With the advent of A-I, finding accurate horticultural information on-line has become even more challenging, so we thought we’d take time to explain which type of on-line sources you can trust and which you probably shouldn’t. Let’s begin by breaking down gardening websites into information sources levels. From most to least trustworthy, those include: Primary (first level), Secondary (2nd level), Tertiary (3rd level), Quaternary (4th level), Quinary (5th level), Senary (6th level), and Septenary (7th level). Once you pass below Level 4 sites, the chance of getting accurate information diminishes greatly.
There are several factors that go into determining if an on-line information source is valid. The first is to understand that some sites are based on actual observations, others are based on in-depth research with top growers/observers, while some are simply based on compilations of others observations, without regard of the authority where the information was sourced or who verified its accuracy.
Who is the steward/manager of the site? This is the single most important factor in determining if the site can be trusted. Site managers range from individuals to botanic garden staff, to academics, and specialty plant societies. Just because an organization is credible, however, doesn’t mean the information is accurate or up-to-date. Many institutions may outsource adding information to staffers, volunteers, and interns, sometimes without accuracy checks. Other sites are managed by encyclopedic, premier level plant authorities like Larry Hatch of Cultivar.org, John Grimshaw of Trees and Shrubs Online, and Alan Weakley of Flora of the Southeast US.


So, how do you determine the knowledge level of the author? The short answer is that it’s difficult. You almost need a high knowledge level of horticulture to know if someone else has a high knowledge level. Just because someone writes a blog, has a Facebook page, or even gardens, doesn’t tell you if they have an actual clue what they’re talking about. The people I trust the most as a source of primary information include: botanic garden staff, plant researchers, plant breeders, plant collectors, plant explorers, and keen amateur gardeners. The more plants a gardener has grown or observed/studied the subject, the more I’d trust their information.
Some folks brag about their credentials, which is often a red flag. If an author has good horticultural credentials, it usually comes through in their writing, without them needing to brag about themselves. One credential red flag that makes readers and nurserymen cringe is when a customer starts a conversation with, “I’m a Master Gardener”. If being an extension service volunteer is all they have to hang their hat on, then everything that comes out after that is just noise. Please don’t take that as an indictment of the program, as yours truly spent years as a Master Gardener in the 1980s, and those who don’t wear it as an introductory knowledge badge, are some of the most amazing people I’ve ever met.
Primary (Level 1) sites are those put together, and regularly updated by a premier level horticulturist/botanist. This could include first hand observation, gardening experience, or someone who regularly communicates with these folks, and records their findings and experience. These sites include some botanical gardens, some specialty nurseries, research sites (mostly academic), and some personal websites or Facebook pages. I particularly value sites that take a “deep dive” into one or more specific plant groups. Some of our favorite sites for trustworthy information include:
Agaveville, Biota of North America Project (BONAP), Cultivar.org, Flora of the Southeast US, Missouri Botanical Garden (MOBOT.org), Trees and Shrubs Online, San Marcos Growers, and hopefully our Plantdelights.com and JLBG.org.
There are number of Specialty Plant Society Sites, of which some are top notch, with regard to having accurate and up to date information. Others are not so dependable. I have asked Larry Hatch of Cultivar.org to help me rank the best.
Top Specialty Plant Society Sites: Alpine Garden Society, American Azalea Society, American Bamboo Society, American Begonia Society, American Boxwood Society, American Camellia Society, American Daffodil Society, American Hemerocallis Society, American Hibiscus Society, American Holly Society, American Hosta Society, American Iris Society, American Ivy Society, American Orchid Society, American Penstemon Society, American Peony Society, American Rhododendron Society, American Rose Society, Brugmansia Growers International, Cactus and Succulent Society of America, Cyclamen Society, Hardy Fern Foundation, International Carnivorous Plant Society, International Clematis Society, International Dendrology Society, International Maple Society, International Oak Society, International Palm Society, Magnolia Society International, North American Lily Society, North American Rock Garden Society, Pacific Bulb Society, and The Gesneriad Society. This list is not intended to be comprehensive.
Secondary (Level 2) sites are those which have first hand observations and/or compilations, but have no overseeing group of knowledgeable botanists/horticulturists to make sure the information is accurate. Secondary sites like INaturalist includes pictures of plants taken around the world. While the site is incredibly valuable, there is no knowledgeable person checking the site for botanical accuracy. On a recent search for Gloriosa superba, we found a large number of the photos so identified are actually a different species, Gloriosa simplex.
