Spatial analyses for nonoverlapping objects with size variations and their application to coral communities

Area of Science:

• Ecology
• Spatial Statistics
• Marine Biology

Background:

• Conventional spatial analysis uses point-based, center-to-center distances, which can be inadequate for non-overlapping, size-varying organisms like corals.
• Understanding inter-object spacing is crucial for analyzing spatial associations and interactions in such ecological communities.

Purpose of the Study:

• To introduce and demonstrate novel pairwise statistics based on minimum distances between objects.
• To compare the utility of minimum distance statistics with conventional point process and grid-based statistics for analyzing spatial patterns.
• To clarify the advantages and limitations of different spatial statistical methods in ecological contexts.

Main Methods:

• Developed new pairwise statistics using minimum distances between objects (approximated as disks).
• Calculated conventional point process statistics and grid-based statistics for comparison.
• Applied these methods to encrusting coral community data.

Main Results:

• Minimum distance statistics revealed aggregation in most coral genera at scales of 1-25 cm.
• Genet-level analysis (considering fragmented colonies as a single genet) showed weak or no aggregation, indicating random distribution and fragmentation as the cause of aggregation.
• Point process statistics indicated larger aggregation scales, influenced by colony size, while grid-based statistics were sensitive to colony size.
• Repulsive effects between aggressive and competitively weak coral genera were quantified.

Conclusions:

• Minimum distance statistics provide a more accurate assessment of spatial patterns for non-overlapping, size-varying organisms.
• Colony fragmentation is a significant factor influencing observed aggregation patterns in encrusting corals.
• The study extends the understanding of spatial pattern analysis for complex ecological systems.