*3.4. Design Experiments*

### 3.4.1. Digital Hive Morphologies

Based on the studies describing the living conditions of honeybees in feral nests [15], we identified our geometrical benchmarks for the overall structure such as the inner nest shape and volume and entrance opening size. Figure 8 demonstrates a design study for generating stand-alone hive morphologies. Load-wise, the inner nest should accommodate approximately 60,000 honeybee workers plus one queen, including their colony's honey, the wax, and other nest materials, weighing as much as 80 kg in total. This cavity should be well insulated to avoid temperature fluctuations within the nest space. For providing the fungal mycelia's scaffold structure, a high porosity for oxygen distribution and structural stability—especially during growth by degradation—are necessary. In consequence, the honeybees' nesting volume should be surrounded by a highly thick and voluminous enclosure of mycelium composite to provide sufficient insulation and mechanical stability.

One of the lessons we learned by exposing bees to mycelium-based composites in the first field experiment was that they removed the soft parts of the material. Therefore, our scaffold serves as a barrier between the bees' nest and the dense yet soft mycelial zones, while allowing hyphal reach. We use honeybees' natural tree nests as a reference for the inner nest geometry of our one-to-one size mycelial hives (further described in the next section). Figure 9 shows two morphologies created using the IS method, which only takes a basic load and support conditions into account in the topology optimisation part and is then edited to have open (Figure 9a) and enclosed (Figure 9b) bee habitation spaces. We show the 1:5 prototypes because this design method development is directly related to the following 3D printing and self-feed procedures.

*3.4. Design Experiments* 

chanical stability.

3.4.1. Digital Hive Morphologies

some colonies initiated their nests there, instead of the targeted mycelium and sensor-attached area. The temperature and relative humidity measurements did not provide a clear distinction between the climate within the mycelium retrofit hives and fully wooden hives, potentially because the nesting spots were different in each hive. Only one of the six hives survived until the following summer season. This survivor hive was one with mycelium attached. We think that the reason is that we populated the hives late in the season, not giving bees enough time to reproduce, collect pollen and nectar, and build wax combs to store enough honey. Additionally, the weather conditions were particularly challenging that year, with heavy rainfall leading to floodings in many adjacent buildings. The colonies and their wax combs were so small that there was too much empty space in the hives before the winter and not enough insulation on the walls.

Based on the studies describing the living conditions of honeybees in feral nests [15], we identified our geometrical benchmarks for the overall structure such as the inner nest shape and volume and entrance opening size. Figure 8 demonstrates a design study for generating stand-alone hive morphologies. Load-wise, the inner nest should accommodate approximately 60,000 honeybee workers plus one queen, including their colony's honey, the wax, and other nest materials, weighing as much as 80 kg in total. This cavity should be well insulated to avoid temperature fluctuations within the nest space. For providing the fungal mycelia's scaffold structure, a high porosity for oxygen distribution and structural stability—especially during growth by degradation—are necessary. In consequence, the honeybees' nesting volume should be surrounded by a highly thick and voluminous enclosure of mycelium composite to provide sufficient insulation and me-

hives.

**Figure 9.** Morphologies created using IS method. (**a**) Open structure for a swarming honeybee colony to nest and 1:5 prototype 3D printed with a wood-infill filament. (**b**) Enclosed structure for a feral honeybee colony to nest and 1:5 prototype 3D printed with a wood-infill filament. brid Superorganisms Diversify Urban Ecological Niches" (Figure 10). **Figure 9.** Morphologies created using IS method. (**a**) Open structure for a swarming honeybee colony to nest and 1:5 prototype 3D printed with a wood-infill filament. (**b**) Enclosed structure for a feral honeybee colony to nest and 1:5 prototype 3D printed with a wood-infill filament.

