Abandoned Wood Ant Nests as Sites for Seedling Germination

Abstract

We compared the species composition of seedlings germinated on ten recently abandoned Formica polyctena nests with the plant species on active nests and the plant species composition of the forest floor around the nests. Compared to no plants at all, when the nests were inhabited, seedlings of 12 different vascular plant species germinated on the nests after their abandonment. Four of them were myrmecochorous, of which two occurred only on the nests. After abandonment, ant activity ceases, the water content of the nests increases, and decomposition of the organic nest material is accelerated, resulting in faster nutrient mineralization. Consequently, more nutrients are available for plants. Our results suggest that abandoned nests can serve as “regeneration niches”, ensuring genetic diversity, especially in plant populations which rely mainly on vegetative propagation. However, more research on a larger scale is needed to assess the role of wood ants and their abandoned nests in forest vegetation development.

1. Introduction

Wood ants, often described as ecosystem engineers and keystone species in forest ecosystems, are one of the most important insect groups, mainly due to their large pop-ulation sizes, their activity, and direct or indirect effects on other insects, plant growth and abundance, and carbon nutrient fluxes [1]. Their nests accumulate more carbon and nutrients (nitrogen, calcium, phosphorus, magnesium, potassium, sodium) than the surrounding soil [2]. Wood ants are described as keystone species because of their role in predator–prey dynamics, seed dispersal, affecting herb and tree growth, and physical and chemical properties of forest soils [1,3,4,5].

The ant Formica polyctena (Foerster, 1850) is one of the five wood ant species belonging to the Formica rufa-group (Formica s. str.) ants occurring in Poland. Their nests are large, aboveground mounds, composed mainly of organic material, such as small twigs, pieces of bark, needles, grass stems, as well as resin particles, collected in the surroundings. The belowground part of the nests consists of mineral soil mixed with organic matter, and the lowest gallery of the mound is formed by tunnels and chambers running deep into the ground [1,6,7]. As F. polyctena is an obligate polygynous species, dispersing mainly by colony splitting, it establishes large networks of connected nests [7].

Myrmecochory, the dispersal of seeds by ants, is a common and well-studied process [5,8], and at least 11,000 plant species spreading mainly through myrmecochory have been identified globally [9]. Most myrmecochoric plant species produce seeds with lipid-rich appendages, so-called elaiosomes. After the seeds are transferred to the ant nest, the elaiosomes are removed and preferentially fed to ant larvae [9,10,11,12,13]. Seeds without elaiosomes are discarded in the abandoned breeding chambers inside the nest or are removed from the nest to refuse piles, where other organic waste from the nest is also stored [5,14]. Some seeds without elaiosomes are also attractive for foraging ants. These seeds are coated with substances, which, through chemical mimicry, induce a grabbing and carrying response in ants and are effectively moved to the nest [11]. Studies assessing myrmecochory focus mainly on observations of active nests and investigate the mechanism of seed dispersal, the role of particular ant species in the dispersal of specific plant species, the efficiency of seed dispersal, or the influence of ant nests on the distribution of plants [5,11,15]. Abandoned wood ant nests have been investigated for their effect on soil because of accumulated organic matter [16,17,18], and their role in vegetation development [19,20,21,22,23].

The aim of this case study was to compare the species composition of seedlings germinated on ten recently abandoned Formica polyctena nests with the plant species composition of the forest floor surrounding the nests.

2. Materials and Methods

The research was carried out in the Stoczek Forest District, a managed forest under the administration of Polish State Forests in the Białowieża Forest Division (12,593 ha), north-eastern Poland, bordering Belarus and covering 1476 ha. We relied on the results of a wood ant inventory conducted in the Stoczek Forest District in 2016 [24], and the monitoring of the nests conducted in 2016–2020 to select ten Formica polyctena nests for research [25]. These nests were active in 2016–2019 and were abandoned before summer 2020. Five nests were located in a fresh mixed deciduous forest and the other five in a fresh mixed coniferous forest. The ant nest size distribution between forest types was similar, and their average volumes did not differ (df = 4.237, p = 0.0836). All the nests were monitored constantly as part of our wood ant studies in the Stoczek Forest District since 2016.

