Earth Worm Experiment

Investigating the Horticultural Ecology of Earthworms

Hugh M. Lewis

In the Fall of 2003, as part of my daughter's science experiment in her 8th Grade science class, we arranged to investigate in a systematic manner whether standard garden earthworms, that we had been cultivating in our mulch heap for the previous year, made a difference in terms of the growth of new plants. We believed that the addition of worms to the soil, all other things being more or less the same, would make some significant difference in terms of the early growth of new plants. Common knowledge held that earthworms were basically good for plants, that they were the "intestines of the soil" and that their excrement was a form of fertilizer that made plants growth faster and healthier than otherwise. We wanted to test this knowledge out for ourselves to see whether or not it was really true or just an unquestioned wives' tale that had been handed down from before anyone remembered.

Our experimental group of new plants contained earthworms added to potting soil, and the control group was without earthworms using the same potting soil. We purchased seeds, plant sets, and standard sized containers from a local nursery. We obtained as well a potting mix soil for seeds that contained organic ingredients and appear to have enough humus to sustain earthworms.

We ended up using several different kinds of plants, including corn, sunflowers, Swiss chard, spinach, and radishes. We picked plants as sets that were available at that time of the year (Spinach and Swiss Chard) and those as seeds that we could be assured of fast germination and early growth (Corn, Sunflowers and Radishes).

Our mulch heap had been developed for over a year, and was a rich source of earthworms of at least two different varieties--the standard, long, night-crawler, and the shorter earthworm that is more commonly found in this vicinity. We dug up a total of about 75 earthworms, distributed between the different experimental containers. We used worms of all sizes, from large to small, and therefore of all apparent ages.

We got off to a relatively late start in the year, October 19th, 2003, and we were delayed about a week by the poor results in germination of the first round of seeds we planted. Mildew developed in our seed bed, and destroyed most of the young seedlings. We changed the original mix to fifty percent loam, taken from the same location for both control and experimental plants. Therefore, the time available for conducting our experiment was limited to about 6 weeks in total, which put a fundamental constraint up what we could do as well as the number of measurements we could ultimately make.

Because it was late in the year, the Sun was relatively low on the horizon and shadows cast by trees and walls traveled within hours across the backyard area. We would have to move all the plants at least three times in the course of an afternoon so that they would get the full benefit of the sun, and in the late afternoon we would move the plants inside, where we kept them under a set of plant-lights that we had rigged. All in all, the conditions entailed a considerable daily output of effort and attention on our part to keep the plants in the maximum amount of sun or available artificial light as much as possible.

Our original addition of earthworms to the experimental containers was a total of more than seventy worms that we dug from our compost heap. We added on two separate occasions more earthworms to each experimental container, calculating the possibility of any losses that may have happened along the way. The total earthworms we worked with were around one hundred in number.

Once germinated, the plants sprouted and grew very rapidly in the next few weeks. We attempted to measure all the plants by their maximum height, each evening, and we missed only a few days between. We would measure to the top of the largest leaf in the first couple of weeks when the plants had just germinated and were only showing their first set of leaves. By the end of the measurement period the plants began taking on a more mature growth form, making measurements easier to do except for the incumbent crowding do to the rapid increase in size of each plant. If this experiment were conducted again, I would make sure to provide more spacing for the individual plants--even providing each plant with their own individual pot.

We moved the plants inside daily at about 5:00 P.M. and kept them under the plant-lights until about 10:30 or 11:00 PM each evening, until the bulbs were turned off. The plants were taken back outside early each morning about about 7:00 or 7:30 AM, and then shifted from location to location to catch the most sun until the following evening.

It was noticed in the first week that earthworms were attempting to leave the pots in the darkness and travel about in the room--we found 5 or 6 dead ones in the carpet, and one or two who had gotten across into the control pots. From then on I made a point of separating the control from the experimental pots by greater distance, and did not notice a great deal more earthworm movement at night, though during the course of the study we found one or two more dead ones in the carpet, and one had even made it as far as the nearby storage closet!

I made a point of "replenishing" the experimental pots periodically with injections of new worms I would dig from the same mulch heap.

Overall, the results of this experiment can be said to be equivocal in terms of the measured growth differentials of most of the young plants. In fact, in a few cases, the control group actually outperformed the experimental group, suggesting that in fact the presence of so many earthworms in the pot may have actually resulted in a competition between the worms in a limited container and the growth of the new plants in the same confined space. The experiment was quite successful in the sense that recovered from the experimental group was most of the worms we originally put into the pot--to the point that they appear to have rapidly adapted to the conditions of the pot environment, and even created for themselves their own little environment. At the same time,  only one or two worms were found in any, or even all of the control pots after the dirt was emptied and carefully, somewhat tediously sifted from the collecting trays.