Here we describe the full-scale mycelium-based hives produced by using the VO method. From summer 2019 to summer 2021, we produced three types of mycelium-based

approx. 5 kg of filament. We soaked the printed parts in the purified water for 10 h each, refreshing the water every 2–3 h. This was a delicate and laborious process which also caused the thin extrusions to lose their stability, weakening the layer adhesion, and yielding rickety parts with decreased structural scaffolding capacity. After the removal of the PVA, we exposed the parts to ultraviolet light on a clean bench for 12 h (to kill the contaminating microorganisms). After 6 h of drying time on a clean bench, in separate clean plastic boxes, we filled the vertical channels of the three hive parts with grain seeds and beechwood dust inoculated with the PO mycelium. After three weeks of growth at room temperature (23 °C), these modules were taken out of their plastic boxes and left to dry in a kitchen oven for six hours. We moved the largest fresh and growing part to the open exhibition space of the Festival Headquarters, at the Vienna Design Week 2019. We let the hive fruit by exposing it to more oxygen, light and water as part of our exhibition "Biohy-

3.4.2. Fully Grown Mycelial Hives

### 3.4.2. Fully Grown Mycelial Hives

Here we describe the full-scale mycelium-based hives produced by using the VO method. From summer 2019 to summer 2021, we produced three types of myceliumbased hives.

In the summer of 2019, our GrowLay Hive-1 was printed in three parts in 32.5 h using approx. 5 kg of filament. We soaked the printed parts in the purified water for 10 h each, refreshing the water every 2–3 h. This was a delicate and laborious process which also caused the thin extrusions to lose their stability, weakening the layer adhesion, and yielding rickety parts with decreased structural scaffolding capacity. After the removal of the PVA, we exposed the parts to ultraviolet light on a clean bench for 12 h (to kill the contaminating microorganisms). After 6 h of drying time on a clean bench, in separate clean plastic boxes, we filled the vertical channels of the three hive parts with grain seeds and beechwood dust inoculated with the PO mycelium. After three weeks of growth at room temperature (23 ◦C), these modules were taken out of their plastic boxes and left to dry in a kitchen oven for six hours. We moved the largest fresh and growing part to the open exhibition space of the Festival Headquarters, at the Vienna Design Week 2019. We let the hive fruit by exposing it to more oxygen, light and water as part of our exhibition "Biohybrid Superorganisms Diversify Urban Ecological Niches" (Figure 10). *Biomimetics* **2022**, *7*, x FOR PEER REVIEW 14 of 21

**Figure 10.** GrowLay Hive-1 (**a**) 3D printed scaffold before the PVA removal. (**b**) A close up picture taken two years after: the powder like mycelium remains on the surface. (**c**) Vienna Design Week, 2019: mycelium kept degrading the hive scaffold and as we misted it, the fruit bodies emerge. (**d**) A picture taken right after the mycelium hive part is removed from its box and placed in the kitchen oven. **Figure 10.** GrowLay Hive-1 (**a**) 3D printed scaffold before the PVA removal. (**b**) A close up picture taken two years after: the powder like mycelium remains on the surface. (**c**) Vienna Design Week, 2019: mycelium kept degrading the hive scaffold and as we misted it, the fruit bodies emerge. (**d**) A picture taken right after the mycelium hive part is removed from its box and placed in the kitchen oven.

Within the GrowLay Hive-2—which had a slightly larger, 45 L inner nest volume

compared to the GrowLay Hive-1—we aimed to introduce a honeybee colony in the spring of 2020 (Figure 11a). Therefore, the hive's stability and durability became priorities. Instead of increasing the structural stability via geometry and density of the printed structure, we kept the PVA, the water-soluble polymer, to support the scaffold with better particle adhesion. For the ease of mycelium infill and handling, we printed this hive in six smaller parts compared to the GrowLay Hive-1, in 45 h using approx. 6 kg of filament. The first prototype of GrowLay Hive-2 was disposed of due to *Trichoderma* fungus contamination. The next one could be produced only in mid-July 2020 which is almost the mid of the bee season. This time, the TV mycelium was grown in the vertical channel with birchwood chips. The inoculation process was in a laboratory but not on a clean bench. After 4 weeks of colonisation, beginning of September 2020, we moved the hive under a small wooden protective shelter into an outdoor setting at the HIVEOPOLIS Honeybee Within the GrowLay Hive-2—which had a slightly larger, 45 L inner nest volume compared to the GrowLay Hive-1—we aimed to introduce a honeybee colony in the spring of 2020 (Figure 11a). Therefore, the hive's stability and durability became priorities. Instead of increasing the structural stability via geometry and density of the printed structure, we kept the PVA, the water-soluble polymer, to support the scaffold with better particle adhesion. For the ease of mycelium infill and handling, we printed this hive in six smaller parts compared to the GrowLay Hive-1, in 45 h using approx. 6 kg of filament. The first prototype of GrowLay Hive-2 was disposed of due to *Trichoderma* fungus contamination. The next one could be produced only in mid-July 2020 which is almost the mid of the bee season. This time, the TV mycelium was grown in the vertical channel with birchwood chips. The inoculation process was in a laboratory but not on a clean bench.

