King's College Cambridge wildflower meadow monitoring data: biodiversity, climate change and society
Data files
Mar 24, 2023 version files 198.45 KB
-
biodiv_spdist_bats.csv
-
biodiv_spdistdata.csv
-
Biodiv_summarystats.csv
-
README.md
-
soil_organic_matter.csv
-
surveyresponses_redacted.csv
Abstract
The biodiversity and climate crises are critical challenges of this century. Wildflower meadows in urban areas could provide important nature-based solutions, addressing the biodiversity and climate crises jointly, and benefitting society in the process. King’s College Cambridge (England, UK) established a wildflower meadow over a portion of its iconic Back Lawn in 2019, replacing a fine lawn first laid in 1772.
We used biodiversity surveys, Wilcoxon signed rank, and ANOVA models to compare species richness, abundance, and composition of plants, spiders, bugs, bats, and nematodes supported by the meadow, and remaining lawn, over three years. We estimated the climate change impact of meadow vs lawn from maintenance emissions, soil carbon sequestration, and reflectance effect. We surveyed members of the university to quantify the societal benefits of, and attitudes towards, increased meadow planting on the collegiate university estate.
In spite of its small size (0.36 ha), the meadow supported approximately three times more plant species, three times more spider and bug species and individuals, and bats were recorded three times more often over the meadow than the remaining lawn. Terrestrial invertebrate biomass was 25 times higher in the meadow compared with the lawn. Fourteen species with conservation designations were recorded on the meadow (six for lawn), alongside meadow specialist species.
Reduced maintenance and fertilising associated with meadow reduced emissions by an estimated 1.36 Mg CO2-e per hectare per year compared with lawn. Relative reflectance increased by 25-34% for meadow relative to lawn. Soil carbon stocks did not differ between meadow and lawn.
Respondents thought meadows provided greater aesthetic, educational, and mental well-being services than lawns. In open responses, lawns were associated with undesirable elitism and social exclusion (most colleges in Cambridge restrict lawn access to senior members of the college), and respondents proved overwhelmingly in favour of meadow planting in place of lawns on the collegiate university estate.
This study demonstrates the substantial benefits of small urban meadows for local biodiversity, cultural ecosystem services, and climate change mitigation, supplied at a lower cost than maintaining conventional lawn.
Methods
Study site
Meadow establishment
The meadow area covers about 40% of the original extent of the King’s College Back Lawn, which was first laid in 1772. The dimensions are 96 x 66 m lawn (0.63ha) and 96 x 37 m meadow (0.36 ha). A soil study commissioned before sowing showed both the topsoil (of 30 cm depth) and subsoil were strongly alkaline (pH 8.4) sandy loams. The topsoil had intermediate fertility (20-27 mg/l extractable phosphorus, 131-167 mg/l extractable potassium, 0.50-0.52% total nitrogen using Dumas method), whilst the subsoil had moderately high fertility (35-54 mg/l extractable phosphorus, 69-129 mg/l extractable potassium, 0.24-0.39% total nitrogen). Thus topsoil removal was not necessary, and seed was sown into glyphosate treated scarified topsoil at 6 g/m2 in October 2019. Three different seeds mixes sourced by Emorsgate were sown: the Great Lawn meadow mix, a perennial meadow species mix intended as the long-term flora of the meadow; a Cornfield Annual mix intended to provide first year colour; and a Supplementary Mix composed of species with lower establishment probability from seed, but high conservation value (Table S3).
Meadow management
The meadow is managed as an East Anglian hay meadow following traditional Lammas practices as far as possible. Hay is cut once a year around August 1st (Lammas day) to a height of c. 350 mm, with one subsequent cut at 350 mm in December, in place of the historical light grazing. Hand weeding was performed through the visitor seasons to remove the occasional individual of undesirable species (mainly Sonchus oleraceus and Cirsium vulgare). No other management or intervention has been practised. Management of the remaining 60% of the lawn continues as before; the lawn is a fine lawn mix with Agrostis stolonifera and Festuca rubra dominant. It is maintained with twice-weekly cuts from March–September, weekly cuts from October–December, dropping to biweekly cuts in January and February. NPK fertiliser is applied at c. 30 g/m2 in spring (8% N, 7% P, 8% K) and winter (3%N, 8% P, 8% K). A selective herbicide (Praxys) is applied to the remaining lawn at the minimum dosage once to twice per year. Insect pesticides are no longer applied, and watering is avoided as far as possible. Fertiliser and herbicide are applied in a directional fashion by ride-on vehicle during suitable weather conditions only to minimise run-off.
