduced/partial irrigation). Plants were top-dressed with 89.1 g (woody plants) or 53.5 g (herbaceous plants) of 20N–4P–7K slow release fertilizer (Osmocote Pro, 8-9 months; ICL, Tel Aviv, Israel) upon planting and between years. Finally, grass areas between block rows were mowed weekly.
Multiple soil samples were collected and mixed at each site and then air-dried for a standard elemental analysis using a Mehlich-3 extraction method (University of Florida Extension Soil Testing Laboratory, Gainesville, FL, USA) as reported in Supplemental Table 1. Macronutrients and micronutrients at both locations were within normal limits with soil pH values of 6.7 and 7.0 in north and northcentral Florida, respectively (University of Florida Extension Soil Testing Laboratory, Gainesville, FL, USA). Phosphorous (P), potassium (K), magnesium (Mg), and calcium (Ca) had high nutrient ranges (Supplemental Table 1). As reported by the Florida Automated Weather Network (2024), field conditions in north Florida were as follows: average monthly rainfall, 0.42 cm; mean minimum and maximum temperatures, 13.6 °C and 26.3 °C, respectively; and relative humidity, 78.7% (Supplemental Fig. 1). In northcentral Florida, the field conditions were as follows: average monthly rainfall, 0.35 cm; mean minimum and maximum temperatures, 15.6 °C and 28.3 °C, respectively; and relative humidity, 81.5%. Of note, plants grown in north Florida received major rain events during February and March of year 2. Similarly, plants grown in northcentral Florida received major rain events during June of year 1, and during February through March of year 2 (Supplemental Fig. 1).
Each month during the first and second year of the study, for each plant within each plot, we measured the plant height (cm), defined as the vertical distance from the soil surface to the apex of the plant’s canopy, the plant canopy width at its widest point (width 1), and the width perpendicular to width 1 (width 2) (cm). We reported the plant width as the greater of these two measurements (width 1 and width 2), representing the maximum extent of the plant’s lateral spread. In addition to these measurements, a growth index was calculated as a composite metric of plant growth. The growth index was determined by summing the plant height and the average of the two perpendicular widths, width 1 and width 2, and then dividing the sum by two. We evaluated plant height, width, and growth index according to Wilson et al. (2020) and Wilson and Knox (2006). We also recorded plant survival at the end of years 1 and 2 after assessing the number of living plants in each plot after the establishment period and before plants went dormant. The survival percentages were calculated based on the original number of plants planted in each plot.
Additionally, every month for 2 years, a floral survey was conducted; the total flowers were counted for each plot at both sites. Capitulate inflorescences (Bidens, Coreopsis, and Gaillardia) were notated as a single flower (Fig. 1).
To evaluate the main effects of provenance (i.e., native status) and irrigation on vegetative traits (growth index, plant height, plant width), we used the largest values per plot in year 1 and year 2; the years were analyzed separately because of varying weather conditions. After assessing the normal distribution of these vegetative traits, we used a linear mixed model with plant provenance (native vs. non-native) and its interaction with irrigation treatment (full vs. partial) as fixed effects and site as a random effect. To assess the main effects of provenance and irrigation on flower counts, we used the total number of flowers in year 1 and year 2, which were analyzed separately, and used a generalized linear mixed effects model with plant provenance and its interaction with irrigation as fixed effects and site as a random effect. For the flower counts model, we accounted for overdispersion with a negative binomial distribution.
To evaluate the number of surviving plants at the end of year 1 and year 2, we fit a generalized linear mixed model with a negative binomial distribution for percentage survival as the response variable, plant provenance and its interaction with irrigation as fixed effects, and site as a random effect.
To evaluate differences between native and non-native species within a genus and the consistency of these responses across planting locations, we generated linear models with effects of provenance and site and their interaction on the growth index separately for years 1 and 2 of the study and separately for each of the 10 genera. Additionally, we performed generalized linear models to evaluate the response of floral abundance for each of the 10 genera separately using provenance, site, and their interaction as fixed effects, and we included a negative binomial distribution to account for overdispersion.
To evaluate species-specific responses to irrigation and the consistency of these responses across planting locations, we generated linear models with effects of irrigation and site and their interaction on the maximum growth index separately for years 1 and 2 of the study and separately for each of the 20 species. Additionally, we performed generalized linear models to evaluate the response of floral abundance for each of the 20 species separately using irrigation, site, and their interaction as fixed effects, and we included a negative binomial distribution to account for overdispersion.