Research Field Laboratory. However, it was too late in the season for a honeybee colony to start building their natural comb structures which require a lot of their energy. To avoid

(Figure 11b).

After 4 weeks of colonisation, beginning of September 2020, we moved the hive under a small wooden protective shelter into an outdoor setting at the HIVEOPOLIS Honeybee Research Field Laboratory. However, it was too late in the season for a honeybee colony to start building their natural comb structures which require a lot of their energy. To avoid risking a foreseeable winter death of such a late-established colony, we left the hive empty (Figure 11b). *Biomimetics* **2022**, *7*, x FOR PEER REVIEW 15 of 21

**Figure 11.** GrowLay Hive-2. (**a**) An exploded axonometric drawing of the whole construction, showing details of the fungal parts. (**b**) Outdoor assembly during late summer 2020. **Figure 11.** GrowLay Hive-2. (**a**) An exploded axonometric drawing of the whole construction, showing details of the fungal parts. (**b**) Outdoor assembly during late summer 2020.

The most recent hive scaffold was printed with clay. The overall form of the digital model was built using the VO method and after determining the toolpath drawing parameters. To achieve the required honeybee inhabitation volume within fabrication constraints (max. printing diameter = 40 cm and heights of each module kept to a maximum of 15 cm to avoid layer collapse), the walls of the mid-body parts had to be thinner, making the inoculation—filling the vertical gaps with mycelium inoculated flax fibres—more difficult, if not impossible in those areas. Furthermore, because we initially added water to adjust the viscosity, the printed clay shrank by nearly 20% during the drying and firing processes, as a result, the honeybee inhabitation volume decreased from 39 L to an average of 30 L. The more we expand the inner nest volume, the thicker the mycelial wall should be to maintain the thermal stability within the hive. This would require either dividing the ring-like modules into printable sizes across their cross sections or using a The most recent hive scaffold was printed with clay. The overall form of the digital model was built using the VO method and after determining the toolpath drawing parameters. To achieve the required honeybee inhabitation volume within fabrication constraints (max. printing diameter = 40 cm and heights of each module kept to a maximum of 15 cm to avoid layer collapse), the walls of the mid-body parts had to be thinner, making the inoculation—filling the vertical gaps with mycelium inoculated flax fibres—more difficult, if not impossible in those areas. Furthermore, because we initially added water to adjust the viscosity, the printed clay shrank by nearly 20% during the drying and firing processes, as a result, the honeybee inhabitation volume decreased from 39 L to an average of 30 L. The more we expand the inner nest volume, the thicker the mycelial wall should be to maintain the thermal stability within the hive. This would require either dividing the ring-like modules into printable sizes across their cross sections or using a larger 3D printer. Additionally, this would result in longer toolpaths, therefore a larger surface area through which the clay would lose water and higher shrinkage rates.