Biodiversity
The study has a before-after-control-impact (BACI) design for the plants, invertebrate, and nematode datasets, with sampling initiated before the meadow was established.
Plants
Botanical surveys were carried out in July for each flowering summer (2020, 2021) and in September for the pre-sowing baseline (2019). Five quadrats 50 × 50 cm were placed every 15 m perpendicular to the edge in both the meadow and the lawn (KBME01-KBME05, KBSO01-KBSO05). The origin of the meadow transect is 15 m from the northern lawn edge, and 5 m from the eastern lawn edge, at latitude 52.204691 °N, longitude 0.115580 °E. The origin of the lawn transect is 15 m from the southern lawn edge, and 5 m from the eastern lawn edge, at latitude 52.204045 °N, longitude 0.115737 °E. Abundance was measured by counting presence in each of 25 equal subdivisions of the quadrat. Mean plant height was recorded. In addition, running checklists of all species present in the lawn and meadow separately were collected over the course of each year, with 2-3 principal recording visits made each year in March, April, and July. Plants were identified as sown or non-sown using the stated seed mix (Table S3). Plant attribute data (distribution, scarcity) were sourced from PLANTATT (Hill et al., 2004). Designated species follow JNCC (2022).
Invertebrates
Above ground invertebrates were sampled by sweep net (July 2020, July 2021, pre-mowing) and pitfall trap (September 2019, 2020, 2021, post-mowing) at five points in both the meadow and lawn. Sweep net transects were 20 paces each, centred on the plant quadrat locations. Sweep net specimens were identified to species level for all taxa in 2020, employing morphospecies names as necessary, and to species for Hemiptera (bugs) and Araneae (spiders) only in 2021. 2020 data were restricted to Hemiptera and Araneae for analysis. Pitfall traps were sited at the centre of the plant quadrat locations. Pitfall trap specimens were weighed in 2021 only. Pitfall specimens were identified to species for all taxa present in 2019, and to species for Hemiptera, Araneae, and Orthoptera only in 2020 and 2021. Spider attribute data (hectad occurrence, habitat preferences) were sourced from British Spiders (2022). Hemiptera habitats were sourced from British Bugs (2022), with hectad distribution data from National Biodiversity Network (2022). Arthropod body size data were compiled from NatureSpot, British Bugs, BugGuide, and Bugwoodwiki (2022); male and female maximum body lengths were averaged. Designated species follow JNCC (2022).
Bats
Bats were surveyed via two unattended ultrasonic recorders (Wildlife Acoustics Song Meter SM4BAT FS Ultrasonic Recorder) placed adjacent to the meadow, and the lawn. Recorders were left for five or six nights each over four recording periods in May, June, July, and October in 2021 only. Audio files were auto-identified to species using Kaleidoscope version 5.4.6 before being checked manually. All records of Barbastella barbastellus were accepted, one record of Myotis bechsteinii was assigned to Myotis daubentonii, Plecotus austriacus records were assigned to Plecotus auritus or Eptesicus serotinus, one Rhinolophus ferrumequinum record was assigned to Pipistrellus pipistrellus. Myotis species are generally considered indistinguishable by audio recording only. The only Myotis species recorded in our dataset was auto-identified as Myotis daubentonii, which was also seen foraging at the river, and so the identity has been retained for analysis. The total number of echolocations recorded over the year in each habitat is used as a proxy for abundance (several passes by the same bat would not be distinguished). Designated species follow JNCC (2022).
