**3. Results**

#### *3.1. Holon Level 1: The Macadamia Nut Industry on the "Big Island"*

Among the significant industrial sectors in the Hawaiian Archipelago is the macadamia nut industry on Hawaii Island (hereafter termed the "Big Island" to avoid confusion). One macadamia nut factory processed 14.1 Gg of macadamia nuts in 2005 (Figure 4). The processing plant's operations relied on very few island imports relative to its direct material inputs (island export-import ratio of 3.4 by mass). The facility derived 60% of its power from an on-site generator that burned 660 Mg of discarded macadamia nut shells and another 20% from 130 Mg of waste fuel oil. Water was extracted from two deep wells on the property, and the nuts were grown primarily in orchards adjacent to the factory, minimizing the distance for material transport. Other industries could, of course, be analyzed in the same way, depending on the availability of data and the purpose of the analysis.

The material flows for individual facilities could be e ffectively characterized by conventional MFAs for enterprise-level holons. However, these analyses alone do not inform how the quantified flows relate to flows for a city, an island or county, and other higher-level holons. Whole island data show that 25.6 Gg of macadamia nuts were harvested on the Big Island in 2005, so the factory processed over half of all macadamia nuts collected on the island in that year. Such a comparison for a single material may be simple and interesting, but much more data are required to make robust assessments about the overall relationships between material flows for an individual enterprise (or industry) and aggregate flows for higher-level holons. For example, the relative contributions of the factory's material flows to the total flows for the State of Hawaii remain unaddressed. This knowledge gap may be filled, however, by utilizing results from other holarchic levels, as shown in the following sections.

**Figure 4.** Material flow analysis for a macadamia nut processing plant on the Big Island in 2005. The units are Gg a<sup>−</sup>1. Based on Houseknecht et al. [40].

#### *3.2. Holon Level 2: Municipalities on the Big Island: Hilo and Kona*

The example holon for level 2 is a municipality. As part of our study, the students conducted MFAs to examine the social metabolism of Hilo and Kona, the two principal urban centers on the Big Island. In this study, Hilo's spatial extent is defined as the Hilo Census Designated Place (CDP), and Kona's extent is defined as the collection of Kailua, Kahaluu-Keauhou, Kalaoa, Holualoa, and Honalo CDPs, based on extensive consultation with geographers and the County of Hawaii planning departments (Although Kona and Hilo are known as the largest municipalities on the Big Island, neither has municipal boundaries per se. It is typical in Hawaii that the more populous islands are designated as counties and are unincorporated. With no official boundaries, a geographical carving out of CDPs had to be created for both Hilo and Kona to conduct the comparison properly, specifically in connection with National Science Foundation gran<sup>t</sup> #0948781 ULTRA-Ex Award. "Human-Nature Interactions in an Urbanized Island Setting: Hilo and Kailua-Kona, Hawai'i as Model Socio-Ecological Systems."). The macadamia nut factory lies outside of our system boundaries for Hilo and Kona. Because more than one example of a municipality holon was addressed, it was possible to compare the two results, as shown in Figure 5.

Significant differences are seen across Hilo and Kona in imports, exports, and domestic extraction. The resident population of Hilo was approximately 20% greater than that of Kona in 2009 [52], but the direct material input to Hilo was 43% greater than that of Kona. One reason for this disparity was Hilo's fossil fuel imports, which were 87% greater than that of Kona. On the other hand, there were several materials that had larger flows for Kona than those for Hilo. Notably, Kona imported, locally extracted, and consumed a larger amount of biomass, for grazing and for food and agricultural products. The larger imports and domestic consumption of biomass may partially be explained by Kona's greater focus on the tourism industry. Interestingly, Kona extracted more biomass, even though Hilo had 13 times more water than the dry, tourism-oriented Kona [26]. Construction minerals accounted for 44% of direct inputs to each municipality, and almost all were used within the cities. Together, the two municipalities consumed 350 Gg of fossil fuels and utilized 1,350 Gg of construction minerals. While separate MFAs for holons at the same holarchic level allow for some comparisons, no

claims with respect to higher-level holons could be made without MFA results for higher-level holons, as was the case with the macadamia nut factory.

**Figure 5.** Material flows (Gg a<sup>−</sup>1) for the Big Island communities of Hilo (left) and Kona (right) in 2009, excluding water. Data from Chertow and Seto [27].

#### *3.3. Holon Level 3: The Big Island*

The example holon for level 3 is an island. An MFA that was completed for the Big Island is shown in Figure 6. The economy of the island was evenly dependent on imports and local extraction in 2009, with the direct material input equaling 5,080 Gg. Fossil fuel imports amounted to 580 Gg, of which 99% was consumed on the island. Construction minerals were the largest material flow, equaling 40% of direct inputs. Practically all non-metallic mineral inputs were consumed on the island. Approximately 72% of the direct material input, a larger ratio compared to Hilo and Kona, was consumed on the Big Island as a whole. This difference is mainly a result of the large local extraction of biomass for grazing on the island.

**Figure 6.** Material flow analysis for the Big Island (Gg a<sup>−</sup>1) for year 2009, excluding water. Data from Chertow and Seto [27].

A comparison of material flows for Hilo and Kona with that of the entire island reveals that while almost half of the island's population resided in Hilo and Kona [52], the two municipalities together accounted for the use of 61% of fossil fuels and 66% of construction minerals consumed on the entire island. Furthermore, the quantity of biomass extracted in the two municipalities combined was practically insignificant for the island overall (0.9% of local extraction and 0.4% of the direct material input to the Big Island).