larger 3D printer. Additionally, this would result in longer toolpaths, therefore a larger surface area through which the clay would lose water and higher shrinkage rates. Our Mycelial Clay Hive is made up of 13 ring modules and took 15 h to print. Before Our Mycelial Clay Hive is made up of 13 ring modules and took 15 h to print. Before printing the clay scaffold, a commercially available stoneware paper clay was hand mixed with 15% of its weight with tap water to reach a suitable printing viscosity. When compared

printing the clay scaffold, a commercially available stoneware paper clay was hand mixed with 15% of its weight with tap water to reach a suitable printing viscosity. When com-

subsequent single-person handling. To avoid cracks, the units were air dried for two days after 3D printing, loosely covered with plastic sheets. They were then fired at 1200 °C. We filled the voids in the modules with pregrown mycelium spawn and ground particles of entire flax plants that had been inoculated with TV's mycelium. The inoculation procedure was carried out as quickly and cleanly as possible without sterile conditions. They were incubated in a 23 °C ambient room temperature for the first ten days. During this time, we used a heat mat connected to a temperature sensor and controller, alternating

to the previous GrowLay Hive modules, the clay ones had a significantly higher weight, which was beneficial for overall stability, but it had to be kept to a minimum for subsequent single-person handling. To avoid cracks, the units were air dried for two days after 3D printing, loosely covered with plastic sheets. They were then fired at 1200 ◦C. We filled the voids in the modules with pregrown mycelium spawn and ground particles of entire flax plants that had been inoculated with TV's mycelium. The inoculation procedure was carried out as quickly and cleanly as possible without sterile conditions. They were incubated in a 23 ◦C ambient room temperature for the first ten days. During this time, we used a heat mat connected to a temperature sensor and controller, alternating the plastic boxes. During this time, one module that had cracked during transportation was stabilised along with its constituent mycelium. *Biomimetics* **2022**, *7*, x FOR PEER REVIEW 16 of 21 the plastic boxes. During this time, one module that had cracked during transportation was stabilised along with its constituent mycelium.

The main challenge we had in making the first two hives was that the aeration of the mycelium inoculate within the scaffold channels was not sufficient. This created a fast formation of the thick mycelial skin on the surfaces, which were exposed to air while leaving the inner areas of the walls only marginally colonised. When hydrophobia and protection from intruding animals are required, this thick leatherlike differentiation of mycelium can be beneficial. However, when it occurs at the intersecting surfaces of modules that are stacked on top of each other, it inhibits the further hyphal growth for biowelding separate modules together. Another problem was the deformation of the module geometries during the incubation period with high moisture levels. This resulted in an ill-defined continuity in the overall hive geometry. We observed that PVA drips out, creating a strong chemical border that blocks the mycelium from penetrating into the extruded scaffold material. So, if GrowLay™ was a decided material, removing the PVA and improving the structural integrity of the hive via the scaffold design variables would have been a better solution. To compensate for these losses of deformation and stability, we used bamboo sticks as an inner reinforcement when we installed the hive in the garden. The hive was disassembled in Autumn 2021, and we observed that several other animals—such as snails, spiders, and soil insects—already occupied the hive. In the clay scaffolds, the mycelium infill grew faster and more uniformly than in previous GrowLay Hives due to the increased porosity created by the toolpath and the micropores formed after firing the fibres out of the modules. Mycelium's robust and uniform growth allowed for the biowelding of modules with a large enough surface area in contact (Figure 12). The main challenge we had in making the first two hives was that the aeration of the mycelium inoculate within the scaffold channels was not sufficient. This created a fast formation of the thick mycelial skin on the surfaces, which were exposed to air while leaving the inner areas of the walls only marginally colonised. When hydrophobia and protection from intruding animals are required, this thick leatherlike differentiation of mycelium can be beneficial. However, when it occurs at the intersecting surfaces of modules that are stacked on top of each other, it inhibits the further hyphal growth for biowelding separate modules together. Another problem was the deformation of the module geometries during the incubation period with high moisture levels. This resulted in an ill-defined continuity in the overall hive geometry. We observed that PVA drips out, creating a strong chemical border that blocks the mycelium from penetrating into the extruded scaffold material. So, if GrowLay™ was a decided material, removing the PVA and improving the structural integrity of the hive via the scaffold design variables would have been a better solution. To compensate for these losses of deformation and stability, we used bamboo sticks as an inner reinforcement when we installed the hive in the garden. The hive was disassembled in Autumn 2021, and we observed that several other animals—such as snails, spiders, and soil insects—already occupied the hive. In the clay scaffolds, the mycelium infill grew faster and more uniformly than in previous GrowLay Hives due to the increased porosity created by the toolpath and the micropores formed after firing the fibres out of the modules. Mycelium's robust and uniform growth allowed for the biowelding of modules with a large enough surface area in contact (Figure 12).