Soil nematodes
Soils were sampled contemporaneously with the pitfall traps in September 2019, 2020, and 2021, and were co-located. Approximately 7 cm width by 10 cm depth of soil was dug and mixed. Nematodes were extracted by wetting 180-200g of soil on top of a paper towel with RO water. The wetted soil was left overnight in a tray covered with an autoclave bag to prevent evaporation. The flowthrough was collected in 1 L glass media bottles (Fisherbrand), and left to settle at a 45° angle for 24 hours. The sediment was pipetted into a 50 mL conical centrifuge tube (Corning) using a soda lime glass pipette (Fisherbrand) and centrifuged at 300 x RCF for 15 minutes. The pellet was transferred to a 1.5 mL microcentrifuge tube (Eppendorf) and centrifuged at 20 000 RCF for 1 minute and snap frozen in liquid nitrogen. The frozen tissue was lysed at 30 Hz in a tissue lyser (Qiagen) for two minutes with one 5mm and two 2mm glass beads (Qiagen). From the samples, DNA was extracted using a ChargeSwitch™ gDNA Micro Tissue Kit mini protocol. Using the well-established 18S RNA primers, NemFopt and 18Sr2bRopt (Waeyenberge et al., 2019) DNA was amplified (Q5® High-Fidelity DNA Polymerase) via PCR and cleaned using the Monarch® PCR & DNA Cleanup Kit 5 μg. The amplified DNA was sent to the GENEWIZ Takely Lab (UK) for next generation sequencing.
Climate change
Carbon sequestration
Soil organic matter (SOM) was measured as a proxy for soil carbon sequestration. Soils were sampled contemporaneously with the pitfall traps in September 2019 and 2021, and were co-located. 7 cm width by 10 cm depth of soil was dug and mixed. For SOM, 100 g of soil from each sample was dried at 70°C for two days, homogenised and sieved (2 mm), then weighed into three pseudoreplicates of 5.00 g each per sample location. SOM was estimated using the loss on ignition method: samples were subjected to 8 hours in a muffle furnace at 450°C and reweighed once cool (Pribyl, 2010). SOM for the meadow and lawn samples were normally distributed and were compared using a t-test. We used a conversion factor of 2 (Pribyl, 2010) to convert from SOM to soil carbon i.e. soil organic matter is 50% carbon, and a literature value for soil density of 1440 kg/m3 for sandy loam (Yu et al. 1993). Above ground dry biomass was estimated for the meadow by counting the hay harvest in bales, weighing a bale, calculating the proportion of water in a bale by oven drying a sample, and multiplying up. These values are not included in the carbon sequestration figures as the pool is short-lived, nevertheless the productivity of the meadow is noted here.
Society
A survey was designed to assess respondents’ opinions of the cultural services provided by meadow and lawn, and respondents’ preferences for meadow and lawn (Appendix S1). Ethics oversight for the survey design and administration was provided by the Cambridge Hub. The survey was administered once in 2021 with responses recorded between 6th February and 26th March. At this time the meadow had had one flowering season and was in a winter dormant period. Given the timing and method of recruitment, respondents are likely to have seen the meadow for themselves, though we did not insist on this. 278 respondents were recruited via mailing lists of the University faculties, colleges, societies, and University affiliated organisations. Respondents were informed of the purpose, methods, and end use of the research and gave their informed consent to their data being collected and used for the purposes described in a privacy notice. An opt-out of having answers quoted was provided. No risks to participants were identified and participants were free to withdraw at any time. A small financial incentive was offered to respondents in the form of an Amazon gift voucher awarded to two randomly chosen respondents. Participants remained anonymous, unless they opted into being contacted for the randomly selected reward. All identifying information was deleted after disbursement of the rewards. Questions were always asked in the same order. The question of preference for lawn, meadow, or a mixture was repeated after the provision of information on the benefits of lawns and meadows. This information consisted of a written summary of the provisioning, regulating, cultural, and supporting ecosystem services derived from wildflower meadows and lawns, and was written by the survey administrator from published peer-reviewed literature. References to the primary sources were provided to participants. Responses were analysed using Wilcoxon signed rank tests, and chi-squared tests of association. Open responses were analysed by identifying and exploring common themes qualitatively.