Tertiary (Level 3) sites are those of other information compilers, where we find far too many errors to include them in Level 1. We still regularly use many of these for our research, but we also are able to spot a percentage of errors on each. Many times, this is simply due to a lack of competent staffing, or the use of volunteers and interns. One of the most prevalent good on-line compiler sites is Kew.org (RHS Kew Gardens)
Quartenary (Level 4 ) sites are hybrids between a compiler site and an expert site. The best example would be Wikipedia. Since most of the information is written by volunteers, some is incredibly accurate, while others is completely outdated and in worse cases, wrong. Wikipedia can be a great place to start, but I always use it to trace back to an original source to confirm the information. Many State Extension Services have sites that would fall into this category, as would PFAF.org (Plants for a Future).
Quinary (Level 5) sites are compiler sites, but with editors, who don’t appear to be high-level gardeners, massage the information to try and make it look first hand. Starting at this level, and going forward are the sites I’d never trust for accurate information. A detailed look you will see that most of the information is simply cut and paste from other site, without any verification of its accuracy. Some of the information here is truly horrific. Sites in this Level 5 group include TheSpruce.com, BirdsandBlooms.com, Beejwala.com, GardeningKnowHow.com, Gardenersworld.com, gardzenonline.com, and Harvesttotable.com. One of the sites has an article on Gloriosa superba, that describes its needs for full or partial shade. Two paragraphs later, the same article says to plant them in at least 6-8 hours of full sun. The article goes on to say that deadheading causes more flowers, which is simply not true, but is “a cut and paste” line that is used with almost all flowering plants. Almost all articles about Lycoris at this level repeat that they flower better in full sun, which is simply not true.
Senary (Level 6) sites are those untrustworthy scraper sites that are done by computers. These sites “scrape” information from a variety of sites, but there is no knowledgeable human to review the information to see if it’s actually accurate. The most notorious of these sites that often comes up in searches is Gardenia.net
Septenary (Level 7) are A-I scraped sites. At the top of most searches now, you’ll find a paragraph or two titled “Overview or A-I Overview”. This should never be trusted as accurate. I find that these sites often pull from more than one site, and so compilations often mix information from a Level 1-6 site, so there is no way to distinguish accurate from inaccurate information. Like the Level 6 sites, there is no knowledgeable human to review the information for accuracy. In addition to the A-I summary at the top of your search, other A-I sites not to be trusted include GregApp.
A-I is a mixed bag of information, some accurate, and some troublingly incorrect. The problem is that most folks have no idea which is which.
Here are a couple of examples of an A-I search. If your search for Ophiopogon corifolius, here is what is A-I generated.
Ophiopogon corifolius is a species of Ophiopogon, commonly known as lilyturf, native to China. (so far, so good) It is a perennial, evergreen plant with grass-like leaves and is part of the Asparagaceae family. (so far, so good) The plant is known for its ability to thrive in shady, dry areas and it has a slow-growing, mound-forming habit. (Wrong…this is simply making an incorrect assumption, based on Ophiopogon japonicus) O. corifolius is actually a stoloniferous (running) species, and since it’s never been cultivated, no one knows its tolerance to drought.
Native Range: China (Guangxi, SW. Guizhou). (correct)
Growth Habit: Perennial, evergreen, slow-growing, mound-forming. (not correct)
Habitat: Thrives in shady, dry areas. (unknown)
Ophiopogon species, including O. corifolius, are often used as groundcover plants, particularly in shady locations. (not correct, since it’s not in cultivation)
Their attractive foliage and relatively low maintenance make them popular for landscaping. (Unknown)
At best, this A-I search yielded information that’s 20-40% correct.
How about this question, Will Lithops tolerate frost?

No, Lithops, also known as living stones, do not tolerate frost. They are native to arid, warm regions of southern Africa and prefer warm temperatures. Exposure to frost can cause their water-filled leaves to freeze, rupture, and rot. It’s best to protect them from freezing temperatures by bringing them indoors or into a greenhouse when temperatures drop below 40°F (4°C), according to some succulent care guides. As we have written many times, Lithops can easily tolerate temperatures to 10 degree F, as long as they don’t get winter rains. ) (This is a 100% fail.)
Let’s try a different question. What was the first mail order nursery to sell ecommerce on-line?
Determining the absolute first mail-order nursery to sell online is difficult, as the transition to e-commerce for nurseries happened gradually and often without formal announcements of being “first”.
However, some pioneers and early adopters in the online plant sales space have been identified:
Garden Crossings: This Michigan-based business started as an online mail-order business in 2002 and later added a physical store.
