Bowl Trapping and Netting differences based on a very large sample

[Sam Droege]

 All

Attached is a detailed report by species on the absolute and relative capture rates of bee species from 22 years of surveys in Maryland and the District of Columbia for bowl trapping and netting techniques.  The report goes into lots of details with tables for all the species, discussions, and interpretations, but a summary of the results is presented below.

It is important to note that these are reported as differences rather than bias.  To understand bias one has to know something about the real population, which we don’t.  Therefore our analyses simply document the differences between survey techniques, how closely each survey techniques results reflect true populations is unknown, but because of the high correlations between the two techniques this is indirect evidence that commonly captured species in surveys are indeed common species in the environment and rarely caught species are also likely to be rare in the environment.

The relationship between captures rates (particularly very low ones) and conservation status needs to be approached cautiously.  We don’t have time here to get into that, but, the general idea is that rarity does not equate necessarily with a species’ population's instability or decline.  It is better to approach rarity in inventories as a plea for more information and considerations.

The raw data are available for those who wish to pursue further analyses.

A Fun Summary of Characteristics of Netting and Bowl Trapping for Bees in Maryland and the District of Columbia

Uber Summary: Data from netting and bowl traps are highly correlated. While species are captured at different rates, there is no evidence that either technique has no chance of detecting a species.  Even with large sample sizes, evidence supports that inventories greatly benefit from using both approaches versus relying only on one.

·         Years, 2002-2024

·         1835 Netting Collection Events

·         6362 Bowl Collecting Events

·         Collecting Event = unique time, date, place, technique, observer combination

·         Total specimens captured by bowls: 139,114.

·         Total specimens captured using nets: 31,337.

·         Number of bowls retrieved within an individual collection event varied from 1 to >230; mean = 15.2, median = 10.

·         Of the 1112 collection events that kept track of time netting, the mean was 41.7 minutes, the median 30 minutes, the range from 2 – 480 minutes, SE = 1.55

·         Ratio of female:male = Netting 70:30, Bowl 76:24

·         There were 3.47 times as many bowl collection events as netting collection events. A mean of 17.1 specimens/collection event were collected during netting and 21.9 specimens/collection event for bowl traps.

·         Correlations between bowl and net captures were highly correlated

·         Total species by net: 356

·         Total species by bowl: 332

·         Number of species caught in bowl traps but not caught while netting: 39 (12%)

·         Number of species caught while netting but not caught in bowl traps: 64 (18%)

·         Number of species caught on 5 or more collection events in bowl traps: 226

·         Number of species caught while netting on 5 or more collection events: 207

·         Percent of rarely caught species (only 1-4 collecting events) for bowl traps: 32%

·         Percent of rarely caught species (only 1-4 collecting events) for netting: 42%

·         Number of species caught in bowl traps but not caught while netting when species restricted to >4 collection events: 9

·         Number of species caught while netting but not caught in bowl traps when species restricted to >4 collection events: 15.

·         Greatest number of collection events missed for a single species: 35 (both techniques had same number)

·         Using differences in proportion of collection events captured,145 species were favored by bowl traps

·         Using differences in proportion of collection events captured, 250 species were favored by netting

·         Species with greatest differentials between netting and bowl trapping had large body sizes for netting and small ones for trapping

·         What follows are patterns of species rates of capture by family and common genera:

o   Andrenidae. 23 bowl/70 net

§  Andrena. 21 bowl/58net

o   Apidae. 37 bowls/60 net

§  Bombus. 0 bowl/11 net

§  Ceratina. 1 bowl/4 net

§  Melissodes. 4 bowl/8 net

§  Nomada. 16 bowl/12 net

o   Colletidae. 0 bowls/23 net

§  Colletes. 0 bowl/12 net

§  Hylaeus. 0 bowl/10 net

o   Halictidae. 63 bowl/45 net

§  Agapostemon. 4 bowl/0 net

§  Halictus. 1 bowl/4 net

§  Lasioglossum. 51 bowl/22 net

o   Megachilidae. 22 bowl/48 net

§  Coelioxys. 1 bowl/9 net

§  Hoplitis. 4 bowl/1 net

§  Megachile. 2 bowl/21 net

§  Osmia. 13 bowl/5 net

·         Melittidae. 0 bowl/4 net

·         Estimated number of species for region generated from netting results: 400.2

·         Estimated number of species for region generated from bowls: 407.2

·         Estimated number of species from combination of netting and bowls: 490.7

Sabbaths

1999, III

 

I dream of a quiet man

who explains nothing and defends

nothing, but only knows

where the rarest wildflowers

are blooming, and who goes,

and finds that he is smiling

not by his own will.

 

      -  Wendell Berry