CAWR

Sampling design

At 3 sites in Orange County, CA (SCR = Sand Canyon Resevoir, UCI = UCI Ecological Preserve, and BC = Bommer Canyon), 2-3 CAWR territories were surveyed. Within each territory, 9 of the most common habitat elements were sampled from the canopy (via visual and vacuum sampling) and ground (via pitfall traps). Sampling occurred at 4 time blocks between 2012 and 2013, and are aggregated across all timepoints. Due to the differences in sampling methods, canopy and ground arthropods are considered separately.

The sampling design is somewhat imbalanced? BC territory 1 had no pitfall trapping and vacuuming occurred only on 2 habitat elements between both blocks (excluded from analyses). In one sampling period, a habitat element is coded as GNCA and it is not clear what this is (Grass Native CA?), so the associated 26 arthropods are currently excluded from analyses. At UCI, there were 3 blocks in 2 of the territories, and the third territory (#5) there are data for only 7 habitat elements (no ARCA or RHIN, which are coincidentally the elements in the mystery BC data…). Also the time blocks for sampling area not totally clear…aggregating them all together and assuming that it is equal across blocks.

Sampling effort

Each habitat element was vacuumed at a constant effort.

Avg plant biomass vacuummed per sample.

Arthropod community

Arthropods were sampled from CAWR territories in order to assess the dietary resources available within their habitats.

Arthropod abundance & biomass

A total of arthropods were sampled representing 24 orders. Within Hemiptera, arthropods were divided into common subgroups, either Heteroptera, Auchenorrhyncha, or Sternorrhyncha. A small number of individuals (36) were unable to be identified to order and are excluded from further analyses.

Each individual was measured in length and used to estimate biomass by order designations (Hodar 1996). Biomass conversion data were not available for 6 arthropod orders (Archaeognatha, Raphidioptera, Unknown, Siphonoptera, Emhemeroptera, Pseudoscorpionida) and are thus excluded from biomass estimates. However, these were among the least common sampled from both canopy and ground surveys (fewer than 15 individuals in total for each order except Archaeognatha which had nearly 300), and are thus should have little impact on the interpretation of results.

Hymenoptera was the dominant arthropod order in terms of estimated biomass in shrub canopies and on the ground. In the canopy, Hemiptera had the second highest biomass (total of 5561 mg across 3 subgroups), while on the ground Isopods were the next biggest slice of biomass. Hymenoptera and Isopoda also ranked 1 & 2 in terms of abundance on the ground, respectively. Hemipterans were the most abundant order in the canopy (21,527 individuals total), with the 3 Hemipteran subgroups ranking in the top 3 (Sternorrhyncha most common). 3 orders were unique to the ground (Chilopoda, Mantodea, Pseudoscorpionida) but very uncommon, and Emhemeroptera was recorded only in the canopy, once.

Site effect?

Are these data representative of the CSS ecosystem? How variable are the arthropod communities among sites?
Comparing blocks or territories within each site?

Using the total arthropod biomass for each Block, separate ANOVAs testing for variation in arthropod biomass by canopy and ground. In canopy, no effect of location on total biomass (P = 0.369), however ground biomass differed with location (P = 0.014). SCR differs from BC and UCI, which is due to high abundance of Hymenoptera in territory 1 block 2.

Mean arthropod biomass by location (mean +/- se), with smaller points showing mean biomass for territories within location.

And a PERMANOVA test showed that there was a significant effect of location on arthropod community composition on both the ground (P = 0, R2 = 0.37) and in shrub canopies (P = 0, R2 = 0.37).

NMDS based on Bray-Curtis dissimilarities of arthropod relative biomass across orders

By territory:

Relative biomass by arthropod order

Does not include orders that compose less than 1% of total biomass on average. Following calculations are at the territory-level, using the mean across blocks in each territory.

CAWR diet

Prey DNA was recovered from ## out of ## total CAWR fecal samples. Sequences were compared to reference sequences in the BOLD database to obtain the closest match in identification. Established cutoffs for percent similarity were used to designate the taxonomic resolution for ids at order, family, genus, or species (Zeale et al 2011; Species id >= 99.3%, genus >= 94.9%, family >= 91%, order >= 85.9%). In cases where sequences matched at an equal similarity to two or more taxa, they were also searched in NCBI and assigned the highest taxonomic level that included all the matches. When the order ID was uncertain, IDs were checked against a phylogeny constructed from the aligned sequences. Sequences that did not match the database above 85.9% similarity were considered to be unknown.

Of a total 327 OTUs were distinguished from fecal samples (n = 29), 302 of which were matched to an arthropod order using BOLD. The most common IDs were: Cyclosa sp., Lepidoptera, Arachnis sp., Diptera, Araneidae, Gryllodes sp., Tipula sp….16 OTUs were unable to be matched to the order-level

Distribution of diet across habitat elements & strata

OPLI = Opuntia littoralis; RHIN = Rhus integrifolia, SANI = Sambucus nigra; ERFA = Eriogonum fasciculatum; ARCA = Artemisia californica; NAGR = Native grass; EXGR = Exotic grass; BRSP = Brassica sp.; BARE = Bare ground

Relative biomass

Canopy:

Ground:

The lepidoptera associated with OPLI is unexpected

Total biomass