Melbourne’s trees and parks get plenty of attention. But thousands of smaller garden beds sit at street level, and their condition is less visible. This report asks a simple question: are the garden beds actually in good condition? It starts with where the beds are, then looks at their condition, the kinds of plants recorded and two possible differences: irrigation and bed size.
The evidence comes from the City of Melbourne’s 2023–2024 inventory, collected by Royal Botanic Gardens Victoria. This is a snapshot of garden beds managed by the municipality, not an assessment of all Melbourne green space today.
We first need to understand the coverage of the inventory. Use the legend to focus on parks or streets, and hover for exact counts.
Reading the chart: Each horizontal bar is one neighbourhood. The two colours show how many of its recorded beds are in parks and streets. The numbers describe inventory counts, not how much garden space each neighbourhood has; a tiny and a large bed both count as one.
The inventory records an overall condition score from 1 (Total failure) to 5 (Excellent). This chart shows the share of those ratings for each bed type, so a large group does not automatically dominate. Bed types with fewer than 10 valid assessments are excluded to avoid misleading 100% bars from very small groups.
Reading the chart: Each row totals 100%. The warm colours represent poorer ratings and the green shades represent better ratings. Comparing percentages shows differences in the mix of ratings, not differences in the number of beds. These observed ratings do not prove that a bed type itself causes better condition.
The inventory distinguishes Indigenous (local to greater Melbourne), Native (Australian, but not necessarily local), Exotic, and Mixed planting classifications. The bars compare only beds with a specified origin.
Reading the chart: Each bar totals 100% of beds with a known origin in that location. Hover on a colour to see its exact percentage and number of beds. These are bed-level origin classifications, not counts of individual plants or species. Comparing bars tells us how the recorded composition differs, not which location is more biodiverse.
The usable irrigation comparison in this dataset is concentrated in street beds, so the chart below intentionally compares streets only. This avoids presenting a Park-versus-Street comparison that the available irrigation records cannot support.
Reading the chart: The line inside a box is the median condition score and the box contains the middle half of observed scores. Overlap indicates that ratings are spread similarly across the two groups. This is not a causal test: the inventory does not tell us whether irrigation was installed because a bed needed it, or whether maintenance, bed type or other conditions explain a difference.
Each dot is an assessed bed. The horizontal axis is its area, and the vertical axis is the condition rating. Use the dropdown above the chart to switch between Both, Park only and Street only. Hover over any point for its size and actual score.
Reading the chart: Area uses a logarithmic axis: 1 to 10, 10 to 100, and 100 to 1,000 each represent a tenfold increase. The dots are nudged slightly vertically to make overlaps visible; recorded scores remain whole numbers. Try the dropdown to see if the pattern looks different in parks and streets. This graphic describes association, not cause.
The five charts make that summary more specific. They show where beds were recorded, how condition varies by bed type, how the planting mix differs between parks and streets, and why irrigation and size alone cannot explain condition. This is a more careful answer than declaring every garden bed healthy or unhealthy. The strongest conclusion is about the variation inside Melbourne’s smaller green spaces—not about what caused it.
The published inventory describes more than 4,100 garden-bed assets. This report starts with 17,721 raw rows and uses 4,040 distinct non-empty asset IDs after removing repeated records; 4,017 of those have a valid overall condition score (1–5). The different counts arise from the chosen asset-ID filter and available rating values; they should not be confused with the official inventory headline. The analysis is at garden-bed level, so plant species listed on separate source rows are not counted as extra beds.
Missing/unknown neighbourhoods and origins are excluded where they are not meaningful categories. For the bed-type chart, categories with fewer than 10 valid ratings are excluded and sample sizes are shown. The irrigation chart is restricted to street beds with Yes/No irrigation data. The inventory covers only the City of Melbourne, was assessed during 2023–2024, and cannot establish causal relationships. Overall condition is a five-level ordered rating, not a continuous biological measure of health. Because repeated asset rows can contain species-level details, this report deliberately focuses on bed-level fields and does not count individual species.
City of Melbourne. (2025). Renewals for Nature: City of Melbourne garden bed inventory 2024 [Data set]. City of Melbourne Open Data Portal. https://data.melbourne.vic.gov.au/explore/dataset/renewals-for-nature-city-of-melbourne-garden-bed-inventory-2024/
I acknowledge the use of ChatGPT solely for the refinement of written work, including grammar correction, clarity, and readability. The tool was used to improve the presentation of the written content without changing its original meaning or intent. I ran and reviewed the visualisations in RStudio and published the report on RPubs.