To determine the significance of fixed effects, all aforementioned models were subjected to an analysis of variance (ANOVA); effects were described as strong (P < 0.01), moderate (0.01 < P < 0.05), weak (0.05 < P < 0.10), or no evidence for an effect (P > 0.10) according to Muff et al. (2022). All analyses used RStudio statistical software (version 2023.06.2 + 561, Mountain Hydrangea; Boston, MA, USA) including the following packages: ‘car’ for ANOVAs (Fox and Weisberg 2019); ‘emmeans’ for estimated marginal means and SEs from the aforementioned models (Lenth 2020); ‘lme4’ for mixed effects models (Bates et al. 2015); ‘MASS’ for GLMs (Venables and Ripley 2002); and ‘ggplot2’ for visualization (Wickham 2016).
There was strong evidence that the provenance of plants (native status to Florida), but not the irrigation treatment, affected measured vegetative traits of plants evaluated in this study (Fig. 2A). Specifically, in year 1 of the study, native plants had a larger growth index than non-native plants (χ2 = 247.05; P < 0.0001). However, irrigation had no effect on growth index in year 1 (χ2 = 1.89; P = 0.17), and there was no interaction between irrigation and provenance in year 1 (χ2 = 0.04; P = 0.85). In year 2, native plants again had a larger growth index (χ2 = 26.46; P < 0.0001), but there was no effect of irrigation (χ2 = 0.21; P = 0.65) or interaction between irrigation and provenance on the growth index (χ2 = 0.01; P = 0.91) (Fig. 2A).
Citation: HortTechnology 34, 5; 10.21273/HORTTECH05470-24
Native plants also had greater plant height than non-native plants in year 1 (χ2 =133.27; P < 0.0001). However, there was no effect of irrigation treatment (χ2 = 0.37; P = 0.54) or interaction between irrigation and provenance on plant height in year 1 (χ2 = 0.04; P = 0.84). In year 2, plant height showed results similar to those of the previous year, with native plants being significantly taller (χ2 = 17.30; P < 0.0001), and with irrigation (χ2 = 0.14; P = 0.71) and the interaction between irrigation and provenance (χ2 = 0.003; P = 0.96) continuing to show no significant effect on plant height.
Finally, in year 1, plant width was greater for native than for non-native plants (χ2 = 214.79; P < 0.0001). There was no effect of irrigation on plant width in year 1 (χ2 = 2.06; P = 0.15) or an interaction between irrigation and provenance (χ2 = 0.02; P = 0.89). Native plants were similarly wider than non-native plants in year 2 (χ2 = 29.30; P < 0.0001), and there was no effect of irrigation (χ2 = 0.18; P = 0.66) or interaction between irrigation and provenance (χ2 = 0.01; P = 0.92).
In year 1, there was strong evidence that the provenance of plants (native status) and the irrigation treatment affected the survival of plants. Specifically, native plants had a higher survival rate than that of non-native plants (χ2 = 27.21; P < 0.0001). Fully irrigated plants also had a moderately higher survival rate than that of non-irrigated plants in year 1 (χ2 = 7.40; P = 0.01), but there was no interaction between irrigation and provenance (χ2 = 0.01; P = 0.92) (Supplemental Fig. 2A). In year 2, native plants had a higher survival rate (χ2 = 30.18; P < 0.0001), but there was no evidence of an effect of irrigation (χ2 = 2.65; P = 0.10), and only a weak interaction between irrigation and provenance on survival rate (χ2 = 3.04; P = 0.08) (Supplemental Fig. 2B).
Provenance (i.e., native status) and planting location differently affected the growth index of plants across genera (Supplemental Table 2). Overall, during the first year, all native plants grew larger than their non-native counterparts, except for the rosemary (Salvia/Conradina spp.) and sage (Salvia spp.) pairs of plants, which did not show differences (Supplemental Table 2). We observed the same general pattern in year 2, with the exceptions that the non-native sage and skullcap (Scutellaria) grew larger than their native congeners (Supplemental Table 2). In year 1, three genera showed an interaction between provenance and planting location, varying from weak to strong, whereas in year 2, six genera showed an interaction between provenance and planting location, varying from weak to strong (Supplemental Table 2). However, in general, the direction of the effect of provenance was consistent across locations but varied in magnitude (Supplemental Table 2).
