Due to the similarities between the areas of hyperaccumulation and dynamic accumulation, and the fact that research of the former is further along than research of the latter, a clear pathway has already been established for researchers of dynamic accumulators to follow. A major component of clearly defining hyperaccumulation has been the establishment of nutrient concentration thresholds that plants must meet in order to be considered hyperaccumulators. Plants that have been proven to meet the thresholds are then entered into an easily searchable database. The database assists researchers in identifying promising plants for further study, and real-world applications for these plants can then be developed. The model hasn’t been perfected yet, and hyperaccumulator researchers are still facing some challenges, such as the existence of multiple competing sets of thresholds, several databases with conflicting criteria for inclusion, and additional quality control issues such as the use of “spiked” growing medium or contaminated plant tissue samples giving inflated nutrient readings (Reeves, Baker, Jaffré, Erskine, Echevarria, van der Ent, 2018; Jansen, Broadley, Robbrecht, and Smets, 2002; Rascio and Navari-Izzo, 2011). But the implementation of nutrient thresholds and curated databases of hyperaccumulator species has gone a long way in advancing the study of these plants.

Robert Kourik, popularly believed to have coined the term “dynamic accumulator” in the 1980s, proposed in 2014 that the USDA-hosted “Dr. Duke’s phytochemical and ethnobotanical databases” could be used to compare nutrient values across thousands of plant species to help identify dynamic accumulators (Kourik, 2014). A year later, Dean Brown pointed out the similarities between dynamic accumulators and hyperaccumulators, and suggested that nutrient concentration thresholds should be set for dynamic accumulators, using ppm concentrations of dried plant tissue samples, consistent with hyperaccumulator thresholds (Brown, 2015).

The “Dynamic accumulator database and USDA analysis” online tool was created in 2020 by Unadilla Community Farm, with support from USDA-NIFA through Northeast SARE under subaward number FNE20-967. This database is intended to provide up-to-date information on all available peer-reviewed data on dried plant tissue concentrations of 20 beneficial nutrients, aggregating data from the USDA-hosted “Dr. Duke’s phytochemical and ethnobotanical databases.” Following Brown’s advice, USDA’s “high ppm” values are used as these correspond with dried plant tissue samples, consistent with hyperaccumulator thresholds. This data is used to calculate nutrient value averages across 7,098 entries. Threshold concentrations are then set across 20 nutrients, following the model used for the identification of hyperaccumulators.

Unlike hyperaccumulators, which accumulate toxic heavy metals or metals of interest to the mining industry, dynamic accumulators are defined by their ability to accumulate nutrients that are beneficial to plants. Therefore threshold concentrations are calculated for 20 nutrients that have been shown to be either essential or beneficial for plant health: N, P, K, Al, B, Ca, Cl, Co, Cu, Fe, I, Mg, Mn, Mo, Na, Ni, S, Si, Se, and Zn (Provin and McFarland, 2018; Uchida, 2000; Vatansever, Ozyigit, and Filiz, 2017; Kiferle, Martinelli, Salzano, Gonzali, Beltrami, Salvadori, Hora, Holwerda, Scaloni, and Perata, 2021; Leyva, Sánchez-Rodríguez, Ríos, Rubio-Wilhelmi, Romero, Ruiz, Blasco, 2011). Dynamic accumulator thresholds of roughly 200% of average concentrations result in a total of 340 plant species that have been shown to achieve nutrient concentrations high enough to qualify as dynamic accumulators. On average, dynamic accumulators currently account for 9.59% of plants in each nutrient category. Plant species that meet dynamic accumulator thresholds are presented in the online tool in a detailed list, along with all available nutrient concentration data, to assist in further research.

When outlining the general model for identifying dynamic accumulators, Kourik (2014) describes that researchers should point to nutrient concentration data, in ppm, “as compared to other plants.” In Kourik’s model, the USDA phytochemical and ethnobotanical databases are used to form the data set of “all known plant tissue nutrient concentrations.” Average nutrient values are calculated across this data set, and dynamic accumulators’ nutrient concentrations are compared to these averages. These comparisons can be expressed as a percent of an average nutrient value or, as Brown (2015) proposes, as a biome-concentration factor (Bf). In this case, the “biome” referred to is the group of plants currently included in the USDA database. Both comparisons are made by dividing the plant’s nutrient concentration by the average across the data set. For example, the fruit of the cucumber plant has been shown to contain 80,000 ppm of nitrogen (N). When compared to the N concentrations of 94 plant entries in the USDA databases, cucumbers contain 336% the N of the average plant, or a Bf of 3.36 – more than 3 times the average. Both methods of nutrient concentration comparison are included in the “Dynamic accumulator database and USDA analysis” online tool.

Since the USDA databases receive regular updates as new plant tissue analyses make their way into peer-reviewed journals, the data set relied on for the study of dynamic accumulators is growing. This means that average nutrient values and biome-concentration factors are constantly changing. For example, archived data is available from the USDA databases from 2013, listing cucumber N concentrations as 476% of the average, or a Bf of 4.76. Almost 10 years later, there have been 2,167 new entries added to the database, and the cucumber fruit’s Bf has been revised to 3.36. This illustrates the “dynamic” nature of the USDA databases themselves, and the importance of stable nutrient concentration thresholds to assist in further studies of dynamic accumulators. The online tool presented here should likewise undergo regular updates to ensure it includes the most up-to-date information on plant nutrient concentrations, averages, and biome-concentration factors. The nutrient concentration thresholds used for the classification of dynamic accumulators should also be periodically reviewed and possibly updated to reflect our growing understanding of plants’ nutrient concentrations, as is done in the field of hyperaccumulators.

An additional consideration with the current method of using nutrient concentration thresholds to identify dynamic accumulators is that it does not take into account the effect of soil nutrient concentrations on plant tissue nutrient concentrations. It has been noted in the field of hyperaccumulator research that growing medium that has been “spiked” or amended to artificially raise a particular nutrient’s concentration in the soil can result in heightened plant tissue concentrations (Reeves et al., 2018). For this reason, studies of dynamic accumulators should report plant tissue concentrations along with nutrient concentrations of the growing medium used. With these two data points, bioaccumulation factors (BAFs) can be calculated, by dividing plant tissue concentrations (in ppm) by “background” concentrations in the soil (also in ppm). BAFs are useful for assessing whether plant tissue concentrations are in fact the result of nutrient accumulation (BAF>1), exclusion (BAF<1), or simply an indication of soil nutrient concentrations (BAF=1). BAFs can vary based on a range of factors, such as overall plant health, growing conditions such as soil moisture, or interrelated soil processes (Sharpley, 1997). For this reason, it is only by reporting BAFs for a plant species across a range of growing conditions and growing media that we can better understand how to effectively use dynamic accumulators to achieve tangible benefits for farmers.