During 2016–2019, when the nests were still active, they were monitored several times for potential plants germinating on the nest surface. In addition, in early summer of 2020, just after abandonment, we identified all species of vascular plants that had germinated on them. We compared the species composition of seedlings germinated on abandoned Formica polyctena nests with the plant species composition of the forest floor around the nests. Data on forest floor vegetation were collected in 2019–2020 during fieldwork related to the monitoring of Formica polyctena nests. At that time, circular study plots were established, with a 20-m radius from the centre of the nest, and floristic inventories of forest floor vegetation using the Londo [26] quantitative scale, within the plots were conducted. The Londo scale is a decimal scale for recording coverage in vegetation analysis [26].

As the Shapiro–Wilk test indicated that the data were not normally distributed (W = 0.742, p < 0.001), we used the Kruskal–Wallis rank sum test for determining whether the number of plant species differed between the sites (active nests vs. abandoned nests vs. forest floor). For a more detailed multiple comparison, we used the Dunn test with a Bonferroni correction (p < 0.05).

The floristic similarity for seedlings germinated on abandoned nests with the plant species composition of the forest floor around the nests was assessed using the Jaccard similarity index (J) [27]. It was calculated separately for all five nests and forest floor in a fresh mixed deciduous forest together and for all five nests and forest floor in a fresh mixed coniferous forest, using the following formula:

$$J\left(A,B\right)=\frac{\left|A\cap B\right|}{\left|A\cup B\right|}=\frac{\left|A\cap B\right|}{\left|A\right|+\left|B\right|-\left|A\cap B\right|}$$

The higher the percentage, the more similar the populations are.

We used the presence or absence of data of the plant species for an indicator species analysis (ISA), where we compared the nests with the surrounding forest floor, using the ‘multipatt’ command in the ‘indicspecies’ package in R [28]. ISA integrates relative species abundance and the frequency of occurrence to produce a maximum indicator value for each species. Each species is then assigned to the one group for which it has the greatest indicator value. Indicator values range from 0 to 100 (or 0 to 1 depending on the scaling approach). A value of 100 represents a perfect indicator species, i.e., a species that occurs exclusively in a group, and has a high relative abundance within that group [28].

All statistical testing was conducted in R 4.1.1 [29].

3. Results

On all nests in both forest types, which were still active in 2016–2019, no seedlings at all were found during monitoring (Figure 1). In total, we found 70 plant species on the forest floor surrounding the nests in the fresh mixed deciduous forest and 66 plant species in the fresh mixed coniferous forest. Seedlings of 12 different vascular plant species germinated on the abandoned nests (twelve species in the fresh mixed deciduous and six species in the fresh mixed coniferous forest,

Table 1. Plant species found on abandoned wood ant (Formica polyctena) nests and in their surroundings on plots of 20-m radius in fresh mixed deciduous and fresh mixed coniferous forests in the Stoczek subdistrict. Species present are marked with x, and obligately myrmecochorous species with an asterisk [30,31].

	Fresh Mixed Deciduous Forest		Fresh Mixed Coniferous Forest
Plant Species	Nests	Surrounding	Nests	Surrounding
* Anemone nemorosa L.	x	x
* Galeobdolon luteum Huds.	x	x	x	x
* Melampyrum pratense L.	x
* Moehringia trinervia L.	x
Convallaria majalis L.	x	x	x	x
Cardamine bulbifera (L.) Crantz	x	x
Maianthemum bifolium (L.) F.W. Schmidt	x	x	x	x
Mycelis muralis (L.) Dumort.	x	x
Oxalis acetosella L.	x	x	x	x
Rubus idaeus L.	x	x	x	x
Urtica dioica L.	x	x	x	x
Viola sp.	x	x

). In both deciduous and coniferous mixed forests, significantly less plant species were growing on abandoned nests than on the forest floor surrounding them (χ² = 25.115, df = 2, p < 0.001, Figure 2). No differences were observed in the number of plant species between active and abandoned nests (p = 0.107), whereas more species occurred on the forest floor compared to the abandoned (p < 0.05) and active nests (p < 0.001).

