Soil types vary considerably and Proteaceae are fine-tuned for their particular soil type while a hybrid rootstock generally has a wider tolerance. Compared to the scion, the rootstock may also have a resistance to nematodes, certain pathogens, drought, high or low ph. etc. In most cases, it is more robust and stimulates the scion (Orothamnus zeyheri, Protea mucronifolia, P. rupicola, P. odorata).
If there is a one-off opportunity for propagating a rare plant, all three methods of propagation using seeds, cuttings and grafting spread the risk. I have found that cuttings and seed viability are sometimes unrewarding from plants in the wild as they may have undergone stresses that are not apparent. Germinated seedlings are sometimes vulnerable to differing conditions from which they evolved and the medium-term survival rate may be low, yet a grafted plant of the same species showed remarkable growth (P. effusa).
Grafting can be used to rescue cuttings that have failed to strike roots. I received material for two proteas (P. effusa and P. witzenbergiana) from a very inaccessible area – seeds and cuttings. The cuttings that were processed had very few striking after eight weeks. The few grafts I had done all failed, leaving me with seeds and iffy cuttings. The cuttings were looking healthy above the medium but with a gentle pull (a callusing stem will swell a bit and have resistance) it seemed I may get two or three strikes out of 15. I then grafted eight of the easiest to pull out and waited. The stems were black and rotting in the rooting medium and would not have rooted although the rest of the stem and leaves above the medium looked good and must have been receiving water in the misting bed through the leaves only. After grafting two died but six started to produce new shoots three months later. The moral of the story is don’t give up – try everything. Incidentally, the seeds then germinated strongly a bit later in early winter, yet I lost most of them to damping off in early summer, leaving me with the grafts that were from the rescued cuttings as well as a few rooted cuttings. Subsequently, I did the same with some very late to root Protea acuminata cuttings so it seems as if this is an option for emergencies.
I have successfully grafted a scion from a plant that was dying from phytopthora (Mimetes stokoei) so grafting for the purpose of rescue is possible. Phytopthora kills the roots, but apparently, the stems are unaffected until all the water reserves are used up. Of eleven grafts from the dying Mimetes, only the topmost scion was successful, the lower ones had probably used up all their reserves.
Quite often the grafted entity can be more successful in the medium term and reach maturity before a cutting and certainly a seed-derived plant, thus allowing faster development of a collection. It is well known that a third-generation manipulated pot plant will give a greater percentage strike rate for cuttings and a grafted entity can yield scions within six months and even a second batch of scions within the first year (Protea mucronifolia, P. rupicola). After a few years from planting out, the grafted plant can start to fade, perhaps due to the rootstock not delivering the correct nutrients (or not the correct amount) and in these instances, the plant should have been planted into a larger pot for specialist feeding. This is an ongoing enquiry for me as to what rootstock to match with specific species.
Orothamnus zeyheri only grows on south-facing slopes at around 700 meters and above in peaty conditions with high winter rainfall plus summer cloud condensation (2000 mm per year), yet grafted plants grow well at sea level and can live without summer irrigation. The opposite also applies as grafted P. sulphurea and P. humiflora do well at sea level while their natural habitat is higher, drier, and hotter with a different soil type.
A grafted motherstock in an orchard that needs little irrigation is a better prospect than the difficulties of maintaining a potted collection on its own roots. There is always a certain amount of annual die-off of plants to anticipate (due to pathogens such as phytopthora and age) so it is wise to ensure that every year you graft enough new plants to maintain the mother stock. This will ensure that you always have young plants developing that will give good scions. An ageing plant produces progressively shorter stems yielding fewer cuttings and scions.
Recently (2020) I have started a study to determine if seeds of grafted Proteaceae are viable, and here I must thank again the Botanical Education Trust of South Africa for their funding to push this enquiry. This is in response to observations that the seeds of my grafted Protea nana and P. aristata germinate poorly and it will take a few years to have any hard results. Hand pollinating will answer if the cause is a lack of pollinators for those species. I have hand-pollinated half of the flowers of a P. nana on one grafted plant as the first step of the experiment as this grafted plant only produced 12 germinated seedlings from 60 flowers while 62 flowers from the wild (thus ungrafted) produced 184 germinations. Subsequent to the hand-pollinating exercise, germination from the hand-pollinated grafted plant versus the same open pollination grafted plant were both very low, so finding a result needs further enquiry across more species and will also require using different rootstock.
As grafting is the radical joining of two plants, it follows that if one of them is not strong and healthy, it will bring down the other so all aspects of grafting, however, small can start some weakening. I shall be discussing even the smallest detail as cross genus grafting of Mimetes onto Leucospermum is unforgiving of weak procedures. Some of these details may be relevant for only a tiny benefit in one species but may be more productive in others.