Three mature L. terrestris per earthworm-present experiment  were added to each earthworm-present jar, and the experiments left for a further 25 days. Experiments were watered every 2–3 days for the duration of the experiment, and earthworms weighed before and after the experiments.Fluorescein diacetate hydrolysis  assays were carried out immediately following completion of the soil-based experiments, on fresh soil sieved to 2.8 mm using a protocol described by Shaw and Burns . The FDH assay colourimetrically measures the hydrolytic capacity of soil as a proxy for biological activity. Briefly, approximately 1 g soil was equilibrated at room temperature with sterile potassium phosphate buffer  and reacted with 0.1 mL of a 1000 μg/mL solution of fluorescein diacetate in acetone for 30 min. To terminate the reaction, the solution was vortexted with 15 mL of 2: 1 chloroform: methanol , before centrifugal separation of the layers. The top layer containing the residual fluorescein was removed, centrifuged at 16,500 g for 5 min before analysis using a PerkinElmer Lambda 25 UV–vis spectrophotometer at 490 nm. Biological activity was expressed as μg fluorescein/g dry soil/30 min, calculated using external calibration with solutions containing known concentrations of fluorescein.Earthworm-plant experimental data were analysed using SigmaPlot for Windows and Excel 2016. Outliers in the data were identified as being those values which were more than 1.5 times the interquartile range above the third quartile or below the first quartile and were removed prior to further statistical analysis. Plant biomass in the presence and absence of earthworms in both the hydroponic and soil-based experiments, and pH in the hydroponic experiments were normally distributed  and were compared using a ttest. The impacts of plants and earthworms on FDH activity in the soils and phytohormone concentration in the hydroponic experiments were determined using two-way analysis of variance  with the presence or absence of earthworms and plants as the two factors and a p value of ≤0.05 taken to indicate significance differences. Phytohormone concentration data from the hydroponic experiments were not normally distributed and could not be normalised using standard transformations. However, because we wished to determine interaction effects between earthworms and plants on FDH activity and the concentration of phytohormones, and because ANOVA is robust to violations of assumptions of normality and equal variance, we analysed our data using two-way ANOVA followed by Holm Sidak post hoc testing, rather than one-way ANOVA on ranks considering first the impact of earthworms and then the impact of plants.

Although the genome of S. alba is not well studied, comparative genome mapping has shown that much of its genome is conserved in the closely related Brassica species B. napus,stacking flower pot tower with many regions of conserved gene order in the well-studied Arabidopsis thaliana . In both A. thaliana and B. napus, a number of studies have identified genes upregulated in response to stress . Therefore, using the B. napus genome as a guide, DNA primer sequences were designed, and tested to see if these primers could amplify S. alba orthologues in both the plant shoots and roots from the hydroponic experiments. Gene expression studies were conducted using quantitative PCR  to quantify levels of expression from genes in the ABA synthesis pathway , and genes upregulated in response to abiotic stresses including NaCl, drought or cold . Full details of the genes, and primer sequences used, can be found in the Supplementary information. Prior to qPCR, primer sequences were blasted against L. terrestris and E. fetida genomes to ensure sequences were specific to S. alba.FT-ICR-MS analysis was used to survey the range of phytohormones in the hydroponic and soil samples. Typical spectra of Fractions 1 and 3 from soil extracts are shown in Fig. 2A and B; analysis of hydroponic solutions resulted in less complex spectra in which phytohormone signals could be easily identified. For Fraction 1, ABA was detected in most  of the hydroponic and soil samples analysed . IAA was detected in only very few samples, despite it being widely reported in similar analysis of plant tissues. IAA is reported to be unstable at certain light levels, pHs and temperatures ; therefore, to determine whether IAA survived our extraction procedure, a small amount of an authentic IAA standard was spiked onto soil and analysed after extraction. IAA was detected at only very low intensity, suggesting that it cannot be extracted and detected effectively using this procedure, despite the low extraction temperatures. Analysis of Fraction 3 yielded a complex spectrum, attributed to the variety of compounds present within soil. However, the high mass accuracy analysis allowed the identification of several compounds related to plant growth including Ade, iP and Z as well as benzyladenine  and other adenosine-related products . Although iP was often present at very low intensity, it was detected in all samples analysed using FT-ICR-MS. As a result of the data from the untargeted FT-ICR-MS analyses, it was decided to proceed to quantification by MRM of Z, iP, Ade and ABA, using labelled standards of each as an internal standard. Analysis of IAA was also attempted, but in most cases this compound was not detected.Addition of known amounts of isotopically labelled internal standards prior to extraction allowed the quantification of levels of the four target compounds . The amount present in each sample  was determined by comparison of the peak areas of the diagnostic product ion, with those of the relevant internal standards. This method accounts for losses during extraction and chromatography, as losses of the labelled standard are assumed to be equivalent to loss of the target compounds.