Four of the species on the nests were obligate myrmecochorous [30,31], while the remaining plant species had different seed dispersal strategies (Table 1). Seedlings of Melampyrum pratense and Moehringia trinervia, both myrmecochorous species, were found only on the abandoned nests and never on the forest floor of fresh mixed deciduous forest (Table 1), whereas other myrmecochorous species, Anemone nemorosa was found on the nests and the forest floor (Table 1).

The Jaccard similarity index of plant species growing on abandoned nests and on the forest floor surrounding nests was 0.18% in mixed deciduous forest and 0.22% in mixed coniferous forest.

In total, 41 plant species were found to have an indicator value in the deciduous mixed forest, 24 species in the coniferous mixed forest, and 17 species were indicator species for both types of forests (see Table S1 available as Supplementary Data). When considering plant species occurring only in the respective sites (indicator value 100%), Melampyrum pratense was an indicator species for abandoned nests in deciduous forests. Galeobdolon luteum, Cardamine bulbifera, and Urtica dioica were indicator species for deciduous forests (of which C. bulbifera exclusively there) and were found in the forest floor and on abandoned nests of deciduous forest plots. Oxalis acetosella, Maianthemum bifolium, Convallaria majalis, and Rubus idaeus were indicator species for all forest floors (of which C. majalis exclusively there), but only for the abandoned nests in coniferous forests (

Table 2. Results of indicator species analysis (ISA) for forest floor around active nests vs. forest floor around abandoned nests. The higher the tabulated value (%), the more the species in question is restricted to the respective group. A value of 100 indicates that the species occurs only in sites belonging to this group. Statistical significances: *** < 0.001, ** < 0.01 and * < 0.05). Myrmecochorous species are marked with +.

	Forest Floor				Nests
	Around Active Nests		Around Abandonded Nests		Active Nests		Abandoned Nests
Species	Deciduous	Coniferous	Deciduous	Coniferous	Deciduous	Coniferous	Deciduous	Coniferous
Melampyrum pratense L. +							100 ***
Galeobdolon luteum Huds. emend. Holub +	85 **		85 **				85 **
Cardamine bulbifera L.	100 **		100 **				100 **
Urtica dioica L.	77 *		77 *				77 *
Oxalis acetosella L.	87 *	87 *	87 *					87 *
Maianthemum bifolium (L.) F.W. Schmidt	92 ***	92 ***	92 ***	92 ***				92 ***
Convallaria majalis L.	100 **	100 **	100 **	100 **				100 **
Rubus idaeus L.	94 **	94 **	94 **	94 **				94 **

). Stellaria holostea, Galium odoratum, Melica nutans, Melittis melissophyllum, Athyrium filix-femina, Geum urbanum, and Phyteuma spicatum were indicator species only for deciduous forest. Indicator species for coniferous forest plots only were Molinia caerulea, Vaccinium vitis-idaea, Calluna vulgaris, and Epilobium angustifolium (see Table S1 available as Supplementary Data). Three of myrmecochorous species had maximum indicator value: Melampyrum pratense for abandoned nests in deciduous forest, Melica nutans for deciduous forest, and Luzula pilosa for deciduous forest and coniferous forest (Table S1 and Supplementary Data).

4. Discussion

In early spring, seedlings can be found on active nests of Formica polyctena and on their margins, as the high concentration of nutrients in the substrate promotes the rapid growth of seedlings on the nests [5]. However, the proximity of the active nest has a negative effect on the growing seedlings, because with the increased activity of the ants, the plants are damaged by workers, and consequently, they die before reaching the generative phase [5]. Our preliminary study revealed the presence of vascular plant seedlings on the nests in the first year after the nest was abandoned by Formica polyctena. However, we did not observe any plants on these nests when they were still inhabited by ants. The negative impact of ants no longer exists on the abandoned nests, and the emerging seedlings have a chance for further growth and reaching the generative phase. The substrate of active nests is more fertile than the soil surrounding them. Nests are rich in organic matter and mineral nutrients, such as Ca, K, Mg, Na, and P [4]. In addition, in active nests the humidity is particularly low and regulated by the outer layer of the nest [32]. However, in abandoned nests, mechanisms regulating humidity cease to exist, water content of nest material increases, the decomposition of nests is accelerated, and the nutrients are mineralized and available for plants [17]. These conditions promote the germination of seeds, as demonstrated in our study. Abandoned nests in dense ground vegetation can serve as ‘regeneration niches’, ensuring genetic diversity, especially in the populations of plants which rely mainly on vegetative propagation.

