One option is to exhaust other methods of gathering data before utilizing the homegrown crowd. For example, the database can import data from relevant plant databases that support the distribution of their data . Another example is to utilize a crowd that does not consist of community members from Mechanical Turk, for example, to crawl community-referenced websites and texts for data. Off-loading the work of the homegrown crowd onto import scripts and non-affiliated crowds allows the homegrown crowd to focus on seeding the plant database with folk knowledge from within their community. The crowd’s collective intelligence should be used to address the challenge of seeding the database with folk knowledge by framing the challenge as a cooperation problem rather than a collection of cognitive problems. Framing database seeding as a cooperation problem requires individuals in the crowd to factor in what other people are doing to make decisions that have mutual advantage. The technology steward can facilitate seeding the plant database with datasets that crowd members may already have. After importing their personal lists, crowd members can choose to add data that benefit themselves and their peers. For example, if the database has a lot of entries for nitrogenfixing plants but few on plants that attracted pollinators, it would be mutually beneficial for the member and the crowd if the member adds data for pollinators rather than nitrogen-fixing plants. Similarly, tomato grow bags crowd members can choose to recruit others for their local plant knowledge that is not already represented in the database . Solutions to cooperation problems often require trust and are often built upon cultural norms and conventions to regulate behavior .

Members of the crowd will have to trust each other to put in data that they believe is of good quality. Adding data of poor quality, including data that does not reflect the community’s cultural norms and values, is mutually detrimental to the person who added the information and to the rest of the crowd because both in-turn use that information to make decisions in their sustainable polyculture designs. The crowd’s collective decisions on individual data points is likely more appropriate for addressing issues of quality control than to seed the entire database. If a datum is under disagreement from a small subset of the crowd , it should be flagged as a point of conflict. A query for this datum should then be dispatched to a well-formed subset of the crowd so that their collective wisdom can be used to gauge which is the correct answer. Search-misses should be addressed by both individual or collective decisions. A search-miss occurs when a user searches the database for a plant or property that the database does not have. The user could then opt to enter the datum, thus creating an individual decision. By entering their own data, some users will join the crowd for their first time, thus growing the size of the crowd. If many misses occur for a specific search criterion, the system could dispatch queries to the crowd for that plant or property. The crowd will collectively decide on a value for this datum, increasing its likelihood of being correct. A dispatching system should not query the crowd for all data points that are search-misses because the crowd could become overwhelmed with requests while the database is sparsely populated. For both users and crowd members, the lack of data in the database during its infancy increases the possibility of fragile engagement. Due to the potential for fragile engagement, the order in which to tool is introduced to various parts of the information ecology may be crucial.CalFlora lists each plant’s associated beneficial organisms, such as bees and butterflies.

CalFlora has some data about plant tolerances, primarily in regard to soil characteristics, temperature, and rain. However, the database provides no ethnobotanical data, such as whether it is food producing. The lack of ethnobotanical information requires users to search other information resources to determine if a plant will provide them with products or services. CalFlora’s advanced search allows a user to search and filter the data based on most available plant data points, making it easy to search for California wild plants in a specific county and microclimate. CalFlora’s data cannot be exported. However, search results, containing taxonomic rank, common name, status, life form, and family, and plant information pages are formatted as text and can be copied and pasted into a spreadsheet, though this process would be long and tedious for collecting large amounts of information. The primary limitation of the CalFlora database is the lack of form, ethnobotanical, and ecosystem service data needed to include California Wild plants in an agroecosystem design. In contrast, the SAGE Plant Database is designed to provide ethnobotanical and ecosystem service data of California wild plants. This is important in agroecosystems because wild plants can be used as alternatives to traditional agriculturally productive plants in effort to cater to the animals and insects that depend on native flora for habitat and food. SAGE is also designed to include form data useful in assessing spatial constraints and opportunities of an agroecosystem design. For example, a user might search plants that create an overstory vertical layer, but without form data they cannot determine which trees would make a good overstory. In addition, while CalFLora has growing condition data limited to wild, unmanaged, or native ecological context, the SAGE Plant Database is designed to include growing condition data beyond CalFlora’s tolerance and soil data so that users can understand how to care for California wild plants in maintained, mixed agricultural landscape.

