Commensal relationships between plants and their dispersers are crucial for community structure and function. While non-human dispersers are well-studied, the role of humans, particularly in relation to non-domesticated plants, remains less understood. Australia, with its historical reliance on non-domesticated plants, provides a unique setting to explore this. This study focuses on the Martu Aboriginal people, a group with a long history of mobile hunting and gathering, examining how their practices affect plant distributions. The impact of human-mediated seed dispersal on wild plant populations is widely debated, with most research focusing on the passive spread of invasive species. However, emerging evidence suggests that sedentary, agriculturally-based indigenous communities may have significantly influenced the distribution of useful, non-domesticated plants. This study aims to determine if highly mobile hunter-gatherers can affect plant distributions through mechanisms beyond agriculture, such as seed dispersal and landscape disturbance. Understanding this interaction is important for several reasons: it can shed light on the transition from foraging to farming, inform biodiversity conservation strategies, and elucidate the formation of Indigenous cultural landscapes. Australia, with its long history of Aboriginal land use characterized by foraging, high residential mobility, and fire-stick farming, is particularly relevant for such an investigation. Aboriginal populations are believed to have influenced the distribution of numerous plants through fire management, bioturbation, and seed scattering. This study builds on existing anecdotal evidence about bush tomato dispersal, using a controlled observation research design to test causal mechanisms. The study will focus on four plant species in a Western Australian desert landscape with a long history of Martu occupation, utilizing ground-based ecological surveys and proxies for seed dispersal and anthropogenic disturbance.

Existing literature highlights the importance of understanding human-mediated seed dispersal, especially in contexts outside of agriculture. Studies on vehicle-mediated dispersal of invasive species and the unintentional spread of seeds on clothing are prominent in this field. Conversely, research on the impact of indigenous communities on plant distribution is growing. Studies have shown that sedentary, agriculturally-based indigenous groups have played a substantial role in shaping the distribution of useful non-domesticated plants, evidenced by higher plant densities near old settlements and archaeological sites. However, research on the influence of mobile hunter-gatherer populations on plant distribution through seed dispersal or landscape disturbance is limited. The importance of this understudied area is clear: it can shed light on the co-evolution of humans and plants, assist in designing better biodiversity conservation policies, and help to understand the long-term interactions between humans and the environment, particularly regarding the transition between foraging and farming economies. Several studies have looked at how Indigenous peoples have influenced plant distribution in Australia, often indirectly through genomic studies of plants suspected to have been dispersed by humans. Anecdotal evidence suggests that the Martu people may have played a significant role in spreading some plants, but there has been a lack of rigorous ecological studies to verify this.

The study employed a hypothetico-deductive methodology with a controlled comparison research design. The research involved a combination of ecological transect surveys and ethnographic observations conducted over a 25-year period in collaboration with Martu communities. Ten 10 km x 10 m belt transects were surveyed to assess the presence and abundance of four plant species: two bush tomatoes (*Solanum diversiflorum* and *S. centrale*), a seed grass (*Eragrostis* spp.), and a forb (*Scaevola parvifolia*). *S. parvifolia* served as a control species due to its low preference among Martu. Predictor variables included remotely sensed environmental and soil data, fire history mapping, and ethnographic observations. Ethnographic data was gathered through 800 days of participant observation during foraging trips, quantitatively recording harvesting activities, yields, and seed dispersal events for each species. The researchers observed high potential for dispersal for *S. diversiflorum*, with significant in-patch and en-route seed dispersal during harvesting and transport. *S. centrale* showed lower dispersal potential due to the edible nature of its seeds. *Eragrostis* had moderate dispersal potential due to seed loss during processing, while *Scaevola* showed little potential for direct dispersal. To quantify the potential for anthropogenic dispersal of *S. diversiflorum*, 12 contemporary dinner camp locations were surveyed. The presence and abundance of *S. diversiflorum* were significantly higher at sites where fruit had been processed compared to control sites. Statistical modeling involved hierarchical generalized linear models (GLMs) using transect ID as a random effect. Predictor variables encompassed dispersal proxies (archaeological/residential site distance, site type, land use intensity, water permanence) and engineering proxies (time since fire, fire season, fire frequency). Model selection used an information-theoretic approach, comparing the global model and a series of best-fit models for presence and abundance. Additional covariates included soil conditions, water availability, and vegetation cover to account for environmental factors influencing plant distribution. Likelihood ratio tests were used to evaluate the significance of the top models compared to null models with only random effects.

Ethnographic observations revealed a high potential for anthropogenic dispersal of *S. diversiflorum*, with substantial seed dispersal occurring during harvesting, transport, and processing. Surveys of food consumption sites showed significantly higher plant density at sites where *S. diversiflorum* fruit was processed. Statistical modeling showed that *S. diversiflorum* presence was strongly predicted by proximity to sites and high contemporary land use near permanent water sources. The engineering model (fire proxies) explained the most variation in *S. diversiflorum* presence compared to other species. For *S. centrale*, dispersal predictors were weak, but fire frequency positively impacted presence. *Eragrostis* presence showed complex interactions between dispersal and fire, being more likely near sites with ephemeral water sources under high contemporary land use. The control species, *Scaevola*, showed no effects of dispersal or landscape engineering, potentially due to the avoidance of sites by major dispersers (birds) due to human presence. The best models for plant presence corroborated the results from the global model, with land use remaining a strong predictor for *S. diversiflorum*, and complex interactions observed for *Eragrostis*. Abundance models indicated *S. diversiflorum* abundance was highest near sites with permanent water, while *Eragrostis* abundance was higher near sites with ephemeral water and intensive land use.

The findings provide strong evidence that highly mobile Martu people have acted and continue to act as significant seed dispersal agents, substantially influencing the distribution of culturally significant plants. The significant impact of Martu practices on *S. diversiflorum* distribution highlights the importance of considering human activities in ecological studies, even in the absence of intentional management or agriculture. The results demonstrate the need to move beyond simplistic foraging/farming dichotomies. Seed dispersal, whether intentional or unintentional, has significant ecological consequences. The varied responses of the four plant species to human activities (seed dispersal and fire) emphasize the complexity of human-environment interactions. The study's results underscore the importance of considering the historical and ongoing role of Indigenous people in shaping ecosystems and biodiversity. Integrating Indigenous Ecological Knowledge (IEK) into conservation and restoration strategies is crucial, acknowledging the diverse ways people interact with and shape their environments.

This research provides compelling evidence for the substantial impact of Martu people's activities, primarily seed dispersal and fire management, on the distribution of key plant species in arid Australia. The study demonstrates the significant role of humans, even highly mobile hunter-gatherers, in shaping plant communities, highlighting the importance of integrating Indigenous Ecological Knowledge into conservation and restoration efforts. Future research could focus on more detailed genetic studies of the focal plant species to further confirm the role of Martu in shaping genetic diversity and expansion, as well as on expanding the study to a broader range of species and landscapes to explore the generality of the findings.

The study's sample size for the dinner camp survey was relatively small, which might affect the statistical power. The reliance on remotely sensed data for fire history reconstruction and other environmental variables introduces some uncertainty and potential for error. The study primarily focused on four key species, limiting the generalizability of the findings to other plants. The study primarily focussed on the dispersal of specific species, but many other species are impacted by the cultural burning practices of the Martu. Future research could focus on a wider selection of species, as well as on exploring other elements of the Martu’s relationship with the land.