Perfect Plants Nursery: This family-owned business launched its e-commerce website in 2015.
Romence Gardens & Greenhouses: They began exploring e-commerce around 2010,
Don Eaton launched Bower & Branch in 2010, an e-commerce platform for independent businesses in the green industry.
So, how did A-I do? I’d call this an epic failure. We (Plant Delights) launched our on-line e-commerce in April 1997. Our neighbors, the former Niche Gardens, launched their e-commerce in 1996. A-I gets a 0% correct for this search.
Dear A-I, How is the best way to eliminate yellow nutsedge from the garden
The best way to eliminate yellow nutsedge from your garden is a combination of preventative measures and targeted treatments, focusing on consistent removal of plants and tubers, and potentially using herbicides. It’s a persistent weed, so patience and persistence are key. (So far, so good.)
Here’s a breakdown of effective strategies:
Composting: Do not add nutsedge tubers to your compost pile, as they may survive and spread when the compost is used. (good)
Monitoring: Continue to monitor your garden for new growth, even after treatment, and address it promptly to prevent further spread. (good)
Patience: Nutsedge is a persistent weed, so be prepared to implement a long-term management plan. (good) Prevent Establishment: (Duh)
- Remove Small Plants: Regularly inspect your garden and remove any young nutsedge plants before they develop tubers (the underground storage structures that allow them to spread). This is most effective when plants have fewer than 5 leaves. (good)
Eliminate Wet Conditions: Yellow nutsedge thrives in moist environments. Improve drainage in your garden by amending the soil with organic matter and ensuring proper watering practices. Avoid overwatering. (This makes absolutely no sense. Yellow nutsedge grows equally as well in wet or dry soils.)
Use Mulch: A thick layer of mulch like wood chips or straw can help suppress nutsedge growth by blocking sunlight and preventing tuber formation. (Whoever wrote this has never worked with yellow nutsedge, which thrives when mulched.)
Check Soil and Materials: Be mindful of where you get topsoil and other materials, as they can introduce new nutsedge tubers to your garden. (that’s actually good advice, but being mindful in and of itself isn’t particularly helpful.) - Control Established Nutsedge:
Hand Pulling: For small infestations, carefully hand-pull the entire plant, including the roots and tubers, to minimize regrowth. Be sure to get as many tubers as possible. (Good)
Summer Cultivation: Regular tilling or digging, especially in the summer when tubers are actively developing, can help bring them to the surface where they can be removed or die off. (OMG…that’s the best way to make yellow nutsedge spread like wildfire. Never, under any circumstances, would I recommend running a rototiller thru a patch.)
Herbicides: Selective Herbicides: For larger infestations, consider using selective herbicides specifically designed for nutsedge control. Look for products containing halosulfuron or sulfentrazone. (Well, halosulfuron can burn back the foliage, but isn’t effective at killing yellow nutsedge, or is sulfentrazone. At best, both can reduce the size of the patch.)
Smothering: Cover the area with cardboard and then a layer of mulch (like grass hay) to deprive the nutsedge of sunlight, effectively smothering it. This method can be more effective when combined with other methods like hand-pulling and herbicide application. (possible, if your don’t have anything planted, and won’t for years.)
We’ll be generous and give A-I a 50% score for the nutsedge question here, but the reply includes some very bad advice.
We hope you find this source evaluation useful as you traverse the online world, looking for accurate plant information.
Thank you for a very interesting and informative post!
Spot on, Tony! Thanks!
Thank you for this evaluation. I have found it increasingly hard to get reliable data on plants and gardening since A-I got involved, and really before that when gardening amateurs with good journalistic and marketing skills start websites without deep subject matter knowledge to back them up. The problem has gone from having too little data to having to dodge a torrent of trash data to find good, truly authoritative information. You have confirmed my suspicions of some of the unreliable sites too. Very helpful article.
Wow, what valuable information! I shared it with my Garden Club group as well!
Thank you for taking the time to pull all this content together, Tony!
What spectacular information to have. Needless to say, I will keep this list and use it to access truthful information from now on. I was unaware of so many of these resources. Thank You!
Thanks for another fantastic article with some new sources I’m eager to check out. Your websites are often the first I search when researching plants, especially newer cultivars. While not everyone’s growing conditions approximate yours, the history, context, and highlighting of unique characteristics, accurate size/age data of un-pruned specimens, and of course good humor should be appreciated by all.