The data found in USDA PLANTS is sourced from an extensive network of federal partners and institutions and is curated by the small National Resource Conservation Service National Plants Data Team. It is an expert resource because the data is derived or validated through research efforts. USDA PLANTS has an extensive list of plant attributes it catalogs including distribution, taxonomy, ecology, legal status, morphology/physiology, growth requirements, reproduction, and suitability/use data. The database can be searched by over 120 attributes. All search results can be exported to a comma separated value file. Although there are nearly 50,000 plants in the PLANTS database, only about 2,000 plants, all of which are plants used in conservation efforts, have defined “characteristics data,” mostly consisting of intrinsic characteristics but also including some tolerances, products, and services . A search for most agricultural plants, like fruit trees, will have hardly any data available. For example, the PLANTS database only returns 28 plants that grow in in the county the Manzanita community is located in that are palatable to humans, not because there are only 28 palatable plants that grow in that county, but because those are the only plants that have “human palatable” data. The primary limitation of the USDA PLANTS Database is the incompleteness of the data set. The USDA PLANTS Database is only populated by staff and partners, and not by average users. Their data population method ensures level of quality control that crowd sourced data could not. However, for the permaculture communities, such a level of quality control is not essential because they are engaging with small-scale systems with lower financial risk than industrial agriculture or regional conservation efforts. The SAGEPlant Database design attempts to address incompleteness by enabling and encouraging users to contribute data that is absent. The USDA PLANTS Database does not catalog region-specific growing requirements, which participants use in both their design and maintenance of permaculture systems. The SAGE Plant Database design includes region specific growing requirements and does so because members of the local community are able to contribute their personal experiences to the data set.The EDIS publication database has over 7,000 peer reviewed publications produced by the University of Florida Institute of Food and Agriculture Sciences. The publications cover a range of topics including, adjacent to, and beyond plants, such as agriculture, community development, ecosystem restoration, grow bags garden consumer information, lawn and garden care, and sustainability. Many articles are about a single species of plant, where as other articles are about a broader topic and provide only a little bit of information about a plant. Single plant publications within this database have the most similar format to the information stored of a plant in a plant database. For example, a publication about white mulberry has a synthesis followed by an attribute list of properties and values and a list of references .

Attribute properties for the white mulberry include taxonomic and distribution information, form characteristics, growing conditions, use and management summary, pests, and diseases. This list of attributes is not exactly the same for each single-plant publication but is representative of the kinds of data found in those publications. The primary way to find information relevant to permaculture design or practices is to use the basic search function. Users can search for any keywords of their choosing from the publications, such as the scientific or common name of a plant, insect, or animal, or land feature . Users can browse articles by topic such as agriculture, community development, environment, or lawn and garden. To browse plants and plant information, a user must look in many places, which can make it hard to find data. For example, the Environment root topic has Plants subtopic, the Agriculture root topic has a Crops subtopic, and the Lawn and Garden root topic has a Landscape Plants subtopic. Although the EDIS publication database is a rich information resource, its primary limitation is the difficulty that a user has in finding the information they are looking for. For example, the EDIS does not hyperlink key terms or concepts across publications.When there is a plant or concept discussed in one publication that is expanded upon in another, it is up to the user to connect the information between the two. Second, because much of the data is written in a manuscript format, the user must spend time reading large chunks of text to locate the information they need or to determine that the information is not present. In contrast, with the plant information in a database format, like SAGE, users can quickly locate and sort through plant data and information.The 127 plant profiles are similar in content to the information found in a plant database. These profiles are brief, typically only single paragraphs, but include information about the plants form, their ideal growing conditions, companion plants, and ecosystem services. Sometimes, these profiles also include information about how the Native Americans used the plant. The ToLN plant information, and particularly the plant profiles, provide visitors with a succinct set of native plant information. However, because the information is limited to native plants, users are unable to explore native plant relationships with nonnative but agriculturally productive plants – a technique often used in permaculture. In contrast, the SAGE Plant Database supports this sort of exploration by featuring a range of non-native plants that are valued for their ecosystem services or human uses in addition to native plants. The ToLN plant information has similar limitations to those of the EDIS publication database in the sense that the data or information contained within those documents are not cross-referenced. Finally, the ToLN plant profiles omit pertinent details regarding the plants growing conditions, uses, or form in attempt to be brief. By filtering for specific properties, the SAGE database can provide data as brief or as extensive as the user requires.The Natural Capital Plant Database is a plant database designed specifically to support practitioners engaging in permaculture projects. Staff and registered contributors provide the data through referencing scholarly resources. The plant attribute data are similar to that of SAGE, including category, characteristics, tolerances, behaviors, human uses, and ecological functions. The Natural Capital Plant Database also lists which user polycultures a plant is a part of, associates of a plant , and compatibilities and incompatibilities with other plants. The Natural Capital Plant Database has four membership tiers, including free and paid. The Annual paid membership allows users to search the plant by site conditions, ecological functions, human uses, and limiting factors. Designer memberships are more expensive and allow users to do customized searches of the database based on their site conditions and download comma separate value reports. Researcher membership are designer memberships with the additional allow users to supply new plant data.