To investigate the animals' (including humans') dependency on diverse microbial

organism and fungal architectures as a biodiversity maximising strategy. In this paper, we showed our hybrid construction method and a design framework for merging two complex and dynamic material systems—the honeybee superorganism and mycelial networks—in order to reassemble a potentially lost link between the social insects' wellbeing and bioactive inner nest spaces. We report our findings and insights on the following top-

**Figure 12.** Mycelial Clay Hive. ics in this paper. **Figure 12.** Mycelial Clay Hive.

**4. Discussion** 

### **4. Discussion**

To investigate the animals' (including humans') dependency on diverse microbial communities in their habitats (our built environment), we use the honeybee as a model organism and fungal architectures as a biodiversity maximising strategy. In this paper, we showed our hybrid construction method and a design framework for merging two complex and dynamic material systems—the honeybee superorganism and mycelial networks in order to reassemble a potentially lost link between the social insects' wellbeing and bioactive inner nest spaces. We report our findings and insights on the following topics in this paper.

The architecture of honeybee nest enclosures and the architecture of their comb construction differ. Nevertheless, mycelial architecture is comparable to both. The honeybee colony builds a custom comb structure within established nest enclosures. These enclosures function as heat and light barriers, and bees themselves engage as material constituents of this dynamic multi-material system. Thus, the nest's material components are the bees' products and also their bodies. Mycelial material systems are similar to honeybee nests in that they are adaptive material systems that are entangled with their surroundings and actively manage their local environment, such as chemical and microbiological conditions, in order to survive. When placed in predetermined nest enclosure formworks or scaffolds, the microscale hyphae span the entire geometry. Like honeybees, mycelium largely remains within the physical domains of its prepared form while being able to adjust its behaviour dynamically. In addition, as mentioned earlier, wood-rotting fungi are already present in the natural tree hollows where robust wild bee colonies live, and mycelium architecture is already present in detectable levels in honeybee habitats and their nearby ecosystems.

We argued that the mycelia have been part of the social immunity of the honeybees while co-occupying the tree cavity nests throughout evolutionary history. However, the concentration of mycelia in the nest enclosure materials is significantly higher in the fungal hives we presented. This could imply that the natural balance of fungal metabolites, honeybee symbionts, and honeybees may not be established in fully grown mycelial hives. Our goal is not to reconstruct a tree cavity, yet we aim to seed sentient material systems in which mycelium takes part and renders it specifically receptive to microbial communities that have coevolved with the honeybee species over millions of years. Nothing exists in isolation in nature, especially in a honeybee hive. The adaptive mechanisms found in living materials arise due to their form, multiresolution microstructures, causing them to behave in nonlinear ways, responding to external stimuli in unpredictable ways. In the targeted microbiology coupling experiments, we isolate organisms and organic materials from the environments in which they occur in harmony with their complex surroundings. This entails increasing the size of the specimens from a Petri-dish scale to a fully functional on-site beehive structure, which alters the length and time scales in which mycelial fungi might act as a symbiotic agent for honeybee and vice versa. For example, we do not know if the antibacterial capacity of mycelia can inhibit the dispersal of beneficial bacteria in the honeybee nest microbiome. To effectively map the beneficial bioactivities of mycelia via designing its scaffolds, various compositions with various fungal species, growth substrates, and morphologies should be rigorously tested in both laboratory and field conditions. To grow many identical repetitions of mycelial parts and full-scale beehive prototypes for such experiments, fungal biofabrication processes should be improved to be more efficient in terms of human labour. However, even though some properties, such as surface qualities, bioactive agents, and densities may vary on different sizes and timescales, the primary goal of providing a good thermal environment for bees must be met. Therefore, the thermal dynamics of this coupled system in different hive morphologies, including heat distribution and water balance, are of particular interest to us.