I’m surprised you make no mention of land grant university sites like NCSU or UF/IFAS. While they can be a conglomeration with occasionally questionable regurgitated characteristics or inevitable errata, the sheer scope is impressive. There are some wonderful compendiums, with specific advice for different growing conditions often found throughout an area, and citations of more in depth coverage and peer reviewed scientific studies. The Raulston image search feature can be invaluable for mature growth characteristics.
Nutsedge is certainly challenging, if not downright impressive for it’s durability. Pulling the shoots only spurs an explosion of new ones, and digging out tiny nutlets seems futile. My understanding is that Halosufurone, etc., along with Glyphosate are indeed systemic, killing the roots/rhizomes and nuts. The issue is if you wait too long, until there are more than 3-5 shoots 3-5″ tall, when they’ve already spread new nutlets that are making their own roots/rhizomes. Halosulfurone apparently has some soil activity, both pre and post emergent, which may or not be desireable if used close to broadleaf ornamentals, etc. Some of the other sedge herbicides are labelled for “over the top” use on certain ornamentals, though perhaps not without some damage.
Even if applied they’re applied perfectly, it’s a multistage and multi year campaign.
I would think successive rounds of tilling, each followed by “stale bedding” and then preferably herbicide treatment or cultivation/flame weeding, etc. of the shoots would be the quickest way to eradicate an established stand. Here’s a link to a similar study:
https://www.agronomyjournals.com/archives/2024/vol7issue12S/PartH/S-8-1-56-657.pdf
Thanks for all y’all’s hard work and willingness to educate.
Great information here! My only comment would be to have put the A-I responses in italics or enclosed within quotation marks in order to make it very clear what was the A-I ‘voice’ and what was the writer’s.
Thanks Tony for the well researched an excellent information and advice.
I really enjoyed this letter.
It made me laugh when you made the comment about Master Gardeners because it also makes me totally tune them out when they begin a sentence with “Well, I’m a master gardener…” I mentioned that on a gardening forum once and was of course immediately attacked by someone who said he was a master gardener and if I just listened to them I might learn something. Never mind my 55 years of gardening experience, my 250+ garden books and the thousand or more magazine articles I have read and learned from. But I’m not a master gardener, so sad!
Your observation about iNaturalist is true. I too use it with care. For example, they correctly identify the Agave plants growing in barren, disturbed ground immediately southwest of the built up area of Real de Catorce, San Luis Potosi, as Agave X madrensis. However, they err in calling the green leaved natural hybrids of the Sierra Madre Oriental, from the Cerro Magueyal, south to near the Peña Nevada, as also belonging to Agave X madrensis {a hybrid of Agave gentryi and Agave lechuguilla first described from near the summit of the Cerro Potosi). The green-leaved Agaves are well-known in cultivation both as named cultivars (Bellville, Green Goblet, etc.) and as numerous unnamed selections of Mexican imports now found along the roadsides of Texas.
We have tried unsuccessfully to find seed of Agave x madrensis, so we have yet to grow it. The green-leaf hybrids are certainly a different taxa, but I’m not aware of anyone who has been able to confirm what species were involved with the cross, although we suspect Agave salmiana and another more winter hardy parent.
Agave salmiana is a possibility, however, it is a cultigen spread by the pre-Columbian and modern makers of pulque. The high elevation green-leaved Agaves common as understory plants in oak forests south of General Zaragoza, NL are far from any pulque-growing areas.
Are your ‘Monterrey Azul’ plants derived from seed collected at the Cerro Magueyal? The Cerro Magueyal is just north of the road from Horsetail Falls to Laguna de Sanchez in Nuevo León.
Unraveling the phylogeny of the plants of the Sierra Madre Oriental is complicated by the frequent and prolonged glacial periods during the last 2.588 million years of the Pleistocene. We are now left with many “tips” of phylogenetic trees with few or no intermediates remaining. Throughout the Pleistocene, glacial periods have taken up 90% of the time & interglacials (like the current one) just 10% of the time.
It wasn’t just the cold and montane glaciers of those glacial periods that wiped out so many phylogenetic ancestors, CO2 (already a trace gas – it is currently about one twenty-fifth of one percent of the atmosphere – compare to argon at almost one percent (0.93%) of the atmosphere) concentrations dropped close to the minimum level below which photosynthesis is not possible – about 190 ppm.
We are uncertain of the exact location of the Agave ‘Monterrey Azul’ seed. We need a DNA researcher willing to sample some of these mystery agaves and track down the parentage as best possible.
Good info. Thanks!