A common quote on the internet, “It’s best to use strongly scented plants for companion planting because they confuse pests looking for their host plant.” This is a commonly voiced opinion, and it usually follows with a list of fragrant plants that make good companion plants, such as marigold, onion, mint, chives, lavender, wormwood and many other herbs.
This seems to makes perfect sense. Insects looking to lay eggs on a cabbage, might be confused when they smell a marigold and leave the area thinking there are no cabbages.
Read any of the thousands of web sites and books about companion planting and they all agree; strongly scented plants are your best option. The only problem is that none of these suggestions are based on science. ………. The facts are going to surprise you.
Companion Plants for Cabbage
If you have ever tried to grow cabbage you know that a number of insects like to lay their eggs on this plant, and the hatched larvae love to eat it. It’s a perfect subject for testing companion plants.
A research project at the Horticulture Research International, Wellesbourne, UK, planted cabbage in a field experiment, and surrounded the cabbage with 24 different kinds of plants including garden bedding plants, weeds, aromatic plants, and popular companion plants. Of these, 20 plants did result in fewer eggs being laid by the cabbage root fly.
The surprising find was that “the aromatic plants; curry, marigold, mint, onion, sage and thyme were not more effective at confusing the root fly than non-aromatic plants.”
Similar findings were found when they tested onions and the onion fly.
Green Leaves are More Important Than Scent
This same group of scientists have carried out a number of other experiments and have concluded that scent has nothing to do with a plants ability to deter insects. The most important characteristic are green leaves.
When plants with reddish leaves were tested, they performed poorly compared to similar varieties with green leaves.
Scent seems to play no part in how these insects select the leaf for handing. The insects simply land on a green leaf and then test it to see if it is a suitable host for egg laying. If not, they leave and try another leaf.
They are not attracted to the color brown and as a result don’t land on brown leaves or the bare ground beside a host plant.
It is leaf color and not odor that makes a good companion plant for the purpose of confusing herbivorous insects.
Will Paper Work as a Companion Plant?
In another, very elegant experiment various crucifers, including cabbage, cauliflower, Brussels sprouts, swedes (rutabaga) and turnips, were grown in clover. The clover acted as a companion plant, decreasing the egg laying of eight pest species from four different insect orders.
In another test area, the clover was replaced with plant models made with green paper. These worked just as well as the clover.
Since the paper produces none of the chemicals that the clover might produce, it clearly shows that it is the green leaves and not any plant chemicals that are confusing the insects.
The Appropriate/Inappropriate Landing Theory
The above described work has lead Finch and Collier to propose the Appropriate/Inappropriate Landing Theory, which in essence says that herbivorous insects find the general location of host plants by smell. As they get close, they land on plants based on leaf color, shape and size, not smell. They then decide if the leaf is appropriate, ie the right kind of host plant, or inappropriate, probably using smell and taste. If it is a host plant, they continue with egg laying. If it is inappropriate, they leave and try again on the same plant, a nearby plant or leave the area.
To Weed or Not to Weed?
Since weeds are mostly green, they make good companion plants. Unfortunately, most gardeners remove the weeds and expose bare soil or a mulch, neither of which is green. This has the effect of making it much easier for insects to find the host plant since there are the only green things left.
From the point of view of reducing insect eggs on host plants, it would be better to leave the weeds. The problem with this is that the weeds also compete with the host crop, reducing production.
Intercropping, growing two crops together may be a good alternative.
What is a Scented Plant?
Humans tend to think that all life forms are similar to ourselves. We smell a marigold and call it scented and then go on to assume every other life form thinks it’s scented too. That is one reason myths about scented plants are common.
But other life forms may not even detect the chemicals we sense from the marigold. And even if they can detect them, they may be interpreted differently. Dung beetles think manure smells great but you don’t see humans flocking to it.
Dogs smell all kinds of things we can’t. Insect pheromones – the sex fragrance of insects – cause male insects to fly long distances but are undetected by us.
Why is it then that we think insects will be affected by plants that we find fragrant?
Are Strongly Scented Plants the Best Companion Plants?
Most discussions of companion plants talk about this subject as if there is one thing to consider. In my previous post, Companion Planting – Truth or Myth?, I pointed out that companion plants can be used in many ways. It is not all about preventing insect damage. So in order to answer the question in the subtitle, we have to know the goals for the planting.
As far as insect damage goes, it seems fairly clear that scent has nothing to do with the prevention of herbivorous insect damage, at least not on crucifers. Any plant with appropriate large green leaves, including weeds, will reduce egg laying on plants.
What about other types of bugs like aphids? They may behave differently.
One thing that is interesting here is the fact that so many sources claim that scent is a big part of this, when it’s not. It illustrates how popular beliefs, and anecdotal reports can confuse gardeners. It also makes me wonder how the books on companion planting got their lists of “the best plants”? Clearly they are not based on facts.
Another key point about using companion plants for producing crops is that insect damage is only part of the picture. It is really all about yield, and none of the studies presented here addressed yield. In cases where yield is important, it may be better to remove weeds and use no companion plants. But we don’t know that, based on the current information.
The more I delve into companion planting, the more complex the picture gets.