The indicator species analysis (ISA) showed that Galeobdolon luteum, Cardamine bulbifera, and Urtica dioica were indicator species for deciduous forests (of which C. bulbifera exclusively there) and were found in the forest floor and on abandoned nests. This seems plausible, since the respective plant seeds are presumably transported from the surrounding forest floor, either by ants or by other means. In contrast, Melampyrym pratense occurred only on abandoned nests in deciduous mixed forests, and not even in the surrounding forest floor. This could indicate that seeds of this species are dispersed for even longer distances. Two other of myrmecochorous plant species: Melica nutans and Luzula pilosa were not found on abandoned nests in deciduous forest and coniferous forest, but there were found on the forest floor in study area. Oxalis acetosella, Maianthemum bifolium, Convallaria majalis, and Rubus idaeus were indicator species for all forest floors, but for some reason, only for the nests in coniferous forests. One reason could be that our observations were conducted just after abandonment, and that a longer time period after abandonment could be needed to reliably adduce the ant effect on plant species composition.

Our observations showed that fewer plant species germinated on abandoned nests than on the surrounding forest floor. The lower number of plant species on abandoned wood ant nests suggests that ants collect and transport only seeds from certain plants into their nests, which is in accordance with results of earlier studies [5,10,13]. Another reason might be that these nests were abandoned just very recently, and with more time elapsing, also more plant species would occur on the nests.

The Jaccard floristic similarity index, assessing the similarity of plant composition on abandoned nests and the surrounding forest floor, using a scale from 0 to 100%, was very low in both forest types, suggesting their species composition was very different.

Species of seedlings growing on the investigated nests represented both obligate myrmecochorous plants (producing seeds with elaiosomes), and those with other strategies for seed dispersal. Non-myrmecochorous seeds without elaiosomes supposedly were placed in the nests by ants [11], but they could be dispersed to that location also via other means, for example, by gravity or wind. Most plant species were also present on the forest floor surrounding the nests, which may indicate that their seeds were moved to the nest by ants. Usually, ants transport diaspores to their nest from a distance of several meters [33]. In the fresh mixed deciduous forest, two plant species whose seedlings were present on the nest were not recorded on the forest floor within a 20-m radius from the centre of the nest. Probably, the seeds of Melampyrum pratense and Moehringia trinervia with elaiosomes were so valuable for ants that they were transported by them from a greater distance, and thus, ants dispersing seeds at longer distances enable plants to colonize new habitats. Ants can increase the regeneration success of myrmecochores by dispersal for distance and placement in a larger spectrum of microsites in contrast to species not adapted for myrmecochory [34]. However, further research is needed on a larger scale to assess the role of wood ants and their abandoned nests in vegetation development.

5. Conclusions

We can conclude that, when compared to active wood ant nests, the number of vascular plants increases immediately after nest abandonment, due to successful germination. Seedlings emerging on abandoned nests likely reach the generative phase and, thus, abandoned ant nests become ecological niches for these plant species. Growing species on the studied abandoned Formica polyctena nests represented both obligate myrmecochorous plants and plants with other strategies for seed dispersal. The seeds deposited in the nests can also potentially play a decisive role in maintaining floristic biodiversity (structure and composition) in the forest. Our study was examining overall plant germination, and not only germination from ant-stored seeds. However, the results could also show an interaction of ant-modified soils benefitting germination of particular plant species aside from ant dispersal. The potential role of wood ants in the plant understory composition can significantly affect the forest-floor plant composition, for example, after natural disturbances in forests. More studies into the plant species compositions of abandoned ant nests and the surrounding forest floor are needed to assess the role of abandoned wood ant nests for vegetation development in the forest ecosystem.