By the final day, we did not know if any earthworms had survived in the pots or not. Even as most of the containers were clear plastic and therefore see-through, we could not clearly make out evidence for living worms except for a few burrow trails and holes on the side. Therefore, when we finally did break out the soil from the pots, thereby destroying the plants (we harvested the spinach and chard for dinner) we were surprised to find most of the worms we had originally put into the containers, or that were part of the complement added subsequently. From the set of plants that held individuals, 3 experimental and 3 controls, we found 12 earthworms from the experimental pots, and only 3 that must have migrated to the control pots. From a second set of clear, long plastic containers that held sunflowers and corn specimens, we counted 26 remaining earthworms from the experimental container, and only 4 small babies were found to possibly have hatched in the control container during the experimental period. In the final set of long red plastic containers holding the Spinach and Chard, we found 64 earthworms in the experimental container, close to the original number we started within that pot, and only 2 worms surviving in the matching control container. By simple Chi Square tables, these differentials were significant, demonstrating that enough of a boundary was kept between experimental and control containers to prevent total contamination of the results.

While we must admit having been surprised and somewhat happy that we could in fact grow earthworms for that period of time, under somewhat trying conditions, and maintain an effective separation between the experimental and control containers in spite of the nocturnal movements of many of the worms, we must also admit that as far as the growth of the plant specimens were concerned, there were few clear noticeable differences between experimental and control sets, except possibly for the chard. 

Analysis of the data, and the average of all the data combined, demonstrated no significant difference in growth between experimental and control groups, and, in fact, showed a slightly better performance for the control group compared to the experimental group.

Overall, presence of the earthworms appeared to make no significant difference for most of the plants during the period of measurement and the early phase of development of these plants, given mitigating conditions of the poor sunlight and artificial light. In a few cases, results suggest that the earthworms may even have hindered the rapid development of the plants rather than facilitated this development.

Only in the case of the Swiss Chard did the experimental plants, on average, appear to do better than the control group, but this difference was so small a factor as to be easily overlooked. Overall, sample sizes were so small that the final averages were affected by the inclusion of extremely small plants that had failed to germinate promptly or properly, and hence were out-classed by their neighbors from the start in terms of competition of scarce light resources. These smaller sample plants should have been removed from the final analysis, though in this case it made little overall difference in the outcomes.

Recommendations to carry this experiment further would be to implement more controlled conditions, at a better time of the year (preferably in the spring or summertime), and to carry the plants over the full life cycle of each plant's development, which would entail setting up full garden facilities where earthworm presence/absence could be managed on a larger scale. I would use more plants of each kind, more different kinds of plants, and for longer periods that ran the length of their life cycle. We used far too few plants to yield any robust data-samples. It would have been better as well to have grown each plant in a container of sufficient size that there would be no crowding or competition between plants in the same space, thus eliminating a possible factor affecting the results that may otherwise contaminate such a study.

In four out of five different plant samples, the average growth for the control (no worm) actually did better than the average growth for the experimental (worm). Only in the case of Swiss Chard did the average measures of the experimental perform in the last half of the experiment better than the measures of the control. In all cases, the differences between the growth of the experimental and control plants could not be said to be significantly different in an unequivocal way, though there was a slight difference. In call cases the growth patterns of the experimental plants tracked very closely their correlated control plants

The only kind of conclusion that can be drawn from this experiment is that not only must the presupposition be questioned that earthworms are by definition good for plants in the soil, but the presence and growth of earthworms in constrained contexts might actually compete with the demands of plants in the same contexts, especially during the early phases of the plant's initial growth and early development. While earthworm ecology may be good for soil ecology overall, and earthworm excrement may be nutrient rich, particularly for certain kinds of plants, the fact of the living presence of earthworms may in fact result in inter-specific competition with plants, if not in predation on the root structures that may be the result of the burrowing activities of the worms. I would consider at this stage this kind of question of earthworm-plant ecology to be fundamentally far too complicated to be easily resolved by such a simple study, though I think the results were robust and reliable enough to draw the general conclusions that we did. It is apparent that probably some kinds of plants do better in the presence of earthworms than other kinds of plants, and it is possible that earthworm activities may actually hinder or hurt the developmental growth of certain varieties of plants, even if their excrement may be nutritive for the soil overall.

Certainly, both plants and earthworms are part of a larger, more complicated soil system that includes the activity of bacteria and other macro and micro-fauna that would be invisible to the naked eye. The principle function of earthworms beyond the process of aerating and turning over soil, must be to function as detrivores in recycling more rapidly decaying vegetation. It is unknown if earthworms would prey upon the stems and leaves of young seedlings and new plants. This kind of activity and predation was not obvious in the plants that we observed in this study, on the other hand it should not be completely discounted as a possibility, even if earthworms are mainly sub-surface and plants generally grow above the surface of the ground.

The nocturnal activities of the earthworms, especially the longer "night-crawlers" came as something of a surprise, especially in terms of their apparently deliberate efforts to leave the confines of their containers and to quest for better ground in other places remote from where they started.


Blanket Copyright, Hugh M. Lewis, 2005. Use of this text governed by fair use policy--permission to make copies of this text is granted for purposes of research and non-profit instruction only.

Last Updated: 03/14/05