Thus, comparing the metabolism of Hilo and Kona with that of the entire island illustrates that much of the island exhibited characteristics of a more rural economy. With MFA data for multiple holarchic levels, researchers may be able to make practical approximations of material flows for holons with data gaps. For example, subtracting Hilo and Kona's quantities for local extraction of food and agricultural products from the total extraction of food and agricultural products on the Big Island yields the result that the distribution of extracted food and agricultural products was 3.9 Gg in Hilo, 4.4 Gg in Kona, and 58.7 Gg across the entire rest of the island. This distribution would have been missed from solely an island-level MFA or a comparison of Hilo and Kona. The estimation technique is not directly applicable to our dataset because the flows for Hilo and Kona were derived together with those for the rest of the island by applying allocation factors to island-level flows (based on ratios of cropland area, building permits, product sales, etc.). The approach would have been practical, however, if the domestic extraction of biomass for Hilo and Kona were calculated using cropland area and granular crop yield estimates, which vary widely by section of the island. It could also be useful in cases where material flow data are documented for a higher-level holon (e.g., island or county) and only for some nested, lower-level holons (e.g., city).

With approximate data for the extraction of food and agricultural products on the rest of the Big Island, comparisons could be made with the macadamia nut factory. The quantity of macadamia nuts processed by the factory accounts for roughly a quarter of food and agricultural products extracted on the island, excluding Hilo and Kona, and a fifth of all food and agricultural products extracted on the island overall. This comparison also shows, however, that even an enterprise such as the macadamia nut factory contributed less than 1% of the total domestic extraction and the direct material input for the Big Island.

#### *3.4. Holon Level 4: The Hawaiian Archipelago*

The example holon for level 4 consists of the six most populous islands of the Hawaiian Archipelago, which account for the vast majority of material flows for the State of Hawaii. Figure 7 illustrates the overall material flows for the state, revealing the archipelago holon's metabolic profile to be very di fferent from that of the Big Island holon. The State of Hawaii had a direct material input of 24.8 Tg in 2007, split evenly between domestic extraction and imports, similar to the Big Island. Unlike the Big Island, however, 94% of the direct material input to the state was consumed within the archipelago system.

Construction minerals were the most prominent materials for domestic extraction and apparent consumption for the archipelago holon. The flows of fossil fuels for the entire state were much larger than for the Big Island holon in absolute terms and in relation to other materials for imports, exports, and apparent consumption. Further analyses of material flows for island-level holons revealed that over two-thirds of all fossil fuel imports are consumed on Oahu. The distribution of fossil fuel consumption among the islands was in line with the population distribution, although there also appeared to be a decrease in consumption per capita with islands with higher population densities. This trend was similar for other material flows as well.

**Figure 7.** Material flow analysis for the State of Hawaii (Tg a<sup>−</sup>1), 2007. Based on Famely et al. [37] and Kanaoka [55].

The quantities for the state's exports may appear to be small when considering that the Big Island's exports alone were comparable to those of the entire state. The comparison of MFAs for the archipelago-level and island-level holons revealed that 46% of all of Hawaii's ocean shipments (sum of imports and exports) were intrastate (inter-island) shipments by mass, mostly consisting of metals. This result is in agreemen<sup>t</sup> with figures from a study by the Hawaii DBEDT [56]. For the archipelago holon, fossil fuel was a significant export, translating to 6% of imported fossil fuels. Much of the exported fossil fuels consisted of naphtha, a solvent and feedstock for plastics, for which no significant market exists in Hawaii [57].

#### *3.5. Holon Level 5: Pacific Ocean Shipping*

The available data for material flows to and from the Hawaiian Archipelago do not permit us to fully reconstruct the material interactions between the archipelago and the planet. One particular material flow, however, can be accessed, plotted, and studied: the flows of petrochemical fuels to the Hawaiian Archipelago from the world's major oil-producing regions (Figure 8).

As in the example of automobile paint in the introduction, conceptualizing the global supply chain for petrochemical fuels as a holon could help understand the dependence of lower-level holons on the holarchic system. As we have noted, petrochemical fuels constitute the most significant direct material input to the State of Hawaii by mass (Figure 7). Because Hawaii has no proven crude oil reserves, the use of petrochemicals on the islands has been completely dependent on imports. The total quantity of imported liquid fuels remained around 50 million barrels per year between 1992 and 2009 (Figure 8). While almost half of the liquid fuel imported to Hawaii in 1992 was from the continental U.S. (i.e., Alaska), that ratio was reduced to 2% by 2009. The energy system at the heart of Hawaiian society has thus become significantly more reliant on material flows from foreign nations than was the case a half-century ago.

**Figure 8.** The flows of petrochemical fuels to the Hawaiian Archipelago from the world's major oil-producing regions. Adapted from Cooke and Parsons [58].

If the material flows for the State of Hawaii were studied as a single-level system without examining their connections to the global economy, these supply chain relationships and associated political and economic dynamics would be overlooked. Studying these relationships is vital to understanding how material flows may change in the future, such as could occur with supply chain disruptions due to economic sanctions. These insights are also important for understanding how decisions made at lower-level holons could shape material flows at higher-level holons. This is especially important now that the Hawaiian state legislature has adopted the goal of generating 100% of its electricity from renewable energy resources by 2045 [59,60].