Concentrations of the target compounds were then corrected to give values per gram of dry soil or per mL of hydroponic solution.Linearity of response for all compounds was demonstrated by injecting solutions containing both isotopically labelled and authentic standards at concentrations ranging from 0.1 to 100 ng/mL. Linearity was above 0.98 for all compounds  across the concentration range except for IAA; this has been attributed to the potential decay of IAA during the extraction process. Results for IAA have thus been reported only as either detected or undetected . Having demonstrated linearity, relative response factors  were calculated for each target compound from the peak area ratio of labelled standard: target compound in a solution containing a known quantity of each. This provides a correction based on the response expected from the target compound compared to the internal standard. Repeatability was determined by three replicate injections of solutions prepared at 0.01 ng/mL and expressed as the % relative standard deviation . The repeatability varied between compounds, with that for iP being comparatively weak but with all other compounds being under 10%. Limits of detection  are defined as the lowest concentration of a compound that can be reliably detected, and limit of quantification  the lowest concentration at which the compound can be reliably quantified. Whilst several studies have reported the extraction and detection of a range of phytohormones from plants , similar analysis in other matrices, such as soil and other growth media, has been neglected. However, the ability to examine the range and concentrations of phytohormones present in soil could potentially open up routes to understanding the influences of below-ground organisms  and rhizobacteria on plant growth. In this study, it has been demonstrated that a range of phytohormones can be extracted from both soil and hydroponic growth solution and detected using untargeted FT-ICR mass spectrometry. Furthermore, three commonly occurring cytokinins  as well as ABA were then quantified by MRM. Limits of detection ranged between 0.08 ng/g fresh soil for Ade and 2 ng/g fresh soil for iP, with much lower concentrations detectable in the hydroponic solution, as a much larger volume of matrix could be processed for analysis. The LoD for soil could potentially be improved by using larger masses of soil, although this was experimentally limited by the need to rapidly remove, without heating, the larger volumes of solvent that would be used. In addition, larger volumes of soil would have increased animal usage in the experiments. The repeatability of analyses varied, with the data for iP being prone to high levels of error but those for ABA and zeatin showing particularly good repeatability and linearity . There is scope for further refining these methods; previous studies report the detection of phytohormones in plant tissue at the picomole per gram fresh weight level .

The fast analysis time  reported here offers the ability to expand the method to examine additional compounds that may be present in soil, such as gibberellins and other auxins. Although these compounds were not detected by FT-ICR-MS in our experiments, and therefore not targeted for quanification, this may not be the case for others using our approaches. Applications for the ability to detect phytohormones in plant growth media are far reaching, including developing understanding of interactions between soil microbes and plants, and understanding the potential uses of plant growth-promoting bacteria, for example in agriculture. Despite the obvious potential value in understanding these relationships in terms of advancing land management practices to maximise crop growth , little research has been previously conducted in this area.Applying the developed analytical methods, experiments were carried out aimed at investigating the impact of the presence of earthworms with and without plants on the concentrations of phytohormones within the growth media. Despite no measurable differences in plant biomass in the presence of earthworms, a significant increase in the presence of ABA was detected when earthworms were present in hydroponic solution together with plants. This suggests that there could be interactions between the earthworms and plants that cause ABA to be produced. The possibility that the presence of earthworms alters the regulation pathways of certain phytohormone-related genes was tested for by molecular biology methods. A search of the earthworm genome for genes related to ABA production revealed no matches, indicating that earthworms are unlikely to be able to directly produce ABA. We hypothesise instead that the increase in ABA we observed in our earthworm-present experiments in hydroponic solution was caused by indirect influence. Further research would need to be carried out in order to fully assess the mechanisms by which earthworms may be involved in ABA  regulation in plants. An increase in ABA production in the presence of earthworms could be attributed to a range of indirect factors including increased competition for nutrients, or the chemical modification of the solution by earthworms. As ABA is frequently associated with abiotic or biotic stress, this seems the most obvious explanation for its increased presence . Analysis of the pH of the solutions did not reveal significant differences, although this is only a very broad measure of the degree to which the earthworms may have altered the environment. It is also possible that the presence of earthworms induced changes in the expression of genes known to be involved in plant stress responses. For example, in addition to affecting plant roots through burrowing, and physiological activities such as excretions ,vertical grow rack there is some evidence that earthworms also feed on living plant root material .

A small-scale study was therefore conducted to see if genes known to be involved in stress responses, or in the biosynthesis of ABA, were upregulated in either the plant roots or plant shoots grown in hydroponic solution in the presence of earthworms. However, only a few genes were tested and in only one case  was a significant difference  seen between the presence/absence of earthworms. Whilst this may be related to the observed differences in ABA concentrations, the metabolic pathway of ABA production is complex and as such this requires further investigation . In particular, transcriptomic-based studies could be employed to assess the effect of earthworms on global plant metabolic pathways. Substantial evidence exists that earthworms benefit crop yields . However, the observations from our hydroponic experiment suggest that under some circumstances, for example in an already stressed system, earthworms may in fact cause greater levels of stress to plants, resulting in higher levels of ABA in the soil. To our knowledge, no study has effectively dismantled the effects of earthworms in systems with limited nutrient availability.