Irrigation and planting location (north or northcentral Florida) differently affected the growth index of plants across species (Table 2). During the first year in north Florida, five species (Spanish needles, Jethro Tull coreopsis, Arizona sun blanket-flower, pardon my pink beebalm, and Sandankwa viburnum) grew larger under full irrigation than under partial irrigation. Similarly, in northcentral Florida, four of these same species, Spanish needles, Jethro Tull coreopsis, Arizona sun blanket-flower, and pardon my pink beebalm, also grew larger under full irrigation than under partial irrigation in year 1 (Table 2). In year 2, only two species showed a weak effect of irrigation and were larger under the partial irrigation treatment as compared with full irrigation treatment (Ruffled Satin® rose of Sharon and spotted beebalm) (Table 2). Across both years of the study, only the false rosemary consistently performed better in northcentral Florida than in north Florida, whereas its non-native counterpart, barbeque rosemary, consistently grew larger in north Florida than in northcentral Florida (Table 2).
Similar to measured vegetative growth responses, native plants produced more flowers than non-native plants; this effect was weak in year 1 (χ2 = 3.34; P = 0.06) but strong in year 2 (χ2 = 18.46; P < 0.0001) (Fig. 2B). For example, native plants produced, on average, 1.3-times more flowers than non-native plants in year 1 and 2.1-times more flowers than non-native plants in year 2. As with vegetative traits, we found no evidence that irrigation affected the total flower density in year 1 (χ2 = 0.35; P = 0.55) or year 2 (χ2 = 1.73; P = 0.19). Finally, there was no significant interaction between provenance and irrigation on total flower density in year 1 (χ2 = 0.01; P = 0.94) or year 2 (χ2 = 0.15; P = 0.70).
Overall, provenance (i.e., native status) affected the total floral abundance of the 10 genera during the first year of the study (Supplemental Table 3); within each genus, native plants produced more flowers than non-native for seven of the 10 genera, with the exceptions of non-native Salvia, Scutellaria, and Viburnum, which produced more flowers than their native counterparts (Supplemental Table 3). Patterns differed slightly in year 2; for six genera, floral abundances were higher in native than in non-native species; however, for the sage and rosemary pairs, non-native species had higher floral abundance in one or both locations, and for two genera (Hibiscus and Monarda), there was no difference between native and non-native species (Supplemental Table 3).
Irrigation and location differently affected flower abundance across the evaluated plants (Table 3). In year 1 of the study, Florida scrub skullcap, pardon my pink beebalm, and barbeque rosemary had greater floral abundances under full water irrigation treatment, whereas swamp rosemallow and azure blue sage presented greater flower abundances under partial irrigation treatments (Table 3). The dwarf Burford holly responded differently to irrigation between sites; it had greater flower abundance under full irrigation treatment in northcentral Florida but greater floral abundance under partial irrigation treatment in north Florida (Table 3). In year 2, three species (barbeque rosemary, dwarf Burford holly, and pardon my pink beebalm), including two of the same species as in year 1, had greater flower density under full water irrigation treatment; however, Ruffled Satin® rose of Sharon had greater abundance under partial irrigation (Table 3). Across both years of the study, a few taxa consistently produced more flowers in north Florida (non-native bidens, barbeque rosemary, Ruffled Satin® rose of Sharon, and big blue sage) than in northcentral Florida; however, three taxa produced more flowers across both years of the trial in northcentral Florida than in North Florida (false rosemary, blanket flower, spotted beebalm) (Table 3).
Unique to this study, we paired pollinator-friendly native and non-native congeneric taxa to assess the effects of plant provenance (native vs. non-native) under different irrigation treatments on plant performance. We found that provenance affected plant growth and flowering more than irrigation for the species evaluated under the environmental conditions reported for the 2-year study. Across both irrigation treatments, Florida native plants (including shrubs and herbaceous plants) performed consistently better overall in terms of vegetative growth and flowering than non-native plants over both years.