We proposed two digital methods for generating nest morphologies. Devised specifically for fused deposition manufacturing, these methods aim to establish bespoke stay-in scaffolds for living mycelia and also for other biodesign narratives. As opposed to the commonly used casting technique where the nutrition substrate and mycelium culture are filled in a mould manually, by using fused deposition techniques to lay the mould as a stay-in scaffold with internal structure, we can radically increase the surface area within a given volume and encourage a more uniform distribution of hyphae within the composite material system. We continue to improve our digital design techniques in order to produce more versatile and efficient design models that facilitate the iterative and exploratory nature of designing with and for other living organisms. We further improve our digital design skills in order to create more flexible and efficient design models that support the iterative and experimental nature of designing with and for other living organisms.

It is an intriguing challenge to develop a fungal bioproduction process and ensure its healthy maintenance as an integral part of another complex entity: the technologically enhanced beehive of HIVEOPOLIS [66]. According to the HIVEOPOLIS design brief, the fungal materiality and scaffold morphologies should synergistically support a variety of physical, mechanical, and chemical properties. First and foremost, a self-sustaining mycelial hive should have at least three mycelium growth qualities to ensure the bioavailability of fungal metabolites and enzymes, as well as their safety in a durable and warm hive structure: (1) maintained healthy mycelium colonisation for its biological activities like enzymes and beneficial volatiles production, (2) thick mycelial skin on spots where water protection is needed, and (3) aerial growth of hyphae in order to mechanically connect separate modules and towards the beehive interior for bees exposure to mycelium in its purer state. In conclusion, the idea of utilising mycelium materials as a living material, coevolving hosts in the nests of other creatures, including humans, is an intriguing novel concept that guides our research. Yet, the meaningful re-integration of other living entities into the bee habitats, and human indoor spaces, is a challenging but promising task. It requires economic, cultural, and technological positioning of biohybrid architectures in human society, and eventually all ecosystems of Earth, this also demands the cultivation of multispecies living narratives and practices in our everyday lives. For us, the challenge was to bring the performance of smaller scale prototypes and qualities of small-scale samples to full-scale prototypes, ready in time for honeybee field experiments which are highly season dependent. When the technical challenges are overcome, more research is needed, however, to discover the long-term events that mycelium may initiate as part of habitat architectures. The increased microbial diversity in terms of quantity—the number of distinct species does not satisfy the goal to reach well-balanced microbial diversity. It is still speculative whether we will be able to effectively modulate the indoor microbiomes of our habitats or other organisms by incorporating dense mycelial networks into our architecture. We need to learn more about "who is there, where they live, and what they are doing" in our fungal designs. State of the art biotechnologies can aid in collecting this information in relation to our design probes. We can then use this knowledge in combination with the tools and methods developed in our fungal architecture community, to re-establish these causalities between geometrical, topological, and tactile aspects of our designs and microbial activities. In general, all these goals require focused groups of biodesign researchers collaborating with people in other scientific and engineering fields.

**Author Contributions:** Conceptualization, A.I. and T.S.; methodology, A.I.; software, A.I.; validation, A.I. and T.S.; formal analysis, A.I.; investigation, A.I. and T.S.; resources, T.S.; data curation, A.I.; writing—original draft preparation, A.I. and T.S.; writing—review and editing, A.I. and T.S.; visualization, A.I.; supervision, T.S.; project administration, T.S.; funding acquisition, T.S. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported by the Field of Excellence "Complexity of Life in Basic Research and Innovation" (COLIBRI) at the University of Graz and the EU H2020 FET-Proactive project "HIVEOPOLIS" (no. 824069).

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Acknowledgments:** The authors thank the Mycology Lab and Dalial Freitak at the Institute of Biology at the University of Graz, and the Institute of Environmental Biotechnology at the Technical University of Graz. Open Access Funding by the University of Graz.

**Conflicts of Interest:** The authors declare no conflict of interest.