Contrary to our predictions, native and non-native plants did not differ in their growth response to irrigation regimes. We had hypothesized that native plants would be better-adapted to typical Florida rainfall patterns, thus requiring less supplemental water in the form of irrigation. However, our overall finding was consistent with that of Scheiber et al. (2008), who evaluated 21 common landscape shrub species under different irrigation frequencies and found that responses to irrigation did not significantly differ between native and non-native species as a group. These findings imply that native plants may benefit from supplemental water for plant growth in ways similar to non-native plants, and that vegetative traits like growth index may be more affected by other abiotic factors such as temperature or soil characteristics and/or by plant genotype than water irrigation treatments, at least under nonextreme water availability conditions (Asseng et al. 2015; Vadez et al. 2023).
Although there was no overall effect of irrigation on vegetative growth across both years, we found that some individual species responded to the irrigation treatment. For example, five of 20 species, both native (Spanish needles) and non-native (Jethro Tull coreopsis, Arizona sun blanket-flower, pardon my pink beebalm, and Sandankwa viburnum) responded positively to full irrigation inputs in the first year of the study. In the second year, growth responses to irrigation inputs were less evident among species, with only the native spotted bee balm and non-native Ruffled Satin® rose of Sharon growing larger under full irrigation. This may have been a consequence of a plant’s establishment period because plants in the first year were actively growing and developing root systems needed for water use efficiency (Dukes et al. 2020) and the formation of symbiotic relationships with microorganisms in the soil. Similarly, we found that irrigation treatment did significantly affect survival of plants in year 1, the establishment year, with greater overall survival under full irrigation than under partial irrigation. In addition, the overall minimal responses to irrigation, especially in year 2, may have been influenced by the use of FFL non-native species as well as native species, which are well-adapted to growing in Florida landscapes (Florida-Friendly Landscaping Program 2024). In fact, non-native FFL plants have demonstrated water use efficiency similar to that of natives when planted in the appropriate place with the provision of water for establishment (Clem et al. 2021).
Finally, the results may be different under more severe water limitations. In this study, the drier partial irrigation treatment (at 10% volumetric content per hour) was chosen to limit water availability while still sustaining plants and reducing mortality across the 2-year study. This indicates that provenance (native status) has a more consistent long-term effect on survival rates and other vegetative traits when compared with that of irrigation (Fig. 2 and Supplemental Fig. 2). Furthermore, our water treatments were selected to represent common watering regimes used in Florida, which experiences periodic droughts during drier times of the year and has high water needs because of low soil water-holding capacity and high solar irradiance. Plants under the reduced irrigation treatment did visually show morphological symptoms of water stress at times during the study (wilt, leaf rolling, leaf senescence, and changes in leaf color); however, physiological and biochemical signs of drought stress (i.e., leaf water potential, shoot-to-root biomass, transpiration rate, stomatal conductance, and ion leakage) (Seleiman et al. 2021) were not measured because they were outside the scope of this study. Specifically, our study aimed to assess two different irrigation regimes that represent common and feasible irrigation strategies, but it did not explicitly seek to assess the effects of drought stress on these plant taxa. If irrigation is removed entirely, then we may see stronger responses to water availability, including outside of the establishment year, but at the risk of plant mortality.
Adaptation to the planting location additionally influenced the individual plant performance in our study. For example, in the first year, 17 of the 20 plant species that were evaluated exhibited higher growth indices in north Florida than in northcentral Florida, but the results in year two reflected a different pattern (Table 2). Of notable exception, false rosemary consistently performed better in northcentral Florida than in north Florida. This native species occurs naturally in sand and scrub xeric habitats, thus thriving in open, sunny conditions and well-drained soils (Dosoky and Setzer 2018; Matrazzo and Bissett 2020). Northcentral Florida has sandy, well-drained soils (Ferrarezi et al. 2020), thus likely explaining the positive response of this species to that location. The majority of the plant species that fared better in north Florida may alternatively prefer siltier soils with greater water-holding capacity typical of north Florida (Florida Natural Areas Inventory 2010) or soils with specific mineral compositions (Supplemental Table 1). Additionally, we noted differences in weather across the locations that could furthermore impact plant performance in north Florida compared with that in northcentral Florida (Supplemental Fig. 1).
Total floral abundance also varied by provenance in ways similar to plant size. That is, native plants produced, on average, 1.3-times more flowers than non-native plants in year 1, and 2.1-times more flowers than non-native plants in year 2. This result is consistent with a 2.0-fold increase in flower density by native plants as compared with that of non-native plants reported by previous studies (Kalaman et al. 2022b; Razanajatovo et al. 2015). Furthermore, the effects of irrigation treatment on flower performance were highly species-specific and location-specific, with some plants producing more flowers under full irrigation and others producing more flowers under partial irrigation, with no overall relationship between the native status and response to irrigation. The greater flower production by some species under partial irrigation suggested compensatory flower production or early flowering under limited water to ensure reproduction (Seleiman et al. 2021; Shavrukov et al. 2017). Other important adaptations of plants to water stress and drier temperatures include robust root growth at early stages of water stress to absorb the water in deep soil, thus producing smaller and thicker leaves, accelerating reproductive growth, and producing thicker stems (Dosoky and Setzer 2018; Fang and Xiong 2015).
Notably, although the effect of irrigation on floral abundance varied across years, locations, and genera, the effect of plant provenance on floral abundance was more consistent and significant (Supplemental Tables 2 and 3). Native plants of multiple genera, including Spanish needles, tickseed coreopsis, blanket flower, and inkberry, outperformed their non-native congeners in terms of floral abundance across both years and locations. However, of interest, in the first year of establishment, some non-native plants produced more flowers than their native congeners, such as the big blue salvia, Malaysian skullcap, and Sandankwa viburnum. In the second year, the big blue salvia produced more flowers than its native counterpart, representing the only non-native species to consistently outperform its native congener across years and locations. Previous studies have also shown that non-native ornamental Salvias are highly attractive ornamentals for different pollinators because they produce large floral displays (Kalaman et al. 2022a). Thus, although native plants produced more flowers overall, the exceptions suggest that some non-native plants can be valuable additions to pollinator gardens.
Selecting native plants can ensure that resources are provided for specialist pollinators that depend on specific native resources and a larger quantity of resources within larger and denser floral displays for more generalist pollinators (Lopezaraiza-Mikel et al. 2007; Memmott and Waser 2002; Seitz et al. 2020). Although floral abundance is one important determinant of pollinator resource availability, other floral traits, including the quantity and quality of resources per flower (i.e., pollen protein content, nectar quantity, and sugar concentration per flower), will additionally determine the resource value of a plant. Thus, additional studies that examine whether overall resource quality and quantity are enhanced in dense floral displays are necessary to comprehensively determine the resource value of native and non-native plants.
Considering the influence of the native status, irrigation inputs, and geographic location on plant size and flower density, our findings have implications for the selection of FFL plants. Our study results provide useful insights that can help landscapers, growers, and stakeholders with their selection and implementation of pollinator plant mixes in residential or agricultural landscapes in Florida. Further work is underway to determine the attraction of these plant taxa to diverse pollinators and changes in floral resources (nectar and pollen) and floral signals (flower color and volatile organic compounds) under induced water stress. Integrating these results with the relative attractiveness of the plants to different pollinator taxa will facilitate the optimization of future plant mixes for different pollinators such as butterflies, managed bees, and wild bees. We concluded that native plants generally outperform non-native plants in terms of size and flower density, and that partial irrigation (10% volumetric content per hour) may be sufficient to sustain plant size and floral density across both native and non-native plant species in typical rainfall years in Florida.
Floral representation of 10 native (N) and 10 non-native ornamental plants selected for this study. Photos courtesy of Joanna J. Silva and Walker Bensch.
Overall effects of provenance (native and non-native) on the (A) plant growth index and (B) total floral abundance in year 1 and year 2 of the study. Different letters indicate significant differences at P < 0.05.
Contributor Notes
We gratefully acknowledge the USDA NIFA Project and FNGLA for funding this project. We thank Kelly Thomas, Andy Woodruff, Shiala Morales, Stan Chabert, and Anthony Brown for their assistance with the field preparation, plant installation and maintenance, and data collection. We gratefully acknowledge financial support from the Fulbright Student Program, which is sponsored by the US Department of State and the Fulbright Commission in Colombia. The contents of this work are solely the responsibility of the author and do not necessarily represent the official views of the Fulbright Program, the US Government, or the Fulbright Commission in Colombia.
J.J.S. is the corresponding author. E-mail: joanna.jaramillo@ufl.edu.
Floral representation of 10 native (N) and 10 non-native ornamental plants selected for this study. Photos courtesy of Joanna J. Silva and Walker Bensch.
Overall effects of provenance (native and non-native) on the (A) plant growth index and (B) total floral abundance in year 1 and year 2 of the study. Different letters indicate significant differences at P < 0.05.
